Functional and inhibitory studies on cystathionine gamma lyase

FUNCTIONAL AND INHIBITORY STUDIES ON CYSTATHIONINE-γ-LYASE (CSE) HUANG SHUFEN (B. Sc. (Hons.), NUS) A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF SCIENCE DEPARTMENT OF CHEMISTRY NATIONAL UNIVERSITY OF SINGAPORE 2009 ACKNOWLEDGEMENTS I would like to thank my supervisor, Dr Tan Choon Hong, as well as Dr Deng Lih-Wen, Dr Jayaraman Sivaraman and Prof Philip Keith Moore for the use of their laboratory facilities and their invaluable guidance and support in this project. My appreciations also go towards Miss Guan Yanyi, Mr Zhao Yujun, Mr Fu Xiao, Mr Tan Yaw Sing and Miss Chua Jia Hui for their effort in synthesizing the inhibitor candidates, Mr Sun Qingxiang for his guidance on X-ray crystallography studies, Ms Wang Xiaoning, Ms Liu Jie and Mr Cheng Fei for their patience in guiding me on molecular cloning and immunoblotting techniques, as well as members of Dr Tan’s, Dr Deng’s, Dr Sivaraman and Prof Moore’s labs for their help in this project. Last but not least, I would like to express my gratitude towards my husband, Mr Bryan Lim for his constant encouragement and emotional support in my pursuance of the Masters degree. i TABLE OF CONTENTS Acknowledgements .............................................................................................................. i  Table of Contents ................................................................................................................ ii  Summary ............................................................................................................................. v  List of Publications ........................................................................................................... vii  List of Tables ................................................................................................................... viii  List of Figures .................................................................................................................... ix  List of Symbols ................................................................................................................ xiv  1. Introduction .................................................................................................................... 1  2. Tissue H2S assay for screening inhibitors of H2S production........................................ 5  2.1. Objectives ................................................................................................................ 5  2.2. Experimental ............................................................................................................ 5  2.3. Results and discussion ............................................................................................. 7  2.4. Conclusion ............................................................................................................. 13 3. Cloning and expression of recombinant human CSE .................................................. 14  3.1. Objectives .............................................................................................................. 14  3.2. Experimental .......................................................................................................... 15  3.2.1. Preparation of recombinant human CSE plasmids ......................................... 15  3.2.2. Mammalian expression of FLAG-tagged CSE ............................................... 16  3.2.3. Western blotting .............................................................................................. 16  3.2.4. Optimization of bacterial expression of human CSE ...................................... 17  3.2.5. Miniaturized assay of H2S production ............................................................ 18  3.2.6. Bacterial expression of human CSE enzyme .................................................. 19  3.2.7. Purification of human CSE enzyme ................................................................ 19  3.3. Results and discussion ........................................................................................... 22  3.3.1. Construction of recombinant human CSE plasmids ....................................... 22  3.3.2. Expression and determination of H2S synthesizing activity of recombinant human CSE ............................................................................................................... 23  3.3.3. Purification of human CSE enzyme ................................................................ 26  3.4. Conclusion ............................................................................................................. 32 ii 4. Purified protein assay screen for inhibitors of H2S production ................................... 33  4.1. Objectives .............................................................................................................. 33  4.2. Experimental .......................................................................................................... 33  4.2.1. Kinetics of H2S production from purified recombinant human CSE ............. 33  4.2.2. Protein H2S assay screen for inhibitors of H2S production............................. 34  4.2.3. Determination of H2S standard curve ............................................................. 35  4.2.4. Optimized protein H2S assay screen for inhibitors of H2S production ........... 36  4.3. Results and discussion ........................................................................................... 37  4.3.1. Kinetics of H2S production from purified recombinant human CSE ............. 37  4.3.2. Protein H2S assay screen of various commercially available inhibitor candidates .................................................................................................................. 39  4.3.2. Protein H2S assay screen of various chemically synthesized inhibitor candidates .................................................................................................................. 42  4.3.4. Optimization of protein H2S assay screen ...................................................... 44  4.3.5. Redetermination of H2S standard curve .......................................................... 49  4.3.6. Redetermination of inhibition potency of inhibitor candidates ...................... 50  4.4. Conclusion ............................................................................................................. 54  5. Elucidation of the three-dimensional structure of human CSE ................................... 55  5.1. Objectives .............................................................................................................. 55  5.2. Experimental .......................................................................................................... 55  5.2.1. Determination of protein homogeneity via DLS experiments ........................ 55  5.2.2 Screening and optimization of crystallizing conditions ................................... 56  5.2.3. X-ray diffraction and structure determination ................................................ 56  5.2.4. Proteolytic cleavage of CSE ........................................................................... 57  5.3. Results and discussion ........................................................................................... 59  5.3.1. Optimization of protein concentration for crystallization studies................... 59  5.3.2. Screening and optimization of crystallizing conditions for CSE .................... 60  5.3.3. Screening and optimization of crystallizing conditions for CSE-inhibitor complexes ................................................................................................................. 64  5.3.4. Proteolytic cleavage of CSE ........................................................................... 69  5.4. Conclusion ............................................................................................................. 71  6. Mechanism of H2S production ..................................................................................... 73  Objectives ..................................................................................................................... 73  Experimental ................................................................................................................. 76  6.2.1. Assay of H2S synthesis from various in vivo sulfur-containing compounds.. 76  6.2.2. Cloning of pET-22b(+)_CSE .......................................................................... 76  6.2.3. Bacterial expression and purification of polyhistidine-tagged (His-tagged) CSE ........................................................................................................................... 76  6.2.4. Optimization of bacterial induction conditions for His-tagged CSE .............. 77  6.2.5. Preparation of mutant CSE clones .................................................................. 78  6.2.6. Bacterial expression of mutant GST-tagged CSE proteins ............................. 78  iii 6.2.7. Optimized procedure for purification of GST-tagged mutant and wild-type CSE ........................................................................................................................... 79  6.2.8. Analysis of protein secondary structure via circular dichroism (CD) measurements ............................................................................................................ 80  6.2.9. Comparison of the H2S synthesizing activities of the CSE mutant proteins .. 80  6.2.10. Kinetics of H2S production under varying exogenous PLP concentrations . 80  6.3. Results and discussion ........................................................................................... 82  6.3.1. Assay of H2S synthesis from various in vivo sulfur-containing compounds.. 82  6.3.2. Cloning of pET-22b(+)-CSE .......................................................................... 84  6.3.3. Bacterial expression and purification of His-tagged CSE .............................. 85  6.3.4. Preparation of mutant CSE clones .................................................................. 86  6.3.5. Bacterial expression and purification of mutant GST-tagged CSE proteins .. 87  6.3.6. Analysis of protein secondary structure via CD measurements ..................... 89  6.3.7. Comparison of the H2S synthesizing activities of the CSE mutant proteins .. 91  Mutant CSE proteins affecting the binding of PLP cofactor ................................ 92  Mutant CSE proteins affecting the activation of L-cysteine substrate.................. 95  Mutant CSE proteins affecting the affinity of the enzyme for L-cysteine.............. 99  6.3.8. Kinetics of H2S production in the presence of varying PLP concentrations 102  6.3.9. Proposed mechanism for catalysis of H2S production by human CSE ......... 107  6.4. Conclusion ........................................................................................................... 112  7. Development of a polyclonal antibody specific towards human CSE ....................... 114  7.1. Objectives ............................................................................................................ 114  7.2. Experimental ........................................................................................................ 115  7.2.2. Immunoprecipitation (IP) of endogenous CSE using rabbit antibody serum 115  7.2.3. Purification of anti-hCSE 1366..................................................................... 116  7.2.4. Probing for endogenous CSE levels in various cell lysates .......................... 116  7.2.5. Immunoprecipitation of endogenous CSE using purified anti-hCSE antibody ................................................................................................................................. 117  7.3. Results and discussion ......................................................................................... 118  7.3.1. Testing of anti-hCSE sera ............................................................................. 118  7.3.2. Purification of antibody serum from rabbit 1366 ......................................... 119  7.3.3. Characterization of purified anti-hCSE antibody ......................................... 120  7.4. Conclusion ........................................................................................................... 123  8. Concluding remarks ................................................................................................... 125  Biblography..................................................................................................................... 127  Appendix 1: Forward and reverse primers used for PCR amplification of CSE ............ 132  Appendix 2: Mutagenic primers used for thermal cycling of mutant strands................. 133  Appendix 3: Mechanism for H2S production as proposed in the Honors project. .......... 135  iv SUMMARY In recent years, increased interest has been directed towards hydrogen sulfide (H2S) as a third gasotransmitter and its role in various neurodegenerative and cardiovascular diseases. Cystathionine-γ-lyase (CSE) is one of the two enzymes believed to be responsible for the endogenous production of H2S. Research has also shown that inhibitors of H2S production are effective in reducing the severity of certain diseases caused by increased endogenous H2S levels. However, these established inhibitors of CSE exhibit low potency, low selectivity and poor cell-membrane permeability. As such, we aimed to develop more specific and potent inhibitors of CSE towards H2S production. To achieve this, various inhibitor candidates were synthesized and tested using a previously established rat liver homogenate assay. An expression and purification system for the human CSE enzyme was also developed to enable a more reliable method of screening of the inhibitor candidates via a purified protein assay, which was optimized for more efficient trapping of evolved H2S in this work. The X-ray crystal structures of the enzyme in its apo- and holoenzyme forms, as well as in complex with one of its inhibitors have also been determined to aid in future rational design of inhibitors. Initial attempts to co-crystallize the enzyme with some of our inhibitor candidates were also performed in this work. Although CSE has been well-known for its role in the transsulfuration pathway, the biochemical role of the enzyme in production of H2S is currently not well understood. Hence, we were also interested in the mechanism for CSE-mediated H2S production. This was achieved via site-directed mutagenesis and kinetic studies on the enzyme. The in v vitro release of H2S from various sulfur-containing compounds present in our bodies was also tested using our purified protein assay. Through these studies, not only were we able to propose a more detailed mechanism for the catalysis of H2S production, we were also able to identify crucial residues which may directly affect the binding of inhibitors as well as certain key functional groups and their distribution within the inhibitor to allow for increased binding affinity to the enzyme. Lastly, a polyclonal rabbit antibody that is specific towards human CSE was also developed to serve as a platform for future studies of the function of the enzyme at the cellular level. vi LIST OF PUBLICATIONS Sun, Q., Collins, R., Huang, S., Holmberg-Schiavone, L., Anand, G. S., Tan, C. H., vanden-Berg, S., Deng, L. W., Moore, P. K., Karlberg, T., and Sivaraman, J. (2009). Structural Basis for the inhibition mechanism of human cystathionine-gamma-lyase: An enzyme responsible for the production of H2S. Journal of Biological Chemistry , 284 (5), 3076-3085. Huang, S., Chua, J. H., Sivaraman, J., Tan, C. H., & Deng, L. W. (2009). Site-directed mutagenesis and kinetic studies on human cystathionine-gamma-lyase reveal interesting insights into the mechanism of H2S production. Paper in preparation. vii LIST OF TABLES Table 1. Percentage inhibition levels for various commercially available compounds assayed at 10 mM L-cysteine, 2 mM PLP and 5 mM test compound concentrations unless otherwise stated................................................................................................................... 8 Table 2. Percentage inhibition levels for various chemically synthesized test compounds assayed at 10 mM L-cysteine substrate, 2 mM PLP and 5 mM test compound concentrations unless otherwise stated. .............................................................................. 8 Table 3. Effect of various buffers on the polydispersity index of the protein solution, as measured by DLS at a protein concentration of 1mg/mL and a temperature of 20 °C. ... 29 Table 4. Effect of increasing sodium chloride concentrations on the polydispersity index of the protein solution. ...................................................................................................... 29 Table 5. A comparison of the kinetic parameters of human CSE when utilizing L-cysteine or L-cystathionine as substrate.......................................................................................... 39 Table 6. Observations upon addition of a mixture of ZnAc and varying concentrations of NaOH, and TCA in the presence of ZYJ4291 as a test compound................................... 48 Table 7. Effect of varying protein concentration on the polydispersity index of the protein. ........................................................................................................................................... 60 Table 8. Data collection and refinement statistics for crystallized human CSE enzyme. 61 Table 9. Correlation between logP values and production of H2S for the various mutated amino acids at 339th position of human CSE .................................................................. 101 Table 10. Relative levels of endogenous CSE in various cell lines. ............................... 122  viii LIST OF FIGURES Figure 1. Inhibition profiles and IC50 values of (A) PAG, (B) BCA, (C) N-Boc-L-cysteine and (D) N-Cbz-D-cysteine determined in the presence of 10 mM L-cysteine substrate... 12 Figure 2. (A) PCR amplification of human CSE for subsequent cloning into pGEX-4T-3, pcDNA3.1(+) and p3xFLAG-CMV-10. (B) DNA gel electrophoresis of restriction enzyme cleaved recombinant CSE plasmids. ................................................................... 22 Figure 3. Western blot analysis of the expression of FLAG-tagged human CSE in 293T cells transfected with recombinant pcDNA3.1(+)-FLAG-CSE and p3xFLAG-CMV-10CSE plasmids. ................................................................................................................... 23 Figure 4. (A) 10 % SDS-PAGE gel showing expression of GST-CSE fusion protein (~66 kDa) under different induction conditions. (B) 10 % SDS-PAGE of total (T), soluble (S) and insoluble (I) fractions of cell lysates from bacteria induced for 3 h at 30 °C or 18 h at 18 °C. ................................................................................................................................ 24 Figure 5. H2S synthesizing activities (expressed as nmol H2S produced per mg total protein) of rat liver homogenate, lysates of 293T cells transfected with pcDNA3.1(+)FLAG-CSE or p3xFLAG-CMV-10-CSE and lysates of bacterial cells transformed with pGEX-4T-3-CSE induced under various conditions. ....................................................... 25 Figure 6. 10 % SDS-PAGE analysis of affinity purification and thrombin cleavage of GST-CSE. ......................................................................................................................... 27 Figure 7. (A) Anion exchange profile of the affinity pure CSE enzyme. (B) Gel filtration profile of the protein after ion exchange chromatography................................................ 31 Figure 8. 10 % SDS-PAGE (A) and 6 % native-PAGE (B) gels of the peak gel filtration fractions and the final purified CSE enzyme after gel filtration chromatography . .......... 32 Figure 9. Relationship between amount of H2S produced in 30 min against amount of purified recombinant CSE added in the presence of 10 mM L-cysteine substrate. .......... 37 Figure 10. (A) Graph of initial reaction velocity against L-cysteine substrate concentration in the presence of 2 mM PLP. (B) Logarithmic Hill plot of lg(V/(Vmax-V)) against lg[S]. ..................................................................................................................... 38 Figure 11. Average percentage inhibition values of various L-cysteine analogues, BCA and PAG assayed at 10 mM concentration in the presence of 10 mM L-cysteine substrate, 2 mM PLP and 10 µg purified CSE. ................................................................................. 40 ix Figure 12. Inhibition profiles and IC50 values of (A) N-acetyl-L-cysteine, (B) Nisobutyryl-L-cysteine, (C) BCA and (D) PAG determined in the presence of 5 mM Lcysteine substrate, 2 mM PLP and 5 µg purified CSE...................................................... 42 Figure 13. Averaged percentage inhibition values of various synthesized inhibitor candidates assayed at 0.1 mM, 1 mM or 5 mM concentration in the presence of 2.75 mM L-cysteine substrate, 2 mM PLP and 5 µg purified CSE (candidates 1 to 6) or 7.5 µg purified GST-CSE (candidates 7 to 9). ............................................................................. 44 Figure 14. Net A670 readings reflecting the distribution of trapped sulfides when different amounts of NaOH were added together with ZnAc.......................................................... 47 Figure 15. Relationship between absorbance at 670 nm and amount of H2S produced. .. 50 Figure 16. Average percentage inhibition values of various synthesized inhibitor candidates assayed at 2.5 mM concentration in the presence of 2.75 mM L-cysteine substrate, 0.5 mM PLP and 7.5 µg GST-CSE. ................................................................. 51 Figure 17. DLS profile and parameters for the purified CSE protein at 5.0 mg/mL. ....... 60 Figure 18. Optimization of CSE crystallizing condition for X-ray diffraction and subsequent structure determination................................................................................... 61 Figure 19. Asymmetric units of human CSE determined in this work (A) and by our collaborator (B). ................................................................................................................ 62 Figure 20. (A) Electron density map around PLP in the CSE holoenzyme. Significant differences in stereo-overlay of peptide chains around the Tyr-114 (B) and Lys-212 (C) residues, shown in green (our structure) and yellow (collaborator’s structure). .............. 63 Figure 21. Absorbance spectra of our purified CSE enzyme before and after L-cysteine incubation, and upon a readdition of equimolar amount of PLP. ..................................... 64 Figure 22. A closed-up view of the active site region of the CSE-PAG complex. ........... 65 Figure 23. Proposed mechanism for the inhibition of CSE by PAG. ............................... 66 Figure 24. IC50 analysis on the inhibition of H2S production from the Y114F mutant CSE protein by PAG. ................................................................................................................ 67 Figure 25. (A) Spherulites of CSE complexed with 5 mM N-isobutyryl-D-cysteine formed in 0.1 M BICINE pH 9, 20 % PEG 6000; (B, C) Needles formed around spherulites 2 weeks later. .................................................................................................. 68 Figure 26. (A) Crystal of CSE complexed with 5 mM N-isobutyryl-L-cysteine in 0.1 M BICINE pH 9, 20 % PEG 6000, 10 mM ZnCl2; (B, C) ZnCl2 crystals at bottom of well.69 x Figure 27. 10 % SDS-PAGE gel analysis on the proteolytic cleavage of CSE. ............... 70 Figure 28. Optimization of the proteolysis of CSE by chymotrypsin............................... 71 Figure 29. Alignment of the amino acid sequences of mouse, rat, human, Dictyostelium (slime mold), yeast and Streptomyces CSE as well as E. coli cystathionine-γ-synthase (CGS) and cystathionine-β-lyase (CBL)........................................................................... 74 Figure 30. Active site of the human CSE enzyme showing the location of crucial amino acids (only side chains shown) which would be studied by site-directed mutagenesis. ... 75 Figure 31. A comparison of the net amount of H2S produced over 30 min by 5 mM of various sulfur-containing compounds. .............................................................................. 82 Figure 32. (A) PCR amplification of CSE for cloning into pET-22b(+). (B) Restriction enzyme cleaved plasmids indicating the presence of CSE insert which was determined to be correct upon sequencing. .............................................................................................. 85 Figure 33. 10 % SDS-PAGE gel analysis showing attempted expression of His-tag CSE in the presence of 0.1 mM IPTG at 20 °C for 20 h (A) and optimization of bacterial expression conditions (B). ................................................................................................ 86 Figure 34. 0.8 % agarose gel showing PCR amplification of various mutant pGEX-4T-3CSE plasmids. ................................................................................................................... 87 Figure 35. 10 % SDS-PAGE analysis of the induction of GST-tagged mutant and wildtype CSE proteins. ............................................................................................................ 88 Figure 36. Proportion of α-helices, β-sheets, turns and unordered regions of GST-tagged mutant and wild-type CSE proteins. ................................................................................. 89 Figure 37. Distances (in angstroms) between the polar contacts of the carboxylic acid side chain of Glu-157 and amino group of Tyr-114 in the CSE holoenzyme (A) and apoenzyme (B). ................................................................................................................. 91 Figure 38. A comparison of the H2S synthesizing abilities from 5 µg of various GSTtagged CSE alanine mutants against wild-type GST-CSE. .............................................. 91 Figure 39. H2S synthesizing activities of mutant CSE proteins with mutagenic alterations to the Lys-212, Tyr-114, Asn-161 and Phe-190 residues, assayed in the presence of 5 µg GST-tagged protein, 0.5 mM PLP and 2.75 mM L-cysteine. ........................................... 94 Figure 40. H2S synthesizing activities of mutant CSE proteins with mutagenic alterations to the Tyr-60, Arg-62, Ser-340 and Thr-189 residues, assayed in the presence of 5 µg GST-tagged protein, 0.5 mM PLP and 2.75 mM L-cysteine. ........................................... 96  xi Figure 41. (left) The human CSE tetramer made up of a dimer of dimers. (right) Magnification of the interactions between PLP and Tyr-60 and Arg-62 from the adjacent monomer in subunits C and D of the enzyme. .................................................................. 97 Figure 42. (left) Schematic representation of the hydrogen bonding network involving Thr-189, Asp-187 and the PLP cofactor. (right) H2S synthesizing activities of mutant CSE proteins with mutagenic alterations to the Asp-187 residue, assayed in the presence of 5 µg GST-tagged protein, 0.5 mM PLP and 2.75 mM L-cysteine................................ 98 Figure 43. H2S synthesizing activities of mutant CSE proteins with mutagenic alterations to the Glu-339 residue. .................................................................................................... 101 Figure 44. Ionic interactions involving Arg-375 before (left) and after (right) the binding of L-cysteine substrate. ................................................................................................... 102 Figure 45. Amount of H2S produced in 30 min under different exogenous concentrations of PLP, assayed in the presence of 7.9 µg GST-tagged CSE and 2.75 mM L-cysteine. 103 Figure 46. (A) Graphs of initial reaction velocity, V against concentration of exogenous PLP determined under various concentrations of L-cysteine substrate. (B) Graphs of initial reaction velocity, V against L-cysteine substrate concentration for the various concentrations of PLP that was added in the assay......................................................... 105 Figure 47. (A) Double reciprocal plots for the various concentrations of L-cysteine substrate that was added in the assay. (B) Secondary plot for determination of the true Vmax and KM values for L-cysteine. ................................................................................. 106 Figure 48. (A) Double reciprocal plots for the various concentrations of exogenous PLP that was added in the assay, up to 75 µM. (B) Secondary plot for determination of the true Vmax and KM values for PLP. ................................................................................... 107 Figure 49. Proposed mechanism for the catalysis of H2S production from L-cysteine by human CSE. .................................................................................................................... 109 Figure 50. Amount of H2S produced in 30 min under different exogenous concentrations of PLP, assayed in the presence of 5 µg GST-tagged Y114F mutant CSE and 2.75 mM Lcysteine. .......................................................................................................................... 112 Figure 51. Probing of different amounts of pure CSE (A) and endogenous CSE from HepG2, 293T and 5 % w/v rat liver homogenate (B) with anti-hCSE sera from either rabbit 1365 or rabbit 1366. (C) Immunoprecipitation of endogenous CSE from 293T cell lysates by utilizing antibody serum of either rabbit 1365 or 1366. ................................ 119 Figure 52. Chromatograph of eluted anti-hCSE from HiTrap Protein A column and volume of 1 M Tris pH 9.0 base needed to neutralize various fractions from the blank run. ......................................................................................................................................... 120  xii Figure 53. Western blot on purified CSE and endogenous CSE levels in various homogenates or lysates utilizing anti-hCSE antibody from different immunization batches............................................................................................................................. 121 Figure 54. A comparison of the relative endogenous CSE levels among various cell lysates, normalized against β-actin. ................................................................................ 122 Figure 55. Immunoprecipitation of endogenous CSE from 293T, HepG2, K562 and U937 cells. ................................................................................................................................ 123  xiii LIST OF SYMBOLS Symbol Significance A670 Absorbance at 670nm BCA β-cyanoalanine BME β-mecaptoethanol CD Circular dichroism CSE Cystathionine gamma lyase DLS Dynamic light scattering DTT Dithiothreitol ECL Enhanced chemiluminescence GST Gluthathione S-transferase His-tag Polyhistidine-tag HRP Horse raddish peroxidase IP Immunoprecipitation IPTG Isopropyl-β-D-thiogalactopyranoside LB Luria-Bertani LB-Amp100 LB broth supplemented with 100μg/mL of ampicillin NNDPD N,N-dimethyl-p-phenylenediamine dihydro-chloride PAG DL-propargylglycine PLP Pyridoxal 5’-phosphate SDS-PAGE Sodium dodecyl sulfate-polyacrylamide gel electrophoresis TBS Tris buffered saline xiv TCA Trichloroacetic acid V Initial reaction velocity ZnAc Zinc acetate xv 1. Introduction Cystathionine-γ-lyase (CSE, EC 4.4.1.1), an enzyme found in mammals and some fungi, is involved in the reverse transsulfuration pathway (Scheme 1) where L-methionine is converted to L-cysteine through a series of metabolic interconversions (Rose & Wixom, 1955). Specifically, the role of CSE in this reaction pathway is to convert L-cystathionine to L-cysteine whilst generating α-ketobutyrate and ammonia. The reaction proceeds via an α, γ-elimination mechanism where the C-γ-S bond of L-cystathionine is specifically cleaved to yield L-cysteine (Flavin, 1971). A defect in this metabolic pathway has been found to be associated with cystathioninuria as well as L-cysteine deficiency and subsequent impairment of glutathione metabolism (Uren, Ragin, & Chaykovsky, 1978, Vina, et al., 1995). In humans, CSE activity was detected in adult liver tissue but not that in fetal liver (Sturman, Gaull, & Raiha, 1970). This was however attributed to posttranslational regulation of CSE gene expression in the developing human liver (Levonen, et al., 2000). Studies by Levonen et al. (2000) had also identified two isoforms of CSE as splice variants of one another; the longer being enzymatically more active than the shorter. Structurally, CSE is composed of four identical monomers of approximately 45kDa with a covalently bound pyridoxal 5’-phosphate (PLP) cofactor in each monomer. The crystal structure of the enzyme was first depicted from yeast CSE by X-ray crystallography (Messerschmidt, et al., 2003). In Messerschmidt’s study, factors affecting the enzymatic specificity of the various transsulfuration enzymes had also been discussed. Recently, studies on the single nucleotide polymorphic variant of human CSE (c.1364G>T, Ser403ÆIle) which had previously been found to be correlated with higher plasma homocysteine concentrations under homozygous conditions (Wang, Huff, 1 Spence, & Hegele, 2004) had shown that the PLP content and steady-state kinetic properties of the polymorphic enzyme was similar to that for the normal Ser403 variant (Zhu, Lin, & Banerjee, 2008). Experiments on the Thr67ÆIle and Gln240ÆGlu mutant CSE proteins which had previously been identified to lead to cystathioninuria (Wang & Hegele, 2003) had also revealed that the affinity of the mutant enzymes for PLP was weakened and that the enzyme activity could be restored by exogenous PLP in this study (Zhu, Lin, & Banerjee, 2008). Scheme 1. Reverse transsulfuration pathway present in mammals and fungi. NH3 OOC S β γ α COO NH3 L-cystathionine L-serine L-methionine H2O Cystathionine γ-lyase (CSE) HS COO NH3 L-cysteine COO + O + NH3 α-ketobutyrate Cystathionine β-synthase (CBS) L-homocysteine Besides the primary role of the enzyme in the conversion of L-cystathionine to L-cysteine, studies have also shown that rat liver CSE can utilize L-cysteine as a substrate for producing H2S gas (Stipanuk & Beck, 1982; Braunstein & Goryachenkova, 1984). This gas which had previously been primarily regarded as an environmental hazard and toxic gas, has recently been recognized as a third gasotransmitter besides carbon monoxide and nitric oxide (Wang, 2002). Other than CSE which accounts for endogenous production of H2S in the liver, kidney, intestine and vascular smooth muscle cells, the in vivo production of H2S has also been attributed to cystathionine-β-synthase (CBS) and 3mercaptopyruvate sulfurtransferase for brain and heart tissues respectively (Kamoun, 2004). Studies by Yang et al. (2008) had specifically shown H2S as a physiologic 2 vasorelaxant and that prononced hypertension was triggered in CSE knockout mice models due to the absence of in vivo H2S production. Characterization of a novel, slowreleasing H2S compound, GYY4137 under in vivo conditions had also shown that the vasorelaxant effect of H2S occurs via the opening of vascular smooth muscle KATP channels (Li, et al., 2008). Imbalances in the endogenous H2S levels have therefore been associated with various diseases such as Alzheimer’s disease (Eto, et al., 2002), pulmonary hypertension (Li, et al., 2005), haemorrhagic shock (Mok, et al., 2004), carrageenan-induced hindpaw oedema (Bhatia, et al., 2005a), acute pancreatitis (Bhatia, et al., 2005b) and endotoxemia (Collin, et al., 2005). In addition, H2S donors such as sodium hydrosulfide or GYY4137 as well as inhibitors of H2S production such as DLpropargylglycine (PAG) and β-cyanoalanine (BCA) have been found to exhibit therapeutic potential in various disease models where the severity of the diseases were found to be alleviated upon administration of these compounds (Szabó, 2007; Li, et al., 2008). H2S donor compounds and inhibitors of H2S biosynthesis may hence provide insights into the mechanisms underlying various diseases, or function as therapeutic drugs. Currently, the two commercially available inhibitors of H2S production, PAG and BCA possess low potency, low selectivity and limited cell-membrane permeability characteristics (Szabó, 2007). Therefore, there is a need to develop more effective inhibitors of H2S production. To achieve our aim, various L-cysteine and L-cystine analogues would first be tested using a rat liver homogenate assay. However, due to many problems with this assay, we proceeded to develop a pure protein assay by cloning, expressing and purifying the 3 human CSE enzyme for subsequent screening of inhibitor candidates. The purified protein would also be utilized for X-ray crystallography studies for elucidation of the three-dimensional structure of human CSE so as to aid in the rational design of inhibitors of H2S production in future. In addition, the expressed protein would also enable us to further explore the functional role of CSE in the catalysis of H2S production which is currently not well understood, as well as gain further insights into the mechanism for production of H2S. These would be achieved via site-directed mutagenesis and kinetic studies. Lastly, a polyclonal antibody which is specific towards the human CSE enzyme would be developed and characterized so as to serve as a platform for future functional studies on this protein. 4 2. Tissue H2S assay for screening inhibitors of H2S production 2.1. Objectives As mentioned in the introduction, there lies a need in developing more selective and potent inhibitors of H2S production since the current commercially available inhibitors, PAG and BCA are relatively weak and less selective. In this section which had been accomplished during the UROPS project, various commercially available and chemically synthesized L-cysteine analogues would be tested for their inhibition levels towards H2S production from rat liver homogenates. Drawbacks of the strategy used for inhibitor design as well as the tissue H2S assay screen would also be discussed. 2.2. Experimental A spectrophotometric assay modified from that described by Stpanuk and Beck (Stipanuk & Beck, 1982) was used for assaying the production of H2S from rat liver homogenates. All experiments on intact animals were undertaken with adherence to guidelines from the local National University of Singapore Institutional Animal Care and Use Committee (IACUC). Upon killing the rats, the livers were removed, cut into small pieces and kept frozen at -80 °C prior to the assay. For each assay, a small portion of the rat liver was thawed and homogenized in appropriate amounts of ice-cold 100 mM KHPO4 pH 7.4 buffer. Stock solutions of PLP and L-cysteine were prepared in 100 mM KHPO4 pH 7.4 buffer. For each test, a duplicate and a baseline control were performed in 1.5 mL cryovial tubes. A negative control experiment without addition of any test compound was also performed. Test compounds were either purchased from commercial sources or chemically synthesized. Trichloroacetic acid (TCA, 10 % w/v, 250 µL) was first added to 5 only the baseline control tubes to stop enzymatic reactions immediately by denaturing protein once the liver homogenate was added. This was followed by the sequential addition of saline (10 μL for test compounds dissolved in 100 mM KHPO4 pH 7.4 buffer; 25 µL for test compounds dissolved in DMSO), PLP (50 mM, 20 μL), rat liver homogenate (5 % w/v, 430 μL), and the test compound (20 μL for compounds dissolved in KHPO4 buffer; 5 µL for compounds dissolved in DMSO) to each tube. For the negative control experiment, the same volume of solvent in which the test compound was dissolved was added instead of the test compound. The tubes were then vortexed and preincubated on ice for 30 min, after which L-cysteine substrate (10 mM, 20 μL) was added. The tubes were parafilmed tightly, gently vortexed and incubated in a 37 °C water bath for 30 min. After incubation, the tubes were cooled on ice. Zinc acetate (ZnAc, 1 % w/v, 250 μL) was added via needle to trap evolved H2S and all enzymatic reactions were stopped by addition of TCA (250 μL) via needle. After centrifuging at 4 °C, 10000 rpm for 2 min, N,N-dimethyl-p-phenylenediamine dihydro-chloride dye (NNDPD, 20 mM, 133 μL) in 7.2 M HCl and FeCl3 (30 mM, 133 μL) in 1.2 M HCl were added for development of methylene blue. The tubes were centrifuged at 4 °C again, at 12000 rpm for 4 min. 300 μL of the supernatant from each tube was loaded into a 96-well microplate, and the absorbance at 670 nm (A670) was measured. The amount of H2S produced was calculated against a calibration curve of sodium hydrosulfide (NaHS: 0-100 μM) and the percentage inhibition of each test compound was then determined. For determination of the IC50 values of potential inhibitors, the assay was performed in varying concentrations of the inhibitor. The IC50 value was then estimated from the graph of percentage inhibition versus inhibitor concentration. 6 2.3. Results and discussion From the commercially available compounds which were tested, N-isobutyryl-L-cysteine, N-isobutyryl-D-cysteine and N-acetyl-L-cysteine were some of the better inhibitors besides PAG and BCA which are the two established inhibitors of CSE (Table 1). Our strategy was hence to synthesize L-cysteine or D-cysteine analogues with modifications to the amino group. Compounds with substituents attached to the sulfhydryl group of Lcysteine were also synthesized and tested, but these were found to be poorer inhibitors compared to L-cysteine analogues with only their amino groups modified by the same or a similar group (Table 2). Modification to both amino groups of cystine also led to a decrease in inhibition potency. These results suggest that a free sulfhydryl group may play a crucial role in the binding of the compound to the enzyme’s active site. Although PAG and BCA do not possess this sulfhydryl functionality, they bind to the enzyme mechanistically through their amino group unlike these test compounds, which we believe would bind to the enzyme through other reversible or non-mechanistic means. As for the N-substituted urea or thiourea L-cysteine derivatives, the urea derivatives generally fared better than their corresponding thiourea derivatives. The inhibition levels were observed to increase when the electron-withdrawing property of the thiourea group was increased (from phenyl-thiourea to (3,5-difluoro)-thiourea to (3,5-bis- trifluoromethyl-phenyl)-thiourea), though such a trend could not be established for the corresponding urea groups. Initially, it was also postulated that D-cysteine rather than Lcysteine analogues, could be better inhibitors of H2S production since both N-isobutyrylD-cysteine and the Cbz-protected D-cysteine analogue exhibited higher inhibition levels than their corresponding enantiomer. However, this was not observed for the Boc7 protected D- and L-cysteine analogues. Due to the low availability and high cost of Dcysteine as a starting material for synthesis of D-cysteine analogues, subsequent cysteine analogues were all synthesized from L-cysteine. Table 1. Percentage inhibition levels for various commercially available compounds assayed at 10 mM L-cysteine, 2 mM PLP and 5 mM test compound concentrations unless otherwise stated. O O O S OH OH OH NH2 NH2 DL-propargylglycine 94.0 % NH2 DL-penicillamine S-methyl-L-cysteine 0.0 % O -21.9 %* O O NC OH NH2 Beta-Cyanoalanine 100.0 % N S OH NH2 S-β-(4-pyridylethyl)-L-cysteine 5.6 % HS OH NH2 DL-homocysteine -106.6 %* O O HS OH NH O HS HS OH NH O N-isobutyryl-L-cysteine O O N-isobutyryl-D-cysteine * * 17.3 % HS 19.9 % OH NH N-acetyl-L-cysteine 24.2 %* * Compounds assayed at 10 mM concentration. Table 2. Percentage inhibition levels for various chemically synthesized test compounds assayed at 10 mM L-cysteine substrate, 2 mM PLP and 5 mM test compound concentrations unless otherwise stated. O HS O O OH NH O N-Boc-L-Cysteine 38.8 % HS O O OH NH O N-Boc-D-Cysteine 11.9 % HS O OH NH N-pivaloyl-L-cysteine 8.1 % 8 O O S OH NH HO NH O O O O S S O HS OH N,S-dipivaloyl-L-cysteine N,N-dipivaloyl-L-cystine -5.3 % 2.0 % O HS O O NH O OH NH CF3 N-(trifluroacetyl)L-cysteine 13.5 %† O OH HS NH O OH NH O S Cl S-benzyl-L-cysteine N-(chloroacetyl)L-cysteine -28.0 %† N-(Butyryl)-L-cysteine 3.6 %† -6.4 % O O HS O O S OH NH2 OH O NH2 HS OH NH OH O O NH O S-(1-phenyl-ethanone)L-cysteine N-Cbz-L-cysteine 17.2 %^ 4.0 % O NH HO O S O S O OH NH O NH HO O 7.8 % S O S O OH O OH NH HS O OH NH N,S-di(benzoyl)-L-cysteine -28.3 %^ O S O NH HO S O Cl N,S-di(4-chlorobenzoyl)L-cysteine N-(2-naphthoyl)-L-Cysteine 8.0 % 14.1 %† OH NH O NH 31.5 % O O S O S N,N-dibenzoylL-cystine N,N-diCbzL-cystine Cl O O O O N-Cbz-D-cysteine 27.0 % O S OH O NH S O N,N-ditosyl-L-cystine 30.0 % 9 O HS O HS OMe N-cyclohexoyl-L-cysteine N-((4-Methoxy-phenyl)thiourea)-L-cysteine -23.3 % 2.2 % OH NH O NH NH O HS NH S OH O OH NH N-(phenyl-urea)L-cysteine 39.4 % O HS NH O NH HO S O S S O N-(phenyl-thiourea)L-cysteine N,N-(diphenylurea)-L-cystine 6.2 % OH NH N H F OH NH HS OH NH S NH NH F F N,S-Bis((3,5-Difluoro-phenyl)thiourea)-L-cysteine -18.9 % F N-((3,5-Difluoro-phenyl)thiourea)-L-cysteine 1.8 % O O S S F F 32.2 % O O S N-((3,5-Difluorophenyl)-urea)-L-cysteine HS -60.4 % S F OH NH HS O NH OH .HCl S-(phenyl-thiourea)L-cysteine hydrochloride -228.9 % NH F NH2 F O O S N H NH NH O S NH OH O HS OH NH O OH NH HS OH NH S NH NH F F 3C N-((2,6-Difluoro-phenyl)thiourea)-L-cysteine -152.7 % CF3 N-((3,5-Bis-trifluoromethylphenyl)-urea)-L-cysteine 35.6 % F 3C CF3 N-((3,5-Bis-trifluoromethylphenyl)-thiourea)-L-cysteine 42.3 % 10 O HS S O OH S NH NH N H O S S O N-((4-acetyl-phenyl)thiourea)-L-cysteine 23.9 % ^ † OH NH NH N,S-Bis((4-acetyl-phenyl)thiourea)-L-cysteine O 11.1 % Compounds assayed at 2.5 mM L-cysteine substrate and test compound concentrations Compounds assayed at 10 mM and 2.5 mM L-cysteine substrate and test compound concentrations respectively. Two of the more potent analogues that were synthesized, N-Boc-L-cysteine and N-CbzD-cysteine, besides the commercially available inhibitors PAG and BCA, were selected for determination of their inhibition profiles (Fig. 1). Although there were compounds which were more potent than these two analogues at 5 mM concentration (for example N((3,5-Bis-trifluoromethyl-phenyl)-thiourea)-L-cysteine and N-(phenyl-urea)-L-cysteine), their inhibition profiles could not be obtained since their solubilities became rather poor when more than 5 mM of these samples were assayed. From the IC50 values, N-Boc-Lcysteine is still a relatively weak inhibitor compared to PAG (IC50 = 0.2 mM) and BCA (IC50 = 0.1 mM). The IC50 value for N-Cbz-D-cysteine could not be determined since precipitation became eminent for inhibitor concentrations beyond 8 mM. 11 A B Percentage inhibition (%) Percentage inhibition (%) 120.0 100.0 80.0 60.0 40.0 20.0 0.0 0.0 120.0 100.0 80.0 60.0 40.0 20.0 0.0 2.0 4.0 6.0 8.0 Concentration of PAG (mM) Concentration of PAG (m M) 2.0 4.0 6.0 8.0 10.0 Concentration ConcentrationofofBCA BCA(mM) (m M) IC50 at 0.2 mM PAG IC50 at 0.1 mM BCA D 80.0 Percentage inhibition (%) Percentage inhibition (%) C 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 8.7 4.0 8.0 12.0 16.0 20.0 24.0 Concentration (mM) Concentrationof of Boc-L-cysteine Boc-L-cysteine (mM) IC50 at 8.7 mM Boc-L-cysteine 40.0 30.0 20.0 10.0 0.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 Concentration of of Cbz-D-cysteine (mM) Concentration N-Cbz-D-cysteine IC50: Cannot be determined; Rsq: 0.983 Figure 1. Inhibition profiles and IC50 values of (A) PAG, (B) BCA, (C) N-Boc-L-cysteine and (D) N-Cbz-D-cysteine determined in the presence of 10 mM L-cysteine substrate. A major problem encountered during the assay was attributed to the poor solubility of most inhibitor candidates. Although this could be alleviated by lowering the concentrations of the test compounds and substrate, the sensitivity of the assay would become compromised since only very little amounts of H2S was produced for the negative control tubes to which the inhibition levels of the compounds would be subsequently computed. As such, an optimized substrate and test compound concentration of 10 mM and 5 mM respectively was utilized for most experiments. Although the dissolution of the poorly soluble compounds in DMSO could also alleviate 12 the problem, some precipitation was still evident upon addition to the rat liver homogenate, particularly for those hydrophobic L-cysteine analogues. As there may be a possibility of inaccurate results due to any undesirable effects of DMSO, the volume of DMSO was kept to a minimum of 1 % v/v in the assay. Nevertheless, a negative control experiment where DMSO was added to this final concentration was included in each assay to ensure that the production of H2S had not been compromised. 2.4. Conclusion The random synthesis of L-cysteine analogues did not serve as a strategic way for the design of inhibitors of H2S production. Moreover, the tissue H2S assay that was used for screening had required large amounts of substrate and test compounds to yield large enough responses for accurate results. This was undesirable not only due to high costs but also false positives which may result due to non-specific inhibition caused by the aggregation of excessive amounts of the test compound to the target protein (McGovern et al., 2002). There was also a possibility of degradation of the substrate or test samples by other enzymes present in the rat liver homogenate, thus affecting the inhibition levels. A more efficient approach would be to base the design of the inhibitor upon the threedimensional structure of the protein, as well as to utilize pure protein instead of rat liver tissue homogenates in the assay. To achieve this, the human CSE gene would be cloned into various vectors for subsequent production of the purified enzyme. 13 3. Cloning and expression of recombinant human CSE 3.1. Objectives Due to the intrinsic drawbacks of the rat liver homogenate assay that was utilized in the preliminary screen for inhibitors of H2S production, an expression and purification system for the human CSE protein had been developed during the UROPS and Honors projects. Besides developing a better assay for the production of H2S and screening of inhibitor candidates, the success in establishing a purification system for human CSE also forms an important basis for the subsequent elucidation of the three-dimensional structure of the enzyme, kinetics and site-directed mutagenesis studies for expounding upon the mechanism for the catalysis of H2S production, as well as in developing a polyclonal antibody which is specific towards this protein. The human CSE gene was hence cloned into 3 different vectors as we would like to determine which would subsequently allow for the most efficient expression of protein. Although the bacterial expression vector pGEX-4T-3 was likely to achieve this aim, the expressed enzyme may not be in the active form since post-transcriptional and post-translational mechanisms are absent in bacterial cells. Mammalian expression vectors were hence included in our choice of vectors to which the human CSE gene would be cloned into. The various vectors that were used would also aid in purification of our protein subsequently. pGEX-4T-3, for instance, allows a GST-tag to be incorporated to the N-terminus of our protein. This tag not only aids in easy purification of the protein via affinity chromatography, but is also believed to improve the solubility of the fusion protein (Donald et al., 1988). The presence of three FLAG epitopes in the mammalian expression vector, p3xFLAG-CMV- 14 10 is also likely to increase the efficiency of purification due to strong interactions with the anti-FLAG antibody that could be used in the purification process. As the other mammalian expression vector pcDNA3.1(+) does not possess any tags for easy purification of the protein, a FLAG epitope would be designed in the forward primer used for PCR amplification of FLAG-CSE so that the expressed protein would subsequently possess the FLAG tag for easy purification. Upon expression of CSE from the various recombinant plasmids, the H2S synthesizing activity of the crude mammalian and bacterial cell lysates would be determined and an expression system which allows for an economical production of large amounts of the protein would be selected for further expression and purification of the enzyme. 3.2. Experimental 3.2.1. Preparation of recombinant human CSE plasmids Polymerase chain reaction (PCR) amplification on human full length CSE cDNA (GenBank accession no. BC015807) obtained from Open Biosystems (cat. no. MHS1010-73982) was performed using 30 PCR cycles (30 s at 94.0 °C, 30 s at 57.0 °C and 2 min at 72 °C). The primers used for the amplification process are listed in Appendix 1. 25 μL of each of the respective purified PCR products were then cloned into pGEX-4T-3 and pcDNA3.1(+) with EcoRI/XhoI sites, and into p3XFLAG-CMV-10 with EcoRI/KpnI sites. The constructs were fully sequenced (1st Base Pte Ltd) and found to contain the desired CSE inserts. 15 3.2.2. Mammalian expression of FLAG-tagged CSE 293T cells were grown at 37 °C in Dulbecco's modified Eagle's medium (DMEM, Sigma) supplemented with 10 % fetal bovine serum (Hyclone), 2 mM L-glutamine (Gibco), 100 units/ml penicillin (Gibco) and 100 μg/ml streptomycin (Gibco) in a humidified atmosphere containing 5 % CO2. For transfection of plasmid DNA into 293T, cells were first seeded onto two 6cm plates to reach a confluency of about 50 % the next day. The transfection mixture was prepared by adding 6 µL of FuGENE 6 transfection reagent and 2 µg of the plasmid DNA (pcDNA3.1(+)-FLAG-CSE, p3xFLAG-CMV-10-CSE and their corresponding mock vectors) to 94 µL of serum free DMEM medium, followed by incubation at room temperature for 15 min. The transfection mixture was then added dropwise to the plated cells, and cells were harvested two days later by trypsinization. For preparation of cell lysates, cells were first washed with phosphate buffered saline (PBS), pelleted by centrifugation, and subsequently lysed in appropriate volumes of mild lysis buffer (150 mM NaCl, 20 mM Tris-HCl pH 8.0, 1 % TritonX-100, 10 % w/v glycerol, 1 mM DTT) supplemented with protease and phosphatase inhibitors (2 mM PMSF, 4 µg/mL Leupeptin, 4 µg/mL Aprotinin, 2 µg/mL Pepstatin A, 2 mM Na3VO4, 10 mM NaF) by passing the mixture through 21 G needle on ice. After incubation on ice for an hour, the lysate was centrifuged at 13000 rpm for 15 min at 4 °C, and the supernatant aliquoted and kept at -80 °C until further use. 3.2.3. Western blotting Protein samples were first separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred onto PVDF membranes using 16 transfer buffer without SDS at 60 mA for 2 h. Blocking was performed by incubating the membrane in 3 % w/v non-fat skim milk prepared in Tris buffered saline (TBS) supplemented with 0.05 % Tween 20 (TBS/0.05 % Tween) at 4 °C overnight. The membrane was then incubated at room temperature (25 °C) for 1 h with the specific primary antibody prepared in 1.5 % non-fat skim milk prepared in 1x TBS/0.05 % Tween. Following that, the membrane was washed thrice with TBS/0.1 % Tween at room temperature for 10 min each on a bench top shaker. Incubation with the respective secondary antibody (prepared in 1.5 % non-fat skim milk in TBS/0.05 % Tween) was performed at room temperature for 1h, after which the membrane was washed thrice with TBS/0.1 % Tween at room temperature for 10 min each on a bench top shaker. Chemiluminescence analysis was then performed by incubating the membrane with enhanced chemiluminescence (ECL) substrates (PerkinElmer or Amersham) for 1 min at room temperature. The image was then developed with X-ray film (Amersham) for exposure. 3.2.4. Optimization of bacterial expression of human CSE 25 ng of the recombinant pGEX-4T-3-CSE plasmid was transformed into competent Escherichia coli (E. coli) BL21 cells. 4 colonies of the pGEX-4T-3-CSE transformed BL21 cells were inoculated into 20 mL of steam-autoclaved Luria-Bertani (LB) broth supplemented with 100 μg/mL of ampicillin (LB-Amp100) and incubated at 37 °C overnight with vigorous shaking. 7 mL of the starter culture and 7 mL of 20 % w/v glucose solution was propagated in 126 mL of LB-Amp100 for cultures to be incubated at 30 °C, while 5 mL of the starter culture and 5 mL of 20 % w/v glucose solution was 17 propagated in 90 mL of LB-Amp100 for cultures to be incubated at 18 °C. Batches of bacteria were then induced with 0.1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) at OD600 0.5 or 1.0. Protein expression was performed at 30 °C for 3 h or 6 h, or at 18 °C for 4 h or 18 h. The bacterial cells were then pelleted by centrifuging 6000 x g at 4 °C for 15 min, and subsequently lysed in lysis buffer (20 mM Tris pH 8.0, 150 mM NaCl, 1 mM EDTA, 2.5 mg/mL lysozyme and 1.5 % w/v sodium sarcosyl) supplemented with protease inhibitors (1 mM PMSF, 2 µg/mL Leupeptin, 2 µg/mL Aprotinin, 2 µg/mL Pepstatin A) for determination of protein solubility and enzyme activity. 3.2.5. Miniaturized assay of H2S production A miniaturized assay from that described in Section 2.2 was used to determine the amount of H2S produced from mammalian and bacterial cell lysates. 100 µL of 10 % w/v TCA was added initially to the baseline control tubes. The reaction mixture consisted of a smaller total volume of 300 µL and comprised saline (60 μL), PLP (30 mM, 20 μL), rat liver homogenate (2.5 % w/v) or either mammalian or bacterial cell lysates (200 μL) and L-cysteine (150 mM, 20 μL). The tubes were parafilmed tightly, gently vortexed, and then incubated in a 37 °C water bath for 30 min. After incubation, the tubes were cooled on ice. After addition of ZnAc (1 % w/v, 100 µL) and TCA (10 % w/v, 100 µL), the tubes were centrifuged at 10000 rpm for 2 min at 4 °C. NNDPD (20 mM, 71.4 μL) and FeCl3 (30 mM, 71.4 μL) were then added for development of methylene blue. The tubes were then treated as described above and the A670 readings of the supernatants were measured. 18 3.2.6. Bacterial expression of human CSE enzyme Four colonies of the pGEX-4T-3-CSE transformed BL21 cells were inoculated into 150 mL of steam-autoclaved Luria-Bertani (LB) broth supplemented with 100 μg/mL of ampicillin (LB-Amp100), and incubated at 37 °C overnight with vigorous shaking. 50 mL of the starter culture and 50 mL of 20% w/v glucose solution were propagated separately into each of two 2.5 L Erlenmeyer flasks, each containing 900 mL of LBAmp100. The mixture was incubated at 37 °C with vigorous shaking until an optical density at 600 nm (OD600) of 0.3 was attained, upon which bacterial growth was continued at 18 °C until an OD600 of 0.5 was attained. IPTG was then added at a rate of 0.1 mM to induce the expression of human CSE at 18 °C for 18 h. The cells were harvested by centrifugation in a Beckman JLA8.1000 rotor at 6000 rpm for 15 min at 4 °C. The pellets were combined and resuspended in 200 mL of the supernatant, divided into five portions, and centrifuged again at 8000 rpm for 15 min at 4 °C using an Eppendorf centrifuge (5804R) equipped with a F-34-6-38 rotor. The supernatants were removed, and the pellets (each containing 400 mL worth of the original cell culture) were kept at -80 °C until further use. 3.2.7. Purification of human CSE enzyme The bacterial pellet was thawed and resuspended in 40 mL of lysis buffer containing 50 mM Tris pH 8.0, 100 mM NaCl, 1 % Triton-X, 5 mM dithiothreitol (DTT), 1 mM PLP and 0.22 mg/mL of protease inhibitor mixture (Cat. no.: P8465, Sigma). Cell lysis was performed by sonication on ice using the 30 % pulsed maximum output of a Sonics Vibracell sonifier equipped with a macrotip. The lysate was then cleared by centrifuging 19 at 18000 rpm for 30 min at 4 °C using a Beckman JA25.50 rotor. The supernatant was introduced to a 100 mL chromatography column containing 5 mL of glutathione sepharose beads previously equilibrated with 40 mL of the lysis buffer. The column was sealed and incubated with slight shaking at room temperature (20 °C) for 30 min. Following that, the flow-through was collected, and the beads were washed once with 40 mL of 20 mM Tris/HCl pH 8.0, 100 mM NaCl, 1 % Triton-X, 1 mM DTT and then twice with 40 mL of 20 mM Tris/HCl pH 8.0, 100 mM NaCl, 1 mM DTT. On-column removal of the gluthathione S-transferase (GST) tag was then performed by adding 10 mL of cleavage buffer (50 mM sodium phosphate buffer pH 8.2, 100 mM NaCl, 1 mM DTT) containing between 5 to 20 units of thrombin protease (Sigma), incubating at room temperature for 15 min, and eluting the cleaved protein solution. This was repeated until minimal protein concentration was detected in the eluates. The eluates were then pooled together and concentrated using an ultra-centrifugal filter (Millipore Amicon Ultra-4 30000 MWCO) at 4 °C. The concentrated eluate was subsequently passed through a HiTrap Q HP anion exchange column (GE Healthcare) connected to a BioRad BioLogic Duo Flow FPLC equilibrated with buffer A (10 mM sodium phosphate buffer pH 8.2 and 1 mM DTT) at a flow rate of 1 mL/min. CSE was subsequently eluted at about 0.15 M NaCl when a linear 50 ml-gradient from 0 to 1 M NaCl at 1 mL/min was applied. Fractions containing CSE were identified by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and then pooled together. The pool was then passed through a Superdex-200 column connected to the BioRad BioLogic Duo Flow FPLC equilibrated with buffer A at a flow rate of 1 mL/min. Peak fractions corresponding to the target CSE protein were pooled together and concentrated to about 5 mg/mL in 10 mM 20 sodium phosphate buffer pH 8.2 and 1 mM DTT using an ultra-centrifugal filter (Millipore Amicon Ultra-4 10000 MWCO) at 4 °C. The protein was kept at 4 °C for short-term storage of up to two weeks, and at -80 °C for long-term storage. 21 3.3. Results and discussion 3.3.1. Construction of recombinant human CSE plasmids PCR amplification of the human CSE gene was rather efficient and gave high yield of the amplified PCR products. DNA gel electrophoresis (Fig. 2A) also showed that the PCR products were of the right size (~1.2 kb). Subsequent restriction enzymes digestion and ligation of the CSE inserts to their corresponding vectors were performed. Successful clones were identified by DNA gel electrophoresis which displayed a ~1.2 kb DNA band after digestion with its corresponding restriction enzymes (Fig. 2B). The accuracy of the CSE insert were fully sequenced and confirmed by DNA auto-sequencing. A (kb) 10.0 8.0 6.0 5.0 4.0 3.0 2.5 2.0 1.5 1.0 B pGEX-4T-3-CSE 1 2 3 4 pcDNA3.1(+)FLAG-CSE 1 2 3 4 p3xFLAGCMV-10-CSE 1 2 3 4 (kb) 10.0 8.0 6.0 5.0 4.0 3.0 2.5 2.0 CSE 1.5 DNA 1.0 CSE insert Figure 2. (A) PCR amplification of human CSE for subsequent cloning into pGEX-4T-3, pcDNA3.1(+) and p3xFLAG-CMV-10. (B) DNA gel electrophoresis of restriction enzyme cleaved recombinant CSE plasmids indicating the presence of CSE inserts which were subsequently sequenced and determined to be correct. (Adapted from Huang & Tan, 2006) 22 3.3.2. Expression and determination of H2S synthesizing activity of recombinant human CSE Transfection of recombinant pcDNA3.1(+)-FLAG-CSE and p3xFLAG-CMV-10-CSE into 293T cells was somewhat successful as can be observed from the presence of a band at approximately 43 kDa when compared to the mock transfection sample (Fig. 3). For pcDNA3.1(+)-FLAG-CSE, lesser amounts of the protein was expressed compared to that for p3xFLAG-CMV-10-CSE as the former transfection experiment had been performed as a trial practice. We hence believe that the transfection efficiency and protein yield would be improved in subsequent experiments. Alternatively, some optimization of the transfection protocol may be performed to obtain larger amounts of the desired protein. Transfection Lane – + – + 1 2 3 4 (kDa) 250 150 100 75 50 37 FLAG-tagged CSE 25 15 10 pcDNA3.1(+)FLAG-CSE p3xFLAG-CMV10-CSE Figure 3. Western blot analysis of the expression of FLAG-tagged human CSE in 293T cells transfected with recombinant pcDNA3.1(+)-FLAG-CSE and p3xFLAG-CMV-10CSE plasmids. Mock transfection of empty pcDNA3.1(+) and p3xFLAG-CMV-10 vectors are shown in Lanes 1 and 3 respectively. Western blot was performed by blocking the membranes at 4 °C overnight followed by incubation with anti-FLAG antibody (1:1000 dilution, Sigma). Horse raddish peroxidase (HRP)-conjugated antimouse antibody (1:10000 dilution, Amersham) was used as secondary antibody. Chemiluminescence was detected using enhanced luminol substrate from PerkinElmer. 23 The bacterial expression of CSE was also successful under various conditions as shown in Fig. 4A. Although a similar expression of CSE had been reported by Steegborn et al. (Steegborn, 1999), only about 10 % of the expressed protein was soluble in their study. Slowing the protein expression rate by utilizing lower concentrations of IPTG and lower expression temperatures did not improve the solubility of their protein significantly. In our study, we optimized the bacterial expression of CSE by varying the optical densities and temperatures to which induction was performed. Optimization of the bacterial induction conditions not only allowed us to explore which would give the highest yield of soluble protein, but also enabled us to determine which condition would allow for the expression of the most active enzyme. The results showed that all of the induction conditions experimented led to expression of large amounts of the fusion protein. There was also no significant difference in the solubility of the expressed protein under different induction conditions (Fig. 4B). B A Temperature 30°C 18°C Induction at OD600 = 0.5 OD600 = 1.0 OD600 = 0.5 Time after 0h 3h 6h 3h 6h 0h 4h 18h induction 0.1mM IPTG - - + - + - + - + - - + - + (kDa) 200 116 97.4 66.2 45.0 Temperature Induction at (kDa) 30°C OD600 = 0.5 OD600 = 1.0 T S I T S 18°C OD600 = 0.5 I T S I 200 116 97.4 GST- 66.2 CSE 45.0 GSTCSE 31.0 31.0 Figure 4. (A) 10 % SDS-PAGE gel showing expression of GST-CSE fusion protein (~66 kDa) under different induction conditions in the absence (-) and presence (+) of IPTG over 6 h at 30 °C and 18 h at 18 °C. (B) 10 % SDS-PAGE of total (T), soluble (S) and insoluble (I) fractions of cell lysates from bacteria induced for 3 h at 30 °C or 18 h at 18 °C. Broad molecular range protein ladder (Biorad) was loaded in Lane 1 for both gels. (Adapted from Huang et al., 2007) 24 Utilization of the above mammalian cell (transfected with pcDNA3.1(+)-FLAG-CSE or p3xFLAG-CMV-10-CSE) and bacterial cell lysates (from soluble lysate fractions of bacteria induced for 3 h at 30 °C or 18 h at 18 °C) in the miniaturized H2S assay showed that the expressed protein was active towards H2S production (Fig. 5). This was significant, particularly for the bacterial lysates since the results implied that the activity of the bacterial expressed GST-fusion protein was not compromised despite the addition of a GST-tag to the protein or the absence of post-transcriptional or post-translational modifications to the expressed protein in the bacterial cells. As large-scale production of the enzyme at relatively low costs was unachievable for mammalian expression systems, subsequent expression of the enzyme for inhibitory and functional studies was carried out using the bacterial expression system. Figure 5. H2S synthesizing activities (expressed as nmol H2S produced per mg total protein) of rat liver homogenate (Lane 1), lysates of 293T cells transfected with pcDNA3.1(+)-FLAG-CSE (Lane 3) or p3xFLAG-CMV-10-CSE (Lane 4) and lysates of bacterial cells transformed with pGEX-4T-3-CSE induced under various conditions (Lanes 5 – 10). Lanes 2, 5, 7 and 9 are the negative controls of lanes 3 and 4, 6, 8 and 10 respectively. Some H2S production was noted in the negative controls as well due to endogenous CSE in 293T cells and bacterial cystathionine-β-lyase which could utilize Lcysteine for production of H2S as well. 25 A comparison between the activities of rat liver homogenate or mammalian and bacterial cell lysates should not be made since the percentage of CSE may not be similar for all three types of lysates and there may also be other enzymes present which could utilize Lcysteine for H2S production. However, as the amount of expressed human CSE was approximately similar for the lysates of bacterial cells which were induced under different conditions (Fig. 4B), a justifiable comparison could be made between the activities of these lysates. From Fig. 5, it was noted that the greatest H2S synthesizing activity was achieved from protein expressed under low induction temperature for the bacterial expression system (Fig. 5, Lanes 5 – 10). This is in accordance with our original expectation since slower rates of protein expression would allow for folding of the polypeptide chain to a three-dimensional structure that is closer to that of the native protein. Our results may also possibly explain the low H2S synthesizing activity of the CSE enzyme prepared by Steegborn et al. since a higher induction temperature of 25 °C had been used in their experiment. This stresses the importance of expressing the CSE enzyme at low temperatures in maintaining the enzyme activity. Subsequent bacterial inductions were therefore performed at 18 °C. 3.3.3. Purification of human CSE enzyme Upon expression of the enzyme in the bacterial cells, the cells were lysed using high energy ultrasound waves to release the expressed proteins. The lysate was then cleared via centrifugation to separate the soluble proteins from bacterial debris as well as the insoluble proteins. As our enzyme was expressed as a fusion protein to GST, incubation 26 of the centrifugal supernatant with glutathione-coated sepharose beads would allow for separation of the fusion protein from other bacterial proteins which do not bind to the beads and are removed in the flow-through. SDS-PAGE analysis of the purification process (Fig. 6, lane 3) showed that about 90 % of our fusion protein was soluble. Some fusion protein was however present in the flow-through (lane 5), most probably due to insufficient glutathione sepharose beads or too short a binding time used. In our experiment, the amount of glutathione sepharose beads had been optimized to about 5 mL as too large a bed volume tended to increase non-specific binding of other bacterial proteins, as well as increase the time and amount of thrombin protease needed for cleavage of the GST-tags subsequently. Moreover, all steps had to be performed within a short period of time to avoid protein degradation by proteases since the purification process had been carried out at room temperature. The binding time was hence kept to 30 min since the amount of protein obtained subsequently was sufficient for further studies. (kDa) 1 2 3 4 5 6 7 8 9 10 97.4 66.2 45.0 GST-CSE CSE 31.0 GST 21.5 Figure 6. 10 % SDS-PAGE analysis of affinity purification and thrombin cleavage of GST-CSE. Protein bands at approximately 66 kDa, 45 kDa and 27 kDa correspond to monomeric GST-CSE, CSE and GST respectively. Lane 1: Low molecular range protein ladder (Biorad); lane 2: total protein of bacterial lysate; lane 3: soluble fraction of lysate; lane 4: insoluble fraction of lysate; lane 5: flow-through; lane 6: wash 1; lane 7: wash 2; lane 8: glutathione sepharose beads before cleavage of GST-CSE; lane 9: glutathione sepharose beads after cleavage of GST-CSE; lane 10: CSE eluates after cleavage of GSTtags (Adapted from Huang et al., 2007) 27 In the wash process, Triton-X served to remove any non-specifically bound bacterial proteins as well as solvate the hydrophobic regions of our target protein for easy dissolution upon removal of the GST-tags. A second wash buffer without addition of this detergent was needed to remove residual Triton-X which could inhibit the action of thrombin protease during the removal of the GST-tag from the enzyme. Upon cleavage of the fusion protein, the GST-tag would remain bound to the glutathione sepharose beads (Fig. 6, lane 9), while CSE would be eluted (Fig. 6, lane 10). This process was initially found to result in extensive protein precipitation which was to be expected since Steegborn’s work had revealed that the CSE enzyme was rather insoluble. Optimization of the cleavage process was hence performed by establishing a suitable buffer for our protein. To achieve this, the extent of protein aggregation upon preparation of the enzyme in various buffers was explored via dynamic light scattering (DLS) experiments (see Experimental section 5.1.1 for procedure). In these buffers, a buffer pH of 8.2 was employed since a pH difference of 2 units from the protein’s isoelectric point, which in our case is 6.2, often improves protein solubility. The protein has also been found to exhibit the highest activity at pH 8.2 in Steegborn’s work. From the polydispersity indices gathered from the DLS experiments, the protein was observed to be less homogenous in Hepes buffer than either Tris or sodium phosphate buffer (Table 3). As polydispersity indices which are smaller than 0.20 were more likely to result in 28 crystallization of the protein subsequently (Adrain et al., 1997), the buffers used in the purification process were either based on Tris or sodium phosphate. Table 3. Effect of various buffers on the polydispersity index of the protein solution, as measured by DLS at a protein concentration of 1 mg/mL and a temperature of 20 °C. (Adapted from Huang et al., 2007) Buffer condition 10 mM Hepes pH 8.2, 1 mM DTT 10 mM Tris/HCl pH 8.2, 1 mM DTT 10 mM Na3PO4 pH 8.2, 1 mM DTT Polydispersity index 0.22 0.16 0.11 Table 4. Effect of increasing sodium chloride concentrations on the polydispersity index of the protein solution, as measured by DLS at a protein concentration of 0.8 mg/mL and a temperature of 20 °C. (Adapted from Huang et al., 2007) Buffer condition 10 mM Na3PO4 pH 8.2, 1 mM DTT 10 mM Na3PO4 pH 8.2, 1 mM DTT, 50 mM NaCl 10 mM Na3PO4 pH 8.2, 1 mM DTT, 100 mM NaCl 10 mM Na3PO4 pH 8.2, 1 mM DTT, 200 mM NaCl 10 mM Na3PO4 pH 8.2, 1 mM DTT, 300 mM NaCl 10 mM Na3PO4 pH 8.2, 1 mM DTT, 400 mM NaCl Polydispersity index 0.10 0.09 0.12 0.10 0.23 0.26 The solubility of the protein can also be improved by the addition of salts to the purification buffers. An optimal salt concentration was hence determined by preparing the enzyme in sodium phosphate buffers containing different amounts of sodium chloride and measuring their polydispersity index values. The results showed that the extent of protein aggregation was fairly similar at sodium chloride concentrations of between 0 to 200 mM (Table 4). Increasing the salt concentrations above 300 mM was however undesirable due to a large increase in the polydispersity index. As such, the salt concentration was kept below 200 mM in the purification buffers. Upon optimization of the cleavage buffer for removal of the GST-tag, the extent of protein precipitation was greatly alleviated. To further minimize the occurrence of protein 29 precipitation, the removal of the GST-tag was performed in a staggered and controlled fashion by eluting the protein once a certain threshold protein concentration was met. The amount of thrombin protease used was also decreased gradually after each elution to slow down the rate of cleavage and prevent unwanted protein precipitation towards the end. From our denaturing SDS-PAGE analysis (Fig. 6, lane 9), it was confirmed that the cleavage process was more or less complete. The faint bands which appeared in the cleavage eluates (Fig. 6, lane 10) were due to some uncleaved fusion protein (~66 kDa), thrombin protease (~37 kDa) and GST-tags (~27 kDa) which had been eluted together with our target CSE enzyme (~45 kDa). Subsequent ion exchange and gel filtration steps served to further purify the protein. Our protein, being negatively charged at pH 8.2, would remain bound onto the positively charged ion exchange media and only be eluted at as the ionic strength of the buffer was increased. This occurred when a conductivity of 20 mS/cm was attained, which corresponded to a sodium chloride concentration of 150 mM (Fig. 7A). Due to some errors of the FPLC machine which caused a sudden increase in sodium chloride concentration gradient, a kink was noted in the major protein peak. The minor peak observed on the chromatograms corresponded to the impurities which had been eluted together with our protein in the previous purification step. 30 A 1.00 0.75 0.50 B 250.0 1.00 100.0% Buffer B 25.0 100.0% Buffer B 200.0 0.75 75.0 Major peak 0.50 0.00 0.0 50.0 10.0 100.0 Kink 25.0 15.0 150.0 50.0 0.25 20.0 Major peak 0.25 Shoulder 50.0 Minor peak 0.0 Minor peak 0.0 0.00 0.0 -5.0 -50.0 AU 0.00 20.00 40.00 Millilitres 60.00 80.00 mS/cm AU 5.0 0.00 20.00 40.00 60.00 Millilitres 80.00 100.00 120.00 mS/cm Figure 7. (A) Anion exchange profile of the affinity pure CSE enzyme. The black line shows the programmed NaCl gradient, while the red line shows the actual gradient reflected by changes in conductivity. Changes in absorbance at 280 nm (A280), which is related to the concentration of the eluted protein, is shown by the blue profile. Pooled fractions which would be subsequently purified by gel filtration are shown by the boxed region of the major peak. (B) Gel filtration profile of the protein after ion exchange chromatography. The blue profile shows changes in A280 value which is related to eluted protein concentrations, while the red profile shows changes in conductivity of the eluted protein. Fractions in the boxed region were analyzed by SDS and native PAGE and subsequently combined. (Adapted from Huang et al., 2007) The pooled fractions from the ion exchange major peak were then passed through the gel filtration column where proteins were separated from large to small ones. The chromatogram still revealed signs of impurities as depicted by the shoulder before the major peak, and the minor peak. Based on the position of these peaks, they most probably corresponded to uncleaved GST-fusion proteins and CSE dimers respectively. The position of the major peak also indicated that our purified enzyme was about 200 kDa, which was close to the actual molecular weight of 180 kDa. Upon confirmation of the purity of the individual major peak fractions via SDS and native PAGE analyses (Fig. 8), the fractions were pooled together and concentrated. The final purified enzyme was about 95 % pure, and about 40 mg of purified CSE protein was gathered from each litre of the original bacteria culture. 31 A (kDa) L 1 2 3 4 5 6 7 8 B 1 2 3 4 5 6 7 8 97.4 66.2 45.0 31.0 Figure 8. 10 % SDS-PAGE (A) and 6 % native-PAGE (B) gels of the peak gel filtration fractions (lanes 1 to 7) and the final purified CSE enzyme after gel filtration chromatography (lane 8). L: low molecular range protein ladder (Biorad) (Adapted from Huang et al., 2007) 3.4. Conclusion In this section, the human CSE gene was successfully cloned into three different vectors. The expression of the protein could be achieved from all three recombinant plasmids, and H2S synthesizing activity was also observed from all three lysates containing the expressed enzyme. However, as the bacterial expression system was more economical, large scale production of the human CSE enzyme was performed using the pGEX-4T-3CSE recombinant plasmid. Both expression and purification procedures have also been optimized in this work. Having obtained the purified CSE enzyme, the protein would be utilized in the next section for determining the kinetics of H2S production, re-determining the inhibition levels of various commercially available and chemically synthesized compounds and optimization of the H2S assay for screening of inhibitor candidates. 32 4. Purified protein assay screen for inhibitors of H2S production 4.1. Objectives As discussed in Chapter 2, there lies a need in developing a H2S assay that is based on the purified human CSE protein to enable a more accurate determination of the inhibition levels of the inhibitor candidates. This work which is performed in the Masters project, would be expounded in this chapter. Previously, studies by Steegborn et al. (1999) had also shown that human CSE displayed only marginal reactivity towards L-cysteine, the supposed substrate for H2S production. Although this reaction had been established for rat liver CSE, no reports had directly proven the catalytic ability of human CSE in converting L-cysteine to H2S. Therefore, we aimed to establish the ability of human CSE in catalyzing the production of H2S as well by exploring the kinetics of this reaction. 4.2. Experimental 4.2.1. Kinetics of H2S production from purified recombinant human CSE The production of H2S from 5 μg of the purified enzyme was determined at 3 min intervals over 30 min at 37 °C with slight modifications to the above assay described in Section 3.2.5. Stock solutions of PLP and L-cysteine were prepared in 10 mM sodium phosphate pH 8.2 buffer. Each reaction mixture, had a total volume of 100 μL, and contained saline (10 μL), PLP (10 mM, 20 µL), CSE (0.125 mg/mL, 40 µL) and sodium phosphate buffer pH 8.2 (10 mM, 30 µL) adjusted to final L-cysteine concentrations of between 0.75 mM to 3.5 mM. Development of methylene blue for absorbance studies 33 was achieved by addition of NNDPD dye (20 mM, 42.9 μL) and FeCl3 (30 mM, 42.9 μL). The initial reaction velocity (U per mg CSE, where 1 U = 1 μmol H2S produced per min) for each substrate concentration was then determined, plotted against the L-cysteine substrate concentration ([S]), and fitted to the Hill equation: V= Vmax [ S ]h K 0.5 + [ S ]h (1) The Hill coefficient, h was determined via the gradient of the logarithmic plot of the Hill equation, ⎛ V lg⎜ ⎜ Vmax − V ⎝ ⎞ ⎟ = h lg[S ] + constant ⎟ ⎠ (2) while the turnover number and catalytic efficiency were calculated using equations (3) and (4) respectively. kcat = Vmax [E]T Catalytic efficiency = (3) k cat k or cat KM K 0.5 (4) The graphical software, SigmaPlot (Systat) was used for all curve fitting and regression analyses. 4.2.2. Protein H2S assay screen for inhibitors of H2S production A similar assay to that performed for determination of the kinetics of H2S production was adapted for screening of inhibitor candidates which were synthesized by Mr Zhao Yujun (ZYJ), Mr Fu Xiao (FUX), Mr Tan Yaw Sing (TYS) and Miss Chua Jia Hui (CJH). Each reaction mixture consisted of 5 µg or 10 µg of the purified CSE enzyme, saline (10 µL), 34 PLP (10 mM, 20 µL), sodium phosphate pH 8.2 buffer, (50 mM, 40 µL), inhibitor candidate (prepared in 10 mM sodium phosphate pH 8.2 buffer, 25 % or 100 % v/v DMSO, 10 µL) and L-cysteine (100 mM, 50 mM or 27.5 mM, 10 µL), dissolved to a final volume of 100 µL. Each test compound was first preincubated with the enzyme on ice for 30 min before the L-cysteine substrate was added at 15 s intervals. The tubes were parafilmed tightly, gently vortexed, and then incubated in a 37 °C water bath for 30 min. After incubation, the evolved H2S was trapped via addition of ZnAc (1 % w/v, 100 µL) and enzymatic actions were stopped via addition of TCA (10 % w/v, 100 µL). The amount of H2S produced was then determined via absorbance measurements of the centrifugal supernatants upon addition of NNDPD and FeCl3 as described previously. 4.2.3. Determination of H2S standard curve A mixture of 0.85 % w/v ZnAc and 3 % w/v NaOH (100 µL) was added to various H2S standards (0-800 µM, 100 µL) prepared in 50 mM sodium phosphate pH 8.2 buffer containing 0.5 mM PLP, 2 mM Tris pH 8.0 buffer, 20 mM NaCl and 0.1 mM DTT. TCA (10 % w/v, 100 µL) was added and the tubes were then treated with NNDPD (20 mM, 42.9 µL) and FeCl3 (30 mM, 42.9 µL). After a short spin, 300 µL of the supernatant was removed for absorbance measurements at 670 nm. It was crucial to ensure that upon addition of FeCl3, the absorbance of the solution was measured in the shortest time possible as the absorbance readings were noted to decrease over time, particularly for high concentrations of the H2S standard. The average A670 value was then plotted against the amount of H2S present for each standard. 35 4.2.4. Optimized protein H2S assay screen for inhibitors of H2S production Stock solutions of PLP and L-cysteine were prepared in 50 mM sodium phosphate pH 8.2 buffer, while stock solutions of the inhibitor candidates were prepared in DMSO. Each tube consisted of 7.9 μg of GST-tagged CSE enzyme (dissolved in elution buffer – 20 mM Tris pH 8.0, 50 mM NaCl, 1 mM DTT buffer), saline (10 μL), PLP (2.5 mM, 20 μL), sodium phosphate pH 8.2 buffer (50 mM, 45 μL), inhibitor candidate (50 mM, 5 μL, added at 15 s intervals) and L-cysteine (27.5 mM, 10 μL, added at 15 s intervals). A set of tubes where no enzyme had been added served as baseline control, while a separate set of tubes where DMSO was added instead of the inhibitor served as the negative control. Each inhibitor candidate was pre-incubated with the enzyme on ice for 15 min followed by an additional 15 min at 37 °C before addition of the substrate. The tubes were then parafilmed, gently vortexed and incubated at 37 °C for 30 min. A mixture of 0.85 %w/v ZnAc and 3 % w/v NaOH (100 μL), followed by TCA (10 % w/v, 100 µL) were then added to each tube at 15 s intervals via needle. A supernatant mix comprising the above reagents (saline, PLP, sodium phosphate pH 8.2 buffer, elution buffer, DMSO, ZnAc, NaOH and TCA) was also prepared. Upon centrifuging the tubes at 13000 rpm for 3 min, the supernatant was removed and 300 μL of the prepared supernatant mix was added to the pellet. NNDPD (20 mM, 42.9 μL) and FeCl3 (30 mM, 42.9 μL) were then added to the tubes containing the trapped H2S. After centrifuging the tubes again at 13000 rpm for 3 min, 300 μL of the supernatant was loaded onto the microplate for absorbance measurements at 670 nm. The amount of H2S produced was calculated against the new calibration curve of sodium hydrosulfide (NaHS: 0-800 μM) and the percentage inhibition of each test compound was then determined. 36 4.3. Results and discussion 4.3.1. Kinetics of H2S production from purified recombinant human CSE In the previous section, bacterial lysates containing the expressed CSE enzyme had been assayed and shown to produce significant amounts of H2S. To establish that the production of H2S from these bacterial lysates was indeed due to the action of human CSE on the L-cysteine substrate, the purified enzyme was utilized in a subsequent assay to determine if L-cysteine could indeed be converted into H2S. Our preliminary results showed that this reaction was indeed possible, and increasing the amount of enzyme in the assay led to a corresponding increase in H2S production (Fig. 9). This was a significant finding that stressed the ability of human CSE in targeting and cleaving C-β-S Amount of H2S produced (nmol) bonds. 60.0 50.0 40.0 30.0 20.0 10.0 0.0 0 4 8 12 16 20 24 Amount of CSE added (μg) Figure 9. Relationship between amount of H2S produced in 30 min against amount of purified recombinant CSE added in the presence of 10 mM L-cysteine substrate. (Adapted from Huang et al., 2007) The sigmoidal graph of initial reaction velocity (V) against L-cysteine substrate concentration ([S]) (Fig. 10A) however suggested that the binding of L-cysteine to CSE could be a cooperative process. This was likely to be expected since our purified enzyme 37 was homotetrameric as observed from the gel filtration and DLS profiles of our enzyme. Fitting of the curve to the Hill equation enabled us to determine the maximum reaction velocity for the enzymatic reaction, Vmax, and the substrate concentration at which half Vmax is attained, K0.5, to be 0.14 U per mg CSE and 2.75 mM respectively. The extent of cooperativity, reflected by the Hill coefficient, h, was determined to be 2.6 from the logarithmic plot of the Hill equation (Fig. 10B). Since the Hill coefficient is less than 4, which is the number of binding sites for CSE, the Koshland-Nemethy-Filmer (KNF) model (Koshland, Nemethy, & Filmer, 1966) as opposed to the Monod-WymanChangeux (MWC) model (Monod, Wyman, & Changeux, 1965) may be employed to account for the cooperative binding of L-cysteine (Bisswanger, 2002). Binding of the first substrate induces a conformational change in other subunits, hence increasing the B 0.5 0.12 0.10 0.08 0.06 y = 0.04 0 .14 x 2.66 2 .75 2.66 + x 2.66 0.02 lg(V/(Vmax-V)) A Initial reaction velocity, V (Umg-1) enzyme’s affinity for more substrate. 0.0 -0.5 [S] (mM) vs Vo (U/mg CSE) x column vs y column y = 2.607x - 1.147 -1.0 -1.5 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 L-cysteine substrate concentration, [S] (mM) -2.0 -0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 lg[S] Figure 10. (A) Graph of initial reaction velocity against L-cysteine substrate concentration in the presence of 2 mM PLP. (B) Logarithmic Hill plot of lg(V/(Vmax-V)) against lg[S]. (Adapted from Huang et al., 2007) The turnover number per site (kcat) and the catalytic efficiency (kcat/KM) for our purified enzyme were calculated to be 0.1 mol of L-cysteine reacted to form H2S per mole of CSE monomeric site per second and 0.04 mM-1s-1 respectively. A comparison of these kinetic 38 parameters to those gathered from various reactions of human CSE revealed that the rate of production of H2S (as observed from the kcat value) for our purified enzyme was about 20 times slower than the rate of conversion of L-cystathionine to L-cysteine, which is the primary reaction catalyzed by the enzyme (Table 5). The catalytic efficiency for the conversion of L-cysteine to H2S was also about 100 times lesser than that for production of L-cysteine. These findings reiterate that CSE is more reactive towards C-γ-S bonds than C-β-S bonds. Nevertheless, as revealed in Zhu’s work, our observations also show that human CSE is a multifunctional enzyme where a secondary reaction of L-cysteine to H2S is possible. The large disparity in the kinetic parameters of CSE when utilizing Lcysteine as a substrate between Zhu’s and our work was most probably attributed to the different experimental setup as well as the different amount of exogenous PLP cofactor that was added in their assay. This would be further explored and discussed in Section 6. Table 5. A comparison of the kinetic parameters of human CSE when utilizing L-cysteine or L-cystathionine as substrate. Substrate KM Vmax -1 kcat -1 kcat/KM Reference (mM) (Umg ) (s ) (mM-1s-) L-cysteine 2.75 0.14 0.1 0.04 Huang et al., 2007 L-cysteine 3.5 0.9 0.7 0.19 Zhu et al., 2008 L-cystathionine 0.5 2.5 1.9 3.75 Steegborn et al., 1999 L-cystathionine 0.4 2.3 1.7 4.25 Zhu et al., 2008 4.3.2. Protein H2S assay screen of various commercially available inhibitor candidates Having established that our purified enzyme was able to catalyze the production of H2S from L-cysteine, a H2S assay which utilized the purified protein for screening of inhibitors was developed. The inhibition levels of some of the commercially available L39 cysteine analogues, together with the two established inhibitors of CSE, PAG and BCA were redetermined using this assay. As observed from our previous inhibitor screen which utilized rat liver homogenates as a source of the CSE enzyme, cysteine analogues with substitutions on the sulfhydryl moiety displayed poor inhibitory effects (Fig. 11). A large improvement in the inhibition potency was however observed for N-acetyl-Lcysteine, N-isobutyryl-L-cysteine, and N-isobutyryl-D-cysteine (Fig. 11). These three Nsubstituted L-cysteine analogues seemed promising since they showed better inhibition levels at 10 mM concentration than PAG in our assay. Percentage inhibition (%) 110.0 99.6 98.7 97.6 99.1 90.0 65.5 70.0 44.5 50.0 30.0 10.0 2.9 -10.0 -8.2 -20.1 -30.0 1 2 3 S 1: H2 N OH 2: NH2 OH NH N-acetyl-L-cysteine 6: O 7 8 S 3: 9 O O OH NH2 HS 4: S-β-(4-pyridylethyl)L-cysteine O NH2 HCl L-cysteine ethyl ester HCl O N HS O OH NH2 O O HS S HCl 6 N S-(2-aminoethyl)L-cysteine HCl S-ethyl-L-cysteine 5: 5 O O Key 4 HS OH NH N-isobutyrylL-cysteine 7: O O OH NH N-isobutyrylD-cysteine 8: H2N OH 9 : β-cyano-L-alanine (BCA) O OH NH2 DL-propargylglycine (PAG) Figure 11. Average percentage inhibition values of various L-cysteine analogues, BCA and PAG assayed at 10 mM concentration in the presence of 10 mM L-cysteine substrate, 2 mM PLP and 10 µg purified CSE. Each compound was prepared in 10 mM sodium phosphate pH 8.2 buffer and assayed in triplicates over three separate experiments. 40 However, a more accurate indication of the potency of these N-substituted L-cysteine analogues should be gathered from their IC50 values instead. The inhibition profiles of Nacetyl-L-cysteine and N-isobutyryl-L-cysteine were hence obtained and their IC50 values were compared to those of BCA and PAG. In this experiment, the amount of enzyme and L-cysteine substrate loaded in the assay were reduced to 5 µg and 5 mM respectively to minimize wastage of enzyme and substrate. The results showed that although the inhibition levels of either N-acetyl-L-cysteine or N-isobutyryl-L-cysteine were higher than that for PAG beyond 5 mM, the IC50 values for these two cysteine analogues were higher compared to PAG’s (Fig. 12). The IC50 value is hence a better indicator of the inhibition potency rather than the inhibition level at a fixed concentration. At this moment, the potency of our best inhibitor candidates was still about 40 times lesser than that for the commercial inhibitors of CSE. Nevertheless, the results suggest that L-cysteine analogues that possess bulky group substitutions on the amino group may be possible lead compounds for the inhibition of H2S production, which had been predicted from the rat liver homogenate assay as well. 41 Percentage inhibition (%) B 110.0 100.0 80.0 60.0 40.0 20.0 0.0 70.0 50.0 30.0 10.0 -10.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 -30.0 Concentration of N-acetyl-L-cysteine (mM) Concentration of N-isobutyryl-L-cysteine (mM) -20.0 IC50 = 4.2 mM C Percentage inhibition (%) 90.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 IC50 = 4.3 mM D 120.0 100.0 80.0 60.0 40.0 20.0 0.0 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 Concentration of BCA (mM) IC50 = 0.1 mM Percentage inhibition (%) Percentage inhibition (%) A 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 Concentration of PAG (mM) IC50 = 0.3 mM Figure 12. Inhibition profiles and IC50 values of (A) N-acetyl-L-cysteine, (B) Nisobutyryl-L-cysteine, (C) BCA and (D) PAG determined in the presence of 5 mM Lcysteine substrate, 2 mM PLP and 5 µg purified CSE. 4.3.2. Protein H2S assay screen of various chemically synthesized inhibitor candidates Various L-cysteine analogues which had previously been tested in the rat liver homogenate assay were re-synthesized and tested using the purified enzyme assay in the presence of 2.75 mM L-cysteine. A factor to note when designing the assay for screening of inhibitors was the amount of substrate that should be loaded since the inhibition levels is expected to differ for different substrate concentrations, assuming the inhibitor acts on the enzyme competitively. A recommended substrate concentration by Eisenthal and Danson (Eisenthal & Danson, 2003) is the KM value of the enzyme for the substrate, 42 which in our case was 2.75 mM. In this assay, rather different inhibition trends for the tested compounds had been observed when compared to the tissue H2S assay (Fig. 13). However, the protein assay was considered more indicative of the true inhibition potency of any candidates since the enzyme had been added in the purified state unlike that from the homogenate. Moreover, as the enzyme utilized in the protein assay was of human rather than rat origin, the observed inhibition potency should be more representative of the true inhibition level when administered to humans subsequently. Three other urea derivatives (ZYJ4226, ZYJ4227, ZYJ4291) had also been synthesized and tested using the newly developed assay. Rather than having a mercapto-acid side chain bound to the urea functionality as in compounds 2 – 4 (Fig. 13), these derivatives had only an acidic side chain bound to the urea group. However, for these 3 urea derivatives, the absence of the sulfhydryl group did not seem to have compromised the activity of these two derivatives by much. Our initial hypothesis from results of the rat liver homogenate screen, that a free sulfhydryl group was necessary in the binding of the compound to the enzyme was hence probably wrong. It would be more likely that this area to which the original sulfhydryl group of L-cysteine occupies is more spatially constrained, and thus the attachment of large bulky groups to the sulfhydryl group would be unfavorable. In this assay, ZYJ4226 and ZYJ4291 were found to display higher inhibition levels than that for ZYJ4227. This was most probably due ZYJ4226 and ZYJ4291 being conformationally more restrained than ZYJ4227, hence leading to a smaller entropy loss encountered during binding of the compound to the enzyme. 43 Figure 13. Averaged percentage inhibition values of various synthesized inhibitor candidates assayed at 0.1 mM, 1 mM or 5 mM concentration in the presence of 2.75 mM L-cysteine substrate, 2 mM PLP and 5 µg purified CSE (candidates 1 to 6) or 7.5 µg purified GST-CSE, which was an equivalent of 5 μg CSE in terms of molar quantity (candidates 7 to 9). Each compound was prepared in DMSO to a final concentration of 2 % v/v (for compounds assayed at 0.1 mM or 1 mM concentration) or 10 % v/v (for compounds assayed at 5 mM concentration) and assayed in triplicates. 4.3.4. Optimization of protein H2S assay screen A major problem encountered whilst using the current protein assay for screening of the inhibitor candidates was that upon addition of trichloroacetic acid for stopping enzymatic 44 actions, the compounds would become protonated and most would precipitate (Scheme 2). Although care had been taken to ensure that the precipitate would not be loaded into the microplate during spectrophotometric measurements, some of these precipitated compounds were found to interfere with the development of methylene blue, leading to a much lower absorbance value that was measured. It was hence likely that the inhibition levels of some compounds had been over-estimated. Scheme 2. A schematic of the protein H2S assay that was used for screening of inhibitor candidates, its drawback and one possible strategy for overcoming this problem. ZnAc Colorless soln ZnAc, TCA o Add dyes 37 C 30min Baseline control CSE + S + I TCA + CSE + S + I Pale blue soln Ppt present White ppt Blue ppt Hopefully…. Strategy 1: Centrifuge and transfer supernatant to new tube before adding dyes. Problem: H2S easily lost during centrifugation Colorless soln Abbreviations: S – substrate I – inhibitor candidate Ppt – precipitate Soln - solutiona Pale blue soln The first strategy that was undertaken during the optimization process was to remove the precipitated compounds prior to addition of the NNDPD and FeCl3 dyes (Scheme 2). To achieve this, ZYJ4291 was used as a test compound and each tube was centrifuged to pellet the precipitated compounds after addition of TCA. A set of negative control tubes where DMSO was added instead of the test compound was also treated in the same manner. The supernatant was then transferred to a separate tube for addition of the dyes (Scheme 2). Although the original interfering effects caused the precipitated inhibitor candidate could be removed by this method, a large amount of H2S was lost upon 45 centrifugation of the test solutions based on a comparison of the absorbance values of test solutions which were centrifuged (Fig. 14, ZnAc + TCA, sulfides trapped in supernatant and pellet) and those which were not (Fig. 14, ZnAc + TCA, total sulfides). To account for this phenomenon, we would have to consider the process with which the evolved H2S is trapped in the assay (Scheme 3). During the assay, ZnAc had been added to trap the evolved H2S gas as insoluble ZnS. However, upon addition of TCA, the reaction mixture becomes highly acidic and the whole equilibrium is shifted backwards. As such, the trapped sulfides would become released again as H2S, and since H2S is only partially soluble in water, it would be easily lost over time and particularly upon centrifugation where the test solution tends to heat up. It is hence crucial to ensure that the reaction mixture remains basic before the dyes are added so that H2S is trapped effectively in the solid state as ZnS. To explore this, a mixture of ZnAc and varying concentrations of NaOH was tested to determine which would be the most effective in trapping H2S as insoluble ZnS. Scheme 3. Process of trapping of H2S gas. H2S (g) H2S (aq) + - H (aq) + HS (aq) + 2- 2H (aq) + S (aq) Zn2+ ZnS(s) 46 0.600 Net A670 values 0.500 0.485 0.488 0.484 0.471 0.459 0.400 0.297 0.300 ZnAc + TCA 0.258 ZnAC + 2.5% NaOH + TCA 0.200 0.150 ZnAC + 3% NaOH + TCA 0.081 0.100 0.007 0.006 ZnAc + 5% NaOH + TCA 0.014 0.000 Total sulf ides Sulf ides trapped in supernatant Sulf ides trapped in pellet Figure 14. Net A670 readings reflecting the distribution of trapped sulfides when different amounts of NaOH were added together with ZnAc. The assay was performed in duplicates in the absence of any inhibitor candidates, and the maximum amount of sulfides that could be trapped was determined by adding the dyes immediately after addition of TCA, without any centrifugation. Sulfides that could be trapped in the supernatant (soluble sulfides) or in the pellet (insoluble sulfides) were assayed from tubes which were centrifuged upon addition of TCA to determine the extent to which H2S had been effectively trapped in the pellet. As revealed by Fig. 14, the maximum total amount of H2S that could be trapped was increased significantly when a mixture of ZnAc and NaOH was used. However, an optimal amount of NaOH could not be determined as yet since all the sulfides were found to be trapped predominantly as insoluble ZnS, with the exception of 5 % NaOH. The original strategy of separating any precipitated compounds by centrifuging may hence be inappropriate since they would coexist with the trapped ZnS in the centrifugal pellet. Nevertheless, the results showed the importance of adding NaOH since this was unlike that when no NaOH had been added, where the sulfides were found mostly in the solution phase and hence could be easily lost with time. Another set of experiment was subsequently designed to determine the optimal concentration of NaOH that would not 47 only trap H2S effectively, but also minimize the precipitation problem during the screening of inhibitor candidates. Table 6. Observations upon addition of a mixture of ZnAc and varying concentrations of NaOH, and TCA in the presence of ZYJ4291 as a test compound. Treatment Observations after adding ZnAc+NaOH Observations after adding TCA pH* ZnAc + TCA White ppt was formed. Ppt remained insoluble. 1 ZnAc + 2.5 % NaOH + TCA Solution turned pale yellow. Solution turned slightly cloudy. 5–6 ZnAc + 3 % NaOH + TCA Solution turned yellow. White ppt was formed in pale yellow^ solution. 8–9 ZnAc + 5 % NaOH + TCA Solution turned yellow. Solution remained yellow^. 13 – 14 * Approximate pH of the reaction mixture after addition of TCA, as estimated by pH indicator paper (It should be noted that the occurrence of any white precipitate could also be due to Zn existing as a complex of Zn(OH2)62+, Zn(OH)4(OH2)22- or Zn(OH)2 under different pH). ^ Due to PLP which had been added in the assay. A possible method for separation of the inhibitor from trapped sulfides may be achieved by careful alteration of the pH such that any added inhibitor remains largely soluble while H2S is trapped as insoluble ZnS. This was accomplished when a mixture of ZnAc and 3 % NaOH was added, and upon addition of TCA, the reaction mixture still remained sufficiently basic (Table 6) for the inhibitor to stay dissolved (Scheme 4). The formation of white precipitate in this case was not due to the inhibitor, but rather the conversion of Zn(OH2)62+ to Zn(OH)2 as the reaction mixture becomes neutralized upon addition of TCA. Upon centrifugation, the supernatant which contained the inhibitor could then be removed to a separate tube, and subsequent addition of the dyes was performed only to the centrifugal pellet which contained the trapped ZnS. A mixture of ZnAc and 2.5 % NaOH was inappropriate since the pH of the terminated reaction mixture was still acidic. As such, most of the inhibitor would be precipitated and coexist with the insoluble ZnS 48 (Scheme 4). For tubes with 5 % NaOH added, although the reaction mixture was still strongly basic for the inhibitor to remain dissolved after addition of TCA, most H2S had been found to be present in the centrifugal supernatant (Fig. 14) rather than in the pellet. This concentration of NaOH was hence inappropriate. Scheme 4. Partition of inhibitor and trapped H2S after addition of TCA Most ideal: Precipitated inhibitor coexisting with trapped sulfides in pellet Inhibitor dissolved in supernatant Soluble inhibitor coexisting with trapped sulfides in supernatant H2S trapped as insoluble sulfides in pellet ZnAc + 2.5% NaOH ZnAc + 3% NaOH ZnAc + 5% NaOH 4.3.5. Redetermination of H2S standard curve As the H2S assay had been optimized for subsequent runs, the relationship between the amount of H2S produced and absorbance at 670 nm had to be reestablished. The H2S standards were dissolved in sodium phosphate buffer containing PLP, Tris, NaCl and DTT rather than water since the former resembled more closely to that of the actual reaction condition in the optimized assay. Similarly, the H2S standards were treated with a mixture of 0.85 % w/v ZnAc and 3 % w/v NaOH instead of ZnAc alone since the former had been established to trap H2S more efficiently than the latter. The data was fitted to a quadratic curve as the A670 readings were noted to deviate from linearity, particularly at high H2S standard concentrations (Fig. 15). 49 3.00 Average A670 value 2.50 2.00 y = (-8.966E-5)x2 + (3.736E-2)x + 4.272E-2 Rsqr = 0.9992 1.50 1.00 0.50 0.00 0 20 40 60 80 Amount of H2S (nmol) Figure 15. Relationship between absorbance at 670 nm and amount of H2S produced. Each experiment was performed in triplicates, and the average A670 value for each H2S standard was determined over three separate runs. 4.3.6. Redetermination of inhibition potency of inhibitor candidates Using the optimized H2S inhibition assay, the inhibitor candidates were assayed again to determine their actual inhibition levels. Due to an improved method of trapping the H2S gas in the optimized assay, the inhibition levels of most compounds were found to be much reduced (Fig. 16) when compared to those obtained previously (Fig. 13). The inhibition of PAG at 2.5 mM concentration was however found to be much higher (Fig. 16) than that obtained previously (Fig. 12D), most probably due to PAG being an irreversible inhibitor which acts on the enzyme mechanistically by forming a covalent bond to an active site amino acid (Steegborn & Clausen, 2000). As such, the period of time and temperature at which the inhibitor is allowed to interact with the enzyme are important factors which can affect the measured inhibition level. In the optimized assay, a 15 min pre-incubation of the inhibitor with the enzyme on ice followed by another 15 50 min at 37 °C, as opposed to a 30 min pre-incubation on ice (experimental condition for obtaining the inhibition level of PAG in Fig. 12D) was thus the most likely reason for a higher inhibition level that was determined for PAG. Figure 16. Average percentage inhibition values of various synthesized inhibitor candidates assayed at 2.5 mM concentration in the presence of 2.75 mM L-cysteine substrate, 0.5 mM PLP and 7.5 µg GST-CSE. Each compound was added as a DMSO 51 stock solution to a final concentration of 5 % v/v and assayed in duplicates in at least three separate experiments. Comparing compounds with a mercapto-acid or an acidic side chain coupled to the urea functionality (Compound 5 versus compounds 8 and 9) or thiourea functionality (Compound 6 versus 14; compound 7 versus 15), compounds comprising the mercaptoacid side chain were found to display better inhibition levels than their correspondents which possessed only the carboxylic acid side chain (but with the same substituent on the other side of the urea or thiourea functionality). As compounds which comprised the mercapto-acid side chain were derived from L-cysteine itself, these L-cysteine derivatives were probably more easily recognized by the enzyme and hence could inhibit the enzyme more strongly. However, compounds such as their correspondents which did not possess the sulfhydryl group were synthetically less challenging than the L-cysteine derivatives and were hence further explored. As observed in the previous screen in Section 4.3.3, the inhibition level of ZYJ4227 was also found to be slightly poorer than that of ZYJ4226 in the optimized assay, most probably due to the flexible acidic side chain that the former compound possessed. An interesting observation from the naphthalene-substituted compounds 8 to 12 was that the positioning of carboxylic acid group three carbons away from the nitrogen atom of the urea or thiourea group seemed to be more optimal towards the inhibition of the enzyme. The inhibition level of ZYJ4291 (compound 10) for instance, was about 2-fold that of ZYJ4226 and ZYJ4227 (compounds 8 and 9). This trend was also observed for TYS1026 (compound 12) which exhibited higher inhibition potency than FUX11156 (compound 11), albeit at a smaller percentage increase, most 52 probably due to the flexible side chain in TYS1026. This optimal arrangement of the acidic group may hence be serve as a guideline for future inhibitor design. At this moment, a direct correlation between the inhibition potency and the type of substituents on the compound still could not be made. Comparing across thiourea compounds with the benzoic acid side chain (compounds 11 and 13 to 16), an increase in the electrophilicity of the benzene ring substituent did not correlate with an increase or decrease in inhibition levels. It would hence be better to base the design of the inhibitor on the three-dimensional structure of the enzyme rather than random syntheses. Recently, the crystal structure of the native CSE enzyme as well as its complex with PAG had been solved by our collaborators (Sun, et al., 2008). Particularly, the three-dimensional structure of the enzyme complexed with PAG would provide great insights into the binding mode of the inhibitor, and these would prove useful in the rational design of potent inhibitors of CSE. A discussion of this would follow in the subsequent chapter. The only hit compound we obtained thus far was CJH1035. This compound had an IC50 of 0.45 mM (tested in the presence of 2.75 mM L-cysteine and 7.5 µg GST-CSE) which was rather close to that found for PAG (0.3 mM) and BCA (0.1 mM). Upon observation, CJH1035 is actually a derivative of β-trifluromethyl-alanine where the L-enantiomer was reported to be an irreversible inhibitor of rat liver CSE as well (Alston, et al, 1981). It would hence be necessary to ensure that the inhibtion of the enzyme by CJH1035 was not due the action of β-trifluromethyl-alanine upon in-vitro cleavage of the urea functionality (though unlikely). Crystallization screens should also be carried out on this compound with the aim of co-cystallization with CSE so that future design of more potent inhibitors can be achieved. 53 4.4. Conclusion In this section, we established human CSE as an enzyme responsible for the production of H2S. Kinetics studies revealed that this reaction is secondary to that for the conversion of L-cystathionine to L-cysteine. The assay for production of H2S had also been optimized for more accurate determination of the inhibition levels of various inhibitor candidates. This optimized assay was based on the usage of basic ZnAc for effective trapping of H2S. With the optimized assay, various inhibitor candidates had also been tested for their inhibition levels, and an initial hit compound with an IC50 value of 0.45 mM had been identified. In the next chapter, the X-ray crystal structure of the human CSE enzyme in both its apo and holo forms, as well as complexed with PAG would be presented. Attempts to facilitate the crystallization process via proteolytic cleavage of the protein would also be discussed. 54 5. Elucidation of the three-dimensional structure of human CSE 5.1. Objectives Thus far, the random synthesis of inhibitor candidates did not serve as a good strategy for inhibitor design. With the development of an expression and purification system for human CSE in the Section 3, the three-dimensional conformation of the enzyme could hence be elucidated via X-ray crystallography (in the Honors project) to aid in future rational design of inhibitors. This structure of the enzyme would also be compared with that determined by our collaborator from the Structural Genomics Consortium. During the Masters research project, our collaborator had also successfully co-crystallized the PAG inhibitor with human CSE. In this section therefore, interesting binding modes of PAG with CSE, and a mechanism for the inhibition process would be discussed. Attempts to co-crystallize the enzyme with various L-cysteine analogues which had previously been tested and shown to exhibit some inhibition, as well as proteolytic cleavage experiments to aid in the crystallization process would also be presented in this chapter. 5.2. Experimental 5.2.1. Determination of protein homogeneity via DLS experiments The purified protein was first centrifuged at 10000 rpm for 3 min at 4 °C. 16 µL of the supernatant was then loaded into the DLS cuvette and a monochromatic beam of light was passed through the cell with the aid of a DLS machine (ProteinSolutions DynaPro). 55 Fluctuations in the light intensity due to scattering by the protein molecules were recorded over twenty measurements, and analysis of the data was subsequently performed using the DLS software, Dynamics v5. 5.2.2 Screening and optimization of crystallizing conditions Suitable crystallizing conditions were screened with commercially available screen kits from Hampton, Jena Bioscience and Nextal using both hanging-drop and under-oil methods at room temperature (20 °C). For the hanging drop method, 1 µL:1 µL or 2 µL:1 µL of protein solution to reservoir buffer were suspended over 500 µL of the reservoir buffer. A mixture of 0.6 µL protein to 0.4 µL reservoir buffer submerged under 20 µL parafilm oil was in turn used for the under-oil method. Upon the discovery of a suitable crystallizing condition, the condition was further optimized via an additive screen and a grid screen. For the additive screen, small amounts of inorganic ions, polymers, detergents, reducing agents or small organic molecules from the commercial Hampton Additive Screen Kit were included in the reservoir buffer. As for the grid screen, the concentrations of precipitant, buffer and salt, as well as the buffer pH were systematically varied around the initial hit condition, and the conditions were rescreened via the hanging drop method. 5.2.3. X-ray diffraction and structure determination Upon successful crystallization of the protein, crystals that were of good quality in terms of size and shape were submerged in a cryo-protectant (0.15 M lithium sulfate, 0.1 M 56 phosphate-citrate pH 4.2, 18 % PEG1000, 3 % 1,6-hexenediol, 20 % glycerol) and subsequently mounted onto a cryo-loop. A complete data set was then collected from the in-house X-ray diffraction machine (Rigaku, R-Axis IV) and then processed with the software, HKL2000 (Otwinowski & Minor, 1997). The structure of the enzyme was solved via the molecular replacement method against the CSE search model structure (PDB: 2NMP) with the aid of MolRep (Murshudov, et al., 1997). Refinement of the protein structure was then performed using the Crystallography and NMR System program (Brunger, et al., 1998), while model building was achieved using the COOT program (Emsley & Cowtan, 2004). 5.2.4. Proteolytic cleavage of CSE The reaction mixture, comprising a total volume of 100 µL, consisted of 100 µg of CSE (1 mg/mL, 100 µL) and 0.5 µg of various proteolytic cleavage enzymes. The proteases were allowed to act on the enzyme at room temperature (20 °C), and subsequent termination of the protease activity was performed by removing 30 µL of the reaction mixture into 2 µL of protease inhibitor mixture (Sigma, 44 mg/mL) and 8 µL of 5x sample loading buffer after 15 min, 45 min and 2 h. A control experiment without the addition of any proteases was also performed to determine the extent of protein degradation throughout the duration of the experiment. Denaturing SDS-PAGE analysis was then performed on the terminated reaction samples to determine a suitable proteolytic cleavage enzyme for CSE. Optimization of the cleavage process by chymotrypsin was performed by incubating the same amount of CSE enzyme with 2 µg or 1 µg of chymotrypsin and terminating the reaction after 15 min, 30 min, 50 min, 75 57 min and 100 min, or incubating 75 µg of CSE (1 mg/mL, 75 µL) with 0.375 µg or 0.1875 µg of chymotrypsin and terminating the reaction after 2 h, 3 h, 4 h and 5 h. 58 5.3. Results and discussion 5.3.1. Optimization of protein concentration for crystallization studies The concentration of the protein solution is one crucial factor for successful crystallization. A super-saturated protein solution would lead to unfavorable protein aggregation and precipitation while solutions which are too dilute would require a long time for nucleation to occur. It was hence necessary to ensure that the purified protein obtained in Section 3 was at its optimum saturation level for spontaneous nucleation to occur. Using a maximum polydispersity index of 0.20 as a guide, a plausible protein concentration for subsequent screening of crystallization conditions was determined to be around 10 mg/mL via DLS experiments (Table 7). However, an initial screen of the crystallization conditions at this protein concentration was accompanied with extensive precipitation in more than 70 % of the conditions that were tested. The protein concentration was hence reduced to about 5 mg/mL in subsequent screens where a more reasonable degree of precipitation in 40 % of the conditions was observed. As the polydispersity index of the protein at 5 mg/mL was significantly lower than 0.20 (Fig. 17), protein crystallization was expected to be likely, according to Adrain et al. (1997). Other parameters which were revealed from the DLS experiment include the hydrodynamic molecular weight and hydrodynamic radius, both of which support the existence of our purified protein as a tetramer in solution. 59 Table 7. Effect of varying protein concentration on the polydispersity index of the protein. Protein concentration (mg/mL) Polydispersity index 0.81 0.10 2.2 0.10 3.0 0.11 4.3 0.12 5.8 0.15 10.6 0.20 15.2 0.24 DLS experiments were performed in 10 mM Na3PO4 pH8.2 buffer, 1 mM DTT at 20 °C (Adapted from Huang et al., 2007) Temperature (°C) Count rate Radius Polydispersity index MW (kDa) % Mass Sos error 20.0 2200786 4.99 0.11 217 100.0 5.59 Figure 17. DLS profile and parameters for the purified CSE protein at 5.0 mg/mL. (Adapted from Huang et al., 2007) 5.3.2. Screening and optimization of crystallizing conditions for CSE The crystallization of CSE had been accompanied with several difficulties due to the high percentage (46 %) of flexible loops possessed by the enzyme, as predicted by an online protein secondary structure prediction software (Rost et al., 2004). Most of the conditions that were provided by the commercial screen kits were hence ineffective. The initial hit condition was obtained from condition 6 (0.2 M lithium sulfate, 0.1 M phosphate-citrate pH 4.2, 20 % PEG1000) of the Nextal JCSG+ Suite screen kit. Interestingly, crystals were formed only when 10 mM L-cysteine was included in the crystallizing condition. This could be due to a stabilization of the polypeptide chains upon addition of the L60 cysteine substrate, hence aiding in the crystallization process. Optimization of this crystallizing condition via the additive screen and grid screen led to the growth of CSE crystals which were of sufficient size and quality of X-ray diffraction (Fig. 18). Some of the crystallographic data and refinement statistics are summarised in Table 8. Additive and grid screen Initial hit condition: 0.2 M Li2SO4, 0.1 M phosphate-citrate pH 4.2, 20 % PEG1000, 10 mM L-cysteine X-ray diffraction Optimized condition: 0.15 M Li2SO4, 0.1 M phosphate-citrate pH 4.2, 18 % PEG1000, 3 % 1,6hexenediol, 10 mM L-cysteine Figure 18. Optimization of CSE crystallizing condition for X-ray diffraction and subsequent structure determination. (Adapted from Huang et al., 2007) Table 8. Data collection and refinement statistics for crystallized human CSE enzyme. (Adapted from Huang et al., 2007) Parameter Space group P42 Unit cell dimensions a= 121.39Å, b= 121.39Å, c= 125.468 Å; α = 90°, β = 90°, γ = 90° No. of monomers in asymmetric unit 4 Refinement resolution 10.0 – 3.0Å R-value 0.22 Rfree 0.29 Overall B-factor 24.2Å2 The three-dimensional structure that was obtained for our enzyme was similar to that determined for yeast CSE (Messerschmidt et al., 2003) and human CSE as elucidated by our collaborator. The asymmetric unit consisted of four monomers arranged as a dimer of 61 dimers. The major difference between our structure and that determined by our collaborator was that our structure was observed to be in a much more open configuration (Fig. 19A), which suggested that weaker interactions existed between the monomers for our structure. A B Figure 19. Asymmetric units of human CSE determined in this work (A) and by our collaborator (B). Figure was produced by Pymol software (DeLano, 2002). A closer look at the active site region of our structure revealed the basis for this open configuration. Our enzyme had been crystallized in the apo-form without PLP unlike that in our collaborator’s study. As such, hydrogen bonding interactions and ionic linkages which were crucial in maintaining the tetrameric conformation were most probably lost. This was particularly so for the hydrogen bonds between the phosphate group of PLP and Tyr-60 as well as Arg-62 from the adjacent monomer, which were necessary for the assembly of each dimer. Due to the absence of PLP, the Tyr-114 and Lys-212 residues were also observed to be displaced from their original positions by as much as 8.6 Å and 7.0 Å respectively when compared to the structure of CSE holoenzyme determined by our collaborator (Fig. 20B and C). These displacements were observed since the π- 62 stacking interactions between the pyridoxal ring of PLP and Tyr-114, as well as the covalent Schiff-base linkage between PLP and Lys-212 would be disrupted upon the loss of the cofactor (Fig. 20A). The electron density corresponding to these regions of Tyr-60, Arg-62 and Tyr-114 (Gly-26 to Asn-65 and Met-110 to Tyr-120) were hence observed to be absent in our CSE apoenzyme structure since these regions would become much more flexible upon disruption of the above mentioned interactions with PLP. Figure 20. (A) Electron density map around PLP in the CSE holoenzyme. Significant differences in stereo-overlay of peptide chains around the Tyr-114 (B) and Lys-212 (C) residues, shown in green (our structure) and yellow (collaborator’s structure). The red ball and stick structure represents the PLP cofactor in the holoenzyme structure. (Adapted from Huang et al., 2007) The crystallization of our enzyme in the apo-form was not due to a loss of the cofactor during the purification process (Sun, et al., 2009). In fact, it was likely to be triggered upon addition of L-cysteine into the crystallizing condition. As shown in Fig. 21, our purified enzyme displayed the characteristic 427 nm absorbance peak which corresponded to the Schiff-base linkage between PLP and Lys-212 (blue line). In-vitro incubation of our enzyme with 10 mM L-cysteine for 3 days and subsequent buffer exchange into 20 mM Na3PO4 pH 7.8 buffer led to a corresponding suppression of the 63 427 nm peak, which indicated a breakage of the PLP-Lys-212 bond (red dotted line). Upon the readdition of an equimolar amount of PLP, the 427 nm peak was regenerated (cyan line). More importantly, it was observed that all of the added PLP had rebound to the enzyme since the 388 nm peak which corresponded to free unbound PLP was absent for this spectrum (cyan line). This revealed that upon a long period of incubation with Lcysteine, which was the case for crystallization, not only would the PLP-Lys-212 covalent Schiff-base linkage be broken, the PLP cofactor would also be released from the enzyme. An explanation for this phenomenon would be provided in the subsequent chapter. Before cysteine incubation After cysteine incubation Readdition of PLP Figure 21. Absorbance spectra of our purified CSE enzyme before and after L-cysteine incubation, and upon a readdition of equimolar amount of PLP. (Adapted from Sun et al., 2009) 5.3.3. Screening and optimization of crystallizing conditions for CSE-inhibitor complexes Attempts to co-crystallize our enzyme with established inhibitors, PAG and BCA had also been performed in our study. Complexes of CSE with these inhibitors were prepared by mixing CSE with varying concentrations of the inhibitors ranging from 1 mM to 10 64 mM. An initial crystallizing condition for CSE-PAG and CSE-BCA was obtained in JenaBioscience Screen 10 Condition B6 (1.5 M lithium sulfate, 0.1 M Tris-HCL pH 8.5) where microcrystals were observed from the under-oil method. However, optimization of this condition by varying the drop-size and ratio of protein to crystallizing solution had only improved the size of the microcrystals slightly. A further optimization of this crystallizing condition was not performed as our collaborator had successfully crystallized and determined the structure for the CSE-PAG complex shortly after that. PAG Y114 PLP K212 Figure 22. A closed-up view of the active site region of the CSE-PAG complex. The active site region of the CSE-PAG complex was observed to be rather similar to that of the native enzyme (Fig. 22). The PLP cofactor was still covalently bound to the Lys212 residue via the Schiff-base linkage, and Tyr-114 was stacked almost parallel above PLP to establish π-π interactions. The mode of inhibition of CSE by PAG was however more clearly elucidated where PAG was shown to irreversibly inhibit the enzyme by covalently binding to the Tyr-114 residue. This was drastically different for PAG complexes with CsdB (a PLP-dependant enzyme found in E. coli) and methionine-γ-lyase (MGL, also a PLP-dependant enzyme found in most prokaryotes), where the amino group 65 of PAG was observed to remain bound to the PLP cofactor (Sun, et al., 2009). The reasons proposed for these observations were that the side chain of His-123 in CsdB (which occupies the same position as Tyr-114 in human CSE) was too far away to establish a covalent bond interaction with PAG, and for MGL, the release of PAG from PLP was disfavored due to the absence of amino acids which could form stabilizing Hbond interactions with the amino group of PAG. Lys212 Lys212 CO2- Internal SchiffH3N+ base with PLP N O OH Step 1 B O P O HO N CH3 H CSE-bound PLP Lys212 H CO2- CO2- H2N N + BH N OH P NH2 Step 2 BH+ OH P N CH3 H CSE-bound PLP N H CH3 Step 3 Lys212 CO2- O NH3+ Tyr114 N N H CO2 CH3 + CO2- H3N - NH2 N Step 5 Tyr114 OH P Lys212 H O N Step 4 OH OH P N H Lys212 OH P N H CH3 CH3 Tyr114 Figure 23. Proposed mechanism for the inhibition of CSE by PAG. Based on the binding mode of PAG in human CSE, we proposed a mechanism to account for the inhibition of CSE by PAG. A base, most probably from one of the active site basic amino acids, would first activate the amino group of PAG (Fig. 23, Step 1) for subsequent binding to PLP via a transaldimination reaction (Fig. 23, Step 2). This is followed by abstraction of a β-hydrogen atom of the bound PAG to yield an activated allene (Fig. 23, Step 3), which is then subjected to nucleophilic attack by the hydroxyl group of Tyr-114 to form a vinyl ether intermediate (Fig. 23, Step 4). Subsequent transaldimination with Lys-212 regenerates the internal aldimine and a covalently bound 66 PAG with Tyr-114 (Fig. 23, Step 5) as revealed from the crystal structure of CSE-PAG complex (Fig. 22). Furthermore, to demonstrate the importance of the hydroxyl group of Tyr-114 on the irreversible inhibition of CSE by PAG, we proceeded to determine the IC50 value for the inhibition of H2S production from the Tyr-114ÆPhe mutant CSE protein by PAG (details on the expression and activity of this mutant CSE protein would be provided in the next chapter). In this mutant protein, as the phenyl side chain of the Phe-114 residue does not possess the hydroxyl substituent for nucleophilic attack of the PAG allene intermediate in Step 4 of the above mechanism, the inhibition of this mutant protein towards H2S production would be expected to be much weaker. Indeed, the IC50 value was increased significantly by about 8-fold when PAG was tested upon the Tyr-114ÆPhe mutant CSE protein (Fig. 24). This hence serves as an additional evidence, besides the crystal structure of the CSE-PAG complex, for the role of Tyr-114 in the inhibition of CSE by Percentage inhibition (%) PAG. 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 0.0 IC50 = 2.3mM 2.0 4.0 6.0 8.0 10.0 Concentration of PAG (mM) Figure 24. IC50 analysis on the inhibition of H2S production from the Y114F mutant CSE protein by PAG. The assay was performed in the presence of 5 µg GST-tagged mutant protein, 0.5 mM PLP and 2.75 mM L-cysteine. 67 In our study, we had also attempted to co-crystallize our enzyme with various L-cysteine analogues which had been shown to possess some inhibition capabilities in Section 4. An initial crystallizing condition was noted for CSE with N-isobutyryl-D-cysteine in Nextal JCSG+ Suite Condition 15 (0.1 M BICINE pH 9, 20 % PEG 6000) from the hangingdrop method. Spherulites were formed about 3 days after setting up for crystallization, and needle clusters were formed subsequently around the spherulites 2 weeks later (Fig. 25). This result however, was not reproducible, and optimization by grid screen and Additive Screen failed to give any crystals. This could either be due to slight differences in the crystallizing solution prepared as opposed to that in the screen kit, or differences in the quality of protein used for crystallization. Slight variations during the purification process could have affected the conformation of the protein, especially in the flexible regions, hence preventing crystallization of the protein-inhibitor complex even under the same crystallizing condition. A B C Figure 25. (A) Spherulites of CSE complexed with 5mM N-isobutyryl-D-cysteine formed in 0.1 M BICINE pH 9, 20 % PEG 6000; (B, C) Needles formed around spherulites 2 weeks later. A preliminary crystallizing condition for CSE with N-isobutyryl-L-cysteine was also obtained in the same condition with the addition of 10 mM ZnCl2 as additive using the 68 hanging-drop method (Fig. 26A). This crystal was most probably protein since salt crystals which formed at the bottom of the well (Fig. 26B and C) were significantly different in appearance and polarizability compared to that in the protein drop. Sadly, this protein crystal had failed to give any meaningful X-ray diffraction data and hence the structure of the inhibitor complex could not be determined. A B C Figure 26. (A) Crystal of CSE complexed with 5 mM N-isobutyryl-L-cysteine in 0.1 M BICINE pH 9, 20 % PEG 6000, 10 mM ZnCl2; (B, C) ZnCl2 crystals at bottom of well. 5.3.4. Proteolytic cleavage of CSE As observed from the structure of human CSE, we believe that the main difficulty in the crystallization of this protein, be it in the native or inhibitor-complexed form, lies in the high percentage of flexible loops possessed by the enzyme. A method of minimizing the extent of protein instability due to flexible loops in the protein structure was to remove these regions by proteolytic cleavage. This method has been found to be a prerequisite step for the crystallization of certain proteins such as Gram-positive β-recombinase (Orth et al., 1999). The aim of this strategy was therefore to cleave the flexible loops which would be much more exposed, prior to the α-helices and β-beta sheet regions which are much more stabilized. A screen was hence performed to determine a suitable proteolytic cleavage enzyme for this process. 69 For the control experiment where no proteases had been added, a slight band was observed under the main protein band at about 45 kDa (Fig. 27). This slight band corresponded to degraded CSE and was present even at the beginning (t = 0min) due to the utilization of an older stock of protein for this experiment. From the initial screen that was performed, trypsin, elastase, thermolysin, pepsin, substilisin and protease K were found unsuitable as the cleavage process had been too rapid, resulting in many proteolytic fragments or even complete cleavage where no protein fragments were observed after the experiment. On the other hand, the action of either Endoproteinase (EP) Lysine-C, EP glutamic acid-C or EP arginine-C was found to be too minimal since the CSE enzyme seemed to have remained intact even after a 2 h incubation with the proteolytic enzymes. As for chymotrypsin, the cleavage process was more controlled since a gradual increase in the amount of proteolytic fragment at approximately 43 kDa was obtained with time. None Protease added 0 Time (min) (kDa) 170 130 95 72 56 43 15 45 Trypsin 120 15 45 Chymotrypsin 120 15 45 120 Endoproteinase Endoproteinase Lysine-C Glutamic acid-C 15 45 120 15 45 120 34 26 Endoproteinase Substilisin Protease added Time (min) (kDa) 170 130 95 72 56 43 34 15 45 Arginine-C 120 15 45 120 Protease K 15 45 120 Elastase 15 Thermolysin 45 120 15 45 120 Pepsin 15 45 120 26 Figure 27. 10 % SDS-PAGE gel analysis on the proteolytic cleavage of CSE. 70 However, as the main body of the CSE enzyme (~45 kDa) was still present despite an incubation of 2 h with chymotrypsin, an optimization of the cleavage process was performed by varying the amount of proteolytic cleavage enzyme used and the incubation times. However, about 50 % of the protein was still observed to be in the uncleaved state despite utilizing longer cleavage times or larger amounts of chymotrypsin (Fig. 28). It was also observed that when very high ratios of protease to CSE were used, the cleavage tended to be less controlled, resulting in the formation of more fragments within a shorter period of incubation time. It may therefore be advisable to perform the cleavage slowly using smaller amounts of chymotrypsin, and purify the various proteolytic fragments subsequently via gel filtration to obtain the desired fragment for crystallization. This experiment was however only a preliminary study to determine the feasibility of proteolysis in facilitating protein crystallization. Much work would hence still be needed to determine a suitable protocol for the cleavage process. Chymotrypsin : CSE ratio Time (hrs) 0 (kDa) 95 72 56 43 34 1 : 200 2 3 4 1 : 400 5 2 3 4 1 : 50 5 (min) 15 30 50 1 : 100 75 100 15 30 50 75 100 26 Figure 28. Optimization of the proteolysis of CSE by chymotrypsin. 5.4. Conclusion In this chapter, the three-dimensional structure of the human CSE enzyme as well as significant differences between the apoenzyme (determined in our study) and 71 holoenzyme (determined by our collaborator) were presented and discussed. The impetus for the formation of the CSE apoenzyme was believed to be the addition of L-cysteine into the crystallization condition which in turn triggered the dissociation of PLP from the purified enzyme. This hypothesis where a dissociation of PLP would occur during the process of H2S production would be further explored subsequently. Having elucidated the active site of the enzyme, various amino acids that could possibly play a role in the production of H2S would also be studied via site-directed mutagenesis in the next chapter. Based on the structure of the CSE-PAG inhibitor complex determined by our collaborator, we had also proposed an inhibition mechanism as well as further investigated the role of Tyr-114 in the inhibition of the enzyme with the aid of site-directed mutagenesis. Although initial crystallizing conditions had also been identified for CSE with Nisobutyryl-D-cysteine and N-isobutyryl-L-cysteine, much work still lies in the optimization of the condition to obtain crystals of good diffraction quality for structure determination. In this section, preliminary experiments for the proteolytic cleavage of CSE to facilitate protein crystallization had also been carried out, though an appropriate protocol remains to be identified. As such, much work still lies in the optimization of the cleavage process. 72 6. Mechanism of H2S production 6.1 Objectives It has been established in Chapter 4 that human CSE is able to utilize L-cysteine as a substrate for production of H2S. We were however also interested on whether the in vivo production of H2S by CSE would be possible from other sulfur-containing compounds present in our bodies. As such, the in vitro release of H2S from various sulfur compounds would be conducted in this section. The results gathered from this experiment could also provide information on the affinity of the enzyme towards certain distribution of functional groups and hence aid in the design of potent inhibitors of H2S production. Previously, we had proposed a mechanism for the production of H2S from L-cysteine (Appendix 3). This mechanism would be further expounded through site-directed mutagenesis experiments to identify amino acid residues which play crucial roles in the production of H2S. These residues were chosen based on their degree of conservation across CSE homologs and other transsulfuration enzymes (Fig. 29), as well as their likely functions as proposed previously by Messerschmidt et al. (2003). For instance, Tyr-114 was believed to activate the incoming substrate for subsequent transaldimination with Lys-212 during substrate binding to the PLP cofactor, while Glu-339 was hypothesized to be the key amino acid in determining whether a β- or γ-lyase reaction would occur upon substrate binding. In this work, other amino acids that could serve as the base for activating the incoming substrate (Fig. 30, purple residues) or affect the binding affinity for the L-cysteine substrate (Fig. 30, green residues), as well as amino acids (Fig. 30, cyan residues) which interact with the PLP cofactor via ionic or hydrophobic interactions would also be studied to determine their roles in the catalysis of H2S production. 73 Figure 29. Alignment of the amino acid sequences of mouse, rat, human, Dictyostelium (slime mold), yeast and Streptomyces CSE as well as E. coli cystathionine-γ-synthase (CGS) and cystathionine-β-lyase (CBL). Amino acid residues which would be mutated in our study are indicated by asterisks. Sequence alignment was performed using ClustalW2 (Labarga et al., 2007). 74 PLP Figure 30. Active site of the human CSE enzyme showing the location of crucial amino acids (only side chains shown) which would be studied by site-directed mutagenesis. The PLP cofactor is shown as grey stick, while the amino acids are colored based on their likely functions or relative position from PLP. Significant hydrogen bonds and polar interactions are depicted by the dotted lines. Figure was produced by Pymol software (DeLano, 2002). As observed in the crystallization of the CSE apoenzyme, prolonged incubation of Lcysteine, which is the substrate for H2S production, with CSE seemed to trigger the dissociation of PLP from the enzyme. Hence, we hypothesized that besides the original role of substrate binding, PLP could also be involved in the production of H2S by being displaced out of the active site during the course of H2S production. For continued catalysis of H2S production, external PLP would hence be required to rebind to the enzyme. A more in-depth study on the kinetics of H2S production would thus be executed 75 in the presence of varying exogenous PLP concentrations to better appreciate the roles played by PLP during the production of H2S. 6.2 Experimental 6.2.1. Assay of H2S synthesis from various in vivo sulfur-containing compounds The production of H2S from various sulfur compounds was assayed in triplicates at 37 °C over 30 min. Each reaction mixture contained saline (10 µL), PLP (2 mM), CSE (5 µg) and 5 mM of various sulfur-containing compounds topped up to a final volume of 100 µL with 0.2 M Tris pH 8.0 buffer. Another set of baseline control tubes were prepared by first adding 100 µL of 10 % w/v trichloroacetic acid to denature the enzyme before addition of the substrate. The procedures for incubation, termination of reaction (with ZnAc followed by TCA), and absorbance measurements (after addition of NNDPD and FeCl3) were similar to that described in Section 3.2.5. 6.2.2. Cloning of pET-22b(+)_CSE The human CSE gene was cloned into the pET-22b(+) vector with EcoRI/XhoI sites using the same protocol as that for pGEX-4T-3-CSE (in Section 3.2.1.) and was fully sequenced. 6.2.3. Bacterial expression and purification of polyhistidine-tagged (His-tagged) CSE 30 ng of the recombinant pET-22b(+)_CSE plasmid was transformed into competent E. coli BL21 cells. The transformants were inoculated into 100 mL of LB-Amp100 and 1 % 76 w/v glucose, and incubated at 37 °C with vigorous shaking until an OD600 of 0.3 was attained, upon which bacterial growth was continued at 20 °C until OD600 reached 0.5. IPTG was then added at a rate of 0.1 mM to induce the expression of human CSE at 20 °C for 20 h. The cells were harvested by centrifugation at 6000 x g for 15 min at 4 °C. The supernatants were removed, and the pellet was resuspended in 10 mL of lysis buffer containing 50 mM NaH2PO4 pH 8.0, 100 mM NaCl, 1 % Triton-X, 10 mM imidazole and 10 mM β-mecaptoethanol (BME). Cell lysis was performed by sonication on ice for 7.5 min. The lysate was then cleared by centrifuging at 10000 rpm for 45 min at 4 °C. The supernatant was introduced separately to 2 mL of nickel and 2 mL of cobalt affinity beads previously equilibrated with 10 mL of the lysis buffer, and then incubated at room temperature for 30 min with slight shaking. The flow-through was collected, and the beads were washed thrice with 5 mL of wash buffer consisting of 50 mM NaH2PO4 pH 8.0, 100 mM NaCl, 1 % Triton-X, 20 mM imidazole and 10 mM BME. Elution of bound proteins was performed using 2.5 mL of elution buffer containing 50 mM NaH2PO4 pH 8.0, 100 mM NaCl, 250 mM imidazole and 10 mM BME. The purification process was monitored by denaturing SDS-PAGE analysis. 6.2.4. Optimization of bacterial induction conditions for His-tagged CSE Five colonies of the pET-22b(+)_CSE transformed BL21 cells were inoculated into 20 mL of LB-Amp100 and incubated at 37 °C overnight with vigorous shaking. 15 mL of the starter culture and 15 mL of 20 % w/v glucose solution was propagated in 270 mL of LB-Amp100. The mixture was incubated at 37 °C with vigorous shaking until OD600 reached 0.1, upon which bacterial growth was continued at 30 °C until OD600 reached 0.5 77 or 1.0. Batches of bacteria were induced with 0.1 mM or 1 mM IPTG at OD600 of 0.5 or 1.0. Protein expression was performed at 30 °C for 4 h or 20 °C for 20 h. 6.2.5. Preparation of mutant CSE clones Mutant strand synthesis was performed using 42 ng of pGEX-4T-3-CSE as template and 125 ng of the respective mutagenic primers as listed in Appendix 2. Sixteen PCR cycles, which consisted of 30 s denaturation at 95 °C, 1 min annealing at 55 °C, and 12 min amplification at 68 °C, were performed using 2.5 U of PfuTurbo DNA polymerase. The non-mutated parental DNA template was then digested using 20 U of DpnI restriction enzyme for 2 h at 37 °C. Transformation of the mutant DNA into competent DH5α cells was performed using 5 µL of the plasmid. 1 colony from each of the transformants was transferred to a new master LB-Amp100 plate, inoculated into 5 mL of LB-Amp100 and cultured overnight at 37 ºC in a shaker incubator. Minipreps were then performed to obtain the mutant DNA from the bacterial cells and the sequence was verified by automated sequencing. 6.2.6. Bacterial expression of mutant GST-tagged CSE proteins 10 ng of each of the mutant and wild-type plasmids was transformed into competent E. coli BL21 cells. The transformants were separately inoculated into 8 mL LB/Amp Amp100 and incubated at 37 °C overnight with vigorous shaking. 6 mL of the starter culture and 6 mL of 20 % w/v glucose solution was propagated in 108 mL of LBAmp100 and incubated at 37 °C with vigourous shaking until OD600 of 0.3 was attained, 78 upon which bacterial growth was continued at 20 °C until OD600 reached 0.5. IPTG was then added at a rate of 0.1 mM to induce the expression of mutant and wild-type CSE at 20 °C for 20 h. The bacteria culture was separated into 50 mL aliquots and the bacterial cells were harvested by centrifugation at 6000 x g for 15 min at 4 °C and subsequently stored at -80 °C until further use. 6.2.7. Optimized procedure for purification of GST-tagged mutant and wild-type CSE The bacterial pellet (from 50 mL of the bacterial culture) was thawed and resuspended in 10 mL of the lysis buffer as described in Section 3.2.7. Lysozyme (0.3 mg/mL, Sigma) was added prior to sonication on ice for 20 min using the 20 % pulsed maximum output of a Misonix XL2010 sonifier equipped with a microtip. The lysate was then cleared by centrifuging at 10000 rpm for 45 min at 4 °C using an AB 50.10A rotor (Thermo Electron Corporation). The supernatant was subsequently introduced to a chromatography column containing 1mL of glutathione sepharose beads previously equilibrated with 10 mL of the lysis buffer. The column was sealed and incubated with slight shaking at room temperature (25 °C) for 45 min. Following that, the flow-through was collected, and the beads were washed thrice with 15 mL of wash buffer (20 mM Tris pH 8.0, 100 mM NaCl, 1 mM DTT) containing either 1 % v/v, 0.5 % or 0.1 % TritonX100 respectively in each of the three washes. Elution of the fusion protein was then performed by adding 10 mL of elution buffer (20 mM Tris pH 8.0, 50 mM NaCl, 1 mM DTT) containing 1 mM of reduced glutathione (Sigma) and incubating at room temperature for 20 min. The eluate was then collected and concentrated using an ultracentrifugal filter (Millipore Amicon Ultra-4 30000 MWCO) at 4 °C. 79 6.2.8. Analysis of protein secondary structure via circular dichroism (CD) measurements Each protein was buffer exchanged into 10 mM sodium phosphate pH 8.2 buffer to a final concentration of about 0.2 mg/mL with the aid of an ultracentrifgal filter (Millipore Amicon Ultra-4 30000 MWCO) at 4 °C. Protein concentrations were then determined with the aid of a NanoDrop ND-1000 spectrophotometer. The spectropolarimeter (Jasco J-810) was then flushed with nitrogen gas, after which the nitrogen gas flow was maintained at a flow rate of 5 L/min. An averaged CD spectrum was then gathered from three separate wavelength scans of 260 nm to 190 nm for each protein. Data analysis was subsequently performed using the CDSSTR algorithm via the online DicroWeb CD analysis software (Whitmore & Wallace, 2004). 6.2.9. Comparison of the H2S synthesizing activities of the CSE mutant proteins The production of H2S from 5 µg of each GST-tagged mutant CSE protein was assayed in a similar way to that described in Section 4.2.4, with the exception of adding NNDPD and FeCl3 directly to the solution for absorbance measurements after the addition of basic ZnAc and TCA. 6.2.10. Kinetics of H2S production under varying exogenous PLP concentrations Stock solutions of PLP and L-cysteine were prepared in 50 mM sodium phosphate pH 8.2 buffer. Each reaction tube consisted saline (10 µL), 3.88 µg of the purified human CSE enzyme, PLP (1.5 µM – 1 mM) and L-cysteine (0.75 – 5 mM) topped up to a final volume of 100 µL with 50 mM sodium phosphate pH 8.2 buffer. The production of H2S was monitored at 5 min intervals over 20 min at 37 °C. Evolved H2S was trapped by addition 80 of a mixture of 0.85 %w/v ZnAc and 3 % w/v NaOH (100 µL) while termination of reaction was performed by addition of 10 % w/v TCA (100 µL). NNDPD dye (20 mM, 42.9 μL) and FeCl3 (30 mM, 42.9 μL) were then added to each tube for absorbance measurements at 670 nm. The amount of H2S produced at each time point was calculated against the calibration curve determined in Section 4.2.3. and subsequently plotted against time to determine the initial reaction velocity (V) under each L-cysteine and PLP concentration. Double reciprocal plots were then obtained by plotting a series of lines under fixed PLP or L-cysteine concentrations to determine the reciprocal values of apparent maximum velocity (1/Vapp), which would then be subsequently plotted against the reciprocal values of PLP or L-cysteine concentrations in a secondary plot to determine the true maximum velocity (Vmax) and Michaelis Menten constants (KM) for PLP and Lcysteine. The graphical software, SigmaPlot (Systat) was used for all curve fitting and regression analyses. 81 6.3. Results and discussion 6.3.1. Assay of H2S synthesis from various in vivo sulfur-containing compounds The sulfur compounds to be assayed were chosen based on their occurrence in humans. Due to the poor solubility of D-cysteine, L-cystine, L-cystathionine, L-homocystine and S(5’-adenosyl)-L-homocysteine at neutral pH or in organic solvents, these compounds were first dissolved in 0.25 N NaOH to obtain a stock concentration of 100 mM before diluting to 10 mM with 0.2 M of Tris pH 8.0 buffer. The pH of the final solution was found to be around pH 8, hence indicating that the added NaOH should not affect the enzyme activity greatly. A control by preparing L-cysteine in the same manner also showed that there were no significant differences in H2S synthesizing activity compared to that prepared in 0.2 M Tris buffer directly (first and last lanes in Fig. 31). Figure 31. A comparison of the net amount of H2S produced over 30 min by 5 mM of various sulfur-containing compounds. From the results, D-cysteine was shown to produce about 4 times less H2S than L-cysteine, indicating that the enzyme is more specific to the L-isomer. Furthermore, the higher 82 reactivity of CSE towards C-γ-S bonds was also revealed by the 2.4-fold increase in amount of H2S produced from DL-homocysteine compared to L-cysteine. This was expected since the elimination of H2S from homocysteine requires the cleavage of C-γ-S bond, while that from cysteine involves breakage of C-β-S bond. The amount of H2S produced from 5 mM L-homocysteine could potentially be higher, since the enzyme is most probably less specific to the D-isomer. For L-cystine and L-homocystine, production of H2S was minimal since they do not possess free sulfhydryl groups. These compounds could however slowly hydrolyze to yield L-cysteine and L-homocysteine respectively, hence accounting for some production of H2S. H2S could also be produced from Lcystathionine since the primary reaction of CSE would convert L-cystathionine to Lcysteine (Equation 5), which could then further react to produce H2S (Equation 6). However, as ammonia is produced in both reactions, the conversion of the L-cysteine product from the first reaction (Equation 5) to H2S in the second reaction (Equation 6) would be much suppressed, according to Le Chatelier’s principle. In this study, no H2S production was observed from L-glutathione, L-methionine, taurine, hypotaurine, Nacetyl-L-cysteine and S-(5’-adenosyl)-L-homocysteine, which was expected since the endogenous production of H2S in humans has been attributed largely to L-cysteine and Lhomocysteine (Chiku et al., 2009) NH2γ β' O H2O, CSE O β' O α' α' + + α HS OH OH S OH NH3 β NH2 NH2 O O L-cystathionine L-cysteine α-ketobutyrate HO β' HS O α' (5) O H2O, CSE OH NH2 L-cysteine H2S + OH + NH3 (6) O pyruvate 83 6.3.2. Cloning of pET-22b(+)-CSE Previously in Section 3.3.3, as the removal of GST-tag from the CSE fusion protein had been found to compromise the final yield of protein due to protein precipitation and degradation, we had attempted to first express the mutant CSE proteins as His-tagged fusion proteins. Removal of the His-tag would not be required since this tag is only 1 kDa in size and thus would not be likely to affect the overall conformation and activity of the protein. However, as this would require the proteins to be expressed from a different expression vector which may or may not succeed, the GST-tag expression system would be utilized instead should the His-tag expression system fail. The human CSE gene was hence cloned into the pET-22b(+) vector which will incorporate a His-tag into the C-terminus of the protein. PCR amplification of the human CSE gene was as efficient as before in Section 3 and gave high yield of the amplified PCR product (Fig. 32A). Subsequent restriction enzyme cleavage and cloning of the CSE insert into the pET-22b(+) vector was also successful as observed from the presence of DNA bands at ~1.2 kb after digestion with the corresponding restriction enzymes (Fig. 32B). The sequence of the CSE insert was also determined to be correct after automated sequencing. 84 A pET-22b(+)-CSE B XhoI /EcoRI cleavage (kb) 10.0 5.0 2.5 1.0 1 + 2 – 2 + (kb) 8.0 6.0 4.0 3.0 2.0 1.5 1 – CSE DNA 10.0 8.0 6.0 5.0 4.0 3.0 2.5 2.0 1.5 1.0 CSE insert Figure 32. (A) PCR amplification of CSE for cloning into pET-22b(+). (B) Restriction enzyme cleaved plasmids indicating the presence of CSE insert which was determined to be correct upon sequencing. 6.3.3. Bacterial expression and purification of His-tagged CSE The bacterial expression of His-tag CSE was first tested on the wild-type enzyme using a protocol based on that for the expression of GST-tagged CSE which had been established during the Honors project. Two different affinity purification methods using Co and Ni affinity resin were tested to determine which would be more effective in purifying the protein. However, SDS-PAGE analysis of the purification process had revealed little differences in the purification efficacy when either resin was used. More importantly, the His-tag CSE seemed to express poorly as observed from the absence of a thick protein band at approximately 45 kDa upon bacterial induction (Fig. 33A). The original protocol for induction of the fusion protein was therefore unsuitable and optimization of the induction condition had to be performed. However, despite varying induction temperatures, OD600 at which bacteria was induced and IPTG concentrations, the expression of His-tagged CSE still failed (Fig. 33B). The purification of CSE mutants would therefore have to be based on GST-tag expression system. A comparison of the 85 H2S synthesizing activities of GST-tagged and GST-cleaved wild-type CSE had revealed that the presence of the GST-tag would not affect the enzyme activity greatly. As such, the mutant CSE proteins would all be purified in the GST-tagged form subsequently to minimize problems with protein precipitation and degradation during thrombin cleavage of the GST-tag as described in Section 3.3.3. A B Time (h) after induction 0.1 mM IPTG (kDa) 200 116 97.4 66.2 t=0 - t = 20 - + Temperature Induction at Time (h) after induction 1.0 mM IPTG 0.1 mM IPTG (kDa) 200 30 °C OD600 = 0.5 t=0 20 °C OD600 = 1.0 OD600 = 0.5 t=4 t=0 t=4 t=0 t = 20 + - - + OD600 = 1.0 t=0 t = 20 + - - + - - + - - + 116 97.4 66.2 45.0 45.0 31.0 31.0 Figure 33. 10 % SDS-PAGE gel analysis showing attempted expression of His-tag CSE in the presence of 0.1 mM IPTG at 20 °C for 20 h (A) and optimization of bacterial expression conditions (B). Broad molecular range protein ladder (Biorad) was loaded in Lane 1 for both gels. 6.3.4. Preparation of mutant CSE clones An amplification time of 1 min/kb of plasmid length, as recommended by the sitedirected mutagenesis kit manual from Strategene, was utilized initially in the thermal cycling conditions. However, this failed to yield any PCR product. The amplification time was subsequently doubled and significant amounts of mutant CSE plasmids could generally be produced (Fig. 34). For certain mutant plasmids such as those for the mutation of Lys212ÆAla (K212A) and Asn161ÆGln (N161Q), smaller amounts of PCR 86 product was obtained most probably due to a higher energy cost of mismatch between the mutagenic primers and the PCR template. Nevertheless, the amount of PCR products produced was sufficient for subsequent steps and hence the thermal cycling conditions were not optimized. Figure 34. 0.8 % agarose gel showing PCR amplification of various mutant pGEX-4T-3CSE plasmids. 6.3.5. Bacterial expression and purification of mutant GST-tagged CSE proteins The protocol for bacterial induction of GST-tagged mutant CSE proteins was based on that established in Section 3, albeit in a smaller production scale. In each induction experiment, a control experiment using wild-type CSE was performed. The overexpression of GST-tagged mutant proteins and wild-type CSE was successful for all the different mutant proteins that were to be expressed as observed from the SDS-PAGE gel analysis of the induction experiment (Fig. 35). 87 Y114F K212A E339A E339Y Y114A K212R E339K R375A S340A E349A IPTG – + – + – + – + – + – + – + – + – + – + (kDa) Y60A E157A N161A D187A Wild – + – + – + – + – + 68 43 E157D D187E R62A T189A S340T IPTG – + – + – + – + – + (kDa) R375K Y60T S209A T211A F190A – + – + – + – + – + F190Y N161Q R62K – + – + – + T189S Wild – + – + 68 43 Figure 35. 10 % SDS-PAGE analysis of the induction of GST-tagged mutant and wildtype CSE proteins. Previously, the purification of GST-tagged CSE had been carried out in a different lab. As such, when the same protocol was utilized for purification of the mutant CSE proteins in the current lab, several problems were encountered. It was hence necessary to reoptimize some of the purification procedures to ensure maximal efficiency and protein yield. In this work, the amount of lysozyme added to aid in the lysing of bacterial cells was optimized to 0.3 mg/mL. High concentrations of lysozyme were found to lyse the bacterial cells too quickly, hence causing large amounts of the fusion protein to be released as insoluble inclusion bodies. Moreover, the sudden lysis of bacterial cells was also accompanied with the release of large amounts of genomic DNA, which caused the lysate to be too viscous for effective sonication. Incubation of the cell mixture with lysozyme before sonication was also found to be unfavorable as the mixture was often too viscous for effective sonication subsequently. Lysozyme was hence added just prior to sonication so that both enzymatic and mechanical actions could operate together to lyse the bacterial cells and solubilize the GST-fusion protein more effectively. In our study, the concentration of reduced glutathione for elution of the bound fusion protein had also been optimized to 1 mM as widespread protein precipitation was 88 observed when high concentrations of reduced glutathione were utilized. A longer incubation time served to compensate for the lower concentration of reduced glutathione used. The elution buffer for this process had also been optimized to a Tris-based buffer rather than the previous sodium phosphate-based buffer to minimize protein precipitation. 6.3.6. Analysis of protein secondary structure via CD measurements To enable a more meaningful comparison of the activities of the mutant CSE proteins subsequently, it was crucial to ensure that any changes in enzyme activity was not a result of destruction in overall protein conformation, which may occur due to mutagenic alterations to the active site amino acids. As such, the secondary structures of the mutant proteins were determined and compared to that of the wild-type protein by means of CD 0.28 0.19 0.23 0.27 0.19 0.23 0.28 0.19 0.24 0.27 0.20 0.21 0.28 0.20 0.24 0.28 0.19 0.21 0.29 0.18 0.23 0.29 0.23 Unordered Turns Strand R62K Y114A Y114F E157A E157D N161A N161Q D187A D187E T189A T189S 0.30 0.30 0.28 0.31 0.27 0.32 Helix 0.30 0.31 0.29 0.17 * 0.30 0.00 * 0.17 0.27 0.19 0.33 0.60 0.22 0.28 0.20 0.80 0.30 R62A 0.22 0.27 0.19 0.31 Y60T 0.32 0.31 Y60A 0.22 0.28 0.19 0.28 0.20 0.22 0.27 0.19 0.40 0.23 0.28 0.60 0.24 0.80 0.20 1.00 0.24 measurements. wild 0.28 0.28 0.27 0.29 0.28 0.29 0.29 0.27 0.28 0.28 0.19 0.18 0.20 0.19 0.20 0.20 0.19 0.20 0.18 0.19 0.22 0.23 0.25 0.23 0.24 0.23 0.25 0.24 0.22 0.31 0.31 0.28 0.29 0.28 0.28 0.28 0.29 0.32 Unordered Turns Strand Helix 0.30 0.23 0.29 0.27 0.18 0.21 0.34 0.19 0.27 0.19 0.23 0.30 0.22 0.26 0.19 0.21 0.34 0.29 0.28 0.19 0.20 0.22 0.40 0.30 1.00 0.00 F190A F190Y S209A T211A K212A K212R E339A E339K E339Y S340A S340T E349A R375A R375K wild Figure 36. Proportion of α-helices, β-sheets, turns and unordered regions of GST-tagged mutant and wild-type CSE proteins. The normalized root mean square deviation for the 89 computed percentages of secondary structures were all well below 0.1, indicating that the results were meaningful and reliable. Overall protein conformation was found to be relatively similar for all proteins with the exception of the E157A and E157D mutants. The results revealed that the overall conformation of majority of the mutant CSE proteins was maintained when compared to that for the wild-type protein. However, significant changes in secondary structure composition, and most probably overall protein conformation, were observed for the E157A and E157D mutant proteins (Fig. 36). As revealed by the crystal structures of both human and yeast CSE holoenzyme, the carboxylic acid side chain of Glu-157 was hydrogen bonded to the amino hydrogen of Tyr-114 (Fig. 37A). However, in the conformationaly much more open CSE apoenzyme structure, this hydrogen bonding was no longer possible as the Tyr-114 residue was too distant from the Glu-157 amino acid (Fig. 37B). In our study, this hydrogen bonding interaction would also be unattainable for the methyl side chain of alanine in the E157A mutant, whilst for the E157D mutant protein, this interaction may likely be affected due to a shorter carboxylic acid side chain in the aspartate amino acid. The observed differences in secondary structures and hence overall protein conformation of the E157A and E157D mutant proteins may therefore be attributed to a a displacement of the Tyr114 residue as a result of a disruption to the above-mentioned hydrogen bonding interaction. 90 A B PLP Figure 37. Distances (in angstroms) between the polar contacts of the carboxylic acid side chain of Glu-157 and amino group of Tyr-114 in the CSE holoenzyme (A) and apoenzyme (B). Figure was produced by Pymol software (DeLano, 2002). Net amount of H2 S produced (nmol) 6.3.7. Comparison of the H2S synthesizing activities of the CSE mutant proteins 60.0 * 50.21 50.0 40.0 30.0 20.0 10.0 0.0 * * * * * * 0.11 0.11 0.28 1.10 0.24 0.05 * * 5.90 5.38 9.88 * 0.04 * 0.46 * 0.16 8.62 * 0.11 * 0.11 Figure 38. A comparison of the H2S synthesizing abilities from 5 µg of various GSTtagged CSE alanine mutants against wild-type GST-CSE. Each assay was performed in duplicates in the presence of 2.75 mM L-cysteine and 0.5 mM PLP. The results are displayed as the mean from three independent runs ± SD. *P < 0.05 compared to wildtype GST-tagged CSE enzyme. An initial mutagenic conversion of the active site amino acids depicted in Fig. 30 to alanine led to several interesting changes in the H2S synthesizing activities compared to 91 the wild-type CSE enzyme. As shown in Fig. 38, most of the alanine mutants displayed a significant decrease in amount of H2S produced compared to wild-type CSE, hence suggesting the importance of these amino acid residues in maintaining the enzyme activity. For the E157A mutant protein, this loss in enzyme activity was expected since the above CD experiments had suggested the likely destruction of the protein conformation upon mutation of the Glu-157 residue. Interestingly, the Glu-339 residue which had previously been attributed to be one of the key amino acid residues in determining the enzymatic specificity of transsulfuration enzymes (Messerschmidt, et al., 2003), had displayed a significant increase in H2S synthesizing activity upon mutation to alanine (E339A). A few other alanine mutants (Fig. 38, S209A, T211A and E349A) were however found to display comparable enzyme activities to that for the wild-type enzyme, hence suggesting that the Ser-209, Thr-211 and Glu-349 residues play less crucial roles in the production of H2S. A further discussion on the mutation of these residues to alanine would be provided below. Having identified crucial amino acid residues which would greatly affect the catalysis of H2S production, a further exploration of the role of these residues was conducted by a subsequent mutation to a corresponding amino acid with similar side chain properties. This would also be further elaborated in each of the sections below. Mutant CSE proteins affecting the binding of PLP cofactor From the initial screen of enzyme activities of various CSE alanine mutant proteins (Fig. 38), the K212A mutant protein was one of the proteins which were originally expected to display a loss in enzyme activity. This was because covalent binding of the PLP cofactor 92 would not be achievable upon mutation of the Lys-212 residue to alanine. Moreover, as the Lys-212 residue is required for a transaldimination reaction with the L-cysteine substrate during substrate binding, this reaction would no longer be possible for the K212A mutant protein. This hypothesis was tested by a corresponding mutation of the Lys-212 residue to arginine (K212R) which possesses similar side-chain properties as the lysine residue (Fig. 39). However, a failure of the K212R mutant protein in restoring the enzyme activity may suggest that the Lys-212 residue is irreplaceable. Indeed, this residue was found to be fully conserved across all CSE homologs and transsulfuration enzymes (Fig. 29), thus explaining the loss of enzyme activity for the K212R mutant protein as well. Previous reports had proposed that the deprotonated phenolic side-chain of Tyr-114 acts as the base for activation of the substrate (by abstracting a proton from the amino group) for subsequent binding to the enzyme in CSE (Messerschmidt et al., 2003) and most other transsulfuration enzymes (Clausen et al., 1996; Clausen et al., 1998). This would imply that activation of the substrate for subsequent H2S production would not be possible if the Tyr-114 residue was mutated to phenylalanine (which lacks the phenolic side-chain). Interestingly though, when this mutation was performed, the H2S synthesizing activity of the Y114F mutant protein was increased greatly when compared to that for the wild-type enzyme (Fig. 39). The loss of enzyme activity for the Y114A mutant is hence most probably attributed to other reasons. According to the crystal structure of the CSE holoenzyme, the Tyr-114 residue was observed to form π-stacking interactions with the PLP cofactor (Fig. 20A). This interaction would however not be attainable upon mutation 93 of Tyr-114 to alanine. As such, it is likely that the region corresponding to the mutated Ala-114 residue becomes much more flexible, and the affinity of the mutant protein for PLP becomes lowered, thus leading to a loss in enzyme activity. A possible reason for the increased activity for the Y114F mutant protein would be provided at a later stage. *# * * * * *# * *# Figure 39. H2S synthesizing activities of mutant CSE proteins with mutagenic alterations to the Lys-212, Tyr-114, Asn-161 and Phe-190 residues, assayed in the presence of 5 µg GST-tagged protein, 0.5 mM PLP and 2.75 mM L-cysteine. The results are displayed as the mean from three independent runs ± SD. *P < 0.05 compared to wild-type GSTtagged CSE enzyme. #P < 0.05 compared to each of the corresponding alanine mutant protein. Other active site amino acids that were studied included the Asn-161 and Phe-190 residues. These residues are located in close proximity to the hydroxyl and methyl substituents of the pyridine ring in the PLP cofactor (Fig. 30). In the crystal structure of CSE complexed with PAG, Asn-161 was found to be one of the eight residues which interact with PAG (Sun, et al., 2008). As the polar amide side chain of Asn-161 is also in close proximity to the hydroxyl substituent in the PLP cofactor, possible hydrogen bonding interactions between these two functionalities are likely to exist as well. 94 Mutation of Asn-161 to alanine (N161A) in our study was found to render the enzyme inactive towards H2S production, probably due to a failure in establishing stabilizing interactions between PLP and Ala-161 in the mutant protein. However, an attempt in rescuing the H2S synthesizing activity of the enzyme by a corresponding mutation of Asn-161 to glutamine (N161Q) had not been successful (Fig. 39). This was probably due to the high degree of conservation of the Asn-161 residue in CSE homologs (Fig. 29). As for Phe-190 which is also conserved in most CSE homologs, no reports had suggested any possible roles for this residue. From our mutagenesis results however, the F190A mutant protein had not been able to catalyze the production of H2S. From a corresponding mutation of Phe-190 to tyrosine, a significant 10-fold increase in enzyme activity compared to that for the alanine mutant suggests that this residue was probably involved in the catalysis of H2S production by forming significant hydrophobic interactions with the PLP cofactor (Fig. 39). Mutant CSE proteins affecting the activation of L-cysteine substrate Having discovered that Tyr-114 was most probably not the amino acid residue for activation of the L-cysteine substrate during the catalysis of H2S production, we proceeded to mutate other residues which were in close proximity to the substrate binding site and that could possibly take on this role of substrate activation. These residues comprised of Tyr-60, Arg-62, Thr-189, Ser-209, Thr-211 and Ser-340. Although most of these amino acids were not sufficiently basic on their own for activation of the L-cysteine substrate, the basicity of these residues could possibly be increased due to the extensive hydrogen bonding network to the PLP cofactor or neighboring active site residues. An 95 initial screen of the alanine mutants of these active site amino acids revealed only a slight decrease in production of H2S from the S209A and T211A mutants (Fig. 38), hence suggesting that Ser-209 and Thr-211 were not responsible for the activation of the substrate. Most of the other residues were however deemed to be crucial in the production of H2S, as observed from their significant loss in enzyme activity upon mutation to alanine (Fig. 38). An experiment where each of these amino acids was converted to a corresponding amino acid of similar side chain functionality was hence conducted subsequently to determine the most likely residue which served as the base for Net amount of H2S produced (nmol) deprotonating L-cysteine in the catalysis of H2S production. 10.0 8.62 8.0 *# 5.13 6.0 *# 2.97 4.0 2.0 0.0 * 1.10 * 1.70 * 0.24 *# 0.80 * 0.46 * 0.05 Figure 40. H2S synthesizing activities of mutant CSE proteins with mutagenic alterations to the Tyr-60, Arg-62, Ser-340 and Thr-189 residues, assayed in the presence of 5 µg GST-tagged protein, 0.5 mM PLP and 2.75 mM L-cysteine. The results are displayed as the mean from three independent runs ± SD. *P < 0.05 compared to wild-type GSTtagged CSE enzyme. #P < 0.05 compared to each of the corresponding alanine mutant protein. For both Tyr-60 and Arg-62, a corresponding mutation to threonine (Y60T) and lysine (R62K) respectively was found to increase the enzyme activity only very slightly (Fig. 96 40). This was likely due to the importance of Tyr-60 and Arg-62 in maintaining the active site dimer formation where the phosphate group of the PLP cofactor forms extensive hydrogen bonds to Tyr-60 and Arg-62 from the neighboring subunit (Fig. 41). As with the Lys-212 residue, the results suggest that these two residues are irreplaceable. From the alignment of CSE homologs and transsulfuration enzymes (Fig. 29), the Tyr-60 and Arg-62 residues were indeed observed to be fully conserved. Mutation of either of these two amino acids may thus weaken the interactions between adjacent monomers and lower the binding affinity of PLP to the enzyme, hence leading to a corresponding decrease in enzyme activity. PLP(C) A C Tyr-60(D) Arg-62(D) Arg-62(C) B D Tyr-60(C) PLP(D) Figure 41. (left) The human CSE tetramer made up of a dimer of dimers. (right) Magnification of the interactions between PLP and Tyr-60 and Arg-62 from the adjacent monomer in subunits C and D of the enzyme. Figure was produced by Pymol software (DeLano, 2002). On the contrary, partial restoration of enzyme activity was observed for the S340T mutant protein, indicating that Ser-340 could be one of the active site amino acids for activating the L-cysteine substrate towards binding. A further comparison with the T189S 97 mutant protein revealed that Thr-189 was the most probable amino acid for this function as the largest restoration in enzyme activity had been observed (Fig. 40). Theoretically speaking, the hydroxyl side chain of the Thr-189 amino acid should not be sufficiently basic on its own for activating the L-cysteine substrate for binding to the PLP cofactor in the enzyme. This basicity was in actuality increased upon formation of a strong hydrogen bond to the carboxylate side chain of the Asp-187 amino acid, which is Lys-212 N H OH 2- O3PO N H H O O O Asp-187 Thr-189 Net amount of H2S produced (nmol) in turn hydrogen bonded to the pyridoxal hydrogen of the PLP cofactor (Fig. 42). 10.0 8.62 8.0 6.0 4.0 2.0 0.0 * * 0.04 0.04 Figure 42. (left) Schematic representation of the hydrogen bonding network involving Thr-189, Asp-187 and the PLP cofactor. Figure was produced by Pymol software (DeLano, 2002). (right) H2S synthesizing activities of mutant CSE proteins with mutagenic alterations to the Asp-187 residue, assayed in the presence of 5 µg GSTtagged protein, 0.5 mM PLP and 2.75 mM L-cysteine. The results are displayed as the mean from three independent runs ± SD. *P < 0.05 compared to wild-type GST-tagged CSE enzyme. As observed by mutagenic studies, mutation of the Asp-187 residue to alanine (D187A) was found to compromise the production of H2S drastically (Fig. 38). This was in accordance with the above observations where mutation of the amino acids which interact with PLP (such as Lys-212 and Tyr-114) to alanine largely led to a loss in enzyme 98 activity, most probably due to a decrease in affinity for the cofactor. More importantly, as discussed above, the hydrogen bonding network between Thr-189 and PLP would be disrupted upon mutation of the Asp-187 residue to alanine. This would in turn lower the basicity of the Thr-189 residue, indicating that the crucial step of substrate activation would not be possible for production of H2S. A subsequent study on the mutation of the Asp-187 amino acid to glutamate (D187E) also strongly suggests the intolerance of the enzyme to amino acid changes at the 187th position as observed from the absence of H2S production despite mutation to an amino acid with similar side chain properties (Fig. 42). Mutant CSE proteins affecting the affinity of the enzyme for L-cysteine Having established the amino acids involved in the activation of the L-cysteine substrate during its binding, it would be interesting to determine the residues which would affect the enzyme’s affinity for L-cysteine as well during the production of H2S. Previously, Messerschmidt et al. (2003) suggested the role played by the Glu-339 (numbering in human CSE) residue in differentiating between the β-lyase and γ-lyase activities of transsulfuration enzymes. In his study, a hydrophobic amino acid at this position would favor the orientation of the substrate in such a way that would promote a β-lyase reaction. We hence proceeded to test this hypothesis by mutating the Glu-333 residue to the less polar alanine amino acid (E339A), and had indeed observed a significant increase of approximately 6-fold compared to that of wild-type CSE (Fig. 38). Similarly, a mutation of Glu-339 to tyrosine (E339Y), which is the corresponding amino acid at this position in bacterial cystathionine β-lyase, led to an increase in production of H2S by about 6-fold as well (Fig. 43). A rough estimation of the KM value for either of these two mutant proteins 99 was deemed to be about 0.16 mM, which was about one-third that for the wild-type enzyme (Fig. 43), hence supporting the claim of an increased affinity for the L-cysteine substrate in the presence of a more hydrophobic residue at the 339th position in human CSE. Interstingly, the production of H2S had not been compromised even upon mutation of the negatively charged Glu-339 residue to the positive charged lysine amino acid (E339K). This indicates that the charge of the amino acid at this position does not play a role in the production of H2S. Rather, it is the degree of hydrophobicity of this amino acid which would determine the enzyme’s activity towards H2S production. As shown in Table 9, the trend in production of H2S could indeed be predicted by the logP value which gives an idea of the hydrophobicity of the mutated residue. For the E339K mutant protein, the enzyme activity was observed to be much lower than that for either E339A or E339Y since the lysine residue is less hydrophobic than alanine and tyrosine. On the other hand, as lysine is more hydrophobic compared to glutamate, the amount of H2S produced was larger compared to that for the wild-type enzyme. Comparing E339A and E339Y, the activity of the E339Y mutant protein was slightly higher, possibly due to a higher logP value for tyrosine as opposed to alanine. 100 * * *# Figure 43. H2S synthesizing activities of mutant CSE proteins with mutagenic alterations to the Glu-339 residue. The results are displayed as the mean from three independent runs ± SD. *P < 0.05 compared to wild-type GST-tagged CSE enzyme. #P < 0.05 compared to the E339A mutant protein. Kinetic parameters for E339A, E339Y and wild-type CSE were determined from 4 µg of the GST-tagged protein in the presence of 2 mM PLP, using a similar protocol to that described in Section 4.2.1. Table 9. Correlation between logP values and production of H2S for the various mutated amino acids at 339th position of human CSE ^ Amino acid at 339th position logP^ Net production of H2S (nmol) Glutamate (wild-type CSE) -3.69 8.62 Lysine (E339K) -3.05 15.54 Alanine (E339A) -2.85 50.21 Tyrosine (E339Y) -2.26 51.33 Values obtained from http://www.ecosci.jp/amino/amino2j_e.html Besides Glu-339, other amino acids were also postulated to play a role in the binding of L-cysteine during H2S production. The guanidine side chain of Arg-375 for instance, was found to be in a proximal distance for forming ionic interactions with the carboxylate group of the substrate. This amino acid, which is fully conserved across all CSE homologs and transsulfuration enzymes, also interacts with the neighboring Glu-349 101 residue via an ionic bonding network as shown in Fig. 44. As predicted, mutation of the Arg-375 residue to alanine (R375A) led to a loss in enzyme activity (Fig. 38) since the stabilizing ionic interaction involving the carboxylate group of the substrate would not have been possible for the alanine amino acid. A failure in rescuing the activity of the enzyme by a corresponding mutation of Arg-375 to the basic lysine amino acid (R375K) further revealed the importance of the highly conserved Arg-375 residue in maintaining the enzyme activity. Only a slight difference in amount of H2S produced was however noted for the E349A mutant protein (Fig. 38), suggesting that Glu-349 plays a less crucial role in the production of H2S. The slight increase in enzyme activity for the E349A mutant protein could be due to a disruption to the original ionic interaction between Glu349 and Arg-375, hence increasing the electron density on the guanidine functionality for a stronger ionic interaction to be formed to the L-cysteine substrate. H Lys-212 O HS H3N O CSE O HS O NH3 NH2 Arg-375 N 2- PLP N OH O3PO N H Figure 44. Ionic interactions involving Arg-375 before (left) and after (right) the binding of L-cysteine substrate. 6.3.8. Kinetics of H2S production in the presence of varying PLP concentrations In the development of an assay for determination of the catalysis of H2S production from rat liver or kidney tissues (Stipanuk & Beck, 1982), the addition of exogenous PLP was found to be crucial for maximal enzymatic activity. As speculated by observations from 102 the crystallization of human CSE in the previous chapter, the PLP cofactor seemed to be mobile and could possibly be released from the active site of the enzyme during the course of H2S production. This was hence one of the possible reasons for increased enzyme activity upon the exogenous addition of PLP. As a preliminary testing, the amount of H2S produced within a fixed time period had indeed been observed to increase upon an initial increase in the concentration of exogenous PLP (Fig. 45). This was however followed by a decrease in H2S production as the concentration of exogenous PLP was further increased. An explanation for this phenomenon could be that under exceedingly high concentrations of PLP, the probability of the L-cysteine substrate being bound to the exogenous cofactor instead of the enzyme would be higher. This would lead to a decreased effective concentration of the substrate, thus lowering the amount of H2S produced. From this experiment, an optimum exogenous concentration of PLP (in the presence of 2.75 mM L-cysteine substrate) had been determined to be about 0.5 mM. This was hence the concentration of PLP that was utilized in the assays that were performed previously for screening the inhibitor candidates (Section 4.3.6.) and the activities of Amount of H2 S produced (nmol) various mutant CSE proteins (Section 6.3.7). 16.0 14.0 12.0 10.0 8.0 6.0 4.0 2.0 0.0 0 400 800 1200 1600 2000 Concentration of exogenous PLP (uM) Figure 45. Amount of H2S produced in 30 min under different exogenous concentrations of PLP, assayed in the presence of 7.9 µg GST-tagged CSE and 2.75 mM L-cysteine. Each data point was displayed as the mean of triplicate readings ± SD. 103 As it was possible for the production of H2S to become limiting over time, a more accurate reflection of the dependence of the catalysis of H2S production on the exogenous PLP concentration should be determined from a plot of initial reaction velocity (V) against concentration of PLP. In addition, to aid in the determination of various kinetic parameters subsequently, a series of initial reaction velocity measurements was performed by keeping the concentration of L-cysteine constant while varying the concentration of exogenous PLP. As observed from the graph of V against concentration of exogenous PLP (Fig. 46A), the rate of production of H2S was also found to increase with an initial increase in PLP concentration for each concentration of L-cysteine that was assayed. Although this increase in reaction velocity could be due to the enzyme being purified in a state that was not saturated with the PLP cofactor, the reaction velocity had been observed to increase progressively beyond 1.5 µM of exogenous PLP concentration (which was the molar concentration of PLP required to fully saturate the amount of enzyme utilized in the assay). As with Fig. 45, the reaction velocity was found to decrease beyond an optimum concentration of exogenous PLP (Fig. 46A). The extent of decrease in reaction velocity was greater under low concentrations of L-cysteine and lesser at high concentrations of L-cysteine. This was likely since under low concentrations of L-cysteine, the enzyme would have to compete more readily with the excess PLP for the limited amount of substrate that was present. However, when larger amounts of L-cysteine were added, the extent to which the effective substrate concentration was lowered would be lesser since a smaller proportion of the substrate would be bound to the excess PLP cofactor. The result of this phenomenon was more clearly expressed in the graph of initial reaction velocity against L-cysteine concentration, 104 where a sigmoidal curve was obtained instead of a hyperbolic curve at a concentration of 1 mM PLP (Fig. 46B). This implied that the sigmoidal curve obtained in Section 4.3.1 for determination of the kinetics of H2S production (Fig. 10A) was probably not a result of cooperativity effects as assumed previously, but rather a result of decreased effective substrate concentration due to binding with exogenous PLP in that assay. B 0.14 0.75mM L-cysteine 0.12 0.10 1.5mM Lcysteine 0.08 3mM Lcysteine 0.06 0.04 5mM Lcysteine 0.02 0.00 0 200 400 600 800 1000 Concentration of exogenous PLP (uM) Initial reaction velocity,V (U/mg) Initial reaction velocity,V (U/mg) A 0.16 0uM PLP 1.5uM PLP 20uM PLP 40uM PLP 150uM PLP 1mM PLP 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 0.00 1.00 2.00 3.00 4.00 5.00 Concentration of L-cysteine (mM) Figure 46. (A) Graphs of initial reaction velocity, V (expressed as U per mg purified human CSE, where 1 U = 1 µmol H2S produced per min) against concentration of exogenous PLP determined under various concentrations of L-cysteine substrate. (B) Graphs of initial reaction velocity, V against L-cysteine substrate concentration for the various concentrations of PLP that was added in the assay. Given that the previous determination of kinetic parameters for H2S production in Section 4.3.1. was conducted under an excessively high concentration of exogenous PLP, it was necessary to redetermine these parameters by taking into account the concentration of PLP that was added in the assay. To achieve this, a series of double reciprocal plots was obtained by keeping the concentration of L-cysteine constant while varying the reciprocal of PLP concentration (1/[PLP]) as shown in Fig. 47A. Due to the decrease in reaction velocity under high concentrations of exogenous PLP, these data points were omitted in the generation of the double reciprocal plots. A secondary plot was then obtained by plotting the y-intercepts which also give the reciprocal values of the apparent maximum 105 velocity (1/Vapp) against the reciprocal of the corresponding L-cysteine concentration (1/[L-cysteine]) as shown in Fig. 47B. The reciprocal of the y-intercept for this secondary plot would give the true maximum velocity, Vmax for the catalysis of H2S production, which was determined to be 0.195 U per mg of human CSE. The true KM value for Lcysteine, taking into consideration the effects of various exogenous concentrations of PLP, can also be determined from the negative reciprocal of the x-intercept to be 1.38 mM. This value was expected to be smaller than that determined in Section 4.3.1. where 2 mM PLP was included in the assay, since a larger concentration of L-cysteine would have to be added previously to attain an initial reaction velocity of half Vmax, given that a portion of the substrate would be bound to the excess PLP cofactor instead. A B 1/V (min.mg.umol-1) y = 6.7497x + 14.881 16.0 y = 6.0906x + 9.2781 12.0 y = 3.3726x + 7.5086 8.0 y = 3.1344x + 6.9234 4.0 0.00 0.20 0.40 1/[PLP] (uM-1) 0.60 0.75mM L-cysteine 1.5mM L-cysteine 3mM L-cysteine 5mM L-cysteine 1/Vapp (min.mg.umol-1) 16.0 20.0 14.0 12.0 y = 7.1184x + 5.1395 R² = 0.9858 10.0 8.0 6.0 4.0 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1/[L-cysteine] (mM-1) Figure 47. (A) Double reciprocal plots for the various concentrations of L-cysteine substrate that was added in the assay. Data points beyond 150 µM exogenous PLP concentration were omitted from this graph due to apparent decrease in reaction velocities under these conditions. (B) Secondary plot for determination of the true Vmax and KM values for L-cysteine. Likewise, by keeping the exogenous PLP concentration constant, a series of double reciprocal plots could also be generated under varying L-cysteine concentrations. A subsequent generation of the secondary plot would then give the corresponding Vmax and KM values for PLP, which were determined to be 0.190 U per mg CSE and 0.48 µM 106 respectively. This value of Vmax is in close agreement to that determined from the former secondary plot for L-cysteine, indicating that the approximation for the true Vmax value in both cases was accurate. A B 1/V (min.mg.umol-1) 16.0 y = 9.3422x + 6.938 y = 7.3216x + 5.5938 12.0 8.0 y = 6.8842x + 5.2344 y = 7.2542x + 5.2083 4.0 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1/[L-cysteine] (mM-1) 1.5uM PLP 20uM PLP 40uM PLP 75uM PLP 1/Vapp (min.mg.umol-1) 7.5 20.0 7.0 6.5 y = 2.519x + 5.2682 R² = 0.9709 6.0 5.5 5.0 4.5 0.00 0.20 0.40 0.60 0.80 1/[PLP] (uM-1) Figure 48. (A) Double reciprocal plots for the various concentrations of exogenous PLP that was added in the assay, up to 75 µM. Plots beyond 75 µM exogenous PLP concentration were omitted from this graph due to apparent decrease in reaction velocities under these conditions. (B) Secondary plot for determination of the true Vmax and KM values for PLP. 6.3.9. Proposed mechanism for catalysis of H2S production by human CSE Taken together, the above site-directed mutagenesis studies on the active site residues of human CSE as well as in-depth kinetic studies have provided greater insights into the mechanism for H2S production. In particular, the double reciprocal plots for both Lcysteine and PLP (Fig. 47A and Fig. 48A) have revealed that the binding of these two substances most probably follows a sequential rather than a ping-pong mechanism. However, as the L-cysteine substrate can only bind to the enzyme through the PLP cofactor, the binding of these two substances is considered to be an ordered rather than a random process. Our kinetic studies have also supported the original hypothesis that the cofactor may be displaced from the enzyme during the course of the reaction, hence the 107 addition of exogenous PLP could enhance the catalysis of H2S production. Bearing in mind these observations as well as various crucial amino acid residues identified in our site-directed mutagenesis studies, a more detailed mechanism for the production of H2S was proposed (Fig. 49). 108 Lys-212 OH H N H2N NH2 NH2 Arg-375 H O Tyr-114 Apo-CSE Thr-189 O H O PLP Asp-187 Lys-212 H2N N OH O H N HS NH2 N Arg-375 H H O Tyr-114 O Thr-189 H Lys-212 O NH2 O NH2 OH 2-O PO 3 HS O OH OH Tyr-114 O NH2 OH OH OH Step 2 Transaldimination HO N H O OH Step 3 α-protein abstraction H N H2N NH2 O NH3 Tyr-114 O N N H 2-O O OH 3PO N H O H Asp-187 Thr-189 - H 2S NH2 Arg-375 N 2-O Step 4 β-cleavage and release of H2S N O Lys-212 O O Thr-189 Asp-187 Arg-375 OH 2-O PO 3 NH2 Arg-375 HS N +H2O - PLP N N 2-O PO 3 Tyr-114 H2N Lys-212 OH OH Thr-189 OH H O O Lys-212 O Asp-187 PLP Step 5 Release of PLP-iminopropionate intermediate and rebinding of PLP + NH3 HS Arg-375 HO N H H NH2 N NH2 Asp-187 O H2N N 2-O PO 3 Step 1 Activation O H H2N O 3PO OH OH HO N H Tyr-114 O Thr-189 OH Asp-187 H Figure 49. Proposed mechanism for the catalysis of H2S production from L-cysteine by human CSE. 109 The enzyme would first bind to the cofactor by forming an internal aldimine through the Lys-212 residue if it was not already saturated with PLP. As the pyridine ring of the cofactor is electron deficient, it would form efficient π-stacking interactions to the electron rich phenolic side-chain of Tyr-114. Hydrogen bonding interactions between the pyridoxal nitrogen group and the carboxylate side chain of Asp-187 would also be formed and this would in turn activate the neighboring Thr-189 residue towards proton abstraction from the amino group of the incoming L-cysteine substrate (Fig. 49, Step 1). As such, the amino group of the L-cysteine substrate would be sufficiently nucleophilic for a subsequent transaldimination reaction with Lys-212 so as to bind to PLP (Fig. 49, Step 2). L-cysteine is also held in the active site of the enzyme with the aid of ionic interactions between its carboxylate group and the guanidine side chain of Arg-375. The amino side chain of Lys-212 would then abstract the α-proton of the bound substrate (Fig. 49, Step 3), and subsequently pass this proton to the sulfhydryl group upon β-cleavage of the substrate in the next step (Fig. 49, Step 4). The influx of electrons into the pyridoxal ring in Step 3 would increase the electron density of the cofactor, hence weakening the πstacking interaction between the neutral cofactor and Tyr-114. This was hence the most likely impetus for the subsequent exit of the PLP ketimine intermediate from the enzyme as this intermediate is not covalently bound to any of the active site residues at this stage. Moreover, as the β- cleavage reaction is deemed to be a much slower process compared to the γ-lyase reaction for CSE (Table 5), it was likely that this intermediate was displaced out of the enzyme by an incoming molecule of PLP in the regeneration of the holoenzyme (Fig. 49, Step 5). The released PLP-iminopropionate intermediate would then be hydrolyzed back to free PLP, pyruvate and ammonia outside of the enzyme. 110 Till this stage, we still have yet to explain the heightened activity of the Y114F mutant enzyme in the catalysis of H2S production. Comparing the side chains of phenylalanine and tyrosine, the former is less electron rich since the para-hydroxyl group in the phenolic side-chain of tyrosine would donate electron density to the aromatic ring through resonance effects. This would imply that the π-stacking interactions between the electron-rich aromatic side-chain of the 114th amino acid and the electron-deficient pyridoxal ring of PLP would be stronger for the wild-type enzyme. Hence, in lieu of the above proposed mechanism, the α-proton of the bound L-cysteine would be less acidic, and the influx of electron density into the pyridoxal ring during abstraction of the αproton (Step 3) would be rendered more difficult for the wild-type enzyme. This is as opposed to the Y114F mutant where weaker π-stacking interactions would ease the influx of electron density into the pyridoxal ring. As such, the catalysis of H2S production for the Y114F mutant occurs more readily than that for the wild-type enzyme. It is also likely that the rate of dissociation of the PLP-imine intermediate (Step 5) is slower for the wildtype enzyme due to the presence of hydrogen-bonding interactions between the parahydroxyl group of Tyr-114 and the hydroxyl group of PLP. Hence, the turnover rate of PLP would be higher for the Y114F mutant and more H2S can be produced within a fixed period of time. A preliminary experiment on the extent to which the production of H2S was dependent on the exogenous PLP concentration for the Y114F mutant enzyme further illustrates the above discussion. As shown in Fig. 50, the amount of H2S produced was observed to increase steadily as well with an initial increase in exogenous PLP concentration. 111 However, unlike that for the wild-type enzyme which only displayed a 1.75-fold increase in H2S production from a concentration of 0 µM to 500 µM PLP concentration (Fig. 45), a much larger increase in enzyme activity of up to 2.3 times was observed for the same concentration range of PLP for the Y114F mutant enzyme. This suggests that the mutant enzyme was more sensitive towards changes in exogenous PLP concentration and that exogenous addition of PLP could greatly improve the enzyme activity by establishing a faster rate at which the end products (H2S, pyruvate and ammonia) of the reaction and the active holoenzyme were generated, which is in accordance to the mechanism proposed Amount of H2S produced (nmol) above. 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Concentration of exogenous PLP (uM) Figure 50. Amount of H2S produced in 30 min under different exogenous concentrations of PLP, assayed in the presence of 5 µg GST-tagged Y114F mutant CSE and 2.75 mM Lcysteine. Each data point was displayed as the mean of triplicate readings ± SD. 6.4. Conclusion In conclusion, not only has the above work provided greater insight into the types of sulfur-containing substrates that could be utilized by CSE for the catalysis of H2S production from L-cysteine as well as the various active site residues which play roles in 112 the production of H2S from L-cysteine, a clearer elucidation on the mechanism of H2S production had also been afforded. The information gathered would also prove useful in the development of inhibitors specific to the catalysis of H2S production. For instance, the assay on various substrates that could produce H2S has provided us with clues on the types and arrangement of various functionalities that could be recognized by the enzyme. The crucial active site residues, namely Tyr-114, Asp-187, Thr-189, Lys-212, Glu-339 and Arg- 375 which had been found to affect the production of H2S could also serve as a guideline for incorporation of various functional groups into the inhibitor to enhance its binding to the enzyme. Most importantly, the finding that the PLP-ketimine intermediate is likely displaced from the enzyme during the course of H2S production would lead us to design the inhibitor in a way which mimics the structure of this intermediate, so that this PLP-lookalike compound can subsequently bind effectively to and inhibit the enzyme. 113 7. Development of a polyclonal antibody specific towards human CSE 7.1. Objectives Thus far, we had optimized an assay for H2S production, determined the threedimensional structure of the enzyme via X-ray crystallography as well as performed an in-depth study on the mechanism of H2S production via site-directed mutagenesis and kinetic studies. As our protein had been purified to a great extent, we proceeded to develop a polyclonal antibody that would be specific towards human CSE. This would be useful since the antibody obtained can be applied in immunofluorescence or immunoprecipitation studies at a later stage to gain a better understanding of the role of the enzyme at cellular level. Although antibodies for human CSE are commercially available, it is beneficial for us to obtain our own antibody, characterize and purify it ourselves. This is due to the high cost, and more importantly, the problem with quality control for many commercial antibodies as raised by Couchman (2009). Our purified enzyme would hence be immunized into two different rabbits for raising polyclonal antibodies which would subsequently be tested for their specificity and efficiency in recognition of the human CSE target. The antibody serum from either rabbit which was determined to exhibit a higher affinity for the CSE antigen would then be purified via Protein A affinity chromatography. This purified antibody would then be utilized in preliminary experiments for probing the levels of endogenous CSE in various cell lines as well as in immunoprecipitation of endogenous CSE from cell lines that were found to possess high levels of the protein. 114 7.2. Experimental 7.2.1. Western blotting The protocol for western blotting was identical to that described in Section 3.2.3. 7.2.2. Immunoprecipitation (IP) of endogenous CSE using rabbit antibody serum 50 μL of preimmunized serum was first incubated with 10 μL of protein A agarose beads (SantaCruz) and 150 μL 1 x PBS at room temperature on an obital shaker for 30 min. The beads were then pelleted at 1000 x g for 3 min and washed thrice with 1 mL of 1x PBS supplemented with 0.05 % TritonX-100. Pre-cleared cell lysate was prepared by introducing the initial lysate to the treated agarose beads and incubating at 4 °C with rotating for 30 min. The pre-cleared lysate was then incubated with 10 µL of the respective rabbit sera as positive control or 10 µL of pre-immunized sera as negative control at 4 °C with rotating for 2 h. Each of the lysate-antibody complex was then transferred to separate tubes each containing 7.5 µL of fresh protein A agarose beads and incubated at 4 °C with rotating for another hour. The agarose beads were then pelleted at 1000 x g for 3 min and washed five times with 1 mL of ice-cold mild lysis buffer, while the supernatant (flow-through) was kept at -80° C. Bound protein was dissociated by incubating the agarose beads with 40 μL of 2 x SDS-PAGE sample loading buffer supplemented with 0.2 M DTT at room temperature for 10min before boiling for 5min. The mixture was then centrifuged at 14000 rpm for 1 min and the supernatant was kept at -20 °C until further analysis by Western blot. 115 7.2.3. Purification of anti-hCSE 1366 The antibody serum was first cleared of particulate matter by centrifuging the serum at 10000 rpm for 10 min at 4 °C and syringe filtering the supernatant through 0.22 μm. Purification was then performed by passing the cleared serum through a HiTrap Protein A column (GE Healthcare) that had been previously equilibrated with 1 x PBS via a peristaltic pump. The column was subsequently washed with 10 mL of ice-cold 1 x PBS to remove unbound proteins, and then attached to an AKTAPrime Plus machine where it was further washed with 10 mL of 1x PBS. Elution of bound proteins was performed using 15 mL of 0.1 M citric acid buffer pH 4.0 followed by 10 mL of 0.1 M glycine-HCl pH 2.7. Fractions corresponding to the antibody peak were immediately neutralized by specific volumes of 1.0 M Tris-HCl pH 9.0 as determined through a blank elution run. These fractions were then pooled and dialyzed against 1 x PBS (supplemented with 0.001 % NaN3 in the final dialysis) at 4 °C. This cycle of antibody binding and elution was repeated until a significant decrease in amount of eluted antibody was observed. The efficiency of purified antibody was then tested by probing various cell lysates and rat liver homogenates with the antibody. 7.2.4. Probing for endogenous CSE levels in various cell lysates Various cells were grown to appropriate confluency, harvested and lysed in appropriate amounts of RIPA lysis buffer. The protein concentration of each lysate was determined by Bradford protein assay and equal amounts of total protein were loaded for Western 116 blot studies. The endogenous CSE levels in these cell lysates were compared with one another and normalized against β-actin. 7.2.5. Immunoprecipitation of endogenous CSE using purified anti-hCSE antibody The procedure for immunoprecipitation where purified antibody was utilized was similar to that described above where the antibody serum was used. 50 µg of preimmunized serum was used for pre-clearing the cell lysate. For immunoprecipitation of endogenous CSE from HepG2 and 293T, 2.1 mg and 1.8 mg of total protein from the pre-cleared lysate were incubated with 15 µg of the purified anti-hCSE antibody at 4 °C for 1.5 h. The lysate-antibody complex was then allowed to bind to protein A agarose beads at 4 °C for 1 h. For immunoprecipitation of endogenous CSE in K562 and U937, 1 mg of total protein from the pre-cleared lysate was incubated with 7.5 µg of the purified anti-hCSE antibody at 4 °C for 1 h. The lysate-antibody complex was subsequently incubated with protein A agarose beads at 4 °C for 1.5 h. Bound protein was dissociated by incubating the agarose beads with 50 μL of 2 x SDS-PAGE sample loading buffer supplemented with 0.2 M DTT at room temperature for 10 min before boiling for 5 min. The mixture was then centrifuged at 14000 rpm for 1 min and the supernatant was kept at -20 °C until further analysis by Western blot. 117 7.3. Results and discussion 7.3.1. Testing of anti-hCSE sera As the human CSE antigen had been immunized into two different rabbits, it was necessary to determine which rabbit serum would have a stronger affinity for probing the CSE protein. An initial Western blot experiment showed that although both antibody sera could probe for the CSE protein, the serum that was obtained from rabbit 1366 displayed a higher affinity for the protein since stronger bands were observed for both purified human CSE (Fig. 51A) and endogenous CSE in human cell lysates or rat liver homogenates (Fig. 51B). In addition, the antibody serum from rabbit 1366 was found to be successful towards immunoprecipitation of endogenous CSE from 293T cell lysates, but not that for the serum of rabbit 1365. This was a significant finding which indicated that the antibody we had obtained was indeed specific towards the CSE protein. We hence proceeded to immunize rabbit 1366 with the CSE antigen twice more to obtain more antibody sera, as well as purify the antibody sera obtained for long term storage. A (kDa) 50 37 50 Pure CSE (ng) 500 250 125 B (kDa) 50 WB: Rabbit 1365 serum 37 50 WB: Rabbit 1366 serum 37 37 IP eluate C (kDa) CSE WB: Rabbit 1366 serum CSE (50ng) (kDa) – + – + 150 100 75 IP: Rabbit 1365 serum 50 WB: Rabbit 1365 serum IP eluate CSE (50ng) CSE WB: Rabbit 1365 serum 150 100 75 CSE 50 37 37 25 25 IP: Rabbit 1366 serum CSE WB: Rabbit 1366 serum Figure 51. Probing of different amounts of pure CSE (A) and endogenous CSE from HepG2, 293T and 5 % w/v rat liver homogenate (B) with anti-hCSE sera from either 118 rabbit 1365 or rabbit 1366. (C) Immunoprecipitation of endogenous CSE from 293T cell lysates by utilizing antibody serum of either rabbit 1365 or 1366. The figure shows the IP eluates obtained for lysates which had been incubated with pre-immunized sera (–) or the rabbit antibody sera (+).The thick bands at about 50 kDa and 25 kDa correspond to the heavy and light chains of the antibody. Membranes were blocked at 4 °C overnight (A and B) or at room temperature for 2 h (C) followed by incubation with the respective primary antibody sera (1:500 dilution). HRP-conjugated donkey anti-rabbit (1:10000 dilution, Pierce) was used as secondary antibody for A and B, while HRP-conjugated goat anti-rabbit F(ab’)2 (1:10000 dilution) was used as secondary antibody for C. Chemiluminescence was detected with ECL substrate (PerkinElmer) and subsequent exposure on X-ray film for 30 s (A and B) or 1min (C). 7.3.2. Purification of antibody serum from rabbit 1366 Purification of each batch of antibody serum was achieved via Protein A affinity chromatography as the antibody would recognize and bind to the Protein A antigen coupled on the sepharose media. The sepharose beads would then washed to remove nonspecifically bound proteins, and subsequent elution of our target bound antibody could then be achieved with the aid of an acidic buffer which served to weaken interactions between the bound antibody and the Protein A antigen. However, as the acidic buffer may be detrimental towards the conformation of the antibody, it was necessary to neutralize the eluted antibody fractions immediately with a basic buffer such as 1 M TrisHCl pH 9.0. As the pH of the eluted fractions would decrease gradually from the original neutral pH of the 1 x PBS binding buffer to the final acidic pH of the citric acid elution buffer, a blank run was performed using the same elution profile but in the absence of any antibody to determine the volume of Tris base required to neutralize fractions corresponding to those of the antibody peak in the first elution run (Fig. 52). Subsequent elution of the bound anti-hCSE antibody was hence subsequently performed by neutralizing the eluted fractions with these specific volumes of Tris base. A total of 264 119 mg of purified anti-hCSE antibody was obtained finally from three batches of antibody sera harvested on three separate days. Tube no. 9 10 11 12 13 14 15 16 17 pH of eluted fraction 8 7 5.5 4 4 4 4 4 4 Vol. of base (uL) 0 0 10 30 50 60 70 80 80 Figure 52. Chromatograph of eluted anti-hCSE from HiTrap Protein A column and volume of 1 M Tris pH 9.0 base needed to neutralize various fractions from the blank run. 7.3.3. Characterization of purified anti-hCSE antibody Having purified the antibody sera, it was necessary to determine if the purified antibody was still functional as there could be conformational changes to the antibody due to the acidic elution buffer utilized in the purification process. As observed from a Western blot on various cell lysates and purified CSE, each batch of purified antibody was found to be still successful in probing for the purified CSE protein (Fig. 53). The sensitivity of the purified antibodies towards endogenous CSE in the cell lysates or tissue homogenate was however found to increase significantly after the second immunization of the rabbit with the antigen. This was expected since repeated immunization would lead to an increase in affinity of the antibody towards the CSE antigen; a process termed affinity maturation (Janeway et al., 2001). 120 (kDa) 5% rat CSE liver HepG2 293T (50ng) 5% rat CSE HepG2 293T (50ng) (kDa) liver 5% rat CSE (kDa) liver HepG2 293T (50ng) 50 50 50 37 37 37 WB: Purified antihCSE batch 1 WB: Purified antihCSE batch 2 WB: Purified antihCSE batch 3 Figure 53. Western blot on purified CSE and endogenous CSE levels in various homogenates or lysates utilizing anti-hCSE antibody from different immunization batches. 1 mg/mL of the respective primary antibodies (1:1000 dilution) were incubated with each membrane at room temperature for 1h. HRP-conjugated donkey anti-rabbit (1:10000 dilution, Pierce) was used as secondary antibody. Chemiluminescence was detected using ECL substrate (PerkinElmer) and subsequent exposure for 5 min on X-ray film. As the antibody obtained after the third immunization (anti-hCSE batch 3) had been determined to be the most sensitive towards the CSE antigen above, subsequent studies for comparing the endogenous levels of CSE present in various human cell lines such as lung embryonic fibroblast cells (IMR90), embryonic kidney cells (HEK293), liver carcinoma cells (HepG2), breast adenocarcinoma cells (MCF7), cervical cancer cells (HeLa), colon carcinoma cells (HCT116), osteosarcoma cells (U2OS) and leukemia cell lines (HEL, REH, NB4, HL60, U937, K562, Jurkat) were conducted using this batch of purified antibody (Fig. 54). As summarized in Table 10, K562 and U937 which were two of the leukemic cell lines tested were observed to possess significantly higher CSE content than the other cell lines. This was an interesting observation since recent studies by Adhikari and Bhatia (2008) as well as Yang et al. (2006) had shown the pro-apoptotic effect of H2S on pancreatic acinar cell and human aorta smooth muscle cells. As CSE is one of the two enzymes responsible for endogenous production of H2S, there could be mechanisms in place in the leukemic K562 and U937 cells to suppress either the production of H2S or the pro-apoptotic effects of the large amounts of H2S produced as a result of the high content of CSE present in these two cell lines. There could also be 121 mechanisms in place in other cell lines which hinder the endogenous expression of CSE or cause the protein to be much more unstable, hence reflecting the much lower CSE content for these cell lines. Pure CSE (50ng) HEK293 MCF7 CSE (kDa) 50 IMR90 HepG2 HCT116 HeLa HEL U2OS REH NB4 U937 HL60 Jurkat K562 WB: Purified antihCSE antibody 37 β-actin 50 WB: Anti-β-actin 37 Figure 54. A comparison of the relative endogenous CSE levels among various cell lysates, normalized against β-actin. 25 µg of total protein was loaded into each well. 1 mg/mL of purified anti-hCSE batch 3 antibody (1:1000 dilution) was used as primary antibody for probing of CSE, while anti-β-actin (SantaCruz, 1:350 dilution) was used as primary antibody for probing of β-actin. HRP-conjugated donkey anti-rabbit (1:10000 dilution, Pierce) and HRP-conjugated anti-goat (1:10000 dilution) were used as secondary antibody for CSE and β-actin respectively. Chemiluminescence was detected using ECL substrate (15 x dilution, Amersham for CSE; PerkinElmer for β-actin) and subsequent exposure for 8 min and 3 min for CSE and β-actin respectively on X-ray film. Table 10. Relative levels of endogenous CSE in various cell lines. Cell line Relative endogenous CSE levels U937, K562 +++++ HEK293, HepG2 ++++ U2OS, HEL, Jurkat +++ HCT116, NB4, HL60 ++ MCF7, HeLa, REH + IMR90 - To further determine if the specificity of the purified antibody had been affected during the purification process, immunoprecipitation experiments were conducted on 293T, HepG2, K562 and U937 cell lysates. As observed from Fig. 55, the specificity of the antibody for CSE was still maintained and immunoprecipitation of endogenous CSE was 122 repeatable across all four cell lines due to the presence of the CSE band in the IP eluates for lysates which had been incubated with the purified antibody ( + lanes). 293T CSE IP eluate (50ng) (kDa) HepG2 CSE IP eluate (kDa) – – + 250 150 100 75 K562 IP eluate (50ng) (kDa) + 250 150 100 75 IP eluate (kDa) – + 50 50 CSE 50 37 37 37 37 25 25 25 25 IP: Purified anti-hCSE antibody CSE – CSE (50ng) + 150 100 75 150 100 75 50 U937 CSE (50ng) CSE CSE WB: Purified anti-hCSE antibody Figure 55. Immunoprecipitation of endogenous CSE from 293T, HepG2, K562 and U937 cells. The figure shows the IP eluates obtained for each of the lysates which had been incubated with pre-immunized sera (–) or the purified anti-hCSE antibody (+).The thick bands at about 50 kDa and 25 kDa correspond to the heavy and light chains of the antibody. Membranes were incubated with 1 mg/mL of the purified anti-hCSE antibody (1:1000 dilution) as primary antibody followed by HRP-conjugated goat anti-rabbit F(ab’)2 (1:10000 dilution) as secondary antibody. Chemiluminescence was detected using ECL substrate (PerkinElmer) and subsequent exposure for 10 min on X-ray film. 7.4. Conclusion The development of a polyclonal antibody which is specific towards human CSE was successfully accomplished via immunization of a rabbit with the purified CSE protein. The antibody serum was shown to be effective in probing of both human CSE (purified form or in cell lysates) and rat liver CSE (from homogenates), as well as in immunoprecipitation of endogenous CSE from 293T cells. This affinity of the anti-hCSE antibody for either rat or human CSE was maintained even after purification of the antibody sera via affinity chromatography as determined by Western blot and immunoprecipitation experiments. In particular, a preliminary study on the relative 123 endogenous levels of CSE in various human cell lines had raised several possibilities with regards to differences in CSE expression, stability and H2S production. The development of this antibody which is specific towards CSE would hence serve as a useful tool in future studies where the role of CSE is further explored at the cellular level. Possible experiments include investigation of the effect of up-regulation or downregulation of endogenous CSE expression on cell survival due to the strong correlation between H2S and apoptosis as identified in recent studies, co-immunoprecipitation experiments for identification of other proteins which could interact with endogenous CSE, or the in vitro purification of endogenous CSE via immunopurification for determination of the presence of post-translational modifications which could potentially affect the enzyme structure or activity. 124 8. Concluding remarks In this work, we had attempted to develop potent and selective inhibitors of CSE towards H2S production as well as gain a deeper understanding on the functional role of this enzyme in the catalysis of H2S production. The assay that was used for screening of inhibitor candidates was originally based on a rat liver tissue homogenate assay. This assay however had its limitations and drawbacks, and thus it was necessary for us to develop a pure protein assay. To achieve this, we had cloned and optimized procedures for expression and purification of the human CSE protein. The purified protein was subsequently shown to be able to catalyze the production of H2S from L-cysteine in an in vitro assay, hence demonstrating that human CSE is a multifunctional enzyme just like rat liver CSE. This purified protein assay had also been optimized for efficient trapping of H2S gas so as enable the determination of inhibition levels for various inhibitor candidates where an initial hit compound of IC50 value of 0.45 mM was identified. To aid in future rational design of inhibitors for the catalysis of H2S production by CSE, the structure of the protein was also elucidated via X-ray crystallography. In particular, both apo- and holo- forms of the enzyme were presented and significant differences in both structures were discussed. We believe that the loss of PLP was triggered by the inclusion of L-cysteine in the crystallization condition for CSE apoenzyme. This served as the main hypothesis in the subsequent study of the kinetics of H2S synthesis where together with data gathered from site-directed mutagenesis studies of the enzyme, a more detailed mechanism for the catalysis of H2S production was proposed. The role of the Tyr-114 residue in the inhibition of CSE by PAG was also further illustrated through the crystal structure of the CSE-PAG complex as well as an IC50 analysis on the inhibition of the 125 Y114F mutant CSE towards H2S production. Preliminary experiments for the cocrystallization of other CSE-inhibitor complexes had also been carried out in this work. Lastly, a polyclonal antibody that is specific towards CSE had also been developed. This antibody was characterized and shown to be effective towards probing of both purified and endogenous CSE, as well as in immunoprecipitation of endogenous CSE from various cell lysates. It would hence serve as a very useful tool for future functional studies of the enzyme where the role of CSE is further explored at the cellular level. 126 BIBLOGRAPHY Adhikari, S., & Bhatia, M. (2008). H2S-induced pancreatic acinar cell apoptosis is mediated via JNK and p38 MAP kinase. Journal of Cellular and Molecular Medicine, 12 (4), 1374 - 1383. Adrain, R., Amare, F. D., & Burley, S. K. (1997). Dynamic light scattering in evaluating crystallizability of macromolecules. Methods in enzymology, 276, 157-166. Alston, A. T., Muramatsu, H., Ueda, T., & Bright, H. J. (1981). Inactivation of gammacystathionase by gamma-fluorinated amino acids. FEBS Letters, 128 (2), 293-297. Bhatia, M., Sidhapuriwala, J., Moochhala, S. M., & Moore, P. K. (2005a). Hydrogen sulphide is a mediator of carrageenan-induced hindpaw oedema in the rat. British Journal of Pharmacology, 145, 141–144. Bhatia, M., Wong, F. L., Fu, D., Lau, H. Y., Moochhala, S. M., & Moore, P. K. (2005b). Role of hydrogen sulfide in acute pancreatitis and associated lung injury. The FASEB Journal, 19, 623-625. Bisswanger, H. X. (2002). Enzyme kinetics: principles and methods. Weinheim: WileyVCH. Braunstein, A. E., & Goryachenkova, E. V. (1984). The β-replacement-specific pyridoxal-P-dependent lyases. Advances in enzymology, 56, 1-89. Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse- Kunstleve, R. W., et al. (1998). Crystallography and NMR System: a new software suite for macromolecular structure determination. Acta Crystallographica Section D, 54, 905921. Chiku, T., Padovani, D., Zhu, W., Singh, S., Vitvitsky, V., & Banerjee, R. (2009). H2S biogenesis by cystathionine gamma-lyase leads to the novel sulfur metabolites, lanthionine and homolanthionine, and is responsive to the grade of hyperhomocysteinemia. Journal of Biological Chemistry, [Epub ahead of print]. Clausen, T., Huber, R., Laber, B., Pohlenz, H. D., & Messerschmidt, A. (1996). Crystal Structure of the Pyridoxal-5'-phosphate Dependent Cystathionine β-lyase from Escherichia coli at 1.83 Å. Journal of Molecular Biology , 262, 202-224 . Clausen, T., Huber, R., Prade, L., Wahl, M. C., & Messerschmidt, A. (1998). Crystal structure of Escherichia coli cystathionine γ-synthase at 1.5 Å resolution. EMBO Journal, 23, 6827-6838. 127 Collin, M., Anuar, F. B., Murch, O., Bhatia, M., Moore, P. K., & Thiemermann, C. (2005). Inhibition of endogenous hydrogen sulfide formation reduces the organ injury caused by endotoxemia. British Journal of Pharmacology, 146, 498-505. Couchman, J. R. (2009). Commercial Antibodies: The Good, Bad, and Really Ugly. Journal of Histochemistry and Cytochemistry, 57 (1), 7-8. DeLano, W. L. (2002). http://www.pymol.org . The PyMOL molecular graphics system. In. Donald, B., Smith, D. B., & Johnson, K. (1988). Single-step purification of polypeptides expressed in Escherichia coli as fusions with glutathione-S-transferase. Gene, 67, 3140. Eisenthal, R., & Danson, M. J. (2003). Enyzme assays : A Practical approach (2nd ed.). Oxford University Press. Emsley, P., & Cowtan, K. (2004). Coot: model-building tools for molecular graphics. Acta Crystallographica Section D, 60, 2126-2132. Eto, K., Asada, T., Arima, K., Makifuchi, T., & Kimura, H. (2002). Brain hydrogen sulfide is severely decreased in Alzheimer’s disease. Biochemical and Biophysical Research Communications, 293, 1485–1488. Flavin, M. (1971). Trans-sulfuration reactions: introduction. Methods in Enzymology, 17B, 416–417. Gaull, G. E., Sturman, J. A., & Raiha, N. C. (1972). Development of mammalian sulfur metabolism: absence of cystathionase in human fetal tissues. Pediatric Research, 6, 538-547. Huang, S., & Tan, C. H. (2006). The design and synthesis of a new class of selective cystathionine-γ-lyase inhibitors. Department of Chemistry, NUS, UROPS report. Huang, S., Sivaraman, J., Deng, L. W., Moore, P. K., & Tan, C. H. (2007). Structural and kinetic studies of cystathionine-γ-lyase (CSE). Department of Chemistry, NUS, Honours report. Janeway, C. A., Travers, P., Walport, M., & Shlomchik, M. J. (2001). Immunobiology: the immune system in health and disease (5th ed.). New York: Garland Publishing. Kamoun, P. (2004). Endogenous production of hydrogen sulfide in mammals. Amino Acids, 26 (3), 243-254. 128 Karlberg, T., Uppenberg, J., Arrowsmith, C., Berglund, H., Busam, R. D., Collins, R., et al. (2006). Crystal structure of human cystathionine gamma lyase. Paper to be published, PDB ID: 2NMP. Koshland, D. E., Nemethy, G., & Filmer, D. (1966). Comparison of experimental binding data and theoretical models iin proteins containing subunits. Biochemistry, 5, 365385. Labarga, A., Valentin, F., Andersson, M., & Lopez, R. (2007). Web Services at the European Bioinformatics Institute. Nucleic Acids Research, 2007, 1-6. Levonen, A. L., Lapatto, R., Saksela, M., & Raivio, K. O. ( 2000 ). Human cystathionine gamma-lyase: developmental and in vitro expression of two isoforms. Biochemical Journal, 347, 291-295. Li, L., Whiteman, M., Guan, Y. Y., Neo, K. L., Cheng, Y., Lee, S. W., et al. (2008). Characterization of a novel, water-soluble hydrogen sulfide-releasing molecule (GYY4137): New insights into the biology of hydrogen sulfide. Circulation, 117 (18), 2351-2360. Li, X., Du, J., Shi, L., Li, J., Tang, X., Qi, J., et al. (2005). Down-regulation of endogenous hydrogen sulfide pathway in pulmonary hypertension and pulmonary vascular structural remodeling induced by high pulmonary blood flow in rats. Circulation Journal, 69 (11), 1418-1424. McGovern, S. L., Caselli, E., Grigorieff, N., & Shoichet, B. K. (2002). A common mechanism underlying promiscuous inhibitors from virtual and high-throughput screening. Journal of Medicinal Chemistry, 45, 1712-1722. Messerschmidt, A., Worbs, M., Steegborn, C., Wahl, M. C., Huber, R., Laber, B., et al. (2003). Determinants of enzymatic specificity in the cys-met-metabolism PLPdependent enzymes family: crystal structure of cystathionine γ-lyase from yeast and intrafamiliar structure comparison. Biological Chemistry, 384, 373-386. Mok, Y. Y., Atan, M. S., Cheong, Y. P., Wang, Z. J., Bhatia, M., Moochhala, S., et al. (2004). Role of hydrogen sulphide in haemorrhagic shock in the rat: protective effect of inhibitors of hydrogen sulphide biosynthesis. British Journal of Pharmacology, 143, 881–889. Monod, T., Wyman, J., & Changeux, J. P. (1965). On the nature of allosteric transitions: A plausible model. Journal of Molecular Biology, 12, 88-118. Murshudov, G. N., Vagin, A. A., & Dodson, E. J. (1997). Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallographica Section D: Biological Crystallography, 53, 240-255. 129 Orth, P., Jekow, P., Alonso, J. C., & Hinrichs, W. (1999). Proteolytic cleavage of Grampositive ß recombinase is required for crystallization. Protein Engineering, 12, 371373. Otwinowski, Z., & Minor, W. (1997). Processing of X-ray diffraction data collected in oscillation mode. Methods in enzymology, 276, 307–326. Rose, W. C., & Wixom, R. L. (1955). The amino acid requirements of man. XIII The sparing effect of cystine on methionine requirement. Journal of Biological Chemistry, 216, 763-773. Rost, B., Yachdav, G., & Liu, J. (2004). The PredictProtein Server. Nucleic Acids Research, 32, W321-W326. Steegborn, C., & Clausen, T. (2000). Specific inhibition of transsulfuration enzymes. Recent Research Developments in Biochemistry, 2, 191-207. Steegborn, C., Clausen, T., Sondermann, P., Jacob, U., Worbs, M., Marinkovic, S., et al. (1999). Kinetics and inhibition of recombinant human cystathionine γ-lyase. Journal of Biological Chemistry, 272, 12675-12684. Stipanuk, M. H., & Beck, P. W. (1982). Characterization of the enzymatic capacity for cysteine desulphydration in liver and kidney of the rat. Biochemical Journal, 206, 267-277. Sturman, J. A., Gaull, G., & Raiha, N. C. (1970). Absence of Cystathionase in Human Fetal Liver: Is Cystine Essential? Science, 169, 74 - 76. Sun, Q., Collins, R., Huang, S., Holmberg-Schiavone, L., Anand, G. S., Tan, C. H., et al. (2009). Structural Basis for the inhibition mechanism of human cystathioninegamma -Lyase: An enzyme responsible for the production of H2S. Journal of Biological Chemistry, 284 (5), 3076-3085. Szabó, C. (2007). Hydrogen sulphide and its therapeutic potential. Nature Reviews Drug Discovery, 6, 917-935. Uren, J. R., Ragin, R., & Chaykovsky, M. (1978). Modulation of cysteine metabolism in mice: effects of propargylglycine and L-cysteine-degrading enzymes. Biochemical Pharmacology, 27, 2807-2814. Vina, J., Vento, M., Garcia-Sala, F., Puertes, I. R., Gasco, E., Sastre, J., et al. (1995). Lcysteine and glutathione metabolism are impaired in premature infants due to cystathionase deficiency. The American Journal of Clinical Nutrition, 61 (5), 10671069. 130 Wang, J., & Hegele, R. A. (2003). Genomic basis of cystathioninuria (MIM 219500) revealed by multiple mutations in cystathionine gamma-lyase (CTH). Human Genetics, 112 (4), 404-408. Wang, J., Huff, A. M., Spence, J. D., & Hegele, R. A. (2004). Single nucleotide polymorphism in CTH associated with variation in plasma homocysteine concentration. Clinical Genetics, 6, 483-486. Wang, R. (2002). Two's company, three's a crowd: can H2S be the third endogenous gaseous transmitter? The FASEB Journal, 16 (13), 1792-1798. Whitmore, L., & Wallace, B. A. (2004). DICHROWEB, an online server for protein secondary structure analyses from circular dichroism spectroscopic data. Nucleic Acids Research, 32, W668-673. Yang, G., Wu, L., & Wang, R. (2006). Pro-apoptotic effect of endogenous H2S on human aorta smooth muscle cells. The FASEB Journal, 20 (3), 553-555. Yang, G., Wu, L., Jiang, B., Yang, W., Qi, J., Cao, K., et al. (2008). H2S as a physiologic vasorelaxant: hypertension in mice with deletion of cystathionine gamma-lyase. Science, 322 (5901), 587 - 590. Zhu, W., Lin, A., & Banerjee, R. (2008). Kinetic properties of polymorphic variants and pathogenic mutants in human cystathionine gamma-lyase. Biochemistry, 47, 62266232. 131 Appendix 1: Forward and reverse primers used for PCR amplification of CSE Recombinant plasmid pGEX-4T-3-CSE Primer sequences (5’Æ3’) ccggaattccatgcaggaaaaagacgcctcctca (forward) cggccgctcgagctagctgtgacttccacttggagg (reverse) pcDNA3.1(+)- ccggaattcatggactacaaagaccatgacggtcaggaaaaagacgcctcctc (forward) FLAG-CSE cggccgctcgagctagctgtgacttccacttggagg (reverse) p3xFLAG-CMV- ccggaattccatgcaggaaaaagacgcctcctca (forward) 10-CSE cggggtaccctagctgtgacttccacttggagg (reverse) pET-22b(+)_CSE ccggaattccatgcaggaaaaagacgcctcctca (forward) cggccgctcgagctagctgtgacttccacttggagg (reverse) 132 Appendix 2: Mutagenic primers used for thermal cycling of mutant strands Mutant plasmid pGEX-4T-3-CSE_Y114F pGEX-4T-3-CSE_ K212A pGEX-4T-3-CSE_ E339A pGEX-4T-3-CSE_ E339Y pGEX-4T-3-CSE_Y114A pGEX-4T-3-CSE_ K212R pGEX-4T-3-CSE_ E339K pGEX-4T-3-CSE_ R375A pGEX-4T-3-CSE_S340A pGEX-4T-3-CSE_ E349A pGEX-4T-3-CSE_Y60A pGEX-4T-3-CSE_D187A pGEX-4T-3-CSE_E157A pGEX-4T-3-CSE_N161A pGEX-4T-3-CSE_S340T Mutagenic primer sequences (5’Æ3’) tgtatggatgatgtgtttggaggtacaaacaggtacttc gaagtacctgtttgtacctccaaacacatcatccataca ctatgtattctgcaacagcatacatgaatggccacagtg cactgtggccattcatgtatgctgttgcagaatacatag ctatttactctggccgcgagcttgggaggattc gaatcctcccaagctcgcggccagagtaaatag gctatttactctggcctatagcttgggaggattcgaaagc gctttcgaatcctcccaagctataggccagagtaaatagc ttgtatggatgatgtggctggaggtacaaacaggtacttc gaagtacctgtttgtacctccagccacatcatccatacaa ctatgtattctgcaacaagatacatgaatggccacagtg cactgtggccattcatgtatcttgttgcagaatacatag agctatttactctggccaagagcttgggaggattc gaatcctcccaagctcttggccagagtaaatagct gaattagtgacacactgattgcactttctgtgggcttagagg cctctaagcccacagaaagtgcaatcagtgtgtcactaattc ctatttactctggccgaggccttgggaggattcgaaag ctttcgaatcctcccaaggcctcggccagagtaaatag cgaaagccttgctgcgcttccggcaatc gattgccggaagcgcagcaaggctttcg ccagcactcgggttttgaagctagccgttctggaaa tttccagaacggctagcttcaaaacccgagtgctgg gagacattattttggtcgtggctaacacttttatgtcacc ggtgacataaaagtgttagccacgaccaaaataatgtctc gcttgtttggatcgcaacccccacaaaccc gggtttgtgggggttgcgatccaaacaagc cgaaacccccacagcccccacccagaaggt accttctgggtgggggctgtgggggtttcg tactctggccgagaccttgggaggattcg cgaatcctcccaaggtctcggccagagta 133 pGEX-4T-3-CSE_R375K pGEX-4T-3-CSE_D187E pGEX-4T-3-CSE_Y60T pGEX-4T-3-CSE_E157D pGEX-4T-3-CSE_R62A pGEX-4T-3-CSE_S209A pGEX-4T-3-CSE_T211A pGEX-4T-3-CSE_T189A pGEX-4T-3-CSE_F190A pGEX-4T-3-CSE_F190Y pGEX-4T-3-CSE_N161Q pGEX-4T-3-CSE_R62K pGEX-4T-3-CSE_T189S ggaattagtgacacactgattaaactttctgtgggcttagagg cctctaagcccacagaaagtttaatcagtgtgtcactaattcc gagacattattttggtcgtggagaacacttttatgtcacc ggtgacataaaagtgttctccacgaccaaaataatgtctc gccagcactcgggttttgaaactagccgttctggaa ttccagaacggctagtttcaaaacccgagtgctggc gcttgtttggatcgatacccccacaaacccc ggggtttgtgggggtatcgatccaaacaagc ctcgggttttgaatatagcgcttctggaaatcccactagg cctagtgggatttccagaagcgctatattcaaaacccgag ggagctgatatttctatgtatgccgcaacaaaatacatgaatggc gccattcatgtattttgttgcggcatacatagaaatatcagctcc gctgatatttctatgtattctgcagcgaaatacatgaatggccacag ctgtggccattcatgtatttcgctgcagaatacatagaaatatcagc tttggtcgtggataacgcttttatgtcaccatatttccag ctggaaatatggtgacataaaagcgttatccacgaccaaa tggtcgtggataacactgctatgtcaccatatttccagcg cgctggaaatatggtgacatagcagtgttatccacgacca ggtcgtggataacacttatatgtcaccatatttccagc gctggaaatatggtgacatataagtgttatccacgacc cgaaacccccacacagcccacccagaagg ccttctgggtgggctgtgtgggggtttcg gcactcgggttttgaatatagcaagtctggaaatcccactaggaattg caattcctagtgggatttccagacttgctatattcaaaacccgagtgc tggtcgtggataactcttttatgtcaccatatttccagcg cgctggaaatatggtgacataaaagagttatccacgacca 134 Appendix 3: Mechanism for H2S production as proposed in the Honors project. 135 [...]... reaction velocity ZnAc Zinc acetate xv 1 Introduction Cystathionine- γ -lyase (CSE, EC 4.4.1.1), an enzyme found in mammals and some fungi, is involved in the reverse transsulfuration pathway (Scheme 1) where L-methionine is converted to L-cysteine through a series of metabolic interconversions (Rose & Wixom, 1955) Specifically, the role of CSE in this reaction pathway is to convert L -cystathionine to... precipitation became eminent for inhibitor concentrations beyond 8 mM 11 A B Percentage inhibition (%) Percentage inhibition (%) 120.0 100.0 80.0 60.0 40.0 20.0 0.0 0.0 120.0 100.0 80.0 60.0 40.0 20.0 0.0 2.0 4.0 6.0 8.0 Concentration of PAG (mM) Concentration of PAG (m M) 2.0 4.0 6.0 8.0 10.0 Concentration ConcentrationofofBCA BCA(mM) (m M) IC50 at 0.2 mM PAG IC50 at 0.1 mM BCA D 80.0 Percentage inhibition... site-directed mutagenesis and kinetic studies Lastly, a polyclonal antibody which is specific towards the human CSE enzyme would be developed and characterized so as to serve as a platform for future functional studies on this protein 4 2 Tissue H2S assay for screening inhibitors of H2S production 2.1 Objectives As mentioned in the introduction, there lies a need in developing more selective and potent inhibitors... endogenous production of H2S in the liver, kidney, intestine and vascular smooth muscle cells, the in vivo production of H2S has also been attributed to cystathionine- β-synthase (CBS) and 3mercaptopyruvate sulfurtransferase for brain and heart tissues respectively (Kamoun, 2004) Studies by Yang et al (2008) had specifically shown H2S as a physiologic 2 vasorelaxant and that prononced hypertension was triggered... inhibitors of H2S production, an expression and purification system for the human CSE protein had been developed during the UROPS and Honors projects Besides developing a better assay for the production of H2S and screening of inhibitor candidates, the success in establishing a purification system for human CSE also forms an important basis for the subsequent elucidation of the three-dimensional structure of... bacterial cell lysates would be determined and an expression system which allows for an economical production of large amounts of the protein would be selected for further expression and purification of the enzyme 3.2 Experimental 3.2.1 Preparation of recombinant human CSE plasmids Polymerase chain reaction (PCR) amplification on human full length CSE cDNA (GenBank accession no BC015807) obtained from Open... at 2.5 mM L-cysteine substrate and test compound concentrations Compounds assayed at 10 mM and 2.5 mM L-cysteine substrate and test compound concentrations respectively Two of the more potent analogues that were synthesized, N-Boc-L-cysteine and N-CbzD-cysteine, besides the commercially available inhibitors PAG and BCA, were selected for determination of their inhibition profiles (Fig 1) Although there... mutant enzymes for PLP was weakened and that the enzyme activity could be restored by exogenous PLP in this study (Zhu, Lin, & Banerjee, 2008) Scheme 1 Reverse transsulfuration pathway present in mammals and fungi NH3 OOC S β γ α COO NH3 L -cystathionine L-serine L-methionine H2O Cystathionine γ -lyase (CSE) HS COO NH3 L-cysteine COO + O + NH3 α-ketobutyrate Cystathionine β-synthase (CBS) L-homocysteine... Percentage inhibition (%) C 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 8.7 4.0 8.0 12.0 16.0 20.0 24.0 Concentration (mM) Concentrationof of Boc-L-cysteine Boc-L-cysteine (mM) IC50 at 8.7 mM Boc-L-cysteine 40.0 30.0 20.0 10.0 0.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 Concentration of of Cbz-D-cysteine (mM) Concentration N-Cbz-D-cysteine IC50: Cannot be determined; Rsq: 0.983 Figure 1 Inhibition profiles and IC50 values... PLP, assayed in the presence of 7.9 µg GST-tagged CSE and 2.75 mM L-cysteine 103 Figure 46 (A) Graphs of initial reaction velocity, V against concentration of exogenous PLP determined under various concentrations of L-cysteine substrate (B) Graphs of initial reaction velocity, V against L-cysteine substrate concentration for the various concentrations of PLP that was added in the assay 105 Figure ... 5.3 Results and discussion 59  5.3.1 Optimization of protein concentration for crystallization studies 59  5.3.2 Screening and optimization of crystallizing conditions for CSE... mutagenesis and kinetic studies on human cystathionine- gamma- lyase reveal interesting insights into the mechanism of H2S production Paper in preparation vii LIST OF TABLES Table Percentage inhibition... when utilizing L-cysteine or L -cystathionine as substrate 39 Table Observations upon addition of a mixture of ZnAc and varying concentrations of NaOH, and TCA in the presence of ZYJ4291