Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Investigating biomolecular structure using fiber-optic UV resonance Raman spectroscopy Addison, Christopher James 2011

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

Item Metadata

Download

Media
24-ubc_2011_fall_addison_christopher.pdf [ 17.5MB ]
Metadata
JSON: 24-1.0062206.json
JSON-LD: 24-1.0062206-ld.json
RDF/XML (Pretty): 24-1.0062206-rdf.xml
RDF/JSON: 24-1.0062206-rdf.json
Turtle: 24-1.0062206-turtle.txt
N-Triples: 24-1.0062206-rdf-ntriples.txt
Original Record: 24-1.0062206-source.json
Full Text
24-1.0062206-fulltext.txt
Citation
24-1.0062206.ris

Full Text

Investigating Biomolecular Structure Using Fiber-Optic UV Resonance Raman Spectroscopy  by  Christopher James Addison  B.Sc. (Hons), University of Victoria, 2002 M.Sc., University of Victoria, 2005  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  The Faculty of Graduate Studies  (Chemistry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  August 2011  © Christopher James Addison, 2011  ii Abstract  The observation of biomolecular structure is critical for the fundamental understanding of biological function. In this work, Fiber-optic UV Resonance Raman Spectroscopy (FO-UVRRS) was employed to study a number of structure-function relationships.  A suite of metal-containing dioxygenase enzymes were studied in order to determine the substrate protonation state in the enzyme-substrate complex. Two enzymes with a high degree of sequence identity were studied, where one naturally incorporates a Fe2+ metal and the other uses Mn2+. Both enzymes react with the same substrate to form an equivalent muconic semialdehyde product. The enzymes are capable of incorporating the non-native metal into the enzymes with no effect on the kinetic properties. The Raman results show that the nature of the metal has no effect on the substrate protonation state.  The degradation of 2-hydroxy-6-oxo-6-phenylhexa-2,4-dienoic acid (HOPDA) by BphD in the biphenyl catabolic pathway was studied using FO-UVRRS. Previous reports suggest that enol-keto tautomerization was a critical step in the mechanism of the carbon-carbon bond hydrolase. Hydrolytically impaired variants of BphD were used to study the binding mode of HOPDA. The results show that HOPDA binds to these variants in a strained-enolate geometry and does not undergo an enol-keto tautomerization upon enzyme binding.  A fundamental FO-UVRRS study of locked nucleic acid (LNA) olgiomers was performed. LNA bases contain a C4’ to O2’ methylene bridge within the furanose ring of the nucleotide. The research results show that incorporation of a LNA induces a conformational  iii change in the glycosyl bond between the backbone and the base. The results also show that incorporation of an LNA base induces changes in the secondary structure of the nucleic acid.  In another study, chemical contamination of synthetic DNA oligomers was observed using FO-UVRRS. The data showed that commercially-available oligomers purified using standard desalting conditions are contaminated with residual benzamide. The results show that a previously assigned –NH2 scissors vibration in the Raman study of oligomers overlaps with a prominent benzamide vibrational band, calling into question these previous assignments. The results also demonstrate that the extent of benzamide contamination varies from sample-tosample and between different commercial sources.  iv Preface  Chapter Two represents the culmination of a collaborative research project between myself, Katherine Yam (Department of Biochemistry and Molecular Biology, The University of British Columbia), Erik Farquhar (Department of Chemistry and Center for Metals in Biocatalysis, University of Minnesota), along with our respective research supervisors. The research program was designed equally by me and Katherine Yam. Erik Farquhar was responsible for obtaining and purifying all relevant protein samples using methodology that had been developed in his laboratory. Katherine Yam performed further purification on the protein, prepared all samples for analysis and acquired the UV-Vis spectra shown here. All Raman experiments, and the subsequent data processing and analaysis, were performed by me. This chapter was written solely by me, with editing and helpful suggestions provided by Katherine Yam, Mike Blades and Robin Turner. This chapter is currently being prepared for publication.  Chapter Three was a collaborative effort between myself, Geoff Horsman (Department of Biochemistry and Molecular Biology, The University of British Columbia) and our respective research supervisors. Geoff Horsman prepared all samples and acquired all UV-Vis spectra shown here. I performed all Raman experiments along with the subsequent data processing and analysis. Density functional theory calculations used to support this work were performed by Steven Hepperle (Department of Chemistry, The University of British Columbia). The manuscript was written equally by me and Geoff Horsman. Editing of the manuscript was performed by all involved in the work. This chapter is currently being prepared for publication.  v Chapter Four was a collaborative effort involving me, Stanislav Konorov (Department of Chemistry and Michael Smith Laboratories, The University of British Columbia) and Georg Schulze (Michael Smith Laboratories, The University of British Columbia). The research plan was formulated equally by Stanislav Konorov, Georg Schulze and I. Experimental data were acquired equally by me and Stanislav Konorov. The manuscript was written equally by Stanislav Konorov, Georg Schulze and I. The manuscript was revised and edited by all authors. Chapter Three has been published previously:  Konorov, S. O.; Schulze, H. G.; Addison, C. J.; Haynes, C. A.; Blades, M. W.; Turner, R. F. B., Ultraviolet resonance Raman spectroscopy of locked single-stranded oligo(dA) reveals conformational implications of the locked ribose in LNA. Journal of Raman Spectroscopy 2009, 40(9), 1162-1171  Chapter Five represents a continuation of the LNA work discussed in Chapter Four. In this case, the research plan was formulated primarily by me, with input from Stanislav Konorov. Experimental data was acquired equally by me and Stanislav Konorov. The manuscript was primarily written by me, with input from Stanislav Konorov and Georg Schulze. The manuscript was revised and edited by all authors. Chapter Five has been published previously:  Addison, C. J.; Konorov, S. O.; Schulze, H. G.; Turner, R. F. B.; Blades, M. W., Residual benzamide contamination in synthetic oligonucleotides observed using UV resonance Raman spectroscopy. Journal of Raman Spectroscopy 2011, 42(3), 349-354.  vi Table of Contents Abstract ..................................................................................................................................... ii Preface...................................................................................................................................... iv Table of Contents..................................................................................................................... vi List of Tables ........................................................................................................................... ix List of Figures .......................................................................................................................... xi List of Abbreviations ............................................................................................................ xxii Acknowledgements.............................................................................................................. xxvi Dedication ........................................................................................................................... xxvii Chapter One:  Introduction....................................................................................................1  1.1  Preamble ........................................................................................................................1  1.2  Molecular Vibrations and the Raman Effect .................................................................2  1.2.1  Classical Derivation of the Raman Effect................................................................10  1.2.2  Raman Selection Rules ............................................................................................13  1.3  Raman Signal Enhancement ........................................................................................15  1.4  Fiber-Optic Probes for Raman Spectroscopy ..............................................................26  1.5  Methods for Biological Investigations.........................................................................33  1.6  Thesis Structure ...........................................................................................................41  Chapter Two:  UV Resonance Raman Investigation of Metal-Containing Dioxygenases ..44  2.1  Introduction..................................................................................................................44  2.2  Materials and Methods.................................................................................................55  2.2.1  Enzyme Preparations ...............................................................................................55  2.2.2  Model Compound Synthesis ....................................................................................55  2.2.3  Preparation of Samples ............................................................................................56  2.2.4  UV-Vis Absorption Spectroscopy ...........................................................................56  2.2.5  UV Resonance Raman Spectroscopy.......................................................................57  2.3  Results and Discussion ................................................................................................58  2.3.1  Spectroscopic Analysis of Free HPCA Substrates...................................................58  2.3.2  Spectroscopic Study of Fe- and Mn-Bound HPCA Model Compounds .................71  2.3.3  Spectroscopic Study of Native and Non-Native HPCD and MndD Enzymes.........82  vii 2.4  Conclusions..................................................................................................................98  2.5  Acknowledgements......................................................................................................99  Chapter Three:  FO-UVRRS Study of a Carbon-Carbon Bond Hydrolase .....................100  3.1  Introduction................................................................................................................100  3.2  Materials and Methods...............................................................................................105  3.2.1  Chemicals and Protein ...........................................................................................105  3.2.2  Preparation of Samples ..........................................................................................105  3.2.3  Dissociation Constants...........................................................................................106  3.2.4  UVRR Measurements ............................................................................................107  3.2.5  Computational Methods.........................................................................................108  3.3  Results........................................................................................................................108  3.3.1  Dissociation constant of S112A variant for HOPDA ............................................108  3.3.2  UV Resonance Raman Spectroscopy of Free Substrates.......................................110  3.3.3  DFT Calculations ...................................................................................................114  3.3.4  UV Resonance Raman Spectroscopy of Enzyme-Substrate Complex ..................118  3.4  Discussion ..................................................................................................................122  3.5  Conclusions................................................................................................................124  3.6  Acknowledgements....................................................................................................125  Chapter Four:  FO-UVRRS of Locked Single-Stranded Oligonucleotides .......................126  4.1  Introduction................................................................................................................126  4.2  Materials and Methods...............................................................................................134  4.2.1  Oligomers and Sample Preparation .......................................................................134  4.2.2  UV Resonance Raman Spectroscopy.....................................................................135  4.2.3  Data Analysis .........................................................................................................136  4.3  Results and Discussion ..............................................................................................136  4.3.1  Evidence of Base Stacking.....................................................................................142  4.3.2  Raman Scattering From the Locked Base..............................................................144  4.3.3  Local and Stacking Effects of Locked Bases.........................................................149  4.3.4  Differential Stacking in DNA and LNA ................................................................155  4.4  Conclusions................................................................................................................157  4.5  Acknowledgements....................................................................................................158  viii Chapter Five:  Residual Contamination Observed in Synthetic Oligonucleotides ............159  5.1  Introduction................................................................................................................159  5.2  Experimental ..............................................................................................................162  5.3  Results and Discussion ..............................................................................................164  5.4  Conclusions................................................................................................................175  5.5  Acknowledgements....................................................................................................176  Chapter Six:  Conclusions................................................................................................177  6.1  UV Resonance Raman Investigation of Metal-Containing Dioxygenases ................177  6.2  FO-UVRRS Study of a Carbon-Carbon Bond Hydrolase .........................................179  6.3  FO-UVRRS of Locked Single-Stranded Oligonucleotides .......................................180  6.4  Residual Contamination Observed in Synthetic Oligonucleotides ............................182  6.5  Summary of Findings and Overall Impact of the Work ............................................183  References..............................................................................................................................185  ix List of Tables  Table 1.1: Summary of common analytical methods for biomolecular structure determination. 39 Table 2.1: Electronic absorption bands of HPCA in various protonation states. Data obtained under anaerobic conditions. Values in parentheses represent extinction coefficients in units of M-1 cm-1. Assignments are based on the notation described by Platt [121]..................... 61 Table 2.2: Band positions and intensities for the four possible ionization states of HPCA. Shifts of vibrations upon deuterium substitution are also noted. Band intensities and shape denoted by: (s) strong, (m) medium, (w) weak, (vw) very weak.......................................... 65 Table 2.3: Band positions and intensities for the Fe-HPCA and Mn-HPCA complex. HPCA is bound to the metal in the trianionic protonation state. Shifts of vibrations upon deuterium substitution are also noted, along with the shift induced between free (trianionic) and complexed HPCA. Band intensities denoted by: (s) strong, (m) medium, (w) weak.......... 79 Table 2.4: Band positions and intensities for enzyme-bound HPCA in the HPCD:HPCA complex. Shifts of vibrations upon deuterium substitution are also noted. Band intensities and shape denoted by: (s) strong, (m) medium, (w) weak, (vw) very weak, (sh) shoulder.. 93 Table 2.5: Band positions and intensities for enzyme-bound HPCA in the MndD:HPCA complex. Shifts of vibrations upon deuterium substitution are also noted. Band intensities and shape denoted by: (s) strong, (m) medium, (w) weak, (vw) very weak, (sh) shoulder.. 95 Table 3.1: Density Functional Theory (DFT) calculations for HOPDA in the enol(ate) (HOPDAe) and keto (HOPDAk) tautomeric states and there comparison to experimentally observed bands. Band intensities and shape denoted by: (s) strong, (m) medium, (w) weak, (vw) very weak. ................................................................................................................................... 115  x Table 4.1: Monomers, dimers, trimers, pentamers and dodecamers used to detect the presence of base stacking in single-stranded nucleic acids and to assess the influence of a locked ribose on its own and neighboring bases. For a given sequence, the locked and unlocked versions of the oligomers had different secondary but identical primary structures. Locked adenine bases are capitalized. The position of the locked base(s), relative to the 5’-end, is specified by the upper case letters in the locked nucleic acid sequences. The inset shows an adenine monomer with locked ribose (C2’ –C4’ bridge)................................................................. 131 Table 4.2: Adenine-related resonance Raman band assignments in nucleic acids, nucleic acid constituents, and their model compounds. ID: dA (deoxynucleoside, deoxyadenosine); rA (nucleoside, adenosine); dAMP (deoxynucleotide, deoxyadenosine monophosphate). Bonds: b = bend; d = deformation/scissors; s = stretch; Pyr = pyrimidine (6-ring); Im = imidazole (5-ring). Comments: w. = with; D = deuterium. ................................................ 147  xi List of Figures  Figure 1.1: Three possible light scattering processes from a molecule-photon interaction. The photon can be Rayleigh scattered, in which the outgoing photon has the same energy as the incident photon. The system can scatter a photon with less energy [h(ν0 – νvib)] than the excitation source in Stokes scattering. Alternatively, a photon of greater energy [h(ν0 + νvib)] can be scattered in the anti-Stokes process. Figure reproduced from [7] with permission. .............................................................................................................................. 5 Figure 1.2: Schematic representation of the (a) Stokes, (b) Rayleigh and (c) anti-Stokes scattering processes. Figure reproduced from [7] with permission. ...................................... 7 Figure 1.3: Raman spectrum of Chloroform obtained with 514.5 nm excitation. The Rayleigh scattered line occurs at 514.5 nm and corresponds to a 0 cm-1 Raman shift. Peaks to the right of the Rayleigh line correspond to Stokes scattering, and the scattered photons have less energy than the incident photons. Note that the Stokes scattering peaks are consistently higher in intensity than the corresponding anti-Stokes scattering peaks. Peaks to the left have higher energy than the incident photon and occur by way of anti-Stokes scattering. The absolute frequency and wavelength axis is also included for reference. Figure reproduced from [5] with permission. .................................................................................... 9 Figure 1.4: FTIR (upper) and Raman spectrum (lower) of oleic acid methyl ester. In the IR spectrum, the carbonyl (C=O) stretch is prominent while the alkene (C=C) stretch is very weak. Conversely, the Raman spectrum yields a strong signal for the alkene stretch and a weak signal for the carbonyl vibration. Figure reproduced from [5] with permission. ....... 15 Figure 1.5: Comparison of (a) resonance Raman and (b) Fluorescence pathways. In both cases the molecule moves to the excited electronic state by the excitation photon (υex). However,  xii in the resonance Raman process the molecule emits the Raman scattered photon (υs) from an excited vibrational level, whereas in Fluorescence spectroscopy the molecule first undergoes internal conversion before emitting the fluorescence photon (υfl). Figure reproduced from [23] with permission. ................................................................................ 18 Figure 1.6: Hypothetical absorption curve and Raman spectra for a molecule having two chromophores in the UV region, denoted by A and B. The intensity of the Raman vibrations is dependent upon the excitation wavelength used, demonstrating the selectivity possible in the resonance Raman process. ............................................................................ 22 Figure 1.7: Excitation of a fluorescent sample with 514.5 and 785 nm excitation. With 514.5 nm excitation, the fluorescence signal is several orders of magnitude higher than the Raman signal. With 785 nm excitation, the excitation photons are not energetic enough to excite the molecule to the higher electronic state necessary for fluorescence. Therefore, the Raman spectrum can be observed. Figure reproduced from [5] with permission. .............. 23 Figure 1.8: Overlap areas for (a) flush and (b) angled fiber-optic probes. The flush probes are suitable for non-absorbing analytes where path length considerations are not paramount. The angled probes are ideally suited for absorbing analytes as the overlap area is maximized and moved closer to the faces of the fibers. This minimizes the path length for the Raman scattered photons. Figure reproduced from [26] with permission. ...................................... 29 Figure 1.9: (a) Side view of an angled probe geometry used in typical FO-UVRRS applications. Raman excitation of the analyte molecule takes place very close the excitation fiber surface. The Raman scattered photons then pass through the side of the collection fiber, reflect off of the mirrored surface and into a guided mode of the collection fiber for transport to the detection system. (b) Top view of the FO probe showing the excitation fiber surrounded by six collection fibers. Figure reproduced from [46] with permission. .................................. 31  xiii Figure 2.1: Ring-cleaving reaction of catechol catalyzed by extradiol and intradiol enzymes. Both enzymes proceed via an iron-alkylperoxo intermediate, but the carbon-carbon bond cleaved depends on the class of enzyme used. Figure reproduced from [64] with permission. ............................................................................................................................ 45 Figure 2.2: Reaction of homoprotocatechuate (HPCA) with HPCD or MndD to yield a muconic semialdehyde product. .......................................................................................................... 46 Figure 2.3: Initial proposals for the catalytic cycle of extradiol ring cleaving dioxygenases. The mechanism proposed transient change in the metal oxidation state from Fe(II) to Fe(III) over the course of the reaction, and subsequent return to the Fe(II) oxidation state at the conclusion of the catalytic cycle. Figure reproduced from [115] with permission. ............ 48 Figure 2.4: Alternative mechanism proposed for the HPCD-catalyzed reaction of HPCA. Unlike the mechanism proposed in Figure 2.3, no formal change in the metal oxidation state occurs over the course of the reaction. Figure reproduced from [73] with permission. Copyright 2008 National Academy of Sciences, U.S.A. ....................................................................... 51 Figure 2.5: Four possible protonation states of HPCA. HPCA has pKa values of 4.2, 9.5 and 12. ............................................................................................................................................... 54 Figure 2.6: Anaerobic solvent-subtracted UV-Vis spectra of HPCA in the four possible protonation states. (a) Neutral HPCA, 50mM MOPS pH 3 (b) Monoanionic HPCA, 50mM MOPS pH 7 (c) Dianionic HPCA, 50mM MOPS pH 11 (d) Trianionic HPCA in 250 mM sodium t-butoxide/t-butanol solution ~ pH 14...................................................................... 60 Figure 2.7: FO-UVRRS spectra of HPCA in each of the four possible protonation states. Left panel: HPCA in H2O. (a) 3.0 mM HPCA, phosphate buffer pH 3.0, 100 mM Na2SO4; (b) 1.0 mM HPCA, phosphate buffer pH 7.0, 100 mM Na2SO4; (c) 0.3 mM HPCA, phosphate buffer pH 11.0, 100 mM Na2SO4 and (d) 0.5 mM HPCA, phosphate buffer pH 14.0, 100  xiv mM Na2SO4. Right panel: HPCA in D2O. (a) 3.0 mM HPCA, phosphate buffer pD 3.0, 100 mM Na2SO4; (b) 1.0 mM HPCA, phosphate buffer pD 7.0, 100 mM Na2SO4; (c) 0.3 mM HPCA, phosphate buffer pD 11.0, 100 mM Na2SO4 and (d) 0.5 mM HPCA, phosphate buffer pD 14.0, 100 mM Na2SO4. Spectra acquired at 248 nm, 20 mW power and 40 s total acquisition time..................................................................................................................... 64 Figure 2.8: Gaussian peak-fitting analysis for different protonation states of HPCA. Left panel: Neutral HPCA (pH 3) in (a) H2O and (b) D2O. Right panel: Monoanionic HPCA (pH 7) in (c) H2O and (d) D2O. The red lines represent individual peaks obtained from peak fitting while the black line is the raw data. ...................................................................................... 67 Figure 2.9: Gaussian peak-fitting analysis for different protonation states of HPCA. Left panel: Dianionic HPCA (pH 11) in (a) H2O and (b) D2O. Right panel: Trianionic HPCA (pH 14) in (c) H2O and (d) D2O. The red lines represent individual peaks obtained from peak fitting while the black line is the raw data. ...................................................................................... 69 Figure 2.10: Fe- and Mn-HPCA model compounds. The metal centers are in the +3 oxidation state and the substrate is bound in the trianionic form.......................................................... 72 Figure 2.11: UV-Vis Spectra of the model compounds (a) 100 µM Fe-HPCA in 50 mM MOPS pH 7.5 and (b) 100 µM Mn-HPCA in 50 mM MOPS pH 7.5. ............................................. 73 Figure 2.12: FO-UVRR spectra of metal-HPCA model compounds. Left Panel: Fe-HPCA model compound in (a) H2O and (b) D2O with 100 mM SO42- internal standard. Right Panel: MnHPCA model compound in (c) H2O and (d) D2O with 100 mM SO42- internal standard. Spectra acquired at 248 nm, 20 mW power and 40 s total acquisition time......................... 75 Figure 2.13: Gaussian peak-fitting analysis of metal-HPCA model complexes shown in Figure 2.12. Left panel: Fe-HPCA complex in (a) H2O and (b) D2O. Right panel: Mn-HPCA  xv complex in (c) H2O and (d) D2O, respectively. The red lines represent individual peaks obtained from peak fitting while the black line is the raw data. ........................................... 77 Figure 2.14: UV-Vis difference spectra of the HPCA substrate bound in the anaerobic enzymesubstrate complex with (a) Fe-HPCD (b) Mn-HPCD (c) Co-HPCD. [Enzyme] = 30 µM, [HPCA] = 20 µM, 50mM MOPS pH 7.5.............................................................................. 83 Figure 2.15: UV-Vis difference spectra of the HPCA substrate bound in the anaerobic enzymesubstrate complex with (a) Mn-MndD and (b) Fe-MndD. [Enzyme] = 30 µM, [HPCA] = 20 µM, 50 mM MOPS pH 7.5. .................................................................................................. 85 Figure 2.16: FO-UVRRS of enzyme samples with (black trace) and without (blue trace) substrate, along with the corresponding difference spectrum. Left Panel: Fe-HPCD in (a) H2O and (b) D2O. Center Panel: Mn-HPCD in (c) H2O and (d) D2O. Right Panel: CoHPCD in (e) H2O and (f) D2O. Spectra acquired at 248 nm, 20 mW power and 40 s total acquisition time. [Enzyme] = 175 µM, and if applicable, [Substrate] = 150 µM. ............... 88 Figure 2.17: FO-UVRRS of enzyme samples with (black trace) and without (blue trace) substrate, along with the corresponding difference spectrum. Left Panel: Mn-MndD in (a) H2O and (b) D2O. Right Panel: Fe-MndD in (a) H2O and (b) D2O. Spectra acquired at 248 nm, 20 mW power and 40 s total acquisition time. [Enzyme] = 175 µM, and if applicable, [Substrate] = 150 µM. ........................................................................................................... 89 Figure 2.18: Gaussian peak-fitting analysis of the enzyme-bound substrate difference spectrum in the diagnostic 1500-1700 cm-1 region. Left Panel: HPCA bound to Fe-HPCD in (a) H2O and (b) D2O. Center Panel: HPCA bound to Mn-HPCD in (c) H2O and (d) D2O. Right Panel: HPCA bound to Co-HPCD in (e) H2O and (f) D2O. The red lines represent individual peaks obtained from peak fitting and the black line is the raw data.................... 92  xvi Figure 2.19: Gaussian peak-fitting analysis of the enzyme-bound substrate difference spectrum in the diagnostic 1500-1700 cm-1 region. Left Panel: HPCA bound to Mn-MndD in (a) H2O and (b) D2O. Right Panel: HPCA bound to Fe-HPCD in (c) H2O and (d) D2O. The red lines represent individual peaks obtained from peak fitting and the black line is the raw data. ............................................................................................................................................... 94 Figure 3.1: Proposed Mechanism of BphD. Absolute conformation of each tautomer is not intended............................................................................................................................... 101 Figure 3.2: Comparison of the planar and non-planar conformations of enzyme-bound HOPDA. (a) 4 µM HOPDA in aqueous buffer (blue), the transient intermediate obtained upon mixing 8 µM BphD with 4 µM HOPDA (black) [161] and a mixture of 8 µM S112A and 4 µM HOPDA (red) [163]. (b) Crystal structures of HOPDA bound to the S112A/H265A variant of BphD (white) as a planar enolate (E:Se), and HOPDA bound to the S112A variant (straw) in a non-planar fashion (E:Sred) [162]. Oxygen and nitrogen atoms are red and blue, respectively. ........................................................................................................................ 103 Figure 3.3: The binding of HOPDA by S112A as monitored by absorption spectroscopy. (a) Titration of 2 µM S112A with increasing concentrations of HOPDA. (b) Plot of concentrations of bound ([E:S]) versus unbound ([S]f) HOPDA. The red line represents the fit of equation (12) to the data (black squares) with [E]0 = 2.0 µM, and Kd = 1.0 ± 0.2 µM. ............................................................................................................................................. 110 Figure 3.4: UVRR spectra of (a) HOPDA pH 6 (b) HOPDA pH 7.5 (c) HOPDA pH 9 (d) 3chloro HOPDA (3CH) pH 7.5. The UVRR spectra were obtained using potassium phosphate (I = 100 mM), 25 mM KNO3, solution pH as indicated, 25 °C......................... 112  xvii Figure 3.5: Resonance Raman excitation profile of HOPDA at accessible wavelengths in the UV region. Markers are identified as follows: + – 1598 cm-1 vibration; ● – 1668 cm-1 vibration; ♦ – 1300 cm-1 vibration; ■ - 1350 cm-1 vibration. .............................................................. 114 Figure 3.6: Simulated Raman spectra of (a) HOPDAe and (b) HOPDAk using the results of the DFT calculations. Spectra were simulated using lorentzian line shapes with a peak width of 50 cm-1. Individual vibrations are shown as grey lines and the sum of the vibrations is the black line............................................................................................................................. 117 Figure 3.7: UVRR difference spectra of (a) E:HOPDA-E (E = 150 µM, S = 140 µM) (b) E:3CHE (E = 150 µM, S = 140 µM) (c) S112A/H256A:HOPDA-S112A/H256A (E = 150 µM, S = 140 µM). The UVRR spectra were obtained using potassium phosphate (I = 100 mM), 25 mM KNO3, pH 7.5, 25 °C................................................................................................... 119 Figure 3.8: Close-up view of the UVRR difference spectra in the 1500-1700 cm-1 region. The solid black line corresponds to the original data, while the dashed grey lines are individual peaks resulting from the peak fitting analysis. The sum of the individual peaks is represented by the solid red line. (a) E:HOPDA-E (b) E:3CH-E (c) S112A/H256A:HOPDA-S112A/H256A. ........................................................................... 121 Figure 4.1: Comparison of 2’-deoxyribose and ribose structure in DNA and LNA, respectively. LNA contains a methylene bridge connecting the 2’ oxygen and 4’carbon....................... 127 Figure 4.2: Helical structure of A-DNA (left), B-DNA (center) and Z-DNA (Right). This image, created by Richard Wheeler (Zephyris) and deposited in the Wikimedia Commons, is used under the Creative Commons Attribution-Share Alike 3.0 Unported license. This image can be found at: http://en.wikipedia.org/wiki/File:A-DNA,_B-DNA_and_Z-DNA.png.......... 128  xviii Figure 4.3: (a) B-form. The dodecamer, at the interrogation temperature of ~5 °C, is in α’B-form [203, 207], as shown here, where the pyrimidine moieties of stacked bases have more extensive overlap than imidazole moieties. (b) A-form. In contrast to α’B-form, there is less overlap between bases in A-form conformations. The slight overlap that occurs is between imidazole and pyrimidine moieties as is evident from the two clearly visible adenines in the figure. However, the imidazole ring is closer to the modified ribose, and expected to be more influenced by the O2’, than the pyrimidine ring........................................................ 138 Figure 4.4: (a) Resonance Raman excitation with 257 nm light reveals a change in intensity when going from monomer to dimers, attributed to changes in energy levels due to base stacking in the dimers. (b) With 244 nm excitation differences in resonance Raman spectra are revealed that may originate from a combination of different energy levels, vibronic coupling, and solvent interactions....................................................................................... 143 Figure 4.5: Resonance Raman excitation with 257 nm light. (a) Locked adenine monophosphate exhibits increased scattering intensities, wavenumber downshifts (pronounced for the ~1481 cm− 1 band), and other imidazole-related spectral changes in Raman bands indicating that the ribose lock affects the glycosyl bond between ribose and adenine base. The spectra are scaled with regard to the ~1337 cm−1 peak to show wavenumber shifts better. (b) Although toned down, the increased scattering intensities and wavenumber downshifts are also evident in LNA oligomers compared to natural analogs (see also Figure 4.6 and Figure 4.7). ............................................................................................................................................. 146 Figure 4.6: Resonance Raman excitation of pentamers with 257 nm light. (a, b) In locked bases, both peak position and peak amplitude, respectively, for the ~1337 cm−1 Raman peak show effects that are nonadditive or only weakly additive. (c, d) In contrast, strong additive (i.e.  xix stepwise) effects are observed for the ~1481 cm−1 peak resulting from bands arising from molecular bonds closer to the ribose ring. .......................................................................... 151 Figure 4.7: Resonance Raman excitation of dodecamers with 257 nm light. (a, b) Peak position and peak amplitude, respectively, for the ~1337 cm−1 Raman band and (c, d) for the ~1481 cm−1 band. Additive changes are most pronounced in (c) where a smaller difference occurs between oligomers with a single locked base (LNAa – LNAg) as opposed to (a) where the difference between LNAa and LNAl and other oligomers with a single locked base is more evident. Note in (a) that where a locked base occurs at a strand end (LNAa, LNAl), changes are reduced compared to the no-lock (dA12) condition and about half as much as that of other single-lock oligomers (e.g. LNAg) indicating a stacking mediated influence on the nearest neighbor in addition to the presence of the lock itself............................................ 153 Figure 5.1: Protected 2’-deoxynucleosides used in the phosphoramidite synthesis. (a) 2’deoxyadenosine with a benzoyl protecting group at the N6 position and (b) 2’deoxycytidine with a benzoyl protecting group at the N4 position.  The protecting group in  each nucleoside is shown in red. The nucleosides are shown without the 2’-deoxyribose moiety for clarity................................................................................................................. 161 Figure 5.2: (a) Raman spectra of standard desalted DNA and LNA and HPLC-purified DNA pentamers purchased from IDT technologies. Raman spectra were obtained with 257 nm excitation. Note the small but distinct differences in the 1609 cm−1 band between all three samples. (b) With excitation at 244 nm, the 1609 cm−1 band in standard desalted DNA is dramatically enhanced, that of LNA less so, while it is absent in HPLC-purified DNA. The approximate enhancement of this band in DNA relative to LNA is 1.5............................. 165 Figure 5.3: The Raman spectrum of benzamide superimposed on the Raman difference spectrum of DNA–LNA purified by standard desalting only. The Raman difference spectrum of DNA  xx – LNA purified by standard desalting followed by HPLC is shown offset (lower). Identification of benzamide (inset shows structure) as the contaminant in standard desalting samples was due to differential UVRRS enhancement and contamination of benzamide in isosequential dodecamers. With excitation at 244 nm, the distortion created by the benzamide contaminant in the difference spectrum of the samples purified by standard desalting contrasts strongly with the difference spectrum of the same samples in the absence of contaminant (standard desalting followed by HPLC). Oligomers were purchased from IDT technologies................................................................................................................. 167 Figure 5.4: The resonance Raman excitation profile of the most prominent benzamide band (1609 cm−1) indicating that it is already considerably enhanced with 244 nm, compared to 257 nm, excitation. Benzamide is a potential contaminant of synthetic oligomers, and UVRRS with excitation near the DNA absorption maximum ca 260 nm may not reveal its presence in such samples, whereas excitation deeper in the UV does (see also Figure 5.2 and Figure 5.5). ................................................................................................................... 169 Figure 5.5: The UV absorption spectrum of 2 mM benzamide showing absorption maxima near 226 and 268 nm. The UV absorption spectrum confirms that UVRRS with excitation near the absorption maximum for DNA and a minimum for benzamide (ca 260 nm) would not provide strong evidence for its presence in synthetic oligomer samples, whereas excitation deeper in the UV would (see also Figure 5.2, Figure 5.4 and Figure 5.6). The DNA and LNA difference spectra (pentamers, standard desalting minus HPLC-purified, scaled by 12.4) are superimposed on the benzamide absorption spectrum for comparison. The DNA difference spectrum is virtually identical to that of benzamide after scaling and exhibits a red shift in the long wavelength absorption maximum, suggesting that the benzamide is hydrogen-bonded to the oligomer. The LNA difference spectrum shows a  xxi weak presence of benzamide as well as an absorption maximum at 260 nm that we attribute to failed oligomer sequences. Oligomer samples obtained from IDT technologies. .......... 171 Figure 5.6: Raman spectra of DNA in varying purification states with 244-nm excitation. (a) Parent spectra of DNA obtained from Operon Technologies. DNA purified using standard desalting or HPLC (top panel); Raman difference spectrum (bottom panel). Prominent positive peaks here are attributed to benzamide contamination. (b) Parent spectra of DNA obtained from Invitrogen. DNA purified using standard desalting or HPLC (top panel); Raman difference spectrum (bottom panel). Benzamide contamination is observed based on the Raman difference spectrum, but not to the same extent as observed in (a).................. 173  xxii List of Abbreviations  2H1C  2-His-1-carboxylate facial triad  3CH  3-chloro-HOPDA  3CHe  3-chloro HOPDA (enol tautomer)  3CHk  3-chloro HOPDA (keto tautomer)  4NC A Asp CCD  4-nitrocatechol Adenine Aspartic acid Charge-coupled device  CD  Circular dichroism  dA  Deoxyadenosine  dAMP dC DFT dG  Deoxyadenosine monophosphate Deoxycytidine Density Functional Theory Deoxyguanosine  DHB  2,3-dihydroxybiphenyl  DNA  Deoxyribonucleic acid  dsDNA  Double-stranded DNA  dT E EPR E:S  Deoxythymidine Enzyme Electron paramagnetic resonance Enzyme-substrate complex  xxiii E:Se  Enzyme-bound substrate, enol(ate) tautomer  E:Sk  Enzyme-bound substrate, keto tautomer  E:Sred  Enzyme-bound substrate, non-planar with redshifted absorption spectrum  E:Sse  Enzyme-bound substrate, strained enol(ate)  E:S-E  Enzyme-bound substrate difference spectrum  f-ESA  Femtosecond excited state transient absorption  F FO FO-UVRR FO-UVRRS FRET  Phenylalanine Fiber-optic Fiber-optic UV resonance Raman Fiber-optic UV resonance Raman spectroscopy Fluorescence resonance energy transfer  Glu  Glutamic acid  His  Histidine  HOPDA  2-hydroxy-6-oxo-6-phenylhexa-2,4-dienoic acid  HOPDAe  HOPDA (enol tautomer)  HOPDAk  HOPDA (keto tautomer)  HPCA  Homoprotocatechuic Acid / homoprotocatechol  HPCD  Homoprotocatechuate 2,3-dixoygenase  HPLC  High performance liquid chromatography  IR LMCT LNA  Infrared Ligand-to-metal charge transfer Locked nucleic acid  xxiv LRET MCP  Luminescence resonance energy transfer Meta-cleavage product  MOPS  3-morpholinopropane-1-sulfonic acid  MndD  Homoprotocatechuate 2,3-dioxygenase  NA  Nucleic acid  NIR  Near infrared  NMR PC  Nuclear magnetic resonance Protocatechol  PCB  Polychlorinated biphenyl  PNA  Peptide nucleic acid  RNA  Ribonucleic acid  RR RREP RRS RS S Sred  Resonance Raman Resonance Raman enhancement profile Resonance Raman spectroscopy Raman spectroscopy Substrate Enzyme-bound substrate, non-planar with redshifted absorption spectrum  SNR ssDNA UV UVRR UVRRS  Signal-to-noise ratio Single-stranded DNA Ultraviolet UV resonance Raman UV resonance Raman spectroscopy  xxv UV-Vis  Ultraviolet-visible  W  Tryptophan  Y  Tyrosine  xxvi Acknowledgements  Acknowledgements relating to each research project are contained in the acknowledgements section at the conclusion of each research chapter.  I have been fortunate to have had wonderful mentors during the course of my doctoral work. Much thanks goes to my mentors and research supervisors, Michael Blades and Robin Turner. Mike and Robin both gave me the freedom to explore my research while providing support when it was needed.  I also wish to thank two other wonderful scientists whom I had the fortune to work with. Thanks to Georg Schulze for his thorough critical evaluation of the science, and to Stanislav Konorov for embodying what a true scientist should be.  Lastly, I wish to thank my family for their patience, understanding, encouragement, and the occasional-but-necessary kick-in-the-butt during this work.  xxvii Dedication  For Joshua  May your yields be high, your products pure and your experiments always successful!  1  Chapter One: Introduction  1.1  Preamble The elucidation of biomolecular structure has been a major focus of scientific research  over the course of the 20th and 21st centuries. The identification of key structural elements as part of structure-function relationships has been critical from a fundamental scientific perspective. A variety of techniques have been developed to obtain this information, including fluorescence spectroscopy, NMR spectroscopy and X-ray crystallography. However, all of these techniques have strengths and weaknesses, and no one methodology is capable of producing a complete correlation of structure and function. A discussion of the advantages and disadvantages of these techniques is provided in Section 1.5.  Vibrational spectroscopy, and Raman spectroscopy (RS) in particular, has proven to be a valuable method for the elucidation of secondary and local (tertiary) biomolecular structure. The vibrational frequencies are dependent upon molecular geometry, bonding arrangements and the local molecular environment [1]. In turn, the vibrational spectra are rich in structural information which can be used to provide unique insights into biomolecular structure and function [2, 3].  As powerful as it is, RS cannot provide complete and irrefutable evidence of biomolecular structure and function. Instead, RS can be employed in conjunction with other  2  techniques to yield results that were previously unknown or inconclusive. This thesis demonstrates structural insights provided to several unique biologically relevant systems using Raman spectroscopy.  1.2  Molecular Vibrations and the Raman Effect At its core, Raman spectroscopy measures the frequency (νvib) of molecular vibrations by  way of their interaction with incident electromagnetic radiation. Molecular vibrations are routinely modelled to consist of two masses (of mass m1 and m2) connected by a spring with a force constant K [1]. The bond distance changes continuously due to the vibration but centers around an average value (the equilibrium distance). When the internuclear separation moves beyond this equilibrium value, the potential energy of the system increases, and a restoring force acts upon the system. The restoring force is proportional to the displacement, leading to a vibration that is sinusoidal or can be described as simple harmonic motion [1]. The vibrational frequency of the molecule can be solved using the appropriate differential equations to yield:  ν vib =  1 2π  K  µ  (1)  Where K is the force constant for the bond and µ is the reduced mass of the system. For masses of m1 and m2, the reduced mass is given by:  3  µ=  m1m2 m1 + m2  (2)  Stronger bonds will have a higher force constant, resulting in an increase in the vibrational frequency. On the other hand, bonds with a greater reduced mass will yield vibrations at lower frequencies. In that sense, Raman spectroscopy is inherently capable of identifying chemical moieties with differing bond strengths (Such as alkane/alkene/alkyne bonds). Similarly Raman spectroscopy can also differentiate moieties with differing reduced masses (such as the alcohol O-H bond vs. the amine N-H bond).  Since the first report of inelastic light scattering by C.V. Raman in 1928 [4], Raman spectroscopy has developed into a mature scientific technique capable of providing characteristic structural information about molecules. Because of its robust nature – its ability to sample a large variety of molecules under different experimental conditions - RS is now routinely used in a multitude of research and industrial laboratory environments. Key technological developments during the past two decades, particularly in the area of laser sources and array detectors, has helped to facilitate this expansion [5, 6].  When monochromatic light of frequency υo interacts with a molecule, the electron cloud of the molecule is modulated at the incident frequency υo and an oscillating dipole moment is formed as a result of the electronic charge displacement [3]. The molecule thus exists in a ‘virtual state’ lying lower in energy than the first excited electronic state of the molecule. This is an important distinction from fluorescence spectroscopy, in which the molecule is excited to a  4  higher electronic state from which the fluorescence photon is emitted [3]. In Raman spectroscopy, the excitation to the non-emitting virtual state and subsequent de-excitation occur coincidentally, in a much faster time frame (ca. 10-14 s) compared to fluorescence spectroscopy. There are three distinct outcomes that can occur involving scattering of the electromagnetic radiation (Figure 1.1).  In the case of Rayleigh scattering, photons are elastically scattered and retain the same energy as the incoming photon. Inelastic scattering of the incident radiation can also occur. Molecular vibrations at frequency νvib induce an oscillation of the molecular electronic cloud at the same frequency. The incident laser light at frequency ν0 induces its own electron cloud oscillation at frequency ν0, with both oscillations coupling at frequencies ν 0 ± ν vib .  5  Figure 1.1: Three possible light scattering processes from a molecule-photon interaction. The photon can be Rayleigh scattered, in which the outgoing photon has the same energy as the incident photon. The system can scatter a photon with less energy [h(ν0 – νvib)] than the excitation source in Stokes scattering. Alternatively, a photon of greater energy [h(ν0 + νvib)] can be scattered in the anti-Stokes process. Figure reproduced from [7] with permission.  When the emitted photons occur at a lower frequency (ν0 – νvib) than the incident light, the process is referred to as Stokes scattering. On the other hand, photons having a higher frequency (ν0 + νvib) than the incident light result from the anti-Stokes scattering process.  6  The amount of energy that will be gained/lost as part of the Raman scattering process is quantized and related to the vibrational levels in the ground electronic state. In Stokes scattering, shown in Figure 1.2(a), the molecule transitions from a lower energy vibrational state (ν0) to the virtual state. Subsequent relaxation of the molecule to a higher energy vibrational state (ν1) results in the Raman scattered photon lower in energy by the difference between the vibrational ground (ν0) and excited vibrational state (ν1) energies.  In the anti-Stokes process, the molecule transitions from a vibrationally excited state (ν1) to the virtual state, with subsequent emission of the anti-Stokes photon with the molecule moving to the ground vibrational state (Figure 1.2(c)). Generally, the intensity of anti-Stokes emission is much reduced compared to the Stokes scattering process [5, 8, 9]. For anti-Stokes emission to occur, the molecule must be in a vibrationally excited state when the incident photon is scattered. According to the Boltzmann distribution, the ground state will be the most highly populated state at room temperature, resulting in the Stokes signal being far greater in intensity than the anti-Stokes process [1, 5, 9].  7  Figure 1.2: Schematic representation of the (a) Stokes, (b) Rayleigh and (c) anti-Stokes scattering processes. Figure reproduced from [7] with permission.  As a result, a Raman spectrum provides direct measurement of the vibrational energies of the molecule, with the Stokes and anti-Stokes Raman vibrations symmetrically distributed about the Rayleigh line (Figure 1.3). A Raman spectrum consists of a plot of intensity vs. energy (wavenumber, cm-1), with Raman-active molecular vibrations (Chapter 1.2.2) yielding a corresponding peak in the Raman spectrum. Instead of plotting as a function of absolute energy spectra are typically presented in terms of ‘Raman shift’, which describes the magnitude of the shift from the Rayleigh scattered line at hν0. The Rayleigh scattered line is not shifted and occurs at 0 cm-1. Plotting Raman spectra as a function of Raman shift yields spectra which are independent of the excitation wavelength used – the magnitude of the Raman shift is dependent  8  upon the molecular vibration and is independent of the excitation frequency, allowing for direct comparison of Raman spectra collected with different excitation frequencies. Spectra plotted as a function of the absolute energy yields plots with differing x-axes, making direct comparison difficult. A sample Raman spectrum of chloroform is shown in Figure 1.3, with different frequency and wavelength axes for comparison:  9  Figure 1.3: Raman spectrum of Chloroform obtained with 514.5 nm excitation. The Rayleigh scattered line occurs at 514.5 nm and corresponds to a 0 cm-1 Raman shift. Peaks to the right of the Rayleigh line correspond to Stokes scattering, and the scattered photons have less energy than the incident photons. Note that the Stokes scattering peaks are consistently higher in intensity than the corresponding anti-Stokes scattering peaks. Peaks to the left have higher energy than the incident photon and occur by way of anti-Stokes scattering. The absolute frequency and wavelength axis is also included for reference. Figure reproduced from [5] with permission.  10  1.2.1  Classical Derivation of the Raman Effect The electric field component (E) of an electromagnetic waves varies with time t  according to:  E ( t ) = E0 cos ( 2πν 0 t )  (3)  Where E0 is the maximum amplitude of the electromagnetic wave and ν0 is the frequency of the wave. When a molecule is irradiated by the electric field a dipole is induced within the molecule. The magnitude of the dipole moment (P) is proportional to the electric field strength:  P = αE  (4)  Where α is the molecular polarizability (vector and tensor notation have been omitted from the polarizability for simplicity). Substituting equation (3) into (4) yields the timedependent dipole moment of the molecule:  P = α E0 cos ( 2πν 0 t )  (5)  For a molecule vibrating with frequency νvib, the nuclear displacement associated with the vibration can be described by:  q(t ) = q0 cos ( 2πν vib t )  (6)  11  The maximum displacement from the equilibrium bond length is denoted by q0. The polarizability of electrons within the molecule (α) will be perturbed by the molecular vibrations, and the relationship is described by:   ∂α   q  ∂q 0  α ( q ) = α0 +   (7)  In the above equation the polarizability of the molecule at the equilibrium position is denoted by α0, the rate of change of polarizability with respect to the nuclear displacement at the  ∂α   , and q represents the displacement from the equilibrium position is denoted by   ∂q  0  equilibrium bond position.  Combining Equations (5), (6) and (7) yields:  P = α E0 cos ( 2πν 0 t )  ∂α  = α 0 E0 cos ( 2πν 0 t ) +   qE0 cos ( 2πν 0 t )  ∂q 0  ∂α  = α 0 E0 cos ( 2πν 0 t ) +   q0 E0 cos ( 2πν 0 t ) cos ( 2πν vib t )  ∂q 0  1   ∂α  = α 0 E0 cos ( 2πν 0 t ) +     q0 E0 cos {2π (ν 0 − ν vib )} + cos {2π (ν 0 + ν vib )}  2   ∂q 0 (8)  12  From Equation (8) it follows that there are three possible radiative processes occurring. The first term describes the Rayleigh scattering process in which an oscillating dipole radiates light at frequency ν0, matching the excitation frequency employed. The second term radiates at the frequency difference (ν 0 − ν vib ) in the Stokes scattering process. The third term radiates at the sum of the frequencies (ν 0 + ν vib ) , corresponding to anti-Stokes scattering. In the context of obtaining vibrational information, Rayleigh scattering is inherently useless and provides no such information. Experimentally, careful attempts are made to avoid observing the Rayleigh line while performing Raman experiments, as it can cause detector overloads due to the high intensity of the Rayleigh scattered light compared to the Raman scattered photons.  The Intensity of a Stokes Raman vibration, IRaman, is given by:  I Raman = k ⋅ (ν 0 − ν vib ) ⋅ I ⋅ α 2 4  (9)  Where k is comprised of various natural constants, ν0 is the excitation frequency and I is the intensity of the incident radiation. The final term represents the molecular polarizability [6, 8]. Of note is the fact that Raman intensity increases by the fourth power as the excitation frequency increases.  13  1.2.2  Raman Selection Rules Infrared (IR) spectroscopy and Raman spectroscopy are similar processes in that they  probe the vibrational states of molecules, and that information is interpreted to provide important structural insights into the molecule. On the other hand, IR and Raman spectroscopy are governed by markedly different selection rules [9].   ∂α   component Vibrations which are Raman active must have a non-zero value for the   ∂q  0  of Equation (8). In other words, the polarizability of the molecule must change with the molecular vibration in order to be Raman active. Conversely, IR-active vibrations must have a change in the dipole moment of the molecule during the molecular vibration. For a symmetric vibration about a functional group, the vibration is typically IR-inactive due to the fact that there is no change in the dipole moment of the molecule. On the other hand, such a vibration can induce a change in the molecular polarizability, resulting in the vibration being Raman active [6, 8]. Likewise, vibrations from functional groups with a centre of symmetry will be Raman active and IR inactive, or vice versa [6, 8].  IR spectroscopy typically presents strong signals for stretching vibrations involving the carbonyl (C=O) functional groups due to a change in dipole moment resulting from the molecular vibration (changing the length of the C=O bond during the vibration induces a change in the dipole moment about the moiety). On the other hand, the alkene (C=C) bond is symmetric with no net dipole moment. Accordingly, a symmetric alkene stretch will have no change in the net dipole moment and will be very weak in the IR spectrum [1, 6, 8].  14  In Raman spectroscopy, species with distributed (diffuse) electron clouds, such as alkene (C=C) moieties, are easily modulated in the presence of an external electric field [6]. Therefore, a stretching vibration induces a change in the electronic density about the bond with a resultant change in the polarizability. Highly polar species, such as the carbonyl (C=O) double bond, do not exhibit strong signals in the Raman spectrum because it is more difficult to distort the asymmetrically-distributed electron cloud [5]. These important differences are demonstrated in the IR and Raman spectrum for oleic acid methyl ester (Figure 1.4):  15  Figure 1.4: FTIR (upper) and Raman spectrum (lower) of oleic acid methyl ester. In the IR spectrum, the carbonyl (C=O) stretch is prominent while the alkene (C=C) stretch is very weak. Conversely, the Raman spectrum yields a strong signal for the alkene stretch and a weak signal for the carbonyl vibration. Figure reproduced from [5] with permission.  1.3  Raman Signal Enhancement While RS is an incredibly powerful technique, it is inherently limited by the weak Raman  scattering process. For excitation frequencies far from any electronic absorption bands, the Raman cross sections are exceedingly small, approximately 1 in 107 photons are Raman scattered, leading to Raman spectra with low signal-to-noise ratios (SNR) [5, 9]. Raman  16  experiments are often limited to cases where high laser power and long acquisition times are possible in order to maximize the observed Raman signal.  An alternative method to dramatically increase the Raman signal is to employ surfaceenhanced Raman scattering (SERS) [10]. However, SERS is limited by the fact that the analyte molecule must reside close to a metal surface in order to be near the highly enhanced electromagnetic fields produced from surface plasmon excitation, which is the heart of the SERS effect [11]. In doing so, serious concerns arise that introduction of the non-physiological element may induce structural changes within the molecule, thereby affecting the interpretation of resultant structural data. Various methods have been employed to generate the SERS substrate, such as self-assembly of noble metal nanoparticles [12, 13], or by nanomanufacturing and nanolithorgraphy [14-17]. The former, while affordable and simple to create, often have limited reproducibility which may limit their use in quantitative situations [18]. On other hand, nanolithography is typically labour intensive which increases the cost of the substrate.  Another method to increase the overall Raman signal is to employ Resonance Raman Spectroscopy (RR / RRS), in which the excitation wavelength is near an electronic transition of the analyte molecule (Figure 1.5(a)) [19-21]. This is in direct contract to conventional RS, where the excitation frequency is not in resonance with a real excited electronic state of the analyte molecule [3, 22].  The pathway for RR and fluorescence spectroscopy differ in that RRS involves scattering of the excitation photon to yield a Raman photon at lower energy (longer wavelength). On the  17  other hand fluorescence spectroscopy involves absorption of the excitation photon resulting in the analyte molecule existing in an electronically excited state. Subsequent vibrational relaxation (internal conversion) and emission of the fluorescence photon occurs at lower energy (longer wavelength) than the incident photon occurs (Figure 1.5(b)). Because of these key differences the Resonance Raman process occurs on a much faster time scale (ca. 10-14 s) compared to the fluorescence process (ca. 10-8 s) [22]  18  Figure 1.5: Comparison of (a) resonance Raman and (b) Fluorescence pathways. In both cases the molecule moves to the excited electronic state by the excitation photon (υex). However, in the resonance Raman process the molecule emits the Raman scattered photon (υs) from an excited vibrational level, whereas in Fluorescence spectroscopy the molecule first undergoes internal conversion before emitting the fluorescence photon (υfl). Figure reproduced from [23] with permission.  19  The intensity of Raman bands was previously given in Equation (9), but the nature of the Raman polarizability tensor (α) and its effect on the Raman signal was not discussed. The Raman polarizability term has a frequency dependence which is governed by [9]:  α=  M me M en  1  M me M en +   ∑ h e  ν em − ν 0 + i ∆ e ν en + ν 0 + i ∆ e   (10)  Where the equation describes the transition from the mth vibrational level to the excited electronic state (e), followed by emission to the nth vibrational level during the Raman scattering process (Figure 1.5). The νem and νen terms represent the frequencies for the m→e and e→n transitions,  e  is the vibronic bandwidth in the excited state and h is Planck’s constant. M is the  transition moment integral resulting from a molecular transition from the mth to the nth vibrational level by way of the polarizability operator [9]:  M me = ∫ Ψ *m µσ Ψ e dτ  (11)  Where Ψ terms represent the total wavefunction of the ground and excited states while µσ is a component of the electric dipole moment. For a vibration to be Raman-active this integral must be non-zero.  20  Under nonresonance conditions the excitation frequency (ν0) does not coincide with the frequency of the transition to an excited electronic state (νem). As a result, the Raman polarizability term does not appreciably vary with frequency and instead Raman intensity changes are dominated by the (ν 0 − ν vib ) term of Equation (9). 4  As the excitation frequency moves into resonance with the frequency of the electronic transition, ν0 = νem and the denominator of the first term in Equation (10) approaches zero. In doing so, the Raman polarizability dramatically increases and intensity changes are now dominated by the polarizability component of Equation (9). Accordingly, resonance enhancements on the order of 104- 106 can be observed [3, 24].  Under resonance conditions the intensity of the Raman vibrations will be greatly enhanced, but the effect is limited to chromophores whose ground state wavefunctions have significant Frank-Condon overlap with the excited state wavefunction [3, 24]. As a general rule, vibrations which distort the ground-state geometry towards that of the excited-state geometry will be enhanced to the greatest extent [3]. From a practical (experimental) perspective, the change in Raman intensity with differing excitation wavelength can be monitored for each Raman-active vibration and a resonance Raman excitation profile (RREP) can be constructed to determine the optimum excitation wavelength for each vibration.  The ability to ‘tune’ and selectively enhance only certain vibrations is particularly advantageous for complex samples [3]. A major hurdle with the use of vibrational spectroscopy  21  for the elucidation of biological structure is the complexity of the resultant spectra [19]. A molecule with N atoms will contain 3N-6 normal modes of vibration (3N-5 normal modes for a linear molecule). Biological samples containing hundreds or thousands of atoms will yield many overlapping vibrational frequencies, complicating the Raman spectrum. This spectral selectivity significantly improves the ability to interpret Raman spectra by eliminating or minimizing overlapping and interfering bands which are not the subject of study (Figure 1.6).  22  Figure 1.6: Hypothetical absorption curve and Raman spectra for a molecule having two chromophores in the UV region, denoted by A and B. The intensity of the Raman vibrations is dependent upon the excitation wavelength used, demonstrating the selectivity possible in the resonance Raman process.  Fluorescence emission often occurs in conjunction with Raman spectroscopy using visible excitation, hindering observation of the Raman scattered photons [3, 5]. One common method to minimize fluorescence is to excite in the near IR (NIR) region, where the excitation energy is too low to excite the molecule into the higher electronic state necessary for the fluorescence pathway (Figure 1.5(b) and Figure 1.7) [5].  23  Figure 1.7: Excitation of a fluorescent sample with 514.5 and 785 nm excitation. With 514.5 nm excitation, the fluorescence signal is several orders of magnitude higher than the Raman signal. With 785 nm excitation, the excitation photons are not energetic enough to excite the molecule to the higher electronic state necessary for fluorescence. Therefore, the Raman spectrum can be observed. Figure reproduced from [5] with permission.  Alternatively, excitation in the deep UV region (sub-260 nm) can also avoid fluorescence complications for very different reasons. While fluorescence is not eliminated in such cases, the Stokes shift of the fluorescence is typically larger than that for the Raman scattered photons, ensuring that the two processes do not overlap spectrally [2, 3, 25]. In addition, the ν4 increase in  24  the Raman signal with decreasing wavelength makes UV excitation a more suitable choice than NIR excitation [26]. That being said, the use of UV excitation can result in further complications such as photodegradation and inner filtering effects, necessitating additional methods to overcome these issues. This will be discussed in the subsequent sections.  In RR applications the use of UV excitation is desirable due to the fact that many biological molecules possess an electronic transition in the ultraviolet region [3]. Therefore, UV excitation in RRS (UVRRS) holds great promise for the elucidation of key structural information. While UVRRS has historically been limited by a dearth of cost effective and robust excitation sources in the UV region [5], the recent development of commercially-available continuous wave (cw) frequency-doubled lasers in the UV region has made the technique available to both academic and commercial laboratories [5, 27].  UVRR is particularly suited for the study of protein structure and function [28-36]. Resonance enhancement of the peptide backbone π→π* vibrations occurs when excitation wavelengths below 220 nm are used. The amide backbone vibrations are particularly diagnostic of protein secondary structure, as the frequencies and intensities of the vibrations are dependent upon changes in the peptide backbone conformation [2]. In particular, previous studies by Chi [30] and Huang [37] obtained UVRR spectra of proteins in predominately alpha helix, beta pleated sheet, and random conformations. Both reports demonstrated that each protein conformation had distinctive Raman vibrations that are diagnostic of the protein conformation. Their results showed that UVRR is complementary, and in some cases superior to, UV-Vis, circular-dichroism (CD) and IR measurements which are commonly used to determine protein  25  secondary structure [30]. Furthermore, protein dynamics have been studied using UVRRS [33], and can provide further insight into protein folding events.  UVRR is also suited to the study of intramolecular and intermolecular DNA structure due to the intense electronic absorptions that occur in the UV region [38]. Previous studies by Shaw investigated DNA structure following exposure to ionizing radiation sources used in radiation therapy for treating cancer [39]. Using calf-thymus DNA (ct-DNA) and short-stranded DNA as model compounds for human DNA, the authors probed changes in DNA structure using difference spectroscopy. In particular, ct-DNA showed radiation-induced changes suggesting disruption of the stacked base configuration between DNA bases on the same strand. Additional changes in the Watson-Crick hydrogen bonds between base pairs on opposite strands were also observed. Another study examining DNA structure changes following exposure to thermal heat cycling was also performed [40], and noted disruption in the hydrogen bonding secondary structure of DNA.  Another interesting application involves the direct observation of oxidative lesions in DNA. 8-oxo-deoxyguanosine (8-oxo-dG) is an oxidative lesion believed to be responsible for mutations and cancer development and has only been indirectly observed previously [41]. Detection of 8-oxo-dG was made using UVRR because 8-oxo-dG has distinct vibrational frequencies compared to deoxyguanosine (dG) [41].  Extending the fundamental studies of proteins and DNA, UVRR has then been applied to the study of bacterial cell growth [42]. Using excitation at 244 nm, changes in Raman signals  26  associated with aromatic amino acids and nucleic acids were used to track metabolic changes associated with bacterial growth. Similarly, UVRR has been used to classify lactic acid bacteria found in yogurt [43], identification of viral proteins [44], and also for the in situ analysis of antimalarial agents in cinchona bark [45].  1.4  Fiber-Optic Probes for Raman Spectroscopy While RRS is a sensitive technique for probing molecular structure, the fact that the  excitation frequency coincides with an electronic transition of the analyte molecule can lead to photoinduced degradation of the sample following long exposure times or high laser power. Experimentally, one method to increase the signal-to-noise ratio (SNR) is to tightly focus the excitation beam to increase the power density at the sample, allowing for efficient collection of the Raman scattered photons from the excitation spot using an objective lens. However, an increase in the power density can in turn cause increased photodegradation of the sample, leading to a Raman spectrum that is contaminated with peaks from the photodegradation product. To some extent, other methods such as cooling, spinning or flowing of the sample can be employed to minimize sample degradation by ensuring a fresh aliquot of sample is continuously exposed to the excitation source as the experiment proceeds. However, those techniques may not always be possible or may require large quantities of sample which may not be available.  Additionally, the requirement that the excitation frequency fall within an absorption band of the sample in RRS can lead to additional complications. Because of the relatively small Stokes shift of the Raman scattered photons these photons also typically fall within the same  27  analyte electronic transition band, leading to ‘inner filter’ or sample self absorption effects [46]. In such cases, the Raman scattered photons are absorbed by the sample, preventing subsequent detection [46-48]. In fluorescence spectroscopy the inner filter effect is avoided by lowering the analyte concentration through dilution [22]. In other words, a delicate balancing act of concentration is required in which too high of a concentration leads to sample self absorption, while too low of an analyte concentration also lowers the signal intensity (where the signal intensity is proportional to the analyte concentration).  Another viable method to minimize sample self-absorption is to minimize the path length that the Raman scattered photons must travel in order to be detected. In conventional (focussing) Raman applications the typical path length may extend millimetres or centimetres, leading to high sample self-absorption and reduced signal intensity. An alternative method is to employ fibre optic (FO) probes for the delivery/collection of excitation/Raman scattered photons [47, 48].  The development of FO probes for Raman spectroscopy was initially focussed on nonresonance applications in the visible and NIR region of the electromagnetic spectrum [49, 50]. In such cases, optical transmission is very efficient and the excitation and Raman scattered photons can safely travel fiber distances of 10-100 m without major signal loss [5]. This is particularly advantageous in industrial applications where the FO probe can be used to analyze hazardous and/or dangerous samples while the instrument and operator can be housed in a safer environment [5]. FO probes are easy to use, and the operator can simply dip the probe into a sample without concerns about focus and optical alignment changing from sample-to-sample [5].  28  A multitude of different FO probe geometries have been previously employed, including 8around-1 and 18-around-1 geometries [5].  The development of appropriate FO probes for UVRRS was limited by the quality of available FO materials. In the early applications of UVRRS, performed with pulsed excitation sources, laser-induced breakdown of the excitation fiber was a major concern [47, 48]. However, two developments allowed for the use of FO probes in UVRRS [24, 47, 51-53]: The development of suitable FO materials which could handle higher power densities, and the development of continuous-wave frequency-doubled laser systems in the deep UV region. With the improved FO material probes can be used in the 200-250 nm region without probe degradation [47]. This region is rich in biological electronic transitions that can be exploited by way of the resonance Raman process [54].  When designing FO probes for UVRR applications two important considerations must be employed: The effective area of overlap between the excitation and collection fibers and the distance travelled in the medium by the Raman excitation and scattered photons (path length). Under non-absorbing conditions, the area of overlap is of the utmost importance while the latter is of little concern. In the flush fiber probe geometry (Figure 1.8(a)) the fibers are cleaved and polished perpendicular to the long axis of the fibers, and the excitation and collection fibers are assembled so that their end faces are within the same plane [55, 56]. The excitation fiber effectively illuminates a semi-infinite volume extending away from the probe face. The overlap area between the excitation and collection fibers is extremely large but is far away from the face of the collection fibers.  29  Mirrored Surface  A) Flush Geometry  B) Right Angle Geometry  Figure 1.8: Overlap areas for (a) flush and (b) angled fiber-optic probes. The flush probes are suitable for non-absorbing analytes where path length considerations are not paramount. The angled probes are ideally suited for absorbing analytes as the overlap area is maximized and moved closer to the faces of the fibers. This minimizes the path length for the Raman scattered photons. Figure reproduced from [26] with permission.  Under resonance conditions the illuminated volume will be severely decreased due to sample self-absorption of both the excitation and scattered photons. Indeed, the illumination area may only extend microns from the excitation fiber face under highly absorbing conditions [26]. Therefore, the path length of the Raman scattered photons becomes a critical consideration and the goal is to minimize the path length of the Raman scattered photons by moving the  30  overlap area as close as possible to the fiber faces. Right-angled FO probes effectively accomplish these requirements in a suitable compact probe design, as shown in Figure 1.8(b).  The development of the angled collection fiber geometry represents an important achievement in FO probe design and is an advantageous departure from the more traditional flush fiber probe geometry [47, 48]. The Raman scattered photons pass through the side of the collection fiber and is reflected into the guided modes of the fiber using the mirrored surface at the bevelled tip of the fiber (Figure 1.9) [48].  Previous research was performed to examine the various parameters associated with angled FO probes for use in UVRRS [24, 26, 47, 48, 57]. In general, angled probes used in FOUVRRS have a number of important features, which are outlined below:  1. A 6-around-1 geometry is employed, in which a single excitation fiber is surrounded by a ring of 6 collection fibers. The excitation and collection fibers have core diameters of 600 µm and 400 µm, respectively. 2. The excitation fiber is cleaved perpendicular to the long axis of the fiber, while the collection fibers are cut and polished at a 45° angle. The collection fibers have a mirrored aluminum surface deposited onto the angled end of the fiber. 3. The collections fibers are placed such that the mirrored surface faces away from the illuminated volume of the sample, and are offset by approximately 100-200 µm from the end of excitation fiber.  31  Figure 1.9: (a) Side view of an angled probe geometry used in typical FO-UVRRS applications. Raman excitation of the analyte molecule takes place very close the excitation fiber surface. The Raman scattered photons then pass through the side of the collection fiber, reflect off of the mirrored surface and into a guided mode of the collection fiber for transport to the detection system. (b) Top view of the FO probe showing the excitation fiber surrounded by six collection fibers. Figure reproduced from [46] with permission.  While there is room for up to 8 collection fibers to be placed around the excitation fiber, in a more traditional 8-around-1 geometry [5], two positions are left vacant to facilitate mass transport through the probe volume. When used in conjunction with sample spinning, this  32  ensures that fresh sample is continuously replenished in the volume of interrogation, allowing for longer total acquisition times with photosensitive samples.  Notwithstanding other complications, the use of a larger illumination area can spread out the excitation beam, effectively reducing the power density delivered to the sample. As the illumination area is increased the intensity of the excitation source can be concomitantly increased, allowing for higher laser powers while maintaining the power density at an acceptably safe level [46, 58]. Therefore FO probes are capable of delivering a greater amount of laser power to the sample while maintaining the same power density [46], in a manner similar to the concept of line focussing instead of spot focussing used in more conventional forms of Raman spectroscopy [59, 60]. Equation (9) demonstrates that the Raman signal intensity is directly proportional to the intensity of the excitation source used. To maximize the signal-to-noise ratio of the resultant Raman spectrum, the maximum laser power that can be employed without sample photodegradation is typically employed [46]. . With FO probes, the illuminated area is defined by the core diameter of the excitation fiber, and the light is assumed to exit the fiber as an approximately radially uniform cylinder over the short distance of the overlap area [61]. While the collection efficiency of the FO probe will be reduced with a larger excitation fiber, this loss is typically made up for by the allowable increase in the overall laser power because the incident light exits the fiber as a uniform cylinder and not as a focussed spot [46]. Indeed, previous studies compared a FO probe using a 600 µm excitation fiber to a 25 µm spot focus used in UVRR microscopy [46]. The results demonstrated that the FO probe allows for a 576 times increase in power while maintaining the same power  33  density, when compared to spot focussing. While the collection efficiency in the larger excitation fiber is reduced by a factor of 3.25, the overall advantage of using the FO probe is still a two orders of magnitude increase in total signal collection compared to conventional spot focussing (for equivalent risk of photodamage) [46].  Therefore, the development of FO probes for use in UVRRS represents a major evolution in the technique itself. Using FO probes allows for the analysis of photosensitive systems which would not otherwise be experimentally accessible.  1.5  Methods for Biological Investigations FO-UVRRS has a number of advantages which makes it amenable to the study of  biological molecules. Many biological molecules possess a strong electronic absorption in the UV region which can be utilized for RR spectroscopy [3, 38]. Large biological molecules, in particular proteins and nucleic acids, can have complex vibrational spectra which can be difficult to interpret [19, 38]. However, with RR spectroscopy the ability to selectively enhance only vibrations associated with the electronic transition can minimize this problem. More importantly, the use of UV excitation sources can also eliminate sample fluorescence which often occurs in visible-excitation RR spectroscopy [3]. Indeed, FO-UVRRS has been employed to study a wide array of biological systems such as enzymes [62-64], neurotransmitters [65] and oligonucleotides [66, 67] which would otherwise be difficult to access experimentally. In addition, water is an inherently weak Raman scatterer [6, 9], allowing for the analysis of  34  biological molecules under physiological conditions without heavy spectral interference from water [3].  However, FO-UVRRS is certainly not the only available technique for providing insight into biomolecular structure. Instead, FO-UVRRS represents an important complement to information provided by other techniques. In order to understand the contributions that FOUVRRS can make, it is important to briefly compare other techniques commonly employed, along with their advantages and disadvantages.  X-ray crystallography has long been considered the gold standard for characterization of biomolecular structure [68-72]. Crystal structures are obtained by exposing a crystalline sample to short wavelength, high energy X-ray sources and monitoring the resultant diffraction pattern. Careful interpretation of the results can yield the three-dimensional structure of a protein with Angstrom resolution. However, quality diffraction patterns and usable experimental results require high quality crystallized samples which may require months of experiments to produce [68-72]. Furthermore, crystallization may be difficult for air-sensitive or otherwise difficult samples, which may affect the resolution of the resultant crystal structure. More importantly, there is always a concern that extrapolation of the solid-state crystal structure to the solution state may not be accurate for all biomolecules.  For some fundamental studies in enzyme-substrate interactions, the observation of the substrate protonation state may be critical for identifying the underlying reaction mechanism [62, 64]. Elucidation of substrate protonation state is routinely based on differential bond lengths of  35  non-hydrogen containing bonds [62]. In the case of the HPCD/MndD enzyme set discussed in Chapter two, the protonation state of the substrate cannot be directly observed using X-ray crystallography [73, 74], meaning only indirect evidence is obtained. Furthermore, crystallization of proteins in enzyme-substrate studies often employ mutant (deactivated) enzymes or inhibitory substrates, in which the crystal structure may not be reflective of the native enzyme-substrate complex.  Fluorescence spectroscopy (FS) is another common technique used to monitor some levels of biomolecular structure, though it is not as informative as crystallography or nuclear magnetic resonance spectroscopy (NMR, see below). Some protein tertiary structure can be characterized using the natural fluorescence of tryptophan residues. With excitation between 280-300 nm and detection between 330-350 nm, the fluorescence signal yields excellent sensitivity to the local tryptophan environment. Modulation in fluorescence intensity and shift in the fluorescence wavelength are indicators of change in the local environment. As such, the technique is commonly used to monitor protein folding/unfolding kinetics. This method is not limited to tryptophan residues because other fluorescent reporter molecules (tags) can be covalently attached to the biological molecule and the resultant fluorescence can be monitored [75].  Alternatively, Fluorescence resonance energy transfer (FRET) [76] or luminescence resonance energy transfer (LRET) [77, 78] can be used to measure distances between a donor and acceptor fluorophore by way of an energy transfer process. This technique is commonly referred to as a “spectroscopic ruler” because of the excellent sensitivity in distance  36  measurements that it can provide, which is particularly valuable for protein folding and binding experiments [79].  Both fluorescence spectroscopy and FRET exhibit excellent sensitivity and selectivity with excellent time resolution, allowing for the study of dynamic processes. However, energy transfer experiments require labelling of the biological molecule with both donor and acceptor fluorophores. Fluorescence spectroscopy experiments can also use a non-native fluorophore for spectroscopic observation. As with all labelling techniques there is a concern that introduction of the labelling reagents may perturb the biological structure being measured. In addition, sites on the interior of a folded protein may not be accessible for labelling, limiting the structures that can be probed.  UV-Vis spectroscopy is routinely employed for analytical purposes due to the fact that chromophores within a protein are capable of absorbing different wavelength of light with different strengths of absorption [1, 22]. Indeed, UV-Vis spectroscopy is routinely used to quantify protein content based on the overall absorption at 280 nm [80, 81]. However, the absorption properties of the chromophores are relatively insensitive to changes in the overall protein structure, leaving the technique information-poor and unsuitable for biomolecular structural determination.  Circular dichroism (CD) monitors the differential absorption of left- and right-circularly polarized light through a sample [82-84]. This is particularly advantageous for protein samples as they contain active chromophores in the amide backbone, aromatic amino acids and disulfide  37  linkages. As a result, CD is one of the most common techniques for determining protein secondary structure. Much research has been performed in analyzing model proteins in different structural conformations - α-Helices, parallel and anti-parallel β-pleated sheets and random coil structures - for unique CD spectral signatures in each conformation [85-87]. CD spectra of unknown proteins are compared to these known model proteins to assign secondary structure. However, problems with CD can arise due to changes in the optical activity of aromatic amino acids or changes in the electronic environment of the protein [85]. Such perturbations may in turn induce changes in the electron distribution of the molecule, and hence the CD spectrum, without changes in the protein secondary structure. The spectral sensitivity of the technique is also limited because of the broad peaks which are diagnostic of the various structural elements within the biomolecule.  Nuclear magnetic resonance (NMR) spectroscopy has been extended to the study of large biological macromolecules due to its ability to provide high resolution structural information [88-93]. NMR is routinely used to investigate 1H, 13C and 15N local environments of biomolecules in aqueous (D2O) solution, which is highly advantageous for the study of proteins and nucleic acid structure. However, NMR spectroscopy is a ‘broad spectrum’ technique in the sense that it is responsive to all of the specific nuclei being probed. NMR spectra are highly complex and interpretation of the data requires high computational power or complex multidimensional experiments in order to simplify the spectra for interpretation. Furthermore, NMR experiments require high sample concentrations in order to yield usable NMR spectra, which may not be accessible for all systems. As a further downside, NMR instruments are large, expensive and employ large magnets in order to produce the high magnetic fields necessary for  38  the experiments. Long exposure times are employed in NMR experiments - typically on the order of hours - limiting the throughput of the technique.  Electron paramagnetic resonance (EPR) spectroscopy has been routinely employed in the study of enzymes, lipid membranes and other biological systems [94-96]. The technique is limited to systems with an unpaired electron such as radicals or paramagnetic transition metals. Given the highly reactive nature of most radical species, it can be difficult to maintain the radical species at a high enough concentration for detection. To facilitate detection of biological molecules that do not have an inherently unpaired electron, spin labels containing an unpaired electron can couple to the system for detection purposes. Spin labels are stable radical species which have a reactive functional group capable of coupling to specific molecules of interest (such as proteins or membranes) [97]. Again, the fact that the system must be labelled can perturb the underlying structure and chemical environment which is being measured.  Alternatively, extended X-ray absorption fine structure (EXAFS) can be used for structural elucidation of metalloproteins [98-104]. While the technique is incredibly powerful, the major drawback is the requirement for a tuneable x-ray source, only available at synchrotron light sources. Given the limited availability of synchrotron sources in general, and the resultant limitation in beam time for researchers to obtain experimental data, the technique has not yet seen widespread use for the study of enzyme structure.  A brief summary of the advantages and disadvantages for each of the techniques discussed here is shown in Table 1.1:  39  Table 1.1: Summary of common analytical methods for biomolecular structure determination. Technique FO-UVRRS  Advantages Highly Label-free.  Disadvantages  selective. Instruments Results built.  often  Few  Structure Sensitivity  custom 2° and local 3°  commercial  are rich in structural options available. Data must information.  be carefully interpreted to obtain structural information.  X-ray  Detailed and complete Requires  Crystallography structural information.  crystal  growth. 2° and 3°  Solid phase structure may be different than solution phase. Hydrogen position inferred.  Fluorescence  Highly sensitive and Often requires labelling of 2° (FRET) and local  Spectroscopy / selective.  molecule  FRET  Limited information.  under  study. 3° structural  40  Technique  Advantages  Disadvantages  Structure Sensitivity  UV-Vis  Universally available. Lack of spectral information. local 3°  Spectroscopy  Quick  to  operate. Limited to certain protein  Easy to perform.  Circular  Sensitive  probe  Dichroism  secondary structure.  residues.  of High  concentrations  analyte required. interference.  of 2° Water  Minimal  selectivity. NMR  Highly  selective  to Complex  certain atoms (e.g. 1H, Costly 13  C).  spectral  results. 2° and 3°  experimental  apparatus. Low throughput. High  concentration  and  purity of samples required.  41  Technique EPR  Advantages Highly  specific  Disadvantages  Structure Sensitivity  – Limited to samples with 2° and 3°  solvents and matrices unpaired electron (organic do not interfere.  radicals  and  transition  metals). Spin-labelling often required  for  observation. of  biological Concentration  radicals  must  be  maintained. EXAFS  Highly  selective  different Well  to Limited and very expensive n/a  elements. instrumentation. suited  determining  for metal  oxidation states.  1.6  Thesis Structure This thesis presents research results from the study of several distinct and biologically-  relevant systems. The common theme amongst all studies is that FO-UVRRS is used to obtain key structural information that is not necessarily available from other techniques.  Chapter two examines a suite of ring-cleaving dioxygenases with an interesting twist: The enzymes are capable of functioning with a native or non-native metal inserted into the  42  enzyme center, with no change in kinetic parameters. Two close enzyme homologues are examined, and it is demonstrated that the insertion of the non-physiological metal into the enzyme has no effect on the substrate structure.  Chapter three presents the results of a UVRR study of a carbon-carbon bond hydrolase (BphD). UVRR spectroscopy was used to characterize the substrate binding mode to a hydrolytically impaired mutant enzyme of BphD. The results demonstrate that the substrate binds in a strained enolate geometry, and that this strain facilitates ketonization in subsequent steps of the reaction.  Chapter four characterizes the structural relationships between synthetic DNA oligonucleotides and their locked nucleic acid (LNA) analogues. This represents the first reported UVRR study of the LNA oligomers and demonstrates that incorporation of a locked base induces a conformational change in the glycosyl bond between the backbone and the base.  Chapter five is an off-shoot of the FO-UVRRS study of locked nucleic acids. The results conclusively show that there is residual contamination in commercially-available synthetic DNA oligonucleotides that has been previously unreported. The fact that the contamination was only observed now, using FO-UVRRS has a great deal to do with the inherent selectivity that is only available with this technique. This result is important because benzamide contaminates not only the Raman spectra of these DNA oligonucleotides, but also the UV-Vis spectrum. This has important implications in the fundamental study of DNA oligonucleotides as the contaminant  43  could adversely affect spectroscopic and thermodynamic measurements involving synthetic oligonucleotides.  Finally, chapter six summarises the discoveries presented in this thesis, and draws together the common threads that exist between them.  44  Chapter Two: UV Resonance Raman Investigation of Metal-Containing Dioxygenases  2.1  Introduction Aromatic ring-cleaving dioxygenases represent an important class of enzymes that  facilitate the aerobic degradation of natural and man-made aromatic compounds [105, 106]. These reactions represent an important mechanism by which organic carbon can be recycled for further use [107], and is accomplished by way of several reaction steps leading to the formation of a 1,2-dihydroxybenzene (catechol) intermediate, with subsequent oxygen incorporation to yield the final degradation products [64].  Ring-cleaving dioxygenases cleave the catechol C-C ring bond by way of two different mechanisms and are classified based on their regiospecificity [105]. Intradiol dioxygenases involve a non-heme Fe(III) metal center to cleave the bond lying between the two hydroxyl groups of the catechol ring, while extradiol dioxygenases cleave the bond adjacent to the hydroxyl moieties by way of a non-heme Fe(II) metal center (Figure 2.1).  The mechanism for both enzymes is believed to proceed via a bridged iron-alkylperoxo intermediate in which the catechol moiety binds in a bidentate, asymmetrical fashion to the iron metal center. The asymmetry in the extradiol dioxygenase results from the monoanionic nature of the bound catechol, with one of the catechol OH groups being deprotonated. In the intradiol dioxygenase the asymmetry is attributed to a trans-donor effect from an equatorial tyrosinate ligand [64].  45  Figure 2.1: Ring-cleaving reaction of catechol catalyzed by extradiol and intradiol enzymes. Both enzymes proceed via an iron-alkylperoxo intermediate, but the carbon-carbon bond cleaved depends on the class of enzyme used. Figure reproduced from [64] with permission.  An examination of available crystal structures from the superfamily of extradiol dioxygenase enzymes indicates that the enzymes possess a common structural motif that binds the Fe(II) metal center within the enzyme active site [108]. This is in spite of the fact that these enzymes catalyze a host of distinct enzymatic reactions such as C-H hydroxylation, epoxidation of C-C double bonds, ring formation and C-C bond cleavage [105, 109]. This motif consists of two His residues and one Glu/Asp residue in what is commonly referred to as a “2-His-1carboxylate facial triad” (2H1C triad).  The two histidine and one carboxylate residues bind a mononuclear divalent metal ion at the active site of the enzyme, at the vertices of a triangular face of an octahedron [109]. The  46 2H1C triad serves to anchor the metal ion to the enzyme, leaving three binding sites representing one face of an octahedron available. This allows for the binding of O2, substrate and enzyme cofactors within the enzyme active site [105, 108, 109]. This is in direct contrast to hemecontaining oxygenases, which only have one available coordination site for O2, with the substrate bound at a separate site within the protein [108]. Substrate binding in the 2H1C active site primes the metal center for subsequent O2 binding [105, 110] and is believed to modulate the redox potential of the metal center to facilitate further reactivity.  Homoprotocatechuate 2,3-dioxygenase (HPCD, from Brevibacterium fuscum) catalyzes the oxidative cleavage of homoprotocatechol (HPCA) to yield a α-hydroxy-δ-carboxymethyl cismuconic semialdehyde product (Figure 2.2) [111]. HPCD belongs to the type I extradiol dioxygenase class, of which other notable members include catechol 2,3-dioxygenase, the dihydroxybiphenyl 1,2-dioxygenases (BphC and DHBD) along with homoprotocatechuate 2,3dioxygenase from Athrobacter globiformis (MndD). The later is particularly interesting because of its high sequence homology to HPCD, but preferentially utilizes a Mn(II) metal center instead of Fe(II) at the enzyme active site.  Figure 2.2: Reaction of homoprotocatechuate (HPCA) with HPCD or MndD to yield a muconic semialdehyde product.  47 Spectroscopic characterization of the HPCD:HPCA enzyme:substrate (E:S) complex indicates that the substrate binds in a highly asymmetrical manner to the Fe(II) metal center, suggesting only one of the catechol hydroxyl groups are ionized upon binding [62, 112, 113]. Substrate binding also releases three solvent molecules from metal coordination, thereby leaving a 5-coordinate metal center which is primed for oxygen binding to facilitate the remainder of the reaction [111, 112]. Proposals for the catalytic cycle of HPCD, shown in Figure 2.3, hypothesized that electron density is transferred from the substrate to the bound oxygen by way of the metal which coordinates the reactions [113, 114]. Subsequent oxygen binding would yield an oxidized metal-superoxide complex with a +3 metal oxidation state.  48  Figure 2.3: Initial proposals for the catalytic cycle of extradiol ring cleaving dioxygenases. The mechanism proposed transient change in the metal oxidation state from Fe(II) to Fe(III) over the course of the reaction, and subsequent return to the Fe(II) oxidation state at the conclusion of the catalytic cycle. Figure reproduced from [115] with permission.  Nucleophilic attack of the superoxide species occurs on the electron deficient substrate at the carbon bearing one of the catechol OH substituents, yielding the alkyl-peroxy intermediate. The reaction intermediate would then undergo heterolytic O-O bond cleavage, with one of the oxygen atoms being inserted into the catechol ring to form a lactone intermediate via the Criegee rearrangement. Subsequent hydrolysis of the lactone intermediate occurs using the remaining oxygen atom, resulting in the semimuconic aledhyde product.  49 The transient kinetics of the enzymatic cycle have been difficult to characterize because the iron is nearly colorless and electron paramagnetic resonance (EPR) silent, precluding two major forms of spectroscopic characterization [106]. Instead, characterization of the reactions often utilized an alternative substrate, 4-nitrocatechol (4NC). 4NC has chromophores that are highly environment and protonation state dependent, and yields large changes in the optical absorption spectrum throughout the reaction cycle. Kinetically, 4NC also reacts much slower than HPCA with HPCD, allowing some of the intermediates to be observed [106, 116]. However, the introduction of a non-physiological substrate for observation once again assumes no perturbations in the mechanism between HPCA and 4NC.  Interestingly, a homologue enzyme from Arthrobacter globiformis, MndD, catalyzes the same reaction. While HPCD has 83% sequence identity to MndD, the enzymes naturally utilize different metals (Fe and Mn, respectively) to orchestrate catalysis [112, 116]. However, the high degree of similarity between the HPCD and MndD enzymes represents an opportunity to test the role of the metal ion in these enzymatic reactions, as demonstrated by Emerson et al. [73]. The tertiary structure of both enzymes are very similar before and after substrate binding [112], and the enzymes both yield the same product when the equivalent substrate is employed [73].  Therefore, Emerson et al. proceeded to investigate the nature of these enzymatic reactions by swapping the native metals in HPCD and MndD. Insertion of the non-native metal (to form Mn-HPCD and Fe-MndD) did not cause substantial disruption in the active-site structure or in the kinetic parameters of the reaction, despite the metals central role in the formation of the enzyme active site structure [73].  50 This result is particularly unusual because of the dramatic difference in the reduction potentials of Fe(II) and Mn(II) (~0.77 V and ~1.5 V, respectively). The two metals differ by ca. 0.7 V in the absence of redox tuning by the protein structure [117]. The identical nature of the enzyme active-site structures in HPCD and MndD precludes differential tuning of the metal reduction potentials [112]. This suggests that no significant change in the metal oxidation state occurs over the course of the reaction, contrary to the mechanism shown in Figure 2.3. Therefore, the redox couple is formed by O2 reduction and HPCA oxidation. Because this is constant between the native and non-native enzymes, the catalytic parameters should remain constant for all enzymes studied.  On the basis of these results, a mechanism in which the metal oxidation state does not change was proposed [73], as shown in Figure 2.4:  51  Figure 2.4: Alternative mechanism proposed for the HPCD-catalyzed reaction of HPCA. Unlike the mechanism proposed in Figure 2.3, no formal change in the metal oxidation state occurs over the course of the reaction. Figure reproduced from [73] with permission. Copyright 2008 National Academy of Sciences, U.S.A.  As with all 2H1C-containing dioxygenases, substrate binding displaces three solvent molecules from the enzyme active site [108]. The substrate binds with one of the catechol  52 hydroxyl groups deprotonated, priming the active site for subsequent oxygen incorporation at the site trans to the carboxylate ligand. Electron density is then transferred to the bound oxygen by way of the metal center to form the reduced superoxide species, which then attacks the substrate to form an intermediate alkylperoxy species (bottom of Figure 2.4). The remaining acidic proton in the substrate is also removed in order to enhance the substrate susceptibility to oxidation in the next reaction step [107]. Subsequent steps involve a rearrangement with O-O bond cleavage, ring insertion to produce a lactone intermediate and subsequent hydrolysis to form the product of the reaction. Two of the predicted intermediates and the enzyme product complex were recently crystallized in HPCD, providing general support to this mechanism [118].  Therefore, the HPCD enzyme mechanism would differ from other members of the 2H1C triad family in one important manner: the electron required for O2 activation comes directly from the substrate and not from the metal center. In other words, the active site contains a reduced metal-superoxide complex [M(II)-O2• -] instead of the more conventional [M(III)-O2• -] structure which has been observed for other oxygen activating metalloenzymes [105]. This hypothesis is supported by the similar kinetic parameters observed for the native (Fe-) and non-native (Mn-) HPCD enzymes, along with the corresponding MndD counterparts (Mn- and Fe-MndD, respectively).  Knowledge of the substrate binding mode in the HPCD and MndD enzyme families would be beneficial toward understanding the mechanism by which these reactions occur and in principle, could allow for the indirect observation of any changes in the metal oxidation state during the early steps of the reaction. Furthermore, spectroscopic observation of the substrate intermediate in the native and non-native enzyme counterparts may be important for providing  53 support to this proposed mechanism and to monitor for changes in substrate structure brought upon by modulation of the metal center. Fiber-optic UV resonance Raman spectroscopy (FOUVRRS) is uniquely suited for the identification of such enzymatic intermediates [62-64] because it is capable of providing key spectroscopic information which can in turn confer structural insights into the molecules under study [6].  Raman spectroscopy (RS) typically suffers from low signal levels generated by way of the Raman scattering process, necessitating high excitation power or long acquisition times [6]. Such experimental conditions could, in turn, lead to photodamage of the system and has limited the applicability of the technique for studying biological phenomena [20, 26].  On the other hand, in Resonance Raman spectroscopy (RRS) the excitation wavelength for the Raman experiment is concomitant with an electronic transition for the molecule under study [19, 20]. This confers enhanced cross sections, often on the order of 104-106 over conventional RS [6]. In addition, molecular vibrations associated with the electronic resonance will be enhanced to a greater extent, reducing ‘spectral clutter’ from non-diagnostic vibrations [19, 20]. RRS can utilize multiple excitation wavelengths to study different molecular vibrations associated with the molecule. Furthermore, many biological molecules of interest possess an electronic absorption in the UV region, and the use of UV excitation in RRS experiments is highly suited [2, 3]. While fluorescence is a long-established issue with RS in the visible and NIR regions, it is avoided completely with UV excitation [25].  However, the fact that the UVRRS process takes place near a real electronic absorption band of the molecule can lead to sample photodegradation. One method to overcome this is to  54 employ fiber-optic (FO) probes for the delivery and collection of the Raman photons [47, 48]. It has been demonstrated that FO probes allow for greater power outputs to be delivered to the sample while maintaining the same power density level compared with “conventional” UVRRS systems [46]. This in turn can confer enhanced signal levels and/or reduced exposure times.  FO-UVRRS has previously been used to study extradiol [62] and intradiol [64] dioxygenases to conclusively identify the protonation state of the bound catechol substrate. Here, the HPCA substrate has four distinct ionization states which are dependent upon the solution pH. The pKa for the acetic acid moiety is around 4.2, while those for the catecholate hydroxyl moieties are 9.5 and 12 [119].  Figure 2.5: Four possible protonation states of HPCA. HPCA has pKa values of 4.2, 9.5 and 12.  In this work, FO-UVRR has been utilized to study the protonation state of the HPCA substrate bound to both native (Fe-HPCD, Mn-MndD) and non-native (Mn-HPCD and CoHPCD, Fe-MndD) enzyme systems. Knowledge of the protonation state of the intermediate is then used to provide further insight into the proposed mechanisms for the HPCD and MndD family of enzymes.  55  2.2  2.2.1  Materials and Methods  Enzyme Preparations Fe-HPCD, Mn-MndD, Mn-HPCD and Fe-MndD were expressed and purified as reported  by Emerson et al. [73]. Co-HPCD was expressed and purified as reported by Gunderson et al. [115]. With the exception of Fe-MndD, all growths and purifications were carried out under aerobic conditions. Fe-MndD was grown anaerobically in a 5 L fermenter (Bioflow 2000 Fermentor, New Brunswick Scientific) which was thoroughly sparged with N2 and the entire growth was carried out under a low flow of N2 gas (less than 0.5 L/min) to exclude O2. Fe-MndD purifications were carried out under anaerobic conditions using N2-sparged buffer to minimize oxidation of the Fe center. Purified enzyme preparations were equilibrated into 50 mM MOPS buffer (pH 7.5) and stored at -80 °C.  2.2.2  Model Compound Synthesis Fe-HPCA and Mn-HPCA model compounds were synthesized using the analogous  procedure describing the synthesis of K3[Fe(catechol)3] [120]. Briefly, 3 mL N2-purged H2O was added to 18 mmol FeCl3•6H2O or MnCl3•4H2O in an anaerobic chamber, and the mixture was brought to pH 7 with 4M NH4OH until a precipitate was observed. The precipitate was collected, rinsed three times with water, and transferred to a 15 mL beaker equipped with a stir bar. 96.5 mg HPCA, resuspended in 1.5 mL H2O, was added to the precipitate. 4 mL of 6 M KOH was gradually added and the dark brown solution was left to stir anaerobically for 3 h. The reaction mixture was then lyophilized to obtain a dark tarry substance, which was rinsed with propanol and dried under vacuum before analysis.  56  2.2.3  Preparation of Samples Aqueous solutions were prepared using water purified on a Barnsted NANOpure UV  apparatus to a resistivity greater than 17 MΩ•cm. Samples for UV-Vis and UVRR spectroscopy were prepared under an inert atmosphere in an Mbraun Labmaster glovebox (Stratham, NH) maintained at less than 4 ppm O2. Buffers and solvents were vigorously bubbled with argon for 20 min, brought into the glovebox, and allowed to equilibrate for 24 h prior to use. HPCA was weighed and purged with N2 before transferring to the glovebox for sample preparation. Purified HPCD (containing Fe, Mn or Co) or MndD (containing Fe or Mn) preparations were transferred into the glovebox, thawed immediately prior to use and diluted to the appropriate concentration in 50 mM MOPS pH 7.5 buffer. Deuterated samples were prepared by buffer exchanging the thawed enzyme into deuterated 50 mM MOPS, pD 7.5. Samples for UVRR spectroscopy were prepared in plastic 0.6 mL microtubes (Axygen Inc., Union City, CA) to a total volume of 300 µL containing 175 µM enzyme (based on target metal content per monomer), and 150 µM HPCA for E:S samples. 100mM Na2SO4 was also included as an internal standard in UVRR samples. High-purity argon was layered on top of the sample before flash-freezing in liquid nitrogen upon removal from the glovebox.  2.2.4  UV-Vis Absorption Spectroscopy Anaerobic preparations of free HPCA at various pH’s were made in a 1 mL gas-tight  cuvette (Hellma, Concord, ON). Spectra were recorded anaerobically using a Varian Cary 5000 spectrophotometer equipped with a thermostatted cuvette holder (Varian Canada, Mississauga,  57 Ontario, Canada) maintained by 25 °C, controlled by Cary WinUV software version 3.00. Contributions from the solvent have been removed from all spectra.  2.2.5  UV Resonance Raman Spectroscopy Resonance Raman experiments were performed on a custom built instrument that has  been previously described [63, 64, 66, 67]. All experiments were performed at 248 nm using a frequency-doubled Ar+ ion laser (Innova 90C, Coherent Inc., Santa Clara, CA). Excitation light was delivered to samples by way of custom-built fiber-optic probes which have been described previously [46-48]. The laser power delivered was measured to be ca. 20 mW. Sample tubes were spun continuously with a custom-built spinner to ensure a fresh aliquot of sample was analyzed over the course of the experiment. A flow of Argon blanketed the samples to ensure that no aerobic enzyme turnover occurred over the course of the experiment.  The Raman scattered photons were collected by the FO probes and passed through a dielectric interference filter (Barr Associates Inc., Westford, MA) to remove any residual Rayleigh scattered light. The Raman spectra were then dispersed using a 1.0 m focal length monochromator (Model 2061, McPherson Inc., Chelmsford, MA) with a 3600 grove/mm holographic grating (Model 8358-1004-0, McPherson Inc). The Raman light was detected using a liquid nitrogen cooled CCD (Spec-10 400B, Roper Scientific, Trenton, NJ) operating at -120 °C. Consecutive acquisitions of 10 s were obtained for each sample. Sequential spectra were used only if no significant changes were observed in the Raman bands. Exposure studies of the enzyme to the Raman excitation source indicated that the enzyme retained 70% activity after 30 s of exposure (data not shown). Therefore, spectra resulting from enzyme-containing samples  58 were limited to a total exposure of 30 s (3 acquisitions). All spectra were normalized to the 980 cm-1 band of the SO42- internal standard prior to obtaining the difference spectrum.  UVRR data analysis was performed using MATLAB 7.0 (The MathWorks, Natick, MA). Peak-fitting was performed using the Interactive Peak Fitter function for MATLAB (Created by Tom O’Haver). Peaks were constrained to a Gaussian peak shape and were fit iteratively until a minimum in the residuals was obtained. Other parameters - peak width, location, and amplitude were left unconstrained during the fitting process. This fitting process was repeated multiple times, and the peak fitting values quoted here represent the average of a minimum of 10 iterative fitting processes. While the uncertainty in the peak fitting process will vary with spectral parameters such as peak resolution and intensity, it is estimated that uncertainties are approximately 2-3 cm-1 for each peak, consistent with previous work [64].  2.3  2.3.1  Results and Discussion  Spectroscopic Analysis of Free HPCA Substrates To determine the protonation state of the enzyme-bound substrate it is first necessary to  obtain the key spectral signatures of the free substrate in varying protonation states. As a primary characterization method the UV-Vis spectrum of HPCA was obtained at different solution pH. Following that, the Raman spectra of the free substrates at different protonation states was also obtained to serve as a baseline measure of the Raman response of the substrate. Furthermore, the samples were obtained in both H2O and D2O solvent in order to observe isotopic shifts resulting from the H/D exchange of exchangeable protons in the substrate. Such  59 information will be key to interpreting the difference spectrum obtained from the enzyme-bound substrate.  2.3.1.1 UV-Vis Spectroscopy HPCA exists in four distinct protonation states, and these states can be distinguished to some extent based on their electronic absorption spectra. Both neutral and monoanionic HPCA exhibit electronic absorptions at 204, 225 and 281 nm (Figure 2.6 and Table 2.1) with nearly equivalent extinction coefficients. Therefore, it follows that deprotonation of the ethanoic acid moiety has little to no effect on the UV-Vis spectrum. This is contary to previous work involving protocatechol (PC), in which a blue shift of approximately 5-10 nm was observed for both the La and Lb bands when moving from the neutral to the monoanionic form (data not shown; nomenclature is based on Platt [121] and is used to describe the total momentum of each electronic state and the degeneracy of that state). This difference is attributed to the fact that PC has resonance structures available in which the negative charge of the carboxylate group can be delocalized into the benzene ring. No such avenues exist in HPCA due to the sp3 hybridized methylene carbon between the carboxylic acid and catechol moieties.  60  Figure 2.6: Anaerobic solvent-subtracted UV-Vis spectra of HPCA in the four possible protonation states. (a) Neutral HPCA, 50mM MOPS pH 3 (b) Monoanionic HPCA, 50mM MOPS pH 7 (c) Dianionic HPCA, 50mM MOPS pH 11 (d) Trianionic HPCA in 250 mM sodium t-butoxide/t-butanol solution ~ pH 14.  On the other hand, deprotonation of one of the catechol -OH groups of HPCA to form dianionic HPCA induces a red-shift in all of the electronic transitions to 212, 241 and 295 nm. In turn, the molar absorptivity of all three transitions increases relative to those for the neutral or monoanionic species. Previous experiments with catechol noted similar trends [26, 62].  61  Table 2.1: Electronic absorption bands of HPCA in various protonation states. Data obtained under anaerobic conditions. Values in parentheses represent extinction coefficients in units of M-1 cm-1. Assignments are based on the notation described by Platt [121]. Assignment  HPCA  HPCA-  HPCA2-  HPCA3-  (nm, M-1cm-1)  (nm, M-1cm-1)  (nm, M-1cm-1)  (nm, M-1cm-1)  Ba,b  204 (17314)  204 (16429)  212 (17386)  227  La  225 (4462)  225 (4653)  241 (5662)  262  Lb  281 (2333)  281 (2288)  295 (3884)  315  Subsequent deprotonation to trianionic HPCA induces a further red-shift in the electronic absorption bands to 227, 262 and 315 nm. Extinction coefficients are not provided for the trianionic species due to the low solubility of HPCA in the t-butoxide/t-butanol solution, causing high background values and high uncertainties associated with these measurements [26].  In the dianionic and trianionic forms the red shift of the electronic transitions moves the La band into resonance with the excitation wavelength used for Raman experiments (248 nm). In turn, ring mode vibrations associated with the electronic transition will become resonantly enhanced [2, 3, 19]. Such ring mode vibrations were previously used to monitor the protonation state of enzyme-bound catechol [62, 64], and stands to be important for monitoring the protonation state of HPCA.  Dianionic and trianionic HPCA are highly sensitive to O2 exposure, resulting in the formation of polymerized quinones [122]. The polymerized degradation product is highly  62 absorbing in the visible spectrum, and takes on a characteristic burgundy colour which is readily observable [122]. To avoid polymerization and maintain substrate integrity, experiments were performed under anaerobic conditions.  2.3.1.2 Raman Spectroscopy Careful consideration of the excitation wavelength is necessary in order to ensure enhancement of the HPCA vibrational bands while minimizing enhancement of enzyme vibrational bands. At shorter wavelengths in the UV region (229 and 238 nm), Raman vibrations from aromatic amino acids become dominant due to their higher Raman cross-section [26, 123]. Excitation at 257 and 264 nm, which may be most favourable to observe the trianionic HPCA species, is overwhelmed by fluorescence from the phenylalanine and tyrosine residues [123]. Therefore, 248 nm excitation represents a compromise between minimizing observed enzyme Raman vibrations and maximizing substrate enhancement.  The Raman spectrum of HPCA was obtained at four different solution pH’s representing the four possible protonation states (Figure 2.7). Each protonation state exhibits unique spectral markers which can make identification of the enzyme-bound substrate protonation state possible. Furthermore, the spectrum was acquired in both H2O and D2O solvents in order to determine isotopic shifts in the Raman vibrations following hydrogen-deuterium exchange. The neutral, monoanionic and dianionic species posses exchangeable protons, and shifts in associated Raman vibrations should be observed. Conversely, trianioinic HPCA does not posses any exchangeable protons and no shifts should occur.  63  64  Figure 2.7: FO-UVRRS spectra of HPCA in each of the four possible protonation states. Left panel: HPCA in H2O. (a) 3.0 mM HPCA, phosphate buffer pH 3.0, 100 mM Na2SO4; (b) 1.0 mM HPCA, phosphate buffer pH 7.0, 100 mM Na2SO4; (c) 0.3 mM HPCA, phosphate buffer pH 11.0, 100 mM Na2SO4 and (d) 0.5 mM HPCA, phosphate buffer pH 14.0, 100 mM Na2SO4. Right panel: HPCA in D2O. (a) 3.0 mM HPCA, phosphate buffer pD 3.0, 100 mM Na2SO4; (b) 1.0 mM HPCA, phosphate buffer pD 7.0, 100 mM Na2SO4; (c) 0.3 mM HPCA, phosphate buffer pD 11.0, 100 mM Na2SO4 and (d) 0.5 mM HPCA, phosphate buffer pD 14.0, 100 mM Na2SO4. Spectra acquired at 248 nm, 20 mW power and 40 s total acquisition time.  The Raman spectra of both the neutral and monoanioinic species are very similar (Figure 2.7(a) and (b), Table 2.2). Both spectra are dominated by a band near 1600 cm-1 which is attributed to catecholate ring vibration modes [124-126]. Neutral HPCA exhibits a strong vibration at 1293 cm-1 which is upshifted to 1297 cm-1 in monoanionic HPCA. This vibration is attributed to a νCO (7a) vibrational mode [126]. A medium intensity peak at 1155 cm-1 in the neutral species shifts to 1152 cm-1 and increases in intensity in the monoanionic species and is assigned to a δOH vibration [126]. In addition, a weak vibration at 1451 cm-1 in neutral HPCA upshifts to 1457 cm-1 and greatly increases in intensity for monoanionic HPCA. This is attributed to a 19b vibrational mode involving νCC and νCO vibrations within the catechol ring. This vibration may represent a potential spectral marker to differentiate between the neutral and monoanionic species.  65  Table 2.2: Band positions and intensities for the four possible ionization states of HPCA. Shifts of vibrations upon deuterium substitution are also noted. Band intensities and shape denoted by: (s) strong, (m) medium, (w) weak, (vw) very weak. HPCA (H)  H:D shift  HPCA(H)  H:D HPCA2shift (H)  H:D shift  HPCA3(H)  H:D shift  Assignment  1624(s)  -7  1620(s)  -7  1608(s)  -1  1606(s)  0  νCC, 8a  1611(s)  -9  1606(s)  -10  1587(m) -4  1575(s)  0  νCC + νCO, 8b  1567(m) -6  1548(s)  0  νCC  1495(w)  1486(w)  0  νCC, 19b  +1  νCC, 19a  1451(vw)  -8  1457(m)  +1  1293(s)  -1  1297(s)  -3  1296(m) -1  1291(w)  0  νCO, 7a  1260(w)  +2  1262(w)  +2  1277(m) +2  1259(w)  0  νCO, 7b  1155(s)  0  1162(m)  +9  1165(w)  1149(m)  +1  1152(s)  +1  1151(m) +2  1156(w)  1122(w)  +10  1120(vw) +10  1116(w)  1118(w)  δOH, B2  -5  +8  δCH, 9a 0  δCH, 9b  As HPCA moves to the dianionic and trianionic forms (Figure 2.7(c) and (d), Table 2.2) the La electronic transition moves into resonance with the excitation wavelength. As a result, catechol ring modes associated with that electronic transition become resonantly enhanced compared to other vibrational modes. Therefore, catechol modes in the 1500 – 1700 cm-1 region become dominant in the Raman spectrum. In the dianionic species (Figure 2.7(c)), the spectrum is dominated by a cluster of peaks in the catechol ring mode region, with the peak at higher energy being most intense. On the other hand, the trianionic species (Figure 2.7(d)) exhibits overlapping vibrations of approximately equal intensity.  66  A closer examination of the catechol ring mode region is necessary to assign the observed vibrations. Accordingly, a Gaussian peak-fitting analysis of the 1500-1700 cm-1 region was performed for all substrate protonation states. The neutral species of HPCA in H2O exhibits two major vibrations of equal intensity in the 1550-1700 cm-1 vibrational region (Figure 2.8(a)). The 1624 cm-1 and 1611 cm-1 vibrations are assigned to 8a and 8b vibrational modes, respectively [124, 126]. The 8a mode is solely a νCC vibration, while the 8b mode includes both νCC and  νCO vibrational components [124, 126]. Both modes are sensitive to deuterium substitution, where downshifts of -7 and -9 cm-1 occur, respectively (Figure 2.8(b)).  Previous UVRR results obtained for catechol [26, 62, 64] assign only a single catechol vibration in this region. However, the results shown here demonstrate that the individual 8a and 8b modes are obscured by the high degree of overlap between them. Furthermore, a closer examination shows that the peak is asymmetrical in nature, strongly suggesting that two overlapping peaks are occurring.  67  Figure 2.8: Gaussian peak-fitting analysis for different protonation states of HPCA. Left panel: Neutral HPCA (pH 3) in (a) H2O and (b) D2O. Right panel: Monoanionic HPCA (pH 7) in (c) H2O and (d) D2O. The red lines represent individual peaks obtained from peak fitting while the black line is the raw data.  Two vibrations are also observed in the same region for monoanionic HPCA. The two strong intensity vibrations are observed at 1620 and 1606 cm-1 for the respective 8a and 8b modes (Figure 2.8(c)), and the vibrations exhibit a downshift of 7 and 10 cm-1 in D2O (Figure 2.8(d)).  68  Monoanionic HPCA exhibits similar ring-mode vibrations compared to the neutral species. This is unsurprising given that the pendant carboxylic acid moiety, removed from the catechol ring, is deprotonated in monoanionic HPCA. This deprotonation would have only minimal effect on the electronic and vibrational properties of the catechol ring. This is reflected in the ring-mode vibrations undergoing an absolute downshift of 4-5 cm-1 between neutral and monoanionic species. The absolute magnitude of downshift following H/D exchange is consistent for both the neutral and monoanionic species.  69  Figure 2.9: Gaussian peak-fitting analysis for different protonation states of HPCA. Left panel: Dianionic HPCA (pH 11) in (a) H2O and (b) D2O. Right panel: Trianionic HPCA (pH 14) in (c) H2O and (d) D2O. The red lines represent individual peaks obtained from peak fitting while the black line is the raw data.  For dianionic HPCA, catechol 8b and 8a modes are observed at 1587 and 1608 cm-1 (Figure 2.9(a)). A new band at 1567 cm-1 evolves following deprotonation of one of the catechol –OH moieties. This band is assigned to a new νCC mode, which was also observed previously  70 for catechol [64] and protocatechol (data now shown). H/D exchange moves the 8b and 8a modes down -4 and -1 cm-1 respectively, while the new νCC mode downshifts by 6 cm-1 (Figure 2.9(b)). A weak vibration at 1495 cm-1 evolves, and is assigned to a 19b mode [64]. The mode is only weakly sensitive to H/D substitution. This mode moves downwards by 9 cm-1 in the trianionic species, consistent with previous observations [64].  Trianionic HPCA contains three vibrations of approximately equal intensity at 1548, 1575 and 1606 cm-1 (Figure 2.9(c)). The trianionic species of HPCA contains no exchangeable protons, and no shifts in any of the vibrational bands was observed between H2O and D2O (Figure 2.9(d)).  The 8a vibrational mode sees a small (2 cm-1) downshift from 1608 to 1606 cm-1 between the dianionic and trianionic HPCA species. The two remaining vibrations at 1548 and 1575 cm-1 undergo downshifts of 19 and 12 cm-1, respectively. Complete deprotonation of both -OH moieties, and subsequent charge delocalization throughout the benzene ring, results in large downshifts in ring-mode vibrations [62].  UVRR spectroscopy allows for the spectroscopic discrimination of the four possible protonation states of HPCA. Each protonation state exhibits unique spectral characteristics which facilitate identification. Therefore, it stands to reason that identification of the protonation state of the enzyme-bound substrate is possible using this technique.  The monoanionic species exhibits two key differences compared to the neutral species. In the first case, the monoanionic species exhibits an enhanced intensity δOH peak at 1152 cm-1  71 compared to the neutral species. In addition, an increase in intensity of the 19a vibrational band near 1455 cm-1 is also observed. While difficult, it may be possible to discriminate between the neutral and monoanionic species based on changes in these two vibrations. All other spectral features remain constant between these two species, making identification between these two species difficult.  Dianionic HPCA sees the evolution of two unique vibrations that could be used for identification purposes. The new vibrational band at 1567 cm-1 is not found in the neutral and monoanionic species, and is downshifted by 20 cm-1 in the trianionic species. Furthermore, the development of a weak band at 1495 cm-1 may also be used for identification purposes, but the low overall intensity of the peak makes it difficult to observe in the difference spectrum.  Finally, trianionic HPCA yields three strong ring-mode vibrations in the 1500-1600 cm-1 region, which is dramatically different than the other species. Furthermore, the trianionic species is unaffected by H/D substitution due to the lack of acidic protons. Based on the above, the unique spectral markers in HPCA stand to be used in UVRR spectroscopy to identify the protonation state of the enzyme-bound substrate.  2.3.2  Spectroscopic Study of Fe- and Mn-Bound HPCA Model Compounds To model the behaviour of HPCA bound to the metal centers in HPCD and MndD, a  Fe(III) tris (homoprotocatecholate) complex (1) was synthesized using published methods [120]. The model compound results in the formation of an Fe(III)-HPCA complex in which the metal center is in the +3 oxidation state and the HPCA ligands are bound in the trianionic form [120].  72  While the metals centers in both HPCD and MndD are expected to have a +2 oxidation state [73], the model compound may be important because it provides qualitative information about changes in the vibrational modes of HPCA upon binding to a Fe or Mn metal center. This information may be critical to understanding substrate vibrational modes observed in the Raman spectrum of the enzyme-bound substrate. In addition to (1), an analogous Mn(III)-HPCA model compound (2) was synthesized using the same methodology.  Figure 2.10: Fe- and Mn-HPCA model compounds. The metal centers are in the +3 oxidation state and the substrate is bound in the trianionic form.  2.3.2.1 UV-Vis Spectroscopy The UV-Vis spectrum of Fe-HPCA and Mn-HPCA complexes show a strong ligand π →  π* absorption near 280-290 nm (Figure 2.11(a) and (b)), along with a much weaker transition in the 330-350 nm region. This result is consistent with previous observations by Salama [127] and  73 Karpishin [120], who observed a strong absorption at 294 nm along with a weak shoulder absorption near 333 nm [120, 127]. Furthermore, previous studies by Horsman et al. [64] for the Fe(III)-catechol complex demonstrated a 18 nm blue shift in the lower energy band, from 308 to 290 nm. Assuming the 290 nm band observed here corresponds to the 315 nm Lb band in trianionic HPCA, this represents a 25 nm blue shift, comparable to these previous reports [64].  Figure 2.11: UV-Vis Spectra of the model compounds (a) 100 µM Fe-HPCA in 50 mM MOPS pH 7.5 and (b) 100 µM Mn-HPCA in 50 mM MOPS pH 7.5.  2.3.2.2 Raman Spectroscopy Further characterization of the model compounds was accomplished using UVRRS (Figure 2.12). The Raman spectra of the model compounds provide only a few observable  74 features, all of which are attributed to catechol ring modes. In all spectra, an unresolved cluster of vibrational bands near 1600 cm-1 was observed, along with weaker features near 1400 and 1500 cm-1 and a broad, unresolved feature in the 1200 cm-1 region. For both Fe-HPCA and MnHPCA, no vibrational shifts beyond experimental uncertainty were observed between samples in H2O and D2O. This is consistent with HPCA bound to the Fe and Mn metal centers in the trianionic form with no exchangeable protons.  75  Figure 2.12: FO-UVRR spectra of metal-HPCA model compounds. Left Panel: Fe-HPCA model compound in (a) H2O and (b) D2O with 100 mM SO42- internal standard. Right Panel: Mn-HPCA model compound in (c) H2O and (d) D2O with 100 mM SO42- internal standard. Spectra acquired at 248 nm, 20 mW power and 40 s total acquisition time.  For (1), the Raman spectrum is dominated by three vibrations of approximately equal intensity in the 1500-1750 cm-1 region at 1572, 1591 and 1609 cm-1 (Figure 2.13, Table 2.3). The 1591 and 1609 cm-1 bands are attributed to catecholate 8b and 8a modes respectively [127],  76 while the 1572 cm-1 represents a νCC mode [127]. These results are generally consistent with previous observations of the [Fe(cat)3]3- complex by Que [128], who noted weak vibrations at 1567 and 1591 cm-1 using 647.1 nm excitation.  Previous observations by Horsman [64] noted only a single vibration at 1568 cm-1 for the [Fe(cat)3]3- complex. The difference in the number of vibrational modes observed between the Fe-cat and Fe-HPCA complex is attributed to symmetry differences between the two substrates [64]. The low or non-existent intensity of the 8b vibration in the Fe-cat complex was ascribed to the fact that it is antisymmetric with respect to the catecholate C2 axis and would not be subject to resonance enhancement via the Franck-Condon mechanism [64]. The HPCA substrate has much lower overall symmetry than catechol resulting from the introduction of the pendant ethanoic acid moiety. As a result, this mode would become visible in the UVRR spectrum.  77  Figure 2.13: Gaussian peak-fitting analysis of metal-HPCA model complexes shown in Figure 2.12. Left panel: Fe-HPCA complex in (a) H2O and (b) D2O. Right panel: MnHPCA complex in (c) H2O and (d) D2O, respectively. The red lines represent individual peaks obtained from peak fitting while the black line is the raw data.  Similarly, the 1609 cm-1 was not observed by Salama [127] or Que [129] for the Fe-Cat complex. One possible explanation is the different excitation wavelengths employed in the studies: Experiments by Salama and Que were performed at 496.5 and 647.1 nm, respectively, which would excite into the ligand-to-metal charge transfer (LMCT) manifold of the complex  78 [129]. Experiments performed in the UV region would resonantly enhance aromatic ligand modes [120, 127]. Assuming the 1609 cm-1 band corresponds to a 8a vibrational mode, this ring mode would see little enhancement using visible excitation when compared to UV excitation. The use of the resonance enhancement effect, coupled with the decrease in the molecular symmetry between catechol and HPCA (see above), could result in this vibrational mode being observed.  On the other hand, the presence of this vibration cannot be attributed to residual, uncomplexed HPCA. At neutral pH, residual HPCA would exist in the monoanionic form which has unique spectral markers at 1606 and 1620 cm-1 (Table 2.2 and Figure 2.7). These spectral markers are inconsistent with those observed here. More importantly, residual HPCA in the monoanionic form would be sensitive to H/D exchange and result in a movement of the band to lower wavenumbers in D2O. Again, this is inconsistent with the data obtained for the complex.  Based on comparisons to the trianionic HPCA species, complexation to a metal center clearly induces an upshift in the catecholate vibrations by +24, +16 and +3 cm-1 for the νCC, 8b and 8a vibrations, respectively (Table 2.3). The large upshift upon complex formation is consistent with previous observations [35, 64]. The large shifts in vibrational modes between free and Fe-bound HPCA may arise for several reasons. The 8b mode has a C-O component which would be directly affected by complex formation with the Fe3+ metal center [64]. The negative alkoxide moieties would directly interact with the metal, distorting the electronic distribution and the vibrational modes involving these moieties [130, 131]. Conversely, the νCC mode observed at 1572 cm-1 may undergo a downshift due to C-C bond length changes upon complex formation [64]. On the other hand, the 8a vibration mode only sees a 3 cm-1 upshift  79 relative to the free trianionic species. This is also consistent with the vibration being primarily a ring-breathing mode, which would be minimally affected compared to other vibrational modes.  Table 2.3: Band positions and intensities for the Fe-HPCA and Mn-HPCA complex. HPCA is bound to the metal in the trianionic protonation state. Shifts of vibrations upon deuterium substitution are also noted, along with the shift induced between free (trianionic) and complexed HPCA. Band intensities denoted by: (s) strong, (m) medium, (w) weak. Fe-HPCA (H)  H:D shift  Fe-HPCA HPCA3-  Mn-HPCA (H)  H:D shift  Mn-HPCA HPCA3-  Assignment  1609(m)  0  +3  1609(s)  -1  +3  νCC, 8a  1591(m)  0  +16  1588(s)  -1  +13  νCC + νCO, 8b  1572(s)  0  +24  1566(m)  0  +18  νCC  1490(m)  0  +4  1493(m)  0  +7  νCC, 19b  1290(m)  0  -7  1290(m)  0  -7  νCO, 7a  1266(m)  0  +4  1267(m)  0  +5  νCO, 7b  1144(w)  0  -12  1143(w)  0  -13  δCH, 9a  Additional features in (1) are observed in the 1200 – 1300 cm-1 region, including a medium intensity feature at 1266 cm-1 and a weaker feature at 1290 cm-1. Both are attributed primarily to catecholate νCO (7a and 7b) stretches [127]. Vibrations at 1262 and 1320 cm-1 were observed previously using visible excitation [132], and were characteristic of Fe3+-catecholate coordination [133]. The 1320 cm-1 feature was not observed here or in previous UVRR studies  80 of the Fe-catechol complex, suggesting this is a LMCT feature which will not be resonantly enhanced under the experimental conditions employed [132].  There are conflicting reports in the literature for the assignment of the vibrations in the 1200-1300 cm-1 region. Salama [127] attributes these to phenolate C-O stretching modes, while Elgren [132] notes that these modes are predominately ring modes with only minor contributions from the C-O bonds, due to the fact that these peaks were relatively insensitive to 18O labelling of the catecholate oxygen atoms [133, 134]. Interestingly, the absolute location of these bonds did shift between free and complexed HPCA, suggesting some contribution from the νCO modes. Therefore the assignment was tentatively made to be from νCO modes, but this may require further investigation.  Two other weaker vibrations at 1422 and 1490 cm-1, assigned to 19a and 19b modes respectively [127, 135], are also observed in the model compound. These bands are upshifted by 22 and 5 cm-1 compared to trianionic HPCA. These bands have been observed previously [129, 133] and are characteristic of Fe3+-catecholate coordination. In particular, the 19b mode is localized between the carbon atoms to which the two catecholate oxygen atoms are attached [127]. This mode can see substantial change in bond length upon complex formation, by approximately 0.04 Angstroms according to Horsman [64], which in turn would perturb the vibrational frequencies.  The vibration spectrum of (2) presents similar vibrational bands compared to (1). For the Mn-HPCA model compound, vibrations were observed at 1566, 1588 and 1609 cm-1. These vibrations have been upshifted by 18, 13 and 3 cm-1 compared to trianionic HPCA. Small  81 differences in locations of the vibrational bands are observed between the model compounds, suggesting some sensitivity to the nature of the metal center. For example, the 1572 and 1591 cm-1 bands in Fe-HPCA are downshifted by 6 and 3 cm-1 respectively in Mn-HPCA. On the other hand, the 1609 cm-1 band shows no dependence on the metal center. The small shift in the first two bands may suggest some νCO character to both vibrations, which would be susceptible to changes in the metal center to which it is bound [64].  The intensity of the ring-mode vibrations differs substantially between Fe- and Mn-bound HPCA. In particular, there is a dramatic change in intensity for the 1609 cm-1 vibration between the two compounds. The exact cause for this is unknown, but may be a result of post-acquisition processing performed on the spectra. Samples of both model compounds were highly absorbing at the excitation wavelength used, leading to a high a degree of sample self-absorption and low overall signal intensities. As a result, Raman vibrations from water dominated the spectrum, with low intensity vibrations from the compound superimposed on the broad water vibration near 1640 cm-1. Therefore, the Raman spectrum of the ‘background’ (water) was subtracted from each spectrum shown here. Small deviations in the fitting parameters for the background spectrum could lead to apparent changes in the intensities of the model compound spectra [136, 137].  In summary, Fe-HPCA and Mn-HPCA model compounds were synthesized and the spectroscopic properties of these compounds were examined. The UV-Vis spectra of the model compound is consistent with previous literature reports [64, 120, 127], confirming that the metalHPCA complexes form with trianionic HPCA. Furthermore, UVRRS analysis of the model compounds demonstrated an upshift in the 8b and νCC vibrational modes upon complexation,  82 also in agreement with previous reports [64] for the Fe-catechol complexes. The 8a vibrational mode was weakly observed in the resultant Raman spectra and was relatively insensitive to metal complexation. Confirmation of the substrate protonation state was based on the fact that no isotopic shifts were observed between H2O and D2O, suggesting that no exchangeable protons are on the substrate. While these results may not be directly transferrable to the enzymesubstrate complex – the complex exists in the +3 oxidation state while it is expected the enzyme will have a +2 oxidation state – these results provide insights which may be extrapolated to the enzyme-substrate complex.  2.3.3  Spectroscopic Study of Native and Non-Native HPCD and MndD Enzymes The HPCD and MndD family of enzymes represents an interesting opportunity to  monitor the reactivity of the enzyme system following incorporation of the non-native metal. For HPCD, samples containing the native (Fe2+) and non-native (Mn2+, Co2+) metals were obtained. MndD was investigated using the native (Mn2+) and non-native (Fe2+) metal center. The substrate protonation state was probed using both UV-Vis and UVRR spectroscopy in order to elucidate the early steps of the enzyme mechanism. Any structural effects arising from the incorporation of the non-physiological metal should be observable, since UVRR can discriminate between each of the substrate protonation states.  2.3.3.1 UV-Vis Spectroscopy UV-Vis spectroscopic observation of an enzyme-bound substrate is routinely used to provide indirect structural information with respect to the substrate [138]. To obtain the UV-Vis  83 spectrum of the bound substrate, spectra of the free enzyme (E) and an enzyme:substrate (E:S) mixture was obtained. While the spectra will be dominated by absorptions due to the peptide bonds and aromatic amino acid residues [138], careful subtraction yields the spectrum of the enzyme-bound substrate. The difference spectra for Fe-HPCD, Mn-HPCD and Co-HPCD are shown in Figure 2.14:  Figure 2.14: UV-Vis difference spectra of the HPCA substrate bound in the anaerobic enzyme-substrate complex with (a) Fe-HPCD (b) Mn-HPCD (c) Co-HPCD. [Enzyme] = 30 µM, [HPCA] = 20 µM, 50mM MOPS pH 7.5.  In general all UV-Vis difference spectra exhibited two absorption bands in the 200 – 400 nm region. The strongest band occurs between 235-245 nm, with a weaker absorption occurring  84 near 295 nm. Absorptions below 230 nm are not shown in the difference spectrum because of high background absorptions resulting from both enzyme and solvent [26]. Likewise, uncertainties in the location of local maxima in the UV-Vis spectra are estimated to be ca. 3-5 nm due to the low intensity and broad nature of the bands.  UV-Vis studies of the MndD class of enzyme yielded comparable results (Figure 2.15). The spectra of the enzyme-bound substrate yields a strong transition near 240 nm, with a secondary transition observed around 295 nm.  85  Figure 2.15: UV-Vis difference spectra of the HPCA substrate bound in the anaerobic enzyme-substrate complex with (a) Mn-MndD and (b) Fe-MndD. [Enzyme] = 30 µM, [HPCA] = 20 µM, 50 mM MOPS pH 7.5.  Direct comparison of the UV-Vis spectra for the enzyme-bound substrate and that for the free substrates can provide some insight into the substrate protonation state. For the free substrate, absorptions at 204, 225 and 281 nm were observed for the monoanionic substrate, while the dianionic species yielded transitions at 212, 241 and 295 nm, and the trianionic species had transitions at 227, 262 and 315 nm (Figure 2.6, Table 2.1). Previous observations of the enzyme-substrate complex in Fe-catechol species demonstrated small (ca. 3-5 nm) blue shifts in the low energy band upon formation of the enzyme-substrate complex [26]. That result compares favourably with the results presented here: The observed shift in the previous work  86 would likely fall within the experimental uncertainties estimated for both experiments (see above).  The model compounds provide a direct measure of the substrate bound in the trianionic protonation state to a Fe3+/Mn3+ metal center but does not model additional enzyme-substrate interactions in the enzyme active site. The model compounds generally exhibited a strong absorption in the 280-290 with a weak shoulder near 330-350 nm (Figure 2.11). The significant difference between the spectra of the model compound and the enzyme-bound substrate should eliminate the possibility of trianionic bonding in the enzyme-bound substrate.  The UV-Vis results correspond closely to the spectrum for dianionic HPCA (Figure 2.6(c)). Interestingly, there is no blueshift in the enzyme-bound spectrum, unlike that observed in the Fe-HPCA model compound. This could be attributed to competing effects which cancel out any perturbations in the UV-Vis spectrum. While binding to the metal center may induce a blue shift in the substrate electronic structure, further interactions between residues in the secondary shell may cause further perturbations [26]. In particular, HPCD has a Tyr257 residue which interacts by way of hydrogen bonding interactions with the O3 of the HPCA substrate [73, 139]. Furthermore, the asymmetric binding of the HPCA –OH groups to the enzyme metal center may also effect the electronic structure. This mode of binding would not be observed in the model compound.  While the UV-Vis results indicate dianionic binding of the substrate to the enzyme, the results are by no means definitive. Therefore, further investigations into the enzyme-substrate structure were performed using UV resonance Raman spectroscopy.  87  2.3.3.2 Raman Spectroscopy Raman spectroscopy of enzyme-containing samples are dominated by signals arising from aromatic amino acid residues which are resonant with the excitation wavelength employed (248 nm) [58, 140-145]. Spectral contributions from the substrate are not readily apparent, and are at least one order of magnitude smaller than the enzyme vibrations [26]. Careful subtraction of the parent spectra generates a difference spectrum that is attributed to the enzyme-bound substrate [62-64]. To do so, the Raman spectrum of the enzyme (E) was subtracted from the Raman spectrum of the corresponding enzyme-substrate (E:S) complex. When obtained under identical experimental conditions, spectral subtraction using the internal standard as a reference yields the difference (E:S–E) spectrum. The parent spectra for the HPCD enzyme set (containing Fe, Mn or Co metal centers) are shown in Figure 2.16. The parent spectra for the MndD enzyme set (containing Mn and Fe metal centers) are shown in Figure 2.17:  88  Figure 2.16: FO-UVRRS of enzyme samples with (black trace) and without (blue trace) substrate, along with the corresponding difference spectrum. Left Panel: Fe-HPCD in (a) H2O and (b) D2O. Center Panel: Mn-HPCD in (c) H2O and (d) D2O. Right Panel: CoHPCD in (e) H2O and (f) D2O. Spectra acquired at 248 nm, 20 mW power and 40 s total acquisition time. [Enzyme] = 175 µM, and if applicable, [Substrate] = 150 µM.  Prominent features in the E:S and E spectra for both HPCD and MndD are attributed to the aromatic amino acid residues phenylalanine (F), tryptophan (W) and tyrosine (Y) [123, 142, 145]. A brief examination of the vibrations follows. In the 950–1100 cm-1 region of the Raman spectra, the 981 cm-1 band is due to the sulfate internal standard used for spectral normalization. The peak at 1010 cm-1 is due to a combination of phenylalaine F1 [140, 142] and tryptophan W16 modes [146]. The vibration at 1040 cm-1 is assigned to a phenylalanine F18a mode [140, 142].  89  Figure 2.17: FO-UVRRS of enzyme samples with (black trace) and without (blue trace) substrate, along with the corresponding difference spectrum. Left Panel: Mn-MndD in (a) H2O and (b) D2O. Right Panel: Fe-MndD in (a) H2O and (b) D2O. Spectra acquired at 248 nm, 20 mW power and 40 s total acquisition time. [Enzyme] = 175 µM, and if applicable, [Substrate] = 150 µM.  The near-equal intensity doublet at 1175 and 1206 cm-1 results primarily from tyrosine residues (Y9a and Y7a modes) [58, 141], with some contributions from phenylalanine (F7a) for the higher energy vibration [140, 142]. The vibration at 1340 cm-1 arises from an unresolved W7  90 Fermi doublet of tryptophan, with an additional W5 vibrational mode observed at 1460 cm-1 [146].  In the 1500-1700 cm-1 region contributions from all 3 residues are noted. The 1550 cm-1 signal results from a tryptophan W3 vibrational mode [146], while the more intense vibration at 1620 cm-1 is due to tryptophan (W1) [146] and tyrosine (Y8a) vibrational modes [58, 140, 141]. Weaker vibrations at 1575 (tryptophan, W2) and 1602 cm-1 (phenylalanine, F8a) are likely obscured by these more intense vibrations [123, 146].  Overall, the Raman spectra of native and non-native enzymes for HPCD and MndD are remarkably similar. This is consistent with the high degree of sequence identity and similarity in the enzyme active site [73]. Furthermore, the data shows that changing the central metal atom causes no major perturbations in the Raman spectra.  The Raman difference spectrum of the bound substrate in HPCD is also shown in Figure 2.16. Typically, all difference spectra show strong vibrations in the 1500 – 1700 cm-1 arising from 3 overlapping vibrations. Minor vibrations around 1450, 1330 and 1000 cm-1 are also observed. All difference spectra exhibit qualitatively similar vibrational spectra, suggesting similar modes of binding for both the native (Fe-HPCD) and non-native (Mn-HPCD and CoHPCD) enzyme systems. More importantly, these spectra all show distinct shifting of vibrational bands between H2O and D2O. This strongly suggests the presence of exchangeable protons in the enzyme-bound substrate and effectively rules out HPCA binding to the enzyme in the trianionic state.  91 Gaussian peak-fitting analysis of the 1500-1700 cm-1 region was performed to identify the diagnostic ring-mode vibrations which are critical to identify the substrate protonation state. In general, all enzyme systems exhibited a strong vibration around 1615 cm-1, in addition to two medium-intensity vibrations around 1550 and 1575 cm-1. For Fe-HPCD, a strong vibration was observed at 1614 cm-1, with two medium-intensity vibrations at 1551 and 1578 cm-1 in H2O. In D2O, the vibrations upshift by 7, 5 and 2 cm-1 respectively (Figure 2.18, Table 2.4).  92  Figure 2.18: Gaussian peak-fitting analysis of the enzyme-bound substrate difference spectrum in the diagnostic 1500-1700 cm-1 region. Left Panel: HPCA bound to Fe-HPCD in (a) H2O and (b) D2O. Center Panel: HPCA bound to Mn-HPCD in (c) H2O and (d) D2O. Right Panel: HPCA bound to Co-HPCD in (e) H2O and (f) D2O. The red lines represent individual peaks obtained from peak fitting and the black line is the raw data.  In the non-native enzyme Mn-HPCD, the dominant vibration occurs in H2O at 1617 cm-1, and upshifts by 6 cm-1 in D2O. The medium intensity vibrations occur at 1550 and 1577 cm-1 in H2O, are upshifted by 5 and 3 cm-1 respectively following H/D exchange (Figure 2.18, Table 2.4). In Co-HPCD, vibrations are observed at 1548, 1574 and 1614 cm-1 in H2O, and the vibrations upshift by 3, 3 and 4 cm-1 in D2O (Figure 2.18, Table 2.4).  93  Table 2.4: Band positions and intensities for enzyme-bound HPCA in the HPCD:HPCA complex. Shifts of vibrations upon deuterium substitution are also noted. Band intensities and shape denoted by: (s) strong, (m) medium, (w) weak, (vw) very weak, (sh) shoulder. FeHPCD:HPCA (H)  H:D shift  MnH:D CoH:D HPCD:HPCA shift HPCD:HPCA shift (H) (H)  Assignment  1614(s)  +7  1617(s)  +6  1614 (s)  +3  νCC, 8a  1578(m)  +2  1577(m)  +3  1574 (m)  +3  νCC + νCO, 8b  1551(m)  +5  1550(m)  +5  1548 (m)  +3  νCC  1459(w)  +4  1454(w)  +6  1452 (w)  +5  νCC, 19b  1336(m)  -2  1338(w)  -3  1340 (w)  -2  νCO, 7a  The Raman difference spectrum for the MndD suite of enzymes is again dominated by vibrations in the 1500 – 1700 cm-1 region. Peak fitting analysis of Mn-MndD indicates a strong 8a vibration at 1622 cm-1 in H2O, which upshifts by 3 cm-1 in D2O. Two shoulder peaks at 1555 and 1582 cm-1 upshift by 4 cm-1 upon H/D exchange (Figure 2.19, Table 2.5). For Fe-MndD, a strong vibration is noted at 1614 cm-1 in H2O, followed by shoulder vibrations at 1577 and 1552 cm-1. Following H/D exchange the intense 8a band upshifts by 3 cm-1, while the shoulder bands upshift by 5 and 4 cm-1, respectively (Figure 2.19, Table 2.5).  94  Figure 2.19: Gaussian peak-fitting analysis of the enzyme-bound substrate difference spectrum in the diagnostic 1500-1700 cm-1 region. Left Panel: HPCA bound to Mn-MndD in (a) H2O and (b) D2O. Right Panel: HPCA bound to Fe-HPCD in (c) H2O and (d) D2O. The red lines represent individual peaks obtained from peak fitting and the black line is the raw data.  The direction and magnitude of the shifts observed following H/D exchange eliminates the possibility that the observed difference spectrum is due to pertubations in the enzyme  95 coordination sphere following substrate binding. For example the 1550 cm-1 tryptophan W3 vibration was reported to be insensitive to H/D exchange at pH/pD 7.4 [36], while Copeland noted a downshift of 6 cm-1 between H2O and D2O [35]. The 1616 cm-1 tryptophan band was observed to downshift by 13 cm-1 in the d5-Trp compound [147], and 9 cm-1 by Copeland following H/D substitution of the indole NH group [35]. Likewise, the Tyrosine Y8a vibrational mode at 1613 cm-1 shifts down by several cm-1 upon H/D exchange [148, 149]. The fact that the bands in the difference spectra show a distinct upshift upon deuteration would seem to preclude the idea that these bands arise from amino acid residues.  Table 2.5: Band positions and intensities for enzyme-bound HPCA in the MndD:HPCA complex. Shifts of vibrations upon deuterium substitution are also noted. Band intensities and shape denoted by: (s) strong, (m) medium, (w) weak, (vw) very weak, (sh) shoulder. Mn-MndD:HPCA (H)  H:D shift  Fe-MndD:HPCA (H)  H:D shift  Assignment  1622(s)  +3  1614(s)  +3  νCC, 8a  1582(m)  +4  1577(m)  +5  νCC + νCO, 8b  1555(m)  +4  1552(m)  +4  νCC  1460(w)  +3  1450(w)  +5  νCC, 19b  1343(w)  -3  1340(m)  -1  νCO, 7a  The upshift to higher energy following H/D exchange in the enzyme-bound substrate is unusual and requires further explanation. The Raman spectrum of the free substrate noted distinct downshifts following exchange, unlike the small upshift observed in the enzymesubstrate complex. Previous studies with 2,3-dihydroxybiphenyl (DHB) bound to the  96 corresponding dioxygenase noted no shift in the 8a vibrational band between H2O and D2O solvents despite having exchangeable protons, contrary to data acquired for the free substrate [26, 62]. A similar phenomenon was also noted for protocatechol (PC) bound to protocatechol dioxygenase, where vibrational bands underwent a distinct upshift following H/D exchange and was contrary to data obtained for the free substrate [150].  The nature of this effect can be attributed to substrate interactions with hydrogen-bonding residues in the enzyme active site [26, 62]. Crystal structures of the E:S complex between 4NC and Fe-HPCD demonstrates a hydrogen bonding Tyr257 residue stabilizes the substrate O3 group [116]. The major bands in the 1500 – 1700 cm-1 are ring vibrational modes with varying degrees of –OH bending displacement. Hydrogen bonding interactions would suppress substrate O-H displacement [26], removing these contributions from the molecular vibrations and in turn affecting the H/D exchange characteristics of the substrate. Furthermore, this behaviour would not be observed in the Fe-HPCA model compounds for two reasons: No hydrogen bonding residues exist in the model compound to mimic this behaviour, and the substrate in the model compound has no exchangeable protons.  The enzyme-bound substrate spectra for Fe-HPCD, Mn-HPCD and Co-HPCD all have common features in the catechol ring mode region which strongly correlated to the spectrum of dianionic HPCA. More importantly, because shifts in the vibrational bands were observed following H/D exchange, this clearly eliminates the possibility of substrate binding in the trianionic protonation state. Similar results are also obtained for Mn-MndD and Co-HPCD.  97 Previous crystallographic observation of the anaerobic HPCD complex suggested asymmetrical binding of HPCA to the iron metal center [106, 112]. In particular, the crystal structure showed ionization of the C3-OH group based on observed crystal structure bond lengths. The results obtained here are consistent with this experimental evidence: Raman spectroscopy shows binding of HPCA bound to the enzyme in the dianionic state, with one of the catechol –OH moieties deprotonated following enzyme binding.  The absolute location of the catechol ring modes in the 1500 – 1700 cm-1 region differs compared to both dianionic HPCA and the Fe-HPCA model compounds. Based on the FeHPCA model compounds, it was expected that complexation with the metal ion would induce large (ca. 10-20 cm-1) upshifts in the νCC and 8b vibrational modes, compared to the corresponding spectrum for the free substrate. However, for the enzyme-bound substrate, a small upshift was noted for the 8a vibrational mode, but the 8b and the νCC vibrations instead showed substantial downshifts of 9 and 21 cm-1 respectively (compared to the dianionic HPCA compound). As with the unusual H/D exchange characteristics of the enzyme-bound substrate, this behaviour is attributed to differences in the binding modes of the substrate in the enzymebound substrate and the model compound. Enzyme-bound HPCA will experience stabilization interactions from the Tyr257 residue in the enzyme active site [116]. The 8b vibrational mode, which has a large component on the C-OH moiety [126], would experience the greatest perturbation, leading to the downshift observed here.  In addition, the asymmetry of the catecholate bonding to the metal center would greatly perturb the frequencies of the affected vibrations. Previous DFT studies of the Fe-catechol system noted a dramatic downshift in the 8a vibrational frequency resulting from C-C bond  98 elongation within the catechol ring [64]. Crystal structures of the anaerobic HPCA:HPCD complex noted asymmetric bond lengths of 2.49 and 1.86 Angstroms for the C-O catecholate moieties [106, 112], which would further perturb the vibrational frequencies. Once again, such behaviour was not observed in the Fe-HPCA model compounds due to the fact the substrate is bound in the trianionic state, with equivalent bond lengths for both C-O moieties [64].  No corrections were made for unbound (free) HPCA, as this only represents 5% of the total HPCA (based on Kd determination, data not shown). Unbound HPCA would exist in the monoanionic form at the solution pH used, which has a distinct vibrational spectrum compared to the bound HPCA. Therefore, given the small concentration of free HPCA and the limited SNR from the difference spectrum, corrections would have a minimal effect on the difference spectrum.  2.4  Conclusions The HPCD and MndD class of enzymes were studied using UV-Vis and UV resonance  Raman spectroscopy. Samples containing the physiological metal (Fe2+) along with nonphysiological metals (Mn2+, Co2+) were studied in HPCD. Likewise, samples of MndD containing the physiological (Mn2+) and non-physiological metal (Fe2+) were also examined. The results demonstrate that incorporation of non-physiological metals into the enzyme active site has no effect on the substrate protonation state in the anaerobic enzyme-substrate complex. This result is consistent with previous observations by Emerson [73], which demonstrated that key kinetic parameters remain unchanged following incorporation of the non-native metal into  99 the enzyme. In turn, these results provide support for the proposed mechanism in which the metal oxidation state does not change during the course of the enzymatic reaction [73].  2.5  Acknowledgements Financial support for this work was provided by the Natural Sciences and Engineering  Research Council of Canada (NSERC), the Canada Foundation for Innovation, the British Columbia Knowledge Development Fund and the Michael Smith Foundation for Health Research. CJA was partially supported by a fellowship provided by NSERC.  100  Chapter Three: FO-UVRRS Study of a Carbon-Carbon Bond Hydrolase  3.1  Introduction The microbial degradation of aromatic compounds is crucial to maintaining the global  carbon cycle [151]. The aerobic degradation of these compounds often proceeds via a catecholic intermediate which is successively transformed by a meta-cleavage dioxygenase and then a meta-cleavage product (MCP) hydrolase [105]. The latter reaction, an unusual carbon-carbon  bond fission, is exemplified by the hydrolysis of 2-hydroxy-6-oxo-6-phenylhexa-2,4-dienoic acid (HOPDA) by BphD in the biphenyl catabolic pathway. The ability of this pathway to degrade polychlorinated biphenyls (PCB) is limited in part because BphD is inhibited by PCB metabolites such as 3-chloro-HOPDA (3CH) [152, 153]. More recently, a homologue of BphD has been shown to be involved in steroid catabolism [154] and appears to be critical for the pathogenesis of Mycobacterium tuberculosis, the etiological agent of tuberculosis [155].  Although the MCP hydrolases contain a ‘catalytic triad’ (Ser112-His265-Asp237 in BphD) and the structural fold of the α/β-hydrolase enzyme superfamily, studies suggest that these enzymes deviate from the classical nucleophilic mechanism of serine proteinases in two respects. First, hydrolysis appears to involve a general base mechanism in which the catalytic His activates water, yielding a gem-diol intermediate rather than the prototypical acyl-enzyme [156-158]; Second, an enol-keto tautomerization [159, 160] has been proposed to precede hydrolysis (Figure 3.1). Indeed, early studies suggested that MCP hydrolases may promote tautomerization by binding the substrate in a strained, non-planar conformation [160].  101  Figure 3.1: Proposed Mechanism of BphD. Absolute conformation of each tautomer is not intended.  We recently demonstrated that BphD-catalyzed hydrolysis involves the rapid transformation of the free HOPDA enolate (λmax = 434 nm) to a transient intermediate possessing a red-shifted absorption spectrum, E:Sred (λmax = 492 nm) [161]. The catalytic importance of this intermediate is supported by its absence when the wild type enzyme was mixed with the inhibitory compound 3-chloro-HOPDA (3CH) [162]. Moreover, E:Sred did not accumulate in the inactive S112A/H265A variant of BphD, but a similar intermediate (λmax = 506 nm) was trapped in the S112A variant, indicating that formation and decay of E:Sred is catalyzed by His265 and Ser112, respectively [163]. The trapped E:Sred intermediate is readily observed as an orange-  102 colored S112A:HOPDA complex, in contrast to the characteristic yellow color of the enolate retained in the S112A:3CH complex (Figure 3.2(a)).  103  Figure 3.2: Comparison of the planar and non-planar conformations of enzyme-bound HOPDA. (a) 4 µM HOPDA in aqueous buffer (blue), the transient intermediate obtained upon mixing 8 µM BphD with 4 µM HOPDA (black) [161] and a mixture of 8 µM S112A and 4 µM HOPDA (red) [163]. (b) Crystal structures of HOPDA bound to the S112A/H265A variant of BphD (white) as a planar enolate (E:Se), and HOPDA bound to the S112A variant (straw) in a non-planar fashion (E:Sred) [162]. Oxygen and nitrogen atoms are red and blue, respectively.  104  Insight into the structure of E:Sred has been obtained from X-ray crystallography [162, 163]. Whereas the crystal structures of S112A:3CH and S112A/H265A:HOPDA revealed substrates bound as planar enolates (E:Se), a non-planar HOPDA was present in the active site of S112A:HOPDA (Figure 3.2(b)). Multiple restrained refinements were used to assess the compatibility of different HOPDA isomers with the X-ray diffraction data. The results indicated these data are fully compatible with the expected conformational parameters of a ketonized isomer (E:Sk). Nevertheless, the refinements did not exclude the 2-enol tautomer. For this isomer, despite tight restraints on the C2=C3 and C4=C5 torsion angles, refinement produced non-trivial deviations from the expected values for those angles: C2=C3, 175 vs. 180 deg; and C4=C5 (12 vs. 0 deg). A strained enolate (E:Sse) more satisfactorily accounts for the red-shifted absorption spectrum of the intermediate, as steric hindrance to planarity about a double bond raises the ground state energy level but not the excited state [164]. For example, twist about double bonds in the retinal chromophore of bacteriorhodopsin has been proposed to explain the red-shifted spectrum of the L photointermediate to λmax = 550 nm from the expected value of  λmax ≈ 491 nm [165]. This distortion of the chromophore has recently been observed crystallographically [166]. By comparison, the disrupted π-conjugation of E:Sk is predicted to result in a blue shift, although proteins can significantly perturb the absorption maxima of bound ligands [167-169].  To further probe the tautomeric state of the non-planar E:Sred intermediate, we applied fiber-optic UV resonance Raman (FO-UVRR) spectroscopy [62, 64] to measure the vibrational spectrum of the non-planar Sred ligand in the S112A:HOPDA complex. This was compared to  105 the spectra of the planar ligands in S112A:3CH and S112A/H265A:HOPDA complexes. The tautomeric states of the bound HOPDAs were assigned using two approaches: i) comparison to experimentally-obtained spectra of enol and enolate tautomers in solution; and ii) comparison to theoretical spectra for enol, enolate, and keto tautomers obtained by density functional theory (DFT) calculations.  3.2  3.2.1  Materials and Methods  Chemicals and Protein HOPDA and 3CH were enzymatically generated from 2,3-dihydroxybiphenyl (DHB) and  4-Cl-DHB, respectively, using 2,3-dihydroxybiphenyl dioxygenase as previously described [161, 170]. The preparation of DHB and chlorinated DHBs has been described elsewhere [171]. All other chemicals were of analytical grade. Wild-type and variant BphDs from Burkholderia xenovorans LB400 were prepared as previously described [161].  3.2.2  Preparation of Samples All work was performed in potassium phosphate buffer, I = 0.10 M, 25 mM KNO3, pH  7.5. Enzyme-substrate complexes for UVRR were prepared by mixing appropriate quantities of each in disposable microcentrifuge tubes to yield 150 µM enzyme and 140 µM substrate in a total volume of 150 µL. Samples containing the substrate only were prepared by mixing the appropriate quantity of substrate in disposable microcentrifuge tubes to yield 140 µM substrate in a total volume of 150 µL. The freshly prepared samples were flash frozen in liquid nitrogen, and thawed at room temperature immediately prior to UVRR analysis.  106  3.2.3  Dissociation Constants The dissociation constant (Kd) of the S112A:HOPDA complex was determined  spectrophotometrically based on its red-shifted absorbance maximum (506 nm) relative to free HOPDA (434 nm). Spectra were recorded using a Varian Cary 5000 spectrophotometer (Varian Canada, Mississauga, ON, Canada) equipped with a thermostatted cuvette holder maintained at 25.0 ± 0.5 °C and controlled by Cary WinUV software version 3.00. The sample buffer contained potassium phosphate, I = 0.1 M, 25 mM KNO3, pH 7.5. Varying concentrations of HOPDA (0.21 to 20 µM) were added to 2.0 µM of S112A BphD, and the absorbance spectrum was recorded at each concentration. The concentration of bound HOPDA, [E:S], was calculated from the absorption at 506 nm after correcting for the absorbance from free HOPDA ([S]f; ε506 = 2.00 mM-1 cm-1). The correction factor and the Kd values were calculated iteratively by fitting equation (12) to the data using LEONORA [172].  [E:S] =  [E]0 [S]f [S]f + K d  (12)  Briefly, the concentration of free HOPDA was calculated from an initial estimate of Kd, and the absorption at 506 nm was corrected for [S]f accordingly. The corrected absorbance was used to determine the fraction of HOPDA bound to enzyme by taking the ratio of the absorbance at 506 nm relative to the maximum change after addition of a large excess of HOPDA. The concentration of bound HOPDA was obtained by multiplying this fraction by the total concentration of enzyme, [E]0. [S]f was then re-calculated as the difference between the total  107 HOPDA concentration and [ES]. Equation (12) was then fit to the corrected data to obtain a better estimate of Kd, and the correction factor and calculations were repeated. The Kd value converged after 3 iterations.  3.2.4  UVRR Measurements The UVRR spectroscopy instrumentation has been described previously [63, 64].  Briefly, UV laser light was generated using a frequency doubled Ar+ laser operating at 248 nm (Innova 90C FreD, Coherent Inc, Santa Clara, CA). Pump beam delivery and collection of the Raman scattered light was accomplished using a custom fiber-optic probe designed for UVRR spectroscopy [47, 48]. The laser power delivered to the sample was maintained at 10 mW. Raman scattered light was measured using six collection fibers, passed through a dielectric stack interference filter (Barr Associates Inc., Westford, MA) to reject Rayleigh scattered light, and directed into a 1.0 m focal length monochromator (Model 2061, McPherson Inc, Chelmsford, MA) equipped with a 3600 groove/mm holographic grating (Model 8358-1004-0, McPherson Inc, Chelmsford, MA). Wavelength-dispersed light was collected with a liquid nitrogen-cooled CCD detector operating at –120 oC (Spec-10 400B, Roper Scientific, Trenton, NJ). A custom designed sample spinner was employed to continually rotate the samples during the acquisition of data to promote convective mixing. Acquisitions of 10 s duration were obtained for each sample to ensure that no photodecomposition of the sample occurred. To further verify sample integrity during the experiment, and to increase the signal-to-noise ratio, successive spectra acquired from a given sample were compared via difference spectroscopy. Spectra were averaged only if no significant changes occurred in the spectra during the UVRR experiment. All spectra were normalized using the 1050 cm-1 peak of the KNO3 internal standard prior to  108 calculating difference spectra. Spectral variation among samples (e.g., varying background and signal levels) was corrected for by mathematically offsetting the spectra to minimize difference between the internal standard peaks in the two spectra. Near-zero residuals in the region of the internal standard indicated high accuracy in the spectral subtraction of the parent spectra (E:S – E). Following each 10 s spectral acquisition, 10 µL of each sample was removed. Activity assays of these aliquots established that the enzyme maintained greater than 80% activity over the course of 30 s of laser exposure.  3.2.5  Computational Methods The Gaussian 03 software package was used for all calculations [173]. All geometries  and vibrational frequencies were calculated using the Kohn-Sham Density Functional Theory (DFT) method B3LYP [174, 175] with the cc-pVDZ [176] basis set. The results of the DFT calculations were used to simulate Raman spectra. The spectra utilized lorentzian line shapes with a full width at half-maximum of 50 cm-1 for all vibrations. The relative intensities of the peaks were based on the classification of the vibrational intensities resulting from the theoretical calculations.  3.3  3.3.1  Results  Dissociation constant of S112A variant for HOPDA The dissociation constant of the S112A:HOPDA complex was measured to determine  how much free HOPDA was present during the UVRR acquisitions and therefore take into account possible signals arising from this species. HOPDA undergoes a large shift in absorbance  109 maximum from 434 nm to 506 nm upon binding to S112A. The absorbance of S112A:HOPDA at 506 nm was measured as 2 µM S112A was titrated with HOPDA (Figure 3.3(a)). At equimolar amounts of S112A and HOPDA, the spectrum resembles that of the S112A:HOPDA complex and possesses almost no HOPDA peak at 434 nm, indicating tight binding between S112A and HOPDA. The binding was quantified from the binding isotherm (Figure 3.3(b)) to yield Kd = 1.0 ± 0.2 µM, which means ~5% of the HOPDA would remain unbound under the conditions of the UVRR experiment.  110  Figure 3.3: The binding of HOPDA by S112A as monitored by absorption spectroscopy. (a) Titration of 2 µM S112A with increasing concentrations of HOPDA. (b) Plot of concentrations of bound ([E:S]) versus unbound ([S]f) HOPDA. The red line represents the fit of equation (12) to the data (black squares) with [E]0 = 2.0 µM, and Kd = 1.0 ± 0.2 µM.  3.3.2  UV Resonance Raman Spectroscopy of Free Substrates Initial FO-UVRR experiments were performed with the free substrate (HOPDA) to  ascertain its vibrational spectrum. In aqueous buffer, the dominant features for HOPDA  111 consisted of a strong vibration observed at 1598 cm-1 together with a weaker, very broad feature at 1668 cm-1 (Figure 3.4(b)). Additional weaker features were observed in the Raman spectrum at 1182, 1350 and 1400 cm-1.  112  Figure 3.4: UVRR spectra of (a) HOPDA pH 6 (b) HOPDA pH 7.5 (c) HOPDA pH 9 (d) 3chloro HOPDA (3CH) pH 7.5. The UVRR spectra were obtained using potassium phosphate (I = 100 mM), 25 mM KNO3, solution pH as indicated, 25 °C.  To observe the enol and enolate forms of HOPDA, Raman spectra were also acquired at pH 6.0 (Figure 3.4(a)) and 9.0 (Figure 3.4(c)) as the pKa of the 2-hydroxyl is 7.3 [153]. The spectra recorded at each of the three pH values were very similar, with the exception that the  113 1598 cm-1 vibration was much reduced in intensity at pH 9.0. In addition, several weak bands become resolved in the 1200-1400 cm-1 region at varying solution pH. However, deprotonation of the 2-hydroxyl did not significantly affect the main diagnostic vibrational frequencies of the observed HOPDA spectrum. Accordingly, the enol and enolate forms of HOPDA are together identified herein as HOPDAe.  The Raman spectrum of 3CH (Figure 3.4(d)) was very similar to that of HOPDA. The peaks at 1598 and 1668 cm-1 observed in HOPDA were upshifted to 1599 and 1678 cm-1, respectively, in 3CH. In addition, the 3CH peak at 1599 cm-1 was relatively more intense than the 1598 cm-1 peak in HOPDA.  One of the main advantages of employing UVRRS is that the technique enables greater sensitivity and selectivity due to the resonance process. Tuning of the excitation wavelength to coincide with an electronic transition within the ligand enhances the latter’s spectrum relative to those of the protein and solvent. In order to determine the optimum excitation wavelength for experiments involving HOPDA, a resonance Raman excitation profile (RREP) was obtained (Figure 3.5). While 257 nm provides the near optimum excitation for the main diagnostic HOPDA vibrations (1598 and 1668 cm-1), excitation at this wavelength often results in fluorescence from phenylalanine and tyrosine residues in the protein [62, 123]. On the other hand, wavelengths further into the UV (229 and 238 nm) yield a Raman spectrum dominated by the aromatic amino acid residues due to their large number and high Raman cross section [62]. Therefore, excitation at 248 nm was used to optimize substrate enhancement while minimizing spectral interference.  114  Figure 3.5: Resonance Raman excitation profile of HOPDA at accessible wavelengths in the UV region. Markers are identified as follows: + – 1598 cm-1 vibration; ● – 1668 cm-1 vibration; ♦ – 1300 cm-1 vibration; ■ - 1350 cm-1 vibration.  3.3.3  DFT Calculations DFT calculations were performed to aid in the vibrational assignments of the ligands as  the keto tautomers, HOPDAk and 3CHk, are not accessible experimentally. In the calculated spectrum of HOPDAe, the strongest features occurred at 1566 and 1625 cm-1 (Table 3.1 and Figure 3.6). These most logically correspond to the bands at 1598 and 1668 cm-1, respectively, in the experimental spectrum. Accordingly, the band at 1598 cm-1 was attributed to a coupled vibration consisting of the C2-O/C2=C3 (enol) and C4=C5 vibrations, which are all in-plane in  115 the free substrate. This is consistent with previous observations by Schiavoni [177], in which the enol(ate) vOCC vibration occurred at 1627 cm-1 for a β-ketoester (acetylacetone). Here, the conjugation of the enol(ate) vOCC vibration with the C4=C5, C6=O and benzene moieties would lead to a downshift of the enol(ate) vibration to 1598 cm-1 when compared with Schiavoni [1]. The greater intensity of the 1598 cm-1 vibration, compared to the 1668 cm-1, also suggests that it is diagnostic of a C=C vibration [178]. The band at 1668 cm-1 was attributed to several overlapping vibrations consisting of C=C (aromatic), C2=C3 and C6=O stretches.  Table 3.1: Density Functional Theory (DFT) calculations for HOPDA in the enol(ate) (HOPDAe) and keto (HOPDAk) tautomeric states and there comparison to experimentally observed bands. Band intensities and shape denoted by: (s) strong, (m) medium, (w) weak, (vw) very weak. Calc. / cm-1  Assignment  Free HOPDA (experimental)  1472 (m)  C5=C6-H bend  1497 (vw)  C=C (arom.)  1566 (s)  C2=C3, C4=C5, 1598  S112A-bound HOPDA  1594 / 1613  HOPDAe  C2-O str. 1599 (w)a  C=C (arom),  1668  1674  1668  1674  C2=C3 and C6=O 1625 (s)a  C=C (arom), C6=O str.  116  Calc. / cm-1  Assignment  Free HOPDA (experimental)  1515 (w)  S112A-bound HOPDA  C=C-H (arom) rocking  1650 (s)  C=C (aromatic)  HOPDAk  stretch 1675 (m)  C3=C4, C2=O str.  1738 (s)  C3=C4, C1=O str.  a  1755 (m)  C2=O str.  1769 (s)  C6=O str.  Given the broad nature of the band at 1668 cm-1, it is assumed that it is a superposition  of multiple vibrational bands, including the calculated peaks at 1599 and 1625 cm-1.  By contrast, the strongest bands in the calculated spectrum of HOPDAk were shifted by 80-200 wavenumbers to higher frequencies, occurring at 1650, 1738 and 1769 cm-1, respectively (Table 3.1 and Figure 3.6). These vibrations were attributed to C=C aromatic stretches, C3=C4/C1=O and C6=O stretches, respectively. Additional medium-intensity bands were calculated at 1675 and 1755 cm-1 resulting from C3=C4/C2=O and C2=O vibrations, respectively. The predicted large upshift of the bands to higher wavenumbers when tautomerizing from the enol(ate) to the keto tautomer is also consistent with previous observations [177], in which the major bands in the keto tautomer were upshifted by over 100 cm-1 when compared to the enolate tautomer.  117  Figure 3.6: Simulated Raman spectra of (a) HOPDAe and (b) HOPDAk using the results of the DFT calculations. Spectra were simulated using lorentzian line shapes with a peak width of 50 cm-1. Individual vibrations are shown as grey lines and the sum of the vibrations is the black line.  118  In summary, the DFT calculations indicate that ketonization should induce dramatic changes in the resultant Raman spectra, with vibrations diagnostic of the enolate tautomer disappearing and new bands for the keto tautomer occurring at much higher wavenumbers.  3.3.4  UV Resonance Raman Spectroscopy of Enzyme-Substrate Complex The enzyme-substrate complex (E:Sred) was characterized using difference spectroscopy  to extract the Raman contributions from the substrate in the complexes [62, 64]. More specifically, nitrate was used as an internal reference so that the enzyme spectrum (E) could be subtracted from that of the E:S complex to yield the spectrum of the bound substrate (E:S - E). For S112A:HOPDA, the low dissociation constant (Kd = 1.0 ± 0.2 µM, see above) limited the proportion of unbound substrate to ~5%. The difference spectrum for S112A:HOPDA – S112A (Figure 3.7(a) and Figure 3.8(a)), corresponding to Sred, was dominated by the same features as observed in HOPDAe, except that they occurred at 1594 and 1674 cm-1. However, two additional bands were observed at 1613 and 1551 cm-1, respectively.  119  Figure 3.7: UVRR difference spectra of (a) E:HOPDA-E (E = 150 µM, S = 140 µM) (b) E:3CH-E (E = 150 µM, S = 140 µM) (c) S112A/H256A:HOPDA-S112A/H256A (E = 150 µM, S = 140 µM). The UVRR spectra were obtained using potassium phosphate (I = 100  mM), 25 mM KNO3, pH 7.5, 25 °C.  120 The difference spectra of both S112A-bound 3CH (1597 and 1692 cm-1, Figure 3.7(b) and Figure 3.8(b)) and S112A/H265A-bound HOPDA (1597 and 1678 cm-1, Figure 3.7(c) and Figure 3.8(c)) were very similar to the free HOPDA (1598 and 1668 cm-1, Figure 3.4(b)) and free 3CH (1599 and 1678 cm-1, Figure 3.4(d)) Raman spectra. This evidence suggests that HOPDA and 3CH are planar enolates when bound to the double and single variants, respectively.  121  Figure 3.8: Close-up view of the UVRR difference spectra in the 1500-1700 cm-1 region. The solid black line corresponds to the original data, while the dashed grey lines are individual peaks resulting from the peak fitting analysis. The sum of the individual peaks is represented by the solid red line. (a) E:HOPDA-E (b) E:3CH-E (c) S112A/H256A:HOPDA-S112A/H256A.  122  3.4  Discussion The dominance of bands around 1600 and 1680 cm-1 in the S112A:HOPDA-S112A  difference spectrum strongly suggests that the bound HOPDA, Sred, predominates as the enol/enolate tautomer. Importantly, none of the three higher frequency bands predicted to occur in HOPDAk were observed (Table 3.1). Moreover, several features of the bands around 1600 and 1680 cm-1 are consistent with strain on the C2=C3 and C4=C5 double bonds of the enzymebound HOPDA, consistent with a crystal structure of the S112A:HOPDA complex [162]. First, the observation of two bands around 1600 cm-1 (i.e., 1594 and 1613 cm-1, Figure 3.8(a)) could be the result of vibrational decoupling due to enzyme-induced strain of the double bonds. Changes in molecular geometry (conformation) are known to influence the vibrational frequency of normal modes [179-182]. More specifically, twisting about a carbon-carbon double bond in an extended conjugated network has been previously reported [183], resulting in a decoupling of the C=C stretch and the C-H bending vibrations, and a concomitant shift of the band to higher frequencies and reduced overall signal intensity. In HOPDAe, the decoupling of the C2=C3 and C4=C5 bonds would yield two discrete molecular vibrations as observed in the Raman difference spectrum of the enzyme-bound ligand. Moreover, decoupling of conjugated C=C vibrations can result in the sum of the intensities for both (decoupled) vibrations being reduced relative to the sum of the coupled vibration [184].  The conclusion that the bound HOPDA is a strained enolate is further supported by the shift of the 1668 cm-1 band in free HOPDA to 1674 cm-1 upon binding to S112A. The band at 1668 cm-1 is primarily attributed to a coupled vibration involving C=C aromatic and C6=O  123 moieties. Strain of the C2=C3 and C4=C5 double bonds is predicted to slightly decrease the conjugation throughout the HOPDA backbone. Such a decrease would enhance the double bond character of the C=C (aromatic) and C6=O moieties, yielding a slightly upshifted band [185].  DFT calculations suggest that the additional feature at 1551 cm-1 is due to an aromatic ring stretch [186]. As this band was not observed in the spectra of HOPDA in solution or bound to the S112A/H265A variant, it is possible that the absorption spectrum associated with the S112A-bound HOPDA enhances the 1551 cm-1 band. Alternatively, the 1551 cm-1 band may be attributed to a weak tryptophan vibrational (W3) mode [143, 144, 187]. The observation of this band may be consistent with the first steps of the proposed mechanism for the BphD reaction, in which a Trp266 and Asn111 residue coordinate via hydrogen bonding to the C2-O moiety of HOPDA [161]. This change in the coordination environment of the Trp266 residue would be reflected in differences between the E:S and E spectra, leading to a signal change in the difference spectrum. The fact that a similar vibration is observed in the difference spectrum for the S112A:3CH complex suggests that 3CH is coordinated at least to some extent in the active site of the enzyme, by way of the Trp266 residue.  Similarly, it is possible that the ca. 1613 cm-1 vibration could also be attributed to a tryptophan (8a) vibrational mode in the S112A:HOPDA difference spectrum [143, 144]. However, if coordination of the substrate to S112A induces changes in both the 1551 and 1613 cm-1 vibrations, then those changes should also be observed in the S112A:3CH spectrum. However, the fact that this vibrational mode is not observed in the S112A:3CH difference spectrum does not support this conclusion.  124  3.5  Conclusions The FO-UVRR spectroscopic data and DFT calculations strongly support the conclusion  that the S112A-bound HOPDA exists predominantly as an enolate tautomer in which the C2=C3 and C4=C5 double bonds are strained. By contrast, the Raman spectra of S112A:3CH and S112A/H265A:HOPDA indicate that bound ligands are in the planar enol(ate) tautomers in these complexes. These conclusions are consistent with the interpretation of the reported crystal structure [162] and the absorption spectroscopy [163]. As the red-shifted electronic absorption spectrum of S112A:HOPDA is similar to that of the catalytic intermediate, E:Sred, and this spectrum does not occur in the S112A/H265A double variant, the presented data further indicate that His265 assists in inducing strain in the C2=C3 and C4=C5 bonds, presumably facilitating tautomerization, but that the Ser112 catalyzes the ketonization. Indeed, we had previously noted that Ser112 is positioned to protonate C5 [163]. This study substantiates the hypothesis advanced by Bugg and colleagues more than a decade ago that the MCP hydrolases promote tautomerization by destabilizing the substrate via double bond strain [159, 160]. Moreover, it confirms our previous proposal that inhibition of BphD by 3CH arises from the enzyme’s inability to impart similar strain on this compound, thereby impeding ketonization [163]. Finally, in obtaining insight into the mechanisms of catalysis and inhibition in another class of enzymes, this study further demonstrates the utility of FO-UVRR spectroscopy as a probe of macromolecular structure and function, particularly involving photolabile metalloenzymes and their bound substrates.  125  3.6  Acknowledgements CJA and GPH contributed equally to this work. Financial support was provided by  Discovery Grants from the Natural Sciences and Engineering Research Council of Canada (to MWB, RFBT and LDE), the Canada Foundation for Innovation (CFI), and the British Columbia Knowledge Development Fund (BCKDF). Instrumentation and infrastructure were provided by the UBC Laboratory for Advanced Spectroscopy and Imaging Research (LASIR) and Laboratory for Molecular Biophysics (LMB). Dr. Shiva Bhowmik helped prepare Figure 1. Computational resources were provided by Westgrid and C-Horse.  126  Chapter Four: FO-UVRRS of Locked Single-Stranded Oligonucleotides1  4.1  Introduction Locked nucleic acids (LNAs) have a number of useful attributes, including a very high  affinity for and an excellent specificity toward Watson-Crick complementary DNA and RNA [188]. Consequently these and other properties are fuelling an increasing interest in biopolymers based on such conformationally restricted nucleosides [189]. The reader is referred to reviews and their references for more information about the range of these and other nucleoside analogs, their conformational properties, and applications in molecular biomedicine [188-191].  LNAs contain one or more [192] chemically modified ribonucleotides with a C4’ to O2’ methylene bridge (Figure 4.1 and schematic inset, Table 4.1) locking the furanose ring in the C3’-endo conformation [192-194]. In this conformation, the furanose ring is puckered such that the C3’ is on the same side of the plane of the ring as the C5’ atom, whereas the C2’ atom is either in the plane of the ring or on the opposite side [195]. Unlike some other modified nucleic acids (NAs), such as peptide nucleic acids (PNA) where the PNA–DNA complex adopts a B-like secondary structure with the sugar in the C2’-endo conformation [196] (Figure 4.2), nuclear magnetic resonance (NMR) spectra indicate that the LNA–DNA hybrid adopts an A-like  1  A version of this chapter has been published:  Stanislav O. Konorov, H. Georg Schulze, Christopher J. Addison, Charles A. Haynes, Michael W. Blades and Robin F. B. Turner (2009). Ultraviolet resonance Raman spectroscopy of locked single-stranded oligo(dA) reveals conformational implications of the locked ribose in LNA. Journal of Raman Spectroscopy, 40(9), 1162-1171.  127 structure (Figure 4.2) very similar to that of the near-canonical A-form of an LNA–RNA duplex [192].  Figure 4.1: Comparison of 2’-deoxyribose and ribose structure in DNA and LNA, respectively. LNA contains a methylene bridge connecting the 2’ oxygen and 4’carbon.  In single-stranded DNA (ssDNA), where base pair hydrogen bonding is absent, secondary structural organization, if any, is effected by base stacking interactions. Secondary structures do indeed occur in some ssDNA at room temperature [197] and further evidence derives from circular dichroism (CD) [198], femtosecond excited state transient absorption (fESA) [199], NMR [197], fluorescence [200], and Forster (fluorescence) resonance energy transfer measurements [200]. Furthermore, the different secondary structures [201, 202] that double-stranded (hybridized) DNA (dsDNA) take on, and that ssDNA may take on, are influenced by sequence [203-205], solvent [206], temperature [207], ligand binding [208] and other factors. Theoretical efforts explain base stacking phenomena in terms of dispersion  128 attraction, short-range exchange repulsion, electrostatic interactions, and π – π interactions [209213]. Electrostatic and dispersion forces generate the stability of stacking in DNA [209, 212]. Dispersion attraction derives from intermolecular coordinated electron motions (induced dipole– induced dipole interactions) that involve a redistribution of electronic densities inside interacting bases and shift the electronic energy levels of base-stacked NAs [214-216].  Figure 4.2: Helical structure of A-DNA (left), B-DNA (center) and Z-DNA (Right). This image, created by Richard Wheeler (Zephyris) and deposited in the Wikimedia Commons, is used under the Creative Commons Attribution-Share Alike 3.0 Unported license. This image can be found at: http://en.wikipedia.org/wiki/File:A-DNA,_B-DNA_and_Z-DNA.png  129 Structural organization is also effected by conformational constraints imposed by the backbone [195, 217]. Nevertheless, in monomeric units the C4’ –C5’ torsion angle γ is not correlated with other torsion angles although all others are mutually correlated [195]. Thus a considerable degree of structural freedom is allowed, although to a lesser extent in longer polymers [195]. Given the rotational freedom that γ allows, three preferred staggered positions per monomeric unit exist [195]. It is therefore of interest to know how a single conformation is selected from at least 3n possibilities in an n-mer (n > 1). Indeed, a number of lines of evidence suggest that the backbone confers little secondary structural order on DNA molecules. When rotational decay times of transient electric birefringence measurements were used to estimate persistence lengths for single-stranded poly(dT), a helical rise of 0.52 nm was obtained [218]. This value is inconsistent with the helical rise in A- and B-form structures where they are ~0.26 and ~0.34 nm, respectively [195, 219, 220]. Structural order in poly(dT) is mainly due to torsional restrictions within the phosphodiester backbone, and these are not sufficiently rigid in themselves to prevent random coil structures, whereas in poly(dA) the much greater intrinsic rigidity is due mainly to purine stacking [218, 221, 222]. f-ESA experiments show a great similarity in excited state lifetimes between poly(dA) and poly(dA)·poly(dT) [223, 224] suggesting that adenines are stacked similarly at room temperature in both single-stranded and hybridized forms. This follows because stacking causes a similar electronic coupling to occur between bases that is reflected in excited state lifetimes. Thus, the better intuitive model for secondary structure in DNA is that of ‘beads on a chain’ (short, rigid segments of the backbone have some effect) rather than ‘beads on a wire’ (the backbone has large-scale rigidity and this is dominant) or ‘beads on a string’ (backbone rigidity, if any, has no effect). In LNA, because of the methylene bridge, the backbone could be more rigid than in DNA and determine the secondary structure of the molecule to a greater, if not complete, extent. In this case, the  130 secondary structures of poly(dT) and poly(dA) should both be consistent with the A conformation. However, it is not clear exactly how the methylene bridge would affect the torsion angle γ although this torsion angle is known to be correlated with the P-O5’ torsion angle α in ADNA [195]. Furthermore, RNA adopts the A-form secondary structure and neighboring oxygens at the C2’ and C3’ positions do have an effect [195], but ring pucker is not conformationally constrained by a bridge.  131  Table 4.1: Monomers, dimers, trimers, pentamers and dodecamers used to detect the presence of base stacking in single-stranded nucleic acids and to assess the influence of a locked ribose on its own and neighboring bases. For a given sequence, the locked and unlocked versions of the oligomers had different secondary but identical primary structures. Locked adenine bases are capitalized. The position of the locked base(s), relative to the 5’-end, is specified by the upper case letters in the locked nucleic acid sequences. The inset shows an adenine monomer with locked ribose (C2’ –C4’ bridge). Name  Sequence  LM  5'-A-3'  LD  5'-Aa-3'  dAMP  5'-a-3'  (dA)2  5'-aa-3'  (dA)5  5'-aaaaa-3'  LPb  5'-aAaaa-3'  LPc  5'-aaAaa-3'  LPab  5'-AAaaa-3'  LPbd  5'-aAaAa-3'  (dA)12  5'-aaa aaa aaa aaa-3'  LNAa  5'-Aaa aaa aaa aaa-3'  LNAb  5'-aAa aaa aaa aaa-3'  LNAc  5'-aaA aaa aaa aaa-3'  LNAd  5'-aaa Aaa aaa aaa-3'  132  Name  Sequence  LNAg  5'-aaa aaa Aaa aaa-3'  LNAl  5'-aaa aaa aaa aaA-3'  LNAgj  5'-aaa aaa Aaa Aaa-3'  LNAdj  5'-aaa Aaa aaa Aaa-3'  LNAdgj  5'-aaa Aaa Aaa Aaa-3'  Because of the great biological importance of secondary structure in NAs – it affects the susceptibility of NAs to UV damage [199, 225, 226] and it influences the critical processes of hybridization, transcription, and replication [198, 227, 228] – it is essential to have a clear understanding of its origins and expression. Yet the relative extents to which the ordered aggregation of bases and the flexibility restrictions of the backbone determine secondary structure are not fully understood. DNA and its homologous LNA provide a unique opportunity to examine conformationally different isosequential oligomers under identical conditions to gain further insight into the differential contributions of base and backbone to the NA structure, and, specifically, how they shape the LNA structure.  As mentioned above, stacked neighboring bases in B-form DNA have considerable geometric overlap, exhibiting twist and displacement to optimize interactions between mostly positive σ- and mostly negative π-bonds [209, 213, 229]. In A-form DNA the degree of geometric overlap is much reduced [220]. Thus, in NAs with different secondary structures such as A- and B-form helices, the relative orientation of the bases and their degrees of neighboring overlap differ [219, 220], affecting stacking interactions as evident from CD [230-232] and fESA measurements [224]. More specifically, the energy levels in stacked and unstacked regions  133 of NAs are different because electronic interactions between bases occur in stacked regions [214, 224] and, within stacked regions, electronic interactions also differ depending on how the bases overlap [224]. Thus, f-ESA shows that long-lived excited states occur in stacked adenines but not in monomers [223, 224] and that these excited states are much longer-lived in DNA, where neighboring bases have a more extensive geometrical overlap than the corresponding LNA where the overlap is less [224]. Because the presence and type of stacking interactions modify electronic energy levels in NAs, they provide potential for probing with resonance Raman spectroscopy in order to elucidate further the nature and origins of secondary structural organization in NAs.  Ultraviolet resonance Raman spectroscopy (UVRRS) depends on transitions between electronic energy levels to furnish the observed enhancement of vibrational bands [9], can be highly sensitive to structural changes [233], and has become a powerful tool for the analysis of genomic structures [234-236]. For example, vibrational coupling between ribose and base rings occurs in nucleosides [233, 237] and UVRRS shows that in-plane purine ring vibrations are affected by ribose ring puckering [233]. UVRRS also demonstrates hypochromic loss in DNA upon thermal denaturing due to stacking disordering and strand separation [236]. The strong dependence of UVRRS intensities on electronic levels also permits the use of different excitation wavelengths to probe differential base stacking-altered electronic levels [214-216] in NAs. UVRRS is therefore expected to reveal modified ring puckering in LNA as well as its stacking differences when compared to B-DNA. However, no literature reports on the UVRRS of LNA yet exist  134 Using fiber optic-based UVRRS [40, 46, 47, 65] to investigate the nature of stacking interactions in selected DNA mono/oligomers and homologous LNA under defined solution conditions we (1) provide further evidence of base stacking in single-stranded NA oligomers; (2) show that the effects of the locked ribose on the glycosyl bond can be detected with UVRRS; (3) show that the local conformational effects of the locked ribose can be distinguished from effects due to changed interactions amongst the bases; and (4) show that the locked base affects the stacking in the nearest neighbors and much less in the next-nearest neighbors. These results, together with extant literature, provide strong evidence that base stacking interactions in LNA cause a conformational change to A-form. We infer this to be due to altered dipole moments, and in adenine stacks also to reduced relative levels of dispersion interactions that occur in bases in LNA compared to bases in homologous B-DNA. Thus, secondary structure in the polymers follows from altered stacking interactions. This insight is of particular importance in examining the question of how primary structure determines secondary structure since relatively more influence should be assigned to stacking effects and their sequence dependence.  4.2  4.2.1  Materials and Methods  Oligomers and Sample Preparation The monomers and oligomers of LNA and DNA used in this study are defined in Table  4.1. For convenience, the atom numbering convention of the (locked) adenine monomer is shown in the inset. Single-stranded monomers, dimers, trimers, pentamers, and dodecamers of DNA (IDT, Coralville, IA) and sequence-homologous LNA (IDT, Coralville, IA, for dodecamers and Exiqon, Vedbaek, Denmark, for monomer, dimers, and trimers) were dialyzed in 10 mM  135 phosphate buffer at pH 6. The buffer solution consisted of 1 M NaCl, 10 mM Na2HPO4, and 1 mM Na2(EDTA) and was filtered through a 0.2 µm syringe filter to remove large dust particles. Stock concentrations of single-stranded oligomers were determined by measuring the absorbance at 260 nm and 80 °C (to minimize hypochromic effects) using a Cary 1E UV–vis spectrophotometer (Varian, Palo Alto, CA).  4.2.2  UV Resonance Raman Spectroscopy Samples of 300, 150, 100, 60 and 25 µM for monomers, dimers, trimers, pentamers, and  dodecamers, respectively, were prepared by diluting the stock solution with buffer and adding NaNO3 to a concentration of 8 mM as an internal standard. These were then divided into 400 µl aliquots. UVRR spectra were acquired with excitation at 257 and 244 nm (with 10 mW at the sample) to facilitate discrimination between isosequential LNA and DNA. Spectra consisted of four successive acquisitions of 20 s duration each.  We used a fiber-optic probe, described in detail elsewhere [47, 48, 57], for UVRRS of DNA and LNA oligomers. Briefly, the fiber-optic probe consists of one central solarizationresistant fiber (600 µm in diameter) that provides excitation light, coupled into the fiber from an intracavity frequency-doubled argon ion laser, to the investigated sample. The excitation fiber is surrounded by six solarization-resistant collection fibers (each 400 µm in diameter). The tips of the collecting fibers are beveled and polished at 45° angles and are coated with aluminum to collect scattered radiation at 90° from the axis of the excitation fiber through the sides of the fibers opposite the beveled ends. Scattering is reflected off the beveled surfaces into internal guided modes of the fibers. The fiber-optic probe tip is immersed in the sample for spectral  136 acquisition. Convective flow of the sample through the probed volume is maintained by spinning the microcentrifuge tube with a custom motorized sample-tube rotator. The collected light is coupled into a 1-m monochromator (McPherson Inc, Chelmsford, MA), which disperses light onto a liquid-nitrogen-cooled CCD detector (Roper Scientific, Trenton, NJ). The spectral resolution of the instrument is 2–3 cm−1. Spectra were collected from samples maintained in an ice bath at a temperature of ~5 °C.  4.2.3  Data Analysis Data were manipulated and analyzed with the Origin 7 (OriginLab, Northampton, MA)  and with the Matlab 7.0 (The MathWorks, Natick, MA) software. Secondary structures were generated and analyzed with the 3DNA 1.5 (Lu & Olson, Rutgers University, NJ) software [220]. Molecular graphics images were produced using UCSF Chimera (Resource for Biocomputing, Visualization, and Informatics, University of California, San Francisco) [238].  4.3  Results and Discussion The question of major interest here is how the secondary structures of NAs are generated  and more specifically how the secondary structure of LNA comes about. Since contributions from both the locked ribose and from stacking interactions are expected, it is worthwhile to consider briefly their anticipated effects on resonance Raman scattering.  At ~5 °C, the temperature at which the samples were interrogated, the dodecamer is expected to be in the α’B-form [203, 207] conformation, a conformation in which there is greater  137 overlap of the pyrimidine than imidazole rings as shown in Figure 4.3(a). Stacking in shorter oligo-adenines is also expected to be in α’B-form, while LNAs are more likely to be in A-form where less overlap occurs between adjacent adenines as shown in Figure 4.3(b). Thus, in B-form DNA the imidazole rings are relatively unperturbed by stacking and not affected by the presence of the additional O2’ while the pyrimidine rings are overlapping and expected to be mostly affected by stacking interactions. In A-form, there is little overlap by the rings and neither imidazole nor pyrimidine rings are expected to be affected by stacking interactions, except indirectly via the effect of the O2’, while the imidazole ring, being closer to the O2’, is expected to be more affected by the locked ribose than the more distal pyrimidine ring. This forms the basis for our expectation to see different effects, stacking versus locking, exhibited by different parts of the base, imidazole versus pyrimidine, as explained in more detail below.  138  Figure 4.3: (a) B-form. The dodecamer, at the interrogation temperature of ~5 °C, is in α’B-form [203, 207], as shown here, where the pyrimidine moieties of stacked bases have more extensive overlap than imidazole moieties. (b) A-form. In contrast to α’B-form, there is less overlap between bases in A-form conformations. The slight overlap that occurs is between imidazole and pyrimidine moieties as is evident from the two clearly visible adenines in the figure. However, the imidazole ring is closer to the modified ribose, and expected to be more influenced by the O2’, than the pyrimidine ring.  For stacked and overlapping bases in B-DNA, attraction forces between σ- and π-bonds should favor a rotation of adjacent molecules so that π-bonds are positioned over σ-bonds and not over other π-bonds. Indeed, calculations of stacked coplanar bases with ~3.4 Å separation  139 between them show that the interaction energy is primarily determined by twist angle rather than displacement [239]. Since the rings are polygons and not circles, rotation would cause the π-bond to be closer to the polygon center than where the ‘target’ σ-bond is located. Thus, the nucleus at the origin of the s-orbital may be attracted by and move closer to the π-bond and hence closer to the polygon’s center. When the nuclei at the polygonal apices move closer to the polygonal center, reductions in bond lengths between them occur. The reduced bond lengths will result in reduced polarizabilities [240] and increases in vibrational force constants compared to the monomer. Indeed, density functional theory calculations show that classical electrostatic interactions have a major effect on the shape and depth of the potential energy surface in two stacked Watson–Crick base pairs [241] and on bond strength (in diatomic molecules [242]). Reduced bond lengths should translate into decreased Raman scattering intensities and frequency upshifts in the vibrational modes pertaining to molecular bonds so affected. A second effect arises from the electronic coupling between adjacent stacked bases. Oligomers with stacked bases have modified electronic states due to induced dipole–induced dipole interactions between the bases [214-216, 224]. Specifically, electron correlation causes a blue shift in absorption intensities [215, 243] indicating that excitation in electronically coupled [244] bases is energetically more costly although the molecule subsequently attains a lower excited state energy [245] (presumably after electron recorrelation) while stacking itself may make the nonplanar excited states [246] of bases more difficult to attain. Therefore both electron correlation and planar confinement suggest that the excitation of stacked bases would be energetically more costly than the excitation of individual bases and that the UVRR scattering intensities of stacked bases should correspondingly be reduced. Note that this hypochromicity is not dependent on base pairing. Furthermore, electron correlation should also lead to somewhat correlated  140 vibrational modes. This mutual restriction would reduce the range of motion in vibrating bonds leading likewise to reduced scattering.  For a locked base, the presence of the additional oxygen in the ribose, the O2’, causes the O4’ to be more electronegative. Because the lone electron pairs of the O4’ have partial pcharacter, and because the O4’ is conjugated with the base attached at C1’, electrons are withdrawn from the imidazole ring [195]. Withdrawing electrons from the imidazole ring should cause an increase in bond lengths in the imidazole, and increased bond lengths should lead to increased polarizabilities of the bonds [240] and hence to increased Raman cross-sections. Consequently, increased intensities due to larger cross-sections and down shifts due to reduced force constants in imidazole-related peaks should occur for adenines in LNA when compared to those in DNA. Note that electron withdrawal by the ribose may also result in a knock-on electron withdrawal by the imidazole ring from the pyrimidine ring such that the imidazole ring becomes more electron-rich than the pyrimidine ring. Since the electron density of adenine rings influences their stacking interactions [197], this change in the electron density of the rings brought about by the lock may modify or alter stacking independently of effects by the methylene bridge on backbone torsion angles. A second consequence of locking the base arises from the increased mass of the ribose due to the presence of the O2’ and the methylene bridge. This reduces the vibrational wavenumber and increases the vibrational amplitude, but not the length, of the glycosyl bond. Thus, a wavenumber downshift, but no significant change in intensity, would be expected for the (resonance) Raman band(s) assigned to the C1’ –N9 bond. In addition, other vibrational modes, i.e. those coupled with the glycosyl bond, may be subject to small changes. Where such small changes affect closely spaced and unresolved Raman bands, peak narrowing and intensity increases will be expected if the separation between the bands  141 decreases while peak broadening and intensity decreases will be expected if their separation increases.  A final effect is specific to resonance Raman spectroscopy and may arise in consequence of either stacking or the locked ribose (or a combination of both). Here band-selective enhancement of nontotally symmetric modes occurs via the B-term of the Herzberg–Teller expansion using the adiabatic approximation [247]. Thus, selective enhancement is due to vibronic coupling between the resonant electronic state and one or more other excited states. The resonance Raman profiles of bands that are enhanced via vibronic coupling should not correlate well with the analyte’s absorption spectrum while those that are enhanced via the Aterm would.  Taken together, one would therefore expect distinctly different trends to emerge in the spectra due to the effects of altered base stacking vs. varying numbers of locked bases. Stronger base stacking interactions would result in decreased intensities and wavenumber upshifts that are not necessarily proportional to the fraction of locked bases in the molecule. Whereas increasing the number of locked bases in a given strand would result in increased intensities and wavenumber downshifts that are expected to be proportional to the fraction of locked bases in the molecule. Even though this picture could be complicated by enhancement contributions due to vibronic coupling, it does provide a useful framework for interpreting spectral differences corresponding to altered stacking or numbers of locked bases.  142  4.3.1  Evidence of Base Stacking Resonance Raman signal intensities are strongly dependent on the analyte’s electronic  energy levels [240], and since these are modified by stacking, base stacking is expected to be reflected in reduced peak intensities and wavenumber upshifts as outlined above. A reduction in Raman scattering intensities with 257 nm excitation is clearly visible between monomer and dimers when accounting for equal concentrations of bases (Figure 4.4(a)). These modified intensities indicate the presence of base stacking and this is consistent with evidence from other fields. Density functional calculations show that base stacking in single-stranded adenine dimers leads to delocalized excited states with long lifetimes [245] and evidence from f-ESA shows that long-lived excited states already occur in single-stranded adenine stacks as short as dimers (both DNA and LNA, with DNA lifetimes being much longer than those of LNA), but are absent in the monomers [224].  143  Figure 4.4: (a) Resonance Raman excitation with 257 nm light reveals a change in intensity when going from monomer to dimers, attributed to changes in energy levels due to base stacking in the dimers. (b) With 244 nm excitation differences in resonance Raman spectra are revealed that may originate from a combination of different energy levels, vibronic coupling, and solvent interactions.  Figure 4.4(b) shows that the energy levels in the LNA and DNA dimers are different from each other and that they also differ from the LNA monomer. Thus differences in energy levels due to the presence of stacking account for the differences in the UVRRS between monomer and dimers given equal base concentrations. Differences in energy levels due to the type of stacking [224] (DNA vs LNA) account for the differences between the dimers. Included in these differences could be modifications by differences in vibronic coupling and the possible differential interactions of the dimers with the solvent; clearly, the spectrum of LD is not any linear combination of the LM and (dA)2 spectra. Therefore, since the same base (adenine) is involved in all the spectra shown in Figure 4.4, panels (a) and (b), and since the Raman bands  144 enhanced by 257 and 244 nm excitation, respectively, are predominantly due to the in-plane base vibrations, we generally conclude that the differences in the spectra indicate that the electronic energy levels of the base are modified in certain ways in the presence of stacking and modified in different ways in the presence of the locked ribose.  4.3.2  Raman Scattering From the Locked Base Locking the ribose causes intensity increases and frequency downshifts in imidazole ring-  related peaks (Figure 4.5(a); Table 4.2). These changes indicate that the effect of the lock (i.e. methylene bridge and O2’) is to withdraw electrons from the imidazole ring or to modify vibrational modes due to the increased mass of the ribose ring, both as explained above. The specific involvement of the lock in causing these changes is substantiated by the greater downshift in that part of the imidazole ring describing the O4’C1’N9C4 torsion angle χ and forming part of the glycosyl bond (Table 4.2), and closest to the ribose (C4N9, C8N9; ca 1481 cm−1), compared with the part of the imidazole not related to χ (C5N7, N7C8; ca 1337 cm−1). The involvement of the lock is further supported by the appearance of a band that includes the glycosyl bond (N9C1’; ca 1180 cm−1). It is also supported by the sharp reduction in the ca 1210 cm−1 band (and possibly that of the 1380 cm−1 band though the assignment of this band is uncertain, see Table 4.2) that suggests decoupling of the C8H bending vibration or a loss of its vibronic enhancement. The same intensity and frequency changes persist for longer oligomers but are more muted (Figure 4.5(b)). Note that the larger scattering intensity of the 1481 cm−1 band in LM compared to dAMP (i.e. both monomers) cannot be attributed to stacking and in the oligomers, where bases in both LNA and DNA are stacked, the larger scattering intensities in LNA can at most be attributed to different stacking in A-form than B-from structures. We  145 address the differential contributions to the UVRR spectra, by base stacking and by conformational restriction of the ribose, in more detail below.  146  Figure 4.5: Resonance Raman excitation with 257 nm light. (a) Locked adenine monophosphate exhibits increased scattering intensities, wavenumber downshifts (pronounced for the ~1481 cm− 1 band), and other imidazole-related spectral changes in Raman bands indicating that the ribose lock affects the glycosyl bond between ribose and adenine base. The spectra are scaled with regard to the ~1337 cm−1 peak to show wavenumber shifts better. (b) Although toned down, the increased scattering intensities and wavenumber downshifts are also evident in LNA oligomers compared to natural analogs (see also Figure 4.6 and Figure 4.7).  147  Table 4.2: Adenine-related resonance Raman band assignments in nucleic acids, nucleic acid constituents, and their model compounds. ID: dA (deoxynucleoside, deoxyadenosine); rA (nucleoside, adenosine); dAMP (deoxynucleotide, deoxyadenosine monophosphate). Bonds: b = bend; d = deformation/scissors; s = stretch; Pyr = pyrimidine (6-ring); Im = imidazole (5-ring). Comments: w. = with; D = deuterium. cm-1 1173  ID dA  Bonds N7C8+N1C2 s;  Comments  Ref.  257-229 nm  [123]b  -4 w. 1,3- 15N; -2 w. 2-13C; -3 w. C8-D;  [248]  C6NH b 1176  rA  Im + N9C1' s  +31 w. C1'-D 1208  dA  C8H+N7C8 s  257-229 nm  [123]b  1213  rA  C8-H b  -1 w. 1,3- 15N; -1 w. 2-13C; -248 w. C8-D;  [248]  +11 w. C1'-D 1250  dA  N1C2+C8N9 s;  [123]b  257-229 nm  C2H b 1253  rA  Pyr+Im  -5 w. 1,3- 15N; -3 w. 2-13C;  -14 w. C8-D;  [248]  -5 w. C1'-D 1309  rA  Pyr+Im  -10 w. 1,3- 15N; -10 w. 2-13C; -17 w. C8-D;  [248]  +17 w. C1'-D 1310  rA  Pyr  -16 w. 1,3- 15N; -17 w. 2-13C; -2 w. C8-D;  [248]  +1 w. C1'-D 1310  dA  C2N3+C8N9 s; C8H b  -3 in D2O; 257-229 nm  [123]b  148  cm-1  ID  Bonds  Comments  Ref.  1336  dA  C5N7+N7C8 s  +5 in D2O; 257-229 nm  [123]b  1337  rA  Pyr  -10 w. 1,3- 15N; -6 w. 2-13C; -6 w. C8-D;  [248]  +2 w. C1'-D 1339  dAMP -C5N7 (39%) +  192-282 nm  [249]a  257-229 nm  [123]b  N7C8 (12%) 1375  dA  1376  rA  Pyr+Im  -3 w. 1,3- 15N; -3 w. 2-13C; -7 w. C8-D; +4 w. C1'- [248] D  1424  dA  N1C6+C6Ns  +2 in D2O; 257-229 nm  [123]b  1428  rA  Im  -15 w. C8-D; -3 w. C1'-D  [248]  1482  dAMP -C2H d (29%)  192-282 nm  [249]a  257-229 nm  [123]b  -16 w. C8-D; -1 w. C1'-D; others  [248]  +14 in D2O; 257-229 nm  [123]b  -13 w. C8-D; -1 w. C1'-D  [248]  192-282 nm  [249]a  -N9C8 (19%) + C8H d (15%) 1482  dA  C4N9 s + C8H b  1485  rA  Pyr+Im  1506  dA  1508  rA  1580  dAMP C5C4 (48%) -  Im  C4N3 (31%) 1580  dA  C4C5+N3C4 s  -5 in D2O; 257-229 nm  [123]b  1583  rA  Pyr  -2 w. C1'-D  [248]  149  cm-1 1603  ID dA  Bonds  Comments  NH2 b + C5C6 s -420 in D2O; 257-229 nm  Ref. [123]b  + C6N s 1603  rA  Pyr+NH2 d  -1 w. C1'-D  [248]  1650  rA  NH2 d  isotope shifts  [248]  a  4.3.3  From reference [237]. b From reference [237] and others, see referring article.  Local and Stacking Effects of Locked Bases Since some of the changes induced by a modified ribose ring are specific to locked bases,  they are expected to vary in a discrete, stepwise manner reflecting discrete changes in the number of locked bases in an oligomer – an additive effect. They are not expected to vary as a function of the position of the locked base within the primary structure. On the contrary, the effects of locked bases on stacking arrangements may not vary in a discrete, additive manner, but be strongly dependent on their positions relative to other bases and relative to the strand ends. It is known that the position of a locked base in a sequence affects the sugar pucker of neighboring bases [192] thus it could affect base stacking properties as well. Therefore, whether the observed changes generally vary in a discrete or continuous manner may provide a means to discriminate between effects due to the presence of the locked base itself and effects of the locked base on the stacking arrangements of its neighbors.  In Figure 4.6(a) and (b) both peak position and peak amplitude, respectively, for the ~1337 cm−1 Raman band show effects of locked bases (i.e. increasing red shift and intensity). However, these are not stepwise therefore they are not indicative of additive effects. The lack of  150 strong additive effects is consistent with the more distal location of molecular bonds whence scattering originates (see Table 4.2). In Figure 4.6(c) and (d) in contrast, additive effects are clearly observed for the ~1481 cm−1 band resulting from molecular bonds closer to the ribose ring. For the 1580 cm−1 Raman band, assigned only to pyrimidine modes (see Table 4.2), unambiguous stepwise changes are also not observed (data not shown). These results provide further strong proof of the presence of the methylene bridge. They are consistent with our expectations that electron withdrawal by the more oxygen-rich locked ribose via bonds close to the C1’N9 glycosyl bond will lead to wavenumber downshifts and intensity increases in the affected Raman bands. However, we cannot rule out that these results arise from relative shifts in unresolved bands due to the additional mass of the methylene bridge (more below).  151  Figure 4.6: Resonance Raman excitation of pentamers with 257 nm light. (a, b) In locked bases, both peak position and peak amplitude, respectively, for the ~1337 cm−1 Raman peak show effects that are nonadditive or only weakly additive. (c, d) In contrast, strong additive (i.e. stepwise) effects are observed for the ~1481 cm−1 peak resulting from bands arising from molecular bonds closer to the ribose ring.  152 Figure 4.7 reveals that the same general trends occur also in longer oligomers. Some mild stepwise changes are evident in Figure 4.7(a) and (b) for the 1337 cm−1 peak position and amplitude. The most consistent stepwise change is seen for the 1481 cm−1 peak position (Figure 4.7(c) and (d)). At first the data in Figure 4.7(a), compared to that of Figure 4.7(c), may appear more ‘stepwise’, but LNAa and LNAl each has as many locked bases as others with a single locked base (i.e. LNAb to LNAg). Thus, we take the smaller shift in peak position relative to (dA)12 for the 1337 cm−1 band to indicate that a terminal locked base is less disruptive of stacking than a nonterminal locked base. We acknowledge the possibility that due to the asymmetry in the backbone terminals a locked base, when at the 5’-end, may have a different effect than when at the 3’-end, however, our current data (Figure 4.7) do not suggest such an effect. In Figure 4.7(c), the differences in peak shift between oligomers with a terminal locked base, LNAa and LNAl, and other oligomers with a single, but nonterminal, locked base are less pronounced. Nevertheless, it appears that stacking effects are also evident here since the peak shift and intensity are relatively more similar to those of (dA)12 than for the other oligomers with locked bases.  153  Figure 4.7: Resonance Raman excitation of dodecamers with 257 nm light. (a, b) Peak position and peak amplitude, respectively, for the ~1337 cm−1 Raman band and (c, d) for the ~1481 cm−1 band. Additive changes are most pronounced in (c) where a smaller difference occurs between oligomers with a single locked base (LNAa – LNAg) as opposed to (a) where the difference between LNAa and LNAl and other oligomers with a single locked base is more evident. Note in (a) that where a locked base occurs at a strand end (LNAa, LNAl), changes are reduced compared to the no-lock (dA12) condition and about half as much as that of other single-lock oligomers (e.g. LNAg) indicating a stacking mediated influence on the nearest neighbor in addition to the presence of the lock itself.  154  We also infer, from the differences between LNAa and LNAl in which the locked base is in a terminal position and other oligomers with a single, more central locked base, that the neighboring base on both sides is affected. Since LNAa and LNAl have a neighbor only on one side, but other single-lock oligomers have neighbors on both sides, the spectral effect compared to (dA)12 is reduced. Thus, the locked base has a short-range effect on its neighbors, extending only to the nearest or at most the next nearest neighbor. This conclusion receives support from the lack of difference between LNAgj and LNAdj where two or five normal bases separate the locked bases, respectively. Taken together, there is a short-range effect that is probably propagated via locally perturbed stacking interactions.  Although these results confirm that a locked base in a nucleotide sequence has a local, spectroscopically observable trait, it should be noted that the changes for the pentamers are larger (e.g. 1481 cm−1 shift of ~1.2 cm−1 per locked base) than those for the dodecamers (~0.6 cm−1). The reduced effect in longer oligomers should be expected, since a smaller fraction of the total number of bases (9 to 17%) is locked compared to the pentamers (20 to 40%). Therefore we consider the pentamer data to provide the more realistic indication of the effects of a locked base on the UVRR spectrum since they are less modified by other, notably stacking, interactions. At 1580 cm−1 the effects are much weaker – some consistent effect for the presence of a locked base is observed, but clear stepwise effects as in Figure 4.6(c) and (d) are not (data not shown).  Another parameter of the UVRR peaks that is modified is the full-width-at-half maximum (FWHM). The ~1337 cm−1 peak in LM is noticeably narrower than that in dAMP (8.18 ± 0.04 cm−1 and 10.0 ± 0.1 cm−1, respectively). Assuming that there are no aggregates, thus no  155 ‘secondary structure’, in those samples and hence that the same environmental conditions occur for both monomers, narrowing of the 1337 cm−1 peak in the LM sample, and indeed in all LNA samples compared to their analogous DNA samples, is attributed to the shifting of one or more unresolved bands under the ~1337 cm−1 peak. Although the increased overlap between these unresolved bands leads to an increase in peak height, the 1337 cm−1 peak in the LM sample has a ~30% larger area than in the corresponding dAMP sample, indicating that an increase in scattering also occurs. This shifting, in addition to possibly originating from the increased mass of the locked ribose, also may be the result of the rearrangement of electron densities in the adenine base precipitated by the lock (more below). However, the observed increase in scattering argues against an increase in mass being solely responsible for the observed spectral effects due to the presence of the ribose lock, but does not rule out a sole origin for these effects from a charge redistribution due to the lock.  4.3.4  Differential Stacking in DNA and LNA A-DNA and B-DNA have rather different stacking geometries (see Figure 4.3(a) and(b)).  A-DNA has a helical diameter a few Å larger than that of B-DNA, the bases are arranged closer to the perimeter of the helix rather than closer to the center of the helix as in B-DNA, and there is less overlap between bases in A-DNA [219, 220] In B-DNA, evidence suggests that interactions between neighboring adenines can extend to several bases (at least two [215] or 3–4 bases [214]) and theoretical studies indicate that tens of bases could be involved [250]. f-ESA studies from our laboratories point to the occurrence of both – charge transfer interactions involve two bases while Frenkel interactions (Frenkel excitons) involve more than two bases [251]. Although, based on f-ESA, interactions between bases in LNA are also observed, these appear to be more  156 short-lived than corresponding ones in B-DNA and suggest that interactions in LNA are weaker or of a different nature.  The question about the cause of structural order in DNA remains despite knowledge about the structures themselves. Although an overall A-type geometry is present in (fully locked) LNA:DNA hybrids, sugar puckers in the DNA strand are in equilibrium with about 60% C2’-endo pucker [192]. Rapid conversion between C2’-endo and C3’-endo pucker occurs because of similar energy minima and a very low energy barrier between them [195]. If conformation were backbone-driven, then the A-form conformation of LNA should enforce a uniform pucker. Hence, there is no rigid relationship between sugar pucker and conformation. In contrast, a harder π shell, induced through electron withdrawal [213] by the locked ribose, may lead to mutual repulsion by the imidazole moieties of adjacent bases. The imidazole ring is already relatively electron rich [197] and electron withdrawal from the ribose ring, as we infer to have observed here, changes the charge distribution on the imidazole ring, the dipole moment of the base, as well as the base’s transition moments. Increased negative charges in the imidazole rings favor a stacking realignment where imidazole–imidazole overlap is further reduced leading to the A-form configuration. Hence, A-from is favored to accommodate both a greater distance between adenine imidazole moieties and short, rigid backbone segments (e.g. the ‘beads on a chain’ model). Repulsion increases the accessible conformational states available in very short oligomers where they may rotate so that the angle between them can be up to 180°; but as oligomers become longer, the conformational freedom of mutually repulsing bases becomes more restricted. In addition, since the pyrimidine moiety is expected to be relatively electronpoor, some association between the imidazole of one adenine base and the pyrimidine of the adjacent base may occur, thus further influencing the evolving secondary structure. A possible  157 test of this interpretation may be provided, by simulation or experiment, through substitution of the C8-hydrogen in DNA with a more electronegative atom to mimic the electron-withdrawal effect of the locked ribose and its implications for the generation of secondary structure in NAs. Theoretical studies, especially concerning electronic distributions and dipole moments, as yet unavailable for LNAs, also will be valuable in this regard. The alternative view is that the backbone conformation is rigid and determines the overall conformation [217].  4.4  Conclusions We have shown that changes in the glycosyl bond when locking a ribose in the C3’-endo  position can be observed with fiber optic-based UVRRS. These changes manifest as stepwise wavenumber downshifts, especially of the 1481 cm−1 band that incorporates part of the glycosyl bond, proportional to the fraction of locked bases in the oligomer. Differences between oligomers where the locked base is situated at the strand terminal and those where the lock is more centrally located indicate that the influence of a locked base extends to, at most, the nextnearest neighbor. This effect and other spectral differences, that occur compared to the monomer when introducing one or more other bases, indicate the presence of base stacking in singlestranded NAs, consistent with other reports [197, 198]. Stacking effects in B-DNA have an aggregate character [252] while in LNA imidazole–imidazole and pyrimidine–pyrimidine interactions are deduced to be repulsive and imidazole–pyrimidine interactions are deduced to be attractive. Taken together, stacking effects are credited with ultimately determining NA conformation in the solution phase while the backbone perhaps serves to maintain structural order in the ground and excited states as well as base sequence, and thus information content.  158  4.5  Acknowledgements We thank Curtis Hughesman and Yee Chee Lim for help in preparing the samples.  Instrumentation and infrastructure were provided by the UBC Laboratory for Advanced Spectroscopy and Imaging Research (LASIR) and Laboratory for Molecular Biophysics (LMB). Funding was provided by the Natural Sciences and Engineering Research Council (NSERC), the Canadian Institutes of Health Research (CIHR), the Michael Smith Foundation for Health Research (MSFHR), the UBC Centre for Blood Research, the Canada Foundation for Innovation (CFI) and the British Columbia Knowledge Development Fund (BCKDF).  159  Chapter Five: Residual Contamination Observed in Synthetic Oligonucleotides2  5.1  Introduction Synthetic oligonucleotides are routinely used in a broad range of applications in  molecular biology, such as primers for polymerase chain reaction, probes for northern/Southern blotting and in situ hybridization, DNA sequencing, site-directed mutagenesis and microarray technology, to name just a few, and a variety of optical spectroscopic methods have employed oligonucleotides, either as probes or as model analytes for biophysical investigations. In particular, ultraviolet resonance Raman spectroscopy (UVRRS) has been used in many studies involving oligonucleotides to take advantage of the increased signal-to-noise ratios, reduced fluorescence backgrounds and the ability to selectively enhance certain molecular vibrations by tuning the excitation wavelength to overlap with electronic transitions of the constituent bases [38, 40, 54, 123, 236, 253]. For example, we recently employed UVRRS to investigate conformational differences between locked nucleic acid (LNA) and DNA oligomers with identical base sequences [66]. It was during the course of this work that we became aware of some important features of commercial oligonucleotide preparations that could potentially impact a variety of other spectroscopic investigations involving synthetic oligonucleotides.  2  A version of this chapter has been published:  Christopher J. Addison, Stanislav O. Konorov, H. Georg Schulze, Robin F. B. Turner and Michael W. Blades (2011). Residual benzamide contamination in synthetic oligonucleotides observed using UV resonance Raman spectroscopy. Journal of Raman Spectroscopy. 42(3), 349-354.  160 An important step in the standard phosphoramdite nucleotide synthesis process [254, 255] is the removal of protecting groups attached to the exocyclic amines of the phosphoramidite bases to prevent any unwanted side reactions from taking place at these sites. In particular, 2’deoxyadenosine (dA) and 2’-deoxycytidine (dC) utilize a benzoyl-protecting group at the N6 and N4 positions, respectively (Figure 5.1). After synthesis, the protecting groups are removed with an ammonium hydroxide treatment to yield the desired oligomer, plus a mixture of truncated oligomer failure sequences, protecting group by-products (benzamide) and silicates from the hydrolysis of the glass supports. Most oligonucleotide producers indicate that the protecting group by-products should be removed as organic salts during the standard desalting purification step. Small amounts that remain are not likely to cause problems in most bioanalytical applications. However, some residual contaminants can be problematic in some applications, and it is important to be aware of their potential implications.  161  Figure 5.1: Protected 2’-deoxynucleosides used in the phosphoramidite synthesis. (a) 2’deoxyadenosine with a benzoyl protecting group at the N6 position and (b) 2’deoxycytidine with a benzoyl protecting group at the N4 position.  The protecting group in  each nucleoside is shown in red. The nucleosides are shown without the 2’-deoxyribose moiety for clarity.  More importantly, in a broader context, in-depth knowledge about the purity of chemical reagents and other components can be critical to explain chemical reactivity or spectroscopy that is observed. A recent report [256] has shown that the observation of a non-transition-metal Suzuki-type coupling was in fact due to trace-level palladium contamination in commercially available sodium carbonate. In addition, catalyzed N-, O- and C-arylation reactions initially  162 thought to be from ferric chloride are now attributed to trace-level copper contamination in the ferric chloride reagent [257]. Indeed, in the latter paper, the authors demonstrated that the observed reactivity varies dramatically with reagents from different manufacturers, suggesting that the extent of contamination is source dependent.  Markovitsi et al. [244] pointed out that pitfalls in experimental protocols are often not explicitly mentioned in the literature, making comparisons between studies on synthetic model DNA oligomers from various groups, and inferences drawn from them, difficult and inconsistent. The purpose of this work is to similarly report how a by-product of the oligonucleotide synthesis that is not always removed under standard desalting conditions leads to complications in UVRRS studies of synthetic polyadenylic oligomers. Due to the resonance enhancement process, the spectrum of this contaminant strongly interferes with the Raman spectrum of DNA and LNA oligomers under certain excitation regimens. Further complications can arise from higher levels of contamination in DNA compared to LNA samples with the same sequence, and possibly between samples of different sequences, as well as a differential electronic enhancement of the contaminant by DNA compared to LNA. Although our focus is mainly on the presence of a contaminant in synthetic oligomer samples and its effects on interpretation of their UVRR spectra, we also note potential implications of contamination for other techniques.  5.2  Experimental DNA and/or LNA oligomer samples were obtained from IDT Technologies (Coralville,  IA), Operon Technologies (Huntsville, AL) and Invitrogen (Burlington, ON). The DNA sample sequences were 5’-AAA AAA AAA AAA-3’ and 5’-AAAAA-3’; the LNA sequences were 5’-  163 AAA-AAA-AAA-AAA-3’ and 5’-AAAAA-3’ (the locked nucleotide is bold-underlined). Samples were purified by using either (1) standard desalting or (2) standard desalting followed by high-performance liquid chromatography (HPLC). Benzamide, phosphate buffer, NaCl and EDTA were all obtained from Sigma-Aldrich (Oakville, ON) and used without further purification. Benzamide standard solutions were prepared by dissolving benzamide in phosphate buffer (as described below) to yield a concentration of 2 mM. Oligomers were centrifuged prior to use and then dissolved to a sample concentration of 60 µM in a buffer solution consisting of 10 mM phosphate buffer (pH 7.0), 1 M NaCl, 1 mM EDTA and 25 mM KNO3. Oligomer concentrations were quantified by UV-Vis spectroscopy using the 260-nm absorption band of the aromatic bases.  UV laser excitation light was generated using a frequency-doubled Ar+ ion laser (Innova 90C FreD, Coherent Inc., Santa Clara, CA). Excitation light delivery and collection of the Raman scattered signal was accomplished using a custom fiber-optic probe designed for UVRR spectroscopy as previously described [46-48]. Laser power delivered to the samples was 20 mW, and samples were stored in a 500-µl centrifuge tube. The sample tube was placed in a rotating sample holder and constantly spun during measurement to ensure that fresh sample was continuously replenished in the analytical volume to minimize any sample photodegradation. The Raman scattered light was passed through a dielectric interference filter (Barr Associates Inc., Westford, MA) to remove Rayleigh scattered excitation light. The Raman scattered light was then dispersed using a 1.0-m focal length monochromator (Model 2061, McPherson Inc., Chelmsford, MA) equipped with a 3600 grove per mm holographic grating (Model 8358-1004-0, McPherson Inc.) and detected using a liquid-nitrogen-cooled CCD detector operating at -120 °C (Spec-10 400B, Roper Scientific, Trenton, NJ). Acquisitions of 10-s duration were obtained for  164 each sample, and sequential spectra were averaged only if no significant changes were observed during the UVRR experiment. All spectra were normalized to the 1050 cm-1 band of the NO3internal standard, prior to computing the difference spectra. The benzamide UV-Vis absorption spectrum between 400 and 200 nm was obtained at 20 °C with a double-beam spectrophotometer (Cary 1E, Varian, Palo Alto, CA) scanning at a rate of 180 nm min-1 with 0.5-nm intervals.  5.3  Results and Discussion Due to the minor resonance enhancement of the benzamide peak near 1609 cm-1 using  257 nm excitation (pentamers, Figure 5.2(a)), a small augmentation of the peak was observed in DNA relative to LNA at 1609 cm-1. This was at first not recognized as resulting from an impurity. The latter band is virtually superimposed on, and originally was attributed to, the NH2 scissoring vibrational mode in adenine ~ 1604 cm-1 [123, 234, 258, 259]. Therefore, the small differences we observed between DNA and LNA at 1609 cm-1 with 257-nm excitation (pentamers, Figure 5.2(a)) were of interest, but initially not cause for suspicion, due to the fact that the DNA helix is in B-form and the LNA helix in A-form.  165  Figure 5.2: (a) Raman spectra of standard desalted DNA and LNA and HPLC-purified DNA pentamers purchased from IDT technologies. Raman spectra were obtained with 257 nm excitation. Note the small but distinct differences in the 1609 cm−1 band between all three samples. (b) With excitation at 244 nm, the 1609 cm−1 band in standard desalted DNA is dramatically enhanced, that of LNA less so, while it is absent in HPLC-purified DNA. The approximate enhancement of this band in DNA relative to LNA is 1.5.  166  It is common in UVRRS studies of DNA to excite near the its absorption maximum at 260 nm [233, 234, 236, 248, 253, 258], but resonance Raman enhancement profiles (RREPs) can provide additional information about, and discrimination between, chromophores [249, 260] and we proceeded to measure the RREPs for both DNA and LNA. We observed a strong enhancement of the 1609 cm-1 band with 244-nm excitation, both in water (pentamers, Figure 5.2(b)) and D2O (data not shown) solvents, which made us question the putative assignment of this band and consider the possibility of contamination. The results suggested that, if a contaminant were present, resonance enhancement varied strongly between 257 and 244 nm. Importantly, neither mass spectra from the manufacturer (data not shown), nor gel electrophoresis performed in-house (data not shown), nor the (DNA - LNA) Raman difference spectra at 257 nm (data not shown) provided obvious evidence of a contaminant. However, discussions with the manufacturer led us to suspect the possibility that the spectral differences we observed were the result of trace amounts of benzamide.  The UVRRS spectrum, at 244 nm excitation, of benzamide is shown in Figure 5.3 (inset shows its structural formula). Of note is the pronounced vibration at 1609 cm-1, attributed to a ring breathing mode [135, 261] (υ9 according to the notation of Palomar [262]). The υ11 NH2 scissoring mode has been observed at 1572 cm-1 by others[263], but in Figure 5.3 is likely obscured by the much stronger peak at 1609 cm-1. A combination of Ph-CONH2 stretch and ring stretch (υ14) is observed at 1411 cm-1, and a weak band at 1450 cm-1, not completely resolved from the υ14 stretch, is a CCH in-plane bending vibration (υ13) [263]. An additional weak vibration at 1500 cm-1 is a ring stretching mode with some CCH in-plane bending (υ12). Bands at  167 1186 and 1141 cm-1 are CCH in-plane bending modes (υ17 and υ18, respectively), and 1005 cm-1 is a ring trigonal bend mode (υ23) [263].  Figure 5.3: The Raman spectrum of benzamide superimposed on the Raman difference spectrum of DNA–LNA purified by standard desalting only. The Raman difference spectrum of DNA – LNA purified by standard desalting followed by HPLC is shown offset (lower). Identification of benzamide (inset shows structure) as the contaminant in standard desalting samples was due to differential UVRRS enhancement and contamination of benzamide in isosequential dodecamers. With excitation at 244 nm, the distortion created by the benzamide contaminant in the difference spectrum of the samples purified by standard desalting contrasts strongly with the difference spectrum of the same samples in the absence of contaminant (standard desalting followed by HPLC). Oligomers were purchased from IDT technologies.  168 Since the initial UVRR studies involving DNA were motivated by structure/spectra relationships for DNA and LNA strands [66], difference spectra (DNA - LNA) for DNA and LNA dodecamers were obtained. Both the difference spectra for dodecamers, which had been purified using standard desalting only, as well as those purified with standard desalting followed by HPLC are shown in Figure 5.3, the latter offset for clarity. Positive peaks suggest features that are only observed in DNA, and negative peaks indicate features that are only observed in LNA samples. Positive peaks in the difference spectrum were observed at 1000, 1140, 1280, 1420 and 1510 cm-1 and a strong feature at 1609 cm-1. These peaks are all observed in the Raman spectrum of benzamide and strongly suggested that the DNA oligomer samples were contaminated with benzamide even after standard desalting purification. With excitation at 244 nm, the distortion in the difference spectrum of the samples in the presence of the benzamide contaminant contrasts strongly with the difference spectrum of the same samples in the absence of contaminant.  Further evidence came from the RREP for benzamide obtained at varying wavelengths in the UV region (Figure 5.4). The RREP shows that benzamide yields little Raman enhancement at 257 nm but, as the excitation wavelength is moved deeper into the UV region, the 1609 cm-1 benzamide band becomes significantly more enhanced, with maximum enhancement occurring using 238 nm excitation. This is consistent with the UV-Vis spectrum of benzamide (Figure 5.5), which has a maximum absorption at 226 nm [263].  169  Figure 5.4: The resonance Raman excitation profile of the most prominent benzamide band (1609 cm−1) indicating that it is already considerably enhanced with 244 nm, compared to 257 nm, excitation. Benzamide is a potential contaminant of synthetic oligomers, and UVRRS with excitation near the DNA absorption maximum ca 260 nm may not reveal its presence in such samples, whereas excitation deeper in the UV does (see also Figure 5.2 and Figure 5.5).  The UV-Vis difference spectrum between contaminated and non-contaminated DNA pentamers is virtually identical to that of benzamide (Figure 5.5). Since the former spectrum was scaled 12.4 times to coincide with the latter, we estimate the concentration of benzamide in this desalted pentamer to be roughly 150 µM. In 60 µM pentamer concentrations, the concentration of adenine base is 300 µM. Thus, the ratio of base to contaminant is ~ 2:1. Note that this is based on a single estimate from a single sample (we do not know, nor have we tried to assess) the sample-to-sample variations in contaminant. The UV-Vis difference spectrum of LNA pentamers  170 shows an absorption maximum coincident with that of benzamide, as well as an absorption maximum at 260 nm that could be attributed to the presence of some failed oligomer sequences. It is interesting that the DNA sample did not show evidence of failed oligomer sequences. It is also interesting that the same contaminated DNA pentamer (i.e. no change in the concentration of contaminant), when hybridized with a complementary LNA strand, showed reduced enhancement of the 1609 cm-1 band compared to when it was hybridized with a complementary DNA strand (data not shown). Neither complementary strand showed evidence of contamination. The data are insufficient to conclusively explain these differences. However, we could hypothesize that benzamide might be present in a combination of free, backbone-bound, basebound and base-intercalated forms and these combinations may differ between DNA and LNA. Furthermore, such forms and their combinations could also be influenced by the parameters of the manufacturing process. It is also possible that the different electronic structure of LNA [224, 251] compared to DNA, may affect the resonance enhancement of bound or otherwise associated benzamide.  171  Figure 5.5: The UV absorption spectrum of 2 mM benzamide showing absorption maxima near 226 and 268 nm. The UV absorption spectrum confirms that UVRRS with excitation near the absorption maximum for DNA and a minimum for benzamide (ca 260 nm) would not provide strong evidence for its presence in synthetic oligomer samples, whereas excitation deeper in the UV would (see also Figure 5.2, Figure 5.4 and Figure 5.6). The DNA and LNA difference spectra (pentamers, standard desalting minus HPLCpurified, scaled by 12.4) are superimposed on the benzamide absorption spectrum for comparison. The DNA difference spectrum is virtually identical to that of benzamide after scaling and exhibits a red shift in the long wavelength absorption maximum, suggesting that the benzamide is hydrogen-bonded to the oligomer. The LNA difference spectrum shows a weak presence of benzamide as well as an absorption maximum at 260 nm that we attribute to failed oligomer sequences. Oligomer samples obtained from IDT technologies.  172 Benzamide contamination was observed in oligomers purchased from all three manufacturers examined as part of this study. Interestingly, the extent of contamination, estimated from the intensity of the benzamide peaks in the Raman difference spectra obtained with 244 nm excitation, varies greatly between manufacturers (Figure 5.6). Oligomers purchased from Operon Technologies (Figure 5.6(a)) and IDT exhibited pronounced benzamide contamination that could be readily observed in the parent spectra, while those purchased from Invitrogen (Figure 5.6(b)) yielded benzamide peaks that could be observed only when the difference spectrum was closely examined. The fact that the extent of contamination can vary so greatly among manufacturers could in turn adversely affect the direct comparison of data obtained using oligomers purchased from different manufacturers.  173  Figure 5.6: Raman spectra of DNA in varying purification states with 244-nm excitation. (a) Parent spectra of DNA obtained from Operon Technologies. DNA purified using standard desalting or HPLC (top panel); Raman difference spectrum (bottom panel). Prominent positive peaks here are attributed to benzamide contamination. (b) Parent spectra of DNA obtained from Invitrogen. DNA purified using standard desalting or HPLC (top panel); Raman difference spectrum (bottom panel). Benzamide contamination is observed based on the Raman difference spectrum, but not to the same extent as observed in (a).  Previous research [123, 234, 258, 259] has assigned the weak vibration near 1600 cm-1 to an -NH2 scissoring. However, any observed vibration in this region can be attributed to benzamide contamination if the samples were purchased commercially and purified using only standard desalting. Therefore, the -NH2 scissoring assignment may need to be re-examined  174 because HPLC-purified oligomers do not show any sign of the -NH2 scissoring vibration (Figure 5.2(b), Figure 5.6(a) and Figure 5.6(b)) at 244 or 257 nm excitation.  For implications of DNA oligomers beyond UVRRS, we noted the following: The UVVis spectrum of benzamide has a second absorption maximum near 268 nm; thus some potential exists for misinterpretation if benzamide is present and the absorption band at 260 nm used for DNA quantification or for physical studies. Quantification errors would affect all researchers reliant upon accurate quantification in their work. In deoxypolyadenylic acid, however, quantification error is possibly negligible. Figure 5.5 shows the UV-Vis spectrum of a DNA pentamer after standard desalting minus the UV-Vis spectrum of the same DNA pentamer after HPLC purification. Although the major absorption peak at 226 nm does not seem to be affected, there is a ~5-nm bathochromic shift in the lower energy absorption peak of benzamide. Thus, the reduction in absorption at 260 nm may come about as a consequence of this shift, leading to reduced overlap between the 226 and 268 nm absorption maxima. Our inference about quantification is therefore based on the resulting near absence of any absorption at 260 nm.  Even though quantification errors may be negligible, the difference spectrum (Figure 5.5) gives rise to other concerns. Given that red shifts are indicative of π → π* transitions [264], the red shift in the 268 nm absorption band probably indicates the S0 → S2 or π (benzene) → π* (C=O-NH2) transition [263], implying redistribution of the electron density away from the ring toward the distal parts of the molecule where they become available to participate in hydrogen bonding [264]. Thus, we infer that the benzamide is forming hydrogen bonds with the oligomer. Another possibility is binding to the backbone, but inconsistent UVRRS intensities of the 1609 cm-1 peak between oligomer samples with different sequences (data not shown) suggest that this  175 is not the case and, moreover, there may be sequence-dependent effects. We consider benzamide binding to the base as more likely. Indeed, its hydrogen donor and acceptor positions (see Figure 5.3, inset) closely mimic those of the complementary base, thymine. Taken together, the situation arises where quantification of oligomers may be accurate, but analytical measurements dependent on hydrogen bonding, such as calorimetric measurements and kinetic studies, may be affected.  5.4  Conclusions Benzamide is used as a protecting group in the phosphoramidite synthesis of  oligonucleotides. While standard desalting used after synthesis to purify the preparation removes any species that may interfere with most assay reactions or other routine uses of the oligomer, it does not completely remove benzamide. The 1609 cm-1 band of residual benzamide contamination can interfere with nucleic acid bands, especially when using excitation wavelengths near 244 nm, where the benzamide band is strongly enhanced. In particular, the 1609 cm-1 band of benzamide could obscure (or be mistaken as) a weak vibration previously attributed to an NH2 scissoring. UV-Vis absorption spectra suggest that the benzamide contaminant is hydrogen-bonded to the oligonucleotides. Interestingly, the UVRRS interference by benzamide contamination varies between DNA and LNA oligonucleotides. Further purification using HPLC is advised for spectroscopic work involving DNA and LNA oligonucleotides as well as any applications that require highly accurate quantitative measurements.  176  5.5  Acknowledgements We wish to acknowledge assistance from IDT Technologies in response to our enquiries  about possible contaminants. Financial support for this work was provided by the Natural Sciences and Engineering Research Council of Canada (NSERC), the Canada Foundation for Innovation, the British Columbia Knowledge Development Fund and the Michael Smith Foundation for Health Research. CJA was partially supported by a fellowship provided by NSERC.  177  Chapter Six:  Conclusions  The overall theme of this work was to employ fiber-optic UV resonance Raman spectroscopy for investigating biomolecular structure. FO-UVRRS was utilized to provide key structural insights in systems involving enzyme-substrate structure and olignucleotide structure involving locked nucleic acids. In addition, because of the resonance enhancement process, previously unreported contamination of synthetic DNA oligonucleotides was discovered.  The chapter begins with a summary of the important findings arising from each of the research chapters, in order to emphasize contributions to their respective fields.  6.1  UV Resonance Raman Investigation of Metal-Containing Dioxygenases The ability of FO-UVRRS to discriminate between protonation states of an enzyme-  bound substrate provided key insights into proposed mechanisms of the HPCD and MndD enzyme set. FO-UVRRS was used to probe the substrate structure in the HPCD:HPCA complex containing the native (Fe2+) and non-native (Mn2+ and Co2+) metals incorporated into the enzyme. The enzyme-substrate structure of a homologous enzyme (MndD) with HPCA was also examined. MndD is particularly interesting because of its high sequence identity to HPCD, yet it naturally incorporates Mn2+ into the enzyme active site instead of Fe2+. Both the native (MnMndD) and non-native (Fe-MndD) HPCA complexes were probed.  178 Initial experiments were performed with the free substrate (HPCA) in both H2O and D2O solvents in all four possible protonation states. Key spectral features for each of the protonation states, along with the corresponding H/D isotopic substitution shifts were identified. In this manner, a spectral library for each protonation state was obtained and this information was used to identify the substrate protonation state in the anaerobic enzyme-substrate complex.  Model compounds were synthesized in order to examine spectroscopic changes in the HPCA diagnostic bands resulting from formation of the metal-HPCA complex. The synthetic methods resulted in complexes containing a Fe3+/Mn3+ metal center with a trianionic HPCA ligand. While this did not correspond to the expected protonation state and metal oxidation state within the enzyme-substrate active site, it nonetheless provided general information about changes in the Raman spectrum of HPCA upon binding to the metal center. In general, UVRR spectroscopy of the model complexes indicated an upshift in HPCA bands within the diagnostic 1500-1700 cm-1 region.  Comparison of the enzyme-bound UVRR data of the substrate with that for the free substrates and the model complexes strongly suggests that the enzyme-bound substrate exists in the dianionic form. Identification was based on a number of factors, including the absolute location of each peak within the Raman spectrum, along with observed H/D isotopic shifts that precludes the trianionic protonation state. These observations are consistent with other extradiol dioxygenase enzymes which have been previously studied.  These results provide confirmation of the mechanism proposed by Emerson et al. [73] in which binding of the HPCA substrate to HPCD or MndD occurs in the dianionic form with one  179 of the catechol –OH moieties deprotonated. The results obtained here also demonstrate that incorporation of a non-physiological metal does not perturb the substrate protonation state. This supports kinetic data obtained by Emerson [73] in which incorporation of the non-physiological metal resulted in no substantial changes observed in the kinetics or structural properties of the enzyme.  6.2  FO-UVRRS Study of a Carbon-Carbon Bond Hydrolase Fiber-optic UV resonance Raman spectroscopy was used to probe the substrate binding  mode to BphD, an enzyme in the biphenyl catabolic pathway. BphD catalyzes the hydrolysis of 2-hydroxy-6-oxo-6-phenylhexa-2,4-dienoic acid (HOPDA) and involves an unusual carboncarbon bond fission. Enol-keto tautomerization has been proposed to precede the reaction, with early studies suggested that tautomerization was promoted by binding the substrate in a strained, non-planar fashion. Observation of the enzyme intermediate using UVRR provided valuable insight into the nature of the reaction.  Two mutant enzymes, S112A and S112A/H265A, were used in this study. Both enzymes are hydrolytically impaired and can be used to effectively trap the substrate in the enzyme active site. In addition, an inhibitory substrate (3-Chloro HOPDA; 3CH) was used to probe the nature of substrate inhibition.  The Raman spectrum of the bound substrate in the S112A:HOPDA complex was compared to that for the free HOPDA substrate. The enzyme-bound substrate retained the major features of the free substrate, along with a new feature observed at 1613 cm-1. This new feature  180 is attributed to twisting about the C2=C3 and C4=C5 double bonds in the enzyme active site, resulting in a decoupling of the vibrational modes.  On the other hand, the Raman spectrum of the bound substrate in the S112A/H265A:HOPDA complex corresponded to that for unbound HOPDA. This data suggested that the substrate is bound to the double mutant enzyme as the enolate with no twisting about the HOPDA double bonds. Similarly, the Raman spectrum of the inhibitory 3CH substrate bound to S112A also exhibits peaks corresponding to the free substrate with no vibrational decoupling.  Taken together, the results indicate that His265 assists in the enzymatic reaction by inducing strain the C2=C3 and C4=C5 bonds, while Ser112 catalyzes the ketonization step. The results are consistent with a previously-proposed hypothesis by Bugg [159, 160] that this class of enzyme promotes tautomerization by destabilizing the intermediate through double bond strain. These results are suggest that the inability to catalyze the reaction of 3CH arises from the enzymes inability to impart double bond strain in the compound [163], impeding the ketonization process.  6.3  FO-UVRRS of Locked Single-Stranded Oligonucleotides A fundamental study of the locked-nucleic acid analogues of DNA was undertaken using  FO-UVRRS, with the subsequent paper being the first reported result of such inquiries. FOUVRRS made fundamental and key structural insights into differences in nucleic acid structure between DNA and the corresponding LNA oligomer containing one or more locked bases.  181  The UVRR results demonstrated that incorporation of a locked base in a nucleic acid oligomer induces wavenumber downshifts and peak amplitude changes for numerous Raman bands, most notably the 1481 cm-1 vibration. A portion of this vibration is due to the glycosyl bond, and suggests a conformational change in the glycosyl moiety between the backbone and the LNA base. Peak intensity and peak location changes occurred in a step-wise fashion, based on the number of locked bases incorporated into the nucleotide.  The extent of base stacking within the oligonucleotide was examined using UVRR. A comparison of monomer and dimer UVRR data showed a clear reduction in peak intensities and wavenumber upshifts, attributed to base stacking within the dimer. An examination of comparable dimers with and without a locked base suggested changes in the electronic energy levels of the base occurs in the presence of base stacking.  Furthermore, the UVRR results demonstrated that locking of the ribose sugar induces Raman band intensity increases and frequency downshifts for bands related to the imidazole ring. This was attributed to electron withdrawal out of the imidazole ring resulting from incorporation of the oxygen-rich locked ribose. This was substantiated by greater downshifts being observed for vibrations in portions of the imidazole ring closest to the glycosyl bond.  The electronic redistribution following incorporation of the locked ribose sugar was shown to have implications on base stacking of single-stranded nucleic acids. Because of increased negative charge within the imidazole ring, LNA tended to adopt an A-form configuration in which the imidazole-imidazole and pyrimidine-pyrimidine base-stacking  182 interactions are minimized, while imidazole-pyrimidine attraction are maximized. On the other hand, DNA tends to adopt a B-form configuration where stacking interactions between neighbouring adenines can extend over several bases. This was also supported by previous reports which suggested that interactions in LNA are weaker than in DNA.  DNA/LNA secondary structure was also shown to be influenced by the incorporation of a locked base into the oligomer. Stacking arrangements were shown to be strongly dependent upon the location of the locked base within the oligomer, where a LNA oligomer with a terminal locked base showed reduced disruption of stacking arrangements compared to an LNA oligomer with a non-terminal locked base. In turn, the results also suggested that the locked base has a short-range effect on its neighbour, extending out by one or two bases.  6.4  Residual Contamination Observed in Synthetic Oligonucleotides The fundamental comparative study of DNA and LNA oligomers also identified residual  contamination of some nucleic acid oligomers with a protecting group reagent (benzamide) resulting from the phosphoramadite synthetic method used for oligonucleotide synthesis. Due to the selectivity afforded to FO-UVRRS, the contamination was not readily observed at wavelengths near the absorption maximum of DNA (ca. 260 nm), but excitation deeper into the UV increased the resonance enhancement of benzamide. At excitation wavelengths near 248 nm, spectroscopic signatures from benzamide were readily observed. An intense benzamide ring mode was observed near 1609 cm-1, which overlaps with an –NH2 scissors mode which has been previously assigned in the literature. These overlapping bands, and the possibility of  183 contamination in commercially-obtained oligomer samples, means that great caution must be used when important research results are based upon this –NH2 band.  Correlation of the UVRR and UV-Vis results suggested that hydrogen bonding between benzamide and the DNA oligomer is occurring, based on a red-shift of the 268 nm electronic absorption. The red-shift of this band suggests redistribution of the electron density towards the distal portions of the molecule involved in hydrogen bonding.  Overall, these results suggest that researchers must be careful to test for benzamide contamination in commercially-prepared oligonucleotide samples. Benzamide contamination could result in erroneous measurements in a host of techniques and types of studies, such as calorimetric measurements, kinetic studies, or other spectroscopic measurements.  6.5  Summary of Findings and Overall Impact of the Work The results presented in this work demonstrated the utility of FO-UVRRS for the  determination of biomolecular structure. Because of the large signal enhancement afforded by the resonance process, and the subsequent selective enhancement of certain vibrational bands, FO-UVRRS was used to study molecular structure in complex biologically-relevant samples.  FO-UVRRS was used to study substrate structure when bound within an enzyme active site, and the vibrational data was consistent with results obtained from UV-Vis spectroscopy. The technique was also used to study overall nucleic acid structure involving non-physiological nucleotides and provided important insights into the structural effects resulting from this  184 nucleotide incorporation. The power of FO-UVRRS was also demonstrated by the observation of residual contamination within commercially-produced oligonucleotides, which had been previously unreported. In all cases, FO-UVRRS provided structural information that was either not available from other techniques, or was complementary to information available from other techniques.  As such, FO-UVRRS represents a valuable and important technique for the elucidation of biomolecular structure in a wide variety of biologically-relevant environments. When used in conjunction with other analytical techniques, FO-UVRRS is capable of making important contributions to fundamental research problems within the field.  185  References 1. 2. 3. 4. 5. 6. 7.  8. 9. 10.  11.  12. 13.  14. 15.  16.  17.  Pavia, D.L., Lampman, G.M., and Kriz, G.S., Introduction to Spectroscopy: A Guide for Students of Organic Chemistry. 2nd ed. 1996, Toronto: Saunders Golden Sunburst Series. Asher, S.A., UV resonance Raman spectroscopy for analytical, physical, and biophysical chemistry. Part 2. Analytical Chemistry, 1993. 65: p. 201A-210A. Asher, S.A., UV resonance Raman spectroscopy for analytical, physical, and biophysical chemistry. Part 1. Analytical Chemistry, 1993. 65: p. 59A-66A. Raman, C.V. and Krishnan, K.S., A new type of Secondary Radiation. Nature (London, United Kingdom), 1928. 121: p. 501-502. McCreery, R.L., Raman Spectroscopy for Chemical Analysis. Chemical Analysis, ed. J.D. Winefordner. Vol. 157. 2000, Toronto: Wiley-Interscience. 420. Long, D.A., The Raman Effect. 2002, Toronto: John Wiley & Sons. Addison, C.J., Fabrication, Characterization and Optical Properties of Threedimensional Colloidal Gold Nanostructures, in Department of Chemistry. 2005, University of Victoria: Victoria, BC, Canada. p. 182. Long, D.A., Raman Spectroscopy. 1977, Toronto: McGraw-Hill International Book Company. Ferraro, J.R. and Nakamoto, K., Introductory Raman Spectroscopy. 1994, Toronto: Academic Press, Inc. Kneipp, K., Kneipp, H., Itzkan, I., Dasari, R.R., and Feld, M.S., Ultrasensitive chemical analysis by Raman spectroscopy. Chemical Reviews (Washington, D. C.), 1999. 99: p. 2957-2975. Raether, H., Surface Plasmons: On Smooth and Rough Surfaces and on Gratings. Springer Tracts in Modern Physics, ed. G. Hohler. Vol. 111. 1988, New York: SpringerVerlag. Grabar, K.C., Freeman, R.G., Hommer, M.B., and Natan, M.J., Preparation and Characterization of Au Colloid Monolayers. Analytical Chemistry, 1995. 67: p. 735-43. Grabar, K.C., Smith, P.C., Musick, M.D., Davis, J.A., Walter, D.G., Jackson, M.A., Guthrie, A.P., and Natan, M.J., Kinetic Control of Interparticle Spacing in Au ColloidBased Surfaces: Rational Nanometer-Scale Architecture. Journal of the American Chemical Society, 1996. 118: p. 1148-53. Brolo, A.G., Arctander, E., Gordon, R., Leathem, B., and Kavanagh, K.L., NanoholeEnhanced Raman Scattering. Nano Letters, 2004. 4: p. 2015-2018. Van Duyne, R.P., Hulteen, j.C., and Treichel, D.A., Atomic force microscopy and surface-enhanced Raman spectroscopy. I. Silver island films and silver film over polymer nanosphere surfaces supported on glass. Journal of Chemical Physics, 1993. 99: p. 210115. Halteen, J.C. and Van Duyne, R.P., Nanosphere lithography: A materials general fabrication process for periodic particle array surfaces. Journal of Vacuum Science & Technology, A: Vacuum, Surfaces, and Films, 1995. 13: p. 1553-8. Menezes, J.W., Ferreira, J., Santos, M.J.L., Cescato, L., and Brolo, A.G., Large-Area Fabrication of Periodic Arrays of Nanoholes in Metal Films and Their Application in Biosensing and Plasmonic-Enhanced Photovoltaics. Advanced Functional Materials, 2010. 20: p. 3918-3924.  186 18. 19. 20. 21. 22. 23. 24.  25.  26. 27.  28. 29.  30.  31. 32. 33.  34.  35.  Campion, A. and Kambhampati, P., Surface-enhanced Raman scattering. Chemical Society Reviews, 1998. 27: p. 241-250. Spiro, T.G., Resonance Raman spectroscopy. New structure probe for biological chromophores. Accounts of Chemical Research, 1974. 7: p. 339-44. Spiro, T.G., Biochemical applications of resonance-Raman spectroscopy. Vibrational Spectra and Structure, 1976. 5: p. 101-20. Spiro, T.G. and Czernuszewicz, R.S., Resonance Raman spectroscopy of metalloproteins. Methods in Enzymology, 1995. 246: p. 416-60. Skoog, D.A., Holler, F.J., and Nieman, T.A., Principals of Instrumental Analysis. 5th ed. 1998, Toronto: Saunders Golden Sunburst Publishing. Morris, M.D. and Wallin, D.J., Resonance Raman spectroscopy. Current applications and prospects. Analytical Chemistry, 1979. 51: p. 182A-192A. Greek, L.S., Development and Characterization of an Optical Fiber Based Instrument for Ultraviolet Resonance Raman Spectroscopy of Biomolecules. 1998, The University of British Columbia: Vancouver. Asher, S.A. and Johnson, C.R., Raman spectroscopy of a coal liquid shows that fluorescence interference is minimized with ultraviolet excitation. Science, 1984. 225: p. 311-13. Barbosa, C.J., Fiber-optic Ultraviolet Resonance Raman Spectroscopy for Biophysical Investigations. 2002, The University of British Columbia: Vancouver. Asher, S.A., Bormett, R.W., Chen, X.G., Lemmon, D.H., Cho, N., Peterson, P., Arrigoni, M., Spinelli, L., and Cannon, J., UV resonance Raman spectroscopy using a new cw laser source: convenience and experimental simplicity. Applied Spectroscopy, 1993. 47: p. 628-33. Asher, S.A., Chi, Z., and Li, P., Resonance Raman examination of the two lowest amide pipi excited states. Journal of Raman Spectroscopy, 1998. 29: p. 927-931. Chi, Z. and Asher, S.A., UV Resonance Raman Determination of Protein Acid Denaturation: Selective Unfolding of Helical Segments of Horse Myoglobin. Biochemistry, 1998. 37: p. 2865-2872. Chi, Z., Chen, X.G., Holtz, J.S.W., and Asher, S.A., UV Resonance Raman-Selective Amide Vibrational Enhancement: Quantitative Methodology for Determining Protein Secondary Structure. Biochemistry, 1998. 37: p. 2854-2864. Myshakina, N.S., Ahmed, Z., and Asher, S.A., Dependence of Amide Vibrations on Hydrogen Bonding. Journal of Physical Chemistry B, 2008. 112: p. 11873–11877. Austin, J.C., Rodgers, K.R., and Spiro, T.G., Protein structure from ultraviolet resonance Raman spectroscopy. Methods in enzymology, 1993. 226: p. 374-96. Balakrishnan, G., Weeks, C.L., Ibrahim, M., Soldatova, A.V., and Spiro, T.G., Protein dynamics from time resolved UV Raman spectroscopy. Current Opinion in Structural Biology, 2008. 18: p. 623-629. Balakrishnan, G., Hu, Y., Oyerinde, O.F., Su, J., Groves, J.T., and Spiro, T.G., A Conformational Switch to beta -Sheet Structure in Cytochrome c Leads to Heme Exposure. Implications for Cardiolipin Peroxidation and Apoptosis. Journal of the American Chemical Society, 2007. 129: p. 504-505. Copeland, R.A. and Spiro, T.G., Ultraviolet resonance Raman spectra of cytochrome c conformational states. Biochemistry, 1985. 24: p. 4960-8.  187 36.  37.  38.  39.  40.  41. 42.  43.  44. 45.  46.  47.  48.  49. 50. 51.  Zhao, X., Chen, R., Raj, V., and Spiro, T.G., Assignment of the 1511 cm^-1 UV resonance Raman marker band of hemoglobin to tryptophan. Biopolymers, 2001. 62: p. 158-162. Huang, C.-Y., Balakrishnan, G., and Spiro, T.G., Protein secondary structure from deepUV resonance Raman spectroscopy. Journal of Raman Spectroscopy, 2006. 37: p. 277282. Benevides, J.M., Overman, S.A., and Thomas, G.J., Jr., Raman, polarized raman and ultraviolet resonance raman spectroscopy of nucleic acids and their complexes. Journal of Raman Spectroscopy, 2005. 36: p. 279-299. Shaw, C.P. and Jirasek, A., The use of ultraviolet resonance Raman spectroscopy in the analysis of ionizing-radiation-induced damage in DNA. Applied Spectroscopy, 2009. 63: p. 412-422. Jirasek, A., Schulze, H.G., Hughesman, C.H., Creagh, A.L., Haynes, C.A., Blades, M.W., and Turner, R.F.B., Discrimination between UV radiation-induced and thermally induced spectral changes in AT-paired DNA oligomers using UV resonance Raman spectroscopy. Journal of Raman Spectroscopy, 2006. 37: p. 1368-1380. Kundu, L.M. and Loppnow, G.R., Direct detection of 8-oxo-deoxyguanosine using UV resonance Raman spectroscopy. Photochemistry and Photobiology, 2007. 83: p. 600-602. Neugebauer, U., Schmid, U., Baumann, K., Holzgrabe, U., Ziebuhr, W., Kozitskaya, S., Kiefer, W., Schmitt, M., and Popp, J., Characterization of bacterial growth and the influence of antibiotics by means of UV resonance raman spectroscopy. Biopolymers, 2006. 82: p. 306-311. Gaus, K., Roesch, P., Petry, R., Peschke, K.D., Ronneberger, O., Burkhardt, H., Baumann, K., and Popp, J., Classification of lactic acid bacteria with UV-resonance Raman spectroscopy. Biopolymers, 2006. 82: p. 286-290. Nemecek, D. and Thomas, G.J., Jr., Raman spectroscopy of viruses and viral proteins. Frontiers of Molecular Spectroscopy, 2009553-595. Frosch, T., Schmitt, M., and Popp, J., In situ UV Resonance Raman Micro-spectroscopic Localization of the Antimalarial Quinine in Cinchona Bark. Journal of Physical Chemistry B, 2007. 111: p. 4171-4177. Barbosa, C.J., Vaillancourt, F.H., Eltis, L.D., Blades, M.W., and Turner, R.F.B., The power distribution advantage of fiber-optic coupled ultraviolet resonance Raman spectroscopy for bioanalytical and biomedical applications. Journal of Raman Spectroscopy, 2002. 33: p. 503-510. Greek, L.S., Schulze, H.G., Blades, M.W., Haynes, C.A., Klein, K.-F., and Turner, R.F.B., Fiber-optic probes with improved excitation and collection efficiency for deep UV Raman and resonance Raman spectroscopy. Applied Optics, 1998. 37: p. 170-180. Greek, L.S., Schulze, H.G., Haynes, C.A., Blades, M.W., and Turner, R.F.B., Rational design of fiber-optic probes for visible and pulsed-ultraviolet resonance Raman spectroscopy. Applied Optics, 1996. 35: p. 4086-4095. McCreery, R.L., Fleischmann, M., and Hendra, P., Fiber optic probe for remote Raman spectrometry. Analytical Chemistry, 1983. 55: p. 146-8. Schwab, S.D. and McCreery, R.L., Versatile, efficient Raman sampling with fiber optics. Analytical Chemistry, 1984. 56: p. 2199-204. Schulze, H.G., Barbosa, C.J., Greek, L.S., Turner, R.F.B., Haynes, C.A., Klein, K.-F., and Blades, M.W., Advances in fiber optic-based UV resonance Raman spectroscopy  188  52.  53.  54.  55. 56. 57.  58.  59. 60.  61. 62.  63.  64.  65.  techniques for anatomical and physiological investigations. Proceedings of SPIE-The International Society for Optical Engineering, 1999. 3608: p. 157-166. Schulze, H.G., Greek, L.S., Blades, M.W., Bree, A.V., Gorzalka, B.B., Klein, K.-F., and Turner, R.F.B., Development of optical fiber sensor probes for rapid remote in-situ spectroscopic measurements of biological samples. Proceedings of SPIE-The International Society for Optical Engineering, 1997. 2982: p. 251-262. Schulze, H.G., Greek, L.S., Blades, M.W., Bree, A.V., Gorzalka, B.B., and Turner, R.F.B., Neurotransmitter measurement with a fiber-optic probe using pulsed ultraviolet resonance Raman spectroscopy. Proceedings of SPIE-The International Society for Optical Engineering, 1997. 2982: p. 180-188. Wen, Z.Q. and Thomas, G.J., Jr., Structural studies of viruses by Raman spectroscopy main title. LX. UV resonance Raman spectroscopy of DNA and protein constituents of viruses: assignments and cross sections for excitations at 257, 244, 238, and 229 nm. Biopolymers, 1998. 45: p. 247-256. Lewis, I.R. and Griffiths, P.R., Raman spectrometry with fiber-optic sampling. Applied Spectroscopy, 1996. 50: p. 12A-30A. Lin, J., Hart, S.J., and Kenny, J.E., Improved Two-Fiber Probe for in Situ Spectroscopic Measurement. Analytical Chemistry, 1996. 68: p. 3098-3103. Schulze, H.G., The Development of a Fiber-Optic Probe for the In-Vivo Resonance Raman Spectroscopy of Neurotransmitters. 1996, The University of British Columbia: Vancouver. Johnson, C.R., Ludwig, M., and Asher, S.A., Ultraviolet resonance Raman characterization of photochemical transients of phenol, tyrosine, and tryptophan. Journal of the American Chemical Society, 1986. 108: p. 905-12. Zhang, D., Hanna, J.D., Jiang, Y., and Ben-Amotz, D., Influence of laser illumination geometry on the power distribution advantage. Applied Spectroscopy, 2001. 55: p. 61-65. Ramsey, J., Ranganathan, S., McCreery, R.L., and Zhao, J., Performance comparisons of conventional and line-focused surface Raman spectrometers. Applied Spectroscopy, 2001. 55: p. 767-773. Setchell, R.E., Laser-induced damage in step-index, multimode fibers. Proceedings of SPIE-The International Society for Optical Engineering, 1992. 1848: p. 15-23. Vaillancourt, F.H., Barbosa, C.J., Spiro, T.G., Bolin, J.T., Blades, M.W., Turner, R.F.B., and Eltis, L.D., Definitive evidence for monoanionic binding of 2,3-dihydroxybiphenyl to 2,3-dihydroxybiphenyl 1,2-dioxygenase from UV resonance Raman spectroscopy, UV/Vis absorption spectroscopy, and crystallography. Journal of the American Chemical Society, 2002. 124: p. 2485-96. Jirasek, A., Horsman, G.P., Vaillancourt, F.H., Barbosa, C.J., Eltis, L.D., Blades, M.W., and Turner, R.F.B., Probing enzyme-substrate interactions using fiber-optic ultraviolet resonance Raman spectroscopy. American Pharmaceutical Review, 2004. 7: p. 49-53. Horsman, G.P., Jirasek, A., Vaillancourt, F.H., Barbosa, C.J., Jarzecki, A.A., Xu, C., Mekmouche, Y., Spiro, T.G., Lipscomb, J.D., Blades, M.W., Turner, R.F.B., and Eltis, L.D., Spectroscopic Studies of the Anaerobic Enzyme-Substrate Complex of Catechol 1,2Dioxygenase. Journal of the American Chemical Society, 2005. 127: p. 16882-16891. Schulze, H.G., Greek, L.S., Barbosa, C.J., Blades, M.W., Gorzalka, B.B., and Turner, R.F., Measurement of some small-molecule and peptide neurotransmitters in-vitro using a fiber-optic probe with pulsed ultraviolet resonance Raman spectroscopy. Journal of Neuroscience Methods, 1999. 92: p. 15-24.  189 66.  67.  68.  69. 70. 71. 72. 73.  74. 75. 76. 77. 78.  79. 80.  81. 82. 83. 84.  Konorov, S.O., Schulze, H.G., Addison, C.J., Haynes, C.A., Blades, M.W., and Turner, R.F.B., Ultraviolet resonance Raman spectroscopy of locked single-stranded oligo(dA) reveals conformational implications of the locked ribose in LNA. Journal of Raman Spectroscopy, 2009. 40: p. 1162-1171. Addison, C.J., Konorov, S.O., Schulze, H.G., Turner, R.F.B., and Blades, M.W., Residual benzamide contamination in synthetic oligonucleotides observed using UV resonance Raman spectroscopy. Journal of Raman Spectroscopy, 2010. Blakeley, M.P., Cianci, M., Helliwell, J.R., and Rizkallah, P.J., Synchrotron and neutron techniques in biological crystallography. Chemical Society Reviews, 2004. 33: p. 548557. Moffat, K. and Wagner, G., Biophysical methods: editorial overview. Current Opinion in Structural Biology, 1995. 5: p. 637-9. Murali, R. and Burnett, R.M., X-ray crystallography of very large unit cells. Current Opinion in Structural Biology, 1991. 1: p. 997-1001. Reinemer, P., Dirr, H.W., and Huber, R., X-ray structure methods for glutathione binding. Methods in Enzymology, 1995. 251: p. 243-54. Sweet, R.M., Introduction to crystallography. Methods in Enzymology, 1985. 114: p. 1946. Emerson, J.P., Kovaleva, E.G., Farquhar, E.R., Lipscomb, J.D., and Que, L., Jr., Swapping metals in Fe- and Mn-dependent dioxygenases: evidence for oxygen activation without a change in metal redox state. Proceedings of the National Academy of Sciences of the United States of America, 2008. 105: p. 7347-7352. Kovaleva, E.G. and Lipscomb, J.D., Intermediate in the O-O Bond Cleavage Reaction of an Extradiol Dioxygenase. Biochemistry, 2008. 47: p. 11168-11170. Selvin, P.R., Lanthanide-labeled DNA. Topics in Fluorescence Spectroscopy, 2003. 7: p. 177-212. Selvin, P.R., The renaissance of fluorescence resonance energy transfer. Nature: Structural and Molecular Biology, 2000. 7: p. 730-734. Selvin, P.R., Rana, T.M., and Hearst, J.E., Luminescence Resonance Energy Transfer. Journal of the American Chemical Society, 1994. 116: p. 6029-30. Selvin, P.R. and Hearst, J.E., Luminescence energy transfer using a terbium chelate: improvements on fluorescence energy transfer. Proceedings of the National Academy of Sciences of the United States of America, 1994. 91: p. 10024-8. Selvin, P.R., Lanthanide-based resonance energy transfer. IEEE Journal of Selected Topics in Quantum Electronics, 1996. 2: p. 1077-1087. Layne, E., Spectrophotometric and Turbidimetric Methods for Measuring Proteins, in Methods in Enzymology, S.P. Colowick and N.O. Kaplan, Editors. 1957, Academic Press: New York. p. 447-455. Stoscheck, C.M., Quantitation of Protein, in Methods in Enzymology, M.P. Deutscher, Editor. 1990, Academic Press: New York. p. 50-68. Mulkerrin, M.G., in Spectroscopic Methods for Determining Protein Structure in Solution, H.A. Havel, Editor. 1996, VCH Publishers: New York. p. 5-27. Cherney, E., The Molecular Basis of Optical Activity. 1979, New York: Wiley Interscience. Ito, H., Arakawa, Y., and I'Haya, Y.J., Linear-response, Fano-DeVoe polarizability theory for biopolymers composed of chiral chromophores: infinite-order, dipole-dipole approximation calculations of circular dichroism and ultraviolet band shapes of alpha -  190  85.  86.  87. 88.  89.  90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100.  101.  102. 103. 104.  helical and beta -pleated polypeptides. Journal of Chemical Physics, 1993. 98: p. 883542. Manning, M.C., Underlying assumptions in the estimation of secondary structure content in proteins by circular dichroism spectroscopy - a critical review. Journal of Pharmaceutical and Biomedical Analysis, 1989. 7: p. 1103-19. Manning, M.C. and Woody, R.W., Theoretical study of the contribution of aromatic side chains to the circular dichroism of basic bovine pancreatic trypsin inhibitor. Biochemistry, 1989. 28: p. 8609-13. Woody, R.W., Improved calculation of the npi* rotational strength in polypeptides. Journal of Chemical Physics, 1968. 49: p. 4797-806. Walters, K.J., Ferentz, A.E., Hare, B.J., Hidalgo, P., Jasanoff, A., Matsuo, H., and Wagner, G., Characterizing protein-protein complexes and oligomers by nuclear magnetic resonance spectroscopy. Methods in Enzymology, 2001. 339: p. 238-258. Palmer, A.G., III, Kroenke, C.D., and Loria, J.P., Nuclear magnetic resonance methods for quantifying microsecond-to-millisecond motions in biological macromolecules. Methods in Enzymology, 2001. 339: p. 204-238. Van Nuland, N.A.J., Forge, V., Balbach, J., and Dobson, C.M., Real-Time NMR Studies of Protein Folding. Accounts of Chemical Research, 1998. 31: p. 773-780. Rastinejad, F., Evilia, C., and Lu, P., Studies of nucleic acids and their protein interactions by 19F-NMR. Methods in Enzymology, 1995. 261: p. 560-75. Petros, A.M. and Fesik, S.W., Nuclear magnetic resonance methods for studying proteinligand complexes. Methods in Enzymology, 1994. 239: p. 717-39. Wishart, D.S. and Sykes, B.D., Chemical shifts as a tool for structure determination. Methods in Enzymology, 1994. 239: p. 363-92. Hunsicker-Wang, L., Vogt, M., and De Rose, V.J., EPR methods to study specific metalion binding sites in RNA. Methods in Enzymology, 2009. 468: p. 335-367. Williams, B.B. and Halpern, H.J., In vivo EPR imaging. Biological Magnetic Resonance, 2005. 23: p. 283-319. Beinert, H., EPR spectroscopy of components of the mitochondrial electron-transfer system. Methods in Enzymology, 1978. 54: p. 133-50. Berliner, L.J., Spin labeling: theory and applications. 1976, New York: Academic Press. Teo, B.-K., EXAFS: basic principles and data analysis. 1986: Springer. Blackburn, N.J., EXAFS studies of copper proteins. 199063-103. Fay, M.J., Proctor, A., Hoffmann, D.P., and Hercules, D.M., Unraveling EXAFS spectroscopy. Analytical Chemistry, 1988. 60: p. 1225A-1226A, 1228A, 1230A-1234A, 1236A-1238A, 1240A-1243A. Lindley, P.F., Crystallographic studies of biological macromolecules using synchrotron radiation. NATO ASI Series, Series C Mathematical and Physical Sciences, 1988. 221: p. 509-36. Bray, R.C., The structures and catalytic mechanisms of molybdenum centers in enzymes studied by e.p.r. and x-ray spectroscopy. 1984707-22. Scott, R.A., Measurement of metal-ligand distances by EXAFS. Methods in Enzymology, 1985. 117: p. 414-59. Bunker, G., Introduction to XAFS: A Practical Guide to X-ray Absorption Fine Structure Spectroscopy. 2010: Cambridge University Press.  191 105.  106. 107.  108.  109.  110.  111.  112.  113.  114.  115.  116. 117.  118. 119.  Vaillancourt, F.H., Bolin, J.T., and Eltis, L.D., The ins and outs of ring-cleaving dioxygenases. Critical Reviews in Biochemistry and Molecular Biology, 2006. 41: p. 241-267. Groce, S.L., Miller-Rodeberg, M.A., and Lipscomb, J.D., Single-Turnover Kinetics of Homoprotocatechuate 2,3-Dioxygenase. Biochemistry, 2004. 43: p. 15141-15153. Emerson, J.P., Wagner, M.L., Reynolds, M.F., Que, L., Jr., Sadowsky, M.J., and Wackett, L.P., The role of histidine 200 in MndD, the Mn(II)-dependent 3,4dihydroxyphenylacetate 2,3-dioxygenase from Arthrobacter globiformis CM-2, a sitedirected mutagenesis study. JBIC, Journal of Biological Inorganic Chemistry, 2005. 10: p. 751-760. Hegg, E.L. and Que, L., Jr., The 2-His-1-carboxylate facial triad. An emerging structural motif in mononuclear non-heme iron(II) enzymes. European Journal of Biochemistry, 1997. 250: p. 625-629. Koehntop, K.D., Emerson, J.P., and Que, L., Jr., The 2-His-1-carboxylate facial triad: a versatile platform for dioxygen activation by mononuclear non-heme iron(II) enzymes. JBIC, Journal of Biological Inorganic Chemistry, 2005. 10: p. 87-93. Arciero, D.M., Orville, A.M., and Lipscomb, J.D., [17O]water and nitric oxide binding by protocatechuate 4,5-dioxygenase and catechol 2,3-dioxygenase. Evidence for binding of exogenous ligands to the active site iron of extradiol dioxygenases. Journal of Biological Chemistry, 1985. 260: p. 14035-44. Groce, S.L. and Lipscomb, J.D., Aromatic Ring Cleavage by Homoprotocatechuate 2,3Dioxygenase: Role of His200 in the Kinetics of Interconversion of Reaction Cycle Intermediates. Biochemistry, 2005. 44: p. 7175-7188. Vetting, M.W., Wackett, L.P., Que, L., Jr., Lipscomb, J.D., and Ohlendorf, D.H., Crystallographic comparison of manganese- and iron-dependent homoprotocatechuate 2,3-dioxygenases. Journal of Bacteriology, 2004. 186: p. 1945-1958. Shu, L., Chiou, Y.-M., Orville, A.M., Miller, M.A., Lipscomb, J.D., and Que, L., Jr., Xray absorption spectroscopic studies of the Fe(II) active site of catechol 2,3-dioxygenase. Implications for the extradiol cleavage mechanism. Biochemistry, 1995. 34: p. 6649-59. Arciero, D.M. and Lipscomb, J.D., Binding of oxygen-17-labeled substrate and inhibitors to protocatechuate 4,5-dioxygenase-nitrosyl complex. Evidence for direct substrate binding to the active site iron(2+) of extradiol dioxygenases. Journal of Biological Chemistry, 1986. 261: p. 2170-8. Fielding, A., Kovaleva, E., Farquhar, E., Lipscomb, J., and Que, L., A hyperactive cobaltsubstituted extradiol-cleaving catechol dioxygenase. Journal of Biological Inorganic Chemistry, 2011. 16: p. 341-355. Lipscomb, J.D., Mechanism of extradiol aromatic ring-cleaving dioxygenases. Current Opinion in Structural Biology, 2008. 18: p. 644-649. Vance, C.K. and Miller, A.-F., Novel Insights into the Basis for Escherichia coli Superoxide Dismutase's Metal Ion Specificity from Mn-Substituted FeSOD and Its Very High Em. Biochemistry, 2001. 40: p. 13079-13087. Kovaleva, E.G. and Lipscomb, J.D., Crystal Structures of Fe2+ Dioxygenase Superoxo, Alkylperoxo, and Bound Product Intermediates. Science, 2007. 316: p. 453-457. Ishimitsu, T., Hirose, S., and Sakurai, H., Microscopic acid dissociation constants of 3,4dihydroxyphenylpropionic acid and related compounds, and 3,4-dihydroxyphenylalanine (DOPA). Talanta, 1977. 24: p. 555-60.  192 120.  121. 122. 123.  124.  125. 126.  127.  128. 129. 130.  131.  132.  133.  134.  135. 136.  Karpishin, T.B., Gebhard, M.S., Solomon, E.I., and Raymond, K.N., Spectroscopic studies of the electronic structure of iron(III) tris(catecholates). Journal of the American Chemical Society, 1991. 113: p. 2977-84. Platt, J., R. , Classification of Spectra of Cata-Condensed Hydrocarbons. Journal of Chemical Physics, 1949. 17: p. 484-495. Raff, R. and Ettling, B.V., Encyclopedia of Chemical Technology. 2nd ed. 1970, New York: Interscience. Wen, Z.Q. and Thomas, G.J., Jr., UV resonance Raman spectroscopy of DNA and protein constituents of viruses: assignments and cross sections for excitations at 257, 244, 238, and 229 nm. Biopolymers, 1998. 45: p. 247-256. Gerhards, M., Perl, W., Schumm, S., Henrichs, U., Jacoby, C., and Kleinermanns, K., Structure and vibrations of catechol and catechol.H2O(D2O) in the S0 and S1 state. Journal of Chemical Physics, 1996. 104: p. 9362-9375. Ramirez, F.J. and Lopez Navarrete, J.T., Conformational and vibrational study on 1,2dihydroxybenzene. Journal of Molecular Structure, 1993. 293: p. 59-62. Greaves, S.J. and Griffith, W.P., Vibrational spectra of catechol, catechol-d2 and -d6 and the catecholate monoanion. Spectrochimica Acta, Part A Molecular and Biomolecular Spectroscopy, 1991. 47A: p. 133-40. Salama, S., Stong, J.D., Neilands, J.B., and Spiro, T.G., Electronic and resonance Raman spectra of iron(III) complexes of enterobactin, catechol, and N-methyl-2,3dihydroxybenzamide. Biochemistry, 1978. 17: p. 3781-5. Que, L., Jr. and Epstein, R.M., Resonance Raman studies on protocatechuate 3,4dioxygenase inhibitor complexes. Biochemistry, 1981. 20: p. 2545-9. Que, L., Jr. and Heistand, R.H., II, Resonance Raman studies on pyrocatechase. Journal of the American Chemical Society, 1979. 101: p. 2219-21. Oehrstroem, L. and Michaud-Soret, I., Fe-Catecholate and Fe-Oxalate Vibrations and Isotopic Substitution Shifts from DFT Quantum Chemistry. Journal of Physical Chemistry A, 1999. 103: p. 256-264. Oehrstroem, L. and Michaud-Soret, I., Quantum Chemical Approach to the Assignment of Iron-Catecholate Vibrations and Isotopic Substitution Shifts. Journal of the American Chemical Society, 1996. 118: p. 3283-4. Elgren, T.E., Orville, A.M., Kelly, K.A., Lipscomb, J.D., Ohlendorf, D.H., and Que, L., Jr., Crystal structure and resonance Raman studies of protocatechuate 3,4-dioxygenase complexed with 3,4-dihydroxyphenylacetate. Biochemistry, 1997. 36: p. 11504-11513. Michaud-Soret, I., Andersson, K.K., Que, L., Jr., and Haavik, J., Resonance Raman Studies of Catecholate and Phenolate Complexes of Recombinant Human Tyrosine Hydroxylase. Biochemistry, 1995. 34: p. 5504-10. Ormo, M., deMare, F., Regnstrom, K., Aberg, A., Sahlin, M., Ling, J., Loehr, T.M., Sanders-Loehr, J., and Sjoberg, B.M., Engineering of the iron site in ribonucleotide reductase to a self-hydroxylating monooxygenase. The Journal of biological chemistry, 1992. 267: p. 8711-4. Varsanyi, G., Assignments for Vibrational Spectra of 700 benzene derivatives, ed. A. Hilger. Vol. 1. 1974, New York: Wiley. Schulze, G., Jirasek, A., Yu, M.M.L., Lim, A., Turner, R.F.B., and Blades, M.W., Investigation of selected baseline removal techniques as candidates for automated implementation. Applied Spectroscopy, 2005. 59: p. 545-74.  193 137.  138. 139.  140.  141.  142. 143.  144. 145.  146. 147.  148.  149.  150.  151. 152.  153.  Jirasek, A., Schulze, G., Yu, M.M.L., Blades, M.W., and Turner, R.F.B., Accuracy and precision of manual baseline determination. Applied Spectroscopy, 2004. 58: p. 14881499. Bisswanger, H., Enzyme Kinetics: Principles and Methods. 2nd ed. 2008, Weinheim: Wiley-VCH. Georgiev, V., Borowski, T., Blomberg, M.R.A., and Siegbahn, P.E.M., A comparison of the reaction mechanisms of iron- and manganese-containing 2,3-HPCD: an important spin transition for manganese. JBIC, Journal of Biological Inorganic Chemistry, 2008. 13: p. 929-940. Asher, S.A., Ludwig, M., and Johnson, C.R., UV Resonance Raman excitation profiles of the aromatic amino acids. Journal of the American Chemical Society, 1986. 108: p. 3186-97. Ludwig, M. and Asher, S.A., Ultraviolet resonance Raman excitation profiles of tyrosine: dependence of Raman cross sections on excited-state intermediates. Journal of the American Chemical Society, 1988. 110: p. 1005-11. Thomas, G.J., Raman Spectroscopy of Protein and Nucleic Acid Assemblies. Annual Review of Biophysics and Biomolecular Structure, 1999. 28: p. 1-27. Rava, R.P. and Spiro, T.G., Selective enhancement of tyrosine and tryptophan resonance Raman spectra via ultraviolet laser excitation. Journal of the American Chemical Society, 1984. 106: p. 4062-4. Rava, R.P. and Spiro, T.G., Resonance enhancement in the ultraviolet Raman spectra of aromatic amino acids. Journal of Physical Chemistry, 1985. 89: p. 1856-61. Su, C., Wang, Y., and Spiro, T.G., Saturation effects on ultraviolet resonance Raman intensities: excimer/YAG laser comparison and aromatic amino acid cross-sections. Journal of Raman Spectroscopy, 1990. 21: p. 435-40. Sweeney, J.A. and Asher, S.A., Tryptophan UV resonance Raman excitation profiles. Journal of Physical Chemistry, 1990. 94: p. 4784-91. Hu, X. and Spiro, T.G., Tyrosine and Tryptophan Structure Markers in Hemoglobin Ultraviolet Resonance Raman Spectra: Mode Assignments via Subunit-Specific Isotope Labeling of Recombinant Protein. Biochemistry, 1997. 36: p. 15701-15712. Okishio, N., Fukuda, R., Nagai, M., Nagai, Y., Nagatomo, S., and Kitagawa, T., Tyrosine phosphorylation-induced changes in absorption and UV resonance Raman spectra of Src peptides. Journal of Raman Spectroscopy, 1998. 29: p. 31-39. Takeuchi, H., Watanabe, N., and Harada, I., Vibrational spectra and normal coordinate analysis of p-cresol and its deuterated analogs. Spectrochimica Acta, Part A Molecular and Biomolecular Spectroscopy, 1988. 44A: p. 749-61. Jirasek, A., Tyrosinate Protonation and Trianionic Substrate Binding in Protocatechuate 3,4-Dioxygenase (Unpublished Work). 2004, The University of British Columbia: Vancouver. Dagley, S., Essays Biochem., 1975. 11: p. 81. Furukawa, K., Tomizuka, N., and Kamibayashi, A., Effect of chlorine substitution on the bacterial metabolism of various polychlorinated biphenyls. Applied and Environmental Microbiology, 1979. 38: p. 301-10. Seah, S.Y.K., Labbe, G., Nerdinger, S., Johnson, M.R., Snieckus, V., and Eltis, L.D., Identification of a serine hydrolase as a key determinant in the microbial degradation of polychlorinated biphenyls. Journal of Biological Chemistry, 2000. 275: p. 15701-15708.  194 154.  155.  156.  157.  158.  159.  160.  161.  162.  163.  164. 165.  166. 167.  Van der Geize, R., Yam, K., Heuser, T., Wilbrink, M.H., Hara, H., Anderton, M.C., Sim, E., Dijkhuizen, L., Davies, J.E., Mohn, W.W., and Eltis, L.D., A gene cluster encoding cholesterol catabolism in a soil actinomycete provides insight into Mycobacterium tuberculosis survival in macrophages. Proceedings of the National Academy of Sciences of the United States of America, 2007. 104: p. 1947-1952. Rengarajan, J., Bloom, B.R., and Rubin, E.J., Genome-wide requirements for Mycobacterium tuberculosis adaptation and survival in macrophages. Proceedings of the National Academy of Sciences of the United States of America, 2005. 102: p. 8327-8332. Fleming, S.M., Robertson, T.A., Langley, G.J., and Bugg, T.D.H., Catalytic Mechanism of a C-C Hydrolase Enzyme: Evidence for a Gem-Diol Intermediate, Not an Acyl Enzyme. Biochemistry, 2000. 39: p. 1522-1531. Li, J.-J., Li, C., Blindauer, C.A., and Bugg, T.D.H., Evidence for a gem-Diol Reaction Intermediate in Bacterial C-C Hydrolase Enzymes BphD and MhpC from 13C NMR Spectroscopy. Biochemistry, 2006. 45: p. 12461-12469. Li, J.-J. and Bugg, T.D.H., Investigation of a general base mechanism for ester hydrolysis in C-C hydrolase enzymes of the alpha /beta -hydrolase superfamily: a novel mechanism for the serine catalytic triad. Organic & Biomolecular Chemistry, 2007. 5: p. 507-513. Lam, W.W.Y. and Bugg, T.D.H., Purification, Characterization, and Stereochemical Analysis of a C-C Hydrolase: 2-Hydroxy-6-keto-nona-2,4-diene-1,9-Dioic Acid 5,6Hydrolase. Biochemistry, 1997. 36: p. 12242-12251. Henderson, I.M.J. and Bugg, T.D.H., Pre-Steady-State Kinetic Analysis of 2-Hydroxy-6keto-nona-2,4-diene-1,9-dioic Acid 5,6-Hydrolase: Kinetic Evidence for Enol/Keto Tautomerization. Biochemistry, 1997. 36: p. 12252-12258. Horsman, G.P., Ke, J., Dai, S., Seah, S.Y.K., Bolin, J.T., and Eltis, L.D., Kinetic and Structural Insight into the Mechanism of BphD, a C-C Bond Hydrolase from the Biphenyl Degradation Pathway. Biochemistry, 2006. 45: p. 11071-11086. Bhowmik, S., Horsman, G.P., Bolin, J.T., and Eltis, L.D., The Molecular Basis for Inhibition of BphD, a C-C Bond Hydrolase Involved in Polychlorinated Biphenyls Degradation: large 3-substituents prevent tautomerization. Journal of Biological Chemistry, 2007. 282: p. 36377-36385. Horsman, G.P., Bhowmik, S., Seah, S.Y.K., Kumar, P., Bolin, J.T., and Eltis, L.D., The Tautomeric Half-reaction of BphD, a C-C Bond Hydrolase: Kinetic and Structural Evidence Supporting a Key Role for Histidine 265 of the Catalytic Triad. Journal of Biological Chemistry, 2007. 282: p. 19894-19904. Williams, D.H. and Flemming, I., Spectroscopic Methods in Organic Chemistry. 5th ed. 1995, London: McGraw-Hill. Hu, J.G., Sun, B.Q., Petkova, A.T., Griffin, R.G., and Herzfeld, J., The Predischarge Chromophore in Bacteriorhodopsin: A 15N Solid-State NMR Study of the L Photointermediate. Biochemistry, 1997. 36: p. 9316-9322. Lanyi, J.K. and Schobert, B., Structural Changes in the L Photointermediate of Bacteriorhodopsin. Journal of Molecular Biology, 2007. 365: p. 1379-1392. D'Ordine, R.L., Tonge, P.J., Carey, P.R., and Anderson, V.E., Electronic Rearrangement Induced by Substrate Analog Binding to the Enoyl-CoA Hydratase Active Site: Evidence for Substrate Activation. Biochemistry, 1994. 33: p. 12635-43.  195 168.  169. 170.  171.  172. 173.  174. 175.  176.  177.  178. 179.  Tian, F., Debler, E.W., Millar, D.P., Deniz, A.A., Wilson, I.A., and Schultz, P.G., The effects of antibodies on stilbene excited-state energetics. Angewandte Chemie, International Edition, 2006. 45: p. 7763-7765. Sekharan, S., Sugihara, M., and Buss, V., Origin of spectral tuning in rhodopsin - it is not the binding pocket. Angewandte Chemie, International Edition, 2007. 46: p. 269-271. Seah, S.Y.K., Terracina, G., Bolin, J.T., Riebel, P., Snieckus, V., and Eltis, L.D., Purification and Preliminary Characterization of a Serine Hydrolase Involved in the Microbial Degradation of Polychlorinated Biphenyls. 1998. p. 22943-22949. Nerdinger, S., Marchhart, R., Riebel, P., Kendall, C., Johnson, M.R., Yin, C.F., Snieckus, V., and Eltis, L.D., Directed ortho-metalation and Suzuki-Miyaura cross-coupling connections: regiospecific synthesis of all isomeric chlorodihydroxybiphenyls for microbial degradation studies of PCBs. Chemical Communications (Cambridge), 19992259-2260. Cornish-Bowden, A., Analysis of Enzyme Kinetic Data. 1995, New York: Oxford University Press. 208. Frisch, M.J., Trucks, G.W., Schlegel, H.B., Scuseria, G.E., Robb, M.A., Cheeseman, J.R., Montgomery, J.A., Vreven, T., Kudin, K.N., Burant, J.C., Millam, J.M., Lyengar, S.S., Tomasi, J., Barone, B., Mennucci, B., Cossi, M., Scalmani, G., Rega, N., Petersson, G.A., Nakasuji, H., Hada, M., Ehara, M., Toyota, K., Fukada, R., Hasegawa, J., Ishida, M., Nakajima, T., Honda, Y., Kitao, O., Nakai, H., Klene, M., Li, X., Know, E., Hratchian, H.P., Cross, J.B., Adamo, C., Jaramillo, J., Gomperts, R., Stratmann, R.E., Yazyev, O., Austin, A.J., Cammi, R., Pomelli, C., Ochterski, J.W., Ayala, P.Y., Morokuma, K., Voth, G.A., Salvador, P., Dannenberg, J.J., Zakrzewski, V.G., Dapprich, S., Daniels, A.D., Strain, M.C., Farkas, O., Malick, D.K., Rabuck, A.D., Raghavachari, K., Foresman, J.B., Ortiz, J.V., Cui, Q., Baboul, A.G., Clifford, S., Cioslowski, J., Stefanov, B.B., Liu, G., Liashenko, A., Piskorz, P., Komaromi, I., Martin, R.L., Fox, D.J., Keith, T., Al-Laham, M.A., Peng, C.Y., Nanayakkara, A., Challacombe, M., Gill, P.M.W., Johnson, B., Chen, W., Wong, M.W., Gonzalez, C., and Pople, J.A., Gaussian 03. 2003, Gaussian, Inc: Pittsburgh, PA. Becke, A.D., Density-functional thermochemistry. III. The role of exact exchange. Journal of Chemical Physics, 1993. 98: p. 5648-52. Lee, C., Yang, W., and Parr, R.G., Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Physical Review B: Condensed Matter and Materials Physics, 1988. 37: p. 785-9. Dunning, T.H., Jr., Gaussian basis sets for use in correlated molecular calculations. I. The atoms boron through neon and hydrogen. Journal of Chemical Physics, 1989. 90: p. 1007-23. Schiavoni, M.M., Di Loreto, H.E., Hermann, A., Mack, H.G., Ulic, S.E., and Della Vedova, C.O., Keto-enol tautomerism in beta -ketoesters: CH3C(O)CHXC(O)OY (X = H, Cl; Y = CH3, C2H5). Vibrational analyses, NMR spectra and quantum chemical calculations. Journal of Raman Spectroscopy, 2001. 32: p. 319-329. Hanai, K., Kuwae, A., Kawai, S., and Ono, Y., Keto-enol tautomerism and vibrational spectra of phenylpyruvic acids. Journal of Physical Chemistry, 1989. 93: p. 6013-16. Pellett, J.D., Sabaj, K.M., Stephens, A.W., Bell, A.F., Wu, J., Tonge, P.J., and Stankovich, M.T., Medium-Chain Acyl-Coenzyme A Dehydrogenase Bound to a Product Analogue, Hexadienoyl-Coenzyme A: Effects on Reduction Potential, pKa, and Polarization. Biochemistry, 2000. 39: p. 13982-13992.  196 180.  181.  182.  183.  184.  185.  186.  187.  188. 189.  190.  191. 192.  193.  Wang, P., Fu, L.-M., Zhang, J.-P., Kakitani, Y., Ishii, H., Nagae, H., and Koyama, Y., Strong carotenoid-to-peptide interaction immediately after triplet excitation triggering conformational changes in photo-reaction center-bound 15-cis-spheroidene as revealed by submicrosecond time-resolved Raman spectroscopy. Chemical Physics Letters, 2008. 458: p. 175-179. Limantara, L., Katheder, I., Scheer, H., Schäfer, W., and Koyama, Y., The T1 and S1 raman spectra of 15N- and 2H-enriched bacteriochlorophyll a: changes in bond order upon triplet and singlet excitation. Chemical Physics Letters, 1996. 262: p. 656-662. Schrader, B. and Ansmann, A., Determination of Conformation of Conjugated Dienes by Means of Vibrational Frequencies. Angewandte Chemie International Edition, 1975. 14: p. 364-365. Mukai-Kuroda, Y., Fujii, R., Ko-chi, N., Sashima, T., Koyama, Y., Abe, M., Gebhard, R., van der Hoef, I., and Lugtenburg, J., Changes in Molecular Structure upon Triplet Excitation of All-trans-Spheroidene in n-Hexane Solution and 15-cis-Spheroidene Bound to the Photo-Reaction Center from Rhodobacter sphaeroides As Revealed by ResonanceRaman Spectroscopy and Normal-Coordinate Analysis. Journal of Physical Chemistry A, 2002. 106: p. 3566-3579. Deng, H., Manor, D., Weng, G., Rath, P., Koutalos, Y., Ebrey, T., Gebhard, R., Lugtenburg, J., Tsuda, M., and Callender, R.H., A Resonance Raman Study of Octopus Bathorhodopsin with Deuterium Labeled Retinal Chromophores. 1991. 54: p. 1001-1007. Dong, J., Lu, X., Wei, Y., Luo, L., Dunaway-Mariano, D., and Carey, P.R., The Strength of Dehalogenase-Substrate Hydrogen Bonding Correlates with the Rate of Meisenheimer Intermediate Formation. Biochemistry, 2003. 42: p. 9482-9490. Weng, K.F., Shi, Y., Zheng, X., and Phillips, D.L., Resonance Raman Investigation of the Short-Time Photodissociation Dynamics of the Charge-Transfer Absorption of the I2Benzene Complex in Benzene Solution. Journal of Physical Chemistry A, 2006. 110: p. 851-860. Ibrahim, M., Derbyshire, E.R., Marletta, M.A., and Spiro, T.G., Probing Soluble Guanylate Cyclase Activation by CO and YC-1 Using Resonance Raman Spectroscopy. Biochemistry, 2010. 49: p. 3815-3823. Jepsen, J.S., Sorensen, M.D., and Wengel, J., Locked nucleic acid: A potent nucleic acid analog in therapeutics and biotechnology. Oligonucleotides, 2004. 14: p. 130-146. Mathe, C. and Perigaud, C., Recent approaches in the synthesis of conformationally restricted nucleoside analogues. European Journal of Organic Chemistry, 200814891505. Karkare, S. and Bhatnagar, D., Promising nucleic acid analogs and mimics: characteristic features and applications of PNA, LNA, and morpholino. Applied Microbiology and Biotechnology, 2006. 71: p. 575-586. Petersen, M. and Wengel, J., LNA: a versatile tool for therapeutics and genomics. Trends in Biotechnology, 2003. 21: p. 74-81. Nielsen, K.E., Rasmussen, J., Kumar, R., Wengel, J., Jacobsen, J.P., and Petersen, M., NMR Studies of Fully Modified Locked Nucleic Acid (LNA) Hybrids: Solution Structure of an LNA:RNA Hybrid and Characterization of an LNA:DNA Hybrid. Bioconjugate Chemistry, 2004. 15: p. 449-457. Petersen, M., Nielsen, C.B., Nielsen, K.E., Jensen, G.A., Bondensgaard, K., Singh, S.K., Rajwanshi, V.K., Koshkin, A.A., Dahl, B.M., Wengel, J., and Jacobsen, J.P., The  197  194.  195. 196.  197.  198.  199. 200. 201. 202. 203. 204.  205. 206.  207. 208.  209. 210.  conformations of locked nucleic acids (LNA). Journal of Molecular Recognition, 2000. 13: p. 44-53. Koshkin, A.A., Singh, S.K., Nielsen, P., Rajwanshi, V.K., Kumar, R., Meldgaard, M., Olsen, C.E., and Wengel, J., LNA (locked nucleic acids): synthesis of the adenine, cytosine, guanine, 5-methylcytosine, thymine and uracil bicyclonucleoside monomers, oligomerization, and unprecedented nucleic acid recognition. Tetrahedron, 1998. 54: p. 3607-3630. Saenger, W., Principles of Nucleic Acid Structure. 1984, New York: Springer. Leijon, M., Graeslund, A., Nielsen, P.E., Buchardt, O., Norden, B., Kristensen, S.M., and Eriksson, M., Structural Characterization of PNA-DNA Duplexes by NMR. Evidence for DNA in a B-like Conformation. Biochemistry, 1994. 33: p. 9820-5. Isaksson, J., Acharya, S., Barman, J., Cheruku, P., and Chattopadhyaya, J., SingleStranded Adenine-Rich DNA and RNA Retain Structural Characteristics of Their Respective Double-Stranded Conformations and Show Directional Differences in Stacking Pattern. Biochemistry, 2004. 43: p. 15996-16010. Vesnaver, G. and Breslauer, K.J., The contribution of DNA single-stranded order to the thermodynamics of duplex formation. Proceedings of the National Academy of Sciences of the United States of America, 1991. 88: p. 3569-73. Crespo-Hernandez, C.E., Cohen, B., and Kohler, B., Base stacking controls excited-state dynamics in A.T DNA. Nature (London, United Kingdom), 2005. 436: p. 1141-1144. Xu, D.-G. and Nordlund, T.M., Sequence dependence of energy transfer in DNA oligonucleotides. Biophysical Journal, 2000. 78: p. 1042-1058. Dickerson, R.E., Drew, H.R., Conner, B.N., Wing, R.M., Fratini, A.V., and Kopka, M.L., The anatomy of A-, B-, and Z-DNA. Science, 1982. 216: p. 475-85. Ghosh, A. and Bansal, M., A glossary of DNA structures from A to Z. Acta Crystallographica, Section D Biological Crystallography, 2003. D59: p. 620-626. Alexeev, D.G., Lipanov, A.A., and Skuratovskii, I., Poly(dA).poly(dT) is a B-type double helix with a distinctively narrow minor groove. Nature, 1987. 325: p. 821-3. Nelson, H.C.M., Finch, J.T., Luisi, B.F., and Klug, A., The structure of an oligo(dA).oligo(dT) tract and its biological implications. Nature (London, United Kingdom), 1987. 330: p. 221-6. Aymami, J., Coll, M., Frederick, C.A., Wang, A.H., and Rich, A., The propeller DNA conformation of poly(dA).poly(dT). Nucleic acids research, 1989. 17: p. 3229-45. Sprous, D., Zacharias, W., Wood, Z.A., and harvey, S.C., Dehydrating agents sharply reduce curvature in DNAs containing A tracts. Nucleic Acids Research, 1995. 23: p. 1816-21. Premilat, S. and Albiser, G., X-ray fiber diffraction study of an elevated temperature structure of poly(dA).poly(dT). Journal of Molecular Biology, 1997. 274: p. 64-71. Demaret, J.P. and Gueron, M., The sensitivity of the B-Z transition of DNA to multivalent cations in high and low concentrations: A unified electrostatic interpretation. Proceedings of the Conversation in Biomolecular Stereodynamics, 11th, Albany, NY, United States, June 15-19, 1999, 2000. Convers. 11: p. 377-382. Sponer, J., Berger, I., Spackova, N., Leszczynski, J., and Hobza, P., J. Biomol. Struct. Dyn., 2000. Conversation 11: p. 383. Hobza, P. and Sponer, J., Structure, Energetics, and Dynamics of the Nucleic Acid Base Pairs: Nonempirical Ab Initio Calculations. Chemical Reviews (Washington, D. C.), 1999. 99: p. 3247-3276.  198 211.  212. 213.  214.  215.  216.  217.  218.  219. 220.  221. 222. 223.  224.  225.  226.  Hobza, P. and Sponer, J., Toward True DNA Base-Stacking Energies: MP2, CCSD(T), and Complete Basis Set Calculations. Journal of the American Chemical Society, 2002. 124: p. 11802-11808. Luo, R., Gilson, H.S.R., Potter, M.J., and Gilson, M.K., The physical basis of nucleic acid base stacking in water. Biophysical Journal, 2001. 80: p. 140-148. Mignon, P., Loverix, S., Steyaert, J., and Geerlings, P., Influence of the pi-pi interaction on the hydrogen bonding capacity of stacked DNA/RNA bases. Nucleic Acids Research, 2005. 33: p. 1779-1789. Buchvarov, I., Wang, Q., Raytchev, M., Trifonov, A., and Fiebig, T., Electronic energy delocalization and dissipation in single- and double-stranded DNA. Proceedings of the National Academy of Sciences of the United States of America, 2007. 104: p. 4794-4797. Kwok, W.-M., Ma, C., and Phillips, D.L., Femtosecond Time- and Wavelength-Resolved Fluorescence and Absorption Spectroscopic Study of the Excited States of Adenosine and an Adenine Oligomer. Journal of the American Chemical Society, 2006. 128: p. 1189411905. Ritze, H.H., Lippert, H., Samoylova, E., Smith, V.R., Hertel, I.V., Radloff, W., and Schultz, T., Relevance of pis* states in the photoinduced processes of adenine, adenine dimer, and adenine-water complexes. Journal of Chemical Physics, 2005. 122: p. 224320/1-224320/9. Ban, C., Ramakrishnan, B., and Sundaralingam, M., A single 2'-hydroxyl group converts B-DNA to A-DNA. Crystal structure of the DNA-RNA chimeric decamer duplex d(CCGGC)r(G)d(CCGG) with a novel intermolecular G.C base-paired quadruplet. Journal of Molecular Biology, 1994. 236: p. 275-85. Mills, J.B., Vacano, E., and Hagerman, P.J., Flexibility of single-stranded DNA: use of gapped duplex helices to determine the persistence lengths of poly(dT) and poly(dA). Journal of Molecular Biology, 1999. 285: p. 245-257. Cantor, C.R. and Schimmel, P.R., Biophysical Chemistry, Pt. 1: The Conformation of Biological Macromolecules. 1980. 341 pp. Lu, X.-J. and Olson, W.K., 3DNA: a software package for the analysis, rebuilding and visualization of three-dimensional nucleic acid structures. Nucleic Acids Research, 2003. 31: p. 5108-5121. Hagerman, K.R. and Hagerman, P.J., Helix rigidity of DNA: the meroduplex as an experimental paradigm. Journal of Molecular Biology, 1996. 260: p. 207-223. Mills, J.B. and Hagerman, P.J., Origin of the intrinsic rigidity of DNA. Nucleic Acids Research, 2004. 32: p. 4055-4059. Crespo-Hernandez, C.E. and Kohler, B., Influence of Secondary Structure on Electronic Energy Relaxation in Adenine Homopolymers. Journal of Physical Chemistry B, 2004. 108: p. 11182-11188. Konorov, S.O., Schulze, H.G., Addison, C.J., Haynes, C.A., Blades, M.W., and Turner, R.F.B., Base Stacking Configuration is a Major Determinant of Excited State Dynamics in A.T DNA and LNA. The Open Spectroscopy Journal, 2009. 3: p. 9-20. Kundu, L.M., Linne, U., Marahiel, M., and Carell, T., RNA is more UV resistant than DNA: The formation of UV-induced DNA lesions is strongly sequence and conformation dependent. Chemistry--A European Journal, 2004. 10: p. 5697-5705. Lysetska, M., Knoll, A., Boehringer, D., Hey, T., Krauss, G., and Krausch, G., UV lightdamaged DNA and its interaction with human replication protein A: an atomic force microscopy study. Nucleic Acids Research, 2002. 30: p. 2686-2691.  199 227. 228.  229.  230.  231.  232. 233.  234.  235.  236.  237.  238. 239.  240. 241. 242.  Neidle, S., Crystallographic insights into DNA minor groove recognition by drugs. Biopolymers, 1997. 44: p. 105-121. White, S., Szewczyk, J.W., Turner, J.M., Baird, E.E., and Dervan, P.B., Recognition of the four Watson-Crick base pairs in the DNA minor groove by synthetic ligands. Nature (London), 1998. 391: p. 468-471. Parthasarathi, R. and Subramanian, V., Stacking Interactions in Benzene and Cytosine Dimers: From Molecular Electron Density Perspective. Structural Chemistry, 2005. 16: p. 243-255. Greve, J., Maestre, M.F., Moise, H., and Hosoda, J., Circular dichroism study of the interaction between T4 gene 32 protein and polynucleotides. Biochemistry, 1978. 17: p. 887-93. Hung, S.H., Yu, Q., Gray, D.M., and Ratliff, R.L., Evidence from CD spectra that d(purine).r(pyrimidine) and r(purine).d(pyrimidine) hybrids are in different structural classes. Nucleic Acids Research, 1994. 22: p. 4326-34. Sprecher, C.A., Baase, W.A., and Johnson, W.C., Jr., Conformation and circular dichroism of DNA. Biopolymers, 1979. 18: p. 1009-19. Toyama, A., Takino, Y., Takeuchi, H., and Harada, I., Ultraviolet resonance Raman spectra of ribosyl C(1')-deuterated purine nucleosides: evidence of vibrational coupling between purine and ribose rings. Journal of the American Chemical Society, 1993. 115: p. 11092-8. Chan, S.S., Austin, R.H., Mukerji, I., and Spiro, T.G., Temperature-dependent ultraviolet resonance Raman spectroscopy of the premelting state of dA . dT DNA. Biophysical Journal, 1997. 72: p. 1512-1520. Mukerji, I., Shiber, M.C., Spiro, T.G., and Fresco, J.R., A UV resonance Raman study of d(A+-G)10, a single-stranded helix without stacked or paired bases. Biochemistry, 1995. 34: p. 14300-3. Mukerji, I. and Williams, A.P., UV Resonance Raman and Circular Dichroism Studies of a DNA Duplex Containing an A3T3 Tract: Evidence for a Premelting Transition and Three-Centered H-Bonds. Biochemistry, 2002. 41: p. 69-77. Tsuboi, M., Komatsu, M., Hoshi, J., Kawashima, E., Sekine, T., Ishido, Y., Russell, M.P., Benevides, J.M., and Thomas, G.J., Jr., Raman and Infrared Spectra Of (2'S)-[2'2H]Thymidine: Vibrational Coupling between Deoxyribosyl and Thymine Moieties and Structural Implications. Journal of the American Chemical Society, 1997. 119: p. 20252032. Couch, G.S., Hendrix, D.K., and Ferrin, T.E., Nucleic acid visualization with UCSF Chimera. Nucleic Acids Research, 2006. 34: p. e29/1-e29/5. Sponer, J., Leszczynski, J., and Hobza, P., Nature of nucleic acid-base stacking: nonempirical ab initio and empirical potential characterization of 10 stacked base dimers. Comparison of stacked and H-bonded base pairs. Journal of Physical Chemistry, 1996. 100: p. 5590-6. Skoog, D.A., Principles of Instrumental Analysis. 3rd ed. 1985, Philadelphia: Saunders College Publishing. Swart, M., van der Wijst, T., Guerra, C.F., and Bickelhaupt, F.M., pi-pi stacking tackled with density functional theory. Journal of Molecular Modeling, 2007. 13: p. 1245-1257. Krapp, A., Bickelhaupt, F.M., and Frenking, G., Orbital overlap and chemical bonding. Chemistry--A European Journal, 2006. 12: p. 9196-9216.  200 243.  244.  245.  246. 247. 248.  249.  250.  251.  252.  253.  254.  255. 256.  257.  Emanuele, E., Markovitsi, D., Millie, P., and Zakrzewska, K., UV spectra and excitation delocalization in DNA: Influence of the spectral width. ChemPhysChem, 2005. 6: p. 1387-1393. Markovitsi, D., Onidas, D., Talbot, F., Marguet, S., Gustavsson, T., and Lazzarotto, E., UVB/UVC induced processes in model DNA helices studied by time-resolved spectroscopy: Pitfalls and tricks. Journal of Photochemistry and Photobiology, A Chemistry, 2006. 183: p. 1-8. Santoro, F., Barone, V., and Improta, R., Influence of base stacking on excited-state behavior of polyadenine in water, based on time-dependent density functional calculations. Proceedings of the National Academy of Sciences of the United States of America, 2007. 104: p. 9931-9936. Marian, C.M., A new pathway for the rapid decay of electronically excited adenine. Journal of Chemical Physics, 2005. 122: p. 104314/1-104314/13. Koningstein, J.A., Introduction to the theory of the Raman effect. 1972, Dordrecht: Reidel. Toyama, A., Hanada, N., Abe, Y., Takeuchi, H., and Harada, I., Assignment of adenine ring in-plane vibrations in adenosine on the basis of 15N and 13C isotopic frequency shifts and UV resonance Raman enhancement. Journal of Raman Spectroscopy, 1994. 25: p. 623-30. Perno, J.R., Grygon, C.A., and Spiro, T.G., Ultraviolet Raman excitation profiles for the nucleotides and for the nucleic acid duplexes poly(rA)-poly-(rU) and poly(dG-dC). Journal of Physical Chemistry, 1989. 93: p. 5672-8. Bouvier, B., Gustavsson, T., Markovitsi, D., and Millie, P., Dipolar coupling between electronic transitions of the DNA bases and its relevance to exciton states in double helices. Chemical Physics, 2002. 275: p. 75-92. Konorov, S.O., Schulze, H.G., Addison, C.J., Haynes, C.A., Turner, R.F.B., and Blades, M.W., Temperature-Dependent Excited State Absorption in DNA and LNA Oligomers Supports an Emerging Model of Excited State Dynamics in DNA. The Open Spectroscopy Journal, 2009. 3: p. 43-51. Hu, L., Zhao, Y., Wang, F., Chen, G., Ma, C., Kwok, W.-M., and Phillips, D.L., Are Adenine Strands Helical H-Aggregates? Journal of Physical Chemistry B, 2007. 111: p. 11812-11816. Sokolov, L., Wojtuszewski, K., Tsukroff, E., and Mukerji, I., Nucleic acid structure investigated by UV resonance Raman spectroscopy: Protonation effects and A-tract structure. Proceedings of the Conversation in Biomolecular Stereodynamics, 11th, Albany, NY, United States, June 15-19, 1999, 2000. Convers. 11: p. 327-334. Atkinson, T. and Smith, M., Solid phase synthesis of oligodeoxyribonucleotides by the phosphite-triester method, in Oligonucleotide Synthesis: A Practical Approach. 1984, IRL Press: Oxford. p. 35-82. Pon, R.T. and Yu, S., Tandem oligonucleotide synthesis using linker phosphoramidites. Nucleic Acids Research, 2005. 33: p. 1940-1948. Arvela, R.K., Leadbeater, N.E., Sangi, M.S., Williams, V.A., Granados, P., and Singer, R.D., A reassessment of the transition-metal free Suzuki-type coupling methodology. The Journal of Organic Chemistry, 2005. 70: p. 161-8. Buchwald, S.L. and Bolm, C., On the Role of Metal Contaminants in Catalyses with FeCl3. Angewandte Chemie, International Edition, 2009. 48: p. 5586-5587.  201 258.  259.  260.  261.  262. 263.  264.  Toyama, A., Ohta, K., and Takeuchi, H., Correlation between the UV resonance Raman intensity and hydrogen bonding of the adenine ring. Journal of Molecular Structure, 2005. 735-736: p. 235-241. Tsuboi, M., Nishimura, Y., Hirakawa, A.Y., and Peticolas, W.L., in Biological Applications of Raman Spectroscopy, T.G. Spiro, Editor. 1987, John Wiley and Sons: New york. p. 109. Wen, Z.Q., Overman, S.A., and Thomas, G.J., Jr., Structure and interactions of the single-stranded DNA genome of filamentous virus fd: investigation by ultraviolet resonance raman spectroscopy. Biochemistry, 1997. 36: p. 7810-7820. Vacque, V., Dupuy, N., Sombret, B., Huvenne, J.P., and Legrand, P., In situ characterisation of peroxybenzimidic acid by FT-Raman and ATR/FTIR spectroscopy. Journal of Molecular Structure, 1996. 384: p. 165-174. Palomar, J., De Paz, J.L.G., and Catala, J., Vibrational study of intramolecular hydrogen bonding in o-hydroxybenzoyl compounds. Chemical Physics, 1999. 246: p. 167-208. Pei, K.-M., Ma, Y., and Zheng, X., Resonance Raman and theoretical investigation of the photodissociation dynamics of benzamide in S[sub 3] state. Journal of Chemical Physics, 2008. 128: p. 224310. Mamedov, F.D. and Salimov, M.A., Effect of hydrogen bonding on the position of absorption bands in ultraviolet spectra of ternary mixtures. Zhurnal Prikladnoi Spektroskopii, 1971. 14: p. 445-8.  

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.24.1-0062206/manifest

Comment

Related Items