Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Enzymes as indices of growth rate and nitrate metabolism in marine phytoplankton Berges, John Alexander 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1994-89329x.pdf [ 7.74MB ]
JSON: 831-1.0053193.json
JSON-LD: 831-1.0053193-ld.json
RDF/XML (Pretty): 831-1.0053193-rdf.xml
RDF/JSON: 831-1.0053193-rdf.json
Turtle: 831-1.0053193-turtle.txt
N-Triples: 831-1.0053193-rdf-ntriples.txt
Original Record: 831-1.0053193-source.json
Full Text

Full Text

ENZYMES AS INDICES OF GROWTH RATE AND MTRATE METABOLISMIN MARINE PHYTOPLANKTONbyJOHN ALEXANDER BERGESB.Sc., University of Guelph, 1987M.Sc., University of Guelph, 1989A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Department of Oceanography)We accept this thesis as conformingTHE UNIVERSITY OF BRITISH COLUMBIAOctober 1993© John Alexander Berges, 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)___________________________Department of OcvJOG cZ1J4WThe University of British ColumbiaVancouver, CanadaDate + /93DE-6 (2/88)UABSTRACTDetermining the in situ rates of growth and nitrogen incorporation of marinephytoplankton is critical to understanding energy transfer and nutrient and carbon cycling inthe world’s oceans. To overcome the limitations in current methods of estimating biologicalrates (i.e. incubations under unrealistic conditions, or inadequate estimates of spatial andtemporal variability) the use of enzyme activity measurements was examined. Becauseenzymes are functional proteins that adapt to suit prevailing conditions, enzyme levels mayprovide an integrated index of in situ rates of phytoplankton metabolism.Nucleoside diphosphate kinase (NDPK) an enzyme which directs cellular energytowards biosynthesis was examined as an index of specific growth rate (JL) in the diatomThalassiosira pseudonana grown under light limitation, NDPK activity was significantly, butwealdy correlated with . Activity per cell rose at high , but also increased at very low .Although of limited value as a predictive index by itself, NDPK may be useful in conjunctionwith measurements of ATP concentration, or adenylate turnover rates.Nitrate reductase (NR), an enzyme specific for nitrate assimilation may be used incalculating rates of nitrate incorporation (JLN) and thus new production, but previousmeasurements of NR have not matched A new assay using bovine serum albumin toprotect the enzyme from proteases was developed that gave close agreement with N in lightlimited cultures of T. pseudonana and Skeletonema costatum. The relationship also held for T.pseudonana during transitions in irradiance, under nitrate limitation (although NR exceeded!LN at low ), during growth on light-dark cycles, different light spectra, in the presence ofammonium, and during nitrate starvation. In each case, NR accurately predicted NR wasclosely related to nitrate incorporation rates in three additional diatom species, but for othertaxa, particularly the Dinophyceae, NR underestimated IN• Preliminary field experimentswere conducted in Monterey Bay, California during a diatom bloom. 1N predicted from NRmeasurements always equalled or exceeded rates estimated by other methods, including ‘5Nincorporation.UIAppendices to the thesis compare and validate different protein assays in marinephytoplankton, provide details of a computer program to automate and collect enzyme kineticdata from a spectrophotometer, and compare methods of fitting rectangular hyperbolae to avariety of oceanographic data.ivTABLE OF CONTENTSAbstract iiTable of Contents ivList of Tables viiList of Figures ixAcknowledgements xviGeneral Introduction 1The importance of biological rates 1Limitations of primary production 4Phytoplankton nitrogen metabolism and the concept of new production 6Traditional methods of estimating biological rates and their problems 10Alternative approaches to estimating growth rates 14Alternative approaches to estimating new production 17Enzyme activity: a theoretical basis for predicting biological rates 18Application of enzyme activity measurements to planktonic organisms 19Organization and goals of this thesis 22Chapter 1: Relationship between nucleoside diphosphate kinase activity and light-limitedgrowth rate in the marine diatom Thalassiosira pseudonana 25Introduction 25Characteristics of NDPK 26NDPK and growth rate 28Materials and Methods 29Culture conditions 29Cell composition 29Cell homogenization and enzyme assay 30Enzyme characterization 31Steady state experiments 31Transition experiments 32Results 32Enzyme characterization 32Steady state experiments 32Transition Experiments 37Discussion 41Enzyme characterization 41Cell composition 43NDPK and growth rate 45Chapter 2: Optimization and validation of an assay for nitrate reductase activity in marinephytoplankton 50Introduction 50The place of NR in algal nitrate metabolism 50Structure and characteristics of NR 54Regulation of NR 56NR assay methods 59Materials and Methods 67General culture conditions 67Enzyme optimization experiments 67Enzyme characterization in different species 71VComparisons of NR activity with growth rate 72Results 73Enzyme optimization experiments 73Enzyme characterization in different species 82Comparisons of NR activity with growth rate 88Discussion 93Enzyme assay optimization 93Enzyme characterization in different phytoplankton species 96Comparisons of NR activity with growth rate 99Remaining problems with NR assays 103Chapter 3: Relationships between nitrate reductase activity and growth rate under steady-statelight or nutrient limitation in Thalossiosira pseudonana 105Introduction 105MR and the control of nitrate metabolism 105Light and nitrate limitation of growth rate 109Cell composition and scaling of enzyme activity 112Materials and Methods 113General culture conditions 113Steady-state light-limited experiments 113Light transition experiments 114Steady-state nitrate-limited experiments 114Scaling of MR activity 115Results 115Steady-state light-limited experiments 115Light transition experiments 118Steady-state nitrate-limited experiments 123Scaling of MR activity 128Discussion 128Variation in cell composition with growth rate 128Variation in MR activity with growth rate 138Scaling of MR activity 142Chapter 4: Effects of light:dark cycles, different light spectra, nitrate exhaustion, andammonium on the relationship between nitrate reductase activity and nitrate incorporationrates in Thalassiosira pseudonana 147Introduction 147Effects of diel periodicity in irradiance 147Effects of different light spectra 148Effects of nitrate exhaustion 148Effects of ammonium 149Materials and Methods 150General culture conditions 150Light:dark cycle experiments 150Light spectra experiments 151Nitrate exhaustion experiment 152Effects of ammonium and ammonium pulsing 152Results 152Light:dark cycle experiments 152Light spectra experiments 159Nitrate exhaustion experiment 163Effects of ammonium and ammonium pulsing 168Discussion 168Effects of diel periodicity in irradiance 168Effects of different light spectra 175Effects of nitrate exhaustion 178viEffects of ammonium and ammonium pulsing.181Implications of regulatory mechanisms 184Chapter 5: Activity and characteristics of nitrate reductase in natural phytoplanktonpopulations from Monterey Bay, California 186Introduction 186Characteristics of coastal upwelling zones 186Materials and Methods 191Modifications to NR assays 191Assay validation and enzyme characterization 191Containment experiments 194Results 195Assay validation and enzyme characterization 195Containment experiments 199Discussion 205Adequacy of the MR assay 205Diel periodicity 206Effects of ammonium 207NR activity and nitrate incorporation rates 208General Discussion and Conclusions 213Experiments following from the thesis 215Applicability to different species 215Non-nitrogen starvation effects 216Ammonium inhibition time series 216Temperature effects 217Metabolic control analyses 217Optimization of assays in natural populations 218Future directions 218Assay sensitivity and single cell analyses 218Automated assays 219Coupling hydrodynamic and biological scales 219Field application: important regions 220Literature Cited 222Appendix A: A comparison of Lowry, Bradford and Smith protein assays using differentprotein standards and protein isolated from the marine diatom Thalassiosira pseudonana264Appendix B: Adapting an LKB Ultrospec II UV spectrophotometer for enzyme kineticanalyses 282Appendix C: Fitting ecological and physiological data to rectangular hyperbolae: acomparison of methods using Monte Carlo simulations 293vuLIST OF TABLESTable 1.1. Characteristics of nucleoside diphosphate kinase (NDPK) from various sources(ISOZYME = isoelectric point of the isozyme where available, otherwise the authorsdescription; MW = molecular weight, * indicates the weight of a monomer; AE =apparent activation enthalpy below/above the transition temperature; -- = not providedby the authors) 27Table 2.1. Characteristics of nitrate reductase from various sources 55Table 2.2. Selected assay mixtures for in vitro or in situ nitrate reductase assays. (DTT =dithiothreitol, CYS = cysteine, FAD = flavin adenine dinucleotide, PVP = polyvinylpyrrolidone, * = NADPH used in place of NADH,? = information not provided byauthors) 60Table 2.3. Effects of addition of 0.1 mM FAD on nitrate reductase activity (determined byNADH oxidation rate or nitrite production rate) in homogenates of Thalassiosirapseudonana either analyzed directly (normal) or desalted using a Sephadex G-25column. Values represent means and standard errors of 3 replicate assays 84Table 2.4. NR activity in various species of phytoplankton using NADPH as a reductant.Activities are expressed as a mean percentage (± standard errors, n = 2 cultures) ofactivity found using NADH 93Table 2.5. Kinetic constants for nitrate reductase from various eukaryotes 99Table 2.6. Representative nitrate reductase activities from eukaryotic microaigae 102Table 3.1. First-order regression parameters for composition versus growth rate relationshipsin light-limited cultures of various marine phytoplankton. P-values represent theprobability that the slope is equal to zero 117Table 3.2. First order linear regression parameters for composition versus growth raterelationships in nitrate-limited chemostat cultures of Thalassiosira pseudonana. Pvalues represent the probability that the slope is equal to zero 126Table 3.3. Comparison of first order linear regression parameters for nitrate reductase activityscaled to different parameters versus growth rate in Thalossiosira pseudonana in light-limited batch cultures (L), nitrate-limited chemostats (N) or both types of culturestogether. P-values represent the probability that the slope is equal to zero 130Table 3.4. Changes in composition with increasing growth rate (irradiance or nutrient supply)for various species under light or nutrient limitation. D = hours of day light (i.e. 24means continuous light), N SOURCE = the nitrogen source used (N03 = nitrate,NH4 = ammonium), VOL = cell volume, C = carbon quota, N = nitrogen quota,C:N = carbon:nitrogen ratio, CHL = chlorophyll a quota, C:CHL =carbon:chlorophyll a ratio, CHO = carbohydrate quota, PRO = protein quota.Responses are defined as increases (+), decreases (-), no change (nc), or complexbehaviour (c) 131Table 3.5. Relationship of nitrate reductase activity with increasing growth rate, andpercentage of nitrate incorporation accounted for by NR (%NR/N) in various speciesunder different limitations. Light is continuous and chemostats are nitrate-limitedunless otherwise noted 140vu’Table A. 1. Comparison of absorbance versus protein content slopes for bovine serum albumin(BSA), bovine gamma-globulin (BGG), aipha-casein, and protein purified fromThalassiosira pseudonana cultures grown under either high or low light. Valuesrepresent mean and standard error of 5 different determinations from separatelyprepared standards. Summaries of statistical comparisons (one-way ANOVA, followedby Tukey’s multiple range test) are provided below the table. Lines join proteins whichare not significantly different from each other at the 95% C.I 273Table C. 1. Results of model-fitting procedures (Lineweaver-Burk, LB; Eadie Hofstee, EH;Hanes-Woolf, HW; Eisenthal and Comish-Bowden, ECB; Cleland-Wilkinson, W;Tseng-Hsu, TH) for Case 1 (CS 1) and Case 3 (CS3) data. Notation describes error asconstant (C) or variable (V) and error levels as a percentage (10, 20 or 50). Error inboth S and V is denoted “XY”. Bold numbers represent medians of estimatedparameters of the data. The true values are = 10, and Km = 2. Numbers inlighter face represent the percentages of estimates which fell outside the ranges of 1.2-2.8 for Km and 6-14 for Vmajc 299Table C.2. Summary of results of fitting procedures for Vm and Km. Conditions markedwith ““ indicate cases where the median of estimates was within 20% of the truevalue (10 or 2) and less than 20% of the estimates fell outside the ranges of 6-14 or1.2-2.8. Conditions marked with ““ indicate that the median was within 20% of thetrue value, and less than 50% of the estimates fell outside the specified ranges 302Table C.3. Comparison of estimates of the parameters of rectangular hyperbolae (Vmu andK) using different fitting methods for three real data sets. Fitting procedures are:Lineweaver-Burke (LB), Eadie-Hofstee (EH), Hanes-Woolf (HW), Eisenthal andCornish-Bowden (ECB), Cleland-Wilkinson (W), and Tseng-Hsu (TH). Data sets arepictured in Fig. C.4, and described in the text 306kLIST OF FIGURESFigure 1. Diagram of the processes involved in new and regenerated production in the upperwater column. See the text for explanations. (PON = particulate organic nitrogen,DON dissolved organic nitrogen) 9Figure 1.1. Nucleoside diphosphate kinase (NDPK) activity versus substrate concentration fora) thymidine 5’-diphosphate (TDP) and adenosine 5’-thphosphate (ATP) inhomogenates of Thalassiosira pseudonana. Curves are fit to rectangular hyperbolae.Km values are 0.24 mM for TDP and 0.86 mM for ATP 33Figure 1.2. Arrhenius plot of NDPK from Thalassiosira pseudonana. The solid linerepresents a squares regression fit to the data. Apparent activation enthalpy is0.84lkJmol’ 34Figure 1.3. Growth rate versus irradiance curve for Thalassiosira pseudonana. curve is fit toa rectangular hyperbola. ILmax = 1.64 d1 and K1 = 23 mol quanta m’ s1. Eachpoint represents a single culture. Error bars represent the standard error of the mean of3 to 6 growth rate measurements 35Figure 1.4. Cell composition versus light-limited specific growth rate in Thalassiosirapseudonana. A) cell carbon quota, B) cell nitrogen quota, C) cell volume, and D) cellprotein quota. Each data point represents the mean of duplicate determinations from asingle culture 36Figure 1.5. NDPK activity versus A) light-limited specific growth rate, and B) growth rate interms of carbon in Thalassiosira pseudonana. Each data point represents a singleculture. Error bars show the standard error of the mean of two enzyme assays or aminimum of three growth rate determinations 38Figure 1.6. NDPK activity versus specific growth rate in Thalassiosira pseudonana. Activityis expressed per unit carbon (A), nitrogen (B), cell volume (C), or protein (D). Errorbars show the standard error of the mean of two enzyme assays or a minimum of threegrowth rate determinations 39Figure 1.7. Cell composition versus time in terms of A) cell carbon quota, B) cell nitrogenquota, C) cell volume, and D) cell protein quota in Th,i1acsiosira pseudonana. (0)Cultures grown nder high light (135 mol quanta m’s’) and moved to low light (15mol quanta m-’s-l at t = 32h (marked by the arrow. (•) Cultures grown under lowlight and switched to high light at t = 32h. Error bars represent standard errors of themean of three replicate cultures. Statistically significant differences (P < 0.05) areindicated by asterisks (*) 40Figure 1.8. NDPK activity scaled to A) cell number, B) cell carbon quota, C) cell nitrogenquota, D) cell volume, and E) cell protein quota, versus time for transition experimentswith Thalassiosira pseudonana. Symbols are the same as in Figure 1.7 42Figure 2.1. Validation of spectrophotometric assay for nitrate reductase (NR). A) Timecourse of reaction before and after addition of 10 mM KNO3 (indicated by arrow). B)Comparison of activity calculated from NADH oxidation rate (corrected for non-nitrate-specific activity), and that cal,culated based on production of nitrite. Regressionequation is: Y = -0.81 + 0.98 X (r’ = 0.99) 74xFigure 2.2. Comparison of NR homogenization and extraction procedures in homogenates ofThalassiosira pseudonana. A) Relative NR activities in samples collected by filtrationonto glass fibre filters, or centrifugation at 7 500 g. In each case, replicate samples (n= 3) were homogenized by grinding or by probe sonication. B) Relative NR activityin the supernatant and pellet fractions of homogenates of cells collected by filtrationand homogenized by grinding. Homogenizations were performed with or without0.1 % Triton X-100. Centrifugations were done at 750 g for 5 mm 75Figure 2.3. Relative NR activity in homogenates of Thalossiosira pseudonana prepared in 200mM phosphate buffer, 50 mM MOPS buffer, 50 mM TRIS buffer, or 50 mMimidazole buffer. In all cases, pH was 7.9. n = 3 for each buffer treatment 77Figure 2.4. Effects of different additions on nitrate reductase (NR) activity in homogenates ofThalassiosira pseuthnana. A) Activity in homogenates with only 200 mM phosphatebuffer and 0.1 % (v/v) Triton X-100 (1) versus: 5 mM EDTA (2), 0.3 g i1 DTT (3),3.0 g l1 PVP (4), or DTI’, EDTA and PVP (5). B) NR activity in homogenatesprepared using only buffer 5, or with additions of 0.1 mM FAD, or 0.2 mMferricyanide (n = 3 in all cases) 78Figure 2.5. Stability of MR activity over time in homogenates of Thalassiosira pseudonanahomogenized without additions (•), with additions of 3% BSA (I) or with additions ofprotease inhibitors as described in the text (•). Points represent means plus standarderrors of 3 separate homogenates 79Figure 2.6. Relative NR activity in homogenates of Thalassiosira pseudonana provided withdifferent reductants: 0.2 mM NADH, 0.2 mM NADPH, or 0.1 mM NADH plus 0.1mM NADPH. Error bars represent standard errors of the mean of 3 separatehomogenates 80Figure 2.7. MR activity in homogenates of Thalassiosira pseudonana before (t = Oh) andafter (t = 48, 96h) freezing and storage in liquid nitrogen. Points represent the meanand standard error of 3 separate samples 81Figure 2.8. MR activity versus substrate concentration for A) KNO3 and B) NADH inhomogenates of Thalassiosira pseudonana. Curves are fit to rectangular hyperbolae.Km values are 0.0165 mM for NADH and 0.0471 mM for KNO3 83Figure 2.9. Comparison of the effects of addition of FAD and FeCN on MR activity inhomogenates of Skeletonema costatum. Activity is expressed relative to the activity inthe homogenate without additions. Trials 1 and 2 represent two separate experimentson two different cultures. Error bars represent standard errors of the mean of 3homogenates 85Figure 2.10. Nitrate reductase (NR) activity versus substrate concentration for: A) KNO3 andB) NADH in homogenates of Skeletonema costatum. Curves are fit to rectangularhyperbolae. Km values are 0.146 mM for KNO3 and 0.0476 mM for NADH 86Figure 2.11 . Comparison of the effects of different activators on MR activity in homogenatesfrom Amphidinium carterae. Additions include 0.1 mM FAD, or 0.2 mM ferricyanide(FeCN). Error bars represent the standard error of the mean of 3 separatehomogenates 87Figure 2.12. Nitrate reductase (NR) activity versus substrate concentration for: A) KNO3 andB) NADH, for homogenates of Amphidinium carterae. Curves are fit to rectangularhyperbolae. Km values are 0.075 mM for KNO3 and 0.150 mM for NADH 88xiFigure 2.13. Growth rate versus irradiance curves for: A) Thalassiosira pseudonana (•), B)Skeletonema costatum (B), and C) Amphidinium carterae (•). Curves are fit torectangular hyperbolae (parameters are given in the text). Each point represents themean and standard error of three growth rate determinations from a separate culture.Note two experiments are included in B) 90Figure 2.14. Nitrate recluctase (NR) activity versus N03 incorporation rate calculated fromgrowth rate and nitrogen quota for: A) Thalassiosira pseudonana (•), B) Skeletonemacostatum (B), and C) Amphidinium carterae (+). Points represent means and standarderror of 2 enzyme measurements from individual cultures. Dashed lines are leastsquares regressions. Solid lines represent the 1:1 relationships. Regression parametersare given in the text 91Figure 2.15. Nitrate reductase (NR) activity versus nitrate incorporation rate (calculated fromcell growth rate and cell nitrogen quota) for 12 species of marine phytoplankton. Solidline represents the least squares regression. Dashed line represents the 1:1 relationship.Points represent mean NR activities with standard error for 2 NR assays from duplicatecultures. 0 chlorophytes, • diatoms, D prasinophytes, B prymnesiophytes, Vcyanobacteria, V dinoflagellates. Species are indicated by abbreviations as outlined iMaterials and Methods. Equation of the regression line is: Y = -8.34 + 0.786 X (r’= 0.71) 92Figure 3.1. Cell composition versus light-limited specific &owth rate for Thalossiosirapseudonana grown between 6 - 120 JLmol quanta m’s1. A) Cell volume, B), cellcarbon quota, C) cell nitrogen quota, D) cell protein quota, and E) cell C:N ratio.Each data point represents the mean of duplicate determinations from a single culture.Open symbols represent three cultures where selenium limitation may have occurred.Lines represent least squares regressions. Parameters are given in the text 116Figure 3.2. Cell composition versus light-limited specific growth rate for Skeletonemacostatum. A) Cell volume, B), cell carbon quota, C) cell nitrogen quota, D) cellprotein quota, E) cell chlorophyll a quota, F) cell C:N ratio, and G) cellcarbon:chlorophyll a ratio. Where there are 2 types of symbols, they represent twodifferent experiments. Points represent the mean of duplicate determinations fromsingle cultures. Lines represent least squares regresssions. Parameters are given in thetext 119Figure 3.3. Cell composition versus light-limited specific growth rate for Amphidiniumcarterae. A) Cell volume, B), cell carbon quota, C) cell nitrogen quota, D) cellprotein quota, E) cell chlorophyll a quota, F) cell C:N ratio, and G) cellcarbon:chlorophyll a ratio. Where there are 2 types of symbols, they represent twodifferent experiments. Points represent the mean of duplicate determinations fromsingle cultures. Lines represent least squares regresssions. Parameters are given in thetext 120Figure 3.4. Nitrate reductase activity versus light-limited growth rate in Thalassiosirapseudonana. A) NR activity versus specific growth rate, and B) NR versus calculatedrate of nitrate incorporation. Each point represents the mean NR activity in a singleculture. Error bars represent standard errors of the mean of two NR assays. Solidlines represent least squares regressions. Dashed line represents the 1:1 relationship.Open symbols represent cultures where selenium limitation may have occurred 121Figure 3.5. Changes in cell composition ,n transition from low light (15 mo1 quanta m2s1)to high light (135 1mol quanta m’s i)(•), or high light to low light (0) inThalassiosira pseudonana. A) cell volume, B) cell carbon quota, C) cell nitrogenquota, D) cell C:N ratio. Transitions were made at the point indicated by the arrow.xiiEach point represents the mean and standard error of three separate cultures. Asterisks(*) indicate significant differences at P < 0.05 122Figure 3.6. Effects of transitions from high light to low light (open symbols) or low light tohigh light (closed symbols) on: A) growth rate and B) nitrate reductase activity (•, 0),or calculated nitrate incorporation rates(•, D) in Thalassiosira pseudonana.Transitions were made at the point indicated by the arrow. Each point represents themean and standard error of three separate cultures. Asterisks (*) indicate significantdifferences at P < 0.05 124Figure 3.7. Cell composition with growth rate in nitrate limited chemostat cultures ofThalassiosira pseudonana. A) Cell volume, B), cell carbon quota, C) cell nitrogenquota, D) cell chlorophyll a quota, E) cell protein quota, F) cell C:N ratio, and G) cellcarbon:chlorophyll a ratio. Each point represents the mean of two determinations froma single culture. Lines represent least squares regresssions. Parameters are given inthe text 125Figure 3.8. Relationship between nitrate reductase activity and: A) specific growth rate, or B)calculated rate of nitrate incorporation in nitrate limited chemostats of Thalassiosirapseudonana. Each point represents the mean NR activity in a single culture. Errorbars represent standard errors of the mean of two NR assays from a single chemostat.Solid lines represent least squares regressions. Dashed line represents the 1:1relationship 127Figure 3.9. Nitrate reductase activity scaled to different parameters versus growth rate ofThalassiosira pseudonana in light-limited batch (•), or nitrate-limited chemostat (0)cultures. A) Per cell volume, B) per g carbon, C) per g nitrogen, D) per g chlorophylla, and E) per g protein. Each point represents the mean of two assays from a singleculture. Lines represent least squares regression fits. Parameters are given in the text. 129Figure 4.1. Growth characteristics of log-phase cultures of Thalassiosira pseudonana grownon 14:10 h light:dark cycles. A) Culture densities, B) relative fluorescence, C) relativefluorescence per cell. Cultures were grown at high light (•, •), or low light (0, D).Points in A) and B) represent single determinations; points in C) represent means oftwo cultures, with standard errors of the mean 154Figure 4.2. Changes in cell composition in log phase cultures of Thqjas4siosira pseudonanagrown on 14:10 h light:dark cycles at low (6 jLmol quanta m’s’, 0) or high (45jmol quanta m2s,•) irradiance. A) Cell volume, B) cell carbon quota, C) cellnitrogen quota, D) cell chlorophyll a quota, E) cell protein quota, F) cell C:N ratio,and G) cell carbon:chlorophyll a ratio. Each point represents the mean of duplicatedeterminations from two separate cultures. Error bars represent standard errors ofmean values 156Figure 4.3. Nitrate reductase activity per cell in log phase cultures of Thalossiosirapseudonana grown on 14:10 h light:dark cycles at low (0) or high (•) irradiance.Each point represents the mean of two separate cultures. Error bars represent standarderrors of mean NR activity 157Figure 4.4. Nitrate reductase activity (0) or calculated nitrate incorporation rate (•) in twolog phase cultures of Thalassiosira pseudonana grown on 14:10 h light:dark cycles.Each point represents the mean of two enzyme assays. Error bars represent standarderrors of mean values 158Figure 4.5. Particulate nitrogen concentration measured (0) or predicted from nitratereductase activity (•) in two log phase cultures (A and B) of Thalassiosira pseudonanaxliigrown on 14:10 h light:dark cycles. Each point represents the mean of duplicatedeterminations 160Figure 4.6. Cell composition in log phase cultires of Thalassiosira pseudoncrna grown underequal irradiance (45 zmol quanta m2s1) of blue, white or red light. A) Cell volume,B) cell carbon quota, C) cell nitrogen quota, D) cell chlorophyll a quota, E) cellprotein quota, F) cell C:N ratio, and G) cell carbon:chlorophyll a ratio. Error barsrepresent standard errors of mean determinations from two separate cultures.Treatments not significantly different from one another at P = 0.05 are joined by lines. 161Figure 4.7. Effects of blue, white and red light on: A) specific growth rate, and B) nitratereductase activity (D) or calculated rates of nitrate incorporation () in log phasecultures of Thalassiosira pseudonana. Error bars represent standard errors of meandeterminations from two separate cultures. Treatments not significantly different fromone another at P = 0.05 are joined by a line 162Figure 4.8. Changes in A) cell number, B) culture fluorescence, and C) fluorescence per cell,in cultures of Thalassiosira pseudonana entering stationary phase, as indicated by thevertical line. Each point represents the mean of three replicate cultures. Error barsrepresent standard errors, or if not seen, are less than the size of the symbol 164Figure 4.9. pH and ambient nutrient concentrations for cultures of Thalassiosira pseudonanaentering stationary phase, as indicated by the vertical line. A) culture pH, B) nitrate,C) silicate, D) phosphate, E) ammonium, and F) nitrite. Each point represents themean of three replicate cultures. Error bars represent standard errors, or if not seen,are less than the size of the symbol 165Figure 4.10. Cell composition for cultures of Thalassiosira pseudonana entering stationaryphase, as indicated by the vertical line. A) Cell volume, B) cell carbon quota, C) cellnitrogen quota, D) cell chlorophyll a quota, E) cell C:N ratio, and F) cellcarbon:chlorophyll a ratio. Each point represents the mean of three separate cultures.Error bars represent standard errors, or if not seen, are less than the size of the symbol.Note that not all measurements were made at each time 166Figure 4.11. Nitrate reductase activity (•) or rate of nitrate incorporation calculated fromgrowth rate and nitrogen quota (0), increase in particulate nitrogen (0), or depletionof nitrate from the medium (A). Each point represents the mean of determinationsfrom three separate cultures of Thalassiosira pseudonana moving from logarithmicgrowth to stationary phase. Error bars represent standard errors, or if not seen, are lessthan the size of the symbol. Note that not all measurements were made at each time. 167Figure 4.12. Cell composition in log phase cultures of Thalossiosira pseudonana grown with75 M ammonium (NH4), 75 M nitrate (N03), or 75 M nitrate with daily pulses of2 M ammonium (P). A) Cell volume, B) cell carbon quota, C) cell nitrogen quota,D) cell protein quota, E) cell chlorophyll a quota, F) cell C:N ratio, and G) cellcarbon:chlorophyll a ratio. Each point represents the mean of two separate cultures.Error bars represent standard errors of mean determinations. Treatments notsignificantly different from one another at P = 0.05 are joined by lines 169Figure 4.13. Effects of growth on 75 M ammonium (NH4), 75 M nitrate (N03), or 75 Mnitrate with daily pulses of 2 M ammonium (P) on cultures of Thalassiosirapseudonana. A) pecific growth rate, B) specific nutrient uptake rates for nitrate ()and ammonium (0), and C) nitrate reductase activity (R) and calculated nitrateincorporation rate (0). Each bar represents the mean and standard error of twocultures, except ammonium cultures which are shown separately in B and C since theirresponses differed between replicates 170xivFigure 5.1. Diagram of phytoplankton processes in a coastal upwelling zone (after Wilkersonand Dugdale 1987). Environmental conditions such as low light or low nutrients causedecreases in rates of phytoplankton physiological processes (“shift-down”), while highlight and high nutrients cause increases in rate processes (“shift-up”) 189Figure 5.2. Study site, Monterey Bay, California. The map at left shows the location ofMonterey Bay (surrounded by box) in relation to the coast of California. “X” symbolmarks the location of the major sampling site 192Figure 5.3. Profiles of temperature (----), salinity ( ) and relative fluorescence (••) atsampling sites in the vicinity of the main sampling location (indicated by the “X”symbol on Figure 5.2) in Monterey Bay, CA. Profiles were taken: A) on day 3, earlyin the bloom, and B) on day 6 at the height of the bloom 196Figure 5.4. Effects of different assay conditions on nitrate reductase activity in naturalphytoplankton populations sampled from Monterey Bay, CA. A) NR activity inhomogenates used directly, or centrifuged to remove filter fibres. B) Effects ofadditions of NADPH in place of NADH, 0.1 mM FAD, or 0.2 mM ferricyanide(FeCN). Each bar represents the mean of three replicate homogenates. Error barsrepresent standard errors of the mean 197Figure 5.5. NR assay validation in natural phytoplankton samples taken from Monterey Bay,CA. A) Linearity of assay with time. B) Linearity of assay with homogenate addition.Lines represent least-squares regression fits to the data. Note that in B) the open point(0) is not included in the regression 198Figure 5.6. Kinetic curves for nitrate reductase activity in natural populations ofphytoplankton from Monterey Bay, CA. A) NR versus NADH concentration, B) NRversus nitrate concentration. Each point represents a single enzyme activitymeasurements from a single homogenate. Curves are fit to rectangular hyperbolae.Km values are 0.021 mM for NADH and 0.307 mM for nitrate 200Figure 5.7. Nitrate reductase activity in natural phytoplankton populations assayed at differenttimes after homogenization. Each point represents the mean of two enzyme assaysfrom different homogenates. Error bars represent standard errors of mean values.. 201Figure 5.8. Changes in biomass and ambient nutrient concentrations in contained naturalphytoplankton populations from Monterey Bay, CA for control cultures (•) andcultures with 5 M ammonium added (0). A) Chlorophyll a, B) particulate nitrogen,C) nitrate, and D) ammonium. Each point represents the mean of two separatecontained cultures. Error bars represent standard errors of the mean, or if absent aresmaller than the symbols. Cultures were grown under natural light and the black baron the time scale indicates the dark period 202Figure 5.9. Changes in nitrate reductase (NR) activity, and specific rates of nitrateincorporation calculated from changes in pari;ulate nitrogen , changes in ambientnitrate concentration, or saturated uptake of ‘‘NO3 (VNO3) for contained naturalpopulations of phytoplankton from Monterey Bay, CA. A) Control cultures withoutany nitrogen additions, and B) cultures receiving 5 jM ammonium additions at t = 0h. Each point represents the mean of determinations in two separate cultures. Errorbars represent standard errors of the mean, or if absent are smaller than the symbols.Cultures were grown under natural light and the black bar on the time scale indicatesthe dark period 203Figure A. 1. Absorbance versus protein content for different pure proteins and purified algalprotein from Thalassiosira pseudonana for A) Bradford, B) Lowry, and C) Smithxvprotein assays. For clarity, only one out of six sets of bovine serum albumin (BSA,•), bovine gamma globulin (BGG, •) and casein () data are shown. Lines representleast squares regression fits to pooled data. Fits to algal protein data for high lightgrown (0) and low light grown (C)) cultures are not shown. Note scale changes forprotein 270Figure A.2. Comparison of absorbance versus protein curves for A) Bradford, B) Lowry, andC) Smith protein assays for BSA samples with (•) and without (0) additions of 0.1mg chlorophyll a in 90% acetone. Points represent the means of two separatelyprepared standards. In all cases, associated error bars are smaller than the symbols.Lines represent least squares regression fits to the data 272Figure A.3. Comparison of protein content (expressed as pg cell-1) for acetone-extracted (C))versus non-acetone-extracted () homogenates of Thalassiosira pseudonana (n = 6 foreach treatment). Error bars represent standard errors of mean protein content 274Figure A.4. Comparison of protein content (expressed as pg cell1) for trichioroacetic acid(TCA) -precipitated () versus non-TCA-precipitated (C)) homogenates ofThalassiosira pseudonana (n = 5 for each treatment). Error bars represent standarderrors of mean protein content 275Figure B. 1. Screen output of absorbance versus time progress curves from the enzymekinetics program. From bottom to top, lines represent additions of 50, 20, 10, 5, 1, or0 mU of lactate dehydrogenase, respectively. Other assay conditions are described inthe text. Note that the screen image was taken only 6 mm. into the reaction 291Figure B.2. Sample data file output from the enzyme kinetic program of cell (sample)number, time (mm.) and absorbance. Cells 1 through 6 represent additions of 50, 20,10, 5, 1, or 0 mU of lactate dehydrogenase, respectively. Other assay conditions aredescribed in the text 292Figure C. 1. Data cases considered. Cases represent geometrically distributed data (Case 1),data where no points are lower than Km (Case 2), data where no points are higher thanKm (Case 3), data where only points higher than 2 x Km or lower than 0.5 X Km areavailable (Case 4), and data where all points fall between 2 X Km and 0.5 X Km (Case5). In each case, data sets of 10 points were generated with Vm = 10 and Km = 2.295Figure C.2. Examples of error levels (as percentages of V and S) assigned using Case 1 as anexample. Constant error levels are set as percentages of 0.5 x Vm. Errors wereassigned in a normal distribution 296Figure C.3. Examples of frequency distributions of Vm and Km estimates for variousfitting procedures for Case 1 data with 20% variable or constant error. Procedures are:0 Lineweaver-Burk, • Eadie Hofstee, V Hanes-Woolf, V Eisenthal and ComishBowden, C) Cleland-Wilkinson, and I Tseng-Hsu. Y-axis scale is relative percentage.True values of Vm and Km are 10 and 2, respectively 301Figure C.4. Examples of real data sets fit to rectangular hyperbolae using different fittingmethods: Lineweaver-Burk, —--- Eadie Hofstee, Hanes-Woolf,Eisenthal and Comish-Bowden, . -. -. -. Cleland-Wilkinson or Tseng-Hsu. A) Nitratereductase activity versus nitrate concentration in extracts of the diatom Thalassiosirapseudonana, B) phosphate uptake versus concentration in the marine macroalgae Fucusspiralis and C) growth rate versus prey concentration for the marine ciliateStrombidium p. feeding on the marine alga Rhodomonas sp. Parameters for each fitare given in Table C.3 305xviACKNOWLEDGEMENTSIt gives me the greatest pleasure to acknowledge the many people who have contributedto the work described in this thesis. In the course of experiments, I received excellenttechnical assistance from Urve Voitk and Steve Ruskey. In their capacity as curators of theNorth-East Pacific Culture Collection, Jeanette Ramirez and Elaine Simons providedphytoplankton cultures. Maureen Soon cheerfully analyzed CNS samples even when facedwith seemingly impossible deadlines. Bente Nielsen provided liquid nitrogen and advice on itsuse. Kedong Yin taught me the way of the Autoanalyzer. Carl Virtanen helped develop anddebug the software described in Appendix C, and programming tips from David Jones werealso valuable. Certain work could not have been accomplished without the loan of equipmentand cold room space by Drs. Tim Parson and Al Lewis. Chapter 5 exists thanks to thegenerosity of Dr. Dick Dugdale, and the help of the members of his research group (inparticular, Dr. Bill Cochlan, whose “field-wise” advice and attention to detail was critical tothe success of experiments), as well as the captain and crew of the R. V. Point Sur. In helpingme find my way through the labyrinths of university bureaucracy, the UBC Oceanographyoffice staff (and especially Chris Mewis) were indispensable. I have benefited from the adviceand critical comments of Dr. Yves Collos (reviewing my proposal), Dr. Steve Huber (NRpurification), Dr. J. N. C Whyte (protein analyses), Dr. David Karl (NDPK activity), and Dr.Quay Dortch (NR assays and relationships with growth and uptake rates).I cannot adequately express my thanks to the members of Oceanography, Botany andZoology who comprise the extended “Harrison Lab”. They have been my harshest critics andmy strongest supporters and must share credit for my successes. I express my appreciation tothree special friends in this group who have been with me since beginning the degree: RobertGoldblatt, Dr. Catriona Hurd and especially Dr. David Montagnes, who has turned me into acoffee snob. I have had the pleasure of working with Anne Fisher for somewhat less time, butshe has proven to be a valuable scientific colleague and a supportive companion. I alsorecognize the debt I owe to Dr. Peter Thompson, whose advice on culturing techniques andanalyses, contemplations on algal physiological processes, superb intuition and practicalapproaches have both helped and inspired me.Throughout my Ph. D. studies, the members of my supervisory committee (Drs. RogerBrownsey, Tony Glass and Sayid Ahmed) have challenged my ideas and helped me presentthem in the clearest manner. I also commend the additional members of my examinationcommittee (Drs. Steve Calvert and Ian Taylor) and my external examiner (Dr. Ted Packard)for facing what must have been a daunting document, and providing useful suggestions forimprovement as well as challenging questions.To my supervisor, Dr. Paul J. Harrison, I express my deepest gratitude. Paul hashelped me to develop and defend my ideas, and has given me every opportunity to join themainstream of the scientific community. His professionalism, patience, good humour, andgenuine concern for his students serve as examples to everyone involved in graduate education.He is also one of the most down-to-earth, committed and unassuming people I know.1GENERAL iNTRODUCTIONA principal goal of biological oceanography is to understand the flows of energy andmaterials through living systems and to understand their interaction with non-living systems(Parsons et al. 1984a). The first and probably most important interface between these systemsoccurs at the level of photoautotrophic organisms, primarily marine phytoplankton. Thus theecology of these organisms has a central role in oceanic processes. Sakshaug (1980)summarized the major goals of phytoplankton ecology: to relate intrinsic algal propertiesquantitatively to growth rates and identify the factors which limit growth.The importance of biological ratesCompared with terrestrial environments, biomass in marine ecosystems is very low, yettotal primary production is equal or greater (Kelly 1989) because the growth rates of marinephotoautotrophs are extremely high. While biomass measurements alone can provideimportant information about terrestrial systems, in the marine environment a preciseknowledge of rates of biological processes is also critical (Longhurst 1984, Valiela 1984, Kelly1989).The rate of primary production, i.e. the rate of increase of biomass (often expressed ascarbon) of the photoautotrophs is clearly an important quantity. Flynn (1988) has argued forthe use of more precise terms such as “photoautotrophic production”, “photosynthesis” orsimply “carbon fixation”, but “primary production” remains in common use. Primaryproduction sets an upper limit on the potential production of commercially important fisheries,and may ultimately be the critical term in setting sustainable harvests (Mann 1984). Despiteall the interactions and variation at points higher in the trophic food web, Iverson (1990) foundthat primary production was the single best indicator of fish production in a wide variety ofmarine systems. As well, concerns about global warming (the anthropogenic introduction ofcarbon into the atmosphere that is thought to lead to global temperature increases, seeBroecker et a!. 1979, Taylor and Lloyd 1992) have renewed efforts to estimate rates ofprimary production more precisely. As Broecker et al. (1979) pointed out, about 45% of the2carbon known to have been introduced to the atmosphere by human activity since the industrialrevolution cannot be accounted for in any current carbon pool. Oceanographers think that thisso called “missing sink” for carbon can be accounted for in the oceans, due to physicalprocesses that sequester dissolved CO2 for long periods of time in deep ocean waters, or byincreases in phytoplankton production that result in increased sedimentation of organic matterand thus increased carbon flux to the ocean floor (see Sarmiento and Siegenthaler 1992).However, terrestrial ecologists believe that carbon dioxide increases have caused a fertilizationeffect on terrestrial plants; the extra carbon may be tied up in increased production of tropicalrainforests and boreal forests, either in biomass or in forest floor litter (Taylor and Lloyd1992). Much uncertainty about the relative importance of phytoplankton in this the globalcarbon cycle remains. Based on a long-term data set of water clarity records, Falkowski andWilson (1992) could not distinguish a systematic increase in ocean production caused byincreased C02, but such measurements may reflect biomass and not necessarily production(see Welsh 1993). Good estimates of biological rate processes in the ocean are needed toresolve these questions.Since primary production is based largely on differences in growth rate rather than inbiomass, growth rates of marine autotrophs are a key factor. Unfortunately, measuring growthrates in the ocean is difficult. Eppley (1981) pointed out that balanced growth (i.e. matchingincreases of carbon, nitrogen and other elements) is rarely, if ever, seen in phytoplankton inthe ocean; temporal and spatial variations in irradiance, nutrient, and biomass lead tomismatches between cell division rates and nutrient uptake rates. Furthermore, the recyclingof nutrients (see Harrison 1992), and the presence of grazers (see Banse 1992) complicatethings further. Often the problem is not one of quantity or quality of data, but of interpretingdata within existing and potentially constraining concepts (see Eppley 1981). There are severalrecent examples of this phenomenon. Traditionally, there has been little doubt that largespecies of diatoms are very important in marine production. Using routine sampling methods,these species have been shown to comprise up to 25% of total net global primary production (aquantity comparable to the production of pine species in temperate and boreal forests, or of3grasses in savannah and cultivated areas), and to be capable of very high growth rates, and,thus, to dominate oceanic production (see Guillard and Kilham 1977, Thomas et al. 1978,Willen 1991). Since diatoms are most abundant in coastal regions, the open ocean areas havetraditionally been considered relative “deserts”, owing to their low diatom biomass and verylow nutrient concentrations (see Ryther 1969). However, there has been a growing awarenessthat small species (so called “picoplankton”, defined as organisms between 0.2 and 2.0 m indiameter) are abundant and may account for up to 80-90% of production in certain freshwaterand marine systems (see Stockner 1988). Previously, due to the concept that large diatomswere important, these picoplanktonic organisms had been under sampled and their trueimportance underestimated by the sampling techniques used. To some degree, concepts havenow shifted to the other extreme; phytoplankton ecologists are now focusing on the verysmallest organisms, the picoplankton (e.g. Li 1986, Stockner 1988). Goldman (1989) hasdemonstrated that, even if they exist in substantial numbers, these small cells can representsubstantially less production than a few large diatoms. Sampling designed with the importanceof picoplankton in mind would probably result in under sampling the large, rarer diatomspecies.A second example of the problem caused by unchallenged preconceptions concernsideas about the growth rates of oceanic phytoplankton species. Low nutrient concentrationswere previously thought to indicate nutrient limitation and thus low growth rates (e.g. Ryther1969). Work on bacteria and protozoa has shown that there is an active recycling of materialsat rates that are very high (see Pomeroy 1974, Azam et a!. 1983). This so-called “microbialloop” has led to the view that nutrients are so rapidly recycled that a “spinning wheel”develops and high regeneration rates support high growth rates. Traditional methods ofassessing growth rate are poorly suited to these situations and may have perpetuated the idea ofslow rates of growth (see Leftley et a!. 1983). Goldman et a!. (1979) have suggested thatbased on composition, cells in the open ocean are, in fact, growing at near maximal rates.This is based on the observation that there is a highly conserved ratio of elements in most ofthe ocean; C:N:P by atoms is usually 106:16:1, the so-called Redfield ratio (Redfield 1958,4but see also Takahashi et a!. 1985). In laboratory culture, organisms appear only to achievethis ratio when they are growing at their maximal (i.e. non-nutrient-limited) rates (McCarthyand Goldman 1979, Goldman 1980, but see also Tett et a!. 1985 and Goldman 1986 fordifferent interpretations of the ratio in the case of light versus nutrient limitation). Theseexamples perhaps illustrate a general feature of science; while paradigms are helpful inframing questions and directing research, they also bias efforts in particular ways (see Kuhn1970). The controversies mentioned above cannot be easily resolved because of theinadequacy of current methods of determining biological rates.Limitations of primary productionBefore examining the methods and the problems with the determination of rates ofgrowth, it is first necessary to understand how growth rates are limited in the ocean. Marineprimary producers are ultimately constrained by the quantity of light energy available and thisis particularly limiting in polar regions during dark periods, in winter temperate regions wheremixing processes drive phytoplankton deep into the water column, and for phytoplanktonpopulations deep in the water column or in turbid waters such as estuaries (see Parsons et al.1984b). However, the quantity of mineral nutrients is more often likely to cause limitation ofproduction (see review by Platt et a!. 1992a). This has been expressed as Liebig’s “law of theminimum”, which states that the material in lowest concentration relative to the requirementwill be limiting (Liebig, 1840). Therefore, identifying the limiting nutrient becomes animportant problem to solve (see e.g. Dugdale 1967, Droop 1973).Thingstad and Sakshaug (1990) have pointed out that there is a problem withterminology between descriptions of so-called “controlling” and “limiting” factors. Forexample, in a chemostat culture where cell growth is regulated by the addition of a singlenutrient (see Rhee 1979), it is the particular nutrient that is “limiting”, but it is actually therate of supply of that nutrient that is “controlling”. In most cases, it is experimentally difficultto separate limitation from control. In this thesis, mindful of the distinction, the term“limiting” will be used. To add another complication, as pointed out by Dugdale (1967), and5re-emphasized by Falkowski et at. (1992), the limiting nutrient may limit growth rate or theultimate biomass achieved by a phytoplankton community, or its effects may only alter speciescomposition. For example, a deficiency of silicate, a nutrient required by diatoms, may notchange the primary production of a system, but it may result in the decline of diatoms, andtheir succession by other species. Alternatively, if only the biomass is limited by a nutrient,primary production can be calculated from a knowledge of light alone, while if growth rateitself is nutrient-limited the case becomes much more complex (see Falkowski et at. 1992).In marine systems, primary production is probably most often limited by nitrogen,phosphate or silicate. It is worth noting that although there are many nitrogen forms, P and Siare available to phytoplankton in seawater predominantly as phosphate and silicate,respectively) (see Raymont 1980, Parsons and Harrison 1983). The classical view of themarine ecosystem held that nitrogen was the limiting nutrient (e.g. Ryther and Dunstan 1971),and indeed there is clear evidence that this is true in many cases (McCarthy 1980, Cobs andSlawyk 1980, Wheeler 1983). However, there is also evidence that phosphate may be limitingin some locations such as estuaries and coastal regions (e.g. Harrison et at. 1990a, Peeters andPeperzak 1990, Fisher et at. 1992). In the view of many geochemists, phosphate should bethe limiting nutrient, based on the results of modeling phosphorus budgets and flux rates (seeSmith 1984), although the turnover rates of phosphate are also very high (see Lean and Cuhel1987). Silicate limitation has also been reported in coastal and polar regions (e.g. Nelson andTreguer 1992, Egge and Asknes 1992, Brzezinski 1992), although its effects are more likely toinvolve shifts in species composition than changes in production (see Dortch and Whitledge1992). At least part of the problem lies in the inadequacy of methods used to determinewhether a particular nutrient is limiting (see discussion in Peeters and Peperzalc 1990).Recently, it has been suggested that iron (or other trace metals, see Morel et at. 1991)may limit primary production in large areas of the ocean (see review by Martin 1992, but seealso Cullen 1991). Iron limitation has been tested in previous enrichment experiments (e.g.Menzel and Ryther 1961, Tranter and Newell 1963, Menzel et at. 1963, but note that Menzeland Spaeth 1962 found different results), but has more recently become a topic of controversy,6brought on by the proposal that iron could be used to fertilize tracts of the ocean and therebyincrease primary production and draw down elevated atmospheric carbon dioxide levels (seeChishoim and Morel 1991). It is unclear whether fertilization results in a change inproduction, or in a shift in species (see Banse 1990, Wells et al. 1991, DiTullio et al. 1993),or changes in preference for nitrogen forms (see Price et al. 1991) There is also disagreementover the method of calculating growth rates in iron enrichment experiments (compare Martinet al. 1991 and Banse 1990). There is little doubt that the poor ability to determine growthrates is a central problem here.Nutrient limitation can prove to be complicated. Howarth and Cole (1985)demonstrated the potential for interactions between nutrients. In seawater, there is evidencethat high sulphate concentrations result in an impaired ability of phytoplankton to take upmolybdenum. Since molybdenum is required for activity of nitrogenase during N2 fixation,and in enzymes involved in nitrate and nitrite reduction, this may result in additional costs tomarine species to perform these functions, but may manifest itself as nitrogen limitation. Thesame sort of interaction may occur between nitrogen and iron (see Price et a!. 1991).Finally, in a radical departure from the N, P, Si, or Fe limiting nutrient paradigm,Riebesell et al. (1993) recently suggested that carbon may be limiting to marinephytoplankton. Because the pH of seawater is usually close to 8.2 (Riley and Chester 1971),and can rise further during blooms, this means that while the total concentration of carbon inseawater is usually adequate to support production, the actual concentration of CO2 is low dueto the chemical equilibrium favoring HCO3 over CO2. Riebesell et al. (1993) argue thatmany species are unable to use HC03 effectively; thus, in some cases carbon may belimiting. To date, this observation has been confined to laboratory cultures, and there hasbeen criticism (see Turpin 1993) of the experiments performed by Riebesell et al. (1993).Pliytoplankton nitrogen metabolism and the concept of new productionNitrogen is considered the principle limiting nutrient in the oceans. Marinephytoplankton cells obtain inorganic nitrogen principally as nitrate and ammonium (see7reviews by McCarthy 1980, Syrett 1981, Collos and Slawyk 1983, Wheeler 1983, Syrett1989). There are also cyanobacterial species that can fix atmospheric nitrogen gas (N2), andalthough this represents less than 1% of oceanic nitrogen globally, it can represent up to 20%in areas such as the Baltic Sea (Howarth et al. 1988). Uptake of forms of dissolved organicnitrogen is poorly understood due to analytical difficulties; urea and ammo acids may beimportant at times, but peptides, proteins, ureides, amino sugars, and pyrimidines and purinesalso contribute (see Antia et a!. 1991 for an extensive review). These organic compounds arealso used by bacteria, and thus may compete with phytoplankton for these nutrients. Paleniket a!. (1989) and Palenik and Morel (1990) have reported that amino nitrogen may be madeavailable for uptake as ammonium by the action of cell surface amino acid oxidases that occurin many marine phytoplankton species.An important division exists between production based on different nitrogen forms; thesource of the nitrogen is related to the ultimate fate and importance of the production.Dugdale and Goering (1967) first drew attention to the concepts of “new” and “regenerated”production. As illustrated in Fig. 1, the euphotic zone in many areas of the ocean can betreated in isolation due to density stratification of the water column resulting in a thermoclineor pycnocine. Nitrogen can enter from upwelling of deep water (principally as nitrate), fromatmospheric deposition (e.g. rain, as nitrate and ammonium, see Paerl 1985, Duce 1986) andfrom the atmosphere as N2 gas for nitrogen fixation. These forms are termed new since theyhave entered the system for the first time. Regenerated nitrogen arises from that incorporatedinto organisms (particulate organic nitrogen, PON) that is released by death or by cell lysis,grazing, or viral attack (see Suttle et al. 1990) principally as ammonium, urea and amino acids(see Harrison 1992). The distinction between these forms lies in the fact that, if systems areassumed to be in steady state, and if they do not “run down”, only an amount of material equalto the new production can be exported from the system via sedimentation, fish production, etc.(see Longhurst 1991, Platt et al. 1992b). This sets practical upper limits on the potential fishharvest, for example, but it also has interesting implications for the global carbon cycle. Sinceexport will result in a loss of carbon from the system, while regeneration is accompanied by8respiration which will return carbon to the atmosphere, new production is a critical factor inthe global CO2 budget. New production can be viewed as a “biological pump”, sequesteringCO2 in the deep ocean for long time scales; somewhat cynically, Platt et a!. (1992b) havelikened it to sweeping the “pollution of the atmosphere... under the rug of the thermocline”.Eppley and Peterson (1979) extended the concept of new production further by noting thatsystems with higher total primary production also have a greater fraction of new production;this is expressed as the f-ratio, the ratio of new production to total production. Based on thisanalysis, productive upwelling zones, where nitrate-rich water is brought to the surface maywell be the most important areas for new production (Dugdale and Willcerson 1992), but thereis also evidence that nitrate is upwelled episodically in vast areas of the open ocean, and sothey may also contribute substantial new production (see Lewis et a!. 1986, Glover et a!.1988, Platt et a!. 1989, Eppley et al. 1990). As in the case for growth rates in general, thereis much controversy which arises because of problems in methodology (see Platt et al. 1989),and the assumption that oceanic regions are in steady state may not be strictly valid (Platt et a!.1989). For example, Jaques (1991) pointed out that the Southern Ocean has additionalcomplications such as the removal of nitrate from the euphotic zone when deep water isformed.9precipitationdeposition(NO,NH)2 fixationamm, photiczoneNO3N PONurea, DON PON(phytoplankton)[Iankton fluxnkt)grazing, deathupwe ingfish,etc.sedimentationNO3 4= remineralizationNEW PRODUCTION REGENERATED PRODUCTIONFigure 1. Diagram of the main processes involved in new and regenerated production in the upper water column.See the text for explanations. (PON = particulate organic nitrogen, DON = dissolved organicnitrogen).10Traditional methods of estimating biological rates and thefr problemsAs previously mentioned, many of the concepts of marine ecosystems and theirfunctioning arise from, or have been biased by, the methods used to estimate biological ratesin natural environments. There are problems common to methods of assessing growth ratesand rates of new production, and problems distinct to each set of methods.l’ IncubationsThe most common method of assessing production rates is by timed incubation.Clearly, given the enormous spatial and temporal variability in the oceans, any attempt to takeserial samples from the ocean and monitor increases in in situ biomass (e.g. particulate carbon,nitrogen, or chlorophyll a), or decreases in in situ nutrients (e.g. nitrate or ammonium) willresult in poor resolution. Thus, incubations of contained water samples are most often used.For growth rates, photosynthesis or carbon fixation is monitored. Harris (1984) cautioned thatgrowth is not simply equal to photosynthesis, nor is photosynthesis simply equal to carbonfixation or oxygen production. Traditionally, oxygen production has been measured, but thismethod gives poor sensitivity and relies on a photosynthetic quotient to convert from oxygenproduction to carbon fixation, which has associated problems of its own (see Laws 1991).However, there is some evidence that, in certain areas of the ocean, the oxygen method maybe better than alternatives (e.g. Harris 1984). Currently, the standard method involves use ofa radioactive isotope of carbon. Radio-labeled bicarbonate is introduced into theincubation vessels and time-dependent incorporation into particulate material is measured (seeSteemann-Nielsen 1952, Parsons et al. 1984a). Many of the problems with this method arisefrom the necessity of collecting and incubating cells in containers; these problems have beenextensively reviewed (see Venrick eta!. 1977, Li and Goldman 1981, Leftley et a!. 1983,Harris 1984, Gieskes and Kraay 1982, Li 1986, Harris et a!. 1989, Collos et a!. 1993). Forexample, the process of collecting cells and placing them in incubation bottles may damagedelicate species (e.g. Krupatkina 1990), and that once inside the bottles, collisions of cells with11the walls of the bottle, or the changes in turbulence scales will cause cell damage (e.g. Allen1977). Upon containment, the parcel of water is separated from processes such as introductionof nutrients, or removal of waste products by advection, and thus the biological behaviour ofthe sample may change markedly. Grazers of phytoplankton may be selectively removed, andthus production can be enhanced (see Collos et at. 1993), or the containment of a grazer mayimprove grazing ability leading to artificially lower rates. Alternatively, removal of grazersmay decrease production by preventing nutrient regeneration. Trace metals found inincubation bottles and sampling equipment (even wire lines) may present additional problems,both by inhibiting sensitive species, and by enhancing production where metals are limiting(see Fitzwater et al. 1982, Price et at. 1986). All of these problems may account for thesometimes dramatic shifts in species composition seen in contained versus natural samples(e.g. Venrick et at. 1977). The volume of water contained in an incubation bottle can alsomake a large difference, especially in oceanic waters where larger bottles give higherproduction estimates (see Gieskes et at. 1979), but the volume effect may not be significant incoastal regions (Leftley et a!. 1983). The length of incubation is also an issue. Collos et a!.(1993) found that containment altered regeneration rates within bottles and also carbon isotopefractionation. They advocated keeping incubations less than 3 h. Shorter times minimizecertain containment problems, but make extrapolations to longer (e.g. 24 h) time scalesdifficult. The light climate during the incubation is also a problem. Traditionally, in situmethods have bottles returned to the same depths from which they were sampled, on some sortof fixed array. The problem here is that the bottle or container may change the natural lightspectrum (e.g. borosilicate glass will screen out UV radiation) (see Smith et a!. 1980), and onoceanographic cruises it is frequently impossible to stop for extended incubation periods. Analternative is the simulated in situ method where samples are placed in shipboard incubators atappropriate light levels. Very often, however, the light is controlled by neutral densityscreening, while in situ, a change in light spectrum towards more blue light would occur.Laws et al. (1990) found that primary production rates estimated using neutral densityscreening were half of those estimated using in situ incubations. Comparisons of the in situ12and simulated in situ methods indicate that there are differences; Lohrenz et at. (1992) foundthat differences were greatest during short term incubations, and advocated longer incubations,but longer incubations would increase many of the containment effects (e.g. grazing)previously discussed. In both methods, cells are held at a single light level, This is notnatural because cells would normally circulate in the upper mixed layer of the ocean and beexposed to a varying light field. There is evidence that this can make a large difference.Ferris and Christian (1991) reviewed seven studies using fluctuating versus steady lightregimes; production in the fluctuating versus the steady light was decreased in 2 cases,enhanced in 3 cases and showed no difference in 2 cases. Mallin and Paerl (1992) reportedthat varying light reduced photoinhibition and resulted in markedly higher growth andproduction rates in estuarine phytoplankton.After incubation, there is the problem of how to collect the cells and removeunincorporated 14C. Collection is usually done by filtration, but there is evidence of problemswith filter retention of cells (e.g. Stockner et at. 1989). Equally, the rupture of delicate cellscan lead to losses, depending on filter type and filtration pressure (Venrick et at. 1977). Acidfuming or acid rinsing of cells can be used to remove the unincorporated label, but there areunaccountable losses using these methods (see Hilmer and Bate 1989). Alternatively, thewhole sample can be acidified and bubbled to remove unincorporated label. Riemann andJensen (1991) reported up to 57% higher incorporation rates using this method. To correct fornon-specific carbon uptake, a control in which light is blocked is often run, but there can besubstantial carbon uptake in the dark (Leftley et at. 1983). Alternatively, the photosyntheticinhibitor 3-(3,4-dichlorophenyl)-1, 1-dimethylurea (DCMU) can be added (Legendre et a!.1983), or blanks filtered without incubation (i.e. a zero time blank) can be used. The issue ofexcretion of dissolved organic carbon (DOC) from phytoplankton could be critical. Ziotnikand Dubinsky (1989) found that light and temperature both caused variation in DOC release inlaboratory cultures of phytoplankton. The range of excretion fell between 1 and 55 % ofprimary production. Harris et at. (1989) showed that up to 50% of carbon fixed could be lost13in an 8-12 h dark period in the tropics. Bacterial respiration of excreted carbon, and grazingby microheterotrophs were thought to be responsible.When added up, the magnitude of the error caused by these problems can be extreme.Li and Goldman (1981) reported that in a test using laboratory cultures, the 4C methodroutinely over-estimated growth for certain species by up to 40%, yet under-estimated otherspecies by 40-100%. They pointed out that the species used were probably hardier than themajority of species in the field, so the situation with natural populations may be even worse.15NIncubationsIn the case of new production determinations, incubations with the stable isotope 15Nas a tracer are used, typically provided as either nitrate or ammonium (see Dugdale andGoering 1967, Dugdale and Wilkerson 1986). As an incubation technique, this method sharesthe same problems of containment effects, and filtration collection illustrated in the 14Cmethod. Added bicarbonate will probably not affect incubations in the 14C method, butadditions of labeled nutrients is a greater problem, particularly in nutrient-depleted waters(Dugdale and Wilkerson 1986). In addition to these problems, there are problems with theloss of 15N label. Bronk and. Glibert (1991, 1993) looked at release of dissolved organicnitrogen (DON) during incubations and concluded that both uptake and release of DON occur.Releases measured were between 5 and 54% of uptake. Collos et al. (1992b) found thatlaboratory cultures released up to 63% of the nitrate they took up in the first hours of the lightperiod as DON, but took up most of this DON again in the following dark period. This couldlead to serious problems with short-term incubations.Finally, it should be pointed out that all of these techniques are populationmeasurements; measuring any of these parameters at the level of a species or individual taxonis impossible without modifications.14Alternative approaches to estimating growth ratesRecognition of the problems outlined above has led to a variety of ingenious ways toovercome them (see Ceccaldi 1981). Furnas (1990) has divided such methods into fourcategories: a) biochemical or cell cycle markers, b) biochemical rate measurements andrelative growth indices, c) cage and bottle incubations, and d) mathematical models.Cell Cycle MarkersA variety of methods are based on determining the frequency of occurrence of cells in aspecific stage of the cell cycle in phytoplankton populations. In general, the higher thefrequency, the higher the growth rate. Such “frequency of dividing cell” methods rely onbeing able to identify a particular stage of the cell cycle (usually by microscopic examinationof stained cells) and on knowing the time spent in this stage, a variable which can be difficultto determine (MacDuff and Chishoim 1982). In theory, no incubation is required, andspecies-specific information can be obtained (Chang and Carpenter 1988). Chang, Carpenterand others have explored this idea in great detail, moving from theory to the lab, to the field(Chang and Carpenter 1988, 1990, 1991, Antia et al. 1990). It has also been applied tocyanobacteria and picoplankton (Li and Dickie 1991 and Vaulot 1992). Flow cytometry canpermit the method to be applied to individual cells (Chishoim et al. 1986).Biochemical Rate Measurements and Relative Growth Rate IndicesBiochemical and growth rate indices have also been a fruitful area of research (note thatenzymatic indices will be considered separately). There have been a variety of methods basedon chlorophyll a fluorescence. Eppley and Sloan (1966) attempted to use light absorption bychlorophyll to estimate production, but species variation, different photosynthesis versusirradiance relationships and temperature effects caused problems. Chamberlin and Marra(1992) have obtained good correlations between natural fluorescence and photosynthetic rates,but a lack of basic understanding of the processes involved has hampered progress. By15enhancing the fluorescence of chlorophyll by adding DCMU, Furuya and Li (1992) found arelationship between fluorescence and photosynthetic capacity, but for different species therelationship showed over 5-fold variation. By stimulating cells with light pulses, informationabout natural maximum photosynthetic rates has also been obtained (Falkowski et al. 1986,Falkowski et a!. 1992). There have been detailed investigations of cell ATP (Sheldon andSuttcliffe 1978, Laws et a!. 1985, Noges 1989) or GTP (Karl 1980) contents (or variousratios) as growth rate indices, but these vary with the specific limitation and are relativelyinsensitive. Content of RNA, or various ratios of RNA, DNA, and protein have beenconsidered in laboratory and field populations (Dortch et a!. 1983, Dortch et a!. 1985) andsensitive and convenient methods to determine DNA and RNA are available (Berdalet andDortch 1991). Unfortunately, these methods are frequently highly specific to individual taxa,and are biased by detrital material (see Dortch et al. 1985).Modified Incubation TechniquesModifications to the traditional 14C method, and variations on incubation techniqueshave been proposed to overcome problems. Jespersen et a!. (1992) concluded from a reviewof carbon-specific incorporation rates that they often gave growth rates which were too high.Use of 3H incorporation into DNA using labeled adenine was proposed, but rates estimated bythis method were usually much too low. Use of3H-labelled thymidine has also beenproposed, but incorporation of this compound in natural samples appears to be exclusively dueto bacteria (Moriarity 1986). Using traditional 14C methods, but looking at pigment-specificactivities can allow carbon-specific and species-specific production rates to be inferred (seeWelschmeyer and Lorenzen 1984, Gieskes and Kraay 1989) Alternatively, DNA-specific dyeshave been used to monitor increases in DNA instead of increases in carbon in incubations (seeFalkowski and Owens 1982). Irwin (1991) proposed that coulometric methods (measuringconsumption of C02) could be used alongside 14C measurements to provide an independentcheck, or on their own when isotope use was inappropriate or difficult. Estimating proteinsynthesis rates using radioactive sulphur has also achieved some success (e.g. Bates 1981,16Cuhel and Lean 1987). In theory, protein is better conserved along food chains and so may bea better index of net phytoplankton production (see Lean et al. 1989). There are, however,methodological problems involving blank activities (Bates 1981). Alternatively, by using post-incubation separations, Madariaga et al. (1991) found population characteristics ofphotosynthesis and growth rate could be determined using the ratio of 14C incorporated intoprotein over that incorporated in low molecular weight metabolites. Another approach togrowth rate in incubations involves removal of grazers. This can be accomplished byfiltration, or selective heterotrophic inhibitors, but a less invasive method is simply to dilutethe sample sequentially and measure grazing effects (see Li 1986). The y-intercept of a plot ofgrowth rate versus degree of dilution should give the growth rate in the absence of grazing,although there are a number of problems, including nutrient depletion in the absence ofregeneration (see Li 1986). Unfortunately, all of these methods still share the variety ofproblems associated with incubations previously discussed. In terms of modifying incubationmethods themselves, Langdon et al. (1992) and Dandonneau and Bouteiler (1992) have bothproposed remote, in situ samplers which would permit collection and inoculation of isotopeinto vessels without the problems associated with sample manipulation. Furnas (1991) hadsuccess using specially constructed diffusion chambers in place of static bottle incubations.Mathematical ModelsBeyond these experimental methods, modeling offers promise as well. Modelspredicting growth from irradiance alone, or combinations of irradiance, nutrients andtemperature are very popular and diverse and include both mechanistic and empiricalformulations (see Laws et al. 1985, Keller 1989, Laws and Chalup 1990, Sakshaug et al.1991). Cullen (1990) reviewed several models that attempt to predict growth rates fromirradiance, concluding that most of the many available mechanistic models are based on nearlyidentical theoretical bases. Species-specific differences and the problem of short-termunbalanced growth present the greatest difficulties. In a different approach, Lande et a!.17(1989) estimated growth rates in situ from depth profiles of cell concentrations and turbulentdiffusion measurements. Such methods will be aided by new techniques that increase thelower limit of nitrate detection (e.g. Garside 1982, Raimbault et al. 1990, McCarthy et a!.1992). When theoretical and empirical models are combined, it may be possible to predictproduction from surface ocean temperature and ocean colour (see Campbell and O’Riely 1988,Prasad et al. 1992) However, these methods may only be sensitive to processes in the verysurface layer, and must still rely on calibration and validation by surface measurements (seePlatt and Sathyendranath 1988).Alternative approaches to estimating new productionFor new production techniques, several alternatives exist to 15N incubations. Sinceexport production is of interest, measuring the rate of organic sedimentation below theeuphotic zone using moored or free-floating sediment traps is a direct approach (Eppley 1989,Platt et a!. 1992b), although there are problems concerning trap recovery and efficiency (seeGardner 1980, Honjo et a!. 1992). Bulk properties such as the increase in oxygen in theeuphotic zone due to photosynthesis, or the decrease in oxygen below the euphotic zone due todecomposition of sedimented material (apparent oxygen utilization) have been proposed (seePlatt et a!. 1992b). These can be given a time scale by measuring the311:He ratio as anindex of time since departure from the surface mixed layer, but may fail in the presence ofsmall scale mixing and heterotrophic activity. They are also difficult to correlate meaningfullywith biological processes on short time scales. By modeling circulation and mixing, the fluxof nitrate from deeper water can be predicted, giving some estimate of new production (Platt eta!. 1989, 1992b). Isotopic disequilibrium is also a promising method. In the isotope pair238j and 234Th, apparently only Th (thorium) adsorbs to organic particles. When theseparticles sink from the euphotic zone, the change in the 238U:4Thratio is therefore apredictor of export of organic carbon (Platt et a!. 1992). In all these cases, however,relationships between carbon and nitrogen must be assumed to allow interconversion. Laws(1991) pointed out that this is not always suitable; C:N ratios frequently exceed those predicted18by the Redfield ratio (i.e. 6.6 C: 1 N). Thus, when based on nitrate data, the export of carboncould be 15-30% higher than would be predicted using the Redfield ratio. The satellite imagemethods discussed above may be extended to estimate new production by assuming arelationship between nitrate and surface temperature (see Campbell and Aarup 1992). In manyof these methods, the question of whether the parameters are being measured on a time scalerelevant to the organisms performing the processes is difficult to resolve.Enzyme activity: a theoretical basis for predicting biological ratesIn addition to methods considered previously, enzymes may provide a unique approachto estimating biological rates. Since enzymes are catalysts of the biological reactions ofinterest, and since they are adapted in character and concentration to meet prevailing demandsof organisms (Hochachka and Somero 1984), the theoretical basis for studying enzymes isstrong. There are however practical limitations.Enzyme activity can be measured in two basic ways: the quantity of enzyme proteincan be estimated by purification, or immunoassay, or the activity of the enzyme can bemeasured. Assay of enzyme concentration using immunoassays is currently complicated andsubject to many of the same problems as the assay of enzyme activity, but offers extremelyhigh sensitivity and is becoming more routine (see Balch et al. 1988, Rosalki 1989, Parker1990; see also Chapter 2). Practically, when enzyme activity is assayed, in order to bereproducible, it is the maximal activity (Vmax) that is measured. This is done with allsubstrates and cofactors at saturating levels, so that under these conditions the maximal activityshould be proportional to the enzyme concentration (see Rossomando 1990). However, anumber of factors may cause this not to be true. To begin with, enzyme extraction maydamage the protein, or assay conditions may not support optimal activity (Newsholme andCrabtree 1986). Furthermore, the in vitro activity of the enzyme may bear little relationship toin vivo activity because in addition to varying enzyme activity by changing enzymeconcentration, activity can be altered by changing the chemical nature of the enzymes (i.e. bymanufacturing proteins with different characteristics that perform the same function, i.e.19isozymes), or by altering the environment of the protein in situ by changing variables such ascell surface to volume ratio, or cell membrane composition (see Raven 1981). On a finerscale, the enzyme itself may be covalently modified in processes such as adenylation orphosphorylation, or non-covalent mechanisms such as allosteric modification, or regulation ofactivity by substrate supply may prevail (Raven 1981). Each of these mechanisms may act ondifferent time scales and may or may not be detected in assays, depending on the precisehomogenization procedures and assay conditions. Despite the variety of methods of regulationand cell metabolic control, there is still evidence that maximal enzyme activity can be used toaccurately estimate maximal in vivo rates of metabolism (see Newsholme and Crabtree 1986).Application of enzyme activity measurements to planktonic organismsThe potential of enzyme determinations has not been lost on oceanographers. In all buta few cases, it is enzyme activity and not enzyme concentration that has been measured (butnote Balch et at. 1988, Orellana et a!. 1988, Wood 1988).ETS MeasurementsIn phytoplankton, electron transport system (ETS) activity has been proposed as anindex of respiration, or general metabolism (Packard et at. 1971, Kenner and Ahmed 1975a,1975b, Packard 1985), and ETS may also be related to growth rate (Martinez 1992). It hasbeen used in marine (Romano et al. 1987, Packard et at. 1988) and freshwater systems (Rai1988), although the precise relationship between ETS and respiration is often in question (seeKenner and Ahmed 1975b, Martinez 1992).Enzymes Involved in Carbon MetabolismAs an estimate of carbon fixation, the enzyme ribulose 1 ,5-bisphosphate carboxylase(RUBISCO) has proved useful. RUBISCO and photosynthesis have been correlated inlaboratory cultures (Hellebust and Terbough 1967, Descolas-Gros 1982, Hobson et at. 1985,Smith and Platt 1985, Rivkin 1990), and in the field (Priscu and Goldman 1983, Li et at.201984). Glover and Morris (1979) found that RUBISCO activity was highly variable in fieldsituations; it could account for less than half the variation in photosynthetic rate. However,Orellana and Perry (1992) showed good correlations between maximal photosynthetic rates andRUBISCO concentration in cultures, using an immunoassay. Assays of some othercarboxylases involved in carbon fixation, phosphoenolpyruvate carboxylase (PEPCase) andphosphoenolpyruvate carboxykinase (PEPCK), have provided insight into rates of C4 carbonfixation, and carbon isotope fractionation (Morris et a!. 1978, Descolas-Gros 1982, DescolasGros and Fontugne 1985, Descolas-Gros and Fontugne 1988)Enzymes Involved in Mtrogen MetabolismEnzymes of nitrogen metabolism have been explored as indices of nitrogenincorporation rates. Nitrate reductase (NR) has been used to predict nitrate incorporation rates(e.g. Eppley et al. 1969, Blasco et al. 1984), but the relationships have been highly variableand generally difficult to relate to other field measurements. NR immunoassays have alsobeen proposed (Baich et al. 1988). Nitrite reductase (NiR) has also been explored, withlimited success (see McCarthy and Eppley 1972). Enzymes involved in ammoniumincorporation have also received attention. Glutamate dehydrogenase (GDH) was initiallyidentified as a key enzyme of interest (see Ahmed et al. 1977), but activity seldom correlatedwell with incorporation rates (McCarthy and Eppley 1972, Dortch et a!. 1979). Subsequently,it was shown that glutamine synthetase (GS) is more likely to be the pathway for ammoniumincorporation (Miflin and Lea 1976). The emphasis shifted to this enzyme both in culture(Falkowski and Rivkin 1976, Thomas et a!. 1984, Everest et a!. 1986, Slawyk and Rodier1986, 1988) and in the field (Clayton and Ahmed 1987). In some cases, GS activity hascorrelated well with ammonium uptake, but this is not true during starvation, or perturbationsin nutrient levels (Slawyk and Rodier 1986).21Other EnzymesOther enzymes have been used as indicators of specific states, rather than indices ofrates. Alkaline phosphatase (APase) has served as an indicator of phosphate deficiency inmarine and fresh waters (Perry 1972, Jansson 1976, Salcshaug et al. 1984, Davies and Smith1988). Cell surface amino acid oxidases have also been studied, as indicators of the ability ofcells to use combined amino nitrogen (Palenik and Morel 1990). As well, Price and Morel(1990) describe a number of other exoenzymes, including metal reductases, and proteases.Enzymes in Other Planktonic OrganismsThe oceanographic use of enzyme activities as biochemical indices of metabolic rateshas not been limited to phytoplankton. Bacterial exoenzymes, involved in degradation oforganic compounds, have been used as indices of bacterial activity (Chrost et al. 1989, Smithet a!. 1992, Martinez and Azam 1993). Aspartate transcarbamylase (ATCase), a key enzymein nucleotide synthesis, has been measured as an index of secondary production in zooplankton(Bergeron 1983, 1986, 1990) although a relationship between ATCase and secondaryproduction has not been shown under controlled conditions. ETS and GDH activities havealso been applied to zooplankton to estimate respiration and excretion rates, respectively(Bidigare and King 1981, Bidigare et a!. 1982, Mayzaud 1987), but there are indications thatthese enzyme indices may be biased by size effects on metabolic rates (see Berges et a!. 1993).As well, digestive enzymes such as amylase and trypsin have been studied in zooplankton in aneffort to estimate feeding rates (Samain et al. 1983, Mayzaud et a!. 1984, Hasset and Landry1990a, 1990b), but there are problems with the complex spatial variability of enzyme activityin the field (see Hirche 1989) and the slow response time of the enzymes to new conditions(see Roche-Mayzaud et al. 1991). Berges (1989) and Berges et a!. (1990) suggested thatnucleoside diphosphokinase (NDPK), might be useful as a predictor of zooplankton growthrate. As with other enzymes, the use of NDPK activity does not necessarily yieldstraightforward results. For example, NDPK response varied with animal growth stage,22perhaps due to differences between stages of growth by cell proliferation versus stages of cellsize increase (Berges 1989).Despite a strong rationale for the use of enzyme activity, and many cases whereenzyme assays have been applied, the success of these enzyme methods has been equivocal.Relationships of enzyme activities with other indices of biological rates have been highlyvariable. A major problem has been that the enzyme index is frequently applied to a fieldsituation before there has been adequate laboratory investigation. Leftley et al. (1983)criticized marine ecologists for their over-enthusiasm in taking fledgling methods to the field.The problem is that in the field there is usually no way to independently measure the biologicalrate of interest, except by using the very methods whose inadequacy prompted development ofthe enzyme methods in the first place. When traditional methods and enzyme-based methodsdisagree, it is unclear which is correct, or in fact if either is correct. In general, detailedlaboratory studies have not been conducted first.Organization and goals of this thesisIn this thesis, work began with laboratory studies using unialgal cultures under steadystate conditions. Complexity of experiments was gradually increased until enough confidencein the techniques and in the relationships between enzymes and their associated rate processeshad been gained so that preliminary field work could be attempted.For these studies, the diatom Thalassiosira pseudoturna was selected as the principalexperimental organism. This diatom occurs as single cells and has regular dimensions,facilitating counting of cells and cell volume determinations. As well, it is fast growing (up tothree divisions per day) and can be easily maintained in the laboratory. Finally, because thisspecies has been the subject of numerous previous studies, it is well characterized and a largebody of specific information is available. The particular clone considered in these studies, 3H,was originally isolated by Guillard in 1958, from a coastal embayment in Long Island, NewYork (see Guillard and Ryther 1962). Thalassiosira pseudonana has worldwide distribution;there are isolates from locations ranging from along the Atlantic seaboard of the U.S., to23European estuaries, tropical Atlantic reefs and Australian coastal waters (see Nelson and Brand1979). However, it may not be the most ecologically relevant species to use, since it is rarelya dominant member of the plankton (but see Guillard and Ryther 1962, and Gallegos 1992 forrecords of blooms), and is usually confined to enriched waters. There is also evidence that the3H isolate has lost heterozygosity at several loci since isolation (Murphy 1978). However,there is no reason to believe that the fundamental physiology of T. pseudonana is differentfrom that of any other diatom. Its availability, ease of culture and the great deal of previousresearch conducted with this organism, make it a prime candidate for a reference species,much as the white rat serves this purpose for medical research.In Chapter 1, the relationship between growth rate and nucleoside diphosphate kinase(NDPK) activity is examined in T. pseudonana. The goals are: a) to characterize the enzymein this species in terms of optimal assay conditions, and kinetic and thermodynamic constants,and b) to determine whether NDPK activity is related to light limited growth rate in apredictable manner that might allow the enzyme to be used to estimate growth rate.In the following chapters, the enzyme nitrate reductase (NR) is examined as a means toestimate nitrate incorporation rates.In Chapter 2, the goals are: a) to develop and optimize an assay for NR activity in T.pseudonana, b) to characterize the enzyme in terms of kinetic constants, substrate specificityand cofactor requirements in T. pseudonana, as well as another diatom Skeletonema costatum,and a dinoflagellate, Amphidinium carterae, c) to validate the NR assays in these three speciesby determining whether MR activity is sufficient to account for observed nitrate incorporationrates in cultures growing on excess nitrate under light limitation, and d) to determine whetherthe assay developed in T. pseudonana is applicable to a range of other phytoplankton species,using the criterion that NR activity must equal or exceed measured rates of nitrateincorporation.Chapter 3 compares MR activity in T. pseudonana in steady state light-limited andnitrate-limited cultures. The goals of this chapter are to determine: a) if MR activity is related24to nitrate incorporation rates, and b) how cell composition changes under different limitingconditions in order to select a scaling variable for enzyme activity.In Chapter 4 more complex, but ecologically relevant cases are considered, where T.pseudonana is grown on light:dark cycles, or under different light spectra, or where cells arestarved of nitrate, or where cultures are provided with ammonium as a nitrogen source. Thegoals of the chapter are to determine in each case whether the different conditions affect therelationships between NR activity and nitrate incorporation rates seen under steady stateconditions.Finally, in Chapter 5, the NR assay is taken to the field in a preliminary study undercarefully controlled conditions. The goals of this chapter are: a) to determine whether the NRassay developed in Chapter 2 can be applied in the field, b) to determine the characteristics ofNR activity in natural populations in terms of kinetic constants, substrate specificity andcofactor requirements, and c) to compare NR activities to other indices of nitrate incorporationrates including nitrate disappearance from the medium, particulate nitrogen increase, or 15Nuptake. These comparisons are made over diel cycles in irradiance, and in the presence orabsence of ammonium.25CHAPTER 1: RELATIONSHIP BETWEEN NUCLEOSIDE DIPIIOSPHATE KINASEACTIVITY AND LIGHT-LIMITED GROWTH RATE IN THE MARINE DIATOMTHALASSIOSIRA PSEUDONANAINTRODUCTIONSelecting an enzyme to serve as an index of growth rate is not a simple matter. Thereis evidence that many different enzyme activities correlate with growth rate in a variety oforganisms. For example, Pedersen et al. (1978) reported that in the bacterium Escherichiacoli, 102 of 140 proteins (representing 2/3 of the protein mass of the cell) catalogued onchromatography plates showed nearly linear increases with increasing growth rate. In yeastcells, Sebastian et al. (1973) demonstrated a correlation between RNA polymerase I activityand growth rate, while Yao et a!. (1985) found that omithine decarboxylase activity in theciliate Tetrahymena thermophila was also correlated with growth. However, becausephytoplankton growth is often unbalanced (see Eppley 1981), increases in cell number may notbe equal to specific rates of elemental increase (e.g. carbon incorporation), or by the rate ofsynthesis of an individual component (e.g. an amino acid). Thus, an enzyme associated withsynthesis of a particular component may not be suitable as a growth rate index under allcircumstances. A more general index of growth rate is desirable.Hochachka and Somero (1984) divided metabolism in animal cells into three blocks: a)a catabolic block where energy was provided as ATP or NAD(P)H (and presumablycorresponding to photosynthetic reactions in plant cells), b) an anabolic block where ATP andNAD(P)H drive basic biosynthetic reactions and chemical and mechanical work, and c) ablock involving growth and integration. Interestingly, in general, growth and integration donot use ATP directly, but instead use other nucleoside triphosphate (NTP) compounds, e.g.GTP for protein synthesis, CTP for synthesis of certain lipid compounds, and UTP forsynthesis of complex carbohydrates (Lehninger 1982, Hochachka and Somero 1984). Thespecialization of these NTP forms probably aids in proper allocation of ATP among differentmetabolic needs (see Atkinson 1977). Furthermore, NTP compounds are also required for the26DNA and RNA synthesis which must accompany growth (Parks and Agarwal 1973). With theexception of ATP and a small portion of GTP, all nucleoside triphosphates are synthesized bynucleoside diphosphate kinases (E.C., NDPK) (Ingraham and Ginther 1978). NDPKcatalyses the reversible reaction:ATP + NDP -* ADP + NTPwhere NDP and NTP are the high energy di- and triphosphate forms of the nucleosidescytidine, guanosine, uridine, or thymidine. It might be hypothesized that the increasedrequirements for NTP compounds at higher growth rates would necessitate increases in NDPKactivity.Characteristics of NDPKNDPK is found in all cells, and has been measured in a wide variety of organisms(Parks and Agarwal 1973). Characteristics of the enzyme are summarized in Table 1.1. Theenzyme is usually found as a hexamer of about 100 kDa, but Jong and Ma (1991) havereported a tetrameric form in yeast. There also appear to be many isozyme forms of theenzyme with distinct characteristics; however, Gilles et al. (1991) have shown that someNDPK isoforms may be an artifact of enzyme purification procedures. In their study, humanerythrocytes were found to contain only one form of NDPK, a hexamer composed of twodistinct polypeptide chains. When purified under denaturing conditions (using isoelecthcfocusing), these subunits dissociated and could randomly re-associate to produce two or moreapparently different enzymes (Gilles et al. 1991), NDPK is relatively non-specific fordifferent nucleoside di-and triphosphates. K values and reaction rates with differentnucleosides are generally within the same order of magnitude (Ingraham and Ginther 1978).The NDPK reaction described above has an equilibrium constant near 1.0; values rangefrom 0.6 in Bacillus subtilis (Sedmaic and Ramaley 1971) to 1.28 in yeast (Parks and Agarwal1973), depending on assay conditions.Table1.1.Characteristicsofnucleosidediphosphatekinase(NDPK)fromvarioussources(ISOZYME=isoelectricpointof theisozymewhereavailable,otherwisetheauthorsdescription;MW=molecular weight,*indicatestheweight of amonomer;AE=apparentactivationenthalpybelow/abovethetransitiontemperature;--=notprovidedbytheauthors).SOURCEISOZYMEMWSUBUNITSKmATPKmTDPAEREFERENCE(kDa)(mM)(mM)(kJ mor1)Bacillussubtilis8.4100--0.15--1.15/2.46SedmakandRamaley1971brewersyeast8.0102—0.31——ParksandAgarwal1973Saccharomycescervisiae704—0.17—JongandMa1991ratliverm6.0----1.660.16--KimuraandShimada1988“c6.018*61.330.19--beefbrain8.6120--0.230.26--Robinsonetal.1981humanerythrocytes5.480--0.20.11--ParksandAgarwal1973“5.893--1.00.055--“6.384--3.00.22--ParksandAgarwal1973“6.880--0.250.20--“7.3100—1.080.301.00/2.03“8.3103—0.170.12—ft--17*6—----Gillesetal.1991EhrlichAscitestumorcells--76——---.Koyamaetal.1984SpinaceaoleraceaI16*62.0--—Nomuraetal.1991II18*60.89----ScenedesmusobliquusI-100--------KleinandFollmann1988“II-100--------28NDPK and Growth RateThe fact that NDPK is a near-equilibrium enzyme (i.e. the reaction is freely reversible)indicates that it is unlikely that the enzyme is substrate saturated in vivo and thus rate-limiting(see Newsholme and Crabtree 1986). This suggest that the maximal activity of NDPK,measured in vitro with saturating substrate cannot be quantitatively related to an in vivo rate.Nevertheless, a correlation between maximal NDPK activity and growth rate might still bepossible. Brown (1991) has argued that there are adaptive pressures on cells to minimize theirprotein content (since protein is usually near the solubility limit within the cell). Thus, if agiven enzyme was not rate-limiting and it was in greater concentration than necessary, therewould be an advantage to reducing its concentration. As a result, even non-rate limitingenzymes should respond as the fluxes through metabolic pathways change.Although NDPK has been measured in a wide range of organisms from bacteria tohigher plants to mammalian cells (Parks and Agarwal 1973), there is only one report of ameasurement in a unicellular autotroph (Klein and Follmann 1988), and no cases ofmeasurements in marine phytoplankton. There has been speculation about the importance ofNDPK in cell growth processes during development (Dickinson and Davies 1971), including acorrelation with growth rate in mammalian tumor cells (Koyama et al. 1984), and evidence ofrelationships between NDPK and growth rate in crustaceans (Berges 1989, Berges et a!. 1990).In multicellular organisms, however, the relationship between NDPK and growth iscomplicated by changes in body size and composition during development. Such relationshipsmay be clearer in a unicellular organism.The objectives of this chapter are to examine: a) the general characteristics of NDPKin Thalassiosira pseudonana, b) the relationship between maximal NDPK activity (whichshould be proportional to enzyme concentration) and growth rate under light (energy)limitation, and c) the relationship between various cell components and growth rate in order todetermine to which biomass parameter NDPK activity is best scaled.29MATERIALS AND METHODSCulture conditionsThe marine diatom Thalassiosira pseudonana (Hustedt) Hasle and Heimdal (3H clone)was obtained from the Northeast Pacific Culture Collection, Department of Oceanography,University of British Columbia. Cultures were grown in semi-continuous batch culture inenriched artificial seawater (ESAW) based on the recipe by Harrison et a!. (1980), withsodium glycerophosphate replaced with an equimolar concentration of sodium phosphate,ferrous ammonium sulphate with an equimolar concentration of ferric chloride and additions ofselenite, nickel and molybdate to achieve 1 nM final concentration. Temperature wasmaintained at 17.5 ± 0.5° C using a circulating water bath. Cultures were grown in 1 L glassflasks, stirred at 60 rpm with Teflon-coated stir bars and bubbled with air filtered through a0.22 m membrane filter. Continuous illumination was provided by VitaliteTM fluorescenttubes and attenuated by distance or neutral density screening to give a range of irradiancesfrom 6 to 120 mol quanta m2 s measured in air inside empty culture vessels using a LiCormodel 185 meter. During the course of experiments, cultures were never dense enough toreduce average irradiance by more than 10%. Growth rates were followed by in vivofluorescence, measured twice daily using a Turner DesignsTM Model 10 fluorometer and cellcounts using a Coulter CounterTM model TAIl equipped with a population accessory. Allsampling was conducted in early to mid logarithmic growth phase.Cell compositionIn all experiments, cell carbon and nitrogen quotas were determined by filteringsamples onto pre-combusted 13 mm Gelman type AE glass fiber filters and analyzing themusing a Carlo Erba CNS analyzer. Samples for protein determination were collected on precombusted Whatman GF/F filters. Homogenates were prepared as described by Dortch et al.(1984). They were ground with 3% trichloroacetic acid (TCA) and solublized in 1 N NaOH.30Protein was determined by the method of Bradford (1976) using the micro-assay procedure ofthe Bio-Rad Protein Assay kit (Bio-Rad Laboratories, 500-0001) with bovine serum albumin(BSA, Sigma Chemical Co. A 7638) as a standard. Cell volumes were calculated fromCoulter Counter measurements and calibrated using 5 m latex microspheres, followingThompson et al. (1991).Cell homogenization and enzyme assaySamples were collected on 25 mm Whatman GF/F glass fibre filters using filtrationpressures less than 100 mm Hg. Filters were immediately placed in 1 mL of ice coldextraction buffer consisting of 50 mM imidazole, pH 7.4, 2 mM dithiothreitol, 2 mM EDTA,1% (w/v) BSA and 0.1% (v/v) Triton X-100. Cells and filters were ground in a 5 mL glass-Teflon tissue homogenizer for 2 mm. Homogenates were centrifuged in a Sorval RCB-2Bcentrifuge at 4°C for 5 mm at 750 g and used immediately in assays. Preliminary experimentsshowed that no NDPK activity remained in the pellet. This might have been anticipated sincethe enzyme from higher plants has been shown to be predominantly in the cytosol (Dancer etal. 1990).Assay conditions were adapted from Berges et al. (1990). Assays were conducted in 1mL volumes in disposable plastic cuvettes. All assay components were obtained from SigmaChemical Co. and were the purest grade available. ADP produced in the NDPK reaction wascoupled to NADH oxidation through pyruvate kinase (PK) and lactate dehydrogenase (LDH)(Agarwal et a!. 1978). Substrate concentrations were optimized by increasing theconcentration of each reaction component until no further increase in NDPK activity wasobserved. This was routinely verified over the course of the experiments in cultures growingat low and high irradiances. Further increases in substrate concentrations were avoided, sincethey also increased rates of side reaction and thus decreased precision. Final concentrations inthe assay were 50 mM imidazole buffer (pH 7.4), 0.2 mM NADH, 20 mM MgCl , 70 mMKC1, 1.1 mM phosphoenol pyruvate, 2.0 mM ATP, 0.7 mM TDP, 10 U lactatedehydrogenase (Sigma L 2500) and 1 U pyruvate kinase (Sigma P-1506). Reactions were31started by adding TDP. Controls were run without homogenate and without TDP, and rateswere corrected accordingly (Agarwal et a!. 1978). Reactions were followed by monitoring thedecrease in absorbance at 340 nm due to NADH oxidation using a LKB Ultrospec II UVspectrophotometer with a six position water-cooled turret interfaced to an IBM personalcomputer (see Appendix B). Typically, it was necessary to monitor reactions for 5 to 10 mmto establish the initial, linear rate of reaction. Temperature was maintained at 17.5 ± 0.1 °C(the growth temperature of the cultures) using a Lauda RM6 water circulating bath. NDPKactivity was expressed in units (U), where 1 U represents the quantity of enzyme catalyzing theconversion of 1 mol of substrate to product per minute, using a millimolar extinctioncoefficient of 6.22.Enzyme characterizationAssays were conducted over a range of ATP and TDP concentrations to determine Kmvalues for the algal enzyme. Data for NDPK activity versus substrate concentration werefitted to a Michaelis-Menten model using a non-linear fitting routine (NONLIN, Wilkinson1990; see also Appendix C). Assays were also conducted over a range of temperatures from10 to 25°C. An Arrhenius transformation was used to calculate an apparent activationenthalpy (iH) of the enzyme (Hochachka and Somero 1984).Steady state experimentsOn six separate occasions, four to six semi-continuous batch cultures were grown atdifferent irradiances ranging from 6 to 120 jmol quanta m2 s. Cultures were acclimatedfor a minimum of 10 generations except in cultures growing at < 0.4, where 6 to 8generations were allowed. Cultures were sampled for cell volume, cell nitrogen, carbon andprotein quotas (i.e. cell contents), and NDPK activity. These parameters were plotted againstgrowth rate and examined using linear correlation analyses (Wilkinson 1990).32Transition ExperimentsOn two occasions, six T. pseudonana cultures were acclimated in the same manner asin steady state experiments, three to 15 j1mol quanta m2 s4 and three to 135 mol quantam2 Samples identical to those in the steady state experiment were taken; then thecultures were transposed. In the first experiment, cultures were sampled at 24 h intervals for72 h after transposition. In the second experiment, sampling continued for 210 h after thetransition. Changes over time in cell carbon, nitrogen and protein quotas, cell volume, growthrate, and NDPK activity were examined.RESULTSEnzyme characterizationKm values for the substrates, calculated from six separate homogenates, were 0.24 ±0.01 mM for TDP and 0.86 ± 0.06 mM for ATP (Fig. 1.1).The slope of the regression lines of Arrhenius plots of log NDPK activity versus theinverse of temperature gave an apparent activation enthalpy of 0.841 ± 0.026 kJ mol1 (Fig.1.2).Steady state experimentsSteady state growth rates versus irradiance data were collected over a period of 18months and demonstrate the constancy of the growth rate-irradiance relationship over theexperimental period (Fig. 1.3), Fitting a Michaelis-Menten type curve to the data gave aof 1.64 d’ and a half-saturation constant (Kj) of 23 mol quanta m2 From Fig.1.3,‘k (the onset of light saturation as defined in Parsons et al. 1984b) was estimated to beapproximately 40 mol quanta m2 s1.Carbon, nitrogen and protein cell quotas (pg cell-i) were not significantly correlatedwith growth rate (P > 0.3, P > 0.06, and P > 0.5, respectively) (Fig. 1.4). However, there33.z0.5 1.0 1.5TDP (mM)0.0100.0080.0060.0040.0020.0000.00.0250.020:::0.005z0.0000 1 2 3 4 5ATP (mM)Fiure 1.1. Nucleoside diphosphate kinase (NDPK) activity versus substrateconcentration for A) thymidine 5’-diphosphate (TDP) and B) adenosine5’-triphosphate (ATP) in homogenates of Thalassiosira pseudonana.Curves are fit to rectangular hyperbolae. Km values are 0.24 mM forTDP and 0.86 mM for ATP.34-1.6-1.7-1.8z -.C-2.00.0033 0.00361/T(K4)Figure 1.2. Arrhenius plot of NDPK from Thalassiosira pseudonana.The solid line represents a least squares regression fit to the data.Apparent activation enthalpy is 0.84 1 kJ mof1.0.0034 0.0035352.00 •. -2 1150irradiance (tmo1 quanta m s)Figure 1.3. Growth rate versus irradiance curve for Thalassiosira pseudonana.Curve is fit to a rectangular hyperbola.max=1.64 d1 and K1 =23 mo1 quanta m2 s1. Each point represents a single culture.Error bars represent the standard error of the mean of 3 to 6 growthrate measurements, or if not seen are smaller than the size of the symbol.50 10036I I14- • A12-o . •• • -10- •. •.8-C.) I4- . BBCd3- B • -BB.2- . B B -bSJ B• IBjul’11 -o 1-.E3 50- y,,’4’’ CE ‘v v40 yV “ V’V -0 V— V V- 30- vv -o VI I I6- DC.)5-. 4- -03- -I I I I0.0 0.5 1.0 1.5growth rate (d1)Figure 1.4. Cell composition versus light-limited specific growth rate inThalassiosira pseudonana. A) Cell carbon quota, B) cell nitrogen quota,- C) cell volume, and D) cell protein quota. Each point represents themean of duplicate determinations from a single culture.37was a significant positive linear relationship between cell volume and growth rate (P < 0.01).If carbon, nitrogen and protein were expressed per unit cell volume (i.e. pg jm3), there weresignificant negative relationships with growth rates for carbon (P < 0.05) and nitrogen (P <0.05), but not protein (P > 0.09). In addition, there were no significant correlations betweenNDPK activity (on a per cell basis) and carbon cell quota, nitrogen cell quota, protein cellquota or cell volume (P > 0.2 in all cases; data not shown).The relationships between NDPK activity on a per cell basis and either specific growthrate or growth rate in terms of carbon (the product of specific growth rate and carbon cellquota, which is analogous to a 14C measurement) were highly variable (Fig. 1.5). However,NT)PK activity was significantly and positively correlated with growth rate. When a linearmodel was used, NDPK activity per cell was significantly correlated with specific growth rate(P < 0.05) and carbon growth rate (P < 0.04). NDPK activity at low growth rates ( <0.4) appeared to increase. Using a quadratic model, the correlation improved (P < 0.01 forboth cases). Expressing NDPK activity per unit cell volume, carbon, or protein did notchange the pattern of the relationship, although the variability increased significantly (Fig. 1.6A, B, D). When NDPK activity was expressed per unit nitrogen, however, the quadratic termin the NDPK-growth rate relationship was no longer significant (P > 0.07), indicating that therelationship was more linear.Transition ExperimentsTransition experiments provided another way to assess whether growth rate and NDPKactivity were related. By measuring NDPK activity in individual cultures before and after atransfer from low to high light or vice versa, changes in enzyme activity could be followed andan approximate time for changes to occur determined. Transition experiments were repeatedtwice. In the first case, the time course was followed for only 72 h, in the second case for 210h. Results were nearly identical in both experiments; for clarity only the results of the 210 htime course are presented. Under steady state conditions, the growth rate of high light cultureswas 1.45 d’ while low light cultures grew at 0.56 d (Fig. 1.7). Cultures were switched at3812 I10 -OM5i1.5 2.0growth rate (d’)>0 12 I10z8-6-4-2 B0 I I0 5 10 15 20growth rate (pg C d)Figure 1.5. NDPK activity versus A) light-limited specific growth rate, and B)growth rate in terms of carbon in Thalassiosira pseudonana. Eachdata point represents a single culture. Error bars show the standarderror of the mean of two enzyme assays or a minimum of three growthrate determinations.390’I I IU- 1.0-•I I0.8-____‘— 0.6-•?_,.+I. I A> 0.4-_0.2-o— II____z4----3-•12 - — B1-C— I I II I I0.20-—z I Z:_-_0.15-V. 0.10- V‘0.05- TI I I2.5-_o 2.0-1.5-•i.o0.5- +I I0.0 0.5 1.0 1.5 2.0growth rate (d’)zFigure 1.6. NDPK activity versus specific growth rate in Thalassiosira pseudonana.Activity is expressed per unit carbon (A), nitrogen (B), cell volume (C)or protein (D). Error bars show the standard error of the mean of twoenzyme assays or a minimum of three growth rate determinations.40a)C)01)4-0C0= C.)0.440a)E30a)__C) 20a)C) 2.001)1.54-1.00.50 50 250I time (h)Figure 1.7. Cell composition versus time in terms of A) cell carbon quota, B) cell nitrogenquota, C) cell volume, and D) cell protein quota in Thalassiosira pseudonana.( 0 ) Cultures grown under high light (135 jmol quanta m2 s’) and moved to lowlight (15 mol quanta m2 s’) at t = 32 h (marked by the arrow). ( • ) Culturesgrown under low light and switched to high light at t = 32. Error bars representstandard errors of the mean of three replicate cultures. Statistically significantdifferences (P < 0.05) are indicated by asterisks (*).87654100 150 20041t = 32 h, and growth rates had changed by 48 h. Composition between treatments wascompared using a repeated measures ANOVA followed by LSD comparisons at the 95% level(Steel and Tome 1980, Wilkinson 1990). Results were similar to the steady state experimentsin that there were no differences between high and low light cells for carbon quota or nitrogenquota. Protein quota differed only in one case. Cell volume was significantly higher in highlight grown cells before the transition, and by 150 h cell volumes in transition cultures wereeffectively the same as those found under the corresponding low or high light steady state.Steady state NDPK activities were significantly higher in high light cultures when expressedper cell, or per unit carbon, or nitrogen, but not significantly different when expressed per unitprotein (Fig. 1.8). Following the transition, although the NDPK activity dropped significantlyfor high light to low light transition, the treatments were not significantly different at the endof the experiment.DISCUSSIONEnzyme characterizationAlthough NDPK is found in a broad range of organisms, its characteristics are verysimilar (Parks and Agarwal 1973). Differences between characteristics of the crude enzymehomogenate from Thalassiosira pseudonana and those published in the literature might havebeen anticipated, since the majority of work has been done on the purified enzyme. Further,the algal NDPK may be a mixture of isoforms. For example, Nomura et a!. (1991) describedtwo NDPK isozymes from spinach (Spinacea oleracea). Despite these considerations, thekinetic and thermodynamic constants appear almost identical to those published (see Table1.1). A Km for TDP of 0.24 mM was obtained for T. pseudonana, which is close to thevalues found in yeast (Jong and Ma 1991) and in human erythrocytes (Agarwal et at. 1978).For ATP, a Km of 0.86 mM was calculated in the present study, which is very near that foundby Nomura et at. (1991) in spinach leaves and within the range found in human erythrocyteisoforms (Agarwal et a!. 1978). Arrhenius plots of the enzymes42— 76z4‘—S0% 3C‘-41.5zbf• 0.9. ‘—S0% 0.6-12z‘-S0%o‘-40.30- z0.20: 015150 200 250time (h)zFigure 1.8. NDPK activity scaled to A) cell number, B) cell carbon quota, C) cellnitrogen quota, D) cell volume, and E) cell protein quota, versus time fortransition experiments with Thalassiosira pseudonana. Symbols are the sameas in Figure 1.7.0 50 10043often display biphasic behavior, although this is dependent on the particular isoform (Agarwalet al. 1978). It is possible that this was obscured in the present study by a mixture of isoformsor because of the relatively low number of temperatures assayed. The p1 7.3 isoform ofNDPK from human erythrocytes shows two phases with a break at 31°C (Agarwal and Parks1971), while for NDPK from Bacillus subtilis a transition occurs at 25°C (Sedmak andRamaley 1971). Both of these transitions occur at or above the highest temperature tested inthe present study, but the activation enthalpy for T. pseudonana (0.841 U mol 1) is close tothe values determined in other species, below their transition points (see Table 1.1). In allcases, the optimal substrate concentrations and assay conditions are remarkably similar;conditions determined for crustacean tissue (Berges et al, 1990) proved optimal for NDPKfrom Thalassiosira pseudonana.Cell compositionNo consistent relationships were found between light-limited growth rate and carbon,nitrogen or protein quota. Cell volume, however, was positively correlated with growth rate.Although there were significant relationships between carbon, and nitrogen per cell volume,these are probably caused by the significant change in volume alone. This illustrates apotential pitfall in using such ratios (see also Packard and Boardman 1988). Raven (1981)points out that variation in cell volume may be an important metabolic adaptation; in order tomaintain the proper cellular concentration of metabolites and catalysts it may be necessary tochange cell volume. Thompson et al. (1991) provide a detailed review of carbon and volumerelationships with growth rate and show that there is a general, positive relationship for avariety of species. For Thalassiosira pseudonana, in particular, the volume-growth raterelationships in the present study agree well with theirs, but in contrast, Thompson et al.(1991) found a strong positive relationship between carbon quota and growth rate which wasnot observed in the present study. The reason for these differences is unclear, although theexperiments in the present study were conducted over a much shorter period of time than thoseof Thompson et a!. (1991), during which large differences in cell volume and carbon quotas44were seen. Such variability may result, in part, from size changes related to the sexual cycleof diatom species, although we observed no evidence of sexual reproduction in any of ourcultures. Various authors have also demonstrated size and carbon quotas which decrease withgrowth rate (Thompson et a!. 1991).There is evidence in the present study of increases in cell volume, and cell quotas ofcarbon, nitrogen and protein at very low growth rates ( < 0.25). This represents novelinformation since there is very little data in the literature for such low growth rates.Thompson et a!. (1991), for example, had only one culture in this range of growth rates, andalthough Sakshaug and Andresen (1986) reported increases in cell carbon and nitrogen quotasat low irradiance in cultures of Skeletonema costatum, this effect was only prominent whencells were grown on light-dark cycles with short day lengths. Increased NDPK activity at lowgrowth rate was also found in the present study. This is not due solely to increases in cellsize, since the pattern persists even if data are scaled to cell volume or carbon, or protein cellquotas (Figure 1.6 A, B, D), but it may be related to cell nitrogen quota.Composition data for the transition experiments agree well with the steady state data;only cell volume changed consistently throughout the transitions. This volume change is inagreement with data presented by Thompson et al. (1991), although these authors also foundsignificant changes in carbon quota. In a similar light transition study, Post et a!. (1985) notedthat in turbidostat-grown cultures of Thaltssiosira weisfiogli, changes in carbohydrateoccurred during light transitions, but no significant changes in protein were found. Similarly,Claustre and Gostan (1987) found changes in volume but not in protein during transitions withIsochrysis and Hymenomonas species.A discussion of the causes and meaning of these cell composition differences is beyondthe scope of this chapter, but they will be discussed further in Chapter 3. However, suchchanges have important implications for selecting a biomass variable on which to scale enzymeactivity. Since there was no indication that NDPK activity was correlated with any index ofcell composition measured, the usefulness of normalizing to facilitate comparisons within thisspecies is questionable. Because NDPK activity per cell varied over a factor of 6 or 7, while45carbon or cell volume only varied by a factor of two, scaling enzyme activity to either carbonor volume does not substantially change the relationship between NDPK and growth rate, andmay in fact add variation to the measurement. In addition, if the relationship between enzymeactivity and carbon, cell volume or nitrogen is complex (e.g. curvilinear), the scaled enzymeactivity becomes much more difficult to interpret (Packard and Boardman, 1988). NT)PKactivity per unit nitrogen appeared more linear and consistent with a monotonic increase inNDPK activity with growth rate. However, the relationship was highly variable; less than25% of the variation in NDPK activity could be attributed to growth rate. Protein iscommonly selected as a scaling variable in enzyme studies, but diatom species have manypotentially interfering compounds such as amino acids, which complicate such a measurement(Dortch et al. 1984; Appendix A). Cell volume may also be unsuitable as a scaling factorbecause of methodological biases. Thompson et al. (1991) speculated that short-term diatomvolume increases (such as those found in this study) are achieved by addition of intercalarybands, added between the valves of the diatom frustule, which would elongate the cell withoutan increase in width. This implies a change in cell geometry which will result in an error ifthe volume is measured by a particle counter such as a Coulter CounterTM (Kubitscheck 1987;Montagnes et a!. submitted). For the present, when a single species is considered, expressingactivity per cell and providing data on cell composition seems to be the most reasonablecourse. In cases where interspecific comparisons must be made the issue is clearly morecomplex and cell nitrogen may offer some promise.NDPK and growth rateAlthough statistically significant, the relationship between NDPK activity and growthrate is relatively poor and therefore of limited use in a predictive sense. Part of the reason forthis variability may be that NDPK has other functions in the cell in addition to NTPinterconversions. NT)PK has been shown to be involved in cell signal transduction andregulatory processes. NDPK may directly interact with membrane G proteins and may beinvolved in the regulation of adenylate cyclase (Jong and Ma 1991). As well, NDPK has been46implicated in activating guanine nucleotide binding proteins (Jong and Ma 1991; Nomura et a!.1991). The importance of these processes in a unicellular organism is unclear. Because theseprocesses are generally associated with the cell membrane, they may involve membrane boundforms of NDPK. Since the cytosolic forms of the enzyme appear to be more abundant, therole of NDPK in NTP interconversions may predominate.Alternatively, it is likely that because NDPK is a near-equilibrium enzyme, it is notoperating at Vmax in vivo. This can be supported by a simple calculation of the maximumNTP requirement of a growing cell. Consider a T. pseuttonana cell of 10 pg C, that isdoubling once a day ( = 0.69 d1). Based on data from the literature, such a cell would beexpected to have approximately 6 pg protein, 7.5 pg carbohydrate, 3.5 pg lipid, and 0.3 pgDNA and RNA (Darley, 1977, Dortch et at. 1984, Harrison et at. 1990b, Laws 1991, and thepresent study). In terms of protein requirements, it is assumed that 2 GTP per amino acidincorporated into protein are required (Morris 1974, Lehninger 1982; note that the requirementis actually slightly higher, but that a small amount of GTP is also produced by the succinylCoA synthase reaction, and via phosphoenol pyruvate carboxykinase). For carbohydraterequirements it is assumed that all cell carbohydrates exist as chrysolaminarin, or otherglucose-based polymers that require one UTP per monomer (Craigie 1974, Darley 1977; notethat this is an overestimate since only 60% of carbohydrate in diatoms exists aschrysolaminarin). It is also assumed that all cell lipid is found in membranes and is in theform of phospholipid, glycolipid, or other lipid forms that require one CTP per molecule fortheir synthesis (Lehninger 1982, Andrews and Ohlrogge 1990; note that these lipid forms mayaccount for less than 50% of the total). Finally, it is assumed that the DNA and RNA contain3 times as much of other nucleosides as ATP (according to Darley 1977, G+C residuesaccount for 37-58% of the total). Given these assumptions, the total NTP requirement of thecell could not exceed approximately 2 x 10-10 mol NTP miir1 cell 1. In the present study,T. pseudonana cells growing at this rate had NDPK activity in the range of 20-50 x 10-10mol NTP min1cell,or at least an order of magnitude higher than the maximumrequirement.47If NDPK is not operating at Vm, then enzyme activity in vivo may be regulated byfactors including substrate concentration, or control mechanisms such as phosphorylation oradenylate energy charge. If NDPK was substrate-limited, i.e. reaction rates were a function ofsubstrate and not enzyme concentration, it might be expected that ATP and NTPconcentrations within the cell would fluctuate as a function of growth rate. For ATP this doesnot always appear to be true; a review by Karl (1980) showed that ATP content per cell isrelatively constant over a wide range of growth conditions for prokaryotes, autotrophs andheterotrophs. While some studies have demonstrated a correlation between growth rate andATP pools, this depends on whether ATP is scaled to cell number or carbon quota, and thereis still controversy (Chapman and Atkinson 1977; Karl 1980, Sakshaug and Andresen 1986).The relationship may also depend on what is limiting growth. Karl (1980) cites data showingthat in the diatom Thalassiosira weissflogii ATP per cell correlates with growth rate undernitrate or phosphate limitation but not when cells are limited by light or ammonium. Laws etal. (1983) showed that for the diatom Thalassiosira weisflogii, the ratio of ATP to carbon wasconstant over a wide range of light- and nutrient-limited growth rates. They speculated thatATP turnover, as opposed to concentration, might be a critical factor. In bacterial systems,the concentration of ATP and other nucleotides are at best a weak function of growth rate(Marr 1991). Karl (1980) suggests that adenine nucleotides are at or near saturating levels formost respiratory and metabolic enzymes. For other nucleotides there is also disagreement.Chapman and Atkinson (1977) found that other nucleotides followed patterns of ATP and didnot vary with growth rate. Interestingly, they attributed this to rapid equilibration of othernucleotide pools and ATP through NDPK. Data presented by Marr (1991) support this view.Alternatively, Karl (1980) demonstrated that certain NTP pools, particularly GTP, vary withbiosynthesis and growth rate. He suggested that the ratio of GTP/ATP might be useful as anindex of growth rate. Pall (1985) also assigned GTP a key regulatory role in anabolicprocesses within the cell. At another level of enzyme regulation, phosphorylation control ofNDPK has been suggested (Pall 1985) but has not been demonstrated,48There is strong evidence that the adenylate energy charge (AEC, defined as the ratio ofthe concentrations of ATP and one-half the concentration of ADP to the total concentration ofATP plus ADP plus AMP) plays a role in controlling NDPK activity. In general, the energycharge varies between 0.7-0.9 in healthy cells (Atkinson 1977, Plaxton 1990). Thompson andAtkinson (1971) have shown that for bovine liver NDPK, activity of the enzyme is maximalwhen the energy charge is near 1.0 and rapidly drops off as the ratio falls. Laws et al. (1983)demonstrated a significant positive correlation between energy charge and growth rate inThalassiosira weisflogii cultures under a variety of limitations. This may explain the pattern inactivity with growth rate observed in the present study. At moderate growth rates (betweenabout 0.5 to 1.0 d) there is little change in the in vitro activity of the enzyme. Over thisrange either substrate concentration or energy charge may be regulating activity. However,Dolezal and Kapralek (1976) showed that in a bacterium grown in a chemostat between 7 to60% of maximal growth rate, there was little change in adenylate levels and no change inenergy charge. A similar response was seen in the diatom Skeletonema costatum where cellcontent of ATP increased only when growth rates were 50% of imax or greater (Sakshaug1977). These two studies may not be directly comparable with Laws et al. (1983) or thepresent study, since they used chemostats and therefore the cells were nutrient rather than lightlimited. At higher growth rates, energy charge may be near its maximum, so that furtherincreases in growth rate may necessitate increases in NDPK concentration. The reason for anapparent increase in nitrogen quota and NDPK activity per cell at very low growth ratesremains unclear.Another possible source of variability in NDPK activity is the stage of cell division.Berges (1989) found strong relationships between NDPK activity and growth rate in the brineshrimp Artemiafranciscana, but such relationships were specific to different developmentalstages. However, Klein and Follmann (1988) showed that for the green alga Scenedesmusobliquus, NDPK activity was constant throughout the cell division cycle.It is apparent that neither NDPK activity nor nucleotide concentrations are entirelysatisfactory as indices of in situ growth rate. If, however, the measurements were combined,49it is possible that their predictive value would improve, particularly if adenylate energy charge,ATP turnover rates, or substrate concentration regulate NDPK activity over a range of growthrates. Furthermore, measurement of nucleotides and NDPK activity could provide insight intothe specific growth rate limitation that cells experience in situ. The present study has examinedonly light-limited growth rates. Since data presented by Karl (1980) suggest that light,phosphorus, nitrate or ammonium limitation result in different ATP-growth rate relationships,examining NDPK activity with respect to these cases would also be interesting.In summary, NDPK in the diatom Thalassiosira pseudonana appears to be relativelysimilar to other NDPK enzymes previously investigated. Maximal NDPK activity is a poorindex of cell growth rate, although the two variables are significantly correlated. Finally,because cell composition varies with growth rate, and because of difficulties in measuring cellvolume or cell protein in phytoplankton species, scaling enzyme activity to different biomassvariables is problematic. In culture, NDPK activity per cell volume appears to be a usefulexpression, but expressing activity per unit nitrogen might also be suitable.50CHAPTER 2: OPTIMIZATION AN]) VALIDATION OF AN ASSAY FOR MTRATEREDUCTASE ACTIVITY IN MARINE PHYTOPLANKTONINTRODUCTiONAs detailed in the introduction, several researchers have noted that NR does not appearto be satisfactory as an index of nitrogen uptake or incorporation rates (see e.g. Eppley et a!.1969, Packard et al. 1971, Collos and Slawyk 1976, Collos and Slawyk 1977, Dortch et a!.1979, Blasco et a!. 1984; but also note the good agreement found by Morris and Syrett 1965,and Hersey and Swift 1976). There are essentially three possible explanations for thesediscrepancies: a) the extractions and assays of NR are inadequate, b) there is no relationshipbetween NR and nitrate incorporation rate, or c) the presence of regulatory mechanisms meanthat the measured maximal activity of NR is not a good indicator of the actual rate of nitratereduction in vivo. In this chapter and the following two chapters, each of these possibilitieswill be examined.In this chapter, an extraction and assay procedure for nitrate reductase (nitrate:nitriteNADH oxioreductase, E.C., NR) is optimized and validated using marinephytoplankton. At this point in the thesis, the specific roles of and place of nitrate reductase inmarine phytoplankton will be only generally outlined; Chapter 3 will address these issues ingreater detail. Similarly, aspects of the regulation of NR will be considered only as theypertain to assay methods; Chapter 4 will deal with these regulatory mechanisms and theirimplications in greater detail.The place of NR in algal nitrate metabolismThe general nitrogen metabolism of microalgae has been considered in the Introduction.Comprehensive reviews of these processes are provided in Morris (1974), McCarthy (1980),Collos and Slawyk (1980), Syrett (1981), Wheeler (1983), Falkowski (1983), and Syrett(1989). In this chapter, only nitrate metabolism will be considered. In addition to the reviewscited above, specific reviews of nitrate metabolism are available for higher plants (Hewitt et51a!. 1976, Guerrero et a!. 1981, Fernandez and Cardenas 1989, Redinbaugh and Campbell1991).The terminology surrounding the uptake, reduction and subsequent incorporation ofnitrate into cellular constituents is confusing, because different authors have chosen differentterms. In this thesis, the following terms will be used to describe the different processeswithin the cell (after Wheeler 1983). Uptake will be used to describe the removal of nitratefrom the medium, whether judged by disappearance from the medium, or appearance withinthe cells. Note that for higher plants, the presence of intercellular spaces, particularly in roottissue make this more difficult to define (see Redinbaugh and Campbell 1991). Assimilationwill be used to describe the conversion of nitrate to nitrite to ammonium to small organicnitrogen components, such as amino acids and small, soluble peptides. incorporation will bereserved for the process in which small organic components are synthesized intomacromolecules, such as proteins and DNA. Functionally, it is difficult to distinguishassimilation from incorporation. For the purposes of this study, nitrogen will be considered tohave been incorporated when it is retained in filtered samples and detectable by carbon-nitrogen analyzers. Note that the rupture of cells during this process would result in anunderestimate of incorporation, while including inorganic nitrogen contained in the vacuoles offiltered cells might result in an overestimate of this process.Nitrate metabolism in eukaryotes begins with the uptake of nitrate into the cell. Thereis relatively little information about this process. Based on electrochemical andthermodynamic considerations, nitrate transport must be an active process (Pilbeam and Kirby1990). There is thought to be a specific nitrate transport protein (also referred to as apermease). Such a protein has been isolated in cyanobacteria (Omata 1991, Lara et a!. 1993),but is poorly characterized in eukaryotes (see Redinbaugh and Campbell 1991, Miyagi et a!.1992). In higher plants, there is evidence for a 2H+: 1N03 symport (Deane-Drummond1990, Collos et a!. 1992c), but a N03:OH- antiport has also been proposed (DeaneDrummond 1990, Lara et a!. 1993). For marine phytoplankton, Falkowski (1975) showedthat there was an ATP requirement for nitrate transport in a marine diatom. There appears to52be a strong dependence of nitrate uptake on sodium in marine species (Syrett 1989). Fororganisms living in an alkaline environment (seawater pH is near 8.0). it has been suggestedthat maintaining gradients of Na+ instead of H+ may require less energy (Lara et al. 1993).Siddiqi et al. (1990) described a two-phase system in barley roots, where there was a high-affinity inducible system operating at low nitrate concentration, and a constitutive transportsystem at high concentration, and there is also evidence of such a system in marine diatoms(Collos et a!. 1992c). A direct role for NR in the uptake of nitrate in higher plants wassuggested, based on membrane associations of NR and the close link between uptake andreduction of nitrate (Butz and Jackson 1977). However, Warner and Huffaker (1989) havedemonstrated that the induction of transport and the uptake kinetics provided no evidence for arole of NR in nitrate uptake. Nonetheless, in certain fungi, NR activity and nitrate uptake arehighly coordinated (Goldsmith et a!. 1973), and evidence from Tischner et al. (1989) showingthat antibodies to NR protein inhibit nitrate uptake in the green alga Chiorella sorokinianasuggest that if NR is not responsible for uptake, the two processes are closely linked.The reduction of nitrate to ammonium proceeds in two steps; a two electron donationcatalyzed by MR (note that earlier ideas about alternate pathways for nitrate reduction, e.g.Dortch et a!. 1979, Clayton 1986, have been largely discredited), followed by a six electrondonation by nitrite reductase (E.C., NiR). Several reviews suggest that NR is the rate-limiting process in nitrate incorporation (e.g. Beevers and Hageman 1980, Campbell 1988,Wray and Fido 1990), but other authors disagree. Noting that in certain species internalnitrate pools do not accumulate, Tischner (1990) argues that nitrate uptake is in fact thelimiting step (but note that Fuggi (1989) presents a mechanism whereby leakage of nitratewould allow NR to be limiting without a build-up of nitrate within cells). However, evenamong those who maintain that uptake is rate-limiting, there is at least recognition that MR is akey point of control of the process (see De la Rosa et al. 1989). Other authors have noted abuild-up or efflux of nitrite from cells under conditions of senescence, low CO2 (Azura andAparicio 1983), or light-dark transitions (Stulen and Lanting 1976, but see also Sanchez andHeldt 1990) and suggest that NiR may be rate-limiting, especially in the dark when the53physiological source of reductant for NiR (ferredoxin) cannot be produced. Generally,however, NiR activity exceeds NR activity by up to a factor of 8, an indication that NiR is notlimiting (Eppley et al. 1969, Aslam and Huffaker 1989; but note that Kessler and Czygan(1968) found similar levels of the two enzymes in green algae). There is no doubt, however,that NiR is also highly regulated in the cell. Evidence from analyses of mRNA and NiRprotein suggest that induction of NR and NiR are nearly simultaneous (Galvan et al. 1992). Infact, nitrate appears to induce NiR as effectively as it does NR (Galvan et al. 1992). Becausenitrite is toxic within the cell, it makes sense that the two enzymes should be closely coupled,and that NiR activity should exceed the activity of NR.Following these reduction reactions, the ammonium produced may be incorporated intoamino acids in one of two processes: into glutamate via glutamate dehydrogenase (GDH, E.C., or into glutamine by the enzyme glutamine synthetase (GS, E.C. and then toglutamate by the enzyme glutamate synthase (GOGAT, E.C. In general, theGS/GOGAT is thought to be the favoured pathway based on evidence from labeling studies offirst products, equilibrium considerations, inhibitor studies, and the high degree of regulationfound for GS (Miflin and Lea 1976, Syrett 1981, Wheeler 1983, Syrett 1989). GDH isgenerally assigned a role in amino acid catabolism for internal reorganization of cell nitrogen(see Syrett 1989, Robinson et a!. 1991), but under certain conditions it may still be importantin assimilation (e.g. Ahmad and Hellebust 1985a, Calhies et a!. 1992), or in cellular control byadjusting the cells’ glutamine/glutamate ratio (see Flynn 1991). The case for the GS/GOGATpathway limiting nitrogen incorporation has also been made. Since GOGAT activity almostalways exceeds that of GS, GS is thought to be rate-limiting (Syrett 1989).In addition to its roles in nitrate assimilation, there is also evidence that NR mayperform other functions in the cell. Jones and Morel (1988) found a cell membrane-associatedNR in the diatom Thalassiosira weisflogii, and hypothesized a role for NR in controllingplasmalemma redox. The presence of a NR of different molecular weight (representing about0.8% of total cell NR) in membrane fractions of Chiorella sorokiniana was also noted byTischner et a!. (1989) and Tischner (1990). Azura and Aparico (1983) showed that high rates54of nitrite excretion occurred under high light and low CO2 conditions in Chiamydomonasreinhardtii. They suggested that nitrate might be acting as an electron acceptor under theseconditions to adjust levels of reducing power in the cells. Castigetti and Smarrelli (1984) andSmarrelli and Castigetti (1988) have suggested that NR may be involved in reducingsiderophores which are responsible for acquiring metals for cell nutrition. This process maybe quantitatively more important in microalgae in metal-deficient aquatic environments (seePrice et a!. 1991) than for higher plants in soil environments.Structure and characteristics of NRDistinct types of NR exist in prokaryotes, where nitrate is used in place of oxygen as aterminal electron acceptor (dissimilatory forms), or in photosynthetic bacteria andcyanobacteria, where nitrate is used as a nitrogen source (assimilatory forms) (Guerrero et a!.1981). The dissimilatory enzymes are classified as to whether or not chlorate inhibits thenitrate reducing activity (type A) or not (type B), and they are smaller enzymes containingmuch more iron than assimilatory forms (Hewitt 1975). The assimilatory enzymes ofcyanobacteria and photosynthetic (and perhaps chemosynthetic) bacteria differ from NR ineukaryotes in that they use reduced ferredoxin as an electron donor and cannot use pyridinenucleotides (e.g. NADH or NADPH).In contrast, NR in eukaryotes is a large, soluble, multi-centered redox enzyme thatexists in three distinct forms (not including isozymes), based on the source of reducing power:NADH-NR (E.C., the most common form, found in higher plants and algae,NAD(P)H-NR (B.C., which is found in higher plants and green algae, and NADPHNR (B.C., which occurs only in fungi (Campbell and Kinghom 1990). Within thesegeneral categories there is evidence of isozymes (Callaci and Smarrelli 1991); Schuster et a!.(1989) for example showed that there were 4 distinct forms of NR in mustard (Sinapis alba)cotyledons. Table 2.1 gives a comparison of the molecular weights and substrate specificitiesof the purified enzyme from different sources. Comprehensive reviews of the structure of NRare provided by Guerrero et a!. (1981), Solomonson and Barber (1989, 1990), and Wray andTable2.1Characteristicsofnitratereductasefromvarioussources.(MOLWT=molecularweight,REDUCTANT=physiologicalelectrondonor,MONOMERS=numberofmonomersinthenativeprotein,--=informationnotdeterminedbytheauthors)MOLWT.ORGANISMREDUCTANTMONOMERSREFERENCE(kDa)CyanobacteriaAnacystisnidukms83ferredoxin1Andriesseeta!.1989Higher PlantsSpinaceaoleracea230NAD(P)H2Hewitt1975Nicotianap!umbaginfolia214NAD(P)H2Moureauxeta!.1989FungiAspergillusnidulans197NADPH2Hewitt1975Funariahygrometrica232NADPH2PadidamandJohn1991MicroalgaeChiorellasp.370NAD(P)H4SolomonsonandBarber1987Ankistrodesmusbraunii500NAD(P)H8SolomonsonandBarber1990Thalassiosirapseudonana330NADH--AmyandGarrett1974(ii (;i56Fido (1990). The functional size of the enzyme ranges from 200 to 500 kDa, although there issome variation according to the method used to determine the size (Solomonson and Barber1990). The enzyme is composed of single polypeptide chains of about 100 kDa each, whichmay associate as dimers, tetramers or octamers, depending on the species (Solomonson andBarber 1990). Hyde et a!. (1991) demonstrated that the functional domains of the enzyme arevery similar across different species, although the sequence homology is not as highlyconserved as in the case of ribulose 1 ,6-bisphosphate carboxylase oxygenase (RUBISCO) forexample (see Newman and Cattolico 1990). Antibodies raised against NR from squash cross-reacted with the NR of most higher plants, but not with that from Chiorella pyrenoides, orfrom the fungus Neurospora crassa (Cherel et a!. 1986). The molecular weight of thepolypeptide predicted from the DNA sequence is very close to that of the NR protein,suggesting that there is little post-translational modification, aside from the insertion ofcofactors (Sherman and Funkhouser 1989). Each polypeptide subunit has three linearlyarranged domains: a flavin adenine dinucleotide (FAD) region nearest the C-terminal end ofthe protein, a central region with heme-iron contained in a b557-type cytochrome, and a N-terminal component containing molybdopterin (Wray and Fido 1990, Solomonson and Barber1990). It is thought that the electron transfer between the cytochrome and molybdopterin isthe rate limiting step in catalysis (Kay et al. 1991). As will be discussed further under thesection on NR Assay Methods, the NR protein exhibits several so-called “partial activities” inaddition to the full reaction that reduces nitrate to nitrite and oxidizes NADH to NAD.In contrast to the reaction catalyzed by NDPK, which is near equilibrium, the NRreaction (nitrate to nitrite) is generally considered to be irreversible, with the equilibriumconstant (Keq) on the order of 1025 (Hewitt 1976).Regulation of NRThe regulation of nitrate reductase is complex and beyond the scope of this chapter, inwhich only a general discussion is given. Excellent reviews are provided by Fernandez andCardenas (1989), Crawford and Davis (1989), Solomonson and Barber (1990), and Crawford57et a!. (1992). As Solomonson and Barber (1990) point out, there is probably no single modeof regulation of NR, but several modes acting simultaneously, or in sequence.The enzyme appears to be largely regulated by synthesis and degradation of the protein(Sherman and Funkhouser 1989). Sequential induction of transcription (i.e. appearance of NRmRNA’ s) and translation (i.e. appearance of NR immuno-reactive protein) followed byincreases in NR activity have been demonstrated for higher plants (Stewart and Rhodes 1977,Lillo 1991, Li and Oaks 1993), green algae (Sherman and Funkhouser 1989, Diez and LopezRuiz 1989), and marine diatoms (Smith et a!. 1992). The factors involved in the regulation ofNR synthesis are still under debate. NR synthesis has long been held to be induced by nitrate(Stewart and Rhodes 1977, Faure et a!. 1991), but there is evidence that nitrate may not berequired. Kessler and Osterheld (1970) found that in the green alga Ankistrodesmus brauniiNR activity increased when ammonium-grown cells were transferred to N-free medium. Thiswas also found by Amy and Garret (1974) in Thalassiosira pseudonana and Diez and LopezRuiz (1989) in a green alga, leading to the idea that ammonium may repress NR synthesis, butnitrate does not induce it. This effect may be isoform-specific; Calacci and Smarrelli (1991)have shown that of three isoforms, only the pH 7.5 variant of NR in soybean is induced bynitrate. It is also worth noting that Oaks et a!. (1990) showed that trace nitrate contaminationof soil was responsible for a “no nitrate” induction of NR in higher plants. Light (Faure et al.1991, Gao eta!. 1992), alternate carbon sources (e.g. citrate in cucumber cotyledons, Stewartand Rhodes 1977), and some alternate nitrogen sources (e.g. uric acid in certain species ofmicroalgae, Syrett and Hipkin 1973) can also induce NR. NR synthesis is repressed by thepresence of ammonium (Syrett 1989) or other nitrogen sources such as amino acids (Liu andHellebust 1974), but Harrison (1976) reported a NR from a marine dinoflagellate that was notrepressed completely by ammonium. Rasjasekar and Oelmuller (1987) also found that therewas a non-repressible NR in corn. In terms of NR regulation by degradation, a wide range ofNR-specific proteases are known from higher plants and fungi (Wallace 1977).Alternatively, there is a range of situations in which NR activity changes in the absenceof protein synthesis or degradation. In some cases it may be difficult to distinguish these58processes. Tischner and Hutterman (1978) initially thought that light activation of ChiorellaNR was dependent on protein synthesis, based on the inhibition of this activation by theprotein synthesis inhibitor cyclohexamide. They later discovered that the action ofcyclohexamide was non-specific; it inhibited the activation mechanism as well. In green algae,cyanide or superoxides produced at times in photosynthesis appear to convert NR to aninactive form (Pistorius et a!. 1976). Blue light, flavins and mild oxidation with ferricyanidecan be used to reactivate the enzyme (Franco et a!. 1987, Corzo and Neiil 1992b). Thismechanism does not appear to operate in marine diatoms (Serra et a!. 1978a). Light activatesthe enzyme, probably through a phytochrome (Ninneman 1987, De la Rosa et a!. 1989), andthere is a diel periodicity in NR (Packard et al. 1971, Smith et a!. 1992). Specific allostericmodification by adenylates was thought to occur (Eaglesham and Hewitt 1975), but laterevidence shows that this may be mediated through other mechanisms such as phosphorylation.Phosphorylation has been shown to occur in the spinach enzyme during light-dark transitions(Huber et a!. 1992a). Tischner (1984) could distinguish two enzyme forms in Chlore!la, a lowactivity form present at the end of the dark cycle, and a high activity form which appearedabout one hour into the light period. This was hypothesized to be an intramolecular change(Tischner 1984), and it may represent a phosphorylation event. Alternatively, adenylates mayplay other roles in enzyme activity modification (Kaiser and Spill 1991, Kaiser et a!. 1992).There may also be a direct inhibition of NR by ammonium, probably through a product ofammonium incorporation (Syrett 1981, Flynn 1991). Larsson et a!. (1985) showed that in thegreen alga Scenedesmus, inactivation of NR was too rapid to be due simply to degradation ofNR and must involve some inactivation mechanism.In addition to protease action, there are reports of proteins in higher plants that bind toNR and irreversibly inactivate the enzyme, but do not appear to be proteases (Solomonson andBarber 1990, Yoshimura et al. 1992).Other controlling factors may include competition between NR and GDH for reductant(Stewart and Rhodes 1977), the glu/gln ratio in the cells (Flynn 1991), and carbon limitation59(Pace et al. 1990) In one fungus, Mo limitation of NR activity has also been demonstrated(Padidam et al. 1991).NR assay methodsAny consideration of NR activity and its regulation is complicated by the bewilderingrange of assays, extraction buffers and assay conditions that have been employed in differentstudies (see Table 2.2). Assays can be broadly divided into two categories, those that useintact cells (in situ assays), and those that use cell homogenates (in vitro assays).In Situ AssaysThe in situ assay is often termed in vivo in the literature. In principle, cells arepermeablized, provided with nitrate and incubated under conditions where nitrite cannot befurther reduced to ammonium. The nitrite produced is then measured colorimetrically(Hageman and Reed 1980). In this thesis, the term in situ will be used in preference to in vivosince cells permeablized in this manner are not usually viable (see discussion in Corzo andNeill 1992b). The term in vivo is probably better reserved for truly non-invasive monitoringprocedures such as nuclear magnetic resonance (see Roberts 1984).There are many variations in the in situ procedure including the permeablizing agentused (freezing, propanol, toluene, Triton X- 100, cetyl-trimethylammonium bromide) and itsconcentration, whether a buffer is used, the concentration of nitrate provided, and whether areductant or carbon source is provided (see Table 2.2). There are several problems with theseassays. For a true measurement of NR activity, it must be assumed that it is the enzymeactivity which limits the reaction rate, not the ability of nitrate to reach the enzyme (transport),or the ability of nitrite to move out of the cell. It is difficult to verify this assumption (seeHog et al. 1983). Reducing power must not be limiting, but NADH is not readily transportedacross membranes; Lillo (1983) found it necessary to add glucose to barley leaves and allowglycolysis to provide reductant. In addition, the assay must be conducted anaerobically and inthe dark to prevent nitrite from being converted to ammonium via NiR (Lillo 1983). UsingTable2.2.Selectedassaymixturesforinvitroorinsitunitratereductaseassays.(DTI’=dithiothreitol,CYS=cysteine,FAD=flavmadeninedmucleotide,PVP=polyvinylpyrrolidone,*=NADPHusedinplaceofNADH;?=informationnotprovidedbyauthors)ORGANISM°CBUFFERpHDfl’2+EDTACYSFADPVPNONADHOThERREFERENCE(wlv)(m(mM)(mM)(SM)(w/v)(mM)(mM)ADDITIONSINVITROASSAYSSphaerostilberepens20100mMP0437.5----1.0--105%100.4*1MMoEssgaouriandBotton19901%casein10mMPMSFSpinaceaoleracea2450mMPIPES7.6--1010------1.00.5liquidN2Kaiseret at.199250MleupeptinSpinaceaoleracea2550mMP0437.51mM--0.11.010--5.00.1ATPSanchezandHeldt1990Lemnagibba2950mMP0437.8----1.0--50--250.410jMIngemarsson1987leupeptinZeamays30100mMP0437.4----1.01.0----11.70.473%caseinPaceetat.1990Hordeumvulgare2725mMP0437.5----1.0------100.20.1%TritonLillo198315%glycerolHordeumvulgare30100mMP0437.5------5.0----100.4--Tischnereta!.1986Chlamy€knwnas2550mMTRIS7.5.1mM--1.0--10--100.2Francoetat.1987reinhardiiChiorellavulgaris2067mMP0437.6------------6.60.6--Pistoriuseta!.1976Skeletonemacostatum25200mMP0437.91.0%0.844-----0.33.610.167liquidN2Clayton1985f.w.phytoplankion25150mMP0437.6------------100.653%tolueneHochmanetat.1986Table2.2(Continued)ORGANISM°CBUFFERpHDT[Mg2EDTACYSFADPVPNOçNADHOTHERREFERENCE(w/v)(mM)(mM)(mM)(SM)(wlv)(M)(mM)ADDITIONSFucusgairdneriiEnreromorphaintestinatis20200mMP0438.21mM1020200mMP0438.01mM10--5.85--290.29----0.6100.2--4.7•-29029--200.42--------0.211——————11--Everest etat.1986--Eppleyetal.1969--Everest etat.1984liquidN2BalandinandAparicio199210MleupeptinThomasandHarrison1988ThomasandHarrison1988INSITUASSAYSChtamydomonasreinhardiii2525mMHEPES7.5—--250——--1.6—2mMtolueneWatt etat.1992--———200—5%propanolCorzoetat.1991----------20--4%propanolSmithetal.1992--0.5------30--10MglucoseCorzoandNeill19921.1%propanolmarinephyroptankion20200mMP0437.91mM----marine phytoplankton20200mMP0437.91.0%30-100--marinephytoptankron?200mMP0437.91mM----Acetabularia20100mMP0437.51mM--0.5mediterraneaMonoraphidiumbraunii0.20.2Skeleronemacosrarum1529100mMPO437.5—Utvarigida30100mMPO438.0--62the in situ assay, Thomas and Harrison (1988) also found that NR activity was very dependenton the length of incubation with the permeablizing agent propanol. As well, Brinkhuis et al.(1989) point out that uptake of nitrite by cells must also be accounted for in this assay.Sawhney et a!. (1978) cautioned that in situ NR assays are not true reflections of what goes onphysiologically; light, ATP concentration and mitochondrial respiration all affected NR inwheat leaves, and all these parameters were altered under assay conditions. Based on theseconsiderations, it was decided that an in vitro assay offered better quantification of NR activityfor the present study.In Vitro AssaysIn vitro NR assays also present difficulties. The in situ assay is often adopted whenactivity cannot be found using an in vitro technique (e.g. Thomas and Harrison 1988, Corzoand Neill 1992b), although Lillo (1983) compared in situ and in vitro assays in barley leavesand found that in vitro activity was up to 5 times higher. The reasons for failure to detectactivity may have to do with problems associated with the stability of the enzyme when it isextracted. Morris and Syrett (1965) and Hersey and Swift (1976) believed that only a portionof NR was recovered on extraction from Chiorella, and two marine dinoflagellates,respectively. Morris and Syrett suggested that NR from nitrogen-deficient cells was evenmore unstable. Eppley et a!. (1969) calculated that only 25% of NR activity necessary tosupport observed rates of nitrate incorporation was recovered from marine phytoplankton. Onthe other hand, it is possible that these assay conditions themselves may have excluded anecessary cofactor, or been conducted under non-optimal conditions.There are many aspects of homogenization techniques that bear consideration. Cellscan be collected by filtration, although this may cause cell rupture and loss of enzyme(Hochman et a!. 1986). Centrifugation is another method of cell collection, although it istime-consuming, and difficult to use with larger volumes of dilute culture; at high speed it maybe no more gentle than low pressure filtration. Cell disruption has been accomplished byfreeze-thawing in liquid nitrogen (Balandin and Aparicio 1992), grinding in a mortar and63pestle (Eppley 1978), homogenizing with a glass-glass or glass-Teflon tissue homogenizer(Hochman 1982), sonicating (Pistorius et a!. 1976), or using a French press pressure cell(Pistorius et al. 1976). For larger enzymes such as NR, however, there is evidence thatsonication may cause damage to the protein (Pistorius et a!. 1976). Hochman (1982) foundthat in a freshwater dinoflagellate, sonication and French press methods both gave much loweractivities than grinding in a glass-glass culture tube.For in vitro assays the buffer into which the NR enzyme is extracted is criticallyimportant. There is extensive evidence that sulthydryl groups in the active site of the proteinmust be protected by thiol compounds such as cysteine, mercaptoethanol or dithiothreitol(Cleland 1964; Newsholme and Crabtree 1986). Eppley et a!. (1969) noted that D’fl’ wasmore effective than cysteine with marine diatoms, and there is also evidence in barley leavesthat the use of cysteine may stimulate thiol proteases within the cell (Tischner et a!. 1986).Phenolic compounds which are common in the tissues of algae (Thomas and Harrison 1988)may also inactivate proteins; use of polyvinyl pyrrolidone to bind phenolic compounds hasbeen recommended (Loomis and Battile 1966, Gegenheimer 1990). Heavy metals as reagentcontaminants or on glassware may also be problematic, and the use of EDTA can overcomesuch problems (Newsholme and Crabtree 1986). Finally, proteolytic inactivation may occurwith NR, since this is known to occur in fungi, higher plants and green algae (Wallace 1977),although it has not been specifically addressed in marine microalgae. A wide range ofprotease inhibitors are available (Gegenheimer 1990), some of which have been used in NRextraction including leupeptin (Wray and Kirk 1981, Ingemarsson 1987), chymostatin (Longand Oaks 1990), and phenylmethyl sulfonyl fluoride (PMSF, Essagouri and Botton 1990).The addition of protein to extraction buffers has also provided protection from proteases. Bothbovine serum albumin (BSA) and casein have been used at concentrations ranging from 0.1-3% (w/v) (Sherrard and Dalling 1978, Ingemarsson 1987, Pace et a!. 1990).Assay conditions for NR are extremely variable (see Table 2.2). In terms of buffers,phosphate has been found to enhance NR activity 10-30% compared with buffers such as TRIS(Serra et a!. 1978a). In fact, Eppley et a!. (1969) and Everest et a?. (1984) cautioned against64the use of TRIS buffer, which gave lower NR activities. There is also a wide range of bufferscurrently available that may have advantages over traditional buffers such as phosphate orTRIS (see Good a al. 1966). The temperature of the assay is another consideration. Intheory, lower activity can be easily amplified by increasing the assay temperature. Thus, NRassays are often conducted at 25-30°C, which can be more than 10°C higher than the in situtemperatures the organisms experienced. In theory, activity can be back-corrected to the insitu rate using an activation energy derived from an Arrhenius plot (Hochachka and Somero1984; see Packard et al. 1971a), but it is necessary to determine the activation energy.Packard et al. (1971a) found that for Skeletonema costatum grown between 16-19°C, activitymeasured at 25°C was only 25% of the activity measured at 15°C. To make matters worse, insome cases authors have not even reported the assay temperature. Clearly, these problems canbe avoided by conducting assays at the in situ temperatures. Additions of MgSO4 have beenreported to increase NR activity, although this response is variable (Eppley et a!. 1969).Findings by Huber et al. (1992a) that Mg2 inhibits the activity of the phosphorylated NRcomplicate this issue. Kaiser et al. (1992) reported that the effect of Mg2 was overcome byadditions of high EDTA (5 mM). In preliminary experiments, the additions occasionallyincreased activity up to 10%, however they also increased assay variability. With the routineaddition of 5 mM EDTA to assays these effects disappeared; thus in the experiments reportedin this thesis, MgSO4was not added. The issue of the concentrations of NADH (or NADPH)and nitrate to be added must be considered. Determination of maximal activity (Vmax) clearlydemands that substrates be saturating, yet frequently authors do not confirm saturation. Areview of the literature shows a wide range of substrate concentrations (Table 2.2). Finally,additions of flavin adenine dinucleotide (FAD) should be considered as they have long beenknown to enhance NR activity under certain circumstances, especially after partial purificationof the enzyme (Evans and Nason 1953).Activation of NR in assays must also be considered. Additions of ferricyanide havebeen shown to increase NR activity, particularly in green algae (Pistorius et a!. 1976). There65is also an activation mechanism involving pre-incubation with cysteine, above and beyond itsrole as a thiol protectant (Smarrelli and Campbell 1980).Optimization of an assay must be conducted carefully. Clearly, it is possible toincrease activity by using a higher temperature, or by shifting pH outside of the physiologicalrange. However, if an index of what is occurring in vivo is required, the physiologicalconstraints of the system must be accepted. In this study, assays were conducted at the in situtemperature and at a pH of 7.9, which may be to the higher end of the cellular norm (seeGuern et al. 1991), but Amy and Garrett (1974) found that there was a broad pH optimum of7-8 for NR from the diatom Thalassiosira pseudonana.The activity to be monitored is also an issue. The overall reaction (NADH as electrondonor, NO3-as acceptor, NADH-NR) is monitored by either measuring nitrite produced, orNADH oxidized (Wray and Fido 1990). In addition to this activity, NR displays severalpartial activities. For example, in place of NADH, reduced flavin mononucleotide, reducedmethyl or benzyl viologens, or reduced bromphenol blue can donate electrons, resulting in thereduction of nitrate (FMN-NR, MV-NR, BV-NR, or BPB-NR, respectively) (Wray and Fido1990). Nitrite produced is monitored in these assays. Alternatively, electron acceptors otherthan nitrate can be used, for example, cytochrome c (NADH-cytochrome c reductase, alsocalled diaphorase), or ferricyanide (NADH-ferricyanide reductase). NADH oxidized orcytochrome c reduced are monitored spectrophotometrically in these assays (Wray and Fido1990). Since some of these activities do not rely on an intact enzyme, they may to someextent overcome the loss of activity due to protease action (in fact, these activities have beenused along with proteases to elucidate the structure of the enzyme, Wray and Fido 1990).Ingemarsson (1987) compared these partial activities in Lemna. He found that FMN-NR,MV-NR and NADH-cytochrome c reductase activities were all higher than NADH-NR by 20-50%, although the general trends in activity were similar. BPB-NR activity may be 10-15times higher than NADH-NR (Wray and Fido 1990). In a freshwater dinoflagellate, MV-NRwas found to be 3-6 times higher than the NADH-NR. activity and BV-NR was 1.5-3 timesgreater (Hochman 1982). However, there may be problems trying to compare activities,66Yamagishi et at. (1988) for example, noted that NR inactivator protein in spinach completelysuppressed NADH-NR activity, but there was no loss of MV-NR activity. Tischner (1984)also noted that MV-NADH and diaphorase activities in Chiorella did not respond to regulatorymechanisms in the same way as NADH-NR activity. Thus, it appears most sensible tomonitor the full NADH-NR activity.Another class of in vitro assays is the immunoassay (see Whitford et a!. 1987). Intheory, immunoassays do not necessarily require a functional protein since the enzyme reactionneed not be run, and immunoassays may be extremely specific, depending on the specificity ofthe antibody employed. Although immunoassays offer extremely high sensitivity, withdetection of enzyme within single cells possible, they still require a means to convert enzymeconcentration to activity (Baich et at. 1988). However, from a practical point of view, if theenzyme is cleaved by proteolysis, or if the site the antibody recognizes is modified by someregulatory feature, the antibody may not recognize it. Alternatively, if the binding site is nonessential for NR activity, then degraded or inactive enzyme may be detected. There are caseswhere the two methods agree well (Maid et at. 1986, Watt et a!. 1992); however there areother examples where the mismatch between activity and immunoprotein level is difficult toreconcile (Sherman and Funkhouser 1989, Smith et at. 1992). Immunoassays and enzymeactivity assay are therefore distinct and complementary approaches which together providemore detailed information than either alone.Since highly purified NR from marine species of phytoplankton was not available, andthus antisera could not be easily raised, the research in this thesis has focused on the usefulnessof in vitro activity assays by measuring the full NADH-NR activity. The goal of this chapteris to provide a framework for the rest of the thesis by determining an optimal assay for NR inmarine phytoplankton. To simplify considerations at this point, only light-limited semi-continuous batch cultures with nitrate as the sole nitrogen source have been used. Scaling ofactivity was not essential since in a pure culture exhibiting balanced growth, the NR activityand the actual nitrate incorporation rates of the algae could be measured and calculated on aper cell basis. The criteria selected for an optimal assay were that: a) activity be as high as67possible within physiological conditions (i.e. saturating substrates), b) activity be stable at leastover the assay period, and c) activity of NR equal or exceed the calculated rate of nitrateincorporation.MATERIALS AND METHODSGeneral culture conditionsAll species used were obtained from the Northeast Pacific Culture Collection(NEPCC), Department of Oceanography, University of British Columbia. Cultures weregrown in semi-continuous batch culture in artificial seawater medium (ESAW, based onHarrison et a!. 1980, modified as in Chapter 1) at 17.5 ± 0.50 C under continuous light, withnitrate (550 M) as the sole nitrogen source, as described in Chapter 1.Enzyme optimization experimentsFor initial experiments to optimize the NR assay, semi-continuous 1 L batch cultures of thediatom Thalassiosira pseudonana were used. These cultures were harvested in mid-log phaseat cell densities of less than 5.0 x i0 cells ml1.Validation ofspectrophotometric assayTo begin with, the NR assay of Eppley (1978) was used. All reagents were obtainedfrom Sigma Chemical Co. (St. Louis, MO) and were of the purest grade available. Theextraction medium (Buffer A) consisted of 200 mM phosphate buffer, pH 7.9, 1 mMdithiothreitol, and 0.3% (w/v) polyvinyl pyrrolidone (PVP). Cells were harvested on 25 mmGF/F filters and homogenized in 1.0 ml of extraction buffer using a glass-Teflonhomogenizer, as described in Chapter 1. Homogenates were clarified by centrifugation at 750g at 4 °C in a refrigerated Sorval RCB centrifuge for 5 mm. The supernatants were removedusing disposable Pasteur pipettes and used immediately for assays. Assays were conducted in1 cm disposable plastic cuvettes in a total volume of 1.0 ml. The assay mixture contained68final concentrations of 200 mM phosphate buffer, pH 7.9, 0.2 mM NADH and 100-400 lhomogenate. Reactions were initiated by adding 10 mM KNO3.For spectrophotometric assays, the oxidation of NADH was followed at 340 nm(Hageman and Reed 1980), using an LKB Ultrospec II spectrophotometer with a temperature-controlled turret, interfaced to an IBM personal computer (see Appendix B). Enzymeactivities were always determined at the in situ temperature. The initial rate of absorbancechange was followed for 5 mm before addition of KNO3; the rate after KNO3 addition wascorrected for this background to give the nitrate-specific rate, which was converted to a rate ofNADH oxidation using a millimolar extinction coefficient of 6.22 (Hageman and Reed 1980).In the same samples, nitrite production was also determined, using a modification of themethod of Eppley (1978). The reaction was stopped at a specific time (generally 10-15 mm)using 2.0 ml of 550 mM zinc acetate. This concentration is substantially higher than that usedby Eppley (1978), but Eppley also added 6.0 ml of ethanol. The method used in this studycompletely stopped the reaction, but minimized the dilution of the homogenate. Thehomogenate was centhfuged at maximum speed in a clinical centrifuge. Samples (0.5 ml) ofthe clear supernatant were removed for assays. Excess NADH which could interfere with thesubsequent assay of nitrite was oxidized by adding 20 d of 125 JLM phenazine methosulphate(PMS, Scholl et a!. 1974). Nitrite produced was determined colorimetrically using 0.5 mleach of sulfanilamide and N-(1-napthyl)-ethylenediamine 2 HC1 solutions and reading theresulting colour at 543 nm (Eppley 1978).Although there is little doubt that the relationship between nitrate reduction and NADHoxidation is stoichiometric (Evans and Nason 1953, Hageman and Reed 1980), this may not betrue for a crude homogenate versus a purified sample of NR. In order to verify this, activitiescalculated using NADH oxidation and nitrite production were compared in eight differentcultures grown to a range of cell densities. Throughout subsequent experiments, theequivalence of these methods was routinely verified, particularly when investigating a newspecies or a new culture condition.69Cell collection, Homogenization, and Assay OptimizationVariations on homogenization and extraction procedures were made. In oneexperiment, 6 samples were collected from a single culture by filtration as previouslydescribed. An additional 6 samples were collected by centrifuging cell suspensions at 18 000 gfor 15 mm at 4 °C. The pellets were resuspended in 1.0 ml extraction buffer. To account forvolume differences and to improve homogenization, dry 25 mm GF/F filters were added tocentrifuged cell suspensions. Three samples from each collection method were ground aspreviously described. The other three samples were sonicated for 60 s using a Branson model20 sonic disruptor with a microprobe, on the 40% power setting. Homogenates werecentrifuged and assayed as previously described. Activities were compared using t-tests (Steeland Tome 1980). A second experiment ascertained whether extraction of NR from celldetritus was complete. Six samples were collected by filtration from a single culture. Threesamples were placed in normal NR extraction buffer (buffer A), and three were placed inBuffer A plus 0.1 % Triton X-100. All samples were ground and centrifuged as previouslydescribed. The resulting supernatants were removed and the pellets were rehomogenized in anadditional 1.0 ml of extraction buffer. NR was determined in each fraction and activitiescompared using an analysis of variance (ANOVA) followed by Tukey’s least significantdifference (LSD) test. For subsequent experiments, homogenates were collected by filtrationand homogenized by grinding with 0.1% Triton X-100 added.The effect of different buffer compounds on NR activity was compared. Four sets ofthree samples each were collected from a single culture and homogenized and assayed in either200 mM phosphate, 50 mM 3-(N-Morpolino)propanesulfonic acid (MOPS), 50 mMTris(hydroxymethyl)aminomethane (TRIS), or 50 mM imidazole buffers, all adjusted to pH7.9. In this experiment, and in subsequent optimization experiments, NR activities werecompared using an analysis of variance (ANOVA) followed by Tukey’s least significantdifference (LSD) test. Subsequent experiments were all conducted using 200 mM phosphatebuffer, pH 7.9.70The effects of additions to the extraction buffer were tested. Triplicate samples from asingle culture were prepared in: a) buffer (200 mM phosphate plus 0.1 % Triton X-100), b)buffer plus 5 mM EDTA, c) buffer plus 0.03 % (wlv) DTT, d) buffer plus 0.3 % (w/v) PVP,or e) buffer plus EDTA, DTT, and PVP. For subsequent experiments the extraction bufferconsisted of: 200 mM phosphate buffer, Triton X-100, EDTA, lYfl, and PVP (buffer B).The effects of various activating compounds on activity were also evaluated. Triplicatesamples from a single culture were prepared and assayed: a) without additions, b) after a 5 mmincubation with 5 mM cysteine, c) with 0.1 mM flavin adenine dinucleotide (FAD), or d)after a 5 mm incubation with 0.2 mM potassium ferricyanide.The stabilities of enzyme extracts were assessed using three treatments. Sets oftriplicate samples from a single culture were homogenized using: a) buffer B alone, b) bufferB plus 3 % BSA, or c) buffer B plus a protease inhibitor mix recommended by Gegenheimer(1990) consisting of 1 mM PMSF, 1 mM benzamide, 1 mM benzamidine-HC1, 5 mM -amino-n-caproic acid, 10 mM EGTA, 1 g ml1 antipain, 1 g ml1 leupeptin, and 0.1 mgm14 pepstatin. Extracts were assayed immediately, or after 20, 70 or 135 mm. Activitieswere compared within each time interval using t-tests (Steel and Tome 1980). Subsequentextractions were performed adding 3% BSA to the buffer (buffer C).The substrate specificity of the enzyme from T. pseudonana was investigated. Assayswere conducted on triplicate homogenates with either 0.2 mM NADH, 0.2 mM NADPH or amixture of 0.1 mM NADH plus 0.1 mM NADPH.The effect of liquid nitrogen freezing on NR activity was assessed. Four sets oftriplicate samples were collected on 25 mm GF/F filters. One set was homogenized andassayed immediately. The other samples were placed in 1.5 ml Eppendorf microcentrifugetubes and placed in liquid nitrogen. One set was immediately withdrawn from the liquidnitrogen, homogenized and assayed. The other sets of samples were withdrawn and assayedafter 48 or 96 h.71Enzyme KineticsFor the T. pseudonana enzyme, assays were performed with constant NADH (0.2 mM)and KNO3 concentrations ranging from 0 to 10 mM, and with constant KNO3 (10 mM) andNADH concentration ranging from 0 to 0.6 mM. K values for nitrate and NADH wereestimated by fitting the data to a Michaelis-Menten model using a non-linear fitting routine(NONLIN, Wilkinson 1990; see Appendix C). Mean Km values for each substrate wereestimated from determinations on at least 3 separate homogenates.Enzyme DesaltingTo assess the effects of endogenous compounds in the homogenate (e.g. endogenousnitrate), triplicate homogenates of T. pseudonana were desalted using Sephadex 0-25 preparedin columns in 30 ml disposable syringes. The resin was equilibrated with buffer, thenhomogenates were centrifuged through the syringes at 40 C in a refrigerated centrifuge at 750g. NR assays were performed on homogenates before and after passing them through thecolumn, with and without additions of FAD.Enzyme characterization in different speciesSemi-continuous batch cultures of the diatom Skeletonema costatum (Greville) Cleve(NEPCC 18) and the dinoflagellate Amphidinium carterae Hulburt (NEPCC 629) were grownas described for T. pseudonana. Cells were harvested, and NR extracted and assayed aspreviously described, except that in the case of A. carterae OF/F filters were replaced with934 AH filters to prevent clogging by this larger species. As well, because S. costatum formschains of cells, it proved necessary to sonicate samples of this species in a bath sonicator inorder to break up chains for cell counts (Falkowski and Stone 1975). Cell volumes for S.costatum were calculated from the linear dimensions determined under the microscope,assuming a cylindrical shape.72For each species, assays were conducted with additions of FAD, activation with FeCN,and NADPH in place of NADH, as described for T. pseudonana.Km values for nitrate and NADH were determined for each species as described for T.pseudonana.Comparisons of NR activity with growth rateFor T. pseudonana, S. costatum, and A. carterae light-limited growth rate experimentswere conducted as described in Chapter 1, except that irradiances ranged up to 220 molquanta m2 s. Growth rates were monitored using fluorescence and growth irradiance curvesprepared as before (Chapter 1). Cultures were harvested in log phase growth. For eachculture, duplicate NR measurements were made and cell nitrogen quotas measured as describedpreviously (Chapter 1). The nitrate reduction rate necessary to support observed growth(nitrate incorporation rate) was calculated from the product of cell growth rate () and nitrogenquota (QN) assuming that nitrate was the only nitrogen source and that nitrogen quota andinternal nitrate pools were constant (after an acclimation period) over the course of theexperiment. This was converted to units of NR activity and the two variables were comparedfor each culture using linear regression analyses (Steel and Tome 1980).A preliminary survey of NR activity was conducted for 12 species: the diatomsThalassiosira weisflogii (Gru.) Fryxell et Hasle (NEPCC 636), Ditylum brightwellii (t. West)Grunow in Van Huerck (NEPCC 8), and Phaeodactylum tricornutwn Bohlin (NEPCC 31); thechlorophytes Dunaliella tertiolecta Butcher (NEPCC 1), and Chiamydomonas sp. (NEPCC73); the prasinophyte Tetraselmis sp. (NEPCC 46); the cyanobacterium Synechococcus sp.(NEPCC 539); the dinoflagellates Prorocentrum minimum (Pavillard) Schiller (NEPCC 623),and Gymnodinium sanguineum Hirasaka (NEPCC 354); and the prymnesiophytes Paviovalutheri (Droop) Hibberd (NEPCC 2), Emiliania huxleyi (Lohm.) Hay et Mohler (NEPCC659), and Isochrysis galbana Parke (NEPCC 633). Cultures were grown on continuous lightin ESAW as previously described in this chapter, but in 50 ml glass tubes without stirring oraeration. Growth rate was monitored by in vivo fluorescence as previously described.73Quadruplicate cultures were acclimated for a minimum of 8 generations, then sampled for NRactivity and carbon and nitrogen content as previously described in this chapter. Nitrateincorporation rate was calculated as for the other species and compared with NR activity usinglinear regression analysis. In addition, the ability of each species to use NADPH in place ofNADH was tested by conducting NR assays with 0.2 mM NADPH in place of NADH.RESULTSEnzyme optimization experimentsValidation ofspectrophotometric assayFigure 2.1 A shows a typical plot of absorbance versus time in the spectrophotometricNR assay. Typically, the rate of decrease in absorbance before addition of nitrate was 20% ofthe rate after nitrate addition. The absorbance decrease was linear for up to 40 mm under theconditions described, and increased linearly with amount of homogenate added.The spectrophotometric assay agreed very well with the nitrite production assay (Fig.2.1 B). The slope of the NADH oxidation vs. nitrate reduction regression was 0.98 (notsignificantly different from 1.0, P <0.001) and the coefficient of determination (r2) was 0.99.This was verified repeatedly during the course of experiments.Cell collection, Homogenization, and Assay OptimizationNR activities in homogenates collected by centrifugation and grinding were notsignificantly different (P >0.5), but centrifuged samples were more variable (Fig. 2.2 A).Regardless of the collection method, sonication gave consistently lower activities than grinding(Fig. 2.2 A; P <0.01 for filtration, P <0.001 for centrifugation). On centrifugation withoutTriton X-100, significant NR activity (-10% of supernatant) remained in the pellet. Additionof Triton removed this activity, although it did not significantly increase activity in thesupernatant (P>0.08).741.8 i IA•••••••••\\—o time (mm)200- 150EI,• 50B04” I I0 50 100 150 200- .. -l -1 12z NO3 reduction rate (mol mm cell ) x 10Figure 2.1 Validation of spectrophotometric assay for nitrate reductase (NR). A) Timecourse of reaction before and after addition of 10 mM KNO3 (indicated by arrow).B) Comparison of activity calculated from NADH oxidation rate (corrected for NADHoxidation in the absence of nitrate), and that calculated based on production of nitrite.Regression equation is : Y = -0.81 + 0.98 X (r2 = 0.99).751400I200160__supernatant‘ 140 1 pellet -200normal Triton X-100Figure 2.2. Comparison of NR homogenization and extraction procedures in homogenatesof Thalassiosira pseudonana . A) Relative NR activities in samples collected byfiltration onto glass fibre filters, or centrifugation at 7 500 g. In each case, replicatesamples (n 3) were homogenized by grinding or by probe sonication. B) RelativeNR activity in the supernatant and pellet fractions of homogenates of cells collected byfiltration and homogenized by grinding. Homogenizations were performed with orwithout 0.1 % Triton X-100. Centrifugations were done at 750 g for 5 mm.filtration centrifugation76NR activity in phosphate buffer was significantly higher than in TRIS or imidazole(P <0.02 and P <0.01, respectively), but no higher than in MOPS buffer (P >0.06) (Fig.2.3).Single additions of EDTA, DTI’, or PVP significantly increased NR activity (Fig. 2.4A, P <0.05 in all cases). NR activity was numerically highest with the addition of all threereagents, but not statistically different from any of the single additions (P >0.08 in all cases).Additions of FAD had no effect on NR activity (Fig. 2.4 B, P >0.2), but cysteine orferricyanide preincubations resulted in significantly lower activities (Fig. 2.4 B, P <0.01 inboth cases).Addition of BSA to the extraction buffer gave over 50% higher NR activity thanhomogenization without BSA, or homogenization with protease inhibitors (Fig. 2.6, P <0.04in both cases). Whereas activity dropped significantly by 60 mm in homogenates without BSA(P <0.05 in both cases), no significant decrease in activity in the BSA extract was seen at thesame time (P <0.05). By 135 mm, however NR activity in the BSA extract had significantlydeclined (P <0.05).Activity of the T. pseudonana enzyme with NADPH as reductant was less than 10% ofthe activity with NADH, and not significantly different from zero (Fig. 2.6, P >0.07).Activity with 0.1 mM NADH and 0.1 mM NADPH was significantly lower than that withNADH alone (P <0.04) and similar to what would be expected with 0.1 mM NADH alone(see Fig. 2.8 A).Filtered samples which were frozen in liquid nitrogen had identical NR activity to thosethat were not (Fig. 2.7, P >0.5). No decrease in activity was seen in samples stored for 48 or96 h (P >0.5 in both cases), although variability of the assay appeared to increase.77140120a)a). 80C-)c’s 60a).—40a)200Figure 2.3. Relative NR activity in homogenates of Thalassiosira pseudonana preparedin 200 mM phosphate buffer, 50 mM MOPS buffer, 50 mM TRIS buffer, or50 mM imidazole buffer. In all cases, pH was 7.9. n = 3 for each buffertreatment.phosphate MOPS TRIS imidazole78240C.)‘-4ci): 120: ::0 1 2 3 4 5120C.)‘-4‘1)j__ __normal FAD FeCNFigure 2.4. Effects of different additions on nitrate reductase (NR) activity inhomogenates of Thalassiosira pseudonana. A) Activity in homogenates with only200 mM phosphate buffer and 0.1 % (v/v) Triton X-100 (1) versus: 5 mM EDTA(2), 0.3 g i DTT (3), 3.0 g PVP (4), or DTT, EDTA and PVP (5).B) NR activity in homogenates prepared using only buffer 5, or with additions of0.1 mM FAD or 0.2 mM ferricyanide (n = 3 in all cases).A7976-I-IS.—‘SC.)Z210-30 60 90 120 150time (mm)Figure 2.5. Stability of NR activity over time in homogenates of Thalassiosirapseudonana homogenized without additions ( • ), with addition of 3 % BSA( • ) or with additions of protease inhibitors as described in the text ( V ).Points represent means plus standard errors of 3 separate homogenates.I I I II.----± --—----5- TTJ_J_080120-.‘CC)80I.,-C).——200Figure 2.6. Relative NR activity in homogenates of Thalassiosira pseudonanaprovided with different reductants: 0.2 mM NADH, 0.2 mM NADPH, or0.1 mM NADH plus 0.1 mM NADPH Error bars represent standard errorsof the mean of 3 separate homogenates.NADH NADPH NADH +NADPH81200-r’ 150Z 500time (h)Figure 23. NR activity in homogenates of Thalassiosira pseudonana before (t = 0 h)and after (t = 48, 96 h) freezing and storage in liquid nitrogen. Points representthe mean and standard error of 3 separate samples.I I I I0 24 48 72 9682Enzyme KineticsFor T. pseudonana, the Km for nitrate was found to be 0.047 (±0.006) mM (Fig 2.8A). For NADH, an inhibition of activity was seen at levels greater than 0.2 mM. CalculatingKm only for concentrations lower than this gave a Km of 0.0 17 (±0.003) mM.Enzyme DesaltingResults of desalting experiments are shown in Table 2.3. Whether assessed by NADHoxidation or nitrite production, Sephadex-treated samples had lower NR activity (P <0.01).Addition of FAD made no difference to homogenates that had not been Sephadex-treated(P >0.4), but activity of Sephadex-treated samples did not differ from that of normally treatedsamples with added FAD (P >0.4).Enzyme characterization in different speciesFor Skeletonema costatum, ferricyanide treatment resulted in an almost complete lossof NR activity (Fig. 2.9). FAD, however, increased NR activity. As Figure 2.9 shows, theenhancement of NR activity was highly variable: in one trial NR increased by 50%, but in asecond trial the increase was over 250%. No activity was found when NADPH wassubstituted for NADH. For S. costatum, the Km for KNO3 was calculated as 0.146 (± .022)mM, and the Km for NADH was 0.048 (±0.005) mM (Fig. 2.10). Unlike the case for T.pseudonana, high NADH levels did not inhibit NR activity.In Amphidinium carterae, FAD addition caused a decrease in activity (Fig. 2.11),although this response differed between homogenates, resulting in high variability. In contrastto the case in other species, ferricyanide had no effect on NR activity (Fig. 2.11, P >0.3). Noactivity was found when NADPH was substituted for NADH (Fig. 2.11). For A. carterae, theKm for KNO3 was calculated as 0.075 (± .012) mM (Fig. 2.12 A). High levels of nitrate(>1 mM) appeared to inhibit NR activity, but this response varied between extracts. The Kmfor NADH was 0.150 (±0.045) mM (Fig. 2.12 B). High NADH levels did not inhibit NRactivity, as in S. costatum, but not T. pseudonana.83.-.()zFigure 2.8. NR activity versus substrate concentration for: A) KNO3 and B) NADHin homogenates of Thalassiosira pseudonana. Curves are fit to rectangularhyperbolae. Km values are 0.0165 mM for NADH and 0.0471 mM for KNO3...AI I I I I543210765432100 2 4 6KNO3 (mM)8 10 12I I-4.—C.)zB...0.0 0.1 0.2 0.3NADH0.4(mM)0.5 0.6 0.7Table 2.3. Effects of addition of 0.1 mM FAD on nitrate reductase activity in homogenates of Thalassiosira 84pseudonana. NR activity was determined by NADH oxidation rate or nitrite production rate and eitheranalyzed directly (Normal) or desalted using a Sephadex G-25 column (Sephadex). Values representmeans and standard errors of 3 replicate assays.NR activity (U m11)Homogenate treatment Assay addition NADH oxidation N02 productionNormal none 2.90 ± 0.13 2.86 ± 0.08Sephadex none 1.13 ± 0.13 1.29 ± 0.17Normal FAD 2.73 ± 0.09 3.40 ± 0.13Sephadex FAD 2.84 ± 0.13 3.31 ± 0.0785300250200.—.—C.)I).-450Figure 2.9. Comparison of effects of addition of FAD and FeCN on NR activity inhomogenates of Skeletonema costatum . Activity is expressed relative to theactivity in the homogenate without additions. Trials 1 and 2 represent two separateexperiments using two different cultures. Error bars represent standard errors of the meanof 3 homogenates.normal FAD FAD FeCNtrial 1 trial 2860z-I—-...—z0 5 10 15KNO3 (mM)20 2.10. Nitrate reductase (NR) activity versus substrate concentration for: A)KNO3 and B) NADH in homogenates of Skeletonema costatum. Curves are fit torectangular hyperbolae. Km values are 0.146 mM for KNO3 and 0.0476 mMfor NADH.0.0 0.1 0.2 0.3 0.4NADH (mM)87120ci)C)‘-4I)80Figure 2.11. Comparison of the effects of additions of activators on NR activityin homogenates from Amphidinium carterae. Additions to the standard buffer are:0.1 mM FAD, or 0.2 mM ferricyanide (FeCN). Error bars represent the standarderror of the mean of 3 separate homogenates.normal FAD FeCN881. I I I II7......11I I I I IzC.)z‘-40z0 5 10 15 20 25KNO3 (mM) 2.12. Nitrate reductase (NR) activity versus substrate concentration for: A) KNO3and B) NADH, for homogenates of Amphidinium carterae. Curves are fit torectangular hyperbolae. Km values are 0.075 mM for KNO3 and 0.150 mM forNADH.0.0 0.1 0.2 0.3 0.4NADH (mM)89Comparisons of NR activity with growth rateGrowth irradiance curves for T. pseudonana, S. costatum, and A. carterae arepresented in Figure 2.13. For T. pseudonana, was 1.84 d-1 and Kj was 33 jmol quantam2s1. For S. costatum, two separate experiments gave very different results. In the firstexperiment, higher growth rates were achieved giving = 1.08 d1 and Kj of 79 molquanta m2 s, while in the second experiment, ILmax was 0.33 d1 and Kj was 25 jmolquanta m2s4. For A. carterae, tmax was 1.07 d1 and K1 was 68 mo1 quanta m2Figure 2.14 shows the comparison between NR activity and the calculated rate ofnitrate incorporation for each species. For T. pseudonana there was a highly significantrelationship (r2 = 0.99, P <0.001) in which the slope of 0.95 (±0.40) was not significantlydifferent from 1.0 (P>0.3). For S. costatum the relationship was not significant (P>0.5),however, if a single culture was dropped from the analysis (see asterisk, Fig. 2.14 B)), therelationship became significant (r2 = 0.95, P <0.02) and the slope of 0.87 (± 0.06) was notdifferent from 1.0 (P>0.05). For A. carterae there was a significant relationship (r2 = 0.71,P <0.04), but the slope of the relationship, 0.19 (±0.06) was significantly lower than 1.0(P <0.01) indicating that NR activity accounted for less than 20% of the calculated nitratereduction rate.In Figure 2.15, the NR activity is compared to the calculated nitrate reduction rate for12 species of marine phytoplankton. The regression is significant (P <0.05), with a slope of0.786 which is significantly lower than 1.0 (P >0.06). This implies that for these speciesthere is a tendency for NR to underestimate the calculated rate of nitrate incorporation.However, for individual species, there is wide variation. Diatoms tend to fall close to the 1:1relationship, but species such as Dunaliella tertiolecta and Emiliania huxleyi show much higherNR activity than can be accounted for by calculated rates. Alternatively, only very low NRactivity was detected in dinoflagellates tested, and no activity was found in the cyanobacteriumSynechococcus sp.As shown in Table 2.4, species from the Chlorophyceae and Prasinophyceae (i.e. greenalgae) were able to use NADPH in place of NADH. Some activity was found in the900 50 100 1502. 50 100 150 200 250 300I0 25 50 75 100irradiance (mo1 quanta m2 s1)Figure 2.13. Growth rate versus irradiance curves for: A) Thalassiosira pseudonana( • ), B) Skeletonema costatum ( I ), and C) Amphidinium carterae ( V ).Curves are fit to rectangular hyperbolae (parameters are given in the text).Each point represents the mean and standard error of three growth ratedeterminations from a separate culture. Note two experiments are includedin B).12591CI—C).0E.—C)zFigure 2.14. Nitrate reductase (NR) activity versus N03 incorporation ratecalculated from growth rate and nitrogen quota for: A) Thalassiosira pseudonana( • ), B) Skeletonema costatum ( • ), and C) Amphidinium carterae ( V ).Points represent means and standard errors of 2 enzyme measurements fromindividual cultures. Dashed lines represent least squares regressions. Solid linesrepresent the 1:1 relationships. Regression parameters are given in the text.0 50 100 150 20020015010050030025020015010050001250100075050025000 250 750 1000 1250calculated N03 incorporation rate (JLmol mind cell’) x i1250 100 150 200 250 30050092100calculated N03 incorporation rate (jmol mm4 rn!4) x i12Figure 2.15. Nitrate reductase (NR) activity versus nitrate incorporation rate(calculated from cell growth rate and cell nitrogen quota) for 12 species ofmarine phytoplankton. Solid line represents the least squares regression.Dashed line represents the 1:1 relationship. Points represent mean NR activitywith standard error for 2 NR assays from duplicate cultures. 0 chiorophytes,• diatoms, E prasinophytes, • prymnesiophytes, V cyanobacteria,‘V dinoflagellates. Species are indicated by abbreviations as explained inTable 2.4. Equation of the regression line is: Y = -8.34 + 0.786 X (r2 = 0.71).0 20 40 60 8093Table 2.4. NR activity in various species of phytoplankton using NADPH as a reductant. Activities areexpressed as a mean percentage (± standard errors, n = 2 cultures) of activity found using NADH.SPECIES ABBREVIATION % ACTiVITYBacillariophyceaePhaeodactylum tn cornutum Pt 4.37 ± 1.59Thalassiosira weisfloggii Tw 2.31 ± 3.99Ditylum brightwellii Db 29.9 ± 16.9ChlorophyceaeDunaliella tertiolecta Dt 93.9 ± 3.4Chiamydomonas sp. Csp 79.0 ± 23.3CyanophyceaeSynechococcus sp. SspPrymnesiophyceaePavlova lutheri P1 35.2 ± 9.6Isochiysis galbana Ig 0.58 ± 8.01Emiliania huxleyi Eh 1.03 ± 0.47PrasinophyceaeTetraselmis sp. Tsp 39.0 ± 14.6DinophyceaeGymnodinium sanguineum Gs 1.69 ± 1.29Prorocentrum minimum Pm 0.30 ± 0.3194prymnesiophyte Paviova lutheri;. however, in all other cases activities were not significantlydifferent from zero (P >0.1 in all cases).DISCUSSiONEnzyme assay optimizationThe spectrophotometric assay for NR is simple, rapid and appears to provide estimatesidentical to those obtained using the nitrite production assay. This was also found by Evansand Nason (1953) and verified by Amy and Garrett (1974) who used the assaysinterchangeably. It offers the additional advantages that time-dependence and linearity of theassay can be confirmed for a single assay.For T. pseudonana, collection of cells by filtration appeared to be as effective andpotentially more reproducible than centrifugation. Filtration is more feasible when dealingwith large-volume samples. Homogenization by sonication gives lower NR activity thangrinding. This is similar to findings by Pistorius et a!. (1976) and Hochman (1982).Hochman (1982) found that NR activity in a freshwater dinoflagellate was improved bygrinding in a glass-glass versus a glass-Teflon homogenizer, but Hochman did not use filteredsamples. The inclusion of a filter with the cells probably improved the grinding technique;certainly Eppley et a!. (1969) and Serra et a!. (1978a) remarked on its favorable effects,noting full release of activity within 30 s. Dortch et a!. (1984) confirmed by microscopy thatthis techniques resulted in complete fragmentation of cells. The issue of whether the resultinghomogenate should be centrifuged must be considered. The risk is that NR might not be fullyreleased from cells that are sedimented, or might be adsorbed to particulate material such asfilter fragments. However, use of Triton X-100 appears to release NR completely from thepellet. As well, having excess protein in the homogenate (as BSA) may mean that non-specificbinding will remove BSA and minimize the loss of the less concentrated NR protein. Theremoval of membrane fractions may also minimize side reactions that might oxidize NADH.95Phosphate buffer clearly gives highest NR activity. MOPS also seems acceptable, butas previously noted (Serra et al. 1978a) TRIS is unsuitable and imidazole worse yet. Good etal. (1966) have pointed out that buffer choice is largely individual to enzymes; there isprobably no single “best” buffer for all circumstances. The vast majority of researchers haveselected phosphate buffer for NR assays (see Table 2.2). In terms of the extraction buffer,additions of EDTA, DTT and PVP seem justified, based on their enhancement of activity.Everest et at. (1984) found that 2-4 mM EDTA gave highest activity in several marinephytoplankton species. In a cyanobacterium, Herrero et a!. (1984) proposed that sulfhydrylcompounds stabilized enzyme activity by keeping NR amino acid residues reduced. OxidizedNR was apparently more susceptible to proteolytic attack; nitrate itself could stabilize activity.Effects of PVP and DTT have been found to enhance NR activity in a variety of organisms,although there are conflicting and species-specific results. For example, Eppley et al. (1969)found that cysteine was sufficient for flagellate NR protection, but that diatom species requiredDTT. On the other hand, in macroalgae Thomas and Harrison (1988) noted that PVP wasrequired to obtain activity in Fucus species, but actually inhibited activity in Enteromorphaspecies. Where possible, and certainly in laboratory studies, assays should be optimized inindividual species. Although combined additions of EDTA, PVP and DTT increased activityno more than single additions, activity was no lower, and therefore all additions were routinelymade in subsequent assays.BSA increased NR activity in T. pseudonana assays, and improved the stability of theenzyme. From the literature it is possible that this is due to protection from proteases thatdegrade MR either specifically or non-specifically (Wallace 1977). The fact that proteaseinhibitors had no effect in the present study does not necessarily refute this idea because theeffectiveness of the tested protease inhibitors varies with the species, and proteases resistant tothe particular inhibitors used in this study may have been present. For example, leupeptin iseffective against proteases found in Lemna (Ingemarsson 1987) and in barley (Hamano et a!.1984), while PMSF is not. But PMSF was effective in preventing loss of MR activity in a96fungus (Essgaouri and Botton 1990). In corn roots, chymostatin was required (Long and Oaks1990). The effectiveness of BSA has also been noted by Ingemarsson (1987) and Tischner etal. (1986), but in these cases casein was also effective; this was not true in the present study(data not shown). Casein also stabilized NR in wheat leaves (Sherrard and Dalling 1978), andin tomato plants (Ramon et a!. 1989). BSA may act by providing a protein in higherconcentration that the proteases can degrade in preference to NR, but there may be othereffects. Schrader et a!. (1974) proposed that casein and BSA stabilized NR in corn by bindinginhibitory compounds. Some enzymes also show increased activity when the concentration ofprotein in the assay is increased. Without added protein, the total protein concentration in theassay is likely much lower than that found in vivo, and this dilution may adversely affectcertain enzymes (Newsholme and Crabtree 1986, Aragon and Sols 1991). Alternatively, BSAmay decrease non-specific adsorption of NR protein, or it may also be effective in bindingphenolics that are not trapped by PVP (Gegenheimer 1990). Even in the presence of BSA, NRwas still not completely stable; after 120 mm, activity had dropped by almost 20%. This is animprovement over the decay seen by Hersey and Swift (1976), where dinoflagellate NRdeclined by half in 2-3 h. Similar time-dependent decreases in activity after homogenizationhave been seen by Essgaouri and Botton (1990) where 50% of fungal NR was lost in 1 h at 20°C. Everest et al. (1984) reported that NR from marine phytoplankton was stable for up to 24h at 0°C, and Harrison (1976) found that NR from Gonyaulaxpolyedra had a half-life of 24-30 h, but in these cases the assay incubations were as long as 1 h. Substantial enzymedegradation may have occurred during this time; subsequent degradation may have happenedmore slowly, giving results which were interpreted as indicating enzyme stability. In T.pseudonana, NR was stable for up to 60 mm and assays could typically be performed withinthe 15 mm of the extraction; thus activity assays were probably better than in previous studies,but there is still room for improvement.The improved assay was reflected in terms of NR preservation in liquid nitrogen.Ahmed et al. (1976) found that ETS activity and GDH activity in whole cells of marinephytoplankton could be preserved without loss for up to one year in liquid nitrogen. In97contrast, Clayton (1986) reported losses of activity in Skeletonema costatum NR of up to 40%immediately after freezing. Such losses were not found in this study and this suggests that infield situations samples could be maintained frozen and await future analyses.Enzyme characterization in different phytoplankton speciesThe use of NADH versus NADPH and the kinetic constants for NADH and nitrateprovide some basis for comparing enzymes from different species. In this study the onlyspecies able to use NADPH were the green algae (chlorophytes and prasinophytes), althoughthere was some evidence that at least one prymnesiophyte might also be able to use NADPH.The cyanobacterium could apparently use neither. This is in accord with a review by Syrett(1981) showing that green algae alone used NADPH. However, Hochman (1982) noted thatNADPH-NR activity represented only 16.5% of NADH activity in the freshwaterdinoflagellate Peridinium cinctum, and Serra et al. (1978) found some low activity (about 16%of NADH) in the diatom Skeletonema costatum. One possible explanation for this discrepancymight involve a membrane bound NADH:NADPH transhydrogenase (see Jackson 1991). Ifsuch an enzyme were present in cell homogenates, added NADPH could be converted toNADH and used by NADH-specific NR. Since this is most likely to happen in crudehomogenates that are not centrifuged, the results of these studies must be considered carefullyand weighed against results obtained with the purified enzyme. The meaning of differences inpyridine nucleotide specificity is uncertain. Evolutionarily, nitrate utilization was probablyfirst dissimilatory, before oxygen was present in high concentration on Earth (see Mancinelliand McKay 1988). Since cyanobacteria do not use pyridine nucleotides, it is likely that thedivision of MR enzymes into NADH and NADPH forms arose later on. Although well beyondthe scope of this study, it is tempting to speculate that different MR forms may have haddifferent functions. Classically, NADPH is thought to be used primarily in biosyntheticpathways, while NADH is used in degradative energy-producing pathways (Hochachka andSomero 1984). This may have represented a division between, for example, assimilatory anddissimilatory nitrate reduction, although some present day bacteria also use NADH as an98electron donor for dissimilatory nitrate reduction (Stouthamer et al. 1980). Evidence of otherfunctions of NR is accumulating (see Jones and Morell 1988, Castigetti and Smarrelli 1986).The fact that fungi have a NADPH-specific NR (Hewitt 1975) is also interesting.In terms of kinetic behavior of NR, there appear to be distinct species differences.There was evidence that greater than 0.2 mM NADH inhibited NR from T. pseudonana, butS. costatum and A. carterae were unaffected even at 0.4 mM. NADH inhibition of NR hasbeen noted by Serra et al. (1978) in S. costatum, but only at concentrations above 0.6 mM.Hochman (1982) found no NADH inhibition of a freshwater dinoflagellate. Alternatively, A.carterae NR appeared to be inhibited by high nitrate. Despite this finding in two separatekinetic experiments, in the growth rate study, NR activities determined at 1 mM and at 10 mMKNO3 were no different. The reason for this difference is unclear. Km values for NADH andnitrate obtained in the present study are well within the range of those previously found (Table2.5). According to Packard (1979) Km values for nitrate are typically between 0.05 and 0.15mM in marine microalgae. All values in the present study fall in this range. Since uptake ofnitrate typically has a much lower Ks (on the order of 0.0001 to 0.01 mM, Syrett 1981) thishas been taken as evidence that there must be a high-affinity uptake system for nitrate.In terms of FAD additions, no difference in NR activity was found for T. pseudonana,but S. costatum activity was increased. This increase was highly variable; from 275% of thecontrol in one experiment to less than 10% in growth rate experiments. This variability hasbeen noted previously in marine phytoplankton species. Eppley et a!, (1969) found no effectsof FAD, but Dortch et al. (1979) found increased activity. Everest et a!. (1984) reporteddifferent effects of FAD with different marine phytoplamkton species. Insight into thesedifferences is provided by experiments in which T. pseudonana NR was desalted using aSephadex column. There was a loss of activity on desalting which could be completelyrestored by adding FAD. This effect has been previously reported by Evans and Botton (1953)in higher plants and in Chiorella by Vennesland and Solomonson (1972). It appears to be aneffect of dissociation of the FAD cofactor from the enzyme protein. This may be a function ofthe species (or even the strain, see Vennesland and Solomonson 1972), but also theTable2.5Kineticconstantsfornitratereductasefromvariouseukaryotes.ORGANISMKmN03KmNADHREFERENCE(mM)(mM)Higher PlantsSpinaceaoleracea0.0130.007WrayandFido1990FungiSphaerostilberepens1.40.032EssgaouriandBotton1990MicroalgaeChiorellavulgaris0.084--Guerreroetal.1981Thalassiosirapseudonana0.0620.015AmyandGarrett1974Thalassiosirapseudonana0.063-0.083--Packard1979Thalassiosirapseudonana0.0400.008SmarelliandCampbell1980Skeletonemacostatum--0.017Clayton1985Skeleronemacostatum0.240.020Serraeta!.1978Dirylumbrightwellii0.1100.020Eppleyetal.1969Peridiniumcinctum0.190.28Hochmau1982‘0100homogenization procedure used. It might be interesting to examine the effect of FAD additionon NR activity after homogenization by French press or sonication. In Amphidinium carteraeFAD addition produced highly variable results; NR activity was numerically lower with FAD,but this difference was not statistically different. It appears that routine FAD addition maytherefore be a sensible precaution, even if not an absolute requirement for assays.No species tested demonstrated increased MR activity with ferricyanide additions, andin fact, the diatoms showed a decrease in activity. NR activation by ferricyanide has beenfound in green algae (e.g. Pistorius et a!. 1976), but could not be demonstrated in S. costatwn(Serra et a!. 1978a), nor in the freshwater dinoflagellate Peridinium cinctum (Hochman 1982).Thus, this form of NR activation may be confined to green algae.Comparisons of NR activity with growth rateWith the improved assay, measured NR activity could be equated with calculated NRincorporation rates in T. pseudonana. This has not been consistently achieved before, andsuggests that improvements in the MR assay are responsible. If this is true, it might bereflected in a comparison of MR activity found in the present study to other MR activitiesavailable in the literature. Unfortunately, this is not possible for a number of reasons. Tobegin with, assays have been conducted on cells grown at different temperatures, underdifferent light levels or light:dark cycles, or with alternate nitrogen sources in the medium.Even when culture conditions are clearly specified, growth rates of cells are rarely provided.To make matters worse, activities are often scaled to cell number without any indication of cellsize or nitrogen content, to variables such as dry cell weight or packed cell volume that arenearly impossible to compare with data from the present study. Furthermore, where activitiesare scaled to cell protein, the variability in protein assays due to the assay used, the extractionmethod used, or the protein standard employed, may make comparisons impossible (seeAppendix A). For example, if protein extraction was incomplete (i.e. the sample washomogenized in distilled water or buffer alone) MR activities might be inflated, while if TCAprecipitation were not used, protein would likely be overestimated, with the result that MR101activity would appear to be lower. In a few cases where a comparison is possible, results arevariable (Table 2.6). For T. pseudonana, NR activities of up to 150 x 10-12 U cell-1 werefound in the present study, which are equal to 0.25 U mg protein-1. These clearly exceedthose of Amy and Garrett (1974) and Smarrelli and Campbell (1980). NR activities in S.costatum were generally less than 120 x 10-12 U cell-1 , or 0.025 U mg protein. Thesevalues exceed Serra et al. (1978a), Clayton (1986) and Smith et al. (1992) (see Table 2.6). Inall these studies, culture growth rates exceeded those in the present study. In fact, based onliterature values, the S. costatum cultures in the present study were not growing optimally;rates of up to 2.0 d-1 should be possible (Salcshaug and Andresen 1986). Despite this,activities in the present study are higher, although cell size differences cannot be discountedbecause adequate information on relative cell sizes is not provided by these authors. Incontrast, Kristensen (1987) reported NR values of up to 150 x 10-12 U cell-’ in the samespecies which are more similar to the results of the present study. Again, details for adequatecomparison are missing. As for T. pseudonana, NR activity equaled or exceeded thecalculated nitrate incorporation rates in this species. With one exception, these rates were veryclosely related. NR activities in A. carterae in the present study were less than 250 x 10-12 UcelP1. Hersey and Swift (1976) showed activities considerably higher: up to 6000 x 10-12 Ucell-1. It is difficult to resolve this difference. Hersey and Swift (1976), grew cultures atgrowth rates of up to 1.2 d-1, or twice as high as those in the present study, but nitrogenquotas were near 25 pg cell-1 and thus smaller than the range of 50-60 obtained in the presentstudy. It is unclear why these differences occurred, but it may be related to the fact that cellsin the Hersey and Swift (1976) study were grown on a light:dark cycle. In A. carterae, incontrast to the other species, NR activity accounted for less than half the calculated nitrateincorporation rates in most cases.These results were mirrored in the multi-species comparisons. It must be emphasizedthat this experiment was meant only as a broad test of the applicability of the NR assaydeveloped for T. pseudonana; the assay was not optimized for any of the survey species. NRactivity was quite close to calculated nitrate incorporation rates in the diatom species tested,Table2.6.Representativenitratereductaseactivitiesfromeukaryoticmicroalgae,determinedbyassayinvitroorinsitu.ORGANISMASSAYTYPEACTIVITYREFERENCEChiorellasp.ChiorellastigmatophoraChiorellavulgarisDunaliellaprimolectaDunaliellatertiolectaStichococcusbacillarisBrachionwnassubmarinaNannochioropsisoculataPlatymonassubcordformisPlatymonastertrathelePaviovalutherilsochrysisgalbanaPorphyridiumpurpureumEmilianiahuxleyiThalassiosirapseudonanaThalassiosirapseudonanainsituinvitroinvitroinvitroinvitroinvitroinvitroinvitroinvitroinvitroinvitroinvitroinvitroinvitroinvitroinvitro0.007Umgproteiif10.0087Umgprotein10.030Umgprotein0.036Umgprotein0.0225Umgprotein10.030Umgprotein0.0058Umgprotein0.0032Umgprotein0.0067Umgprotein212x102Ucel10.00Umgprotein0.010Umgprotein0.0050Umgproteiif1212x10-12Uce1I0.0108Umgprotein0.09UmgproteinHochmanetal.1986Everestetal.1986MorrisandSyrett1965Everest eta!.1986Everesteta!.1986Everesteta!.1986Everesteta!.1986Everesteta!.1986Everest etal.1986SyrettandHipkin1973Everesteta!.1986Everesteta!.1986Everest eta!.1986Kristiansen1987AmyandGarrett1974SmarelliandCampbell1980Table2.6.(Continued)ORGANISMASSAYTYPEACTIVITYREFERENCEThalassiosiraoceanicainvitro0.058x1012UcellEppleyandRenger1974Skeletonemacostatuminvitro21.6x10-12Ucel1Clayton1985Skeletonemacostatwninvitro150x10.12Ucell4Kristiansen1987Skeletonemacostatuminvitro0.0020Umgprotein4Serraetat.1978Skeletonemacostatuminsitu25x1042UcelltSmithetat.1992Phaeodactylumtncornutuminvitro0.0050Umgprotein4Everest etal.1986Ditylumbrightwelliiinvitro0.0033UmgproteinEppleyetat.1969Ainphidiniumcarteraeinvitro6000x1012Uce1lHerseyandSwift1976Cachoninanieiinvitro4000xiO’2Ucell1HerseyandSwift1976Gonyaulaxpolyedrainvitro1670xi012Ucell4Harrison1976()104but NR activity overestimated the rates for most of the prymnesiophytes and one of thechiorophytes. In contrast, there was insufficient NR activity to account for observed rates ofnitrate incorporation in Tetraselmis sp., and in the dinoflagellates, the activity was near zero.NR has previously been found to be difficult to extract from certain dinoflagellates. Herseyand Swift (1976) detected no activity in Pyrocystis noctiluca or Dissodinium lunula, whileGymnodinium sanguineum was also a problematic species (Cochlan, Dortch and Doucette,unpublished). However, this appears to be species specific since Harrison (1976) andHochman (1982) found MR activity in two different dinoflagellate species. It is noteworthythat in the present study MR activity was detected in Paviova lutheri, a species in whichEverest et a!. (1986) failed to find activity.Remaining problems with NR assaysIt is clear that for certain species, MR assays remain problematic. This may be afunction of the fact that the MR assay was only fully optimized for one species. Clearly, theamount of work necessary to do this for every species would be enormous and is beyond thescope of this study. For cyanobacteria, as demonstrated in this study, NADH and NADPHcannot be used as electron donors for MR. According to the literature, these species requirereduced ferredoxin (Guerrero et a!. 1981). Provision of this compound, however, would alsosupport NiR activity. This might decrease the MR activity measured using nitrite production,since the NiR enzyme could then use ferredoxin and reduce nitrite to ammonium (Wray andFido 1990).It appears that, at least for diatoms and particularly for T. pseudonana, the assay isacceptable. In the following two chapters, the relationship between MR activity and nitrateincorporation rates under different conditions will be explored in this species.105CHAPTER 3: RELATIONSHIPS BETWEEN NITRATE REDUCTASE ACTIVITYAND GROWTH RATE UNDER STEADY-STATE LIGHT OR NUTRIENTLIMITATION IN THALASSIOSIRA PSEUDONANAINTRODUO’IONFrom Chapter 2, it is apparent that for T. pseudonana under steady state lightlimitation, NR activity matched observed rates of nitrate incorporation quite closely. In thischapter, the goal is to explore, in detail, the relationships between growth rate, cellcomposition and nitrate reductase activity under conditions of steady state light limitation,transitions in irradiance and steady state nitrate limitation, and to determine whether NRadequately predicts nitrate incorporation rates under these situations. For light-limited cases,cell composition data from S. costatum and A. carterae were also used. Chapter 4 considersthe response of NR in more complex cases (i.e. diel cycles of irradiance, different lightspectra, nutrient starvation and growth on ammonium), when enzyme regulatory mechanismsmight become important.NR and the control of nitrate metabolismIf NR is to be useful in predicting rates of nitrate incorporation, it must bequantitatively related to these rates. From a general point of view, in the majority of cellularreactions, total concentration of enzymes is generally positively correlated with steady staterates of metabolism (Acerenza and Kacser 1990). For a particular pathway, however, the onlytrue prediction of in vivo rate is the rate of the limiting enzyme in the pathway (but see alsoChapter 1 for a discussion of the usefulness of non-rate-limiting reactions). It is important torecognize that a given enzyme may only be limiting in a reaction sequence under a particularset of conditions because more than one enzyme may be involved in rate control as conditionschange. A great deal of effort has been expended by biochemists trying to elucidate whichsteps of pathways control observed fluxes. Recent summaries of their conclusions are given byCrabtree and Newsholme (1985), Kacser and Porteous (1987), Hofmeyr and Cornish-Bowden106(1991), and Savageau (1991). Specific treatments in plant systems can be found in Preiss andKosuge (1976), Davies (1977) and Raven (1981). One formalization of these theories,metabolic control theory, suggests that the degree of control of a pathway by a specific step (i)can be represented by a flux control coefficient (Ci), such that:Ci = oYIYovi/viwhere öY/Y is the rate of change of the flux, and öV/V is the rate of change of the enzymeactivity at that step. Thus if the enzyme is a rate-limiting step, C1 should be close to 1.0, i.e.a change in enzyme activity results in a proportional change in flux.Although it is rarely certain that in vitro rates of enzyme activity equal rates in vivo,Newsholme and Crabtree (1986) give examples from the animal literature of how maximalactivities of enzymes can be used to predict fluxes through metabolic pathways in vivo, Intheory, this can be most easily accomplished for enzymes that are rate-limiting. A number ofcharacteristics may give an indication that a given enzyme is a rate-limiting step, althoughthese characteristics do not in themselves constitute proof. Rate-limiting enzymes tend to havecomplex structure (e.g. multiple subunits), and have complex (i.e. allosteric) properties (Preissand Kosuge 1976, Davies 1977). They often catalyze non-equilibrium reactions withsubstantial, negative standard free energy changes (AG°’) (Crabtree and Newsholme 1985).As well, such enzymes are frequently subject to reversible covalent regulation (e.g.phosphorylation /dephosphorylation mechanisms) (Raven 1981). Sometimes, the complexnature of rate-limiting enzymes (e.g. allosteric or covalent regulation) means that simple Vmaxassays may not adequately represent what is happening in situ. As Newsholme and Crabtree(1986) point out, it is difficult to predict a priori whether a given enzyme will be useful inestimating metabolic fluxes. Thus, the validation of enzyme indices relies on empiricaldemonstrations that in vivo fluxes and in vitro enzyme activities correspond, i.e. that C = 1.0.There is a great deal of evidence from higher plants and algae suggesting that NR is infact rate limiting for nitrate incorporation. To begin with, NR fits many of the characteristics107listed above for a rate-limiting enzyme. NR has a complex structure and properties (seeGuererro et a!. 1981, Campbell 1988, Solomonson and Barber 1990, Crawford et al. 1992,and Chapter 2). The reaction NR catalyzes is non-equilibrium (Keq is on the order of 1025 to1040), and G°’is very large and negative (-140 to -230 kJ mo11) (Hewitt et a!. 1976,Solomonson and Barber 1990). There is also evidence that NR is regulated by phosphorylation(Huber et a!. 1992a). Furthermore, there are data showing that nitrate pools accumulatewithin tissues and cells, and that NR activities are usually much less than the activities ofenzymes elsewhere in the pathway of nitrate assimilation. Limitation of nitrate incorporationby NR would suggest that nitrate should accumulate within cells, while the downstreamproduct, nitrite, would seldom occur within cells. It is important to note that analyses of suchinternal pools may be deceptive. Ovadi (1991) points out that the mean distances betweenenzymes in cells is often on the order of the size of a typical tetrameric protein. This degreeof crowding may mean that substrates are “channelled”, or effectively passed from enzyme toenzyme with no apparent change in intermediate pools sizes. It is clear that this phenomenonis not always important (e.g. Stitt 1991), but it should be borne in mind. The build-up ofnitrate pools and the absence of intracellular nitrite have been generally confirmed in higherplants (Campbell 1988). This is also true of marine phytoplankton, although it must bestressed that virtually nothing is known about the compartmentalization of nitrate pools in thesespecies (e.g. cytoplasmic versus vacuolar pools). Nitrate pools of substantial size have beenreported in nitrate-limited chemostat cultures of Skeletonema costatum (Dortch et a!. 1979,Dortch 1982, and Thoresen et al. 1982) , and in nitrate-sufficient cultures of Thalassiosirapseudonana (Dortch et a!. 1984). In some cases, nitrate can represent up to 55% of nitrogenin cells of S. costatum, although this is more normally only a few percent (e.g. in nitrate-sufficient T. pseudonana nitrate pools were 2.8% of particulate nitrogen). Because theconcentrations of these pools were larger that the measured Km for nitrate of NR, it wasconcluded that NR was probably operating at fl2I Vm in these cells (Dortch et a!. 1979).Slawyk and Rodier (1986) also measured nitrate pools in Chaetoceros affinis, and found thatNR did not correlate with internal nitrate concentrations, suggesting that the enzyme was not108likely under substrate control. Sciandria (1991) also found nitrate pools in the dinoflagellateProrocentrum minimum, but only when nitrate was supplied continuously. Collos and Slawyk(1980) failed to find evidence of internal nitrite pools in S. costatum. Further down thepathway of incorporation, ammonium pools do not generally accumulate (Collos and Slawyk1980), although there may be problems with methodology for ammonium pools (Dortch 1982,Thoresen et al. 1982). In higher plants, low NR activities relative to other enzymes in thepathway of nitrate incorporation (e.g. NiR and GS) have also been noted (see Guererro et a!.1981, Crawford et al. 1992, and Chapter 2). For marine phytoplankton, this may also be truefor many species, but there are very few data (see review by Cobs and Slawyk 1980). Thehigh degree of regulation of nitrate reductase, including evidence of translational,transcriptional, covalent and allosteric control, also suggests that it is a control point of nitratemetabolism (see Solomonson and Barber 1990, Crawford et a!. 1992).On the other hand, there is evidence that, in specific cases, NR may not be ratelimiting. Guererro et a!. (1981) point out that this is especially true in more structurallycomplex organisms such as higher plants, where translocation and storage become moreimportant. The transport of nitrate into the cell may also be a control point. Guerrero et al.(1981) suggested that ammonium inhibition of uptake of nitrate likely took place at thetransport step in microalgae, and in some higher plants. Ingemarsson (1987) in Lemna, andWatt et a!. (1992) in Chiamydomonas reinhardtii both present evidence that nitrate uptake islimiting, particularly at low nitrate supply. Sanchez and Heldt (1990) have also suggested thatNR may be the rate-limiting step, but that supply of NADH (i.e. substrate limitation) and notNR concentration may be the limiting factor; maximal NR activity would not correlate wellwith incorporation rates if this was true. Limitation at the nitrite reductase (NiR) step mayalso occur. Seith et a!. (1991) have pointed out that in higher plants NiR is regulated to a veryhigh degree, suggesting that it has a rate-limiting function. Virtually nothing is known aboutthe regulation of NiR in marine phytoplankton. Redinbaugh and Campbell (1991) note thatbecause NiR and NR are co-induced in higher plants and apparently closely co-regulated,109distinguishing which is the rate-controlling step may be difficult. Coregulation is beneficial,since nitrite is toxic within cells and it must therefore be removed quickly. Another option,however, if NiR were rate limiting, would be to remove nitrite by extracellular excretion.Martinez (1991) investigated N-starved cultures of S. costatum and showed that nitriteexcretion into the medium occasionally occurred on resupply and uptake of nitrate, but nitriteexcretion also occurs during light to dark transition (see Collos and Slawyk 1980). Regulationat the ammonium assimilation step may also occur. Seguineau et al. (1989) proposed that GSplayed a key role in nitrate incorporation and nitrogen metabolism in general in Dunaliellaprimolecta, based largely on the very high degree of regulation of the enzyme. Alternatively,it may be that the entire pathway adapts to the prevailing nitrate incorporation rate. Stewartand Rhodes (1977) have shown that NR and GS activities closely parallel each other in higherplants, and a close coordination of NiR and GS has also been demonstrated (Weber et al.1990). From theoretical considerations, Brown (1991) reasoned that having single rate-limiting steps in metabolic pathways was wasteful. Because there are constraints on totalprotein content of cells, enzymes which are in excess of routine requirements should be underselective pressure to decrease in concentration. As a result, it would be expected that controlof a pathway would be distributed throughout the pathway, although this may be complicatedby differential turnover rates of proteins, or other means of regulation (Brown 1991).Light and nitrate limitation of growth rateIt is also important that an enzyme index of nitrate incorporation rate should respondpredictably under different limitations on growth and metabolism. As described in the GeneralIntroduction, light- and nitrate-limited growth are probably two very common situations in themarine environment. These limitations may have different characteristics and require differentculturing techniques to study them.Light-limitation restricts the rate at which photosynthesis can provide the cell withenergy in the forms of ATP and reductant (NAD(P)H), and fixed carbon. There are a varietyof acclimations a cell can make to low irradiance so that it can overcome these limitations110(Richardson et al. 1983, Zevenboom 1986, Falkowski and LaRoche 1991a). These involvechanges in cell composition (e.g. Post et al. 1985, Goldman 1986, Claustre and Gostan 1987)including the pigments and proteins of the photosynthetic apparatus (Prezelin 1981, Richardsonet a!. 1983). Light limitation experiments are usually performed in batch culture and, aspointed out by Rhee (1979), these closed systems represent an ever-changing environment inthat nutrients decline, biomass increases, and waste products build up. However, by keepingcultures in logarithmic growth phase, the effects of these changing conditions can beminimized, so that a steady state is approached (Rhee 1979). A distinction between the effectsof irradiance and the effects of changing growth rate on cell parameters must be made underthese conditions. While, in absolute terms, it is irradiance effects that are being considered,changes in irradiance are reflected in changes in growth rates. For comparisons with nitrogen-limited cases, it is useful to consider growth rate as an independent variable. Therefore, in thepresent study, to simplify the discussion, relationships between cell composition or enzymeactivity and growth rate, not irradiance, will be considered. A good relationship between lightand growth rate has been described for the species considered (see Chapters 1 and 2). As longas regions of photoinhibition are avoided, regressing a variable against light as opposed togrowth rate will only affect the shape of the curves and not the trends in responses (see e.g.Zevenboom 1986). Another approach to investigate cell responses to light is to performtransition experiments in which cells are switched from one light level to another. Thisrequires repeated measurements to monitor changes in cell composition and rates ofmetabolism in cultures following the transition until a new steady state is reached.The situation under nutrient limitation is more complex. Sciandria (1991) draws animportant distinction between “limitation”, the restricted supply of a nutrient, and “starvation”the removal of a nutrient. As will become apparent, these two situations may be verydifferent. A nutrient-limited cell may be able to make a range of acclimations to lownutrients, including changes in composition and photosynthetic parameters (Goldman 1980,Herzig and Falkowski 1989, Lewitus and Caron 1990, Cullen et a!. 1992) but these strategiesmay not be available to a starved cell. According to Sciandria (1991), a limited cell will show111cell quotas for the limiting nutrient less than the maximum quota, while a starved cell willshow quotas near the minimum possible (see also Goldman 1980). However, this may not begenerally true. Harrison et al. (1977) compared three diatom species under either starvation orlimitation for ammonium or silicate. Skeletonema costatum and Chaetoceros debilis bothshowed almost two-fold higher silica quotas under silica starvation versus silica limitation, butThalassiosira gravida showed slightly higher silica quotas under limitation versus starvation.On the other hand, C. debilis and T. gravida had much higher nitrogen quotas when starved ofammonium than when limited by it, while no differences in nitrogen quotas were seen inammonium-starved versus ammonium-limited cultures of S. costatum. Harrison et a!. (1977)speculated that because cell division in starved cells stopped one or two divisions after nutrientexhaustion, any cell composition changes had to occur in this period. In chemostat cultures,cells could divide for more than 10 generations before achieving a steady state; thus they had agreater scope for modifying their cell quotas. In any case, a lack of knowledge of the rangesof cell quotas may make application of such criteria impossible in most cases. Cullen et al.(1992) have compared these situations in terms of their implications for photosyntheticacclimation, and illustrate some of the differences. The case of starvation is distinct and willbe reserved for consideration in Chapter 4. In order to investigate limitation, the chemostat isa convenient tool. Chemostats provide more realistic levels of nutrients and allow precisecontrol of growth (by dilution rate) and biomass (by the concentration of the limiting nutrient)(Rhee 1979, but see also Burmaster 1979, and Di Toro 1980 for more mathematical treatmentsof chemostat properties), but they represent a steady state that is controlled by a single factor.Such a state is virtually never achieved in nature; apparent steady states are usually the resultof a combination of nutrient supply rates, loss terms such as sinking, and trophic interactions(see Rhee 1979 and Eppley 1981). Because of the nature of a chemostat, transitions in nitratelimitation are relatively easily accomplished, but in practice the transient states followingtransitions are more difficult to follow. Each sampling will change culture volume and thusdilution rates of the cultures, and although this can be minimized by growing very largecultures (> 6 L), it may become impractical. Furthermore, the more times the culture must112be sampled after the transition, the greater the problem becomes. For this reason, suchtransitions have been avoided in the present study.There is also an important interaction between light and nitrogen limitation.Photosynthetic acclimations are constrained by nitrogen availability (see Cullen et al. 1992).For each sub-saturating irradiance level, a range of nitrogen-limited growth rates are possible.Thus a culture may be nutrient-sufficient at a given irradiance, but if the irradiance wereincreased, the same nutrient supply rate might be insufficient. Alternatively, a nitrogen-limited culture may be photoinhibited by an irradiance that would normally be tolerated undernitrogen sufficiency. Furthermore, photosynthetic carbon and nitrogen metabolism areinextricably linked (see Turpin 1991).Responses to light and nutrient limitation are species specific. To account forinterspecific differences and the interaction of light and nitrogen, the concept of relativegrowth rate, defined as 14/1Lmax (where Liflax is the maximum growth rate at a givenirradiance) has been developed (Goldman 1980, but see also Tett et a!. 1985). In this study,because a single organism is used and nitrogen limitations are performed at a single irradiance,use of specific growth rate as opposed to relative growth rate is justified.Cell composition and scaling of enzyme activityIn comparing light and nutrient limitation, the issue of cell composition becomesimportant. Different growth rates affect the composition of cells and there is evidence thatthese effects differ between light and nutrient limiting conditions (see Rhee 1979, Goldman1980, Goldman and Mann 1980, Morris 1981, Sakshaug and Andresen 1986, Sakshaug et a!.1991, Laws and Chalup 1990, Thompson et al. 1991). As indicated in Chapter 1, the variableto which enzyme activity is scaled can affect the interpretation of the results. For example,Dortch et al. (1979) grew S. costatum at two nitrate-limited growth rates (0.8 and 1.6 d1) andcompared NR activities. When NR activity was scaled to cell volume, activity was higher inthe lower growth rate culture, but when NR activity was scaled to chlorophyll a, the highgrowth rate culture showed higher enzyme activity. Scaling NR activity to cell nitrogen or113protein quotas resulted in no difference in NR activity between the cultures. For culture work,this is not a problem, since NR can be compared directly with measured or calculated rates(see Chapter 2). In the field, however, this is not so, and clearly the issue of a scaling factormust be resolved.MATERIAL AN!) METHODSGeneral culture conditionsCultures were obtained from the North Eastern Pacific Culture Collection (NEPCC)and maintained on artificial medium (ESAW) at 17.5 °C under continuous light, as previouslydescribed (Chapters 1 and 2).Steady-state light-limited experimentsFor the cultures of Thalassiosira pseudonana, Skeletonema costatum, and Amphidiniumcarterae used in the growth rate experiments described in Chapter 2, specific growth rates, cellvolumes, and carbon (C), nitrogen (N) and protein quotas were determined as previouslydescribed (Chapter 1). In addition, for S. costatum and A. carterae, chlorophyll a (chi a)quotas were measured by fluorometric methods after extraction in 90% acetone (Parsons et a!.1984a). C:N molar ratios and C:chl a weight ratios were also calculated. These constituentsand ratios were plotted against specific growth rate, and analyzed by linear regression analysesusing SYSTAT MGLH routines (Wilkinson 1990).For T. pseudonana, NR activity data were taken from Chapter 2, plus six additionalcultures grown and measured as before. NR activities were plotted against specific growthrates or nitrate incorporation rates calculated as the product of cell nitrogen quota and cellspecific growth rates.114Light transition experimentsTransition experiments were conducted as described in Chapter 1. Six 1 L cultures ofT. pseudonana were established, three at low light (15 mol quanta m2s4) and three at highlight (90 mol quanta m2s1). These cultures were acclimated for a minimum of 8generations, sampled at 0 h, and 24 h, then transposed (i.e. high to low irradiance, H—*L, orlow to high irradiance, L—>H) and sampled again at approximately 24 h intervals for threemore days. At each sampling, cell volumes and numbers, cell carbon, nitrogen and proteinquotas, and NR activities were determined as before and C: N ratios calculated (Chapters 1, 2).Specific growth rates () were calculated from changes in cell numbers between samplings.Cultures were maintained in logarithmic growth phase, diluting with fresh medium asnecessary. Nitrate incorporation rates were also determined as before (Chapter 2, except thatcell numbers instead of fluorescence were used to calculate growth rates). Within each timeinterval, differences in cell constituents, growth rates, NR activities and nitrate incorporationrates were tested using paired t-tests, as before (Chapter 1).Steady-state nitrate-limited experimentsFor three separate experiments with 7’. pseudonana, six nitrate limited chemostats wereset up in 1 L flasks that were mixed with Teflon-coated stir bars and magnetic stirrers.Chemostats were provided with an inlet for medium, an overflow for excess culture and asampling port. Chemostats were run by positive pressure; new medium was pumped into theculture using a multi-channel peristaltic pump (Manostat model 1OA) and excess culture wasforced out the outflow by pressure. Under these conditions, once steady state is achieved,growth rate is a function of the rate of new medium addition (dilution rate), and total biomassin the culture is set by the concentration of the limiting nutrient in the added medium (Rhee1979). Thus, growth rate can be controlled by adjusting the rate of the peristaltic pump.Medium (ESAW) was prepared as before, except that nitrate concentration was lowered to 40M, and bicarbonate additions were doubled (to 4 mM) to prevent possible carbon limitation115in the cultures. In each experiment, medium for six cultures was provided from a common 20L reservoir. Cultures were judged to have reached steady state when the cell fluorescence andthe concentration of phosphate (a non-limiting nutrient) of the outflow remained constant overtwo days. At this point, cultures were sampled for cell volume and numbers, cell carbon,nitrogen, chi a and protein quotas and NR activities, as described previously. C:N ratios,C:chl a ratios, and the rate of nitrate incorporation were calculated. Cell constituents wereplotted against specific growth rates (dilution rates/culture volume) and analyzed by linearregression analyses, as before. NR activities were plotted against specific growth rates, orcalculated nitrate incorporation rates.Scaling of NR activityFor T. pseudonana light- and nitrate-limited experiments, NR activity was scaled tocell volume, or cell carbon, nitrogen, chi a or protein quotas. These activities were plottedagainst specific growth rates and analyzed by linear regression analyses. In each case,regressions were performed for light-limited cultures alone, nitrate-limited cultures alone, andboth sets of cultures together. Regression slopes and intercepts were compared following Steeland Tome (1980).RESULTSSteady-state light-limited experimentsResponses of cell constituents to differences in growth rate varied with the speciesexamined. Composition data versus growth rate in T. pseudomzna is presented in Fig. 3.1. Inone replicate experiment of three cultures which used a separate batch of seawater medium, thecultures showed unusual behaviour; cells became elongated and considerably greater involume. This appeared to correspond to limitation by selenium, as previously described byPrice et al. (1987). These cultures were excluded from regression analyses, but are shown asopen symbols in Fig. 3.1. A summary of regression results is presented in Table 3.1.116( I I I60 a A15 B‘S V10- V V -5V-Zo.000..c1-. •. . -‘I 6-2- -AAE10- A AAA A -A A5- A A A -A I I0.0 0.5 1.0 1.5 2.0growth rate (d’)Figure 3.1. Cell composition versus light-limited specific growth rate forThalassiosira pseudonana. A) cell volume, B) cell carbon quota, C) cellnitrogen quota, D) cell protein quota, and E) cell C:N ratio. Each data pointrepresents the mean of duplicate determinations from a single culture.Open symbols represent three cultures where selenium limitation may haveoccurred. Lines represent least squares regressions. Parameters are given inTable 3.1.Table 3.1. First-order linear regression parameters for composition versus growth rate relationships in light-limited cultures of various marine phytoplankton. P-values represent the probability that the slope isequal to zero.SPECIES PARAMETER DATA SLOPE INTERCEPT r2 P-valueSETSThalassiosira cell volume all 5.48 22.7 0.67 < 0.001pseudonanacarbon all 1.30 5.06 0.41 < 0.02nitrogen all -- -- -- > 0.4protein all -- -- -- > 0.4C:N ratio all -- -- -- > 0.4Skeletonema costatum cell volume 1 91.9 80.9 0.66 <0.052 -- >0.3all -- -- -- >0.1carbon all -- -- -- > 0.2nitrogen 1 -- -- -- > 0.32 1.52 0.98 0.86 <0.01all — -- — >0.3chl a 1 -2.88 0.95 0.77 < 0.022 -0.31 0.28 0.69 < 0.04all -0.72 0.51 0.46 < 0.02protein all -- > 0.2C:N ratio all — > 0.3C:chl a ratio all -- > 0.3Amphidinium carterae cell volume 1 184 317 0.84 <0.012 218 362 0.34 <0.05all 252 325 0.48 < 0.001carbon 1 — — — >0.32 -- -- -- >0.1all -- -- -- >0.9nitrogen I -- -- -- > 0.52 -- -- -- >0.9all -- -- -- >0.3chla 1 -- -- -- >0.12 -- -- -- >0.4all -3.38 5.07 0.23 < 0.04protein 1 -- -- -- > 0.12 -- -- -- >0.7all -- -- -- >0.06C:Nratio 1 -- -- — >0.42 4.48 3.81 0.68 <0.001all 2.71 5.00 0.79 < 0.004C:chl a ratio 1 88.2 27.6 0.73 <0.042 -- -- -- >0.07all 80.5 23.8 0.33 < 0.01117118Significant and positive relationships were seen only for cell volume (Fig. 3.1 A, Table 3.1)and carbon (Fig. 3.1 B, Table 3.1), indicating that cells growing faster have higher volumesand carbon contents. Whereas in T. pseudonana replicate experiments gave similar results, inthe other two species there was high inter-experimental variation. For this reason,experiments were usually considered separately. A relationship between carbon quota andgrowth rate was not seen in S. costatum (Fig. 3.2 B, Table 3.1) or in A. carterae (Fig. 3.3 B,Table 3.1), but there was a significant increase in cell volume in one experiment with S.costatum, and in both A. carterae experiments (Fig. 3.2 A, Fig. 3.3 A, Table 3.1). In one S.costatum experiment, there was a significant and positive relationship between cell nitrogenquota and growth rate (Fig. 3.2 C, Table 3.1). For both S. costatum and A. carterae, chl asignificantly decreased as growth rate increased (Fig. 3.2 E, Fig. 3.3 E, Table 3.1). As well,for A. carterae, both C:N and C:chl a ratios increased with growth rates (Fig. 3.3 F, G, Table3.1), although such a phenomenon was not seen in the other species.Data in Figure 3.4 is repeated from Chapter 2 in a slightly different form. In T.pseudonana, NR activity was positively related to growth rate (NR = 7.46 + 41.6 (± 7.86), r2 = 0.67, P < 0.002) (Fig. 3.4 A). As for the composition data, the cultures that werepotentially selenium-limited were left out of the analysis. The relationship between NRactivity and the calculated rate of nitrate incorporation was not different from the 1:1relationship (NR = 4.49 +0.98 (±0.03) nitrate incorporation rate, r2 = 0.98, P < 0.001;Fig. 3.4 B). In this case, selenium-limited cultures were included, as no significantdifferences from other cultures were seen.Light transition experimentsComposition data from the transition experiments did not follow all of the trends seenin the steady-state experiments. Cell volumes were no different between H—*L and L—Hcultures at any point during the experiment, although there was a non-significant trend towardL—*H cells becoming larger, and H—>L cells were significantly smaller after the transition(Fig. 3.5 A, P < 0.01). Some significant differences in carbon were seen. In agreement with119‘1’: _t___.. A• 40- 000 0 -I-:i: B-. 4_v-• D12- -Eg10 *****1EG:-0- -I IC)0.0 0.2 0.4 0.6 0.8 1.0growth rate (d’)Figure 3.2. Cell composition versus light-limited specific growth rate forSkeletonema costatum. A) cell volume, B) cell carbon quota, C) cellnitrogen quota, D) cell protein quota, E) cell chlorophyll a quota, F)cell C:N ratio, and G) cell C:chlorophyll a ratio. Where there are 2 types ofsymbols, they represent two different experiments. Points represent themeans of duplicate determinations from single cultures. Lines representleast squares regressions. Parameters are given in Table 3.1.120200- I -‘450 - I B -300 • I-i -150- r-i0 L -Eu p -!60- vi Iv -V° 0 V20s .ioo-0 -= 0- •, ‘ . iØ -C I I I I8-A I E4o- -+?150- G0. rate (d1)Figure 3.3. Cell composition versus light-limited specific growth rate forAmphidinium carterae. A) Cell volume, B) cell carbon quota, C) cellnitrogen quota, D) cell protein quota, E) cell chlorophyll a quota, F)- cell C:N ratio, and G) cell C:chlorophyll a ratio. Where there are 2 types ofsymbols, they represent two different experiments. Points represent themeans of duplicate determinations from single cultures. Lines representleast squares regressions. Parameters are given in Table 3.1.121—0I——1)0.—C.)zC—C.)E0E..-C.)zFigure 3.4. Nitrate reductase activity versus light-limited growth rate inThalassiosira pseudonana. A) NR versus specific growth rate, and B) NRversus calculated rate of nitrate incorporation. Each point reresents the meanNR activity in a single culture. Error bars represent standard errors of the meanof two NR assays. Solid lines represent least squares regressions. Dashed linerepresents the 1:1 relationship. Open symbols represent cultures where seleniumlimitation may have occurred.0.0 0.5 1.0—1growth rate (d )1.52001501005002001501005000 50 100calculated NO3 incorporation (mol2.0150 2001 1 12mm ceif ) x 10122EC)0)o 871.600.8‘ 0—o 10—o 80.4c-)time (d)Figure 3.5. Changes in cell composition on transition from low light to highlight ( • ) or high light to low light ( 0 ) in Thalassiosira pseudonana.A) Cell volume, B) cell carbon quota, C) cell nitrogen quota, D) cell C:N ratio.Transitions were made at the point indicated by the arrow. Each pointrepresents the mean and standard error of three separate cultures. Asterisksindicate significant differences (P < 0.05).0 2 3 4123steady-state experiments, H—>L cells had higher carbon quotas than L—+H cells before thetransition, but changes in carbon quotas followed no pattern during the transition. In contrastto what was found in steady-state, L—’H cells were significantly higher in nitrogen than H—+Lcells, before the transition. By the end of the experiment, the L—H cells had lower nitrogenquotas and H—>L cells had significantly higher nitrogen quotas (Fig. 3.5 C). Largely becauseof differences in nitrogen, a similar pattern was seen in the C:N ratio (Fig. 3.5 D).Figure 3.6 A shows the clear transition in growth rates when cultures were transposed.There was some indication that cells moved from low to high light actually increased theirgrowth rate above that which the high light-grown cells had shown, but this differencedisappeared by the end of the experiment. In terms of calculated rates of nitrate incorporation,a similar trend was seen (Fig. 3.6 B), and throughout the transition experiment, the NRactivity matched the nitrate incorporation rate almost perfectly.Steady-state nitrate-limited experimentsChemostats stabilized within 5-6 days in all experiments. Examination of nutrientconcentrations showed no nitrate in the outflow, except in the case of the two highest dilutionrates in each experiment where nitrate was between 0.4 and 2 M, and nitrite wasapproximately 0.2 M.Relationships between growth rate and cell composition were quite different duringnitrate limitation than for steady-state light-limited experiments. Significant negativerelationships were found for cell volume and carbon versus growth rate (Fig. 3.7 A, B, Table3.2) in contrast to the positive relationships seen earlier. Once again, nitrogen and proteinshowed no significant relationships with growth rate. C:N ratios significantly declined asgrowth rate increased (Fig. 3.7 F, Table 3.2). Chl a content of cells decreased as growth ratesincreased (Fig. 3.7 D, Table 3.2), which, combined with carbon decreases, lead to significantdecreases in the C:N ratio with increasing growth rate (Fig. 3.7 G, Table 3.2).NR activities in chemostat cultures were positively related to growth rate (NR = 22.7+ 23.9 (± 6.95) , r2 = 0.43, P < 0.004) (Fig. 3.8 A, and positively related to calculatedI I124A* * * **1.6“—‘ 1.2a)‘ 0.8010.4a)Io c 80—‘. .do 0., 0 —> C) ).—C)—C)c‘ .E 40Z-EC)cIC)time (d)Figure 3.6. Effect of light transitions on: A) growth rate, and B)nitrate reductase activity ( • , 0 ), or calculated nitrate incorporation rates( • , LI ) in Thalassiosira pseudonana. Transitions were made at the pointindicated by the arrow. Each point represents the mean and standard errorof measurements made in three separate cultures. Asterisks indicate significantdifferences (P < 0.05). Open symbols represent high to low light transitions.Closed symbols represent low to high light transitions.* *I I01 3 4125E I I60-r’ E 4020B:12 V0 84ba 0Cc c 3C)o 2-00 AD—.) 4-.0.0_I I E•o 2 • IE j I16128..30015020C.)0.0 0.5 1.0 1.5 2.0growth rate (d’)Figure 3.7. Changes in composition with growth rate in nitrate-limitedchemostat cultures of Thalassiosira pseudonana. A) cell volume, B)- cell carbon quota, C) cell nitrogen quota, D) cell chlorophyll a quota,E) cell protein quota, F) cell C:N ratio, and G) cell C:chlorophyll aratio. Each point represents the mean of two determinationsfrom a single culture. Lines represent least squares regressions.Parameters are given in Table 3.2.126Table 3.2. First-order linear regression parameters for relationships between cell composition and growth rate innitrate-limited chemostat cultures of Thalossiosira pseudonana. P-values represent the probability thatthe slope is equal to zero.PARAMETER SLOPE 1NTERCEP r2 P-valuecell volume -10.8 40.2 0.46 < 0.002carbon -5.80 11.9 0.41 < 0.005nitrogen > 0.8chl a 0.049 0.033 0.73 < 0.001protein > 0.7C:Nratio -4.84 11.8 0.37 <0.008C:chl a ratio -215 299 0.54 < 0.001127Figure 3.8. Relationship between nitrate reductase activity and A) specific growth- rate, or B) calculated rate of nitrate incorporation, in nitrate-limited chemostatsof Thalassiosira pseudonana. Each point represents the mean and standard error ofduplicate NR assays from a single chemostat. Solid lines represent leastsquares regression fits to the data. Dashed line represents the 1:1 relationship.Parameters are given in the text.0.0 0.5100C.1I—C)5025N—C,100) 75C)1.0growth rate (d1)1.5 2.00 25 50 75 100- ... -1 -1 12calculated NO3 incorporation rate (mo1 mm cell ) x 10128rates of nitrate incorporation (NR = 24.2 + 0.485 (± 0.125) nitrate incorporation rate, r2 =0.48, P < 0.002) (Fig. 3.8 B). However, in contrast to light-limited experiments, the slopeof the relationship was less than 1:1; NR activity was significantly higher than the calculatednitrate incorporation rate at low growth rates.Scaling of NR activityNR activity scaled to cell volume (Fig. 3.9 A), carbon quota (Fig. 3.9 B), nitrogenquota (Fig. 3.9 C), or protein (Fig. 3.9 D) were all significantly related to growth rate (Table3.3), but this was not true for NR activity scaled to chi a (Fig. 3.9 D)). Where regressionswere significant, no significant differences were found between relationships for light-limitedversus nitrate-limited cultures (P > 0.5 in all cases).DISCUSSIONVariation in cell composition with growth rateLight-limited culturesSince few of the cultures were light-saturated (see Chapter 2, Fig. 2.13), therelationship between composition and growth rate and between composition and irradianceshould be very similar (this would not be true in light-saturated cultures, since irradiance couldincrease without an increase in growth rate). There was variability between species and withintrials, and not all data were available for all species, but some trends emerged. Table 3.4gives a summary of previous studies where composition and growth rate (or irradiance) havebeen related. It is important to note, however, that these data combine continuous light withlight:dark-grown cultures. As Sakshaug and Andresen (1986) have demonstrated, dielperiodicity has profound effects on trends in cell composition and must be cautiouslyinterpreted. The state of the culture in some studies in the literature have been poorly defined.Lewitus and Caron (1990) demonstrated that the trends in senescent cultures of Pyrenomonas129C2L)b100z zo 150C-)-r’ o15001000500oCZ60z200.0 1.0 2.0growth rate (d’)Figure 3.9. Nitrate reductase activity scaled to different parametersversus growth rate of Thalassiosira pseudonana in light-limited batch( 0 ) or nitrate-limited chemostat ( • ) cultures. A) Per cell volumeB) per g carbon, C) per g nitrogen, D) per g chlorophyll a and E) perg protein. Each point represents the mean of two assays from asingle culture. Lines represent least squares regression fits. Parametersare given in Table 3.3.432100.5 1.5130Table 3.3. Comparison of first-order linear regression parameters for nitrate reductase activity scaled to differentparameters versus growth rate in Thalassiosira pseudonana in light-limited batch cultures (L), nitrate-limited chemostats (N) or both types of cultures together (both). P-values represent the probability thatthe slope is equal to zero.SCALING DATA SETS SLOPE INTERCEPT P-valuePARAMETERcell volume L 1.25 0.81 0.48 < 0.006N 1.12 0.50 0.75 < 0.001both 1.33 0.53 0.63 < 0.001carbon L 6.34 2.96 0.44 < 0.02N 6.57 1.91 0.59 < 0.001both 6.78 2.12 0.56 < 0.001nitrogen L 49.6 5.18 0.69 < 0.001N 23.9 30.3 0.30 < 0.02both 34.0 22.0 0.49 < 0.001chla N >0.6protein L 13.9 11.0 0.32 < 0.05N 7.33 10.2 0.34 < 0.02both 12.3 9.51 0.37 < 0.001Table3.4.Changesincompositionwithincreasinggrowthrate(irradianceornutrientsupply) forvariousspeciesunder lightornutrientlimitation.D=hoursofdaylight(i.e.24meanscontinuouslight),NSOURCE=thenitrogensourceused(N03=nitrate,NH4=ammonium),VOL=cellvolume,C=carbonquota,N=nitrogenquota,C:N=carbon:nitrogenratio,CHL=chlorophyll aquota,C:CHL=carbon:chlorophyllaratio,CHO=carbohydratequota,PRO=proteinquota.Responsesaredefinedasincreases(+),decreases (-),nochange(nc),orcomplexbehavior(c).SPECIESDNVOLCNC:NCUEC:CHLCHOPROLIPIDREFERENCE(h)SOURCELight-LimitedChae:oceroscalcitrans24N03+noHarrisonetat.1990Chaetocerosfurcellatus12N03++-Sakshaugetat.1991Ditylumbrightwetlii24N03++EppleyandSloan1966Phaeodaciylumtricornutum24N03++nc-+Geider etat.1985Thatassiosiranordenskioetdii12N03++-Sakshaugetat.1991Thatassiosira pseudonana24N03+noncHarrisonetat.1990Thalassiosirapseudonana24N03++ncncncpresent studyThalassiosirarotula12N03+----Rivkin1989Thatassiosiraweisflogii12N03+++-+LawsandBannister1980Thala.ssiosiraweisflogii24N03+Lawsetat.1983Thalassiosiraweisflogii24N03-+noPostetal.1985Skeletonemacostatum24N03--no-cSaicshaugandAndresen1986Skeletonemacostatum24N03+nc+nc-ncncpresent studyDunaliellatertiolecta12N03no-----Rivkin1989Hymenomo,ws elongata24N03++no+-noClaustreandGostan1987Isochtysisgalbana24N03no-noClaustreandGostan1987Isochrysisgatbana24N03noncHarrisonetat.1990Paviovatutheri24N03--ChalupandLaws1990Pyrenomonassatina24N03+++noncLewitusandCaron1990Amphidiniumcarterae24N03+ncnc+-+ncpresent studyTable3.4(continued)SPECIESDNVOLCNC:NCHLC:CHLCHOPROLIPIDREFERENCE(h)SOURCENitrogen-LimiKed24Chaetoceros debitis24NH4+++-+-Harrisonetat.1977Dirylumbrighrwellii24N03++EppleyandSloan1966Phaeodaczylumtncornutum24N03+++Marsot etat.1991Skeletonemacostatum24NH4+++*+-Harrisonetal.1977Thalassiosiraatleni12N03nc++LawsandWong1978Thalassiosiragravida24NH4+++-+-Harrisonetat.1977Thalassiosirapseudonana24N03+-+-+-CaperonandMeyer1972Thatasciosirapseudonana12N03++-LawsandCaperon1976Thalassiosirapseudonana24N03-Goldmaneta!.1979Thalassiosimpseudonana24N03--nc-++ncpresentstudyThalassiosiraoceanica12N03c++-+-EppleyandRenger1974Thalassiosiraweisflogii12N03nc+-+-LawsandBannister1980Thalassiosiraweisflogii24N03-Lawsetat.1983Thatassiosiraweisflogii12NH4nc+-+LawsandBannister1980Dunalieltarertiotecta12N03++-LawsandCaperon1976Dunatiellatertiolecta12N03c++LawsandWong1978Dunaliellaterriolecra24N03-Goldmanetat.1979Dunaliellatertiolecta24N03++-+-CaperonandMeyer1972Isochrysisgatbana12NH4--Davidsonetat.1989Isochrysisgatbana24N03nc-+-+-HerzigandFalkowski1989Paviovalutheri12N03nc++LawsandWong1978Pavtovalutheri12N03nc+-++deMadriagaandJoint1992Pavtovalutheri24N03+ChalupandLaws1990Paviovatutheri12N03++-LawsandCaperon1976133sauna often differed from those found during logarithmic growth, and thus, this may beanother source of variation in the reported results. The composition data can also be comparedto those found in Chapter 1, since experimental conditions were identical.Volume increased with growth rate in most cases, as was seen in Chapter 1. This hasbeen found for several other species (Table 3.4). Thompson et a!. (1991) came to a similarconclusion in a multi-species study (including T. pseudonana), and an extensive review of theliterature. Thompson et a!. (1991) also noted that the relationship of cell volume to irradianceand growth probably becomes species-specific above saturating irradiance, since the onset ofphotoinhibition is quite species-specific. Cell volumes may be under the control of internalions (Riisgard et a!. 1980), low molecular weight metabolites (Ahmad and Hellebust 1985b) orcarbon content (Claustre and Gostan 1987), all of which may increase with photosynthetic rateat higher irradiances. The size decreases observed in the present study are very similar tothose reported by Thompson et at. (1991), who measured volume changes on a scale of hours.This is surprising since diatom cells are bounded by a rigid silica frustule. There was evidencefrom one species, Ditylum brightwellii, that the size difference was due to addition ofintercalary bands, resulting in a longer cell of similar width. Hitchcock (1983) found that atsaturating irradiance and continuous light, the relationships between cell volume and othercomponents was quite strongly conserved between species; thus volume may follow othertrends. Thompson et at. (1991) speculated about the ecological advantages of smaller cellsize, although on the contrary, larger cells may have a superior ability to absorb light (the so-called “package effect”, see Raven 1986), and a proportionally lower respiratory rate (Taguchi1976, Geider et a!. 1986). Thus there may be equal physiological advantages to increasingcell size at low irradiance. Larger cells have increased capacities to store carbon (which couldlater be catabolized for energy) and this may be an advantage for cells that inhabitenvironments with fluctuating irradiance (see discussion in Thompson et a!. 1991).It is important to recognize that in diatoms there is a size change associated with sexualcycles; since cells decrease in width due to repeated asexual division, cells undergo sexualreproduction which restores their cell width (see Werner 1971a, 1971b). Over the course of134experiments in this thesis, such cycles in cell size were not noted for T. pseudonana, butbetween different experiments, maximum cell volume varied from 25 to 60 m3. Costello andChisholm (1981) have shown cyclic trends in cell volume in the diatom Thalassiosiraweisfloggii, but these volume changes do not appear to correspond to cycles of sexualreproduction. Armbrust and Chisholm (1992) also found changes in cell volume with growthrate in T. weisfloggii, but noted that these changes occurred only in maximal, light-saturatedgrowth rates. There was also a high degree of variability between clonal cultures. As noted inChapter 1, it is also important to consider that the absolute volumes determined by the CoulterCounter are probably not correct. Changes in volume may still be real, but they might alsoresult from cell shape changes.An increase in carbon with growth rate was seen in the present study for only onespecies, T. pseudonana, and this was not seen in the experiments in Chapter 1. Cell volumesand carbon quotas in Chapter 1 were also greater than those in the present chapter; sincegrowth conditions were identical, the reason for these differences is not known. A generalincrease in carbon quota with growth rate has been seen in many studies (Table 3.4), althoughthere are exceptions. Thompson et al. (1991) recently reviewed the literature and reportedsimilar results for a wide range of species. They noted that cell carbon versus growth raterelationships were more variable between species than those found for cell volume versusgrowth rate. As discussed in the preceding paragraph, the differences in carbon may drive thedifferences in volume, or may be a consequence of them. The increase in carbon has beenattributed to increases in carbohydrate or lipid (e.g. Claustre and Gostan 1987), but with theexception of Post et al. (1985) this is not borne out by the few available studies where eitherno change, or decreases in lipid and carbohydrate with growth rate were noted (Table 3.4).For some of these studies where measurements were made on a light:dark cycle, thisdifference may explain the discrepancy. Several authors including Laws and Caperon (1976),Smith and Geider (1985), and (Geider 1992) have noted that at lower growth rates, respirationis a larger fraction of carbon fixed. If this was true, it would suggest that carbon quotasshould decrease at lower growth rate, as observed.135Nitrogen quotas were generally more variable than carbon quotas in all speciesexamined. Changes in nitrogen quotas were significantly related to growth rate in only onecase; N quotas increased with growth rate in one experiment with Skeletonema costatum.There are a range of responses of nitrogen to growth rate reported in the literature (Table 3.4),with no change or an increase commonly found. The decrease in chi a seen with increasinggrowth rate would probably not produce a change in nitrogen, since very little nitrogen isaccounted for in chlorophyll (Dortch et al. 1984), although there may be more nitrogenassociated with light harvesting complexes (Prezelin 1981). Cell protein quota was not relatedto growth rate in the present study, which is in agreement with the data for nitrogen quotaversus growth rate. Protein quota might be expected to vary with growth rate in the same wayas nitrogen quota, since a substantial proportion of cell nitrogen is found in protein (seeAppendix A). However, Morris (1981) in a review of protein synthesis showed that proteinsynthesis and nitrogen metabolism are often uncoupled.The C:N ratio is a function of both carbon and nitrogen quotas and has been usedfrequently as an index of cell nutrient status. As summarized in Goldman (1980), the observedRedfield ratio, representing idealized elemental ratios in balanced growth, predict a C:N ratioof 6.6:1 (mol:mol). Given a 50% minimum protein content in a cell, it is unlikely that C:Nratios could fall much below 3.7 (Goldman 1980). Given that increases in carbon quota butnot in nitrogen quota were observed in the present study, an increase in the ratio would bepredicted, but this was seen only for A. carterae. According to a survey of the literature, therelationship between C:N ratio and growth rate varies between studies (Table 3.4). Turpin(1991) argued that C:N should correlate with irradiance (and therefore growth rate), becauseprotein synthesis mechanisms saturate at lower irradiances than photosynthesis, and Nassimilation outcompetes CO2 fixation for reducing power. This may be what is happening inA. carterae cultures.Variations in chi a quota with growth rate were similar for the species considered in thepresent study (note that chi a data were not available for light-limited T. pseudonana cultures).As would be anticipated, cells growing at high irradiance exhibited decreases in their chi a136quotas (see Richardson et al. 1983). This is consistent with the majority of studies (see Table3.4), although there are exceptions. It has been noted that approximately half the decreases inpigment on transition from low to high irradiance is due to dilution of chi a due to celldivision, while the other half is probably due to chi a degradation (Falkowski and La Roche199 la). The C: chi a ratio provides another way to look at these data. There was no trend forS. costatuin data, but this may have been due to a limited number of data points. For A.carterae, there was a significant increase in C:chl a with growth rate. The data provided byother studies agree with this trend (Table 3.4). The absolute magnitude of the ratio variesbetween 16 to 285 (Banse 1977, Rieman et al. 1989) in most field samples, and the results ofthe present study fall well within this range.Some of the trends noted for steady state cultures were found for transitionexperiments. Claustre and Gostan (1987) pointed out that the two situations are not identicaland differences should be expected. For example, Thompson et al. (1991) showed that carbonquota increased with growth rate, but this was not true in a transition experiment. However,L—*H cells did increase their carbon quotas and H—*L cells decreased carbon quotas ontransitions. As shown previously (Chapter 1), cell volume responded as predicted by steadystate experiments in the transition, although this trend was not statistically significant in thepresent study. The response of carbon was not seen in either data set, although in the presentexperiment the cultures began the transition with the high light cultures having significantlygreater carbon quotas. There was a significant change in cell nitrogen quotas which was notreflected in steady state nitrogen data. This would be consistent with a decrease in nitrogencontent with growth rate increases and might be mediated by decreases in pigment/proteincomplexes. It is possible that in a transition these differences are magnified. C:N ratiosresponded as would be expected from the A. carterae data set, indicating an increase in C:Nratio with growth rate. This was largely driven by the changes in nitrogen quotas.137Nitrate-limited culturesThe responses of cell constituents to changes in growth rates under nitrate limitationdiffered markedly from those found for light limitation.Cell volume decreased as growth rate increased, a finding which is not supported in theliterature. Cultures were examined microscopically for evidence of cell clumping, and CoulterCounter size distributions were studied for evidence of increasing size spread which mightindicate clumping. Based on these examinations, no obvious clumping of cells was seen, but arelatively low percentage of clumped cells might still have caused an apparent increase in cellvolumes, and a decrease in cell numbers, which in turn would have increased calculated cellquotas. In the marine alga Heterosigma akashiwo, Thompson et al. (1991) showed thatnutrient limitation over-rode the effects of irradiance on cell volume; ammonium-limitedcultures showed no change in cell volume with growth rate increases. With limitation by iron,increases in cell volume have also been found in a marine dinoflagellate (Doucette andHarrison 1990).Carbon quotas decreased as growth rate increased. A similar trend towards increasingcarbon quota with increasing nutrient limitation has been noted by Rhee (1980), but theliterature also documents many variations in the relationship (Table 3.4). There are likelyinterspecific differences in these relationships (Rivkin 1989). As well, Laws and Caperon(1976) point out that some of this variation may come about because of differences inmethodology; some chemostat studies have been run as cyclostats with a light:dark cycle, andthe effects of this on composition have been documented. Furthermore, as reviewed in Darley(1977), the nitrogen source makes a significant difference; cells grown in nitrate have higherchl a, phosphorus and ATP quotas than those grown in ammonium (see also Zevenboom 1986,Thompson et al. 1989). In theory, nitrogen-limited cells would have sufficient energy tocontinue to fix carbon, but could not incorporate nitrogen, and thus, an increase in carbon orlipid would be expected. Such an increase is not well supported in the literature (Table 3.4).Cullen et a!. (1992) suggest that carbon storage products in nitrogen-limited cells will increase138until a steady state is reached. Thereafter, all components will increase at the sameexponential rate.Surprisingly, the nitrogen quota of cells was apparently independent of growth rate(and therefore nitrogen limitation). This is at odds with the literature which unanimouslyagrees that nitrogen quotas increase with growth rate (Table 3.4, Goldman and Mann 1980,Zevenboom 1986, Turpin 1991). It may be that the significant differences in cell volumefound in the present study play a factor. Certainly, if there were clumping of cells this wouldhave produced such results; the most nitrogen-limited cultures would appear to have largercells with more nitrogen; however, as previously noted, clumping could not be confirmed.Nitrogen per unit cell volume did significantly increase with growth rate. A lack ofrelationship between protein and growth rate was also seen. Given the expectation thatnitrogen quotas increase with growth rates (Turpin 1991), this is also at odds with the findingsof the present study. A clumping problem might again be invoked.In contrast, C:N ratios fell with increasing growth rate (i.e. as cells became less N-limited), in good agreement with the literature (Table 3.4, Rhee 1979, Goldman 1980, Morris1981, Laws and Chalup 1990). Marsot et al. (1991) did report a positive relationship betweencell C:N ratio and growth rate, but their cultures were also at markedly different densities,suggesting that true steady-states may not have been achieved. This decrease in C:N ratio isnormally attributed to a decrease in nitrogen quota, which was not seen. However, ifclumping did occur, the C:N ratio would not be affected by it. Goldman et a!. (1979) foundthat C:N ratios approached the Redfield ratio (6.6) only under nutrient sufficiency, a resultreflected in the chemostat data in the present study.Chl a quotas in nitrate-limited cultures increased with growth rate, as indicated in themajority of studies (Table 3.4, Turpin 1991). Herzig and Falkowski (1989) have reviewed theprocesses of pigment reduction under nitrogen limitation. If cell clumping had occurred, itmight have been anticipated that cell chl a would rise at low growth rate. However, if thisdid occur, it may only have decreased the slope of the chl a versus growth rate relationship.C:chl a ratios decreased with growth rates, again agreeing well with the majority of studies139(Table 3.4, Goldman 1980), and the model of Laws et al. (1985). Sakshaug et a!. (1991)found that this trend persisted regardless of the light level used, or the daylength, but notedthat the precise relationship changed.Variation in NR activity with growth rateLight-limited culturesNR activity in the present study was positively correlated with growth rate, and verystrongly and quantitatively related to nitrogen incorporation rate. Thus, the NR activity ismuch more strongly related to growth than to factors such as cell size or composition. Thereare virtually no systematic laboratory studies of variation in NR activity with growth rate;most authors have chosen to investigate simple presence or absence of NR activity (e.g.Everest et al. 1986), NR activity in field situations (e.g. Packard et a!. 1971, Blasco et al.1984) or NR activity in cultures in transient states (e.g. Dortch et a!. 1979, Clayton 1986,Smith et a!. 1992). Data from studies in which nitrate incorporation rates and NR activitywere compared are summarized in Table 3.5. It is evident that few other studies have foundstrong relationships between NR and nitrogen incorporation rates. In fact, only Moms andSyrett (1965) and Kristiansen (1987) found NR activity sufficient to account for observednitrate incorporation rates, and relatively few studies have compared cultures at different light-limited growth rates. The good correlations found in the present study are likely the result ofimprovement to the NR assay (see Chapter 2).A good agreement between NR activity and growth rate under light limitation might nothave been anticipated. If cultures were light-limited, this would be evident in a limitation ofenergy, and a decreased ability to fix carbon, but this would not necessarily affect nitrogenuptake or incorporation. However, there is extensive evidence that nitrogen and carbonmetabolism are very tightly coupled (Sawhney et a!. 1978, Bassham et al. 1981, Geider 1992).As Turpin (1991) points out, because cell protein contents are high in algae, over 50% of allalgal carbon is integrally coupled with nitrogen metabolism. Pace et a!. (1990) and Kaiser andTable3.5.Relationshipofnitratereductaseactivitywithincreasinggrowthrate,andpercentageof nitrateincorporationaccountedforbyNR(%NR/N)invariousspeciesunder differentlimitations.Lightiscontinuousandchemostatsarenitrate-limitedunlessotherwisenoted.SPECIESCONDiTIONSRELATIONSHIP%NR/NREFERENCELight-limitedChaetocerosaffinisnitratespike--50SlawykandRodier1986Skeletonemacostatumnitratespike--50Smithetal.1992Skeletonemacostatumsteadystate--<80Clayton1985Skeletonemacostatum12:12light:dark--10-80Kristiansen1987Skeletonemacostatumsteadystatepositive100-200presentstudyDitylumbrightwelliisteadystate--25Eppleyeta!.1969Thalassiosirapseudonanasteadystatepositive100present studyChiorellavulgarissteadystate—>100MorrisandSyrett1965Gonyaulaxpolyedra12:12light:darkpositive50Harrison1976Amphidiniumcarteraesteadystatepositive20presentstudyNitrate-LimitedChaetocerosaffinischemostat--338SlawykandRodier1986Thalassiosirapseudonanachemostatpositive83-190present studyThalassiosiraoceanicachemostatnegative--EppleyandRenger1974Skeleronemacostatumchemostatspositive87-176Dortcheta!.1979ChiorellastigmatophorachemostatsnegativeEverest eta!.1986Chiorellavulgarisstarvation--10-12MorrisandSyrett1965Gonyaulaxpolyedrastarvationpositive<10Harrison1976141Brendle-Behisch (1991) have shown that there is a close coupling between photosynthesis andnitrate reduction (but not necessarily nitrate uptake) in higher plants. Work by Turpin andcolleagues (Elrifi and Turpin 1986, Turpin et al. 1988, 1990) has demonstrated that when N-limited cultures of Selenastrum minutum receive nitrogen, photosynthesis is repressed. This isprobably due to a shortage of ribulose 1,5 bisphosphate (RUBP), brought about by the removalof intermediates in the tricarboxylic acid (TCA) cycle in order to provide so-called “carbonskeletons” with which to combine nitrogen to produce amino acids. Dark respiration alsoincreases under these conditions as carbon stores are metabolized to replenish TCA cycleintermediates. Interestingly, in cells that are limited by the supply of carbon dioxide, nitratecan be taken up, but it will not be reduced until carbon is available (i.e. futile nitrate reductiondoes not occur). This has also been demonstrated in higher plants such as maize seedlings(Pace et al. 1980). Flynn (1991) has suggested that the glutamine:2-oxoglutatate (ctketoglutarate) ratio may be an important parameter in these processes, since this appears to betrue in bacterial systems. If carbon became limiting the ratio would fall, but in a nitrogen-limited situation the opposite would happen. Sensitivity of biochemical and gene regulatorymechanism to this ratio could direct carbon towards nitrogen incorporation in times of nitrogensufficiency. When nitrogen was limiting, cells would store carbon as carbohydrate for lateruse (Flynn 1991). This theory also makes predictions consistent with ammonium inhibition ofnitrate uptake, which will be discussed in Chapter 4. Alternatively, it has been proposed thatnitrate itself may activate cytosolic protein kinases. These in turn would inhibit sucrosephosphate synthase and activate phosphoenolpyruvate carboxylase, resulting in diversion ofcarbon from sucrose synthesis to amino acid synthesis (Van Quy et a!. 1991, Campigny andFoyer 1992).Transient experiments showed that the adaptation of nitrate reductase activity to newphotosynthetic regimes and growth rates occurred quickly and within a day. There is evidencethat it took up to 3 days after the transition in irradiance before a new steady state was reached(but note the apparent over-compensation of L—*H cultures), but throughout the period, NRand nitrate incorporation rates were closely coupled. For culture work, where transitions142between steady states are made, this implies that “prehistory” of cells may not be so important(contrast with Dortch et at. 1979, Blasco et at. 1984). Most previous work on transitions hasinvolved spike additions of nitrate or nitrogen starvation (see Clayton 1986), and frequently itis not clear that cells were initially in a steady state. Under these conditions any time-dependent measurements (i.e. 3 h nitrate uptake) might not be expected to correlate well withan instantaneous enzyme measurement. Measurements of NR activity during irradiancetransitions have not previously been made.Nitrate-limited culturesAs was the case under light limitation, NR in nitrate-limited chemostat cultures waswell correlated with growth rate. However, the relationship was not 1:1 with calculatednitrate incorporation rates; NR at low dilution rates was much higher than the calculated rates,while NR at higher dilution rates was numerically lower than the calculated rate (althoughactivity was variable, and in individual cultures NR was rarely statistically different from thenitrate incorporation rate at dilution rates > 0.5 d4). There was always some nitrate andnitrite detected in the outflow of the chemostats run at highest dilution rates. This could haveoccurred if the dilution rate was very close to the growth rate; minor fluctuations in either thepump rate or the growth rate of the cells may have resulted in periods when dilution wasgreater than growth. This would mean that there was a small loss of cells and as a resultgrowth rates calculated from dilution rates might have been over-estimated. This might helpexplain the lower than expected NR activities in these cultures.In the literature, there is wide variability in the trends in NR activity found in nitrogenlimited cultures. The chemostat experiments of Eppley and Renger (1974) and Everest et at.(1986) both showed negative relationships; NR increased as growth rate decreased. Althoughnot strictly applicable in this chapter, Morris and Syrett (1965) and Harrison (1976) bothreported that NR could not account for observed rates of nitrate reduction in starved culturesof microalgae. Results of Dortch et at. (1979) and Slawyk and Rodier (1986) are more similarto the present study. Although results from these two studies are based on only three cultures143in total, they indicated that NR activity at low dilution rates exceeded calculated incorporationrates (174 and 338%), while at high dilution rates NR was closer to the calculated rate (87%).Some authors have invoked an alternate nitrate reduction mechanism to account fordiscrepancies (e.g. Eppley et al. 1969, Clayton 1986, Slawyk and Rodier 1986), but thisremains dubious.NR activity in the present study exceeded that needed to account for observed rates ofnitrate incorporation at low nitrate-limited growth rates. Nitrogen-limited cells develop theability to rapidly take up limiting nutrients (so-called “surge uptake”, see Conway et al. 1976,Conway and Harrison 1977, McCarthy and Goldman 1979, Dortch et al. 1991a). It ispossible that cells at low growth rates maintain NR at higher levels than needed in anticipationof periods of rapid uptake (see Slawyk and Rodier 1986). Alternatively, Ingemarsson (1987)suggest that at low growth rates in the duckweed, Lemna, it is the flux of nitrate (i.e. thetransport step) and not the NR activity that is limiting. This would be in accord with datafrom Dortch et al. (1979) showing a correlation between NR activity and internal nitrateconcentration, but not with those of Collos and Slawyk (1977) where this relationship was notseen.For diatom species, long periods of nutrient limitation may not be commonlyexperienced in the field. Typically, diatoms are first in successional patterns; they dominate athigh nutrient concentrations due to their rapid division rates (see Guillard and Kilham 1977).Later, as nutrients are exhausted, other species such as flagellates replace the diatomcommunity, which often sinks to the pyconocline. Thus, nutrient-limited chemostats run atlow dilution rates may have little relevance for diatom species (see also Rhee 1979,Zevenboom 1986).Scaling of NR activityGiven the range of differences in cell composition, and the different responses of cellconstituents to different limitations, it might have been anticipated that NR activity scaled to agiven biomass variable would correlate poorly with growth rate. In fact, this is not so; in all144cases except chl a, scaled NR activity was significantly and positively related to growth rates• and the light-limited cultures were no different from the nitrate-limited chemostat cultures.The variability was high, however, typically oniy 50-60% of the variance in scaled NR activitywas explained by growth rate. As discussed in Chapter 1, this may be the result of thevariation in the biomass measurement increasing variability in the scaled enzyme data. Thescaling problem is not an issue in the laboratory, but it becomes critical in the field. Scalingof NR to carbon is problematic because of the large amounts of detrital carbon found in marinewaters (see Banse (1977) for a discussion of the problem with reference to C:chl a ratios).Nitrogen potentially suffers the same problem, although these may not be insurmountable;Dugdale and Wilkerson (1991) found that nitrogen could be used as a scaling factor fornitrogen uptake without apparent interference from non-phytoplankton nitrogen. Chi a iseasily measured and correlates well with living phytoplankton biomass. However, it varieswith irradiance and nitrogen level, and it is the one scaling variable in the present study wherea significant relationship was not found. This is not surprising; cells growing slowly wouldhave low NR in both light- and nutrient-limited cases, but under light limitation chl a quotawould be high (due to light acclimation), while under nitrate limitation it would be low (i.e.cells would be chiorotic). Scaling to protein seems logical and is frequently done, but asdiscussed in Chapter 1 and Appendix A, it is uncertain what different spectrophotometricassays actually measure. The relationship between NR scaled to protein and growth rate wasalso one of the poorer relationships found. Cell volume may offer an alternative, but thiswould requires tedious microscopic measurements, which have large errors associated withthem. Light microscope measurements can be affected by halos around small particles; thisresulted in volume estimates of 2 m diameter latex bead standards that were up to 50%greater than the true values (Montagnes et a!. submitted) and may constitute a significant biasin cell volume measurement. If only bulk measurements of processes are required, it ispossible to scale measurements per litre or m3 of seawater, but this will provide noinformation about the physiological state of organisms in water masses with different145biomasses. As is the case for nitrate uptake rates, scaling NR activity to particulate nitrogenseems to be the most practical course.It is important to note that these conclusions apply, for the moment, only under steadystates, or perhaps light transitions between steady states. Some non-steady states will beconsidered in Chapter 4. There is ample evidence that different relationships between growthand cell composition can result from day:night cycles (Sakshaug and Andresen 1986), culturesenescence (Lewitus and Caron 1990), and temperature (Thompson et al. 1992). As well,limitation other than light or nitrate probably results in different patterns, as demonstrated forammonium and phosphate (e.g. Laws et at. 1985) and iron (e.g. Doucette and Harrison 1990).However, the fact that NR activities in selenium-limited cultures were related to nitrateincorporation rates in the same way as other cultures suggest that this may not be such aproblem.In a review of composition and metabolism, Madraiga and Joint (1992) concluded thatchanges in composition vary with the specific limiting factor, but that physiologicalmeasurements are more related to growth rate differences. Thus, to make the method asapplicable as possible, it may ultimately be more useful to scale NR activity to anotherphysiological measurement, perhaps one that changes minimally or predictably with growthrates. At the present time, too little is known to make a recommendation, but indices such aselectron transport system (ETS) activity (see Packard 1985, Martinez 1992) might havepotential.In summary, in this chapter, strong relationships between NR and growth rates andrates of nitrate incorporation have been demonstrated under steady state culture conditions.The relationship in Thalassiosira pseudonana is better under light limitation than nitratelimitation, where NR activity tends to exceed nitrate incorporation rates at low growth rates.Low nitrate-limited growth may not be a common situation for marine diatoms and thus maynot be ecologically relevant and of less importance to the use of NR activity in the field.These findings suggest that the control of nitrate reduction may well be at the level of theenzyme under steady state conditions. The 1:1 relationship between NR activity and nitrate146incorporation (particularly under light limitation) implies a control coefficient (Ci) near 1.0(Crabtree and Newsholme 1985), which suggests that NR activity can indeed be used toquantitatively predict metabolic rates in vivo. Enzyme scaling to biomass parameters issomewhat problematic since cell composition changes with growth rate are different dependingon the specific limiting factor. However, this appears to be severe only in the case of chi a.It is suggested that NR be scaled to particulate nitrogen, based on the problems found inaccurately measuring alternatives such as carbon, cell volume, or protein.147CHAPTER 4: EFFECTS OF LIGHT:DARK CYCLES, DLmRENT LIGHTSPECTRA, NITRATE EXHAUSTION, AND AMMONIUM ON THE RELATIONSHIPBETWEEN NITRATE REDUCTASE ACTIVITY AND NITRATE INCORPORATIONRATES IN THALASSIOSIRA PSEUDONANAINTRODUCTIONIn this chapter, the effects of several environmental influences that have been shown toplay a role in the regulation of NR activity will be considered. These include diel periodicityin irradiance, different light spectra, nitrate exhaustion, and the influence of ammonium. AsGuerrero et al. (1981) point out, these features (and others) that regulate MR activity alsoaffect the capacity of cells to assimilate nitrate. In each case, the goal of these experiments isto determine how these factors influence the relationship between MR activity and nitrateincorporation rate, and whether they pose problems for the use of MR as an index in the field.Effects of Did Periodicity in frradianceWith the exception of polar regions in certain periods of the year, diel periodicity is themost noticeable feature of irradiance cycles in the ocean (see Parsons et a!. 1984b). Suchcycles have profound influences on aquatic algae including effects on division cycles, taxis,photosynthesis, cell composition, and enzyme activity (Chisholm 1981, Prezelin 1992), Thereare many different diel patterns displayed, and these often appear to be taxa-specific. Forexample, dinoflagellates generally appear to have cell division phased near the light-darktransition, but diatoms such as Thalassiosira weisflogii display diel peaks in division frequencyat midday and midnight (Chisholm 1981). Diel periodicity of nutrient uptake has beenfrequently demonstrated in microalgae in culture (e.g. Eppley and Renger 1974, Syrett 1981),and in field populations (see MacIsaac 1978, Manasneh and Basson 1987, Cochlan et al. 1991,Vincent 1992). Of course, photosynthesis is light dependent. From the relationships betweenMR and nutrient incorporation, and MR and carbon assimilation in photosynthesis previouslydemonstrated and discussed (Chapter 3), it is not unexpected that did periodicity will also148influence MR activity. This has been demonstrated in higher plants at the level of the enzymeactivity (see Lillo 1983, Campbell 1988), the abundance of the enzyme protein (e.g. Oaks etal. 1990) and the rates of transcription and translation (see Lillo and Ruoff 1989, Deng et a!.1991). As well, similar results have been found for green algae (see Velasco et al. 1989),macroalgae (e.g. Gao et a!. 1992) and other microalgae (e.g. Eppley et a!. 1971, Packard etal. 1971a, Hersey and Swift 1976, Harrison 1976, Smith et a!. 1992), as well as in fieldpopulations (e.g. Packard and Blasco 1974, Collos and Slawyk 1976). Specific light-activating mechanisms for MR are also known (see Hug and Hunter 1991, Kaiser and BrendleBehisch 1991, Riens and Heldt 1992, Kaiser eta!. 1992, Huber et a!. 1992a, 1992b).Effects of different light spectraUnlike the full white light spectrum common in laboratory experiments and in theterrestrial environment, the light spectrum in the ocean is biased towards the blue because longwavelength light is effectively absorbed by water. This shift towards the blue increases in thedeep ocean, depending on the water clarity and the abundance of suspended matter (seeParsons et a!. 1984b). It has often been observed that long-term growth under blue light leadsto an increase in total protein content in higher plant (Duke and Duke 1984, Barro et a!. 1989)and algal cells (Wallen and Geen 1971, Morris 1981, Rivkin 1989, Kowallik et a!, 1990,Apparicio and Quinones 1991). This may well influence nitrogen metabolism and thus affectnitrate incorporation. However, blue light may also have specific and possibly differenteffects directly on the MR molecule, perhaps mediated through blue light receptors involvingphytochromes or flavin (see Azura and Aparicio 1983, Duke and Duke 1984, Ninneman 1987,Solomonson and Barber 1990, Hug and Hunter 1991, Lopez-Figueroa and Rueliger 1991).Effects of nitrate exhaustionIn natural populations, as in cultures, microalgae pass through several growth phases;an initial period of slow growth (lag phase), a period of logarithmic growth (log phase), aplateau of biomass (stationary phase) and a later period of decline (senescence) (Fogg 1975).149The transition to stationary phase is caused by a limitation of some necessary requirement,often a nutrient. In this case, the effect is similar to that of nutrient starvation. Up to thispoint in the thesis, care has been taken to see that all cultures have been in logarithmic growthphase, where experiments are most reproducible (see Rhee 1979). However, in the field, cellsmay face periods of nitrate starvation and thus may be in different growth phases. It becomescritical to understand how NR will respond when cells become nitrogen-starved. There aredata suggesting that there is a rapid decline in NR which occurs in step with decreases innitrate assimilation (e.g. Morris and Syrett 1965, Hersey and Swift 1976), or a gradual declinein NR, which occurs more slowly than the decrease in nitrate assimilation (e.g. Syrett andPeplinska 1988) There are even reports of transient increases in NR when nitrogen runs out(e.g. Kessler and Oesterheld 1970, Slawyk and Rodier 1986, Watt et at. 1992).Effects of ammoniumIn previous chapters of the thesis, experiments involved cultures which had been grownwith nitrate as the sole nitrogen source. This is not true in the natural environment wheresources such as ammonium (Wheeler 1983) and organic nitrogen (Antia et at. 1991) are oftenpresent. Since ammonium is more reduced than nitrate, it has been argued that ammoniumshould be a preferred nitrogen source since it requires less energy to use, and thus confers agrowth advantage (Syrett 1981, 1989) However, Thompson et a!. (1989) failed todemonstrate such an advantage in T. pseudonana cultures. Ammonium has been shown toinhibit the uptake of nitrate in some studies (e.g. Syrett 1981, Dortch et a!. 1991b, Cochlanand Harrison 1991a) but not in all cases (see Dortch 1990). There is strong evidence thatammonium is able to suppress NR activity in higher plants (Ingemarrson 1987, Solomonsonand Barber 1990) and algae (Morris and Syrett 1965, Serra et at. 1978b, Dortch et a!. 1979,Flynn et a!. 1993), although there are exceptions (Harrison 1976, Collos and Slawyk 1980).Whether the inhibition by ammonium of nitrate uptake and the inhibition of NR activity arecoordinated is an important question if NR is to be used as an index of nitrate incorporation.In higher plants, the two processes appear to be uncoupled in the short term. Lee and Drew150(1989) reported that nitrate influx to barley roots was inhibited within 3 mm, a much shorterresponse time than is typically found for NR activity. Larsson et al. (1985), for example,argued that the inhibition of uptake was much faster than the inhibition of NR in the green algaScenedesmus. Blasco and Conway (1982) suggested that the inhibitory effects of ammoniumon the two processes were independent in natural populations. How the inhibitions of nitrateuptake and NR activity are mediated remains unclear, but it is generally thought that someproduct of ammonium assimilation, such as glutamine, is responsible (Syrett 1981, 1989,Clarkson and Luttge 1991). Other mechanisms that have been proposed include a directinfluence of ammonium on NR (Florencio and Vega 1982), or an interaction betweenammonium and nitrate mediated by the links to carbon metabolism (see Flynn 1990).In this chapter nitrate reductase activity and nitrogen incorporation rates are comparedin cultures of T. pseudonana that have been: a) grown on light:dark cycles, b) grown underwhite, blue or red light, c) starved of nitrogen, or d) grown on (or in the presence of)ammonium. The goal of the study was to determine whether these conditions prevent the useof nitrate reductase as an index of nitrate incorporation.MATERIALS AND METHODSGeneral culture conditionsCultures of Thalassiosira pseudonana were obtained from the NEPCC and maintainedon artificial seawater (ESAW, modified as before) as described previously (Chapter 1). Asbefore, cultures were grown at 17.5°C, stirred and bubbled with filtered air.Light:dark cycle experimentsCultures were grown in 6 L glass flat-bottomed boiling flasks at 16°C in anenvironmental chamber. Irradiance on a 14:10 h light:dark cycle was provided by fluorescentlights (Vitalites). Four cultures were grown through a minimum of eight generations, two at45 mol quanta m2 s’, two at 6 mol quanta m2 s. Growth rates were monitored by151fluorescence measurements taken within 1 h of 10 OOh each day, or by cell counts using aCoulter Counter (see Chapter 1). Culture medium was identical to that previously described(Chapter 1), except nitrate concentrations were reduced from 550 to 225 M. Cultures inlogarithmic growth phase were sampled every 3 h over a 24 h cycle. At each sampling, 25 mlsamples were filtered (25 mm GF/F) and frozen for nutrient analyses later. Dissolved nitrate,ammonium, silicate, and phosphate were analyzed within one month using a TechniconAutoAnalyzer IT® and nutrient chemistry as described by Freiderich and Whitledge (1972).Nitrite was measured as described previously (Chapter 2). Samples were also taken andanalyzed for fluorescence, cell numbers, cell volumes, and carbon, nitrogen, protein and chl acell quotas, as previously described (Chapters 1, 2 and 3). Molar ratios of carbon:nitrogenand weight ratios of chlorophyll a:carbon were calculated. NR activity was determined at eachsampling as previously described (Chapter 2). Nitrate incorporation rates were calculatedfrom the change in particulate nitrogen in the cultures over each 3 h sampling interval, andcompared with NR activities.Light spectra experimentsSix cultures of T. pseudonana were grown in 1 L glass flasks in a water bath, aspreviously described (Chapter 1). Nitrate concentration in the medium was full ESAWenrichment, 550 M. Two cultures were screened with blue-coloured filters (Roscolux # 69),two with red-coloured filters (Roscolux # 19), and two remained in full white light.Continuous irradiance was adjusted with neutral density filters and distance so that each culturereceived equal quantum irradiance of 45 mol quanta m2 Cultures were allowed toacclimate for a minimum of 8 generations, and they were sampled in logarithmic growth phasefor cell numbers, cell volumes, and carbon, nitrogen, protein and chl a cell quotas, aspreviously described (Chapters 1, 2 and 3). Molar ratios of carbon:nitrogen and weight ratiosof chl a:carbon were calculated. NR activity was determined as previously described (Chapter2). Nitrate incorporation rates were calculated from the nitrogen cell quotas and growth rates,as described before (Chapter 2). For cell composition, growth rate and NR data, blue-, white-152and red-light treatments were compared using one-way ANOVA designs, followed by Tukeymultiple comparison tests, with a set at 0.05.Nitrate exhaustion experimentThree 1 L cultures of T. pseudonana were grown under continuous irradiance at 115mol quanta m2s1. Culture medium was as described previously, but nitrate concentrationwas one-fifth normal, or 110 M. Cells were maintained in logarithmic growth phase for 8generations, then allowed to grow into stationary phase. Beginning on day 3, for 5 days,samples were taken daily for nutrients (nitrate, ammonium, phosphate, silicate, and nitrite),cell numbers, cell volume, carbon, nitrogen, protein and chl a quotas. C:N and C:chl a ratioswere also calculated. On days 3, 4, and 6, NR activity was measured as before (Chapter 2).For day 3, during logarithmic growth, nitrate incorporation rate was calculated as before(Chapters 2 and 3); for days 4-7 it was estimated from depletion of nitrate from the medium,or increase in particulate nitrogen in the culture. Changes in cell composition over time wereevaluated by performing linear regression analyses of composition versus time and comparingthe slopes of these regressions with zero using t-tests (Steel and Tome 1980, Wilkinson 1990).Thus, increases in cell composition would be represented by regressions with slopes greaterthan zero and decreases by regressions with slopes less than zero. This is a conservativetechnique, since non-linear changes may also have occurred that might not be detected in alinear model, but the relatively few points in time meant that more complex models could notbe judged statistically.Effects of ammonium and ammonium pulsingSix 1 L cultures of T. pseudonana were grown under continuous irradiance at 115 molquanta rn2 51 Cultures were grown on ESAW as described previously, except that thenitrogen source was either 75 jLM nitrate for two cultures (N03 treatment), or 75 Mammonium (added as ammonium chloride) for two cultures (NH4 treatment). Two additionalcultures were grown on nitrate-enriched medium (75 1LM), but each day a pulse of ammonium153sufficient to bring the ambient concentration to 2 M was added (P treatment). This treatmentwas chosen because 2 M is a common level of ammonium in many areas of the ocean (seeMcCarthy 1980), and because a level of 1-2 iM ammonium is generally thought to affectnitrate uptake and nitrate reductase activity (see Syrett 1981, but see also Dortch 1990). Cellswere maintained in logarithmic growth phase for 8 generations, then sampled for cell numbers,cell volumes, and carbon, nitrogen, protein and chlorophyll a quotas, as previously described(Chapters 1, 2 and 3). Molar ratios of carbon:nitrogen and weight ratios of chlorophylla:carbon were calculated. NR activity was determined at each sampling as previouslydescribed (Chapter 2). Nutrients were sampled at 16 h before the experiment and immediatelybefore NR samples were taken. From the changes in nitrate or ammonium concentrations,rates of nutrient uptake were calculated and expressed as specific daily rates (i.e. d4, as forgrowth rate). Nitrogen incorporation rates were calculated from the change in particulatenitrogen in the cultures over each 3 h sampling interval. For composition, growth rate and NRdata, N03, NH4 and P treatments were compared using one-way ANOVA designs, followedby Tukey multiple comparison tests, with a. set at 0.05.RESULTSLight:dark cycle experimentsPrior to the experimental period, growth rates based on increases in fluorescence or oncell numbers were identical: 0.90 (±0.02) d1 for high-light-grown cultures, and 0.13(± 0.02) d4 for low-light-grown cultures. Figure 4.1 A shows the increase in cell numbers ineach of the four cultures. Over the 24 h sampling period, it was intended that culture densityremain < 6 x i05 cells m14 so that cultures would remain in logarithmic growth phase. Forone of the high-light grown cultures, cell density approached this limit, so after the 9 hsampling, the culture was diluted approximately by half. Despite this disturbance, growthrates and compositional trends in this culture were no different than its replicate. Whenmeasurements were made once a day at the same time of day, growth rates using in vivo154CI1)0C0‘-401000_ 700500t)0‘‘ 300C‘1075—‘ 3283 120 5 10 15time (h)Figure 4.1. Growth characteristics of log-phase cultures of Thalassiosira pseudonanagrown on 14:10 h light:dark cycles. A) Culture densities, B) relative fluorescence,C) relative fluorescence per cell. Cultures were grown at high light ( • , • )or low light ( 0 , E ). Points in A) and B) represent single determinations;points in C) represent means of two cultures, with standard errors of the mean.Note that the vertical axes for A) and B) are logarithmic. The solid black barindicates the dark period.20 25155fluorescence were identical to those based on cell numbers. However, if a shorter time scalewere used, growth rates were very different between the two methods. As illustrated in Figure4.1 A and B, the increases in cell numbers during the day were much less than the increases influorescence. Conversely, at night, fluorescence changed little, but cell numbers increased.This resulted in a clear pattern in which fluorescence per cell increased in the light period anddecreased in darkness (Fig. 4.1 C). This pattern was much less pronounced in the low-light-grown cultures. No pattern in cell division was found; cell numbers increased evenly in lightand in darkness at similar rates.In terms of cell composition, there was little variation in low-light-grown cultures forany parameter measured (Fig. 4.2), although protein quotas of low light cells tended to behigher in the dark (Fig. 4.2 E). For high-light-grown cultures, however, distinct diel patternswere seen in cell volume, cell carbon quota, and in the C:N and C:chl a ratios (Fig. 4.2 A, B,F, and G). Cell volume showed peaks in the middle of the light period, and at the beginningof the dark period (Fig. 4.2 A). Cell carbon increased during the light period and decreased inthe dark (Fig. 4.2 B). C:chl a ratio followed a pattern similar to cell volume (Fig. 4.2 G),while C:N ratio followed a pattern similar to carbon quota (Fig. 4.2 F). Comparing low- andhigh-light cultures, cell volume and carbon quotas tended to be higher in the low lightcultures, but the peak values of the high light culture were equal to those in cells grown underlow light. Cell nitrogen and chl a quotas were uniformly higher in the low-light cultures, andC:chl a ratios were higher in the high-light cultures.In terms of NR activity per cell, there was a diel periodicity in both low and high-lightcultures (Fig. 4.3). There were two peaks, one at the middle of the light period, and a secondbefore the beginning of the light period. NR activities at these peaks were higher in high-lightthan in low-light grown cultures, but NR fell to nearly equal levels at other times. In highlight cultures, NR activities per ml of culture matched rates of particulate nitrogen increasethroughout the diel cycle extremely closely in one culture (Fig. 4.4 A). In the other culture,the match was good, except in the first three sampling periods when NR activity exceededparticulate nitrogen increases (Fig, 4.4 B). This was during the period when the culture156S.- 24o0>22—‘ 7o -—5C)o oE ‘1.64-.o 12-i 0.80.3. 0.2oC)1.2o0.8—‘ 4-•• 0.4SoS‘—‘o Q.,—‘ 4z 9rJ0‘- 30200c-)time (h)Figure 4.2. Changes in cell composition in cultures of Thalassiosira pseudonanagrown on 14:10 h light:dark cycles at low (6 mol quanta m2 s_i, 0 ) orhigh (45 itmol quanta m2 s1, • ) irradiance. A) Cell volume, B) cell carbonquota, C) cell nitrogen quota, E) cell chlorophyll a quota, E) cell protein quota,F) cell C: N ratio, and G) cell C:chlorophyll a ratio. Each point represents themean of duplicate determinations from two separate cultures. Error barsrepresent standard errors of mean values, or where absent are smaller than thesymbols.0 5 10 15 20 2515740.—C.)z10time (h)Figure 4.3. Nitrate reductase activity in log-phase cultures of Thalassiosirapseudonana grown on 14:10 h light:dark cycles at low ( 0 ) or high( • ) irradiance. Each point represents the mean of two separatecultures. Error bars represent standard errors of mean NR activity.0 5 10 15 20 2515840300I 20C0.— I 10—C00C) CoC20.—010ztime (h)Figure 4.4. Nitrate reductase activity ( 0 ) or calculated nitrate incorporationrate ( • ) in two log-phase cultures (A and B) of Thalassiosira pseudonana- grown on 14:10 h light:dark cycles. Each point represents the mean of twoenzyme assays. Error bars represent standard errors of mean values.0 5 10 15 20 25159densities exceeded 6 x 10 5 cells m1’. It was hoped that nutrient concentration in the mediumcould be used to provide an independent estimate of nitrate incorporation rates, however,owing to the high dilutions necessary in order to measure nitrate in these cultures, the resultingconcentrations varied widely and were not suitable for this purpose (e.g. For nitrate, sampleshad to be diluted about 1:20 to bring them within the linear range of the colourimetricreaction. Since the routine resolution of the AutoAnalyzer during the experiment wasapproximately 1 M, only differences greater than about 20 M could be reliably detected;this is on the same order as the nitrate concentration changes observed in the cultures).Nitrate, phosphate and silicate never neared depletion, and only low levels (< 0.5 M) ofnitrite or ammonium were recorded. For the low-light cultures, particulate nitrogen variedwidely as well (Fig. 4.5 A, B). As a result it became difficult to compare NR activity directlyto increases in particulate nitrogen. As an alternative, NR activity and particulate nitrogen foreach culture in each sampling period were used to estimate the particulate nitrogenconcentration at the next sampling period. As shown in Fig 4.5, these predictions werecertainly within the ranges of increases observed, given the high variability of the data. Thissuggests that NR activities were reasonably close to those necessary to account for theparticulate nitrogen increases observed.Light spectra experimentsCell composition differed between blue-, white- and red-light treatments (Fig. 4.6).Cell volumes were higher in blue light than in white light and higher in white light than in redlight (Fig. 4.6 A). Carbon quotas were greater in blue light than red light, but white lightcultures were not different from either blue of red light cultures (Fig. 4.6 B). No significantdifferences in nitrogen quota, protein content, or C:N ratios were seen (Fig. 4.6 C, D, F). Inthe cases of chl a quota and C:chl a ratio, blue-light cultures were significantly higher thanwhite- or red-light cultures (Fig. 4.6 E and G).Under equal quantum irradiance, blue light cultures grew significantly faster than whiteor red light cultures (Fig. 4.7 A). As well, NR activities and calculated rates of nitrate1600.460.420.—‘-4C.)C)0‘.40.34C.)0.3215 20 25time (h)Figure 4.5. Particulate nitrogen concentration measured ( 0 ) or predicted fromNR activityty ( • ) in two log phase cultures (A and B) of Thalassiosirapseudonana grown on 14:10 h light:dark cycles. Each point represents theI I‘‘AI I II I II I0.30I I I0 5 10mean of duplicate determinations.161Figure 4.6. Cell composition in log-phase cultures of Thalassiosira pseudonanagrown under equal quanta (45 mol quanta m2 s’) of blue , white, or red light.A) Cell volume, B) cell carbon quota, C) cell nitrogen quota, D) cell proteinquota, E) cell chlorophyll a quota, F) cell carbon:nitrogen ratio, and G) cellcarbon:chlorophyll a ratio. Error bars represent standard errors of meandeterminations from two separate cultures. Treatments not significantlydifferent from one another at P = 0.05 are joined by lines.C)CCCCC)C)0C0E00zL)E322801081.61.4C 1.2CI0.30.20C)0.4o3020KT-—C:TDG:=blue white red1621.4 AB80C) —C)60Ii:__Z blue whiteFigure 4.7. Effects of blue, white and red light on: A) specific growth rate, andB) nitrate reductase activity (I_____ ) or calculated rates of nitrate incorporation) in log-phase cultures of Thalassiosira pseudonana. Error bars representstandard errors of the mean of two separate cultures. Treatments not significantlydifferent from one another at P = 0.05 are joined by a line.I-I-red163incorporation were higher for blue light (Fig. 4.7 B). Under blue and white light, NR activityand calculated nitrate incorporation rates were not different (P > 0.5 in both cases). In redlight cultures, NR activities were significantly lower that the calculated rates (P < 0.03).Nitrate exhaustion experimentCultures entered stationary phase in terms of fluorescence data after the third day of theexperiment (i.e. the log-normal plots of fluorescence or cell numbers versus time ceased to belinear, see Fig. 4.8 A), but cell numbers continued to increase until day 5 (Fig. 4.8 B). Thisresulted from a decrease in fluorescence per cell (Fig. 4.8 C). pH over the experimentalperiod remained constant at about 8 (Fig. 4.9 A). Nitrate was the first nutrient exhausted onday 5 (Fig. 4.9 B; note that in this case and in the cases following nutrient exhaustion appearsto occur earlier on the figures, due to the wide scale); low levels of silicate persisted until day6 (Fig. 4.9 C) and phosphate was never exhausted (Fig. 4.9 D). Low levels of nitrate andammonium were seen (Fig. 4.9 E, F), but nitrite had disappeared by the end of theexperiment.Cell volume was relatively constant over the experiment, although there was a slight,but not statistically significant decline (Fig. 4.10 A, P > 0.05). Cell carbon quota increased(Fig. 4.10 B, P < 0.001), which coupled with decreases in nitrogen quotas (Fig. 4.10 C, P <0.05), resulted in an increase in the C:N ratio over time (Fig. 4.10 E, P <0.001). Chl aquotas declined slightly (Fig. 4.10 D, P < 0.05), and C:chl a ratios increased with time (Fig.4.10 F, P < 0.001).NR activity fell over the course of the experiment (Fig. 4.11). NR activity generallyfollowed rates of nitrate incorporation calculated from depletion or increase in particulatenitrogen, but NR was still detectable on day 6, at which point nitrate had been exhausted andthere were no further increases in particulate nitrogen.164C3OOOIII300I 2 4time (days)Figure 4.8. Changes in A) cell number, B) culture fluorescence, and C)fluorescence per cell, in cultures of Thalassiosira pseudonana enteringstationary phase (indicated by the vertical line). Each point representsthe mean of three replicate cultures. Error bars represent standard errors,or if not seen, are less than the size of the symbol. Note that the verticalaxes for A and B are logarithmic.5 6 716510987_12040Z 012080400160.80.4CZ 0.0time (d)Figure 4.9. pH and ambient nutrient concentrations for cultures of Thalassiosirapseudonana entering stationary phase as indicated by the vertical line. A) CulturepH, B) nitrate, C) silicate, D) phosphate, E) ammonium, and F) nitrite. Each pointrepresents the mean of three cultures. Error bars represent standard errors ofmean values, or if not seen are smaller than the symbols.0 1 2 3 4 5 6 7166a)C.)00_____________________ __________________________CaC)a)C)toCa.4-0C)Figure 4.10. Cell composition for cultures of Thalassiosira pseudonana enteringstationary phase, as indicated by the vertical line. A) Cell volume, B) cell carbonquota, C) cell nitrogen quota, D) cell chlorophyll a quota, E) cell carbon:nitrogen ratio,and F) cell carbon:chlorophyll a ratio. Each point represents the mean of threeseparate cultures. Error bars represent standard errors of the mean, and if not seenare smaller than the size of the symbol. Note that not all measurements were madeat each time.E0a)C)a)C)to‘SCa0a)to00E00z424038161281. I I I I IA:——=—:toto0CaC)0 1 2 3 4 5 6 7time (d)167a)-..I I I I I I Io —0 0.1.—a)- 4.-C C).EE2OI________________________________0 1 2 3 4 5 6 7Z time (d)Figure 4.11. Nitrate reductase activity ( • ) or rate of nitrate incorporationcalculated from growth rate and nitrogen quota ( 0 ), increase in particulatenitrogen ( E ), or depletion of nitrate from the medium ( L ). Each pointrepresents the mean of determinations from three separate cultures ofThalassiosira pseudonana entering stationary phase. Error bars representstandard errors of mean values, or if not seen are smaller than the symbols.Note that not all measurements were made at each time.168Effects of ammonium and ammonium pulsingIn cultures grown on ammonium, cells were significantly greater in volume and carbonquota than either those grown on nitrate or pulsed with ammonium (Fig. 4.12 A, B). Nosignificant differences in nitrogen or protein quotas were found, but ammonium-growncultures tended to have numerically higher quotas (Fig. 4.12 C, D). There were nodifferences found in chi a quotas, or in C:N or C:chl a ratios (Fig. 4.12 E, F, G).Ammonium-grown cultures grew at significantly higher rates than nitrate-grown orammonium-pulsed cultures (Fig. 4.13 A). In terms of nutrient use and NR activities, the twoammonium-grown cultures behaved differently, and so are presented separately (Fig. 4.13).Ammonium was exhausted in one culture (NH4-1), but remained above 6 M in the other(NH4-2). Nitrate levels in the ammonium-grown cultures were on the order of 1 JLM, due tobackground contamination in the NaCl in ESAW. For the same reason, ammonium levels inthe nitrate grown culture were up to 0.5 M. Nitrate-grown cultures used only nitrate, anddid so at rates consistent with their growth rates (Fig. 4.13 B). Both ammonium-growncultures used ammonium, but a significant use of nitrate was seen in the culture in whichammonium was exhausted (NH4-2). NR activity and calculated nitrogen incorporation rateswere not different for nitrate-grown and ammonium-pulsed cultures (Fig. 4.13 C, P > 0.5 inboth cases). In the ammonium-grown culture where ammonium was not exhausted, no NRactivity was detected, however, in the other ammonium culture, NH4 had significant NRactivity (Fig. 4.13 C).DISCUSSIONEffects of Diel Periodicity in IrradianceThe finding that in vivo fluorescence and cell number show different patterns ofincrease within a diel cycle is important for determination of culture growth rates. Clearly,169Figure 4.12. Cell composition in log-phase cultures of Thalassiosira pseudonanagrown with 75 jM ammonium (NH4), 75 tM nitrate (N03), or 75 tMnitrate with daily pulse of 2 jtM ammonium (P). A) Cell volume, B)cell carbon quota, C) cell nitrogen quota, D) cell protein quota, E) cellchlorophyll a quota, F) cell carbon:nitrogen ratio, and G) cell carbon:chlorophyll a ratio. Error bars represent standard errors of meandeterminations from two separate cultures. Treatments not significantlydifferent from one another at P = 0.05 are joined by lines.LKB0‘S00C.)ci)C)ba‘S000E0S0zc-)4842E 36024ci)C.)ci)‘S162.82.4_— 8705040.30__CD:I-I-T‘4F-r GNH4 N03 P170Figure 4.13. Effects of growth on 75 jiM nitrate (N03), 75 jiM ammonium (NH4) or75 jiM nitrate plus daily 2 jiM pulses of ammonium (P) on cultures of Thalassiosirapseudonana. A) Specific growth rate, B) specific nutrient uptake rates for nitrate) and ammonium L ), and C) nitrate reductase activity ( ) and calculatednitrogen incorporation rate____). Each bar represents the mean and standard errorof two cultures, except ammonium cultures, which are shown separately in B and Cbecause the replicates behaved differently.-z0II.s1.81.61.4321012080400AB000Qa)0‘—400Ca)a).z0E NH4-1 NH4-2 N03 P171fluorescence measurements must be made at the same time each day, or erroneous estimatesmay result. Interestingly, there appeared to be no diel periodicity in the division cycle ofThalassiosira pseudonana under these conditions. As Chishoim (1981) and Prezelin (1992)point out, division patterns are very specific to the taxonomic group considered. Chishoim(1981) found that the diatom T. weisflogii had two daily “bursts” of growth centered at middayand midnight. If this is also true for T. pseudonana, it is possible that the peak of division wassimply spread out enough that individual peaks could not be discerned. Nelson and Brand(1979) have studied cell division patterns in 13 species of marine phytoplankton, including 7clones of T. pseudonana. They found that periodicity of division was species and clonespecific. In at least 4 clones of T. pseudonana division rates were nearly constant throughoutthe light:dark cycle.Patterns of cell volume over diel cycles have been investigated previously. AsChisholm (1981) reports, patterns in diatoms are often complex (see discussion of cell volumein Chapter 3). For T. weisflogii, there was a bimodal pattern with maximum volumesoccurring at mid-morning and just before the light-dark transitions. This was correlated withcell division cycles. A similar bimodal pattern, with different timing was seen here, but therewere no associated division cycles; if a division cycle was involved, the cell volume might beexpected to nearly double. The increases observed were only 20%, but as previouslymentioned, Coulter Counter volumes may be suspect (see Chapters 1 and 3). In the flagellatesHeterocapsa sp. and Heterosigma akashiwo, cell volume monotonically increased in light,peaked just into the dark period, then decreased until the next light period began (Latasa et a!.1992, Berdalet et a!. 1992). Eppley and Coatsworth (1966) reported a similar pattern inDunaliella tertiolecta, which appeared to correlate with cell division, as have Marsot et a!.(1992) in dialysis cultures of Phaeodactylum tricomutum. There are few data for other cellcomposition parameters. It is interesting to note that cell carbon data, in the present study,closely follows the pattern reported for cell volumes in the majority of studies. Morris (1981)summarized data showing that for several species of green algae, carbohydrate contentincreased during the day, while protein increased at night because nitrogen is incorporated with172carbon into protein. Such a pattern would correlate well with carbon quotas in the presentstudy. Increases in protein at night in low light cultures in the present study also tend tosupport this idea, but a similar pattern was not seen at high light. Working with S. costatum,Smith et al. (1992) reported that there were midday peaks in C:N ratios. However, theseauthors also exposed the algae to transitions in nutrient availability, which may have causeddifferences. Eppley and Coatsworth (1966) reported diel variation of up to 20% in C:chl aratios in D. tertiolecta, similar to the present study.Differences between high and low light grown cultures were consistent with previousresults for chl a and C:chl a data (Chapter 3), but were quite different for carbon, nitrogen andcell volume data. Cell volume and carbon were previously found to increase with light-limitedgrowth rate, but here they were either not different, or were lower at high light than at lowlight. There were also differences in nitrogen quotas that were not previously observed. Thereasons for these different findings are unknown, but clearly there are dangers in makinginferences about cell composition in cells grown on a light: dark cycle from cultures grown oncontinuous light.NR activity per cell showed a double peak in the diel cycle. This has not previouslybeen noted, although a similar pattern is visible in data from cultures of Isochrysis galbana ina recent study (Flynn et al. 1993). Three patterns have been commonly found. The first,where activity shows a peak in the light period and very low levels at night, has beendemonstrated in higher plants (Deng et at. 1991), macroalgae (Gao et at. 1992) andmicroalgae (Packard et al. 1971a, Hersey and Swift 1976, Collos and Slawyk 1976, Harrison1976, Velasco et al. 1988, Smith et at. 1992). Martinez et at. (1987) found a variation on thispattern in natural populations of marine phytoplankton; NR activity showed two peaks in themiddle of the light period, with a decrease in activity at solar noon. This was attributed tophotoinhibition of nitrate uptake during the period of highest irradiance (Martinez et at. 1987).The second pattern is a monotonic increase in the light and a decrease in the dark, reported inbarley leaves (Lillo 1983) and natural phytoplankton populations (Manasneh and Basson 1987).Finally, increases in activity just before dawn and declines during the day have been found in173Emiliania huxleyi in chemostats (Eppley et a!. 1971), and in natural marine phytoplanktonpopulations (Packard and Blasco 1974). There is also a report of this pattern in tobacco plants(Roth-Berjerano and Lips 1970), although the pattern varied with the season and sampling wastoo infrequent to provide good resolution. There may be at least two reasons for thesedifferences. Lillo (1983) found that the particular assay used (e.g. in situ versus in vitro)could give very different patterns; there is certainly great diversity among assays in theliterature (see Chapter 2). In the present study, use of high EDTA may have activated dark-inactivated NR (see Kaiser et a!. 1992), but this is unlikely to have made much of a differencebecause such a mechanism only operates on very short time scale (i.e. minutes). Secondly, thesampling frequency of many studies may be insufficient to catch both the peaks. For example,Eppley et a!. (1971) sampled irregularly every 5-6 h, Smith et a!. (1992) sampled every 4-8 h,Manasneh and Basson (1987) sampled at 6 h intervals, and Harrison (1976) sampled onlyevery 12 h. Packard and Blasco (1974) also had restricted time series and samplings. Fromdata in the present study, and these considerations, the seemingly confficting results of manyof these reports can be reconciled. It is also interesting that Cobs et a!. (1993) have showndiel, midday peaks in RUBISCO activity in natural flagellate populations. This concurs withthe nitrogen-carbon coupling discussed in Chapter 3.NR correlated very well with calculated rates of increase of particulate nitrogen in mostcases. Apparently, NR activity exceeded calculated rates as culture densities became high (seeFig. 4.4 B), although the reasons for this are unknown. As cultures reached high densities,pH increased, indicating a possible carbon limitation (nitrate remained above 400 JLM andsilicate and phosphate were both in excess). Compared with activity measured duringlogarithmic growth, NR activity in these cultures nearly doubled (data not shown). Althoughsuch circumstances are not likely to occur in natural waters, it is an interesting result whichshould be pursued. The correlation between NR and calculated rates agrees with the findingsof Eppley et a!. (1971), but although they found a similar pattern, they could not account formore than 25% of calculated incorporation rates with NR activity. This was probably due to apoor extraction of the enzyme. Similarly, Collos and Slawyk (1976) found a correlation1 74between increase in particulate nitrogen and NR activity on a diel cycle, but NR activity couldonly account for 12% of the particulate nitrogen increase. As previously discussed, thecalculated incorporation in this study may involve both uptake and assimilation (see Chapter3), but it does clearly suggest that cells are taking up nitrate in the dark. It has been arguedthat nitrate reduction would not proceed at night because the energy must be derived fromphotosynthesis (Morris 1981) or because photosynthesis must provide carbon skeletons toattach nitrogen (Syrett 1981, 1989). However, breakdown of cell storage products couldprovide both these requirements (see Turpin et al. 1988, Turpin 1991). The case for nitrateuptake being restricted to the light is better established in higher plants (see Oaks et al. 1990),but it is much less certain in the algae. In dinoflagellates, Hersey and Swift (1976) reportedthat Amphidinium carterae and Cachonina nei did not assimilate nitrate in the dark, butHarrison (1976) and MacIsaac (1978) found that Gonyaulax polyedra did; under nitrogenstarvation, cells were capable of meeting 50% of their nitrogen requirements by dark uptake.Because Packard and Blasco (1974) found little periodicity in NR for G. polyedra, theyhypothesized that this species may have competitive advantages in terms of being able to takeup nitrate in the dark. There are likely species-specific differences: Eppley et a!. (1971)reported that the diel periodicity of nitrate and ammonium uptake was more pronounced in thediatom S. costatum than in the prymnesiophyte Emiliania huxleyi. As Cochian et al. (1991)discuss, the absence of nitrate uptake at night may be largely a misconception. They showedthat nitrate was taken up at night by natural populations at 15-16% of the maximum daytimerate. Marsot et al. (1992) found that in dialysis cultures of Phaeodactylum tricornutum, nitrateuptake in the dark was almost half of that in the light. It has also been suggested that NiRshould be less active in the dark because it relies on ferredoxin for reducing power, andferredoxin is apparently only available when photosynthesis proceeds (see Guerrero et al.1981, Martinez 1991). However, Huber et a!. (l992a) found that changes in the activity ofNR from spinach leaves were far greater than changes in NiR over a day-night cycle.There has been a great deal of work specifically on the activation/inactivation of NRfollowing transitions between light and darkness. It is probable that more than one mechanism175is involved in this process. There are reports that NR activation requires synthesis of newprotein (Lillo and Ruoff 1989, Velasco et al. 1988), but there are also reports of increases inNR activity occurring in a matter of minutes following a transition from dark to light (seeKaiser et al. 1992, Riens and Heldt 1992). Such rapid transitions have been shown to involvephosphorylation (Huber et al. 1992a, 1992b, MacKintosh 1992, see also Budde and Randall(1990) for a review of phosphorylation mechanisms). On a longer time scale, however, thedegradation and synthesis is almost certainly involved (see Lillo 1991). Smith et a!. (1992)found that in S. costatum, there was a diel cycle of NR synthesis in which peaks in NR-mRNAwere followed by peaks in NR protein. However, a peak in NR activity preceded these peaks,suggesting that an activation of pre-existing NR enzyme was also involved. Deng et a!. (1991)suggested that these synthesis/degradation patterns may be controlled by the presence ofglutamine, or another nitrogen metabolite. Regardless of the precise mechanisms involved, onthe 3 h time scale measured, NR activity was an adequate predictor of nitrate incorporationrates in the present study.Effects of different light spectraThe results of the light spectra experiment are difficult to compare with other literatureresults because different authors have combined the light treatments with day:night cycles, orhave used either equal energy, or equal quanta of light. As well, some have worked withsaturating irradiances, while others chose lower irradiances. As Morris (1981) points out, theeffects of irradiance level probably have a greater influence than the spectral composition ofthe light.In terms of composition, most work supports the idea that cells grown under blue lightaccumulate protein, while cells grown under red light accumulate carbohydrate (Morris 1981,Barro et a!. 1989, Kowallik et a!. 1990, Grotjohann and Kowallik 1989). Rivkin (1989)demonstrated this in Dunaliella tertiolecta and Thalassiosira rotula, and found in addition thatcarbon quotas were higher in red and blue light-grown cells than in white light-grown cells.This contrasts with results obtained here, but note that Rivkin (1989) used equal176photosynthetically usable radiation (PUR), a measurement that is lower than photosyntheticallyavailable radiation (PAR) which is what was measured in the present study (see Parson et a!.1984b). In addition, Rivkin grew cells on a 12:12 light:dark cycle. Grotjohann et al. (1992)found that Chiorella kessleri grown in blue light had up to 50% more protein than cells grownin red light, and 60% more reaction centres per chi a. Higher protein quotas in blue light-grown cells were not found in the present study. There is also some disagreement on theeffects of different light spectra on photosynthetic pigments. Rivkin (1989) found that chi awas greatest in white or blue light-grown cells and lower in red light-grown cells. This is theopposite of what was found in the present study. Wallen and Geen (1971) found that bluelight grown cells of D. tertiolecta and T. pseudonana had higher chi a, as did Senge andSenger (1991) in three species of green microalgae, and Hermsmeier et at, (1991) inScenedesmus obliquus. However, Barro et a!. (1989) found that soybeans grown in blue orwhite light had less chi a than those grown in red light. Thus, there appears to be a lack ofconsensus on this point. The significantly greater cell volume and carbon quotas of blue lightcells may be consistent with the general tendency of cells growing at higher light-limited ratesto have higher volumes and carbon quotas (see Chapter 3), since blue light cells also grewfaster.The finding that cells grown in blue light grew significantly faster than cells grown onequal quanta of other light spectra is in good agreement with the findings of Wallen and Geen(1971) where growth of D. tertiolecta and T. pseudonana was about 20% greater on blueversus white light, but it contrasts with studies by Rivkin (1989) and Grotjohann et at. (1992)where no differences were found. One reason for these differences may be the irradiance levelused. For T. pseudonana, irradiance in the present study was 45 mo1 quanta rn_i i, whichis probably not saturating for growth (see Fig. 1.3, Chapter 1). Although it is difficult tocompared since Wallen and Geen (1971) measured irradiance in energy versus quantum units,using approximate conversions in Parsons et at. (1984b) given irradiances of 30-40 molquanta m’ Rivkin (1989) grew some of his cultures at saturating irradiance (120 molquanta m’ s-i), but he also provides data from much lower irradiances (12 and 40 mol177quanta m4 s1); the trends in growth rate at low irradiance are no different from those at highirradiance. However, the possibility remains that the effect is species specific. Such growthrate differences may be reflected in other metabolic rates. In macroalgae, blue light has beenfound to increase photosynthetic rates by causing a surface acidification of plants leading toincreases in CO transport and thus photosynthesis (Forster and Dring 1992, Schmid andDring 1993). It is uncertain whether such a mechanism operates, or would be useful in aunicellular organism. Photosynthesis has also been shown to increase under blue light inmicroalgae (Wallen and Geen 1971, Senge and Senger 1991) In addition, blue light appears toenhance rates of cell respiration (Kowallik et a!. 1990) , although this has not been found in allcases (Wallen and Geen 1971). It has been proposed that the lower carbohydrate quotassometimes seen in blue light-grown cells are due to these increased respiration rates, which arein turn the result of increased glycolytic activity. Increased glycolytic activity may be due tolight activation of enzymes such as phosphofructokinase (PFK), as has been demonstrated inChiorella kessleri (Grotjohann and Kowallik 1989, Kowallik and Grotjohann 1988).NR activity correlated very well with calculated incorporation rates, except in red lightgrown cells, where NR was too low to account fully for observed rates. The reason for thediscrepancy is unknown, but it may relate to the absence of blue light; there are indicationsthat blue light has particular effects on NR. Duke and Duke (1984) and Ninneman (1987)have reviewed the specific effects of blue light on NR. These effects appear to involveincreases in NR synthesis and are mediated by phytochrome (especially at low irradiance), orby flavin (at higher light). In green algae, where a cyanide-based inactivation mechanism hasbeen demonstrated, blue light appears to reverse this inactivation (Solomonson and Barber1990). Azura and Apparicio (1983) proposed that the NR activation had to do with balancingcellular redox levels under higher energy blue light. In this scheme, NR would have asecondary non-assimilatory role in using NADH, a process which would be reflected in nitriteexcretion from the cell. Work with the green alga Monoraphidium braunii has establishedthat: a) blue light activation of NR is connected with the cyanide inactivation mechanism inthis species and apparently involves flavin (Navarro et a!. 1991), b) that activation of nitrate178and nitrite transport also occurs (Aparicio and Quinones 1991), and c) that effects similar tothose of blue light can be seen under low CO2 availability (Quinones and Aparicio 1990).Low CO2 availability would also imply that reductant could not be used to fix carbon, andthus NR activity increases could be used to control redox levels under these conditions. Sinceno evidence of a cyanide-inactivation mechanism was apparent in diatoms (see Chapter 2), thismay not be a feature of NR activity in the present study. The blue light induction of NR in thegreen macroalga Ulva rigida has been shown to be dependent on photosynthesis (Corzo andNeil 1992b) raising the possibility that it is a response to increased growth rate and thereforenitrate incorporation rates. In any case, blue light does not affect the close relationshipbetween NR activity and incorporation rate. It is unlikely that the disagreement between NRin red-light-grown cells and nitrate incorporation rates poses a serious problem in the field,since it is difficult to envision a set of circumstances where cells would be exposed to red lightalone.Effects of nitrate exhaustionAs was the case for T. pseudonana grown on light:dark cycles, fluorescence and cellnumbers in the cultures moving into stationary phase did not agree. There was a gradualdecline in fluorescence per cell until a minimum was reached, and this did not correspond to achange in chi a. Thus, the usefulness of fluorescence as a biomass indicator in other than logphase cultures is questionable.The lack of increase in pH suggests that carbon dioxide did not become limiting to thegrowth of the cultures; had this occurred, pH would have risen as CO2 levels were reducedand the carbonate equilibrium shifted (Riley and Chester 1971). The nutrient data showed thatat the onset of stationary phase, first nitrate and then silicate were exhausted.In terms of cell composition, the increase in C:N ratio was a clear indication ofnitrogen starvation, and has been noted in several species of diatoms (Dortch et a!. 1984), andin Phaeodactylum tricomutum cultures (Syrett et a!. 1986). In the case of the T. pseudonana,the increase of C:N ratio found by Dortch et at. (1984) was from 8.3 to 18, almost exactly the1 79same as in the present study. The ratio was driven largely by an increase in carbon quota. Aspreviously suggested (Chapter 3), this is indicative of cells which are still activelyphotosynthesizing, but can fix no more nitrogen (see also Syrett 1981, Davidson et al. 1993).Interestingly, however, comparable changes in cell volume were not seen. This may be anacclimation to nutrient stress, as discussed in Chapter 2. Hersey and Swift (1976) noted slightdecreases in protein on nitrate exhaustion in two dinoflagellate species, and a similar result hasalso been found in Chiamydomonas reinhardtii (Watt et al. 1992). Protein data were notcollected in the present study, but a decrease in nitrogen quota was seen which would beconsistent with a decrease in protein. Decreases in nitrogen quota on starvation for nitratewere also seen in T. pseudonana by Parsiow et al. (1984), in several diatom species (Dortchet al. 1984), and in Micromonas pusilla upon nitrogen starvation (Cochlan and Harrison1991b). In terms of chi a, little change was noted in the starving cultures. This was also trueof nitrogen-starved C. reinhardtii cells (Watt et al. 1992). According to data reviewed bySyrett (1981), a decrease in chl a in nitrogen-starved cells would be expected. It is possiblethat this may have occurred over longer time periods than were used in the present study.NR activity fell as stationary phase was reached, and this decline was comparable todeclines in the rates of nitrate incorporation and nutrient depletion. Three patterns of NRactivity in response to nitrogen starvation have been noted in the literature. In some cases, NRactivity increases after nutrient exhaustion. This has been found in Chiorella (Kessler andOsterheld 1970), C. reinhardtii (Watt et al. 1992), cyanobacteria (Bednarz and Schmid 1992)and in a survey of six species of marine phytoplankton (Hipkin et a!. 1983). In some cases,this increase even occurred in cultures that had previously been grown on ammonium and hadshown no NR activity previously. It is unclear why this would occur, but suggestions rangefrom a simple de-repression of nitrate reductase synthesis once ammonium is removed, to thepresence of oxidative pathways within the cell that provide nitrate in the absence of a nitrogensupply (Funkhouser and Garay 1981, Watt et a!. 1992). A second pattern of response tonitrogen starvation is a slight increase in NR activity in the first hours of nitrate exhaustion,180followed by a decline. This has been observed in Chaetoceros affinis (Slawyk and Rodier1986), and P. tricornutum (Syrett and Peplinska 1988). In the present study, this may havehappened, but daily sampling would not detect such a short term change. Many cells developa “rapid uptake” ability in this time period (see Hipkin et at. 1983, Slawyk and Rodier 1986,Cochian and Harrison 1991b, Dortch et al. 1991a, Martinez 1991). An increase in NRactivity may be part of this response, however, Parsiow et a!. (1984) reported thatdevelopment of nitrate uptake in T. pseudonana took 24 to 48 h after nitrate exhaustion, so thetime scales of these two processes may not match. Parsiow et at. (1984) also report nitriteexcretion into the medium following nitrate exhaustion, but this was not observed in thepresent study. Clayton (1986) also documents a high degree of cellular reprocessing ofnitrogen after nitrate exhaustion in S. costatum, in which NR might play some role. A thirdpattern observed is a constant decline in NR activity after nitrogen depletion. This can be agradual process that happens on a longer time scale than the cessation of nitrate uptake, as seenin Gonyaulax polyedra (Harrison 1976), or in natural marine phytoplankton populations(Eppley et at. 1969), or it may be more rapid, and correspond to decreases in uptake rates asseen in Chiorella vulgaris by Morris and Syrett (1965), two dinoflagellates by Hersey andSwift (1976), or Chiorella soroidniana (Tischner and Lorenzen 1980). The decrease in NRactivity is thought to be the result of enzyme degradation (Syrett 1981). Hersey and Swift(1976) hypothesized that NR might be less stable in the absence of nitrate and so be susceptibleto degradation. In a cyanobacterium, Hererro et at. (1984) suggested that there were twostages to NR degradation: an oxidation of the enzyme in the absence of nitrate, followed byprotease attack. They noted that this was occasionally accompanied by an actual short termincrease in NR synthesis. Aside from Morris and Syrett (1965) and Hersey and Swift (1976),NR activity in cultures depleting nitrate has usually not been well correlated with decreases innitrate incorporation rates. As previously noted, however, in the case of Morris and Syrett(1965), NR activity was only sufficient to account for 10-12% of the actual incorporation. Inthe present study, declines in incorporation and NR activity are closely and quantitativelymatched.181Effects of ammonium and ammonium pulsingIn general, the ammonium pulsed cultures behaved exactly as the nitrate grown culturesin all respects. It is possible that the ammonium pulses were too low to have any effect, sincethere appears to be a threshold for inhibition effects (Syrett 1981, Dortch 1990). At thegrowth rates observed, ammonium would have been taken up within 2-3 h after the addition.Ammonium grown cells had larger cell volumes and carbon quotas. Since ammoniumgrown cultures also grew more quickly, these increases in cell volume and carbon quotas maybe a reflection of increased growth rate, as previously discussed (Chapter 3). Nitrogen quotasand C:chl a ratios were no different. These results are identical to those found for ammoniumgrown versus nitrate-grown cultures of T. pseudonana, grown at saturating irradiance byThompson et al. (1989). However, Thompson et a!. (1989) also found higher C:N ratios andhigher chl a quotas at light saturation in ammonium grown cells, results that were not observedin the present study. In contrast, Darley (1977) summarized data from the diatom Ditylumbrightwellii showing that nitrate grown cells contained more carbon, nitrogen and lipid thanammonium grown cells, although growth conditions in the majority of these experiments werenot well documented. Paasche (1971) reported higher protein quotas in ammonium- versusnitrate-grown Dunaliella tertiolecta, but although numerically larger protein quotas were seenin ammonium-grown cells than in those grown on nitrate in the present study, they were notsignificantly different. In higher plants, increases in protein content when grown onammonium has also been found (Barraro et a!. 1989). Flynn (1990) suggests that nitrategrown cells are more nitrogen-stressed than those grown on ammonium. The data from thepresent study suggests that nitrate grown cells do share some of the characteristics of nitrate-starved cells, when compared with ammonium-grown cells.Ammonium grown cultures also grew faster than nitrate or pulsed cultures. This wasalso found by Paasche (1971) in Dunaliella tertiolecta, and Thompson et a!. (1989) in T.pseudonana, and is consistent with the idea that because ammonium is a more reduced form ofnitrogen, it is therefore less energetically costly to grow on than nitrate (Syrett 1981).182However, Thompson et a!. (1989) found that growth rate differences were only seen atsaturating irradiance, and in the present study irradiance was close to, but probably notsaturating. Arguments about energy advantages of ammonium over nitrate should not hold athigh light when energy is no longer limiting. Thompson et a!. hypothesized that there may becompetition between photosynthesis and nutrient uptake for reductant within the cells, whichcould result in slower growth and perhaps larger carbon quotas if photosynthesis were moresuccessful at using reductant. NR activities were very close to nitrate use rates and calculatedrates of incorporation in both nitrate and ammonium pulsed cultures. The ammonium pulsesappeared to have little effect. In one ammonium culture (NH4-1), where ammonium did notbecome depleted, no NR activity was observed. This is consistent with the vast majority ofthe literature that suggests that ammonium completely inhibits NR activity above 1-2 M (seeSyrett 1981, 1989). However, Zehr eta!. (1989) reported that cultures of T. pseudonana andDunaliella tertiolecta grown on ammonium were able to take up and reduce nitrate, whichsuggested that these species possess a constitutive for of NR. Zehr et al. (1989) speculated thatthey were able to detect nitrate reduction where other studies had not because their methodused the radioisotope ‘3N, as opposed to less sensitive assays for NR activity. In the otherculture (NH4-2), ammonium was exhausted and significant NR activity was found. Larsson eta!. (1985) found similar results in Scenedesmus obtusiusculus: after ammonium exhaustion NRactivity rapidly appeared. However, in this case, there was also excess nitrate in the medium.Alternatively, Morris and Syrett (1965) and Zeiler and Solomonson (1989) found that inChiorella vulgaris, an increase in NR activity followed ammonium exhaustion even when therewas no nitrate present. Solomonson and Barber (1990) concluded, based on several species,that nitrate may not be necessary to induce synthesis of NR, but that removal of ammoniumrepression is sufficient.Although not considered in the present study, many other authors have investigated theeffects of ammonium additions to cultures growing on nitrate. It appears that in the shortterm, ammonium may shut down nitrate uptake before it has an effect on NR activity.Pistorius et a!. (1979) found that in Chiorella, nitrate uptake ceased within 5 mm of the183ammonium addition, but NR was still active up to 60 mm later. Similar results have beenfound by Tischner and Lorenzen (1979), and for higher plants as well (see Ingemarrson(1987), Lee and Drew 1989). Serra et a!. (1978a) working with S. costatum, showed that thedecrease in MR activity was not simply an arrest of enzyme synthesis; additions ofcyclohexamide to block protein synthesis did not show the same degree of inhibition.Alternatively, Hersey and Swift (1976) found that the decrease in MR was very well correlatedwith loss of nitrate uptake ability in two dinoflagellates pulsed with ammonium. In othercases, the situation is not so clear. Larrson et a!. (1985) found a rapid cessation of nitrateuptake, but a slow decrease in MR activity in Scenedesmus obtusiusculus only when cultureswere bubbled with CO2. When cultures were bubbled with air instead, MR activity and nitrateuptake closely paralleled one another. No satisfactory explanation could be found, but Collos(1989) also reported that ammonium inhibition was more severe in C02-bubbled versus air-bubbled cultures. Flynn (1990) has proposed that an interaction with carbon, as previouslydiscussed (Chapter 3) may be involved. Furthermore, as McCarthy (1981) points out,simultaneous use of nitrate and ammonium has been noted in many cases, includingdinoflagellates (Harrison 1976). In the macroalga Ulva rigida, ammonium did not eliminatenitrate uptake or MR activity even when it was supplied at 3-4 times the concentration ofnitrate (Corzo and Neill 1991). Tischner and Lorenzen (1979) reported that ammoniumadditions could actually induce MR activity in Chiorella sorokiniana. There are even reportsof nitrate inhibition of ammonium uptake (see Collos and Slawyk 1980), and nitrate preferenceover ammonium (Proctor 1957, Dortch 1990). Although it has been held that ammoniumconcentrations greater than 1-2 M will inhibit nitrate uptake (see e.g. Packard and Blasco1974, Syrett 1989), Dortch (1990) extensively reviewed the field data from marineenvironments and concluded that the evidence did not support this point of view. Moreover,the distinction between “inhibition” and “preference” has rarely been adequately assessedbecause the uptake rates with either ammonium or nitrate alone and both together must bemeasured. True cases of ammonium inhibition (i.e. a specific direct effect on nitrate uptake184versus indirect effects of ammonium preference) were most often seen at low light, or whennitrogen was sufficient (Dortch 1990).It appears that the inhibition of nitrate incorporation by ammonium (whether at thelevel of uptake or reduction) is reflected in reduction in NR activity, although inhibition ofnitrate uptake can occur more rapidly than inhibition of NR activity. Changes in NR activityare likely mediated by changes in enzyme protein at longer time scales, but perhaps by aninactivation mechanism at scales of minutes to hours. Although at short time scales nitrate-uptake inactivation mechanisms may operate, which would be poorly reflected in NR activity,most measurements in the field are on much longer time scales. Thus, even under thesecircumstances, NR activity may be a useful index of nitrate incorporation rates.Implications of regulatory mechanismsThere is clear evidence that NR activity is regulated by more than one mechanism.There is the temptation to assume that all of these mechanisms are important and thatregulation must necessarily be more complicated than it appears. However, as pointed out byOttaway (1988), many enzyme regulatory mechanisms may be redundant (i.e. a “belt andbraces” situation), or they may be part of a “fossil record” of the way the enzyme wascontrolled at different stages in evolution. For nitrate reductase, an enzyme which may haveevolved for using nitrate as a terminal electron acceptor before becoming involved inassimilatory pathways, and especially if NR retains additional functions, these regulatoryconsiderations are legitimate concerns.In summary, despite temporal changes in nitrate incorporation rates caused byperiodicity in irradiance and nitrate starvation, NR activity closely followed these changes, atleast on a scale of hours to days. This suggest that NR activity could be useful in naturalenvironments with similar scales of variability. Different light spectra did not alter therelationship between nitrate incorporation and NR activity, except under red light, which is acondition unlikely to be found in ocean environments. Finally, although ammonium doesinhibit NR activity, the inhibition appears to be instep with changes in nitrate uptake. Pulses185of ammonium, which may be common in the marine environment (see Goldman 1986), do notappear to influence the relationship either. Based on the results of these experiments, itappears that in the majority of cases, MR activities can adequately predict rates of nitrateincorporation.186CHAPTER 5: ACTiVITY AN]) CHARACTERISTICS OF MTRATE REDUCTASE INNATURAL PHYTOPLANKTON POPULATIONS FROM MONTEREY BAY,CALIFORMAINTRODUCTIONIn previous chapters, work has focused on nitrate reductase activity in phytoplanktonmono-cultures. With some confidence in the NR assay (Chapter 2), and the relationshipbetween NR activity under steady-state and non-steady state conditions (Chapters 3 and 4), theresults of studies in which the NR assay was applied to field situations are described in thischapter. The question of where to try NR assays in the marine environment deserves carefulconsideration. Ideally, a prime location would have relatively few species, and preferablydiatoms, since there is interspecific variability in the NR assay, and the best relationshipsbetween NR activity and nitrate incorporation rates were found in diatoms. The biomass andgrowth rates should both be high to permit the greatest analytical sensitivity. Furthermore,high nitrate and low ammonium concentrations would simplify experiments at this stage.Coastal upwelling zones, and the California current upwelling system in particular, asdiscussed below, meet these criteria very well.Characteristics of coastal upwelling zonesThe general physical mechanisms of upwelling have been well described (Boje andTomczak 1978, Codispoti 1983, Parsons et at. 1984b, Valiela 1984, Mann and Lazier 1991).Coastal upwellings arise due to an interaction of winds, Coriolis force and coastal morphology.The California coast provides an excellent example. Wind circulation patterns in summer arefrom north to south, that is, along the coast. Due to Coriolis and frictional forces, an Ekmanspiral develops; as depth increases, ocean water is displaced at an angle to the wind direction,i.e. offshore (Pond and Pickard 1978, Mann and Lazier 1991). As a result, surface water ismoved offshore which results in a change in barotropic forces and results in deep off-shorewater being forced up along the coast (Codispoti 1983). This is not simply a dilution of187surface waters, as in other types of upwelling, but an actual replacement and can be recognizedby a decrease in surface temperatures. Typically, waters rise from depths less than 250 m, andare confined to relatively narrow coastal regions, on the order of 25 1cm, although thebiological effects of upwelling may influence a much wider region. In a review of theCalifornia current system, Bernal and McGowan (1981) showed that based on over 20 years ofdata, the area of nitrate enrichment is up to 300 km wide. Upwelling intensities are generallyon the order of i- cm s1, but vertical velocities of i02 to 10-1 cm s1 have beenmeasured. In temperate zone upwellings, water is colder and is rich in nutrients; often nitratelevels of > 25 M are measured. Ammonium and nitrite levels are usually low (< 2 PM),although in specific cases they may be higher where there are areas deficient in oxygen (e.g. inthe Peru upwelling, Codispoti and Packard 1980), or where turbid waters lead to reducedphotosynthesis and higher grazing and ammonium regeneration rates (e.g. Whitledge 1981).Because of the variability in wind intensity and direction, there is a very high degree of spatialand temporal variability in upwelling, which makes such regions very difficult to study.Furthermore, in addition to the coastal upwelling itself, there is evidence that cyclonic eddiescan form in the California current system, which result in localized upwellings (see Pond andPickard 1978) and increased spatial variability (Traganza et a! 1981). Other localizedphenomena include shear-induced turbulence, and island and seamount effects that causevertical mixing (Bernal and McGowan 1981). The importance of smaller scale upwellings hasonly been recognized since synoptic coverage using satellite images have been available (e.g.Traganza et a!. 1981).Upwellings may be particularly important in global nitrogen and carbon cycles. Sincenitrate is high and ammonium is low, new production (i.e. the primary production based onnitrate, see Introduction and Dugdale and Goering 1967) is high, with f-ratios on the order of0.7 to 0.75 (Eppley and Peterson 1979). Furthermore, they are critical regions for worldfisheries. The abundance of large diatoms (see Semina 1968, Parsons and Takahashi 1973,Hecky and Kilham 1974, Guillard and Kilham 1977), and the spatial variability, whichprevents grazing zooplankton populations from controlling primary producers, may result in188shorter food chains where energy transfer to higher trophic levels (e.g. fish) is very large(Ryther 1969).In biological terms, upwelling areas show distinct characteristics. Margalef (1967) firstdescribed in detail the pattern of species succession. MacIsaac et al. (1985) modeled thispattern by dividing upwelling regions into four zones, each further from the upwelling centre.In newly upwelled water (zone 1), nutrients are high, but phytoplankton biomass and growthrates remain low for some period after reaching the surface. Factors such as metal availabilityor toxicity are thought to play a role in this delayed response of cells to increased light (seeSunda et a!. 1981). In the second zone, cell division rates and photosynthetic rates haveincreased dramatically. The species found in this zone are principally diatoms, particularlyChaetoceros, Thalassiosira, and Skeletonema species (see also Guillard and Kilham 1977 foran excellent review of the particular species present in different geographic regions). In zone3, nutrients begin to become depleted, particularly nitrogen, although there is also evidencethat silicate may become limiting (see Dugdale 1972). The number of species increases, butthe abundance of individual species falls by a factor of 10-100. At this stage, large chain-forming diatom species dominate (e.g. Chaetoceros). In the last zone, nutrients are very low,biomass decreases and growth rates are again low. Diatoms decline in importance and arereplaced with large motile dinoflagellates, although some diatom species persist (e.g.Rhizosolenia, Hemiaulus and Mastogloia).Biological adaptation rate is thought to play a significant role in these systems.Wilkerson and Dugdale (1987) advanced a conceptual “conveyor belt” model, represented inFigure 5.1. In this model, the ability of species to “shift-up” (i.e. increase their specific ratesof nutrient uptake and growth) from previously limiting conditions on exposure to resources isa critical parameter. Cells from deep water have abundant nutrients, but are light-limited.They undergo a shift-189Figure 5.1. Diagram of phytoplankton processes in a coastal upwelling zone (after Wificerson and Dugdale 1987).Environmental conditions such as low light or low nutrients cause a decrease in the rates ofphytoplankton physiological processes (“shift-down”), while high light and high nutrients cause anincrease in rate processes (“shift-up”).190up as light increases after upwelling, then face a “shift-down” (i.e. a decrease in rates ofnutrient uptake and growth) as nutrients are depleted. After this, cells also undergo a secondshift-down in light as they sink from the euphotic zone (Fig. 5.1). The shift-up sequence hasbeen studied off Point Conception, California, where the entire cycle is completed in 5 to 7days (e.g. Dugdale and Wilkerson 1989). Using the stable isotope 15N as a tracer, biomassspecific nitrate uptake rates were at first low, but increased dramatically after upwelling.Carbon fixation rates followed nitrogen uptake increases but only after a slight time lag. Thephenomenon has also been demonstrated off the coast of Peru (e.g. MacIsaac et al. 1985), andWashington and Oregon (e.g. Dortch and Postel 1985, Kokkinakis and Wheeler 1987).However, Garside (1991) demonstrated, using a simple model, that the shift-up phenomenoncould also occur simply because uptake rates are normalized to particulate nitrogen (note thatDugdale and Wilkerson (1991) argue that this does not happen in practice; chi a-specific ratesshow the same pattern).The high spatial and temporal variability of upwelling zones constitute a seriousdisadvantage for study; repeated monitoring of populations over time is difficult. One solutionis to mark discrete parcels of waters (“drogues”) with drifter buoys, and follow them for someperiod of time (e.g. Wilkerson and Dugdale 1987). However, such an approach is rarelyconvenient, and some exchange with surrounding waters is inevitable. As an alternative,samples can be collected and maintained on deck under suitable temperature and lightconditions (e.g. Wilkerson and Dugdale 1987). Not only does this ensure that the samephytoplanklon assemblage is being sampled, but it makes repeated, frequent sampling verysimple. There are dangers that natural populations may respond differently when they arecontained than they do in situ (see the General Introduction), but it has previously been shownthat the trends in populations that have been contained follow those of cells in tracked droguesreasonably well (Wilkerson and Dugdale 1987).In this chapter, a field study was conducted in the upwelling region of Monterey Bay,California to test how well the newly modified NR assay (Chapter 2) worked in the field witha natural phytoplankton assemblage. This was achieved by comparing rates of nitrate uptake191and incorporation with NR activity in natural populations under near-natural conditions.Characteristics of NR in these populations, and the effects of diel periodicity and ammoniumadditions were also examined.MATERIALS AND METHODSData were collected aboard the R.V. Point Sur during May 1993, in conjunction withthe second cruise in the Shift-Up-93 program. Samples for assay optimization and NR activitycharacterization were taken during an initial survey off the coast of California in Monterey Bay(Fig. 5,2). Sampling for time series experiments was conducted on 11 May, 1993 at Station41(36° 47.77’ N, 121° 54.75’ W), indicated by the cross in Figure 5.2. At each samplingsite, vertical profiles of temperature, salinity, and in vivo chlorophyll a fluorescence weremade using a SeaBird CTD probe. Ten L Niskin water bottles equipped with silicone rubbersprings and fittings were used to sample water from the 50% light penetration depth (usually 3m).Modifications to NR assaysThe constraints of ship-board equipment and the low biomass relative to laboratorycultures made changes to the previous NR assay necessary. Filtration and homogenizationwere performed as described before, and the same extraction buffer was used (Chapter 2). Arefrigerated centrifuge was not available, and thus, homogenates were used directly. As aresult, the homogenates contained glass fibres from the filters. Homogenate volumes used inassays were corrected for the volume occupied by the filter fibres (see below). Time-stoppedassays measuring nitrite production were performed as described in Chapter 2, and sampleswere incubated at the in situ temperature in flowing seawater incubators on deck.Assay validation and enzyme characterizationIn order to determine whether the use of uncentrifuged homogenates made a difference,a preliminary experiment was conducted. The stainless steel rotor tubes of a clinicalULatitudej,,.,—z-—0(o-to-occ0>-m:N:::.-:s:::::.-:::::::::::0—‘.:.-;.-.:.....v...6T0,193centrifuge, and 15 ml glass centrifuge tubes were held on ice until immediately before use.Six 500 ml samples were filtered and homogenized. Three were used directly for assays,while three were centrifuged for 5 mm at full speed, and the supematant then used in NRassays. Temperature in the chilled tubes remained below 4°C during this procedure. Sincethe homogenates of the uncentrifuged samples contained glass fibres from filters, a volumecorrection for these samples was necessary. The volume of glass fibres was estimated fromthe scale on the glass centrifuge tubes in the three centrifuged homogenates, and averaged 0.2ml for a filtered sample homogenized in 1 ml of extraction buffer. The activities incentrifuged and uncentrifuged samples were compared using Student t-tests (Steel and Tome1980).For subsequent experiments, assays were conducted on uncentrifuged homogenates. Asin Chapter 2, the effect of different substrates and activators of NR were verified. Threeseparate homogenates were prepared and assayed with: a) 0.2 mM NADPH in place ofNADH, b) with addition of 0.1 mM flavin adenine dinucleotide (FAD), or c) after incubationwith 0.2 mM ferricyanide (FeCN). Results were compared using a one-way ANOVAprocedure, followed by Tukey’s multiple comparison technique at the 95% confidence level.To verify linearity of the assay with time, replicate samples from a single homogenatewere assayed for periods of 0, 15 or 45 mm. Nitrite produced was plotted against theincubation length and the data was analyzed by linear regression. To test linearity of the assaywith the amount of homogenate added, NR activity was measured in samples with between 0and 800 l of homogenate added. Again, results were analyzed using linear regression. Insubsequent experiments, 30 mm incubations were used and homogenate additions were 500 l.Enzyme kinetics were studied by performing assays of NR activity in replicate samplesfrom single homogenates, and varying NADH concentration (0-0.4 mM) or KNO3concentration (0-20 mM). Enzyme kinetic constants were estimated as before (see Chapter 2,Appendix C).To investigate possible degradation of NR activity with time, three individual sampleswere assayed immediately after homogenization, and at 15, 30 and 60 mm after194homogenization. NR activities were compared using a one-way repeated measures ANOVA,followed by Tukey’s multiple comparison technique at the 95% confidence level.Containment experimentsAt Station 41, samples were collected from the 50% light penetration depth and placedinto four acid-cleaned 20 L polyethylene containers (LMG Reliance). Two of the containersreceived no additions (control). The other two received additions of ammonium chloride tobring ambient concentrations up to 5 M.Initial samples (t = 0 h) were taken directly from the Niskin sample bottles.Containers were placed in a seawater-cooled deck incubator and irradiance was adjusted to50% of surface irradiance using neutral density screening. Sampling was repeated atapproximately 4 h intervals for 32 h.At each time, samples for nutrient analyses (nitrate and ammonium), particulatenitrogen, and chi a were taken. Ammonium samples were analyzed by hand within 24 h usingthe method of Parsons et al. (1984a) and measuring samples in a 10 cm cuvette in a HewlettPackard model 8452A spectrophotometer. A 50°C water bath was used to accelerate colourdevelopment. Samples for nitrate were frozen for later analysis using a TechniconAutoAnalyzer II (see Freiderich and Whitledge 1972). Single 460 ml samples were takenfrom each container and NR assays performed within 1 h. Nitrate uptake experiments wereperformed with samples collected in 280 ml polycarbonate bottles. Incubations for 15N uptakeexperiments were started by adding 2.0 M 15N-labeled nitrate (99 atom %, CambridgeIsotope Laboratories). Enriched samples were placed in deck incubators for 4-5 h.Incubations were terminated by filtration onto 25 mm pre-combusted GF/F filters, and frozenfor later analyses. Dried filters were analyzed for 15N enrichment and particulate nitrogen(PN) content with a Europa Scientific RoboPrep Tracermass mass spectrometer. Particulatenitrogen-specific uptake rates (VNO3) were estimated according to Dugdale and Wilkerson(1986). NR activity was normalized to particulate nitrogen, and particulate nitrogen-specific195rates of nitrate disappearance (A NO3-) and increase in particulate nitrogen (APN) were alsocalculated.As well, samples for species determination were taken before and after the experiment,preserved in 2% formalin, and qualitatively examined to determine the dominant taxa.RESULTSThe cruise took place during an upwelling event; there were high winds for the firsttwo days, then conditions calmed. Figure 5.3 A shows a vertical profile of the water columntypical of the early part of the cruise. Thermal stratification was weak and the water columnwas relatively well mixed. There was a small peak in biomass, indicated by in vivofluorescence, near 10 m. By the sixth day, there was evidence of stratification (i.e. wanner,less dense water near the surface) and structure in both salinity and temperature profiles (Fig.5.3 B). The fluorescence peak was greater, and had deepened to almost 20 m, indicating anincrease in phytoplankton. Microscopic examination showed that the species in this bloomwere predominantly diatoms, with approximately 80% of cells being (in order of abundance)Chaetoceros, Rhizosolenia, Skeletonema, and Nitzchia spp.Assay validation and enzyme characterizationNR activity in uncentrifuged homogenates was not different from activity in samplesthat had been centrifuged (Fig. 5.4 A, P > 0.5). NR activity using NADPH wasapproximately 15% of the activity with NADH as electron donor (Fig. 5.4 B). FAD additionsgave numerically higher NR activity, but this was not significantly greater than NR activitywithout FAD (Fig. 5.4 B, P > 0.2). On the other hand, FeCN addition gave significantlylower NR activity (Fig. 5.4 B, P < 0.00 1).In the NR assay, nitrite production was linear up to at least 45 mm (Fig. 5.5 A).Homogenate additions of up to 400 1il gave linear results, but there was evidence of nonlinearity at additions of 800 l (Fig. 5.5 B).196E34.0salinity (ppt)I I I I0 1 234relative fluorescence34.0salinity (,ppt)I I I I0 1 23 4relative fluorescenceI5Figure 5.3. Profiles of temperature (•------), salinity ( ) and relative fluorescence( ) at sampling sites in the vicinity of the main sampling station (indicated bythe “X” symbol on Fig. 5.2) in Monterey Bay, CA. Profiles were taken: A) onday 3, early in the bloom, and B) on day 6 at the height of the bloom.temperature (°C)5 10 15 5temperature (°C)010 15010203010203040 405033.55033.534.5J534.5197C6C— 128.0zFigure 5.4. Effects of different assay conditions on nitrate reductase activityin natural phytoplankton populations sampled from Monterey Bay, CA.A) NR activity in homogenates used directly, or centrifuged to removefilter fibres. B) Effects of additions of NADPH in place of NADH, 0.1 mMFAD, or 0.2 mM ferricyanide (FeCN). Each bar represents the mean of threereplicate homogenates. Error bars represent standard errors of the mean.Acentrifuged uncentrifugednormal NADPH FAD FeCN1980Ea)C.)0a).—.—C.1.—0z0 10 20 30time (mm)40 5040030020010001086_4200 200 400 600homogenate added (id)Figure 5.5. NR assay validation in natural phytoplankton samples taken fromMonterey Bay, CA. A) Linearity of assay with time. B) Linearity of assaywith homogenate addition. Lines represent least squares regression fits tothe data. Note in B) the open point ( 0 ) is not included in the regression.Each point represents a single determination.800199Enzyme kinetic analyses showed inhibition of NR activity at NADH levels of 0.4 mM,and also lower NR activity at 0.2 mM levels (Fig. 5.6 A). Km values for NADH were 0.021mM, whether calculated with or without the 0.2 mM point. For KNO3,a Km of 0.307 mMwas calculated (Fig. 5.6 B).NR activity had significantly declined by 30 mm after homogenization (Fig. 5.7).Since the assay itself took 30 mm, the resolution of this decline was poor. No difference wasfound between the initial NR activity and activity in assays performed after 15 mm (P > 0.5).Containment experimentsRates of increase of chl a and particulate N were identical in control and ammoniumspiked containers over the 36 h of the experiment (Fig. 5.8 A, B). Growth rates () wereestimated at 0.82 d1 and were the same whether based on increases in chl a or particulate N.Over the same time period, nitrate concentrations decreased in both sets of containers, butnitrate fell to lower concentrations in control containers (Fig. 5.8 C). Ammonium graduallyincreased with time in control containers, reaching 0.45 M by 36 h (Fig. 5.8 D).Ammonium declined steadily in containers with ammonium added until 28 h whenconcentrations were no different from control containers.NR activity showed a diel periodicity in both control and ammonium-spiked containers,but because ammonium also had an effect on NR activity, the trend was clearest in controlcontainers (Fig. 5.9 A, B). Activity was low at the beginning of the light period, rose to apeak and then declined by the beginning of the dark period (Fig. 5.9 A). Towards the end ofthe dark period, NR activity increased, but fell once again by the first sampling of the nextlight period.NR activity in ammonium-spiked containers decreased relative to activity in controlcontainers. With the exception of the 8 h sampling time when NR activity was not differentfrom the control, MR activity in the ammonium-spiked containers was always lower thancontrols, and declined continuously to undetectable levels by 20 h (Fig. 5.9 B). ThisC—I0zC—C)zI I I200..864-.2-.A0-I I I0.0 0.1 0.2 0.3 0.4 0.5NADH (mM)I I I I I8- -6- -4- -B0 5 10 20KNO3 (mM)Figure 5.6. Kinetic curves for nitrate reductase activity in natural populationsof phytoplankton from Monterey Bay, CA. A) NR versus NADHconcentration, B) NR versus nitrate concentration. Curves are fit torectangular hyperbolae. Km values are 0.021 mM for NADH and 0.307 mMfor nitrate. Each point represents a single determination.I I I I I1520125I—E.15>0105time (h)Figure 5.7. Nitrate reductase activity in natural phytoplankton populationsassayed at different times after homogenization. Each point representsthe mean of two enzyme assays from different homogenates.Error bars represent standard errors of mean values.I I0 20 40 602023OE10564+2z0_____________0 10 20 30time (h)Figure 5.8. Changes in biomass and ambient nutrient concentrations incontained natural phytoplankton populations from Monterey Bay, CAfor control cultures ( • ) and cultures with 5 M ammonium added( 0 ). A) Chlorophyll a, B) particulate nitrogen, C) nitrate, andD) ammonium. Each point represents the mean of two separatecontained cultures. Error bars represent standard errors of the mean,or where absent, they are smaller than the symbols. Cultures were grownunder natural light and the black bar on the time scale indicates thedark period.I ID2032o1•—‘%---.4-;z —E_ .EEcl)1‘—‘-‘ 21.t < C°100 10 20time (h)Figure 5.9. Changes in nitrate reductase (NR) activity, and specific rates ofnitrate incorporation calculated from changes in particulate nitrogen,changes in ambient nitrate concentration, or saturated uptake of 15N03(VNO.) for contained natural populations of phytoplankton from MontereyBay, CA. A) Control cultures without any nitrogen additions, and B) cultureswith 5 M ammonium added at t = 0 h. Symbols representmean values of determinations in two separate cultures. Error bars representstandard errors of the mean, or where absent are smaller than the size of thesymbols. Cultures were grown under natural light, and the black bar on thetime scale indicates the dark period.AB30204corresponded to the time when ammonium concentrations fell below 2 M (Fig. 5.8 D). Afterthis, NR activity increased and was not different from activity in control containers at the lasttwo sampling times.Diel patterns of APN, AN03 and VNO3 were somewhat different from those foundfor NR activity in control containers. These rates were highest at the first sampling of the dayand fell continuously during the light period (Fig. 5.9 A). There were increases in ratesduring the dark period, but activities tended to reach peaks at the beginning of the light versusthe end of the dark period; the same pattern was observed in NR activity.Estimates of nitrate incorporation rates (APN, iNO3 and VNO3) generally agreedwell with each other in the control containers, although AN03 rates did exceed otherestimates between 20-30 h. In ammonium-spiked containers, different estimates of nitrateincorporation rates also agreed at most points in time. There was significant nitrateincorporation in the presence of ammonium until the 12 h sampling. AN03 and VNO3 werezero at 12 and 16 h samplings. APN rates did not fall to zero, indicating that growth wasbeing supported by ammonium incorporation over this period. After 16 h, when ammoniumwas still greater than 2 ,1M, rates of nitrate incorporation rose again.NR activity was equal to or greater than nitrate incorporation rates in all but one casefor control containers (Fig. 5.9 A), and in all but 2 cases for ammonium-spiked containers(neglecting APN rates, Fig. 5.9 B). On average, NR activity was 254% of zPN rates, 285%of tNO3 rates, and 285% of VNO3 rates in control containers, while in ammonium-spikedcontainers, NR activity averaged 139% of LPN rates, 166% of zNO3 rates, and 233% ofVNO3 rates. However at individual sampling times (e.g. t = 0 h), the rates agreedconsiderably better (Fig. 5.9 A). In fact, at the first sampling of each day (t = 0, t = 24 h,both occurring at 10: 00 local time), the NR activity agreed well with the other estimates.205DISCUSSiONAdequacy of the NR assayNR activity and assay characteristics in natural populations showed many similarities toresults using laboratory cultures. Use of uncentrifuged samples probably resulted in some lossof precision because the correction of homogenate volumes for the presence of filter fibres wassomewhat crude, but despite this problem, assay results were no different whetherhomogenates were centrifuged or not. The assay was linear with time and linear withhomogenate added for all but the highest additions. Since high concentrations of NADH andKNO3 were added, this non-linearity was unlikely to be the result of substrate depletion, butcould be due to end product inhibition. Particularly in dense blooms, attention must be paid tothis factor or NR assays will underestimate activity. Activity of homogenates held on icedeclined in 30 mm. This suggests that assays, which were conducted over 30 mm, may haveunderestimated true NR activity. The magnitude of this error would be difficult to judge, butthe fact that activities at 0 and 15 mm were not different gives some confidence in the assay.The Km for NADH estimated in this chapter is very similar to that calculated forThalassiosira pseudonana cultures in the present study and the majority of algal speciesexamined by others (see Chapter 2 and Table 2.5). For nitrate, Km values appear to besubstantially higher than those found for T. pseudonana (Chapter 2), but closer to values foundfor S. costatum in culture in the present study, and by other researchers (see Chapter 2 andTable 2.5). NR kinetic constants have previously been determined in field populations, butonly for nitrate. Packard and Blasco (1974) reported Km values of 0.3 14 mM in a mixedpopulation of Gonyaulax polyedra from an upwelling off Northwest Africa. This is close tothe value found in the present study, but the Monterey Bay population was diatom-dominated.K values for nitrate in natural diatom-dominated populations averaged 0.082 (±0.052) mM,with a range of 0.042 to 0.201 mM (Packard and Blasco 1974). There is a possibility thatintracellular nitrate may bias the calculation of kinetic constants, but in the present study, NR206activity in samples without added nitrate was no different from blanks, suggesting that this didnot happen. In terms of the assay, 10 mM nitrate additions appear adequate, but there may bea problem with NADH levels used; 0.2 mM apparently caused some degree of inhibition.Inhibition of NR activity in certain species was seen previously at concentrations of NADH 0.4mM or greater (see Chapter 2), but it may be that inhibition occurs at lower levels in differenttaxa. Based on data presented in Figure 5.6 A, the problem of NADH inhibition does notappear to be severe, but it is critical that this be verified in each field study. This has not beendone in the past.Die! PeriodicityThere was strong evidence of diel periodicity in NR activity in control containers. Thepattern was remarkably similar to that found for laboratory cultures of T. pseudonana inChapter 4, despite the fact that sampling frequency in the field study was 4-5 h versus 3 h inthe laboratory study. Diel periodicity has previously been investigated in natural populations.Packard and Blasco (1974) noted a 100% increase in NR activity just before dawn in naturalpopulations from a California upwelling, and Eppley et al. (1970) found evidence of amidnight low in NR activity and a pre-dawn rise, but, as discussed in Chapter 4, thesedifferent patterns are likely due to different or irregular sampling times. For example,evaluating the data from Eppley et at. (1970) more closely, there is some evidence of a noonpeak in activity as well as the pre-dawn rise, but the sampling was intermittent, with gaps ofmore than 6 h. Manasneh and Basson (1987) found that diel periodicity in natural populationsin the Red Sea could be accounted for purely on the basis of biomass changes, evidentlycoupled to diel vertical migration of dinoflagellates. This was not the case in the present study(note that NR activity in Figure 5.9 is normalized to particulate N). Packard et al. (1971a)also found that diel periodicity in populations from upwelling areas did not correlate withbiomass. Irregular sampling cannot be invoked as an explanation for the pattern seen byMartinez et al. (1987). In this study, samples were collected every 1-2 h over a diel cycle, but207no pre-dawn rise in NR activity was seen, and a double peak in activity was found on eitherside of solar noon.Collos and Slawyk (1976) showed that did periodicity in NR activity was wellcorrelated with internal nitrate concentration, and nitrate uptake, although these were greater inmagnitude than NR activity. Diel periodicity in nitrate uptake and incorporation were alsofound in the present study. The patterns were somewhat different from those found for NRactivity, but comparisons are difficult; NR activity is based on an assay at a single point intime, while the other rates are averages of 4-5 h periods,The diel pattern was less obvious in ammonium-spiked containers. Collos and Lewin(1974) could not detect a diel periodicity in NR activity in populations of surf zone diatoms,and they attributed this to a diel periodicity in ammonium concentration which may haveobscured the pattern of NR activity.Effects of ammoniumAmmonium additions at the 5 MM level caused a decrease in nitrate incorporation rates;however, the effect was relatively minor compared to laboratory results (see Chapter 4);nitrate uptake rates in ammonium-spiked containers were oniy slightly lower than those incontrol containers after 4 h, and only at 12 h did rates drop to near zero. Furthermore, at 20 hnitrate uptake rates began to rise again, when there was still 2 MM ammonium present. Theseresults re-enforce the conclusions of a review by Dortch (1990), and argue against the conceptthat nitrate uptake is always inhibited by ammonium in natural populations (Syrett 1981). Ashas been shown in many studies with higher plants, the responses of nitrate uptake toammonium depend to a high degree on previous growth conditions and prior exposure toammonium (Clarkson and Luttge 1991).NR activity was apparently slower to respond to the introduction of ammonium thanwas the rate of nitrate uptake (see discussion in Chapter 4), but this was difficult to resolve dueto the different time scales involved (i.e. the instantaneous enzyme measurement versus 4-5 haverage uptake rates). Evidence of ammonium inhibition of NR in natural populations is208mostly limited to observational, rather than experimental data. For example, Packard andBlasco (1974) present data showing that a decline in NR activity correlated with a decline innitrate uptake with increasing ammonium concentrations above 0.2 M in samples fromcoastal Greece. However such correlations do not prove a causal relationship. Furthermore,as Dortch (1990) has pointed out, true ammonium inhibition of nitrate uptake (i.e. an indirectinteraction in which decreases in nitrate uptake vary with ammonium concentration) cannot bedistinguished from a simple preference for ammonium (i.e. a direct interaction in whichdecreases in nitrate uptake are independent of ammonium concentration) under theseconditions.NR activity and nitrate incorporation ratesStudies comparing NR activity and nitrate uptake or nitrate incorporation rates havegenerally shown reasonable correlations, but poor quantitative relationships. In most cases,NR activity is too low to account for observed nitrate incorporation rates. For example,Eppley et al. (1970) found good correlations between uptake and NR activity, but NR activityonly accounted for an average of 15% of the nitrate uptake rates. Collos and Slawyk (1976)found reasonable correlations between uptake, incorporation and NR activity, but NR averagedonly 12% of the other rate estimates. In other cases, results are less consistent. Collos andLewin (1974) found NR activity was between 5 and 90% of nitrate uptake rates measured insurf zone diatoms. Working in freshwater lakes, Wynne and Berman (1990) demonstrated thatNR activity ranged from 1 to 50% of nitrate uptake rates. In a study using large shipboard-contained populations, Wilkerson and Dugdale (1987) found that NR activity was almost anorder of magnitude lower than rates determined by 15N uptake, while Dortch and Postel(1989) showed that 15N-nitrate uptake rates in coastal Washington and Oregon waters alwaysexceeded NR activity by at least a factor of 2. The sampling location may also play a role.Blasco and Packard (1974) showed that in data from Californian and Northwest Africanupwellings, NR activity and uptake rates measured with ‘5N were only well correlated inupper waters; the relationship became poorer with increasing depth.209Other authors have demonstrated correspondence between uptake and NR activity, buthave also shown variability in the relationship. Collos and Slawyk (1977) found that the ratioof nitrate uptake to NR activity ranged from 0.03 to 10.8 in natural populations in the CostaRica dome upwelling. They compared 12 h average rates of 15N uptake versus NR activitymeasured at noon, so the mis-match of time scales may be responsible for some of thevariability. Collos and Slawyk (1977) also tried to correct for internal nitrate pools byestimating an in vivo reduction rate. This gave a range of nitrate uptake:NR activity ratios of0.1 to 11.2, most of which were greater than 3 (i.e. NR activity only accounted for a third ofobserved nitrate uptake rates). Comparing ‘5N uptake to NR activity in hundreds of samplesfrom Peru and California upwellings gave significant relationships (Blasco et al. 1984,MacIsaac et a!. 1985). Furthermore, regressions analyses indicated that NR activity (onaverage) exceeded nitrate incorporation rates. This is encouraging, but the relationship onlyexplained 44 to 69% of the variance, and NR activity only equaled or exceeded uptake rates in10 of 23 cases. In contrast, in the present study, NR activity was almost always in excess ofthat needed to account for nitrate uptake and incorporation by a factor of 2, on average.There are a number of possible explanations for the differences between NR activityand incorporation rates. As previously noted, the difference in time scale between NR activityand the other methods is a critical factor. Trying to determine uptake rates on shorter timescales would not be practical; differences in PN or nitrate concentrations would quicklyapproach limits of detection. However, NR assays could be performed more frequently(perhaps hourly or even half-hourly), allowing a better comparison to be made. Changes inambient nitrate and 15N uptake may not be equivalent to NR activity if internal pools werebeing formed. However, if this were true, uptake rates would be expected to exceed NRactivity instead of what was actually observed. It could be hypothesized that nitrate uptakewas a limiting factor. However, given the facts that ambient nitrate in the present study wasalways high, and that half-saturation constants for nitrate uptake are very low (see Packard eta!. 1979), this explanation seems less likely. Furthermore, several previous studies in210upwelling zones have demonstrated the formation of internal nitrate pools (e.g. Collos andSlawyk 1976), indicating that nitrate uptake exceeded nitrate incorporation.In previous studies, a wide variety of other explanations have been considered based onthe premise that rates of NR activity are incorrect. Most of these studies have been forced toreconcile lower NR activity with higher nitrate uptake rates. It has been proposed that NR ismore a function of external nitrate concentration (Kristensen 1987), internal nitrateconcentration (Collos and Slawyk 1976, Dortch and Postel 1989), or simply biomass(Manasneh and Basson 1987) than of nitrate incorporation rates. Others have argued thatammonium inhibition (Blasco et al. 1984) or light limitation (Blasco and Packard 1974) affectthe relationship. Kristensen (1987) has also pointed out the high degree of species variability.From the results of the laboratory experiments (Chapters 2-4) and the diatom communityfound in the present field study, these possibilities appear to be less likely.The effects of “nutrient prehistory” are often invoked (e.g. Wynne and Berman 1990),but there is no clear understanding of the potential magnitude of such effects, or of the timescales of adaptation in NR activity. Blasco et al. (1984) concluded that NR activity was anindex of nitrate assimilation before the time of sampling while 15N provided an estimate ofnitrate assimilation in the 6 hours following sampling. In the present study this cannot besupported; there is no evidence of consistent time lags between 15N uptake rates, NR activitiesand increases in particulate N.Another factor that should be addressed is the potential for interference from organismsthat reduce nitrate in a dissimilatory pathway. For example, Packard et al. (1978) noted thatNR activity from deep water (250 m) of the Peru upwelling where oxygen was low was due tobacterial populations. However, this is unlikely to constitute a significant interference for anumber of reasons. Dissimilatory nitrate reduction occurs when bacteria use nitrate in place ofoxygen as terminal electron acceptor; it does not occur under aerobic conditions. Alllaboratory cultures were bubbled and photosynthesis should ensure that culture medium wassupersaturated with oxygen. Due to mixing, the surface waters of Monterey Bay were alsooxygen-rich. Even if dissimilatory nitrate reduction were occurring, it is unlikely that the NR211assay would detect bacterial nitrate reductase. NADH cannot be directly used by bacterial NR(see Stouthamer et at. 1980), and although assays exist where NADH is added, thesetechniques rely on an intact bacterial respiratory electron transport chain to provide electronsfrom NADH to the cytochrome moiety of the MR enzyme (see Stouthamer et a!. 1980). It hasbeen demonstrated that when cells are homogenized with a detergent such as Triton X-100 (aswas the case in the present study), the electron transport system is not functional, and NADHcannot support nitrate reduction (Stouthamer et al. 1980). Thus, the potential for bacterialinterference in the MR assay is minimal.Finally, it must also be noted that there are likely errors in the estimated rates of nitrateincorporation and uptake. As pointed out in Dugdale and Wilkerson (1986), there are severalpotential problems with the 15N methods used in the present study. Additions of nitrate aresaturating; thus rates must be regarded as representing a maximum potential rate, but this isunlikely to be a problem in the present study because ambient nitrate levels were always high.The incubation period is a problem, since containment effects become important, or a surgeuptake phenomenon may cause rates to be over-estimated, but incubations were short (4-5 h)and cells were not nutrient-deficient, so these may not be serious problems. There is also apotential loss of labeled nitrogen (see Kokkinakis and Wheeler 1987, Bronk and Glibert 1991,1993). This has been attributed to increases in the pool of dissolved organic nitrogen, andbacterial uptake, and appears to be a particular problem in oligotrophic waters whenincubations longer than 1 h are used