UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Calcium pyrophosphate dihydrate crystal deposition disease: characterization and IgG binding properties… Winternitz, Charles Ilderton 1994

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

Item Metadata

Download

Media
831-ubc_1994-0134.pdf [ 5.01MB ]
Metadata
JSON: 831-1.0087420.json
JSON-LD: 831-1.0087420-ld.json
RDF/XML (Pretty): 831-1.0087420-rdf.xml
RDF/JSON: 831-1.0087420-rdf.json
Turtle: 831-1.0087420-turtle.txt
N-Triples: 831-1.0087420-rdf-ntriples.txt
Original Record: 831-1.0087420-source.json
Full Text
831-1.0087420-fulltext.txt
Citation
831-1.0087420.ris

Full Text

CALCIUM PYROPHOSPHATE DIHYDRAT E CRYSTA L DEPOSITIO N DISEASE: CHARACTERIZATIO N AND IgG BINDING PROPERTIES O F MONOCLINIC CALCIUM PYROPHOSPHATE DIHYDRAT E CRYSTAL S by CHARLES ILDERTON WINTERNIT Z B.Sc. (Pharm.), The University of British Columbi a A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE i n THE FACULTY OF GRADUATE STUDIE S Faculty of Pharmaceutical Science s (Division of Pharmaceutics an d Biopharmaceutics ) We accept this thesis a s conforming to the required standar d UNIVERSITY O F BRITISH COLUMBI A January 199 4 Charles Ilderton Winternitz, 199 4 In presentin g thi s thesi s i n partia l fulfilmen t o f th e requirement s fo r a n advance d degree a t th e Universit y o f Britis h Columbia , I  agre e tha t th e Librar y shal l mak e i t freely availabl e fo r referenc e an d study . I  furthe r agre e tha t permissio n fo r extensiv e copying o f thi s thesi s fo r scholarl y purpose s ma y b e grante d b y th e hea d o f m y department o r b y hi s o r he r representatives . I t i s understoo d tha t copyin g o r publication o f thi s thesi s fo r financia l gai n shal l no t b e allowe d withou t m y writte n permission. Department o f fHM{fQ(lc£UTlCfil  bCiEh)C£S> The Universit y o f Britis h Columbi a Vancouver, Canad a Date /&  ¥,/m DE-6 (2/88 ) ii ABSTRACT Calcium pyrophosphate dihydrat e (CPPD) crystal deposition disease i s caused by CPPD crystals which ar e released into the synovial fluid of affecte d joints. Thes e crystals are highly surface activ e and ar e quickly coated by proteins present i n the joint cavity, particularly immunoglobulin G (IgG). Through a  series of events the coated crystals interact with neutrophil s causing their lysis and the resultant releas e of lysosomal enzymes into th e joint cavity, which leads to inflammation an d pain for the affecte d patient . Calcium pyrophosphates hav e been reported to exist in a variety of hydrate an d polymorphic forms including triclinic, monoclinic and hexagona l dihydrates (t-CPPD , m-CPPD and h-CPPD respectively), and a n orthorhombic tetrahydrate (o-CPPT) . I t was reported that h-CPPD forme d from th e dehydratio n of o-CPPT. A t room temperature an d pressure th e stable crysta l form i s t-CPPD and only t-CPPD an d m-CPPD crystals hav e ever been isolated from joints. Previou s efforts t o study the crystals whic h cause this disease have concentrated on the stable t-CPPD form eve n thoug h m-CPPD has als o been shown to have inflammatory potential . Th e purpos e of this work was to synthesize m-CPPD, o-CPPT and h-CPPD crystals an d characterize thei r physicochemical properties an d then to begin studying the role of m-CPPD crystals in CPPD crystal deposition disease by comparing the adsorption of IgG protein onto the surface o f m-CPPD and t-CPPD crystals . A reliable method for the synthesi s of m-CPPD crystals was obtained . Synthesis o f orthorhombic calcium pyrophosphat e tetrahydrat e (o-CPPT ) was also achieved but the method proved unreliable an d only a limited supply of o-CPPT crystal s was available . A s a result h-CPPD crystal s were no t iii synthesized. Th e t-CPPD, m-CPPD, and o-CPPT crystals were characterize d and compared using X-ray powder diffraction, ligh t and scanning electro n microscopy, an d Fourier transform infrare d spectroscopy . Differen t method s of calculation failed to confirm th e unit cel l dimensions of m-CPPD crystals . Both the m-CPPD and o-CPPT crystals were needleshaped whil e the t-CPPD crystals appeare d a s parallelograms under the microscope. Th e m-CPP D crystals had a  smaller particle siz e compared to t-CPPD by a factor o f about 10, meaning they had a  larger surface are a per unit volume than the t-CPP D crystals. Therma l analysi s using differential scannin g calorimetry confirme d the hydration level of the crystals . Zet a potentials for m-CPPD and t-CPP D crystals in deionized water were found to be -18.8 mV and -35.3 mV respectively. The solubilities of triclinic and monoclinic CPPD were measured i n terms of Ca2+ concentratio n in solution which was measured using either a n ion selective calcium electrode , atomic absorption spectroscopy, o r complexiometric titration with EDTA. I t was found tha t m-CPPD crystal s were 1 0 times more soluble at low pH than a t neutral pH. Th e solubility of both m-CPPD and t-CPPD crystals was found to increase with temperature . Throughout the entire temperature range studied m-CPPD crystals had a higher solubilit y than t-CPPD crystals confirming that m-CPPD is the metastable form o f CPPD. Throug h extrapolation o f the data using a Van' t Hoff plot the transition temperature between the two polymorphic forms wa s estimated to be 163 ° C. Adsorption of IgG to m-CPPD and t-CPPD crystals surfaces wa s studied using fluorescent labelle d FITC-IgG. Th e binding isotherms produced were shown to fit the Freundlich equation an d the bindin g constants were explained in terms of heterogeneous binding . iv TABLE OF CONTENT S Page ABSTRACT i i TABLE OF CONTENTS i v LIST OF TABLES i x LIST OF FIGURES x i LIST OF ABBREVIATIONS xv i ACKNOWLEDGEMENTS xvi i INTRODUCTION 1 1. CPPD crystal deposition disease 1 2. Synthesis o f CPPD crystals 4 3. Characterization o f crystals 5 3.1. X-ray powder diffraction (XRPD ) 5 3.2. Thermal analysi s 7 3.3. Zeta potential measurements 8 3.4. Solubility 9 3.5. Fourier Transform infrare d spectroscop y (FTIR) 1 0 4. Protein adsorptio n to crystals 1 0 4.1. Protein adsorptio n t o MSUM and CPPD 1 1 4.2. Protein adsorptio n to hydroxyapatite (HA ) 1 3 5. Objectives 1 4 5.1. Specific objectives 1 5 V EXPERIMENTAL 1 7 1. Materials 1 7 2. Equipment 1 9 3. List of Buffers 2 0 4. Methods 2 1 4.1. Synthesis of crystals 2 1 4.1.1. Calcium dihydrogen pyrophosphate (CDPP ) 2 1 4.1.2. Triclinic calcium pyrophosphate dihydrat e (t-CPPD) 2 2 4.1.3. Monoclinic calcium pyrophosphate dihydrat e (m-CPPD) 2 2 4.1.4. Orthorhombic calcium pyrophosphat e tetrahydrate (o-CPPT ) 2 3 4.1.4.1. Generation of first  seed crystal batches 2 3 4.1.4.2. Generation of the second seed crysta l batches 2 4 4.1.4.3. Further attempt s a t o-CPPT synthesis 2 4 4.1.5. Hexagonal calcium pyrophosphate dihydrat e (h-CPPD) 2 5 4.2. Characterization o f crystals 2 5 4.2.1. X-ray powder diffraction (XRPD ) 2 5 4.2.1.1. Determination of unit cel l parameters fo r m-CPPD 2 6 4.2.1.1.1. Growth of m-CPPD crystals 2 6 4.2.1.1.2. Calculation methods 2 6 4.2.2. Differential scannin g calorimetry (DSC) 2 7 vi 4.2.2.1. Weight los s determinations 2 7 4.2.2.2. Calculation o f heats o f dehydration 2 8 4.2.3. Fourier transfor m infrare d scan s (FTIR ) 2 8 4.2.4. Scanning electro n microscop y (SEM) 2 8 4.2.5. Zeta potentia l determination s 2 9 4.2.6. Particle siz e analysi s 2 9 4.2.7. Solubility determination s 2 9 4.2.7.1. Measurment o f calcium 2 9 4.2.7.1.1. Calcium electrod e 2 9 4.2.7.1.2. Atomic absorptio n spectrophotometry 3 1 4.2.7.1.3. Complexiometric titratio n wit h EDTA 3 1 4.2.7.2. Effect o f pH 3 2 4.2.7.3. Effect o f temperature 3 3 4.2.7.4. Determination o f possible phase change s during solubilit y studies 3 5 4.2.8. Effect o f relative humidit y 3 5 4.3. IgG binding t o CPPD crystal s 3 6 4.3.1. Measurement o f FITC-IgG 3 6 4.3.2. Incubation o f crystals wit h FITC-IgG 3 7 4.4. Statistica l analysi s 3 9 RESULTS 4 1 1. Synthesis o f crystals 4 1 1.1. Determination o f unit cel l parameters o f m-CPPD 4 2 1.1.1. Growth o f single crysta l 4 2 vii 1.1.2. Calculation methods 4 2 2. Characterization o f crystals 5 5 2.1. Thermal Analysis 5 5 2.1.1. Heat of dehydration 6 2 2.2. Fourier transform infrare d scan s (FTIR) 6 2 2.3. Scanning electron microscopy (SEM) 6 2 2.4. Zeta potential determinations 6 2 2.5. Particle siz e analysis 6 2 2.6. Solubility determinations 6 9 2.6.1. Measurment o f calcium 6 9 2.6.1.1. Calcium electrode 6 9 2.6.1.2. Atomic absorption 7 0 2.6.2. m-CPPD solubility dependence on pH 8 0 2.6.3. CPPD solubility dependence on temperature 8 0 2.6.4. Determination o f possible phase changes durin g solubility studies 8 0 2.7. Effect o f relative humidity 8 1 2.7.1. Calcium pyrophosphate dihydrat e 8 1 2.7.2. Anhydrous calcium pyrophosphate 8 1 3. IgG binding to CPPD crystals 9 7 3.1. Indirect method of measurement 9 7 3.2. Direct method of measurement 9 8 DISCUSSION 11 5 1. Synthesis o f crystals 11 5 1.1. Determination of unit cel l parameters 11 7 2. Characterization o f crystals 11 8 viii 2.1. Thermal analysis 11 8 2.2. Fourier transform infrare d scan s (FTIR) 12 0 2.3. Scanning electron microscopy 12 1 2.4. Particle siz e analysis 12 1 2.5. Zeta potential determinations 12 1 2.6. Solubility determinations 12 2 2.6.1. Comparison of calcium measurement methods 12 2 2.6.2. m-CPPD solubility dependence on pH 12 4 2.6.3. m-CPPD solubility dependence on temperature 12 5 2.6.4. Determination o f possible phase changes durin g solubility studies 12 6 2.7. Effect o f relative humidity 12 7 3. IgG binding to CPPD crystals 12 7 4. Summary 13 3 REFERENCES 13 4 ix LIST OF TABLE S Page Table 1 Table 2 Table 3 Table 4 Table 5 Table 6: Table 7: Table 8: Table 9: X-ray powder diffraction pea k data for t-CPPD 4 7 X-ray powder diffraction pea k data for m-CPPD 4 9 Production of o-CPPT crystals 5 0 X-ray powder diffraction pea k data fo r o-CPPT 5 2 Zeta Potential measurements fo r Minusil powder , t-CPPD, and m-CPPD crystals in distilled water a t room temperature 6 8 CPPD solublity of m-CPPD in glycine-HCl buffers o f different p H measured usin g the calcium electrod e 7 2 Regression equations for standard curv e data prepare d using calcium chloride or calcium di-orthophosphat e i n water or pH 2.70 glycine-HCl buffer, measure d using th e calcium electrod e 7 3 Comparison o f calcium concentration s determine d usin g different standar d curve s with the calcium electrod e 7 4 Effect o f incubation time on m-CPPD solubility in different glycine-HC l buffers a s measured by a calcium electrode a t 31° C 7 5 X Table 10 : Compariso n of calcium concentration s determined usin g standard curv e A (CaCl2«2H20 in pH 2.70 glycine-HC l buffer) an d standar d curv e B (calcium tetrahydrogen di -orthophosphate i n pH 2.70 glycine-HCl buffer) usin g atomic absorption spectroscop y 7 8 Table 11 : Effec t o f incubation time on CPPD solubility in distille d water a t 45° C as measured by atomic absorption 7 9 Table 12 : CPP D solubility dependence on temperature i n pH 2.7 0 glycine-HCl buffer measure d using Ca2+ electrod e 8 6 Table 13 : Temperatur e dependenc e o f CPPD solubility in glycine-HCl buffer a t pH 2.70 with the calcium ion concentratio n measured by atomic absorption spectroscop y 8 7 Table 14 : Temperatur e dependenc e of CPPD solubility in Tris-HC l buffer a t pH 7.20 and calcium ion concentration measure d using atomic absorption spectroscop y 8 8 Table 15 : CPP D solubility in distilled water, calciu m ion concentration measured by atomic absorption 8 9 Table 16 : Temperatur e dependenc e of CPPD solubility in glycine-HCl buffer a t pH 2.70 and calcium ion concentratio n measured using EDTA complexiometric titration 9 0 Table 16 : Indirec t binding data for FITC-IgG to CPPD crystals a t 37° C 10 2 Table 17 : Direc t binding data for FITC-IgG to CPPD crystals a t 37° C 10 5 Table 18 : Bindin g constants fo r FITC-IgG binding to m-CPPD an d t-CPPD a t 37° C derived from Freundlic h isotherm 11 4 xi LIST OF FIGURE S Page Figure 1 : Crysta l structure o f calcium pyrophosphate dihydrat e (CPPD) 2 Figure 2 : X-ra y powder diffraction patter n of calcium tetrahydroge n di-orthophosphate 4 4 Figure 3 : X-ra y powder diffraction patter n of calcium dihydroge n pyrophosphate 4 5 Figure 4 : X-ra y powder diffraction patter n of t-CPPD 4 6 Figure 5 : X-ra y powder diffraction sca n of m-CPPD 4 8 Figure 6 : X-ra y powder diffraction patter n o f a: o-CPPT from 5 ° to 40° 20 and b: from 10 ° to 40° 20 using an expanded scal e 5 1 Figure 7 : X-ra y powder diffraction scan s for a ) the crystal s though t to be partially transformed t o h-CPPD and b) th e crystal s whose pattern coul d not be identified 5 3 Figure 8 : Plo t showing the d-spacings observe d from  m-CPP D powder patterns, those calculated for monoclini c magnesium pyrophosphat e dihydrat e based on published single crystal data (Ok a and Kawahara 1982) , and thos e calculated for m-CPPD based o n reported uni t cel l dimensions (Mandel et al . 1988) 5 4 Figure 9 : Differentia l scannin g calorimetry scans for a : t-CPPD an d b: m-CPPD from 150 ° to 380° C at a  heating rate o f 10° per minute 5 7 Xll Figure 10 : Differentia l scannin g calorimetry scans for calciu m tetrahydrogen di-orthophosphat e from 50 ° to 250° C at a heating rate o f 10° per minute 5 8 Figure 11 : Differentia l scannin g calorimetry scans of a: calcium dihydrogen pyrophosphate an d b: o-CPPT from 30 ° to 500° C at a  heating rate of 10° per minute 5 9 Figure 12 : Differentia l scannin g calorimetry scans showing the effects o f storage on the hydration levels of o-CPPT crystals, a : an o-CPPT sample a t room temperature 6 0 Figure 13 : Differentia l scannin g calorimetry scans of crystals tha t resulted durin g attempts to synthesize h-CPPD. a : o-CPPT crystal s that were kept in a  vacuum ove n for 1 8 hours 6 1 Figure 14 : Standar d curv e showing the relationship between th e heat o f melting of indium standard s (kcal) , and the pea k area o f the melting endotherm produced on the differential scannin g calorimetry apparatu s use d to calculate AHdehydration for m-CPPD crystal s 6 3 Figure 15 : Fourie r transform infr a re d scan s for a : t-CPPD 6 4 Figure 16 : Scannin g electron microscope photograph o f t-CPPD 6 5 Figure 17 : Scannin g electron microscope photograph o f m-CPPD 6 6 Figure 18 : Scannin g electron microscope photograph o f o-CPPT 6 7 Figure 19 : Standar d curv e showing dependence of of calcium electrode readings on calcium ion concentration usin g CaCl2«H20 i n distilled water a s a calcium standard 7 1 Figure 20 : Standar d curve s for atomic absorption spectroscop y determinations o f calcium ion concentrations using fiv e different Ca 2+-buffer system s 7 6 Figure 21 : Th e 95% confidence interval s of the slopes obtained fro m the standar d curves which were determined using atomi c absorption spectroscop y and different Ca 2+-buffer systems 7 7 Figure 22: Effec t o f pH on the m-CPPD solubility given as log of the JWM Ca2+ concentration . Previousl y published dat a fo r t-CPPD crystal s (Bur t an d Jackson 1987 ) are shown 8 3 Figure 23 : Solubilit y (given as Ca2+ io n concentration) of t-CPPD an d m-CPPD in glycine-HCl buffer a t a  pH of 2.70 as a function o f temperature, measured using the calciu m electrode 8 4 Figure 24 : Van' t Hoff plot of the solubilit y data presented i n Figure 2 3 with the linear regression curve fits for t-CPP D and m-CPPD 8 5 Figure 25 : X-ra y powder diffraction sca n of m-CPPD incubated a t p H 2.68 for 1 7 hours a t 30°C and allowed to air dry overnigh t at room temperature 9 1 Figure 26 : X-ra y powder diffraction scan s of a: m-CPPD and b: t-CPPD crystals incubated fo r 4 days a t 60° C in glycine-HCl buffer a t pH 2.70 buffer an d allowe d to air dry a t room temperature 9 2 XIV Figure 27 : Differentia l scannin g calorimetry scans of a: t-CPPD an d b: m-CPPD crystal s which had been incubated in distille d water a t 45° C for 1  week, then ai r dried a t room temperature. Th e heating rate was 10 ° per minute an d ranged from 100 ° to 500° C 9 3 Figure 28: X-ra y powder diffraction scan s for anhydrou s calciu m pyrophosphates produce d by heating a : m-CPPD an d b: t-CPPD, in a 400° C furnace fo r 30 minutes 9 4 Figure 29: Differentia l scannin g calorimetry scans of a: m-CPPD an d b: t-CPPD crystal s which had been dehydrated in a 400° C furnace fo r 30 minutes fo r the relative humidit y experiment. Scan s were conducted from 30° to 500° C at a heating rate o f 10° per minute 9 5 Figure 30: Differentia l scannin g calorimetry scans of a: dehydrated t-CPPD an d b: dehydrated m-CPPD afte r the y had bee n stored in a  ZnS04«7H20 desiccato r a t 24° C (90 % relative humidity) fo r 5  months. Th e samples wer e scanned from  130 ° to 500° C at a  rate o f 10° C per minute 9 6 Figure 31 : Standar d curv e for the fluorescence intensit y of [FITC-IgG] solutions in Hanks balanced salt s solution a t 37° C 10 1 Figure 32 : Bindin g curve of FITC-IgG to m-CPPD crystals a t 37° C measured by the indirect method 10 3 Figure 33: Bindin g curve of FITC-IgG to t-CPPD crystal s a t 37° C measured by the indirect method 10 4 XV Figure 34 : Bindin g curve for FITC-IgG to m-CPPD crystals a t 37° C as measured by the direc t method. Inse t shows th e binding curve for FITC-IgG to m-CPPD a t FITC-Ig G concentrations o f 1 - 200 pig/mL. Erro r bars indicate ± 1 S.D 10 6 Figure 35 : Bindin g curve for FITC-IgG to t-CPPD crystal s a t 37° C as measured by the direct method. Inse t shows the binding curve for FITC-IgG to t-CPPD a t FITC-Ig G concentrations o f 1 - 200 jig/mL. Erro r bars indicate +  1 S.D 10 7 Figure 36: Scatchar d plo t of the direct binding data o f FITC-IgG to m-CPPD crystal s a t 37° C. Erro r bars indicate ± 1 S.D 10 8 Figure 37: Scatchar d Plot of the direc t binding data o f FITC-IgG to t-CPPD crystal s a t 37° C. Erro r bars indicate ± 1 S.D 10 9 Figure 38: Fi t of Langmuir isotherm t o direct binding data of FITC-IgG to m-CPPD crystals a t 37° C 11 0 Figure 39: Fi t of Freundlich isotherm t o direct binding data of FITC-IgG to m-CPPD crystals a t 37° C I l l Figure 40: Fi t of Langmuir isotherm t o direct binding data of FITC-IgG to t-CPPD crystals a t 37° C 11 2 Figure 41 : Fi t of Freundlich isotherm t o direct binding data of FITC-IgG to trCPPD crystals a t 37° C 11 3 xvi LIST OF ABBREVIATION S AA: Atomi c absorption spectroscop y ATP: Adenosin e triphosphat e Ca di-ortho: Calciu m tetrahydrogen di-orthophosphat e CDPP: Calciu m dihydrogen pyrophosphat e CPPD: Calciu m pyrophosphate dihydrate crystal s DSC: Differentia l scannin g calorimetry EDTA: Ethylenediaminetetraaceti c aci d FITC-IgG: Fluorescei n isothiocyanate conjugated immunoglobuli n G Fab: Antige n binding fragment o f immunoglobulin G protein Fc : Complemen t binding fragment o f immunoglobulin G protein FT1K: Fourie r transform infrare d h-CPPD: Hexagona l calcium pyrophosphate dihydrate crystal s HA: Hydroxyapatit e HBSS: Hank s balanced salts solutio n IgG: Immunoglobuli n G protein m-CPPD: Monoclini c calcium pyrophosphate dihydrate crystal s m-MPPD: Monoclini c magnesium pyrophosphate dihydrate crystal s MSUM: Monosodiu m urate monohydrate crystal s o-CPPT: Orthorhombi c calcium pyrophosphate tetrahydrate crystal s PMN: Polymorphonuclea r leukocyte s PPi: Pyrophosphat e ion PTFE: Pol y tetrafluoroethylen e SDS-PAGE: Sodiu m dodecy l sulfate -  polyacrylamide gel electrophoresi s SEM: Scannin g electron microscopy t-CPPD: Triclini c calcium pyrophosphate dihydrate crystal s XRPD: X-ra y powder diffractio n xvii ACKNOWLEDGEMENTS I would like to thank my supervisor, Dr . Helen Burt, for her patien t guidance an d encouragement throughou t the course of this work. I  would also like to thank my committee members: Dr . Stelvio Bandiera, Dr . Kathleen MacLeod, Dr. Alan Mitchell, and Dr. James Trotter , for their tim e and guidance . I would als o like to thank Ben Clifford an d Tony Fu from th e department o f Chemistry a t U.B.C. for their assistance with the atomi c absorption apparatu s an d the FTIR instrument respectively . Thanks goes also to the Health Research Foundation of the Pharmaceutical Manufacturer s Associatio n of Canada an d the Medical Research Counci l of Canada who funded a  portion of this work . To my mates in the basement and friends throughou t the building, a special thanks for making the time spent on the project enjoyable . To my family, fo r their encouragement an d suppor t I give my thanks. To my soon to be bride, Lorraine, for her warmth, her comfort , an d he r smile, I am thankful . For Joe Carter's World Series clinching home run, I  am happy . 1 INTRODUCTION 1. CPPD crysta l deposition diseas e In 196 1 Daniel McCarty discovered the condition that has sinc e become known as calcium pyrophosphate dihydrat e (CPPD) crystal deposition diseas e (Brown et al. 1963 ; Kohn et al. 1962) . Th e disease affect s synovia l joints an d may result in painful acut e attack s lasting from 1  day to 4 weeks, or it ma y produce chronic , arthritis-like symptoms . Althoug h other joints can be affected, th e knee joint is the one most commonly involved (McCarty 1976). Because of its clinical similarity to gout, which is caused by monosodium urate monohydrate (MSUM ) crystals, CPPD crystal deposition disease is frequently referre d t o as "pseudogout " (Ryan and McCarty 1989) . Th e syndromes associate d with the deposition of hydroxyapatite (HA ) and other basic calcium phosphates (BCP ) in the joints are termed BCP crysta l deposition diseases . CPPD crystals (Figure 1 ) form primaril y in the articula r cartilage of joints (McGuire et al. 1980) . Thes e crystals can then be released or "shed " into the synovia l fluid by a number of proposed mechanisms suc h a s underlying trauma to the bone, cartilage breakdown, or changes in synovia l fluid io n concentrations. I n the synovial fluid the crystals become coated wit h IgG, complement components, albumin an d other proteins . Thi s promote s phagocytosis of the crystals by polymorphonuclear leukocytes (PMN) or neutrophils, which have receptors for the F c portion of the IgG molecule. McCarty and Kozin (1975)proposed the "rupture from within " hypothesis fo r the interaction of CPPD crystals with neutrophils, a s follows. Th e phagocytosed crystal s are encased in newly formed phagosomes withi n 2 H 20 Ca 2+ Ca 2+ Figure 1 : Crysta l structure o f calcium pyrophosphate dihydrate (CPPD ) 3 the neutrophils . Followin g fusion o f lysosomes with the phagosome i t ha s been suggested tha t lysosomal enzymes within the phagolysosome diges t th e proteins bound to the crysta l surface . Wit h the crystal surface exposed , th e CPPD crystals bind to the phagolysosome membrane an d caus e it to rupture, a process referred t o as membranolysis. Th e resultant release of enzymes into the cytoplasm leads to neutrophil cytolysi s with release of enzymes into the extracellula r fluid an d subsequent tissue damage (Kozin and McCart y 1976a; McCarty and Kozin 1975). Although only the triclinic and monoclinic forms o f CPPD have been found in  vivo,  man y other calcium pyrophosphates hav e been identifie d (Brown et al. 1963) . Calciu m pyrophosphate ca n exist as the anhydrous for m and a s a t least two known hydrates, al l of which show different polymorphi c forms. O f the two forms of CPPD found in joints, the triclinic form appear s to be the more stable form sinc e in water a t 45°C the monoclinic form i s reported to transform int o the triclinic form (Brown et al. 1963) . Accordingly , the triclinic form predominates in synovial fluid by a 3:1 ratio over the monoclinic form, althoug h only the monoclinic form ha s been found in th e dura mater (Mandel et al. 1988) . To date, most studies of this disease have concentrated on the more stable triclinic form o f CPPD. However , the monoclinic form ha s been shown to possess significan t inflammator y potentia l (Roch-Arveille r et al. 1990 ; Watanabe et al. 1992 ) and is present in affected joints. A s a result, it s possible contribution to the CPPD crystal deposition disease process shoul d not be ignored . Of central importance to future studie s of the mechanism o f m-CPPD induced inflammation i s the use of crystals with well characterized bulk an d surface properties , and for which the protein binding characteristics ar e 4 understood. Hence , the overall objective of this work is to synthesiz e m-CPPD crystal s and determine the physicochemical an d protein bindin g properties o f the crystals . 2. Synthesis o f CPPD crystal s In order to characterize m-CPPD crystals and compare them wit h t-CPPD crystals reliable methods of synthesis must be developed. Severa l methods for the synthesis of these crystals have been reported . Cheng and Pritzker (1981 ) crystallized various forms of CPPD an d calcium pyrophosphate tetrahydrate (CPPT ) crystals from solution s of sodium pyrophosphate an d calcium chloride by varying pH, temperature , concentrations of pyrophosphate (PPi ) and Ca2+, an d Mg2+ io n concentrations. I t was found tha t Mg 2+ ions influenced whic h form of CPPD crystallized. A t low Mg2+ concentration s (<0.4 mM) the triclinic form wa s favored bu t a t higher concentrations (0.5 - 0.9 mM) the monodinic form wa s favored. In vitro  CPP D growth studies have been conducted in an attemp t t o mimic physiological growth conditions using collagen gel matrices unde r simulated physiological conditions of pH and ionic strength (Mande l et at. 1990; Mandel an d Mandel 1984) . I n both these cases the author s found tha t an amorphous calciu m phosphate formed initially . Thi s was followed b y conversion to the monodinic form (abou t 6  weeks later) then to the triclini c form (1 3 weeks). Hear n and Russell (1980) studied CPPD formation fro m Ca2+ an d PPi solutions of pH 7.4 in multisolute buffer. The y also found tha t neither the monodinic nor triclinic forms of CPPD formed initiall y but rather , were preceded by the formation o f either amorphou s calciu m phosphates o r orthorhombic CPPT. Th e observation that the metastable monoclinic-CPP D 5 forms firs t wa s supported by Derfus et  al. (1992) who looked at a  minera l forming portion of porcine hyaline articula r cartilage in 1  mM ATP (source of PPi) and observed the formation o f m-CPPD. Thes e crystals were identifie d using FTTR spectroscopy from a  KBr pellet and compensated polarized ligh t microscopy. Methods for the synthesi s of monoclinic and triclinic CPPD were reported by Mandel et al. (1988) . Whil e attempting to determine the X-ray diffraction patter n o f a sample of orthorhombic calcium pyrophosphat e tetrahydrate (o-CPPT ) under vacuum, the authors reported the formation of an hexagonal form of CPPD following the loss of two moles of water from th e tetrahydrate. Thi s means that CPP D can exist in three differen t polymorphi c forms althoug h the hexagonal form ha s never been isolated from huma n tissue. Monoclini c and triclinic CPPD, and o-CPPT ,  from whic h h-CPPD is produced, ar e al l synthesized via the acid intermediate calciu m dihydroge n pyrophosphate (CDPP) . 3. Characterization o f crystal s 3.1. X-ray powder diffraction (XRPD ) X-ray powder diffraction pattern s giv e information abou t the distanc e between paralle l planes of atoms within the crystal lattice. Th e distance s between planes are called d-spacings an d ar e measured in angstrom unit s (A). Th e d-spacings fo r each crystal ar e determined by how the atom s ar e arranged within the crysta l an d the siz e of the atom s themselves. A s a resul t the XRPD pattern produced fo r each crystal type reveals a  se t of d-spacings that i s unique for that crysta l type . Th e XRPD pattern then can be used a s a "fingerprint" t o identify th e crysta l in question . 6 The basic method involves reflecting monochromatic parallel X-ray beams off a powdered sampl e a t various incident angle s of 8. Thes e inciden t rays can reflect of f planes near the surface o f the crystal or deeper down. These rays wil l therefore trave l different pat h lengths within the crystal . Th e result is that the reflected ray s will be out of phase a t most angles of 0 and therefore wil l undergo destructive interference an d will not be detected. A t some angles of 0 the extra path length that a  ray must travel when it reflect s off a plane below another reflecting plane is exactly equal to the wavelengt h of the incident X-rays, or a multiple thereof. I n such cases the reflected ray s will not undergo destructive interference an d can be detected. Th e relationship between the incident angl e 0 and the d-spacings of the crystal i s given by the Bragg equation : nX = 2d sin0 where X is the wavelength of the incident X-ray beam, d is the distanc e between paralle l planes, 0 is the angle between the incident beam an d th e crystal plane, and n is the order of the diffraction o r multiples of the wavelength distanc e that i s added to the path length of a reflecting bea m (Willard et al. 1981) . At least twenty-five calciu m pyrophosphates including the monoclini c and triclinic forms o f CPPD have been identified (Brow n et al. 1963) . Th e powder X-ray diffraction pattern s were reported for these compounds and th e space group, lattice constant s an d density of the t-CPPD crystals were also given (Brown et al. 1963) . The X-ray diffraction pattern s fo r al l three polymorphic forms o f CPPD have been reported by Mandel et al. (1988 ) including their lattice constants . The lattice constants for t-CPPD were reported by Mandel in a  previous paper (1975) using a  single crystal X-ray diffractometer. Th e lattic e 7 constants for monoclinic CPPD were determined by using a mathematica l iteration t o compare the diffraction pattern s of t-CPPD and m-CPPD and tes t different possibl e solutions until a  set of lattice constants coul d be found fo r monoclinic crystals that would fit its pattern (Mande l et al. 1988) . Th e constants for hexagonal CPPD were determined using a  single crysta l diffractometer b y the method of least squares fit with 9 centered reflection s (Mandel et al. 1988) . Th e lattice constants reported for the t-CPPD agre e well with those reported by Brown et al. (1963) . Mandel et al. (1988 ) reported the density to be 2.56 gem"3 for t-CPP D and 2.66 gem-3 for m-CPPD. Value s of 2.46 genr3 for t-CPPD an d 2.47 gem"3 for m-CPPD were reported by Brown et  al. (1963). Al l these density values were calculated . 3.2. Thermal analysi s Thermal analysi s gives information abou t the thermal event s occurring in a sample of a solid. Differentia l scannin g calorimetry (DSC) uses two pans, one containing sample and one empty (reference). Th e two pans ar e kept a t the sam e temperature a s they are being heated an d the difference i n energy required to keep the sample pan a t the same temperature a s the reference i s measured a s heat flow to the sample (Willard et al. 1981) . I n thi s way endothennic an d exothermic reactions can be monitored a t temperature s within the heating range. Endothermi c reactions include events such a s melting or dehydration of the sample , while events that give off heat, suc h as recrystallization, woul d be recorded a s an exothermic reaction . Although wate r makes up 12.42 % of the CPPD weight stoichiometrically, Brow n reported the triclinic form to contain 12.83 % wate r and the monoclinic form to contain 14.69 % water. Bur t an d Jackson (1987 ) 8 characterized syntheti c triclinic CPPD crystals. Th e crystals were heated i n a differentia l scannin g calorimeter (DSC) from 150-350°C . Th e first mole of water of hydration was lost between 19 0 and 275°C and the second mole was lost between 285 and 350°C. 3.3. Zeta potential measurement s The surface charg e of the crystals is believed to play a major rol e in binding interactions of the crysta l with membranes (Bur t et al. 1986 ) and proteins (Kozin and McCarty 1976b) . Whe n suspended in a  polar liquid suc h as water, a  charged crystal will attract ions of opposite charge to its surface . As a result, the effective charg e of the crystal is reduced by these counterions . This is the principle in the measurement of zeta potential (Q. According to the DLVO theory (Derjaguin, Landau , Verwey and Overbeek) the electric double layer surrounding a  charged particle is composed of a tightly bound layer of counterions and a  diffuse laye r of counterions an d co-ions. I f the crysta l moves through the solution , possibly in response to an electric field, the diffuse laye r is sheared off but the tightl y bound layer is not. I n other words, the crystal will move through the liquid with the tightly bound layer intact. Th e zeta potential then, is the potentia l between the edge of the tightly bound layer (shear plane) and the bulk solution (Martin 1993) . Experimentally the zeta potential is measured by determining th e velocity of the crysta l in the suspending medium under a n applied voltage. This is referred t o as the electrophoretic mobility of the particle and can be converted to the zeta potential by the Helmholtz-Smoluchowski formula : 4JIT|V 9 where y\ is the viscosity of the liquid, v is the electrophoretic mobility , and D is the dielectric constant of the liquid (Ducheyne et al. 1992 ; Rao 1972). Burt an d Jackson (1987 ) determined th e zeta potential of t-CPPD i n varying concentrations of phosphate buffer fro m 0  mM (distilled water) to 10 mM at 37°C. Th e zeta potential was found to be dependent on buffe r concentration. I n distilled water the zeta potential was -33.6 ±2.1 mV, reached a  maximum of-43.7 ± 3.0 mV as the buffer concentratio n wa s increased, an d then decreased between phosphate concentrations o f 3.0 an d 10.0 mM. 3.4. Solubilit y The solubility of triclinic CPPD in buffers o f varying pH at 37°C was determined by measuring Ca2+ level s using atomic absorption spectrometr y (Burt an d Jackson 1987) . Maxima l solubility was obtained after 1 6 hours of tumbling. Th e solubility was shown to increase a s the pH was lowered fro m 8.2 to 2.3 and als o increased slightl y from p H 8.2 to 10.3. Other groups (Bennet t et  al. 1975; Brown and Gregory 1976; Xu et al. 1991a; Xu et al. 1991b ) have determined th e solubility of triclinic an d monoclinic CPPD. CPP D samples were incubated in solvent for up to 3 weeks before being measured. Th e amount of CPPD in the supernatan t wa s determined by either measuring the dissolved Ca 2+ ion s using atomi c absorption techniques o r through the use of radioactive 45 Ca. Alternatively , the pyrophosphate concentratio n was sometimes measured using a pyrophosphate (PPi ) assay kit (P-7275, Sigma Chemicals) or by hydrolyzing the PPi to phosphate an d using a colorimetric method (Chen et al. 1956) . Since CPPD produces Ca 2+ an d PPi ions in solution it is not surprising tha t the presence of either of these ions in the solution decreased the equilibriu m solubility of CPPD (Bennet t et  al. 1975) . Th e presence of Mg2+ ion s which bind to PPi, or albumin which binds to Ca2+, increased CPPD solubility since they reduce levels of free Ca 2+ o r PPi in the solven t (Bennett et  al. 1975; Brown and Gregory 1976) . A n increase in ionic strength of the solven t increased CPP D solubility (Bennett et  al. 1975) . Yeas t pyrophosphatase, a n enzyme that hydrolyzes pyrophosphate ions also enhanced CPPD equilibriu m solubility which may be important in terms of removal of crystals from th e cartilage of affected patients . Th e enzyme's activity was stimulated by Mg2+ and suppressed by Ca2+ ions . Maxima l solubility effects du e to the enzym e occurred a t a  pH value of 7.4 (Xu et al. 1991a; Xu et al. 1991b) . 3.5. Fourier Transform infrared spectroscop y (FTIR ) Infrared spectroscop y can give information abou t the types of bonds present in a molecule. I n FTIR, a Fourier Transform interferomete r produce s an infrared absorptio n spectru m (Willar d et al. 1981) . Th e reported infrare d scans of t-CPPD crystal s (McCarty et al. 1966 ; Okazaki et al. 1976 ) found i n human tissue match t-CPPD standards with strong bands a t position s indicating the presence of water and P-0 bonds . 4. Protein adsorptio n t o crystal s Absolom an d Neumann (1988) studied the adsorptio n of the protein s fibrinogen, IgG , human albumin , an d bovine albumin to solid polymer substrates an d showed that protein adsorption increased with increasin g hydrophobicity of the proteins. The y suggested that once a particle is coated with protein its properties become governed by the protein coating and not by the particle itself . The binding of proteins to crystals is a very important facto r i n studying the interactions of crystals with cells since these bound protein s have been shown to have a modulating effect o n the inflammatory respons e produced by the crystals (Hasselbacher 1982) . Th e study of plasma protei n binding to crystals is a good model for protein binding from synovia l flui d since synovial fluid i s an ultrafiltrate o f plasma (Markowit z 1983 ; Revell 1982). I n inflammatory state s the protein concentration in synovial flui d becomes elevated with increased levels of immunoglobulins, lipoproteins , fibrinogen, an d variable levels of complement components (Markowitz 1983; Revell 1982) . Th e following summarizes som e of the work that has been done to characterize the binding of proteins to CPPD, MSUM, and HA. 4.1. Protein adsorption to MSUM and CPP D In som e early studies of protein binding to MSUM crystals the protein s considered included Human Cohn Fraction II (CF-II) , a source of IgG, as well as egg-white lysozyme (LYS), beta-lactoglobulin (BLG), bovine seru m albumin (BSA), and ovalbumin (OVA) . Whe n comparing binding affinities i t was found tha t CF-I I (IgG) > LYS > BLG > BSA > OVA for MSUM crystals (Kozin and McCarty 1976b) . Kozin et al. (1979) found tha t IgG bound to MSUM decreased crysta l induced neutrophil cytolysi s but increased non-cytolytic lysozyme release by neutrophils, whereas the following proteins: IgM, human seru m albumi n (HSA), BLG, and BSA bound to MSUM but inhibited non-cytolytic lysozyme release. However , in these studie s the MSUM crystals were heated prio r to use in order to remove pyrogens. A s shown by Burt (1983), this treatmen t would have resulted in dehydration of the samples and significant change s in the surface an d bulk crysta l properties . 12 Terkeltaub et  al. (1983) studied the plasma protein binding to MSUM crystals. The y found tha t both anionic and cationic polypeptides bound suggesting that the binding mechanism ma y not be solely dependent on charge. Th e plasma proteins Clq, Clr , Cls , fibronectin, fibrinoge n an d kininogen al l bound to MSUM. However , albumi n did not bind in significan t amounts (Cheria n and Schumacher 1986 ; Terkeltaub et al. 1983) . The binding of high density lipoproteins (HDL) and low density lipoproteins (LDL) to t-CPPD or MSUM crystals has been found to decreas e the neutrophil cytolysi s that is caused by these crystals (Burt et al. 1989) . There i s some evidence to suggest that this inhibition is caused by the apolipoprotein B  component of LDL (Terkeltaub et  al. 1984) . Thi s study also found tha t apolipoprotei n A-I and apolipoprotei n E bound to the MSUM surface a s detected using 2-D gel electrophoresis and Western Blot . The binding of IgG to crystals has been characterized in severa l reports. Ig G has been shown to bind more extensively to MSUM than t o t-CPPD (Kozin and McCarty 1976b) . Th e binding appeared to be charge dependent with the relatively positive Fa b portion of the IgG binding to the negative crysta l surface leavin g the Fc portion exposed (Hasselbacher 1979 ; Kozin and McCarty 1976b ; Kozin and McCarty 1980) . I t is thought that th e exposed Fc portions of the bound IgG molecules enhance phagocytosis of the crystals by macrophages through interaction with Fc receptors on the membrane (Kozi n and McCarty 1976b) . Th e binding of IgG does not appea r to be affected b y other proteins a s pre- or co-incubation of crystals with othe r proteins produced little or no effect o n IgG binding (Kozin and McCart y 1980). Th e binding of IgG to MSUM or t-CPPD als o results in increased crystal-induced neutrophi l responses such a s superoxide anion production , 13 degranulation, etc . (Burt an d Jackson 1993 ; Nagase et al. 1989) . Ther e hav e been no studies o f the binding characteristics o f IgG to m-CPPD. 4.2. Protein adsorptio n t o hydroxy apat i te (HA ) The binding of proteins to hydroxyapatite crystal s has been studie d more extensively than protein binding to MSUM or CPPD. HA can be used as the stationary phase in HPLC columns to separat e proteins. Bindin g in this case occurs through a  nonspecific attractio n between positive charges from amin o groups on proteins and the negativ e surface charg e of HA. Th e carbonyl groups on amino acids can act in two ways. The y can be repelled by the negative charge of HA, or clusters of carbonyls, that ar e found on acidic proteins, can interact in a specific manne r with site s on HA (Gorbunoff an d Timasheff 1984 ; Raj et al. 1992) . Fujisawa an d Kuboki (1991) determined the binding of some bovine derived proteins : phosphophoryns, bone Gla protein (BGP), osteomycin, an d bone smal l proteoglycan II to HA crystals. Phosphophoryn , i n particular, i s a highly phosphorylated protei n of dentin which binds calcium ions. The y observed tha t al l four proteins had binding constants in the order of 10"7 M, but tha t acidi c proteins had a  lower binding capacity, presumably due to a repulsion of negative charge on the carbonyl rich protein. The y found evidence of competition fo r binding sites between BGP, phosphophoryns, an d PGII a s well as preferential protei n adsorption on the (100) face of the HA crystal where Ca 2+ ion s are exposed in a  rectangular arrangement . Fujisaw a and Kuboki (1991) suggested that there may be specific binding of proteins to HA. Nyberg(1990) studie d the abilit y of mineral dust s with adsorbe d proteins to enhance superoxide production by human monocyt e derive d macrophages. The y found IgA and IgG to be strong enhancers of the crysta l induced macrophage response . Titanium dioxid e (Ti02) is a material used to make denta l implant s and its biocompatibility with the bone with which it comes in contact can be influenced b y the binding of proteins onto the Ti02 surface . Ellingse n (1991 ) compared the binding of serum proteins to Ti02 with their binding to HA and found tha t both materials bound the same three proteins, namely albumin , prealbumin an d IgG. The studie s above used various methods for identifying proteins . Rocket gel electrophoresis (Hasselbache r 1979) , radiolabelling with 125 I (Kozin and McCarty 1980) , O-Farrel ge l electrophoresis (Terkeltau b et  al. 1988), use of the fluorescent probe fluorescein wit h the protein (Fujisawa an d Kuboki 1991) , SDS-PAGE an d Immunoelectrophoresis (Ellingse n 1991) , and 2-dimensional ge l electrophoresis using Western Blot (Terkeltaub et  al. 1984 ) are al l methods by which bound and unbound proteins can be identified. Th e gel electrophoresis methods allo w for separation an d identification o f different protein s in a  mixture whereas the use of fluorescent o r radiolabel s provide a rapid method for assessing the amount of a known protein that i s present in a  sample . In this work, fluorescent labelle d IgG was used to determine th e characteristics of binding of IgG to CPPD crystals. 5. Objective s Monoclinic CPPD crystals play an important role in the pathogenesi s of CPPD crystal deposition disease . Furthermore , proteins bound to th e surface o f inflammatory crystal s suc h as MSUM and t-CPPD have been shown to modulate crysta l induced neutrophil responses . Ther e is little or no 15 information i n the literature, however, on the physicochemical characteristic s of m-CPPD an d on the binding of proteins such as IgG to the crysta l surface . 5.1. Specific objective s Synthesis ofCPPD  crystals 1) Obtai n reliable methods for the synthesis of triclinic, monoclinic and hexagonal forms o f CPPD. Characterization of  CPPD crystals 1) Characteriz e the crystals using basic techniques including X-ray powder diffraction, DSC , FTIR, surface are a an d particle siz e measurements, zeta potential, scannin g electron microscopy (SEM) and optical microscopy. 2) Determin e the lattice constants for m-CPPD crystals . 3) Determin e the solubility of the m-CPPD crystals over a pH range between 2  and 10 . 4) Measur e the solubility of triclinic, monoclinic, and hexagona l CPPD at differen t temperature s t o determine which forms ar e stable at each temperature an d what , if any, ar e the transitio n temperatures between the forms . 5) Construc t a n hydration phase diagram o f the m-CPPD to determine which hydrate i s stable under differen t relativ e humidity environments . 16 Protein Binding  to  Crystals 1) Measur e the binding affinity o f IgG protein to t-CPPD an d m-CPPD crystal s EXPERIMENTAL 1. Materials Materials used in this thesis project were obtained from th e followin g suppliers: Acetone, reagent grade, Fisher Scientifi c Calcium acetate , Fisher Scientifi c Calcium carbonate , BDH Inc. Calcium chloride dihydrate, Fisher Scientifi c Calcium chloride monohydrate, Sigma Chemical s Calcium tetrahydrogen di-orthophosphate , BDH Chemicals Ltd., Poole England Concentrated hydrochlori c acid (35.4% w/w), BDH Inc. Ethylenediaminetetraacetic acid , Sigma Chemical Compan y Glacial acetic acid, BDH Inc. Glucose, BDH Chemicals Ltd., Poole England Glycine, Sigma Chemicals, St. Louis, MO. Human IgG - FITC conjugate, Sigm a Immuno Chemicals , St . Louis, MO. Hydroxy naphthol blue, Mallinckrodt Chemica l Works, St. Louis Indium standards , packed into sealed aluminum DSC pans in October, 199 0 Lithium chlorid e monohydrate, BDH Inc. Magnesium chloride , BDH Inc. Magnesium sulfat e heptahydrate , Mallinckrodt Chemica l Works, St. Louis, MO. Minusil powder, supplie d by Zeta Meter Inc. Nitrogen N.F. Medical Gas, Linde, Union Carbid e Orthophosphoric aci d (85%), BDH Inc., and Fisher Scientifi c Potassium bromide , Ventron, Alfa Inorganics, Beverly MA. Potassium carbonate , BDH Inc. Potassium chloride , BDH Chemicals Inc. Potassium dihydroge n phosphate, Fisher Scientific, Fairlawn , NJ . Silica gel, Davison Chemical , Baltimore, MD. Sodium bicarbonate , The Nichols Chemical Company Ltd., Montreal, Canad a Sodium chloride , BDH Chemicals Inc. Sodium chromat e dihydrate (Na2Cr203«2H20), Sigma Chemical s Sodium dodecy l sulfate (SDS) , Sigma Chemical s Sodium hydroxide, Fisher Scientifi c Sodium nitrite , Allied chemicals, Morristown, NJ . Sodium phosphat e dibasic (Na2HP04), Fisher Scientifi c Strong ammonium hydroxid e solution , Mallinckrodt, Paris , Kentuck y Trizma Base, (Tris(hydroxymethyl)aminoethane), Sigm a Chemicals , St. Louis, MO. Water, distille d and deionized via a  Milli-RO Water System, Millipore Corp. Zinc sulfate heptahydrate , BDH Inc. 2. Equipmen t Balances, Mettler Corporation models AJ100, AE163, and PJ30 0 Pipettes, variable volume Pipetman from Gilso n Co. Hot plate/stirrer, Cornin g model PC-351 Vacuum oven, National Appliance Compan y Air pump, model 0211-V45N-G8CX Gast ai r pump (Benton Harbor , Michigan) with an Emerson motor model SA55JXGTD-414 4 (St. Louis, MO.) Filters, Pyrex medium grade ground glass vacuum filters,  15 0 mL an d 3000 mL capacity, ASTM # 10-15 Micro-centrifuge tubes , 1. 5 mL capacity tubes with caps, Elkay Products, Shrewsbury, MA. Furnace, Temco electric furnace, mode l RCE Water circulators, Haake Dl wit h Haake EK12 cold finge r Scanning Electron Microscope, Hitachi S-570 Vacutainer tubes , Becton and Dickinson silicone coated Vacutainer tubes , number 643 3 pH Meter, Fishe r Accumet pH Meter Model 610A Water baths, Aquatherm G-8 6 water bath, New Brunswick Scientifi c Company, New Brunswick, NJ., and Shaking Waterbath-25, Precision Scientific Company , Chicago , IL. 3. List o f Buffer s 20 Hanks balanced salt s solution (HBSS) Salt NaCl KC1 Na 2 HP0 4 KH2PO4 MgS04«7H20 mM 137 5.4 0.33 0.44 0.41 Salt CaCl2 Glucose MgCl2 NaHC03 mM 1.3 5.6 0.5 4.2 Glycine-HCl-NaOH buffers (Creed y 1977) • 0. 1 M Glycine - 7.507 g glycine, 5.844 g NaCl to 1  litre with distille d water • 0. 1 M HCl - diluted from 35.4 % w/w concentrated solutio n with distilled wate r • 0. 1 M NaOH - 4.0 g NaOH dissolved and made up to 1  litre in distille d water Buffer preparatio n involved adding sufficient amount s of either NaOH or HCl solution to the glycine buffer t o achieve the desired pH. Tris(hydroxymethyl)aminomethane -  HCl buffer, p H 7.2 (Creedy 1977) • 0. 2 M Tris base solution - 24.228 g Tris base to 1  litre with distille d water • 0. 1 M HCl - diluted from 35.4 % w/w concentrated solutio n with distilled wate r To produce a  pH 7.2 solution roughly 45 mL of the HCl solution was mixed with 25 mL of the Tris base solution an d diluted to 100 mL with distilled wate r 21 4. Method s 4.1. Synthesis o f crystal s 4.1.1. Calcium dihydroge n pyrophosphate (CDPP ) The methods used to synthesize the calcium pyrophosphates studie d required the prior synthesis of the acid intermediate calciu m dihydroge n pyrophosphate (CDPP) . Th e chemical formula o f CDPP is CaH2P207 an d i t has a  molecular weight of 216.04 g/mol. Th e method for its synthesis is given below (Brown et  at. 1963 ; Mandel et  al. 1988) . In a  heavy-duty 600 mL pyrex glass beaker, 25 0 mL of 85% orthophosphoric aci d was boiled a t 210°C with vigorous stirring. Thi s proces s removes the water from the phosphoric acid a t 150° C and converts it to pyrophosphoric aci d a t 200°C. Whe n the bubbling was no longer vigorous and almos t completed, and the temperature wa s stable a t 210° C, two 40 g quantities of calcium tetrahydrogen di-orthophosphat e (C a di-ortho) were added to the hot acid one gram a t a  time, waiting for the powder to dissolve completely before the next addition . Th e temperature wa s maintained a t 210°C. Befor e th e second batch was added, some CDPP previousl y synthesized (~lg ) was mixed into the batch to act as seed crystals . Th e second batch of Ca di-ortho was added a t a  rate o f about 0.5 g every 30 seconds. A s more calcium tetrahydrogen di-orthophosphat e wa s added, th e CDPP began to crystallize out eventually forming a  milky white thic k suspension. This suspension was filtered ho t using a  medium grade glass groun d filter (AST M # 10-15) . Th e filter ha d been preheated by running about 50 mL of hot orthophosphoric aci d through prio r to the suspension an d usin g aluminum foi l wrapped aroun d the funnel t o retain the heat . Onc e the filte r had cooled , the crystal s were washed in acetone and refiltered. Thi s proces s was repeated unti l the adsorbed aci d was removed as evidenced by the eas e with which the crystals could be resuspended . 4.1.2. Triclinic calcium pyrophosphate dihvdrat e (t-CPPD ) Triclinic CPPD was prepared a s given below (Burt and Jackson 1987) . In a  300 mL beaker, 12 5 mL of distilled wate r was heated to 60°C with slow stirring. T o this was added 0.85 mL of concentrated hydrochloric acid and 0.385 mL of glacial acetic acid. I n a  second beaker 0.72 g of calcium acetate was dissolved in 24 mL of distilled water and heated to 60°C. Another 0.72 g of calcium acetate was added to the first  beaker , dissolve d by increasing the stirring rate an d 2.4 g of CDPP was rapidly added to the beaker. Whe n dissolution was complete the stirring rate was reduced fo r about five minutes. Th e contents of the second beaker were then adde d to the first ove r a period of about 30 seconds with rapid stirring . A  white gel forme d which occupied abou t half the volume of the liquid. Th e beaker was the n removed from hea t an d allowed to stand covered and undisturbed for severa l hours a t room temperature durin g which time the gel collapsed and t-CPP D crystallized. Th e crystals were filtered  wit h a  medium grade glass ground filter (AST M # 10-15 ) and washed three times with distilled water and thre e times with acetone. Thi s procedure produced about 1. 5 g of product. 4.1.3. Monoclinic calcium pyrophosphate dihvdrat e (m-CPPD ) The synthesi s of m-CPPD was based on the procedure reported by Mandel et at. (1988) . To 106 mL of distilled water a t 50° C was added 6.96 mL of a 0.337 M MgCl2 solution with moderate stirring . Th e temperature wa s increased to 80°C and 2.0g of CDPP crystals were added a t a  rate o f ~0.1g every 15 seconds (each addition was allowed to dissolve before the next one was added). Followin g this, 8.25 mL of a 2.33 M calcium acetat e slurry at 80°C was poured into the CDPP solution. T o ensure total transfer som e of the CDPP solution was used to rinse out the calcium acetat e beaker an d poured back into the CDPP solution. See d crystals of m-CPPD were then adde d if they were available . Stirrin g was maintained fo r one minute, the sti r ba r removed an d the flask covere d and allowed to stand a t room temperatur e until the gel , which had formed afte r th e addition of calcium acetate , collapsed an d m-CPPD crystallized. Th e crystals were then filtered usin g a medium grade glass ground filter (ASTM #10-15), washed 3  times wit h acetone and allowe d to air dry. Th e pH of the crystal solution afte r eac h ste p in the synthesi s procedure was noted, and the time needed for the gel to collapse in unseeded batches was recorded . The stability of a batch of m-CPPD crystals was monitored over severa l months by obtaining X-ray diffraction pattern s of m-CPPD samples a t intervals. 4.1.4. Orthorhombic calcium pyrophosphate tetrahydrate (o-CPPT ) The synthesi s of o-CPPT was based on the method of Mandel et al. (1988). Wit h moderate stirrin g 1.10 g  of CDPP was added to 200 mL of a 0.05 M solution of ammonium hydroxide . Stirrin g was slowed for 90 minutes. A gel formed which was left undisturbed fo r 3 hours following which th e solution was stirred vigorously for 2 minutes. Th e mix was then covere d an d left t o stand a t room temperature unti l the crystals formed. Th e crystal s were filtered usin g a medium grade glass ground filter (AST M # 10-15), washed three times with distilled water and acetone and allowe d to air dry . 4.1.4.1. Generation  of  first seed  crystal batches The method o f Mandel et al. (1988 ) was modified in attempts t o produce crystals which could be used as seed crystals in future batches . On e modification wa s to acidify the solution with concentrated hydrochlori c acid 24 to pH 4 after th e addition of CDPP, while another was to refrigerate th e solution before the gel collapsed an d recrystallization occurred . I n one case the crystal s were filtered an d the filtrate wa s stored in the refrigerator . Crystallization occurred in this filtrate solutio n and these crystal s wer e filtered an d used as seed material because X-ray analysis indicated th e presence of o-CPPT. On e batch that had been acidified faile d t o show crystallization of any kind when kept a t room temperature for 1  week. I t wa s refrigerated an d yielded a very small quantity of crystals which were insufficient fo r X-ray diffraction identification . Thes e crystals were also used as seed material in the following syntheses . 4.1.4.2. Generation  of  the second seed crystal batches Two sets of three batches were produced. On e set was acidified t o pH 4 following the addition of CDPP, while the second set was not. Withi n eac h set one batch was seeded with the crystals obtained from th e filtrate, on e was seeded with the unidentified crystals , and the other was unseeded. Al l batches were refrigerated . From these six batches the crystals from th e unseeded batch of the acidified se t were thought to be o-CPPT and were used as seed crystals fo r further o-CPP T synthesis . 4.1.4.3. Further  attempts  at  o-CPPT synthesis Five batches were set up a s follows: i) seede d (second seed crystal batches) and refrigerated ; ii) seede d (second seed crystal batches) and the gel left t o collapse a t room temperature; hi) unseede d an d refrigerated ; iv) unseede d an d the gel left t o collapse a t room temperature ; v) unseeded , not acidified to pH 4, and the gel left t o collapse at room temperature. Further attempt s to synthesize o-CPPT were conducted by acidifyin g the CDPP solution to pH 4, seeding with o-CPPT, an d allowing the gel to collapse under refrigeration . The effect o n hydration level of storing o-CPPT for 3 days in a  silica gel desiccator or at 50°C for either 30 minutes or 2 hours was studied using thermal analysis . 4.1.5. Hexagonal calcium pyrophosphate dihydrate (h-CPPD ) The hexagonal form o f CPPD (h-CPPD) was produced when o-CPP T was put under vacuum (Mandel et al. 1988). Unde r these conditions th e tetrahydrate los t two moles of water of hydration an d the lattice rearrange d into the h-CPPD form . In an attempt to produce h-CPPD, about 1  gram of o-CPPT was put in a plastic weigh boat and placed in a vacuum oven without heat for 18 hours. The crystals were then placed in a silica gel vacuum desiccato r for 5 days. Any change in the identity of the crystals was monitored using X-ray powder diffraction an d differential scannin g calorimetry after th e 1 8 hours in the vacuum oven , and afte r th e fourth an d fifth da y in the desiccator . 4.2. Character izat ion o f crystal s 4.2.1. X-rav powder diffraction (XRPD ) X-ray powder diffraction (XRPD ) scans were obtained using a Rigaku Geigerflex X-Ra y Diffractometer Syste m (Rigaku Corporation, Tokyo, Japan) with a  biplanar goniometer , interfaced wit h a  Dexton PCII 286 computer by a D/max-B controller . Th e X-rays used were Ka radiation from a  copper targe t with a  nickel filter. Crysta l samples were packed flat int o an aluminu m sample holder . I n cases when limited sample was available a  thin film of crystals was sprinkled onto double sided Scotch tape attached to the holder . Samples were scanned from 5  to 40 degrees 26 at a scan speed of 3° per minute an d a  step size of 0.05 degrees 28. Th e X-ray tube was operated a t a potential o f 40 kV and a current of 20 mA. XRPD scans were obtained for samples of calcium tetrahydrogen di -orthophosphate (use d to synthesize CDPP), CDPP, m-CPPD, t-CPPD, o-CPPT, an d vacuum oven-treated o-CPPT . 4.2.1.1. Determination  of  unit cell  parameters for  m-CPPD 4.2.1.1.1. Growth  of  m-CPPD crystals The preferred metho d of determining unit cell parameters for m-CPPD would be to grow a  crystal large enough to do single crystal X-ray diffractio n analysis. Attempt s were therefore mad e to grow suitable crystals . Th e supernatant o f the suspension from which m-CPPD crystals had been grown was pipetted into two glass Pyrex test tubes. Th e tubes were then seede d with m-CPPD an d allowed to stand undistured a t room temperature wit h slow evaporation of the supernatant . Other attempt s involved the use of m-CPPD seed crystals in the synthetic procedure until a  crystal of large enough size could be recovered . 4.2.1.1.2. Calculation  methods X-ray powder diffraction dat a fo r m-CPPD was compared with both single crysta l and powder diffraction X-ra y data for monoclinic magnesium pryrophosphate dihydrat e (Oka and Kawahara 1982 ) to see if the two crysta l systems might be isostructural . To confirm previousl y published uni t cel l parameters fo r m-CPP D (Mandel et  al. 1988 ) which had not been determined through singl e crysta l X-ray diffraction work , a  computer program was written by Dr. J. Trotter , 27 U.B.C. Department o f Chemistry. I t was used to calculate the X-ray powder pattern tha t would result from th e unit cell dimensions reported. Thi s calculated patter n was compared with the observed XRPD pattern of m-CPPD. 4.2.2. Differential scannin g calorimetry (DSC) Differential scannin g calorimetry (DSC) scans were run with a  DuPon t series 99 thermal analyze r and 910 DSC. Th e samples were run under N2 gas at a  pressure of 20 p.s.i. Th e heating rate was always 10°C per minut e and the samples were weighed (between 2 and 8 mg) into crimped open aluminum pan s unless otherwise noted. Althoug h not possible with al l samples, some were held between 100 ° - 150°C for 30 minutes prior to thei r run in an attemp t t o remove surface adsorbe d water . DSC scans were obtained for calcium tetrahydrogen di-orthophosphat e (used to synthesize CDPP), CDPP, m-CPPD, t-CPPD, o-CPPT, an d vacuu m oven-treated o-CPPT . Sinc e it was found tha t relatively mild condition s (storage in a silica gel disiccator) could affect th e hydration level of o-CPPT a variety of storage conditions were investigated an d their effects monitore d using DSC. 4.2.2.1. Weight  loss  determinations In som e cases weight loss is reported a t certain temperatures o r a t th e completion of runs. I n these instances sample s had been held at the reporte d temperature fo r 1 0 minutes prior to being removed from th e DSC an d weighed. Thi s was to allow any vaporized water to escape the pan an d prevent an y possible condensation of the water vapor a s the pan cooled. I f the run had not been completed the pan was returned to the DSC, the sampl e was allowe d to re-equilibrate a t the appropriat e temperature, an d the ru n was then continued . 4.2.2.2. Calculation  of  heats of  dehydration The DSC thermograms could be integrated with an Apple II persona l computer to measure peak area. Indiu m standard s o f different mas s wer e used to calibrate the system. Indiu m has a  molecular weight of 114.8 g/g atom, a  melting point of 155°C, and its heat of fusion (AH* ) is 0.775 kcal/g atom (Weast 1967) . Thes e data were used to construct a standard curv e relating the enthalpy of melting (kcal) to the peak are a obtained. Weighed sample s of m-CPPD in open aluminum pan s were first held a t 130°C for 30 minutes to remove adsorbed water. Th e temperature wa s increased to 380° C and the pans reweighed. Th e final sample weight was divided by the molecular weight of anhydrous calcium pyrophosphat e (254.12 g/mol) to determine the number of moles of CPPD that had bee n present. Knowin g the peak areas and the number of moles present, th e molar heat of dehydration o f m-CPPD could be calculated . 4.2.3. Fourier transform infrare d scan s (FTIR) Samples of t-CPPD, m-CPPD, and o-CPPT were compressed to wafer s with KBr at a  pressure of 17 000 psi. Th e wafers wer e scanned using a Perkin Elmer 171 0 FTIR a t a  scan range of 400 to 4000 cm-1. 4.2.4. Scanning electron microscopy (SEM) Scanning electron microscopy (SEM) was performed on samples of t-CPPD, m-CPPD, and o-CPPT. Th e samples were attached to the surface of a metal stu b with graphite paint an d gold coated using a Hummer sputte r gold coater in an argon atmosphere . The y were then examined using a Hitachi S-57 scanning electron microscope. 4.2.5. Zeta potential determination s The zeta potentials of t-CPPD an d m-CPPD were determined usin g a zeta meter (ZM-80 , Zeta Meter Inc., New York) fitted wit h a  Unitron stereomicroscope. Th e zeta potential of Minusil powder was also determine d as a  reference standard . Measurements were conducted under a  field voltage of 200 V. Th e average time for selected particles of a similar size to travel acros s one ful l division (160 pxa) was determined by measuring the cumulative time for te n particles to travel this distance. Thi s was repeated 1 0 times so that a  total of 100 particles were timed. I n the case of m-CPPD it was not possible to measure in sets of 10 due to thermal overturn s o the measurements wer e done as 20 sets of 5 similarly sized particles each . 4.2.6. Particle siz e analysi s The particle size distributions of t-CPPD and m-CPPD wer e determined usin g a light microscope (Leitz Wetzlar, Germany) with magnification o f 130X, and a  glass-slide micrometer with etched division s down to 10 jim. Sample s of crystals were placed on a glass slide with th e micrometer visible . Th e number of crystals within given size ranges wer e counted. A  total of 542 crystals of t-CPPD and 548 crystals of m-CPPD wer e counted. 4.2.7. Solubility determination s The determination o f CPPD solubility was based on measuring calciu m ion concentrations in solution . 4.2.7.1. Measurment  of  calcium 4.2.7.1.1. Calcium  electrode An MI-600 calcium electrod e (Microelectrodes , Inc., Londonberry, N.H. ) connected to an Accumet pH meter 915 (Fisher Scientific) was used t o measure Ca 2+ io n concentration i n solution. Th e electrode gives readings i n mV proportional t o the log of Ca2+ io n concentration. Standard s of CaCl2 ,2H20 solution s in distilled water a t concentrations ranging from 10" 7 to 0.1 M Ca2+ wer e made up and read by the calcium electrode to assess th e range within which the electrode response was linear . The buffer chose n to conduct experiments using the electrode had to be at a  pH at which the CPPD solubility would be within its sensitivity range. The desired solubility of CPPD was one that produced calcium ion concentrations o f slightly greater than 1  x 10 _y M, where y is an integer. Th e reason for this is that the electrode readings ar e proportional to the log of calcium concentration . A s a result, the difference i n the readings that resul t for [Ca 2+] between 0.1 and 0.5 M, for example, is much greater than th e difference betwee n 0.5 and 1  M. Four glycine-HCl buffers wit h pH's between 1.2 7 an d 3.02 were made up and used to make standard curve s with Ca di-ortho as a Ca2+ io n source. The solubility of m-CPPD in each buffer wa s measured in triplicate using the initial electrode methods outlined above . Thi s data could then be used to select an appropriat e buffer . Monoclinic CPPD solutions of known calcium concentration wer e prepared in the following manner. A  sample from a  given batch of m-CPPD was dehydrated usin g DSC. Th e percentage o f water in the sample wa s calculated from weigh t loss during dehydration. Usin g this information i t was possible to calculate how much calcium a  m-CPPD sample of known weight from th e same batch should contain . Th e samples of known calcium concentration wer e then produced by dissolving known amounts of m-CPPD in buffer a t pH 2.70. Thre e standard curve s were produced: CaC^^H^Oi n distilled water ; CaCl 2«2H20 i n glycine-HCl buffer,pH 2.70 ; and Ca di-ortho in glycine-HCl buffer , p H 2.70. Th e spiked samples were read using the Ca 2+ electrode and the readings obtained were analyzed using the regressio n equations calculate d from th e three different standar d curves . Th e standar d curve which had a  regression equatio n which could most accurately measur e calcium concentration s from th e spiked CPPD solutions was chosen as the standard curv e system tha t would be used for the CPPD solubilit y experiments. 4.2.7.1.2. Atomic  absorption  spectrophotometry A Perkin-Elmer 305A Atomic Absorption Spectrophotometer wit h a n air-acetylene flame was used to measure Ca 2+ ion s in solution. Th e instrument give s absorbance readings directly proportional to calcium ion concentration. Si x buffer system s were tested: CaCl 2 " 2 0^ i n distilled water; CaCl 2«2H20 i n glycine-NaOH buffer, p H 9.0; CaCl2»2H20 i n glycine-HCl buffer, p H 2.70; Ca di-ortho in glycine-HCl buffer, p H 2.70; CaCl2»2H20 in Tris base-HCl buffer, p H 7.2; and Ca di-ortho in Tris base-HCl buffer , p H 7.2. Th e regression equations produced by the standar d curves from thes e buffer system s were analyzed to determine the sensitivit y of the atomic absorption methods to the buffer used . Solutions of known calcium concentration, a s prepared in Sectio n 4.2.7.1.1., were also employed to compare the two calcium salt s used to make standard curve s in terms of their accuracy in measuring calcium ion concentrations o f CPPD solutions . 4.2.7.1.3. Complexiometric  titration  withEDTA Ethylenediamine tetraaceti c aci d (EDTA), which complexes Ca2+ i n a 1:1 ratio, was titrated wit h calcium using hydroxy naphthol blue as a n indicator to determine Ca 2+ io n concentration . The EDTA solution was produced by dissolving about 0.73 g EDTA in distilled water . Th e solution was then made up to 500 mL with distille d water using 1  M NaOH to keep the pH above 12 . Standardizatio n wa s achieved by accurately weighing about 0.01 g of CaC03, adding 2 mL of 1 M HC1 and stirring until effervescence (releas e of CO2) stops. Distille d wate r was used to bring the volume up to about 15 0 mL and 100 mg of hydroxy naphthol blue was added. Th e pH was adjusted abov e 12 with NaOH. Th e solution was titrated to a blue endpoint with EDTA. The calcium from CPP D solutions was determined by diluting 5.00 mL of CPPD solutions (accurately measured using a glass pipette) to abou t 100 mL with distilled water. Abou t 10 0 mg of hydroxy naphthol blu e indicator was added an d the pH was adjusted abov e 12 with NaOH. Th e solution was then titrated with the EDTA to a blue endpoint . 4.2.7.2. Effect  of  vH To determine the appropriat e incubation or equilibrium tim e fo r solubility measurements, exces s m-CPPD crystals were added to tube s containing 4 glycine-HCl buffers rangin g in pH from 1.2 7 to 3.02 an d incubated a t 30° C for 1 8 and 24 hours. Calciu m concentrations wer e measured using the calcium electrode . The solubilities of m-CPPD crystals in glycine-HCl-NaOH buffers a t pH's of 2.7, 4.5, 6.0, 7.4 , and 9.5 and a temperature of 37°C were determine d using atomic absorption spectroscopy (AA) to measure calcium ions in solution. A n excess of crystals were placed in a  12 mL Kimax tube wit h PTFE lined caps and tumbled end over end in a 37°C oven using a Fishe r Dynamix tumbler a t 1 0 rpm. Th e tubes were done in triplicate and th e incubation time was three days . A t the completion of the 3  days the tube s were centrifuged i n a  model CL International Clinica l Centrifug e (International Equipmen t Co. , Needham Hts. , MA.) at 145 0 rpm fo r 5 minutes. Sample s of supernatant were diluted with distilled water an d the absorbance values determined usin g AA. 4.2.7.3. Effect  of  temperature Solubilities of t-CPPD and m-CPPD were determined in either pH 2.7 buffer o r distilled water and the calcium concentrations were determine d using the 3  methods described in section 4.2.7.1 . In the first serie s of experiments, the calcium electrode was used to measure calciu m ions in solution and the solubilitie s were determined in a glycine-HCl buffer a t pH 2.7. Exces s of either t-CPPD or m-CPPD crystal s were placed in 12 mL Kimax glass screw cap tubes with PTFE lined caps, along with 6 mL of buffer. Th e tubes were then secured into a  tube rack, lai d on their side, and submerged in an Aquatherm wate r bath with shaker (New Brunswick Scientific , New Brunswick, N . J.). Th e tubes were shaken a t 15 0 rpm a t constant temperatures o f between 30.7° and 61.1° C (± 0.2°C) for 1 8 hours. Th e tubes were centrifuged a t 1450 rpm an d the supernatan t analyzed using the calcium electrode. Experiment s were done in triplicat e with two control tubes containing buffer only . In a  modification o f the method of agitation of the tubes the Fishe r Dynamix tumbler was placed in an aquarium which had a  Haake Dl wate r circulator an d a  Haake EK 1 2 cold finger fo r temperatures below room temperature. Experiment s were done in glycine-HCl buffer, p H 2.70, with a n incubation time of 4 days. Th e experiment was done in triplicate an d calciu m ion concentrations wer e measured using the calcium electrode . The second series of experiments used AA to measure calcium ions in solution. Exces s CPPD in tubes containing glycine-HCl buffer a t pH 2.70 were incubated a t 4 different temperature s concurrently . Tube s were placed in the Aquatherm bat h a t 57.1 C , and in a  Shaking Waterbath-25 (Precisio n Scientific, Chicago , IL.) at 30.0°C. Tube s were tumbled end over end in a n isothermal oven at 37.3° C using a Fisher Dynamix tumbler a t about 1 0 rpm. The tubes a t 3.3°C were tumbled en d over end in a  refrigerator a t about 10 rpm. Th e sample preparation fo r this low temperature wer e done in a  5°C cold room in an attempt to avoid excessive amounts of CPPD going into solution initially. Incubatio n time was 3 days. Tube s were centrifuged a s described above , supernatants dilute d 400 fold with distilled water an d analyzed by atomic absorption spectroscopy . A similar solubility experiment was conducted using AA to tes t solubilities a t temperatures o f 31.5°, 36.5°, and 44.7° C using a Tris buffer , pH 7.2. Du e to the lower CPPD solubility at this pH compared to pH 2.70, the samples did not have to be diluted prior to measurement . To determine the appropriate incubation or equilibration time fo r solubility measurements i n distilled water, exces s t-CPPD or m-CPPD crystals were added to 30 mL distilled water in Erlenmeyer flasks,  stoppered , and placed in a  45° C Aquatherm wate r bath an d shaken a t 150 rpm. Samples were removed at 25, 48, 73, and 96 hours, centrifuged an d analyze d byAA. The effec t o f temperature o n CPPD solubility in distilled water wa s determined usin g AA to measure calcium ions in solution . Experiment s wer e done in triplicate using erlenmeyer flasks containin g an excess of crystals in 30 mL of distilled water .  Thes e flasks  wer e secured in the Aquatherm wate r bath a t temperatures o f 45° and 61° C and incubated for 4 days. A  5 mL aliquot of suspension was pipetted into Kimax tubes, centrifuged a t 2000 rpm in a Beckman GPR centrifuge fo r 5  minutes an d the supernatants analyze d by atomic absorption . 35 A third serie s o f experiments sough t to measure CPP D solubilit y usin g EDTA complexiometri c titratio n t o measure calciu m ion s in solution . Exces s m-CPPD o r trCPPD crystal s wer e place d i n Kimax tube s containin g 6  mL o f glycine-HCl buffe r a t a  pH o f 2.70. Th e tube s wer e place d i n a n Aquather m water ba t h a t constan t temperature s o f 33.2°, 49.1° , 57.5°, and 61.1 ° C  wher e they were shake n a t 15 0 rpm fo r 3  days. Th e tubes wer e centrifuge d a t 1450 rpm an d th e calciu m io n concentration i n the supernatan t wa s measured usin g EDTA complexiometric titration . 4.2.7.4. Determination  of  possible phase  changes  during  solubility studies XRPD scan s wer e performe d o n m-CPPD crystal s whic h had bee n shaking in a  buffe r o f pH 2.6 8 a t 30° C for 1 7 hours t o determine whethe r an y phase change s ha d occurred . XRP D analysi s wa s als o carried ou t o n m-CPP D crystals whic h ha d bee n tumbled i n a  pH 2.7 0 buffer fo r 4  days a t a temperature o f 60°C. Both m-CPP D an d t-CPP D crystal s wer e incubated i n distille d H2 O fo r 1 wee k a t a  temperature o f 45°C. Followin g this the y were analyze d b y DS C to asses s possibl e changes i n hydration leve l of the crystals . 4.2.8. Hydration phas e diagra m Accurately weighed sample s o f m-CPPD an d t-CPP D wer e place d i n previously weighed disposabl e Nunclo n plasti c petr i dishe s (InterMed ) an d stored a t 24° C in desiccator s o f constant relativ e humidity . Th e constan t relative humidit y environment s wer e achieve d by maintaining saturate d solutions o f different salt s i n th e differen t desiccators . Th e salt s an d th e relative humidit y values the y produce a t 24 ° C are give n below : LiCl»H20 (15% ) CaCl 2«H20 (32% ) K2CO3 (43% ) Na 2Cr203«2H20 (52% ) NaN0 2 (66% ) ZnS0 4«7H20 (90% ) In additio n to these a  silica gel disiccator was used to maintain a  3% relative humidity environment . The samples were allowed to equilibrate in their environments an d changes in hydration level were monitored by measuring weight loss and by using DSC. The experiment was repeated using m-CPPD and t-CPPD crystal s which had been dehydrated by being placed in a 400°C furnace fo r 3 0 minutes an d store d in a  vacuum desiccator . Prio r to being placed in thei r relative humidity chambers the crystals were characterized using XRPD an d DSC. 4.3. IgG binding to CPPD crystal s 4.3.1. Measurement of FITC-IgG Fluorescein isothiocyanate isomer conjugated t o IgG (FITC-IgG) was obtained a s a concentrated solution of 19 mg/mL from Sigm a Immunochemicals, St . Louis, MO. Dilution s were made with HBSS to produce FITC-IgG solutions of concentrations between 0.1-4000 jig/mL. Th e fluorescence o f these solutions were measured using a Shimadz u Spectrofluorophotometer RF-54 0 (Kyoto, Japan), and a  Shimadzu micro cell cuvette (204-27125) at a n excitation wavelength of 480 nm an d an emissio n wavelength o f 515 nm. Standar d curve s of fluorescence v s FITC-IgG concentration wer e constructed using FITC-IgG concentrations o f between 0. 1 and 10 0 jiglmL. Nonspecific bindin g of IgG to experimental equipmen t was a  potentia l source of error in these experiments . Thi s is especially true for very low IgG concentrations wher e binding to glassware could result in a  significant erro r in the actua l IgG concentration. T o determine the minimum Ig G concentration a t which this non-specific binding effect become s negligible , IgG solutions of concentrations 0.1 , 1 and 1 0 jig/mL wer e prepared. Eac h of these were spli t into two portions, one of which was used a s a  control while the other went through al l the pipetting and incubation steps that would be used in the actua l binding experiment. A t the completion of the procedur e the fluorescence remainin g in the IgG solutions that had been through al l the experimental procedures was compared to the fluorescence remainin g in th e control tubes . 4.3.2. Incubation o f crystals with FITC-Ig G Fresh batches of t-CPPD and m-CPPD were synthesized an d characterized usin g XRPD and DSC before being used for binding experiments. Fo r each IgG concentration less than 100 0 jig/mL, 1  mL of IgG solution was added to 100 ±0.1 mg of either m-CPPD or t-CPPD crystals i n polypropylene micro centriguge tubes (Elkay Products, Shrewsbury, MA.) . Experiments using FITC-IgG concentrations greater than 100 0 jiglmh wer e done using 0.5 mL of FITC-IgG solution and 50 ± 0.1 mg of crystals. These tubes were capped and shaken to suspend the crystal s and the n tumbled en d over end in a  37°C oven at abou t 8 rpm for 1  hour. Th e experiment wa s done in triplicate with 2 control tubes containing FITC-Ig G solution only . The amount of IgG bound to crystals was analyzed using two methods. In the first "indirect " method, the supernatant s wer e analyzed for FITC-Ig G by measuring fluorescence usin g an absorbanc e wavelength of 480 nm an d a n emmission wavelength of 515 nm. Th e concentration of unbound IgG ([ub-IgG]) was obtained from th e standard curv e produced from th e contro l tubes an d the bound concentration ([b-IgG D was given by: [b-IgG] = [IgG] - [ub-IgG] For the indirect method of measurement t o be used effectively th e crystals in each tube had to bind enough IgG to cause a  measureable dro p in the supernatant IgG concentration. T o determine the amount of t-CPPD crystals needed, tubes containing either 50 or 100 mg of t-CPPD were incubated with IgG solution at concentrations of 10, 20, or 40 jwg/mL and th e concentration of bound IgG was determined using the indirect method. A similar method was used to determine the amount of m-CPPD needed pe r tube. I n this case 25, 50, or 100 mg of m-CPPD crystals were incubated wit h IgG concentrations o f 20, 40, 60 or 100 jig/mL. To determine the effec t o f incubation time on the amount of IgG that would bind to crystals, 10 0 mg of t-CPPD crystals were suspended in 1  mL of a 10 0 jig/mL solutio n of FITC-IgG and tumbled end over end in a 37° C oven. The unbound FITC-IgG concentration was determined by measuring th e fluorescence remainin g in the supernatan t afte r 0 , 1, 2, 3, and 20 hours incubation time . In the secon d "direct" method the bound FITC-IgG was measured. Th e CPPD crystals were incubated a s above but the supematants were discarded . The CPPD crystals were then washed twice with HBSS to remove any FITC-IgG not tightly bound to the crystals . Eac h wash step involved vortexing the crystals in 0.5 mL HBSS, centrifuging, an d removing the supernatant . Th e bound IgG was then measured by eluting the protein from th e crysta l usin g 0.5 mL of a 2% sodium dodecy l sulfate (SDS ) solution and vortexing to suspend the crystals. Th e tubes were then boiled for 1  minute an d centrifuged. Th e amount of FITC-IgG which had been bound to the crystal s was determined by analyzing the supernatan t fo r FITC-IgG using th e spectrofluorophotometer. Th e standard curve from the control tubes wa s used to determine the bound IgG concentration while the unboun d concentration was given by: [ub-IgG] = [IgG] - [b-IgG] The effect o f the number of HBSS washing steps on the amount of FITC-IgG bound to t-CPPD crystal s was determined by washing 10 0 mg t-CPPD crystal pellets, which had been incubated in 1  mL of a 100 /zg/mL FITC-IgG solution , eithe r 1 , 2, or 3 times with HBSS before the SDS treatment an d measurement o f [b-IgG] by the direct method was carried out . To confirm tha t the SDS procedure removed al l the bound protein , crystal pellets of t-CPPD which had been previously used for direct method analysis were treated with SDS and boiled again. Th e SDS supernatant wa s analyzed to determine i f any residual IgG was present . Micro-centrifuge tube s containing 10 0 mg t-CPPD and 1  mL of a 100 jig/mL Ig G solution were incubated a s above. Th e tube and content s were treated a s in the experiments outlined abov e but al l supernatants an d washes were saved . An y IgG that remained in the tubes following the HBS S washing was removed using the techniques describe d for the direct method . For each tube al l of the saved samples were analyzed for FITC-IgG conten t and these amount s were summed an d compared to the amoun t of FITC-IgG originally in each tube (100 /zg). 4.4. Stat ist ical analysi s Best line curve fits wer e calculated using the linear least-square s regression metho d with the computer programs CA-Cricke t Graph III for th e Macintosh and/o r Microsoft Excel . Th e regression dat a calculated wit h 40 Microsoft Exce l was presented i n the form o f an ANOVA table which include d upper and lower slope limits (p<0.05) which were used to generate th e confidence interval s show n in Figure 21. 41 RESULTS 1. Synthesis o f crystal s Figures 2  to 5 show the XRPD scans of Ca di-ortho, CDPP, t-CPPD, an d m-CPPD respectively. Table s 1  and 2 give the peak positions in degrees 28, the d-spacings (A) , and the relative peak intensities for the XRPD patterns of t-CPPD an d m-CPPD respectively . The m-CPPD crystals were synthesized in an acidic environment. Th e pH was lowest after CDP P addition (pH 2.2) and the gel from whic h m-CPPD recrystallized had a  pH of 4.4. Store d dry the crystals retained the XRPD pattern typical of m-CPPD (Figure 5) for a t least 1 8 months. One batch of o-CPPT was successfully synthesize d but the syntheti c method proved unreliable. Th e generation o f the second set of seed crystal s produced the following results , as determined by XRPD. Th e seeds recovered from th e filtrate produce d a  mixture of o-CPPT and m-CPPD, while the seed s from th e refrigerated sampl e produced t-CPPD crystals . Th e batch whic h was not acidified an d unseeded produced a  mixture of crystals, but th e acidified an d unseeded batch produced o-CPPT crystals which were long an d needleshaped wit h a  dendritic appearance . Thes e were the crystal s used a s seed crystals in the experimen t whose results ar e given in Table 3. Samples 1  and 3 were o-CPPT, but al l other samples were eithe r mixtures of o-CPPT and t-CPPD or t-CPPD only. Th e XRPD scan of o-CPPT is shown in Figure 6 and the peak data i s given in Table 4. The attemp t to synthesize h-CPPD involved placing 1 g of o-CPPT in a vacuum oven for 1 8 hours, followed by a silica gel vacuum desiccato r for 5 42 more days . Th e XRPD scans that resulted afte r thes e treatments ar e show n in Figure 7 . Storag e in the vacuum oven resulted in a peak a t 8.3° 26 on the XRPD scan that would correspond to a d-spacing reported fo r h-CPP D (Mandel et al. 1988 ) but conversion was stil l incomplete a s evidenced by the peak a t 7.8° 26 typical for o-CPPT. Th e silica gel vacuum desiccato r however , resulted i n an XRPD scan that did not correspond to any published pattern s for the calcium pyrophosphates . Furthe r attempt s to synthesize h-CPPD by dehydration of o-CPPT were not possible due to the limited suppl y of o-CPPT. 1.1. Determination o f unit cell parameters of m-CPP D 1.1.1. Growth of single crysta l All attempts to grow a crystal of m-CPPD large enough to be used fo r single crystal X-ray analysis were unsuccessful . 1.1.2. Calculation method s Figure 8 shows the location of XRPD peaks calculated for monoclini c magnesium pyrophosphat e dihydrat e based on single crystal data (Oka an d Kawahara 1982 ) as well as the calculated positions of the peaks of m-CPPD assuming the uni t cel l dimensions reported by Mandel et al (1988) are correct . The figure als o shows the observed powder pattern peaks of m-CPPD. I n th e case of the magnesium pyrophosphat e a  match would occur if each peak in the pattern corresponde d to a peak in the observed m-CPPD pattern althoug h the d-spacings would not be identical since Ca2+ an d Mg2+ ions are not the same size . Fo r the calculated pattern of m-CPPD a match to the experimental pattern would occur if every peak found on the powder patter n matched exactl y with a  peak from th e calculated pattern . Th e reverse woul d not be necessary since a calculated peak may translate to a powder peak of very small intensity and therefore woul d not be observed in the powde r pattern. Neithe r pattern matched conclusively . 44 Intensity (cps) 4000 -n 3000-2000-1000-0 _JL 5 1 0 15 20 J_ 25 30 i 35 40 -4 45 50 degrees 2  8 Figure 2 : X-ra y powder diffraction patter n of calcium tetrahydroge n di-orthophosphate 45 Intensity (cps) 400-, 300-200-100-Q lf ftfy'K#»»,4»H>' gw j ,wi 10 15 I T — r 20 2 5 vCiwiJLt^^JW 30 35 4 0 degrees 2  0 Figure 3 : X-ra y powder diffraction patter n of calcium dihydroge n pyrophosphate. 46 Intensity go o n (cps) 600-400-200-0 > w * l i w y y i N < > u •>..- .A J ^ - ^ - - . J , .^ X-JiJvUv„JlJu J^..^«.j%.J 10 1 5 2 0 2 5 3 0 3 5 40 degrees 2  8 Figure 4: X-ra y powder diffraction patter n of t-CPPD Q Pn Pi o 1 a T3 M CD & o •43 fe •i-H <D T3 £ O & >> <S u 1 X! T H CD r—4 *0 £ CD o c CD i-r .CD «*H <D PH 1 b£ .3 o p. • id >> - u "co CD 1—1 . 'aS tf •< he d 03 CO TJ CD CN CO CD CD Q 03 CD P^  o o T—1 o 00 i—1 T—1 00 o Oi d r H r H O 00 to q to o o d o T—I lO o t > iO i q CO* I—1 CN l O 00 Tt; d lO t > d oo l O d o o d I—1 oo 00 t > co t > 00 d o to d T—1 "^ LO o CN d O i O i -*" to CM d lO 00 d i—i to to tr-^ rj5 lO t > t ^ i H LO ^* LO CO 05 1—1 to 00 oo ->* oo I—1 ^* o LO i-I CM t ~ O I—1 00 05 CO OS d T H Tj< o -*' o © CN CN 00 O i LO co CN LO CO o 00 d CN O i o Tj< CN ^ CO CN I—1 t ~ ^ od LO CO d <N o I—1 t ~ " * oo i—i ^ CO o I—1 d <N I—1 i-l o CO 1—1 CN 00 o 00 CO LO CNJ co o -tf t > CN <N I—1 o CO o 1—1 CO oo © co I—1 <N I—1 oo LO LO 00 <N CO i—l o Tj< ^ q CO* CT5 i—l 00 q co lO O l 00 CN i—l o CNJ lO q CN Tt< co 00 03 q cxi o q oi CN i H o CNJ CD t > CM* ^ T—\ d i—i 05 t > (N lO q <N* CO CO i H "tf ^ CN CN t > (N lO 00 CM* co i H o • ^ iO CO CM* 00 l O 1—1 00 CO <N o "*. co co 00 i H oo t > oo CM q <M* to i H -*" co i—l o i H o i q CN o 05 d CO i q CN o ^ d co o i H -tf -*" o "*. CN to "*. t > co i H * S CN C O 00 C N CN CM * CO C N iq t > co d lO l > oo o q oq oq ' o to (N t > od o i co c o <M C O <M ( N O O 1—1 X! 44 03 ft <*-! o 03 ft , v CD CM o lO LO CN i—l 03 <u ft CD CO 03 r O o CO ft a o cfa T5 CU + J 03 3 o r -H OS o >^  -p w S CD el • * - t CD > •43 03 i — i r V rf ^-v CO CO OJ i H + i w a t o SH PQ OS 48 Intensity 25 0 (cps) 200 150-100-50 -0 •MH^WI WJVUMMJYUHJ^ IM\KJUL 10 1 5 2 0 2 5 3 0 3 5 4 0 degrees 2  0 Figure 5: X-ray powder diffraction sca n of m-CPPD 49 Table 2: Peak 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 X-ray powder diffraction pe a Degrees 20 12.00 14.55 19.20 19.55 23.60 23.90 24.15 25.85 27.40 29.00 29.30 29.90 31.95 33.75 35.00 37.05 39.00 d-spacing (A) 7.37 6.08 4.62 4.54 3.77 3.72 3.68 3.44 3.25 3.08 3.05 2.99 2.80 2.65 2.56 2.42 2.31 k data fo r m-CPP E Rel. Intensity a 100.00 13.45 46.38 7.47 8.20 10.44 5.25 8.08 33.12 4.28 81.43 6.80 7.57 30.84 8.18 9.55 16.06 Reference0 d-spacing 7.37 6.08 4.62 3.74 3.43 3.22 3.03 2.86 2.77 2.63 2.56 2.41 2.30 Rel. Int. a 100 5 80 10 5 80 90 5 10 60 40 10 40 a Relativ e intensity calculated from cps of peak cps of base peak (12.55° 20) X100 b (Brown et  al. 1963 ) Table 3: Productio n of o-CPPT crystals Sample 1 2 3 4 5 Seeded Yes Yes No No No Storage Fridge Rm temp Fridge Rm temp Rm temp Acidified Yes Yes Yes No Yes Yield (g) 0.5601 0.4956 0.5027 0.6469 0.4039 Crystal phase s produced o-CPPT o-CPPT, t-CPP D o-CPPT o-CPPT, t-CPP D t-CPPD 51 Intensity 3 0 oo (cps) 2500 2000-1500-Intensity (cps) a Figure 6 : X-ra y powder diffraction patter n of a: o-CPPT from 5 ° to 40° 20 and b: from 10 ° to 40° 20 using an expanded scal e 52 Table 4 Peak 1 2 3 4 5 6 7 8 9 : X-ra y powc Degrees 20 7.80 15.60 23.45 23.90 26.30 27.95 30.35 31.45 39.60 er diffraction pea k data for o-CPPT d-spacing (A) 11.33 5.68 3.79 3.72 3.39 3.19 2.94 2.84 2.27 Rel. Intensity 3 100.00 1.65 4.87 1.12 3.14 0.86 1.20 4.36 1.68 Reference" d-spacing 11.4 5.95 3.75 3.37 3.18 2.94 2.83 2.26 Rel. Inta 100 20 20 40 10 80 10 5 a Relativ e intensity calculated from cps of peak cps of base peak (12.55° 29) X100 b (Brown et al. 1963 ) 53 Intensity 1250 1 (cps) 1000-750-500-250-a 15 2 0 2 5 3 0 3 5 4 0 degrees 2 8 Intensity (cps) 400 300-200-100-degrees 2 0 Figure 7 : X-ra y powder diffraction scan s for a ) the crystal s thought t o be partially transformed t o h-CPPD and b) th e crystals whos e pattern could not be identified . 54 O O (D 0 0 O <x> • • ! n nn cd n • o o • • d-spacings (A ) • 8 • Calculated position s o f m-MPPD peak s Peak position s from  m-CPP D powder pattern s Calculated position s o f m-CPPD peak s base d on reporte d spacing s Figure 8 : Plo t showing the d-spacings observed from m-CPPD powde r patterns, those calculated for monoclinic magnesiu m pyrophosphate dihydrat e based on published singl e crystal dat a (Oka and Kawahara 1982) , and those calculated for m-CPP D based on reported unit cel l dimensions (Mandel et al. 1988) . 55 2. Character izat ion o f crystal s 2.1. Thermal Analysi s The DSC scans for t-CPPD an d m-CPPD are shown in Figure 9. The total weight loss from t-CPP D samples was 13 % and from m-CPPD it was 14.7% . Whe n the surface adsorbe d water was removed from m-CPPD the weight loss was 13.7% . Bot h t-CPPD and m-CPPD crystals gave DSC scans that showe d endothermic peaks each corresponding to the loss of one mole of water of hydration. The scan of Ca di-ortho is shown in Figure 10 . Th e total weight loss from C a di-ortho was found to be 11% . Stoichiometricall y the weight loss would have been expected to be 7.1% for the monohydrate. Figur e 11 a gives the DSC scan of CDPP which also showed weight loss on heating. Th e tota l weight loss was 10.7% , of which 3.5% occured before 300° C and th e remaining 7.2 % was lost between 300° and 450° C over the large endothermi c peak. Thi s 7.2% loss would correspond stoichiometricall y to the loss of 1 mole of water of hydration. The thermogram fo r o-CPPT is shown in Figure l i b . I n the case of o-CPPT the total weight loss was found to be between 22.0 and 22.8%. Eac h mole of water of hydration correspond s to 5.5% of the weight of the sample . Weight loss measurements indicate d that 1  mole of water was lost before 105 ° C. A n additional 2  moles of water were lost between 105 ° and 250° C, and a fourth mol e was lost between 250 ° and 450° C. The effects o f storage conditions on the hydration level of o-CPPT crystals ar e given in Figure 1 2 which shows a : a sample store d a t room temperature; b: sample store d in a  silica gel desiccator for 2  days; c: a sampl e stored a t 50°C for 30 minutes; and d: a sample stored a t 50°C for 2 hours. 56 During DSC scans the original sample had a total weight loss of 20.7%. Afte r storage in the desiccator the weight loss fell to 9.3%. Sample s which had no t been stored in the desiccator but were stored at 50° C for 2 hours prior to DSC analysis had a  total weigh t loss of 20.7%, of which 5.2% occured during th e isothermal hold . Thi s 5.2% weight loss is reflected in the absence of the entire first endothermi c peak on the DSC scan . The attempts to produce h-CPPD yielded crystals whose DSC scans ar e shown in Figure 1 3 (a-c). Figur e 13 a shows the DSC scan that resulted whe n o-CPPT crystals were stored in a  vacuum oven for 1 8 hours without heat o r drying agent. Th e sample showed a weight loss of 19.3% water. Figur e 13b shows the effect o n the DSC scan of further storag e of the crystals in a  silica gel desiccator under vacuum for 4 days. Th e total weight loss of these crystals was 9.3%. O f this 6% occured between 255° and 450° C which would represent 1  mole of water o f hydration. Furthe r desiccato r storage for 2 4 hours produced the sca n shown in Figure 13c . Thi s sample had a  total weigh t loss of 8.5%. 57 a ENDOi 150 200 250 30 0 Temp. °C 350 400 Figure 9: Differentia l scannin g calorimetry scans for a: t-CPPD an d b: m-CPPD from  150 ° to 380° C at a  heating rate o f 10° per minute. 58 ENDOi 50 100 150 Temp. °C 200 250 Figure 10 : Diflferentia l scannin g calorimetry scans for calciu m tetrahydrogen di-orthophosphat e from 50° to 250° C at a  heatin g rate of 10° per minute . 59 100 20 0 30 0 40 0 50 0 Temp. °C Figure 11 : Differentia l scannin g calorimetry scans of a: calcium dihydroge n pyrophosphate an d b: o-CPPT from 30° to 500° C at a heatin g rate o f 10° per minute . 100 20 0 30 0 40 0 50 0 Temp. °C Figure 12 : Differentia l scannin g calorimetry scans showing the effects o f storage on the hydration levels of o-CPPT crystals, a : an o-CPP T sample at room temperature; b: o-CPPT stored in a silica gel desiccator for 2  days; c: an o-CPPT sample stored a t 50° C for 30 minutes; and d: an o-CPPT sample stored a t 50°C for 2  hours. The heating rate was 10° per minute . ENDOi • i  '  |  '  r  |  '  i 1 | , •  i  ' , | ' ' •  '  1 100 20 0 30 0 40 0 50 0 Temp. °C Figure 13 : Differentia l scannin g calorimetry scans of crystals that resulte d during attempts to synthesize h-CPPD. a : o-CPPT crystals tha t were kept in a vacuum oven for 1 8 hours; b: crystals in "a" afte r further storag e in a silica gel desiccator under vacuum fo r 4 days; c: crystals in "b" after being kept in a desiccator fo r a n additional 24 hours. Th e scan rate was 10° per minute . 61 62 2.1.1. Heat of dehydration The standard curv e relating heat of reaction (melting , kcal) to peak area o f the melting endotherm fo r indium i s shown in Figure 14 . Th e heat of dehydration (AHdehydration ) was calculated for both of the endothermic peak s on the m-CPPD scans . Th e peak between 220° and 280°C was found t o have a AHdehydration of 3.55 ± 0.18 kcal/mole, while the value for the peak between 280° and 360°C was found to be 5.50 ± 0.24 kcal/mole . 2.2. Fourier transform infrared scan s (FTIR ) The FTIR scans of t-CPPD, m-CPPD, and o-CPPT are shown in Figure 15 . 2.3. Scanning electro n microscopy (SEM ) The SEM photographs of t-CPPD, m-CPPD, and o-CPPT are shown in Figures 16-1 8 respectively. 2.4. Zeta potential determination s The zeta potentials for Minusil powder, t-CPPD, and m-CPPD ar e given in Table 5. 2.5. Particle siz e analysi s The following crystal siz e distributions were obtained for t-CPPD an d m-CPPD crystals . I n the case of t-CPPD, 45% of the crystals were less tha n 10 pan., 45% were between 1 0 and 50 pan, and the remaining 10 % were greate r than 5 0 pan. Fo r m-CPPD, no crystals were observed greater than 20 pan, 80% were less than 1 0 pan, and 20% were between 1 0 and 20 pan. Heat o f melting (kca l xl05 ) 8-, y = 4.65 e-9x +  2.16 e-7 r 2 =  1.0 0 0 500 0 1000 0 1500 0 Peak are a o f melting endother m Figure 14 : Standar d curv e showing the relationship between the heat of melting of indium standard s (kcal), and the peak area of the melting endotherm produced on the differentia l scannin g calorimetry apparatus used to calculate AHdehydration f° r m-CPPD crystals . i 1  i  r ~ 3200 240 0 200 0 160 0 1200 800 400 c m -1 Fourier transform infra re d scans for a: t-CPPD; b: m-CPPD; and c : o-CPPT crystals . Sample s were mixed with KBr an d compressed into disks at a  pressure of 17 000 psi prior to analysis. 65 Figure 16 : Scannin g electron microscope photograph of t-CPPD. 66 Figure 17 : Scannin g electron microscope photograph of m-CPPD. 67 Figure 18 : Scannin g electron microscope photograph of o-CPPT Table 5: Zet a Potential measurements fo r Minusil powder, t-CPPD, an d m-CPPD crystals in distilled water a t room temperature Sample Zet a Potential (± 1 standard deviation ) Minusil -29. 2 ± 3 mV a t-CPPD -35.3±4mV a m-CPPD -18. 8 ± 1.9 mV a Literatur e values are -29 ± 1 mV for Minusil an d -33.6 ± 2.1 mV for t-CPP D (Burt an d Jackson 1987 ) 2.6. Solubility determination s 2.6.1. Measurment o f calcium 2.6.1.1. Calcium  electrode A standard curve for CaC^H^O i n distilled water is shown in Figur e 19 and indicates linearity of electrode response over calcium ion concentrations of 10"5 - 10"1 M. Glycine-HCl buffers a t four differen t pH' s yielded the values given in Table 6  for m-CPPD solubilit y at 31°C. To assess the accuracy of the electrode system for measuring Ca 2+, solutions of known calcium concentratio n prepared from m-CPP D were required. T o accurately determine the weight fraction o f Ca2+ i n m-CPPD the water conten t of a m-CPPD sample was determined. Th e weight loss on heating was 14.9% and therefore 85.1 % of m-CPPD from thi s batch wa s assumed t o be anhydrous calciu m pyrophosphate. Th e amount of Ca2+ i n the m-CPPD sample s was then calculated . Three standard curve s were constructed an d best fit regressio n equations were calculated fo r each, and these were used to determine Ca 2+ concentrations in the solution s of known calcium concentration . Th e buffe r systems an d the regression equation s they generated ar e shown in Table 7. Four sample s of known calcium concentration were analyzed by the calcium electrod e and calcium concentrations were determined from eac h of the 3  different standar d curves . Th e results ar e shown in Table 8. The calcium concentration s determine d fro m standar d curv e 3 (CaCl2*H20 i n glycine-HCl buffer p H 2.70 ) gave values closest to the actua l calcium concentration s an d was used in subsequent calcium measurment s using the calcium electrode . Using glycine-HCl buffers o f pH between 1.2 7 an d 3.02 the solubility of m-CPPD was assesse d afte r 1 8 and 24 hours using a calcium electrode a t a temperature o f 31°C. Th e results ar e expressed a s mM Ca2+ io n concentration an d are given in Table 9  and show that afte r 1 8 hours th e calcium level s are no longer rising. Therefor e i t was determined that a n 1 8 hour incubation time was sufficient t o determine CPPD solubility in glycine-HCl buffers i n this pH range. 2.6.1.2. Atomic  absorption The following five buffer systems : C a di-ortho in pH 2.70 glycine-HCl buffer, CaCl 2«2H20 i n pH 2.70 glycine-HCl buffer, CaCl 2»2H20 i n p H 7. 2 Tris-HCl buffer, CaCl 2»2H20 in distilled water, an d CaCl2»2H20 i n pH 9. 0 glycine-NaOH buffer, wer e used to construct standard curves for calcium ions using AA. Thes e are shown in Figure 20 . Fo r each of these curves a  linea r regression equatio n was generated an d a  95% confidence interva l of each of the slop e values was calculated. Thes e confidence interval s ar e shown in Figure 21. Two m-CPPD samples of known calcium ion concentration in glycine-HCl buffer, p H 2.70, were analyzed by AA. Th e calcium concentrations wer e determined fro m tw o standard curve s which were prepared usin g CaCl2»2H20 (Curv e A) or Ca di-ortho (Curve B) in pH 2.70 glycine-HCl buffer. Th e results ar e given in Table 10. The effec t o f incubation time on CPPD solubility in distilled water a t 45° C was assesse d using AA and the results ar e shown in Table 11. Calcium electrod e reading s (mV ) 100-, 50 --50 y =  27.40LOGKx) +  107.2 0 r 2 =  0.99 1E-05 0.000 1 0.00 1 0.0 1 0.1 [Ca2+] (M) Figure 19 : Standar d curv e showing dependence of of calcium electrod e readings on calcium ion concentration using CaC^H^O i n distilled water a s a calcium standard . Table 6: CPP D solublity of m-CPPD in glycine-HCl buffers o f different p H measured using the calcium electrod e pH 1.27 1.97 2.62 3.02 Concentration of Ca2+ i n mM (± 1 S.D.) 310.1 (10) 62.6 (3) 23.6 (2) 8.9(1) 73 Table 7 : Regressio n equation s for standard curve data prepared usin g calcium chloride or calcium di-orthophosphate in water or pH 2.70 glycine-HCl buffer, measure d using the calcium electrode Curve 1 2 3 Nature of solution CaCl2 in distilled H 2 0 Ca di-ortho in pH 2.70 CaCl2 in pH 2.7 0 Best fit equation ([Ca2+] in M) mV = 35.854 log[Ca2+] + 66.070 r 2 =  0.998 mV = 21.011 log[Ca2+] + 31.129 r 2 =  0.995 mV = 33.486 log[Ca2+] + 55.043 r 2 =  0.994 Table 8: Compariso n o f calcium concentrations determined using differen t standard curves with the calcium electrod e Sample 1 2 3 4 mV reading -13,-14,-13 -12,-13, -13 -10, -11, -10 -8, -8,- 8 Calcium concentra l Actualb 10.0 8.6 10.7 11.7 Curve 1 6.1 (39) 6.4 (26) 7.4 (31) 8.6 (27) ion in M (% error)a Curve 2 7.7 (23) 8.2 (4) 10.6 (1) 13.7 (-17) Curve 3 8.5 (15) 8.9 (-3 ) 10.4 (3) 12.2 (-5) a Calculate d a s the % difference betwee n the measured an d the actua l calcium concentratio n in each solutio n " Calculate d from known weight of calcium from m-CPPD in solutio n Table 9: Effec t o f incubation time on m-CPPD solubility in differen t glycine-HCl buffers a s measured by a calcium electrode at 31° C. Buffer p H 1.27 1.97 2.62 3.02 Calcium concentration in mM (± 1 S.D.) 18 hr incubatio n 213.8 (8) 62.6 (3) 23.6 (2) 10.4 (1) 24 hr incubatio n 183.3 (13) 66.3 (3) 23.6 c 10.7 (1) c N o S.D. could be calculated because al l readings were the sam e Atomic absorptio n unit s 76 Ca di-ortho , pH 2.7 0 A CaC12, 2.7 0 o CaC12 in Tris , pH 7. 2 • CaC12 in d-H2 0 • CaC12 in Glycine-NaOH , p H 9  « [Ca2+] (M ) Figure 20 : Standar d curve s for atomi c absorption spectroscop y determinations o f calcium ion concentrations using five differen t Ca2+-buffer systems . Slope (x 10-5 ) 25-i 22.5-20 -17.5-1 5 -12.5 T -L i en ft .a CM r—I T 1 T 1 1 o CMa •-& .3 CM t—1 (d U 1 CM £ £M .3 CM t—I (d U 1 CM £ P. .a >-• o • A T J od 1 CM C-^ X CK .» £ .g CM t—1 od Ca2+ - buffer syste m Figure 21 : Th e 95% confidence interval s of the slopes (shown as range s which are hoped to contain the true slop e values) obtained fro m the standar d curves which were determined using atomic absorption spectroscopy and differen t Ca 2+-buffer systems . 78 Table 10 : Compariso n of calcium concentrations determined using standar d curve A (CaCl2«2H20 in pH 2.70 glycine-HCl buffer) an d standar d curve B (calcium tetrahydrogen di-orthophosphate i n pH 2.7 0 glycine-HCl buffer) usin g atomic absorption spectroscop y Sample 1 2 Calcium concentration in mM (% error)a Actual0 226.4 249.7 Curve A 245.4 (8.4) 287.2 (15) Curve B 222.7 (1.6) 260.6 (4.4) a Calculate d a s the % difference betwee n the measured an d the actua l calcium concentratio n in each solutio n 0 Calculate d from know n weight of calcium from m-CPPD in solutio n Table 11 : Effec t o f incubation time on CPPD solubility in distilled water a t 45° C as measured by atomic absorptio n Time (hours ) 25 48 73 96 Calcium concentration in jiM (± 1 S.D.) m-CPPD 120.0 (30) 130.9 (30) 141.7 (30) 163.1 (30) t-CPPD 7(1) 11.3 (0.4) 16.4 (2) 29.3 (4) 2.6.2. m-CPPD solubility dependence on pH The solubility of m-CPPD at different p H values is shown in Figure 22 along with previously published data for t-CPPD (Burt and Jackson 1987) . 2.6.3. CPPD solubility dependence on temperatur e The calcium electrode was used to measure Ca 2+ io n concentration a t various temperatures i n a glycine-HCl buffer, p H 2.70, and solubilitie s (measured a s Ca2+ io n concentration) as a function o f temperature fo r t-CPPD an d m-CPPD ar e shown in Figure 23. I n Figure 24 this solubilit y data is shown as a Van't Hoff plot with best line fits calculate d for eac h crystal type. I f these curve fits were extrapolated they would converge a t a point which corresponds to a temperature of 163°C. A second series of calcium electrode experiments were conducted wit h all sample tubes being tumbled in a water bath a t different temperature s an d the results ar e shown in Table 12. Atomic absorption experiments to measure CPPD solubility a t different temperature s yielde d the results in Tables 1 3 and 14 . CPPD solubility in distilled water was determined by measuring Ca 2+ concentration using AA and the results ar e given in Table 15. The method of EDTA complexiometric titrations to measure calciu m ion concentration yielde d the solubility results given in Table 16 from a glycine-HCl buffer p H 2.70 at various temperatures . 2.6.4. Determination o f possible phase changes during solubility studie s The XRPD scan for m-CPPD which had been incubated a t 30°C for 17 hours in a glycine-HCl buffer o f pH 2.68 is shown in Figure 25 . Th e XRPD patterns followin g a n incubation of four day s in pH 2.70 glycine-HCl buffe r showed no change for eithe r m-CPPD or t-CPPD (Figure 26 a and b). 81 Thermal analysi s of t-CPPD and m-CPPD after 1  week incubation i n distilled wate r a t 45°C showed that they did not change their degree of hydration durin g incubation. Th e t-CPPD sample was found t o be 13.0% water while the m-CPPD sample was 15.0 % water. Th e DSC scans for thes e samples ar e show n in Figures 27a and 27b respectively . 2.7. Hydration phase diagra m 2.7.1. Calcium pyrophosphate dihvdrat e The initial sample weights of the m-CPPD and t-CPPD sample s store d in different relativ e humidity environments ranged from 0.7364 8 g to 0.78198 g. B y five months of incubation a t 24° C the samples had apparentl y reached equilibrium , since the weight change of the sample s between tw o weighings from th e third month to the fifth mont h was less than 0.15%. Th e exception to this were the samples in the NaN02 chambe r (RH = 66%) which showed very large weight gains of up to 2% between 3- 5 months. Th e petr i dishes in this chamber however had turned yellow indicating that they had taken up som e component of NaN02 fro m thei r environmen t an d this, rathe r than water uptake by the sample , is probably what caused the excessiv e weight change . During the course of the five month incubations, none of the sample s changed thei r hydration levels. A s measured by thermal analysi s the t-CPP D samples were 14. 2 ± 0.2% water, while the m-CPPD samples were 15. 5 ± 0.4% water by weight . 2.7.2. Anhydrous calciu m pyrophosphat e Samples of t-CPPD and m-CPPD were dehydrated by heating a t 400° C for 30 minutes. Th e XRPD and DSC scans for the anhydrous product s use d are shown in Figures 2 8 and 29. Therma l analysi s showed these samples to be less than 1 % water. Initial sampl e weights for the anhydrous samples stored in differen t relative humidity environments ranged from 0.553 6 g to 0.9196 g. B y 5 months incubation the samples had apparentl y reached equilibrium sinc e successive weighings one month apart showed weight changes of less tha n 0.17% and in most cases less than 0.05%. Agai n the exception was the NaN0 2 chambe r (RH = 66%). There were no changes in hydration levels of the samples . Eve n at a n RH of 90% the maximum wate r uptake was less than 2.8 % for the anhydrou s sample which had been produced from m-CPPD and 1.1% for the anhydrou s sample produced from t-CPPD . Thi s range of weight gain is likely to reflec t only water adsorption . Th e DSC scans for the samples stored in the 90% RH chamber are shown in Figures 30 a and b. log ( [Ca2+ ] pM) 5-, 4 -2 -1-0 d> • i 4 D o o D D ° nr 6 8 i 10 pH i 12 • t-CPP D o m-CPP D Figure 22 : Effec t o f pH on the m-CPPD solubility given as log of the jiM Ca2+ concentration . Previousl y published data for t-CPP D crystals (Bur t an d Jackson 1987 ) are shown . Ca2+ concentratio n (M ) 0.015 -, 0.0125 -0.01-0.005 T u O-r ^ 0.0075 -P ^  ^ T § D 1 -»• D T Jo !  o ± • • • T J -O D 30 40 50 - r -60 Temp (°C) o • m-CPPD t-CPPD 70 Figure 23 : Solubilit y (given as Ca2+ io n concentration) of t-CPPD an d m-CPPD in glycine-HCl buffer a t a  pH of 2.70 as a function o f temperature, measured using the calcium electrode . Erro r bar s indicate ± 1 S.D. of triplicate samples . 85 log [Ca2+] (2.000) -(2.100) (2.200) -(2.300) - 0.638 r 2 =  0.67 1 y =  -586.833 x -  0.233 r 2 =  0.76 1 1 1  1  1 0.0029 0.003 0 0.003 1 0.003 2 0.003 3 O lo g m-CPP D • lo g t-CPP D 1/T Figure 24 : Van' t Hoff plot of the solubility data presented in Figure 23 with the linear regression curve fits fo r t-CPPD an d m-CPPD. Table 12 : CPP D solubility dependence on temperature i n pH 2.70 glycine-HCl buffer measure d using Ca2+ electrod e Temperature (°C) 15 29.7 45.2 59.7 Concentration of Ca2+ i n mM (± 1 S.D.)a t-CPPD 10.0 (0.5) 16.9 6.9 (0.8) 8.0 m-CPPD 18.3 19.6 (1.0) 11.2 9.5 (0.6) a Error s may not be reported either because only two tubes were read or al l readings were identica l 87 Table 13 : Temperatur e dependenc e of CPPD solubility in glycine-HCl buffe r at pH 2.70 with the calcium ion concentration measured by atomic absorption spectroscopy . Temperature (°C) 3.3 30.0 37.3 57.1 Concentration of Ca2+ i n mM (± 1 S.D.) t-CPPD 15.1 (0.8) 14.9(1.1) 13.5 (0.6) 14.3 (1.3) m-CPPD 21.5 (1.4) 18.7 (0.6) 18.6 (1.1) 15.7 (0.8) Table 14 : Temperatur e dependenc e of CPPD solubility in Tris-HCl buffer a t pH 7.20 and calcium ion concentration measured using atomi c absorption spectroscopy . Temperature (°C ) 31.5 36.5 44.7 Concentration of Ca2+ i n pM ( ± 1 S.D.) t-CPPD 40(5) 48(5) 64(9) m-CPPD 336(52) 206 (11) 350 (26) Table 15 : CPP D solubility in distilled water, calcium ion concentratio n measured by atomic absorption . Temperature (°C) 45 61 Concentration of Ca2+ injiM  ( ± 1 S.D.) t-CPPD 29.3 (4) 17.2 (3) m-CPPD 163.1 (30) 129.3 (26) 90 Table 16 : Temperatur e dependenc e of CPPD solubility in glycine-HCl buffe r at pH 2.70 and calcium ion concentration measured using EDTA complexiometric titration Temperature (°C ) 33.2 49.1 57.5 61.1 Concentration of Ca2+ i n mM (± 1 S.D.) t-CPPD 9.9 (0.1) 8.7 (0.2) 9.5 (0.2) 9.3 (0.2) m-CPPD 14.8 (0.2) 13.1 (0.2) 11.4a 11.8 (0.6) a Valu e is the averag e of two tubes only 91 Intensity 20 0 (cps) 150-100 50-0 hWrtU A U X J J V ^ 10 1 5 2 0 2 5 3 0 3 5 degrees 2 0 40 Figure 25: X-ra y powder diffraction sca n of m-CPPD incubated a t pH 2.68 for 1 7 hours a t 30°C and allowed to air dry overnight a t room temperature. 92 Intensity 25 0 -, (cps) 200-150-100-Intensity (cps) a 5 0 -0 *^%>WWa ^ LJILLI 10 1 5 2 0 2 5 3 0 3 5 4 0 degrees 2 0 500-, 400-300-200-100-0 ^**4*-vJ ALJMJ^J^J^ 10 1 5 2 0 2 5 3 0 3 5 4 0 degrees 2 0 Figure 26 : X-ra y powder diffraction scan s of a: m-CPPD an d b: t-CPPD crystals incubated for 4 days at 60°C in glycine-HCl buffer a t p H 2.70 buffer an d allowed to air dry at room temperature . 93 ENDOi 100 200 300 Temp. ° C 400 500 Figure 27: Differentia l scannin g calorimetry scans of a: t-CPPD and b: m-CPPD crystals which had been incubated in distilled water a t 45° C for 1  week, then ai r dried at room temperature. Th e heating rate was 10 ° per minute and ranged from  100 ° to 500° C. 94 Intensity (cps) degrees 2 0 Intensity (cps) degrees 2 0 Figure 28 : X-ra y powder diffraction scan s for anhydrous calciu m pyrophosphates produced by heating a : m-CPPD and b: trCPPD, in a 400° C furnace fo r 30 minutes. 95 a ENDO i b - i |  i  |  • • |  •  | — i J  >  |  «  |  •  [ — i |  «  | 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 Temp. °C Figure 29 : Differentia l scannin g calorimetry scans of a: m-CPPD an d b: t-CPPD crystals which had been dehydrated in a 400° C furnace fo r 30 minutes for the relative humidity experiment . Scans were conducted from 30° to 500° C at a  heating rate of 10° per minute. 96 f a END0 4 100 200 300 Temp. °C 400 500 Figure 30: Differentia l scannin g calorimetry scans of a: dehydrated t-CPP D and b: dehydrated m-CPPD afte r the y had been stored in a ZnSOWH20 desiccato r a t 24° C (90 % relative humidity) for 5 months. Th e samples were scanned from  130 ° to 500° C at a rate of 10° C per minute . 3. IgG binding to CPPD crystal s The standard curv e of emission fluorescence intensit y versus FITC-IgG concentration i s shown in Figure 31 and is described by the linear regressio n equation: y = 0.963 [FITC-IgG] -1.075 r 2 =  0.999 where [FITC-IgG] is in units of jig/mL. 3.1. Indirect method of measuremen t In the indirect method of measurment, the amoun t of free FITC-Ig G remaining in the supernatan t following incubation of crystals with FITC-IgG solutions was measured an d used to calculate the amount of FITC-IgG bound to the crystals . Th e incubation time for the IgG binding studies wa s optimized by measuring fluorescence intensitie s of supernatants followin g crystal-FITC-IgG incubation times between 1-2 0 hours . Fluorescenc e intensities did not significantly chang e afte r 1  hour and al l incubations were subsequently carried out for 1  hour. Solutions of FITC-IgG concentrations between 0.1 and 1 0 fig/mL wer e analyzed to determine how much IgG would be removed from th e supernatant durin g the course of a binding experiment due to adsorption to the surfaces o f tubes an d glassware. I t was found that a t FITC-Ig G concentrations of 10 fig/mL th e FITC-IgG that bound to the apparatus wa s insufficient t o significantly alte r the concentration of FITC-IgG remaining in solution. Durin g subsequent studie s 1 0 jig/mL wa s the minimu m concentration o f FITC-IgG used . Studies were carried out to determine the minimum weight of crystals per binding incubation experiment , suc h that differences i n the amount of 98 bound FITC-IgG could be detected using the indirect method of measurment . This weight was found to be 100 mg for both m-CPPD and t-CPPD. The binding data for IgG to both m-CPPD and t-CPPD is given in Table 16 . Th e data is shown as plots of amount of FITC-IgG bound per mg of crystal versus IgG concentration for m-CPPD crystals in Figure 32 and fo r t-CPPD crystal s in Figure 33. 3.2. Direct method of measuremen t In the direc t method, the amoun t of FITC-IgG bound to the crysta l pellet following incubation of crystals with FITC-IgG solutions was measured . The first par t of this experiment was to determine if the number of washing steps affected th e amount of IgG that was bound to the crystals before SD S was used. Fo r this experiment al l tubes contained 10 0 mg of t-CPPD suspended in 100 p,g/mL o f an FITC-IgG solution. Crysta l pellets were washed with HBSS either 1 , 2, or 3 times, the concentration of eluted protei n was measured, an d found no t to be affected b y the number o f washes. Th e second part of the experiment was to determine if a single SDS treatmen t was sufficien t t o elute al l of the bound FITC-IgG. Thi s was done by treating the crysta l pellet with SDS for a  second time and measuring any FITC-IgG subsequently released from the crystals . Th e amount of FITC-IgG remaining bound following the first SD S elution treatment was found t o be negligible. The binding data are given in Table 1 7 for IgG binding to both t-CPP D and m-CPPD crystals . The direc t binding curve for the m-CPPD data is given in Figure 3 4 while the curve for t-CPPD is in Figure 35. Th e data for both crystal type s were then plotted according to the Scatchard equatio n a s given below, [bound] [bound ] [S ] [unbound] K d K d where [unbound] is the concentration of FITC-IgG remaining unbound in the supernatant an d is subtracted from th e original FITC-IgG concentration t o give the crysta l bound protein ([bound]). Th e equilibrium dissociatio n constant is Kd and [S] is the total concentration of crystal protein interactio n sites (Hutchens an d Yip 1990) . Th e Scatchard plots for both m-CPPD an d t-CPPD are shown in Figures 36 and 37 respectively . The binding data were then fitted  t o the Langmuir an d Freundlic h equations. Figure s 38 and 39 show the fit of the m-CPPD binding data to the Langmuir an d Freundlich isotherms respectively, while Figures 40 and 41 show the fit of the t-CPPD binding data to the Langmuir an d Freundlic h isotherms respectively . The Langmuir equatio n takes the form : .. Ymb c y ~  1 + be where y is pig of protein bound per mg of crystal, c is the equilibriu m concentration o f FITC-IgG, ym i s the fig of protein needed to bind to 1  mg of crystal in order to produce a  monolayer, an d b is a constant relating th e relative rates of association and dissociation of protein to crystal (Marti n 1993). Th e Langmuir equatio n ca n be written in its linear form as : c 1  c y ~ bym +  y m The linear form of the Freundlich equation is: log y = log k + - lo g c 100 where y and c  are the same as above and k and n are constants tha t describ e the binding (Adamson 1990) . Th e calculated values for these two constant s with respect to IgG binding to t-CPPD an d m-CPPD are given in Table 18. In a  series of recovery experiments, the total amount of IgG accounted for in the supernatant , HBS S wash an d bound to crystals was determine d and compared to the initial amount of FITC-IgG added to the tubes. I n al l cases, the total IgG accounted for was >92% of the initial weight of IgG added to the tubes . Intensity 10 0 -i 75-50-25-y = 0.963x - 1.075 r 2 = 0.999 50 T -75 1 100 12 5 [FITC-IgG] ng/mL Figure 31 : Standar d curve for the fluorescence intensit y of [FITC-IgG] solutions in Hanks balanced salts solution a t 37° C. Table 16 : Indirec t binding data for FITC-IgG to CPPD crystals a t 37° C [FITC-IgG] (jig/mL) 10 20 30 40 50 60 80 100 150 200 Amount bound (fig/mg) 0.029 0.067 0.083 0.086 0.070 0.094 0.134 0.083 0.132 0.067 m-CPPD [unbound] (jig/mL) 7.1 13.3 21.7 31.4 43.0 50.6 66.6 91.7 136.78 193.3 errora (#g/mL) 0.1 0.3 0.7 1.5 1.1 2.2 1.0 3.5 2.0 5.5 Amount bound (#g/mg) 0.016 0.034 ~ 0.047 — 0.055 0.040 0.063 0.199 0.081 t-CPPD [unbound] (fig/mL) 8.4 16.6 ~ 35.3 — 54.5 76.0 93.7 130.1 191.9 errora (jig/mL) 0.6 0.1 ~ 0.1 — 0.9 1.0 0.4 10.5 2.4 a error s reflect ±1 standard deviation of reported unbound concentration s Amt FITC-Ig G bound /  mg m-CPPD (^g/mg ) u.z-0.15-0 .1 -0.05-0 -| o 1 \—o—\ • I o 1 T o < T o r [  1 > o 1 1 0 50 10 0 15 0 20 0 250 [FITC-IgG] (jig/mL) Figure 32 : Bindin g curve of FITC-IgG to m-CPPD crystals a t 37° C measured by the indirect method. Erro r bars represent ± 1 S.D. of triplicate samples . 104 Amt FITC-Ig G bound /  mg t-CPPD (jig/mg) 0.125 -, 0 .1 -0.075 -0.05-0.025 -0 T S n o 1  J 1 T n ± 0 50 10 0 a 150 200 250 [FITC-IgG] (jigfmL) Figure 33 : Bindin g curve of FITC-IgG to t-CPPD crystals a t 37° C measured by the indirect method. Erro r bars represent ±  1 S.D. of triplicate samples . Table 17 : Direc t binding data for FITC-IgG to CPPD crystals a t 37° C [FITC-IgG] (fig/mlj) 10 20 30 40 50 60 80 100 150 200 732 1000 1220 2000 4000 Amount bound (jfg/mg) 0.0052 0.0109 0.0142 0.0184 0.0133 0.0190 0.0203 0.0257 0.0328 0.0424 0.163 0.23 0.2014 0.41 0.718 m-CPPD [unbound] (#g/mL) 9.48 18.91 28.58 38.16 48.67 58.10 77.97 97.43 146.72 195.76 715.7 977 1199.86 1959 3928.2 errora (jig/mL) 0.05 0.13 0.10 0.08 0.10 0.23 0.25 0.09 0.41 0.21 2.9 2 0.96 1 10 Amount bound (#g/mg) 0.0056 0.0127 — 0.0168 — 0.0236 0.0210 0.0279 0.0372 0.0397 — 0.106 — 0.166 0.284 t-CPPD [unbound] (fig/mL) 9.44 18.73 « 38.32 « 57.64 77.90 97.21 146.28 196.03 — 989.4 — 1983.4 3971.6 errora (jig/mL) 0.03 0.08 ~ 0.09 ~ 0.18 0.15 0.19 b 0.33 — 1.0 --1 3 a Error s reflect ±1 standard deviation of reported unbound concentration s D Valu e is the averag e of two tubes only 106 Amt FITC-Ig G bound/m g m-CPP D (^g/mg ) 0.05 - . 1-, 0.75-0.5-0.25-0.04 -0.03 -0.02 -0.01 -T O 1 o o * 0-£— 0 - 1 1  1  1 — 50 10 0 15 0 20 0 T O J. 250 1000 2000 3000 4000 5000 [FITC-IgG] (jig/mL) Figure 34: Bindin g curve for FITC-IgG to m-CPPD crystal s a t 37° C as measured by the direct method. Inse t shows the binding curv e for FITC-IgG to m-CPPD a t FITC-IgG concentrations of 1 -200 jig/mL. Erro r bars indicate ± 1 S.D. of triplicate samples . 107 Amt FITC-Ig G boun d /  mg t-CPPD (fig/mg) 0.4 -. 0.3-0.2-0 . 1 -0 0.05 - • 0.04 -0.03 -0.02 -0.01 -0 • 0 • • • — i 1  1 1  1 0 5 0 10 0 15 0 20 0 25 0 • T • 1 1000 2000 3000 4000 5000 [FITC-IgG] (jig/mL) Figure 35: Bindin g curve for FITC-IgG to t-CPPD crystals a t 37° C as measured by the direc t method. Inse t shows the binding curve for FITC-IgG to t-CPPD at FITC-IgG concentrations of 1 -200 jig/mL. Erro r bars indicate ± 1 S.D. of triplicate samples . 108 [Bound FITC-IgG ] [Unbound FITC-IgG ] 0.07 -, ).05 -. V 0.06 0 0.04-0.03-0.02-0.01 T T -r 1 0 25 T -o-X 50 i 75 100 [Bound FITC-IgG ] (/zg/mL ) Figure 36: Scatchar d plot of the direct binding data of FITC-IgG to m-CPPD crystals a t 37° C. Erro r bars indicate ± 1 S.D. of triplicate samples . 109 [Bound FITC-IgG ] [Unbound FITC-IgG ] U.UB " 0.06-0.04-0.02-0 -T a ± T a ± - r a a a \n\ 1 f-QH 1 \—a-\ I 0 10 20 30 [Bound FITC-IgG ] (fig/mL) Figure 37: Scatchar d Plot of the direct binding data of FITC-IgG to t-CPPD crystals a t 37° C. Erro r bars indicate ± 1 S.D. of triplicate samples. 110 [FITC-IgG] (jtgfmL) y% FITC-Ig G bound per m g m-CPP D 7000-1 6000 5000-2000-1000 & r T o JL I T o 1 1000 I o 1 2000 3000 4000 5000 [FITC-IgG] (jtig/mL ) Figure 38: Fi t of Langmuir isotherm to direct binding data of FITC-IgG to m-CPPD crystals a t 37° C. Erro r bars represent ± 1 S.D. of triplicate samples . log (jig  FITC- IgG bound / mg m-GPPD) 0-, -0.5-- 1 -•1.5-- 2 --2.5 y =  0.808LOG(x) -  3.119 r 2 =  0.98 1 l I  l  I 10 10 0 100 0 1000 0 [FITC-IgG] Qig/mL) Figure 39: Fi t of Freundlich isotherm t o direct binding data of FITC-IgG to m-CPPD crystals a t 37° C. Erro r bars represent ±  1 S.D. of triplicate samples . [FITC-IgG] fcg/mL) jig FITC-Ig G bound pe r m g t-CPP D 20000 15000 10000 -5000-0 100 0 200 0 300 0 400 0 500 0 [FITC-IgG] (jig/mL ) Figure 40: Fi t of Langmuir isotherm to direct binding data of FITC-IgG to t-CPPD crystal s a t 37° C. Erro r bars represent ±  1 S.D. of triplicate samples . T • T D 1 113 log (fig PITC-IgG bound /  mg t-CPPD) 0 -0.5-- 1 -•1.5-- 2 --2.5 y =  0.604LOG{x) -  2.757 r 2 =  0.98 9 l 10 — I 1  1 100 100 0 1000 0 [FITC-IgG] (jig/mL) Figure 41 : Fi t of Freundlich isotherm t o direct binding data of FITC-IgG to t-CPPD crystals a t 37° C. Erro r bars represent ± 1 S.D. of triplicate samples . Table 18 : Bindin g constants for FITC-IgG binding to m-CPPD and t-CPP D at 37° C derived from Freundlich isother m Crystal m-CPPD t-CPPD k (jug bound/mg crystal ) 0.0008 0.0017 n 1.24 1.66 DISCUSSION 1. Synthesis of crystal s Reliable methods of synthesis of t-CPPD and m-CPPD were obtained . Figures 2  and 3 give the XRPD scans used to identify Ca di-ortho and CDPP, two compounds needed to synthesize both forms o f CPPD. Figur e 4 along with Table 1  give the XRPD scan and peak information fo r synthesize d t-CPPD while Figure 5  along with Table 2 give the X-ray data for synthesize d m-CPPD. Al l of the major peak s (relative intensity > 10) in the XRPD pattern obtained from synthesize d t-CPPD and m-CPPD crystals appear in the reference patterns for t-CPPD and m-CPPD respectively. Thi s identifie s the synthesized crystal s a s t-CPPD or m-CPPD and confirms tha t reliabl e methods have been obtained for their synthesis . A s reported by Brown et aZ.(1963), m-CPPD crystallized a t low pH. Althoug h reported to be th e metastable polymorphic form o f CPPD at room temperature (Brown et al. 1963), m-CPPD does not convert to t-CPPD when stored dry at room temperature fo r a t least 1 8 months indicative of a large activation energ y barrier to conversion of m-CPPD to t-CPPD. In al l procedures which ultimately yielded o-CPPT crystals th e amorphous gels had been acidified prior to crystallization. Refrigeratio n o f the gels also tended to favor o-CPPT crystallization as shown by the result s in Table 3. Th e original synthesis by Mandel et al. (1988) showe d tha t o-CPPT crystallized from a  gel at a  pH of 9.7. Brow n et al. (1963 ) showed that o-CPPT crystallized in a dilute solution of an electrolyte in the pH range of 3 to 5, and that even slight warming of the gel mixture tended to cause the 116 formation o f m-CPPD crystals. Base d on this, the modifications o f acidification an d refrigeration o f the gel were attempted . The XRPD scans are shown in Figure 6 (a and b) and were confirme d to be o-CPPT using their peak data (Table 4), which matched those for th e standard XRPD pattern of o-CPPT crystals (Brown et al. 1963 ) i n terms of peak position (d-spacings). Th e position of the peaks in the reference patter n are also shown in Table 4. Th e synthesized o-CPPT crystals had a  long, needleshaped crysta l habit and were not ground prior to XRPD analysis. I t is likely that this caused preferred orientatio n of the crystals during sampl e packing for XRPD analysis an d this is why the base peak is more than 2 0 times more intense than the next highest peak (Cullity 1978). The attempt to produce h-CPPD was conducted in two steps. Th e o-CPPT crystals were treated under relatively mild conditions at firs t (vacuum oven for 1 8 hours), then more intensive dehydrating condition s (silica gel vacuum desiccato r for 5 days). Figur e 7a shows that the mil d conditions were converting the o-CPPT to h-CPPD as the XRPD scan of the crystals shows the emergence of a peak at a  position, 8.3° 20, which corresponds to the reported base peak of h-CPPD (Mande l et al. 1988) . Th e smaller peak, a t 7.8° 20, was also seen in the o-CPPT pattern indicating tha t conversion was stil l incomplete. Unde r more severe dehydrating condition s however, the crystals lost more than two moles of water of hydration per mole of calcium pyrophosphate an d h-CPPD was not formed. Figur e 7b shows th e XRPD pattern o f the crystals that were formed in the silic a gel desiccator . This sample could not be identified. I n this pattern the base peak has moved position to a higher value of degrees 20 which corresponds to a smalle r d-spacing. Th e dehydration of the crystals appeared t o result i n rearrangement o f the lattice, reducing the distance between the lattice plane s 117 responsible fo r the base peak. Ther e is some evidence that these crystals ar e a monohydrate for m of calcium pyrophosphate (see section 2.1., Characterization of crystals, thermal analysis) . 1.1. Determination o f unit cell parameter s Without the availability of m-CPPD crystals large enough for singl e crystal analysis , only the calculation methods could potentially be used to determine the unit cel l constants for m-CPPD crystals. Thes e results ar e shown in Figure 8 and do not confirm m-CPPD unit cell dimensions. I f the m-CPPD crystal s an d m-MPPD crystals were isostructural thei r XRPD patterns would be similar but the m-CPPD pattern would be offset t o large r d-spacings sinc e the Ca 2+ io n is larger than the Mg2+ ion . Ther e is no match between the m-MPPD and m-CPPD patterns because a t the low d-spacin g range (4 • 5 A) there are a  relatively large number of m-MPPD peak s compared to m-CPPD peaks. Furthermor e the m-MPPD peak at a  d-spacin g of 6.5 A has no corresponding peak in the m-CPPD pattern . An alternative method was to calculate a  theoretical powder patter n for m-CPPD based on the unit cel l dimensions reported previously (Mandel et al. 1988) . Agai n no match was achieved because peaks a t d-spacings of 4.55, 4.6, and 5.3 A on the observed XRPD pattern of m-CPPD do not correspond to any peaks on the calculated pattern . Therefor e th e unit cel l dimensions of the m-CPPD crystals synthesize d coul d not be determined, no r could the previously reported dimensions be confirmed. I t is possible only to state tha t the synthesized crystal s have an XRPD pattern which matches the pattern of crystals reported to be monoclinic CPPD. 118 2. Characterization o f crystal s 2.1. Thermal analysi s Throughout the course of the thermal analysi s experiments i t was assumed tha t sampl e weight loss on heating was the result of water loss either from th e surface o f crystals or from within the lattice. On heating, the t-CPPD an d m-CPPD crystals both showed two endo thermic peaks corresponding to the loss of 1 mole of water of hydration each (Figure 9) . I n the case of t-CPPD the first mole of water was los t between 180 ° and 280° C and the second was lost between 280° and 350° C, while for m-CPPD the first  mol e of water was lost between 200° and 280° and the secon d between 280° and 350° C. Stoichiometrically the calcium pyrophosphate dihydrate s shoul d contain 12.4 % water. Triclini c CPPD crystals contained 13 % water an d m-CPPD crystal s containe d 14.7 % water. Followin g removal of surfac e adsorbed water (crystal s held isothermally between 100 ° and 150 ° C for 30 minutes) the weight loss for m-CPPD was 13.7%. I t is likely that th e additional 0. 6 - 1.3% water conten t in both crystals was water tightly bound to the crysta l surface , rathe r tha n water incorporated within the crysta l lattices. Figure 1 0 shows the DSC scan of Ca di-ortho. A  large endothermi c peak is observed between 140 ° and 200° C with a  total weight loss of 11% . The crystals seem to exist as a hydrate with about 1. 5 moles of water pe r mole of Ca di-ortho in the crysta l lattice . Figure 11 a and the correspondin g weight loss of 10.7% for CDP P indicate that CDPP is also hydrated with 1.5 moles of water per mole of CDPP. Figur e l i b show s the DSC scan for o-CPPT. Weigh t loss dat a 119 confirm th e crystal s are the tetrahydrate. Th e first endothermi c pea k corresponds to the loss of 1 mole of water, the second peak to an additiona l two moles of water, whil e the fourth mole of water is liberated over a large temperature rang e from 250 ° to 450° C. A t about 250° C there was a  very small exothermic peak which may indicate some recrystallization o f the sample just prio r to the loss of the last mole of water of hydration. A s shown in Figure l i b th e o-CPPT crystals begin to dehydrate a t lower temperature s (below 100 ° C) than eithe r m-CPPD or t-CPPD. Figur e 1 2 illustrates how different condition s can cause dehydration of o-CPPT crystals. Compariso n of DSC scans of o-CPPT crystals stored a t room temperature (Figur e 12a ) with DSC scans from o-CPP T crystals stored in a  silica gel desiccator for 2  days (Figure 12b ) reveals that the desiccated crystal s produced a  DSC scan in which the firs t 2  endothermic peaks are missing. A s noted above these two peaks corresponded to a total loss of 3 moles of water of hydration, meanin g that storag e in a  silica gel desiccator for 2 days converted the o-CPPT to a monohydrate. Storag e of o-CPPT crystals a t 50° for 30 minutes an d 2 hours caused some dehydration of the crystals. Afte r 3 0 minutes a t 50° the first endothermic peak was greatly reduced (Fig 12c) and by 2 hours it was completely eliminated (Fig 12d) . Sinc e the first peak corresponds to the loss of 1 mole of water of hydration, the sample that remained afte r storag e a t 50° C for 2  hours contain s three moles of water per mole of calcium pyrophosphate. Attempts to synthesize h-CPPT yielded crystals for which therma l analysis result s ar e given in Figure 13 . Storag e of o-CPPT crystals in a vacuum ove n for 1 8 hours caused some dehydration a s shown by the greatl y reduced first endothermi c peak in the DSC scan of these crystals (Fig 13a), as compared t o the DSC scan of o-CPPT (Fig lib), and by the fact that thes e 120 crystals were determined to contain only 19.3% water. I t is these crystal s which yielded the XRPD scan shown in Figure 7a . Ther e is some evidence that dehydration may have caused some transformation o f o-CPPT into h-CPPD sinc e the peak a t 8.3° 28 in Figure 7a corresponds to the pea k reported by Mandel et al. (1988 ) for h-CPPD. Sinc e the data appea r t o indicate that dehydratio n under vacuum preferentially cause s loss of the firs t endothermic peak corresponding to 1  mole of water (Figure 13a) it is interesting to speculate a s to whether the new hexagonal form describe d by Mandel et al. i s really a dihydrate or whether i t may be a trihydrate. Th e designation of the crystal s as a dihydrate by Mandel et al. was based on density measurements an d single crystal X-ray data. Under more rigorous dehydrating conditions in the silic a gel desiccator, the partially dehydrated crystal s were dehydrated furthe r a s evidenced by the loss of the first 2  endothermic peaks (Fig . 13 b,c). Sinc e the data indicates that the first two endothermic peaks correspond to a total of 3 moles of water of hydration it is likely that the remaining crystals ar e a monohydrate form o f calcium pyrophosphate . 2.2. Fourier transform infrare d scan s (FTIR ) As expected the FTIR scans of t-CPPD, m-CPPD, and o-CPPT are quit e similar (Fig 15 a-c), because the crystals al l have simila r bonds. Th e O- H stretch bands due to water can be seen a t 3200 cm"1, and bands due to P- 0 bonds can be seen aroun d 120 0 cm"1. Th e bands a t 160 0 cm-1 occur in th e unsaturated regio n where double bond frequencies commonl y fall (Willard et al. 1981)an d may be caused by P=0 bonds . Th e pyrophosphate molecul e is known to be quite variable in terms of bond lengths an d angles betwee n phosphorous an d oxygen atoms (Mande l 1975 ) and this may cause th e 121 variation in the spectr a of the three crysta l types in the region just below 1200 cm-1. 2.3. Scanning electro n microscop y From the SEM photographs o f t-CPPD, m-CPPD, and o-CPPT (Figs 16-18 respectively) an estimation of relative crystal siz e can be obtained. Th e t-CPPD crysta l shows a parallelogram shap e and has a  long diagonal of about 60 pan. Th e m-CPPD an d o-CPPT crystals are needleshaped wit h lengths of about 3 /mi for m-CPPD an d 1 0 -15 pan for the o-CPPT crystals . 2.4. Particle size analysi s Particle size analysis confirms th e relative cystal sizes of the m-CPP D and t-CPPD crystals . Th e t-CPPD crystal s are 5 to 10 times longer tha n m-CPPD an d a s a result would be expected to have a  much lower surface are a per unit weight than the m-CPPD crystals , given that the densities of the crystal types, 2.56 g cm-3 for t-CPPD an d 2.66 g cm-3 for m-CPPD, are simila r (Mandel et al. 1988) . 2.5. Zeta potential determination s The zeta potential determinations fo r t-CPPD an d m-CPPD were carried ou t in saturated solution s of the crystal s in deionized water. Sinc e the solubilitie s of m-CPPD an d t-CPPD in water ar e very low, the concentrations o f counterions wil l therefore als o be extremely low. Henc e th e zeta potentials for the crystal s suspended in water will closely reflect th e surface charg e of the crystals . Th e results of the zet a potentia l determinations fo r Minusil an d t-CPPD agre e well with values previousl y reported. Th e zeta potential of m-CPPD was measured to be -18.8 mV 122 compared to -35.3 mV for t-CPPD, and therefore m-CPP D crystals have a much lower surface negativ e charge than do t-CPPD crystals . Sinc e crysta l interactions wit h membranes an d proteins can be affected b y charge (Bur t and Jackson 1987 ; Kozin and McCarty 1976b) , this difference i n zet a potential may be reflected i n different bindin g properties that m-CPPD an d t-CPPD crystal s may have with membranes or proteins. 2.6. Solubility determination s 2.6.1. Comparison of calcium measuremen t method s Three methods of calcium measurement were employed in this work , each having their benefits an d limitations for use. Th e first metho d attempted wa s an ion specific electrode for calcium. Th e electrode measure s the electric potential (mV) across a liquid ion-exchange membrane responsiv e to calcium ions. Electrod e responses ar e directly proportional to the logarithm o f the calcium ion activity in the test solution . Figur e 1 9 shows that th e calcium electrod e used gave good linear correlation with the log of calcium ion concentration within a  range of 10"5 -10"1 M when standar d solutions of CaCl2»2H20 in distilled water were used. Th e major drawbac k of ion specific electrode s ar e the relatively large errors that ca n be involved, up to ± 4n% per mV uncertainty in measurement, wher e n is the charge of the ion. Thes e error s can be due to chemical interferences suc h as complexatio n or electrode interferences a s a  consequence of imperfect io n selectivity (Bauer et at. 1978) . A s a result measurements o f calcium in solution using the electrode showe d significant variation s dependin g on the buffer an d calciu m source used to make up the calcium solutions . Atomic absorption readings of calcium ions in solution als o were affected b y other ions in the solutio n medium. A s Figure 21 shows the thre e 123 standard curves , CaCl2*2H20 in glycine-HCl buffer (p H 2.70), Ca di-ortho in glycine-HCl buffer (p H 2.70), and CaCl2-2H20 i n Tris base buffer (p H 7.2) produced using AA could be described using regression equations that wer e not significantly different . Th e regression equations that describe d th e standard curve s CaCl2»2H20 i n glycine-NaOH buffer (p H 9.0) an d CaCl2#2H20 i n distilled water however , were not different fro m eac h other but had significantl y lower slopes than the first three curves . Difference s i n AA readings may be due to interference fro m othe r ions in the flame o r possibly the complexation of calcium ions by phosphates to form calciu m phosphates tha t do not absorb radiation a t the wavelength used in th e experiment (Bauer et al. 1978) . Al l calcium solution s had to be diluted to less than 0.00025 M in order to be within the range measureable by the AA instrument employe d for this experiment . A s a result saturated CPP D solutions in glycine-HCl buffer a t a  pH of 2.70 had to be diluted 400 fold before being analyzed by the AA instrument. Thi s added dilution step may have contributed to the variability of the results . Monoclinic CPPD solutions of known calcium concentration could be prepared in glycine-HCl buffer a t pH 2.70 because the m-CPPD solubility was high enough in this buffer t o make the process feasible. Tabl e 1 0 shows the results obtained when two such solutions were made up and measured usin g AA. Th e readings obtained were fitted t o the regression equations calculate d for two different standar d curves , CaCl2#2H20 in pH 2.70 glycine-HCl buffe r (curve A) and Ca di-ortho in pH 2.70 glycine-HCl buffer (curv e B). Thes e results showe d that using standard curv e B gave a more accurate estimatio n of actual calcium concentration in buffer an d therefore, sinc e the only difference betwee n the standard curves was the calcium sal t used, Ca di-ortho was used to make the standard curves for further A A experiments. 124 Table 1 1 shows that the CPPD samples incubated i n distilled water a t 45° C had not reached thei r equilibrium solubilit y after 4  days. I n the case of the m-CPPD the measured calcium concentrations ar e not significantl y different fro m eac h other, but nevertheless sho w a trend toward increasin g calcium concentratio n eve n after 4  days. Thi s is in contrast to the solubilit y data observed in pH 2.70 buffer wher e equilibrium was achieved by 18 hours (Table 9). Thi s was probably due to the higher solubility of CPPD in pH 2.70 buffer givin g rise to faster dissolutio n rates an d therefore equilibriu m ma y have been established more rapidly . The method of complexiometric titration with EDTA to measure calcium io n concentration relies on the use of a calcium specifi c indicator . The indicator chose n was hydroxy naphthol blue (USP 1990 ; Goettsch 1965) . When in solution at a  pH between 1 2 and 1 3 in the presence of calcium thi s indicator gives a  reddish pink colour to the solution . A t the end point of EDTA titration the indicator, strippe d of calcium which has bee n preferentially complexe d by EDTA, turns a  dark blue colour. Th e drawbac k to this method is that solution s for titration need to be at a  high pH (above 12) to ensure that the hydroxy naphthol blue and the EDTA are in thei r anionic form an d ar e able to complex calcium. Solution s of CPPD in pH 2.70 buffer, therefor e nee d to be alkalized to above pH 1 2 for the method to be used. Thi s change in pH may cause calcium ions in solution to precipitate ou t as calcium phosphates an d may result in reduced measurements o f calcium ion concentration . 2.6.2. m-CPPD solubilit y dependence on pH Figure 2 2 shows the effect o f pH on m-CPPD solubility . A s was th e case for t-CPPD, the solubilit y of m-CPPD was much greater a t low pH tha n 125 it was a t neutral pH. Th e solubility range was from abou t 15 mM Ca2+ a t low pH to about 1  mM Ca2+at neutra l pH. 2.6.3. m-CPPD solubilit y dependence on temperatur e The effect o f temperature on the solubility of both m-CPPD an d t-CPPD in a  glycine-HCl buffer a t pH 2.70 as measured by the calciu m electrode is shown on Figure 23 . Th e solubility of both forms increased wit h temperature by a total of about 0.0025 M Ca2+ betwee n 30° and 60° C. A t every temperature investigated the solubility of m-CPPD was found to be greater than the solubility of t-CPPD crystals , confirming that the m-CPP D crystals ar e the metastable form of CPPD in this temperature range. I f a temperature wer e found a t which both forms had equa l solubilities then thi s would be the transition temperature between the two forms (Haleblian an d McCrone 1969) . To determine the transition temperature th e solubility data wa s plotted accordin g to the Van't Hoff equation in Figure 24 with log calcium concentration plotted against reciprocal temperature in degrees Kelvin to give a  linear plo t for both crystal types (Martin 1993) . Regressio n equation s were calculated to describe best line fits through these points. I f extrapolate d these lines would converge a t a  1/T value corresponding to a transitio n temperature o f 163° C. Thi s means that a t atmospheric pressure t-CPPD was the stabl e polymorphic form of CPPD below 163° C but above it m-CPP D would be the stabl e form. Sinc e the calcium pyrophosphates ar e stil l crystalline a t over 1000 ° C (Haleblian an d McCrone 1969) the transitio n temperature between t-CPPD and m-CPPD is below the melting point of t-CPPD an d so the polymorphic system could be classed as enantiotropi c (Haleblian an d McCrone 1969) . Thi s classification i s somewhat hypothetical , 126 however, sinc e both forms o f CPPD begin to dehydrate near the transitio n temperature an d would be completely dehydrated a t the meltin g temperature. The values reported here for t-CPPD solubilit y are simila r to those reported in the literature . Bennet t et  al. (1975) reported that t-CPP D solubility at 37° C in a 0.1 M Tris-HCl buffer a t a  pH of 7.4 had a solubility of 70 fiM calcium . Bur t an d Jackson (1987) reported the solubilit y of t-CPPD a t 37° C to be 42 fiM in a 0.05 M glycine-HCl/NaOH a t pH 7.4, and 12.6 mM in a 0.05 M glycine-HCl/NaOH buffer a t pH 2.4. Other attempt s to measure CPPD solubility (Tables 1 2 to 16) using the different calciu m measurement methods and using different solvent s produced variable an d inconclusive results. The y are included here fo r completeness to document al l attempts to measure the effect o f temperature on the solubilit y of m-CPPD and t-CPPD crystals . 2.6.4. Determination o f possible phase changes during solubility studie s The m-CPPD an d t-CPPD crystal s used in the solubilit y experiment s did not undergo any phase changes during their incubations in buffer. Thi s was verified by the fact tha t ther e was no change in the XRPD patterns o f the crystals afte r a  1 7 hour incubation in pH 2.68 glycine-HCl buffer a t 30° C (Figure 25) or after a  4 day incubation in pH 2.70 glycine-HCl buffer a t 60° C (Figure 26) . Furthermore , thermal analysi s of t-CPPD an d m-CPPD crystal s which had been incubated fo r 1  week in distilled water a t 45° C, revealed tha t no change in the hydration level of the crystals had occured . 127 2.7. Hydration phase diagra m A method of constructing an hydration phase diagram (Haleblian 1975 ) is to store samples of crystals a t different relativ e humidities, a t constan t temperature an d pressure, and let them equilibrat e a t a  stable hydratio n level. Th e hydration level of the crystals can then be reported a s a function of the relative humidity . Storage of m-CPPD and t>CPPD as well as anhydrous calciu m pyrophosphate produced no change in the hydration level of the crystal s a s indicated by weight change during the course of the experiment , an d by the DSC scan of the anhydrous sample s of calcium pyrophosphate which had been stored in the 90% relative humidity chamber (Figure 30) and show no endothermic peaks due to dehydration. Th e results indicate that th e activation energy required to change the hydration level of the crystals to the stable hydration level was too great for the process to occur spontaneously . Dehydration o f the m-CPPD and t-CPPD crystals produced anhydrou s calcium pyrophosphates with XRPD scans shown in Figure 28 . Th e peaks that resulted from th e dehydration of the CPPD samples match the reporte d pattern for p-calciu m pyrophosphate (Mcintos h and Jablonski 1956) . 3. IgG binding to CPPD crystal s The standard curve of emission fluorescence intensit y as a function of FITC-IgG concentration is linear over the concentration range of 0.1 to 100 ng/mh FITC-IgG . Indirect binding data for both m-CPPD and t-CPPD crystals is reported in Table 1 6 as pig of FITC-IgG bound per mg of crystals a s well as th e concentration o f FITC-IgG (jig/raL)  that remains unbound a s a  function o f the original FITC-IgG concentration. Th e amount of FITC-IgG bound per mg of 128 crystals is also shown graphically in Figure 32 for binding to m-CPPD crystals and in Figure 33 for binding to t-CPPD crystals. Th e error in measurement o f unbound FITC-IgG was never more than 7 % of the initia l FITC-IgG concentration but a t high FITC-IgG concentrations thi s translate s to a relatively large error in the measurement o f bound FITC-IgG. Nevertheless the data shows that m-CPPD crystals bind about twice as much FITC-IgG per mg of crystal as do t-CPPD crystals . Direct binding data for both m-CPPD and t-CPPD crystals is reported in Table 1 7 as jig of FITC-IgG bound per mg of crystal a s a function o f original FITC-IgG concentraion. Th e binding curves generated from this dat a are shown in Figure 34 for m-CPPD and in Figure 35 for t-CPPD. A t low initial FITC-IgG concentrations there is no significant differenc e i n the amount of FITC-IgG bound per mg of m-CPPD or t-CPPD, but a t highe r initial FITC-IgG concentrations the m-CPPD is shown to bind roughly twice as much IgG per mg as t-CPPD crystals . The amount of FITC-IgG that binds to the crystal surface depend s on the affinit y th e FITC-IgG has for the crystal surface an d on the surface are a available for binding (Martin 1993) . Th e surface activit y of proteins can be governed by several factors suc h as protein size (larger molecules have more contact points), charge, and structural aspect s of the protein (Brash an d Horbett 1987) . Wit h IgG adsorption to crystals, charge has been shown to play a significant rol e as the relatively positive Fab portion of IgG binds to negatively charged surface s (Hasselbache r 1979 ; Kozin and McCarty 1980). Based on this one would expect that the FITC-IgG molecules should have a greater affinit y fo r the t-CPPD crystals than the m-CPPD crystal s sinc e the zeta potential measurements showe d the t-CPPD crystal surface t o have a greater negative charge than m-CPPD. Th e fact tha t m-CPPD crystals bind 129 more FITC-IgG per mg of crystal indicates that charge is not the sole determinant o f binding affinity o f FITC-IgG to CPPD. Base d on the marke d differences i n habit an d size range of m-CPPD compared to t-CPPD, m-CPPD will have a  much greater surfac e are a per mg than t-CPPD and therefore wil l bind more FITC-IgG. The measurement o f FITC-IgG binding by the direct method showe d that th e amount of FITC-IgG bound per mg of crystals was in the order of 3-5 times less than the amoun t of bound FITC-IgG measured using the indirec t method. A  possible explanation for this is that when incubated with crystal s the FITC-IgG molecules could either be free i n solution, bound tightly to crystals, or loosely associated with the crystals and the proteins bound to them. Becaus e washing steps would remove al l but the tightly bound FITC-IgG the direct method measures only the tightly bound FITC-Ig G molecules. Th e indirect method, however, subtract s the free FITC-Ig G in the supernatant fro m th e original FITC-IgG present an d assumes the remainder , which is made up of both tightly bound FITC-IgG and loosely associated FITC-IgG, is the bound amount . A s a result the indirect method reported a higher amoun t of bound FITC-IgG per mg of crystals than did the direc t method. Scatchard plots are perhaps the most widely used method fo r extracting binding data from drug-recepto r binding studies (Nogrady 1988) . Under a  given se t of binding conditions, that is the drug binding to a specifi c receptor with a  specific binding constant , the plot is linear. I f there ar e multiple binding sites involved with different bindin g constants then th e Scatchard plo t will not be linear. Scatchar d plots are only valid if the bindin g sites ar e independent an d no cooperative interactions exis t between the m (Nogrady 1988) . 130 Scatchard plots have also been used to study protein adsorption t o solid surfaces (Blanco et al. 1989 ; Hutchens an d Yip 1990). A  linear Scatchard plo t would indicate that the protein bound to the surface via a single class of operational binding sites. Whe n Scatchard plots were used to describe the direct binding data of FITC-IgG to m-CPPD (Figure 36) and t-CPPD (Figure 37) they were found to be non-linear indicatin g nonindependent or nonidentical binding sites. Thi s is not an unexpected result sinc e different crysta l faces of the same crystal represent differen t shear planes of the crystal lattice. A s a result the faces possess differen t physicochemical characteristic s suc h as different surfac e free  energies due to the expression of different atom s on their surface , which may lead to preferential bindin g of proteins on some faces. Thi s was found to be the case for bone acidic proteins which adsorbed preferentially on the (100) face of hydroxy apatite crystals (Fujisawa an d Kuboki 1991) . In a n effort t o characterize the binding isotherms the data was firs t fitted t o the Langmuir equation . Figur e 38 shows the fit of the m-CPP D binding data to the Langmuir equation, an d Figure 40 shows the fit of the t-CPPD binding data to the Langmuir equation . Th e nonlinear nature of these curves shows that the Langmuir equation does not adequately describ e the binding of FITC-IgG to CPPD crystals . Th e Langmuir equatio n assume s hard non-deform able spheres adsorbing to a homogenous surface withou t interactions between the adsorbin g molecules (Brash and Horbett 1987) . This may not be a realistic model to describe protein adsorption to surface s (Brash and Horbett 1987 ) however since adsorbed proteins are flexible, an d once adsorbed can unfold an d attach to the surface wit h many contact points. These contact points may take a  variety of forms suc h as covalent bonds, coulombic attraction, hydrogen bonds, and hydrophobic interactions. Th e 131 number o f contact points formed is an equilibrium process and as a result th e strength with which a  protein is bound to a surface ma y change over time as more or less contact points ar e formed. Th e flexibility an d complexity of the proteins als o allows for potential interactions between adsorbed proteins . Given this it is not surprising that the Langmuir equatio n was inadequate i n describing the adsorptio n of FITC-IgG to CPPD crystals . An alternative wa s to describe the adsorption of FITC-IgG to CPPD crystal surfaces usin g the Freundlich equation (Bras h and Horbett 1987 ) which has traditionally been used to describe heterogeneous bindin g (Adamson 1990) . Th e binding data for FITC-IgG binding to m-CPPD (Figur e 39) and t-CPPD (Figure 41) showed a good fit to the Freundlich equation . The constant k  from th e Freundlich equation relates to the capacity of binding while the inverse of n relates to the affinit y o f binding to the surfac e (Adamson 1990) . At concentrations of FITC-IgG from 1 0 - 200 /zg/mL, the amount s of FITC-IgG bound to m-CPPD and t-CPPD crystals are very similar (Table 17), whereas a t protein concentrations greate r than 732 fig/mL, m-CPP D binds significantly greate r amount s of FITC-IgG than t-CPPD. Th e binding curves for m-CPPD and t-CPPD (Figures 34 and 35) show no evidence of reaching maximum binding of IgG even at very high protein concentrations of 4000 jiglmL. Give n that the Freundlich equatio n describes heterogeneou s solute adsorptio n to surfaces, i t is likely that FITC-IgG adsorbs to the crysta l surfaces wit h different affinitie s o n different region s of the sam e crystal . The Freundlich equation constants given in Table 1 8 indicate tha t FITC-IgG has a  higher affinit y fo r m-CPPD than t-CPPD but that trCPP D has a  greater capacit y for binding FITC-IgG than doe s m-CPPD. Howeve r this interpretation o f these constants maybe complicate d by several factors . 132 The crystals differed i n their crysta l habits an d protein adsorption may have occurred preferentially on certain faces of the crystals due, for example, to differing charg e densities on the different faces . A s discussed above , adsorption of FITC-IgG to the crystal s may have occurred with differen t affinities. Hence , a single binding affinity constan t and binding capacit y value for each crystal type may not accurately represent the nature of the heterogeneous binding of FITC-IgG to m-CPPD and t-CPPD crystals . 133 4. Summar y 1. Reliabl e methods for the synthesi s of t-CPPD and m-CPPD crystal s were obtained. Crystal s of o-CPPT were synthesized but the metho d proved unreliable . 2. Th e t-CPPD and m-CPPD crystal s were characterized by XRPD, thermal analysis , FTIR, SEM, zeta potential measurements, an d particle siz e analysis . 3. Th e solubility of m-CPPD was shown to be in the order of 15 times greater a t low pH than a t neutral pH. 4. Th e solubility of both t-CPPD and m-CPPD were shown to increase with temperature. Withi n the temperature range studied (3 - 61°), m-CPPD was shown to be the metastable form of CPPD by virtue of its having a  higher solubility than t-CPPD. B y extrapolation, th e transition temperature between m-CPPD an d t-CPPD was estimate d to be 163° C. 5. Bindin g curves were constructed to describe the binding of IgG to both t-CPPD an d m-CPPD crystals . Mor e IgG binds to m-CPPD crystal s than to t-CPPD per mg of crystal. Th e binding was shown to be heterogeneous ove r the crysta l surfaces by non-linear Scatchar d plot s and by the fit of the binding data to the Freundlich equation . REFERENCES U.S. Pharmacopeia XXII/National Formular y XVII.Rockville, MD: United States Pharmacopeia l Convention , Inc. , 1990 Absolom DR, Neumann AW. Modification o f Substrate Surface Propertie s through Protein Adsorption. Colloids and surfaces  1988;30:25-45 . Adamson AW. Physical Chemistry  of  Surfaces. (5th  ed.)  New York: John Wiley & Sons, Inc., 199 0 Bauer HH, Christian GD , O'Reilly JE. Instrumental Analysis  .Boston : Allyn and Bacon, Inc., 197 8 Bennett RM, Lehr JR, McCarty D J Jr . Factors Affecting th e Solubility of Calcium Pyrophosphate Dihydrat e Crystals . The Journal of  Clinical Investigation 1975;56:1571-1579 . Blanco R, Arai N, Grinberg N, Yarmush DM, Karger BL. Role of association on protein adsorptio n isotherms. J3-lactoglobulin A adsorbed on a weakly hydrophobic surface. Journal of  Chromatography 1989;482:1 -12. Brash JL, Horbett TA, ed. Proteins at  interfaces.  Physicochemical  and Biochemical Studies.  Washington : American Chemical Society , 1987:(Comstock MJ, ed. ACS Symposium Series ; vol 343). Brown EH, Lehr JR, Smit h JP, Frazier AW. Preparation an d Characterization of Some Calcium Pyrophosphates. Agricultural and Food Chemistry  1963;ll(3):214-222 . 135 Brown WE, Gregory TM. Calcium Pyrophosphate Crysta l Chemistry . Arthritis and  Rheumatism  1976;19(3):446-463 . Burt HM, Evans E, Lam EWN, Gehrs PF, Herring FG. Membranolytic effect s of monosodium urate monohydrate : influence o f grinding. Journal of Rheumatology 1986;13(4):778-783 . Burt HM, Jackson JK. Characterization an d Membranolytic Effects o f Triclinic Calcium Pyrophosphate Dihydrate Crystals . Journal of Rheumatology 1987;14(5):968-973 . Burt HM, Jackson JK. Enhancement o f crystal induced neutrophil response s by opsonisation of calcium pyrophosphate dihydrate crystals . Annals of the Rheumatic Diseases  1993;52:599-607 . Burt HM, Jackson JK, Rowel l J. Calcium Pyrophosphate an d Monosodiu m Urate Crysta l Interactions with Neutrophils: Effect o f Crystal Size and Lipoprotein Binding to Crystals. Journal of  Rheumatology 1989;16(6):809-817. Burt HM, Kalkman PH, Mauldin D. Membranolytic effects o f crystalline monosodium urate monohydrate. Journal of  Rheumatology 1983;10(3):440-448. Chen PS Jr., Toribara TY, Warner H. Microdetermination o f Phosphorus. Analytical Chemistry  1956;28(11):1756-1758 . Cheng P, Pritzker KPH. The Effect o f Calcium an d Magnesium Ions on Calcium Pyrophosphate Crysta l Formation in Aqueous Solutions . Journal of  Rheumatology 1981;8(5):772-782 . Cherian PV, Schumacher HR Jr. Immunochemical an d Ultrastructura l Characterization o f Serum Protein s associated with Monosodium Urat e Crystals (MSU) in Synovial Fluid Cells from Patient s with Gout . Ultrastructural Pathology  1986;10:209-219 . Creedy J. A Laboratory Manual  for  Schools and  Colleges.London:  Heineman n Educational Books, 197 7 Cullity BD. Elements of  X-ray diffraction.  (2nd  ed.)  Reading , MA: Addison-Wesley Publishing Company, Inc., 1978 Derfus BA , Rachow JW, Mandel NS, Boskey AL, Buday M, Kushnaryov VM, Ryan LM. Articular Cartilag e Vesicles Generate Calciu m Pyrophosphate Dihyrate-Like Crystal s in Vitro. Arthritis and Rheumatism 1992;35(2):231-240 . Ducheyne P, Kim CS, Pollack SR. The effect o f phase differences o n the time-dependent variation of the zeta potential of hydroxy apatite. Journal of Biomedical Materials  Research  1992;26:147-168 . Ellingsen JE. A Study on the Mechanism o f Protein Adsorption to TiO Biomaterials 1991;12:593-596 . Fujisawa R , Kuboki Y. Preferential Adsorptio n of Dentin an d Bone Acidic Proteins on the (100) Face of Hydroxyapatite Crystals . Biochimica et Biophysica Acta  1991;1075:56-60 . Goettsch RW. New Indicator for the Complexometric Titration of Calcium with EDTA. Journal of  Pharmaceutical Sciences  1965;54(2):317-318 . Gorbunoff MJ , Timasheff SN . The Interaction of Proteins wit h Hydroxyapatite. III . Mechanism. Analytical Biochemistry 1984;136:440-445. Haleblian J , McCrone W. Pharmaceutical application s of polymorphism. Journal of  Pharmaceutical Sciences  1969;58(8):911-929 . 137 Haleblian JK. Characterization o f habits and crystalline modification o f solids and their pharmaceutical applications . Journal of Pharmaceutical Sciences  1975;64(8):1269-1288 . Hasselbacher P . Binding of IgG and Complement Protein by Monosodium Urate Monohydrate an d Other Crystals . The Journal  of  Laboratory Clinical Medecine  1979;94(4):532-541 . Hasselbacher P . Crystal-Protein Interactions in Crystal-Induced Arthritis. In: Weissmann G , ed. Advances  in  Inflammation Research.  Ne w York: Raven Press, 1982 : vol 4). Hearn PR, Russell RGG. Formation of Calcium Pyrophosphate Crystal s in Vitro: Implications for Calcium Pyrophosphate Crysta l Depositio n Disease (Psuedogout) . Annals of  Rheumatic Diseases  1980;39:222-227 . Hutchens TW, Yip T-T. Protein interactions with immobilized transitio n metal ions: quantitativ e evaluations of variations in affinity an d binding capacity. Analytical Biochemistry  1990;191:160-168 . Kohn NN, Hughes RE, McCarty DJ Jr., Faires JS. The Significance o f Calcium Phosphate Crystal s in the Synovial Fluid of Arthritic Patients: Th e "Psuedogout Syndrome." II. Identification o f Crystals. Annals of  Internal Medicine  1962;56(5):738-745 . Kozin F, Ginsberg MH, Skosey JL. Polymorphonuclear Leukocyt e Response s to Monosodium Urate Crystals : Modificatio n by Adsorbed Seru m Proteins. The Journal of  Rheumatology 1979;6(5):519-526 . Kozin F, McCarty DJ Jr. Protei n Adsorption to Monosodium Urate , Calciu m Pyrophosphate Dihydrate , and Silica Crystals. Arthritis and Rheumatism 1976a ; 19(3):433-438. Kozin F, McCarty DJ Jr. Protein Binding to Monosodium Urat e Monohydrate, Calciu m Pyrophosphate Dihydrate , an d Silicon Dioxide Crystals. I. Physical Characteristics. The Journal of  Laboratory Clinical Medecine  1976b;89(6):1314-1324 . Kozin F, McCarty DJ Jr. Molecular Orientation of Immunoglobulin G Adsorbed to Microcrystalline Monosodium Urat e Monohydrate. The Journal of  Laboratory Clinical  Medecine  1980;95(l):49-58 . Mandel GS, Halverson PB, Rathbum M , Mandel NS. Calcium Pyrophosphat e Crystal Deposition: A Kinetic Study Using a Type I Collagen Gel Model. Scanning Microscopy  1990;4(1):175-180 . Mandel GS, Renne KM, Kolbach AM, Kaplan WD, Miller JD, Mandel NS. Calcium Pyrophosphate Crysta l Deposition Disease: Preparation an d Characterization o f Crystals. Journal of  Crystal Growth  1988;87:453 -462. Mandel NS. The Crystal Structure o f Calcium Pyrophosphate Dihydrate . Acta Crystallographica  1975;B31:1730-1734 . Mandel NS, Mandel GS . A Model for Human Calciu m Pyrophosphate Crysta l Deposition Disease: Crystallization Kinetic s in a  Gelatin Matrix . Scanning Electron  Microscopy  1984;4:1779-1792 . Markowitz H. Synovial Fluid Proteins. In: Dalt Y, ed. Proteins  in  Body Fluids, Amino  Acids,  and  Tumor  Markers:  Diagnostic and  Clinical Aspects. Ne w York: Alan R Liss, Inc., 1983: Martin A. Physical Pharmacy.  (4th  ed.)  Philadelphia: Lea & Febiger, 199 3 McCarty DJ Jr. Calciu m Pyrophosphate Dihydrate Crysta l Depositio n Disease -1975. Arthritis and  Rheumatism  1976;19(3):275-285 . 139 McCarty DJ Jr., Hogan JM, Gatter RA, Grossman M. Studies in pathologica l calcifications i n human cartilage . I . Prevalence and types of crystal deposits in the menisci of two hundred fifteen cadavera . The Journal of Bone and  Joint  Surgery  1966;48-A(2):309-325 . McCarty DJ, Kozin F. An overview of cellular and molecular mechanisms i n crystal-induced inflammation. Arthritis  and  Rheumatism 1975;18(6):757-764. McGuire MKB, Hearn PR, Russell RGG. Calcium Pyrophosphate Crystal s (Their Relevance to Calcium Pyrophosphate Dihydrate Crysta l Deposition Disease, Psuedogout, Chondrocalcinosi s an d Pyrophosphat e Arthropathy): Biochemical and Physiological Aspects. In: Maroudas A, Holborrow EJ, ed . Studies  in  Joint Disease  I.  Tunbridg e Wells, Kent: Pitman Medical Ltd., 1980 : 117-156. Mcintosh AO, Jablonski WL. X-ray diffraction powde r patterns o f the calciu m phosphates. Analytical Chemistry  1956;28(9) : 1424-1427. Nagase M, Baker DG, Schumacher HR Jr. Immunoglobulin G  Coating on Crystals an d Ceramics Enhances Polymorphonuclear Cel l Superoxid e Production: Correlation with Immunoglobulin G  Adsorbed. The Journal of  Rheumatology 1989;16(7):971-976 . Nogrady T. Medicinal chemistry:  a  biomedical approach.  (2nd  ed.)  New York: Oxford University Press, 198 8 Nyberg P, Klockars M. Effect o f Immunoglobulins on Mineral Dust-Induce d Production of Reactive Oxygen Metabolites by Human Macrophages . Inflammation 1990 ; 14(6621-629). Oka J, Kawahar a A. The Structure o f Synthetic Dimagnesiu m Diphosphate(V) Dihydrate. Acta Crystallographica  1982;B38:3-5 . 140 Okazaki T, Saito T, Mitomo T, Siota Y. Pseudogout: clinical observations an d chemical analysi s of deposits. Arthritis and  Rheumatism 1976;19(3):293-305. Raj PA, Johnsson M, Levine MJ, Nancollas GH. Salivary Statherin . Dependence on Sequence, Charge , Hydrogen Bonding Potency, an d Helical Conformation fo r Adsorption to Hydroxyapatite an d Inhibitio n of Mineralization. The  Journal of  Biological Chemistry 1992;267(9):5968-5976. Rao SR. Surface Phenomena.London:  Hutchinso n Educational Ltd. , 197 2 Revell PA. Examination o f Synovial Fluid. In: Berry CL, ed. Bone  and Joint Disease. Heidelberg : Springer-Verlag, 1982 : Current Topics in Pathology; vol 71. Roch-Arveiller M, Legros R, Chanaud B , Muntaner O , Strzalko S, Thuret A, Willoughby DA, Giroud JP. Inflammatory reaction s induced by various calcium pyrophosphate crystals . Biomedicine and  Pharmacotherapy 1990;44:467-474. Ryan LM, McCarty DJ Jr. Calcium Pyrophosphate Crysta l Depositio n Disease; Psuedogout; Articular Chondrocalcinosis . In: McCarty DJ Jr. , ed. Arthritis  and  Allied  Conditions.  Philadelphia : Lea and Febinger , 1989: 1711-1735. Terkeltaub R, Curtiss LK, Tenner AJ, Ginsberg MH. Lipoproteins Containin g Apoprotein B are a  Major Regulato r of Neutrophil Responses to Monosodium Urate Crystals . The Journal of  Clinical Investigation 1984;73:1719-1730. Terkeltaub R, Tenner AJ, Kozin F, Ginsberg MH. Plasma Protein Binding by Monosodium Urate Crystals . Analysis by Two-Dimensional Ge l Electrophoresis. Arthritis and  Rheumatism  1983;26(6):775-783 . 141 Terkeltaub RA, Santoro DA, Mandel G, Mandel N. Serum an d Plasma Inhibi t Neutrophil Stimulatio n by Hydroxyapatite Crystals . Arthritis and Rheumatism 1988;31(9):1081-1089 . Watanabe W, Baker DG, Schumacher HR Jr. Comparison of the acut e inflammation induce d by calcium pyrophosphate dihydrate , apatit e and mixed crystals in the rat ai r pouch model of a synovial space . Journal of  Rheumatology 1992;19:1453-1457 . Weast RW, ed. Handbook of  Chemistry and  Physics.  48th  ed.  Cleveland : The Chemical Rubber Company, 1967 : Willard HH, Merritt LL Jr., Dean JA, Settle FA Jr. Instrumental Methods  of Analysis. (6th  ed.)  Belmont : Wadsworth Publishing Company, 198 1 Xu Y, Cruz T, Cheng P, Pritzker KPH. Effects o f Pyrophosphatase o n Dissolution o f Calcium Pyrophosphate Dihydrate Crystals . The Journal of  Rheumatology 1991a ; 18(1):66-71. Xu Y, Cruz TF, Pritzker KPH. Alkaline Phosphatase Dissolves Calciu m Pyrophosphate Dihydrate Crystals . The Journal of  Rheumatology 1991b;18(10):1606-1610. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0087420/manifest

Comment

Related Items