Vitamin C Inhibits the Enzymatic Activity ofStreptococcus pneumoniae Hyaluronate Lyase

Abstract

Enzyme activity measurement showed thatl-ascorbic acid (vitamin C (Vc)) competitively inhibits the hyaluronan degradation by Streptococcus pneumoniae hyaluronate lyase. The complex crystal structure of this enzyme with Vc was determined at 2.0 Å resolution. One Vc molecule was found to bind to the active site of the enzyme. The Vc carboxyl group provides the negative charges that lead the molecule into the highly positively charged cleft of the enzyme. The Vc ring system forms hydrophobic interactions with the side chain of Trp-292, which is one of the aromatic patch residues of this enzyme responsible for the selection of the cleavage sites on the substrate chain. The binding of Vc inhibits the substrate binding at hyaluronan 1, 2, and 3 (HA1, HA2, and HA3) catalytic positions. The high concentration of Vc in human tissues probably provides a low level of natural resistance to the pneumococcal invasion. This is the first time that Vc the direct inhibition on the bacterial “spreading factor” was reported, and Vc is also the first chemical that has been shown experimentally to have an inhibitory effect on bacterial hyaluronate lyase. These studies also highlight the possible structural requirement for the design of a stronger inhibitor of bacterial hyaluronate lyase.1f9g Enzyme activity measurement showed thatl-ascorbic acid (vitamin C (Vc)) competitively inhibits the hyaluronan degradation by Streptococcus pneumoniae hyaluronate lyase. The complex crystal structure of this enzyme with Vc was determined at 2.0 Å resolution. One Vc molecule was found to bind to the active site of the enzyme. The Vc carboxyl group provides the negative charges that lead the molecule into the highly positively charged cleft of the enzyme. The Vc ring system forms hydrophobic interactions with the side chain of Trp-292, which is one of the aromatic patch residues of this enzyme responsible for the selection of the cleavage sites on the substrate chain. The binding of Vc inhibits the substrate binding at hyaluronan 1, 2, and 3 (HA1, HA2, and HA3) catalytic positions. The high concentration of Vc in human tissues probably provides a low level of natural resistance to the pneumococcal invasion. This is the first time that Vc the direct inhibition on the bacterial “spreading factor” was reported, and Vc is also the first chemical that has been shown experimentally to have an inhibitory effect on bacterial hyaluronate lyase. These studies also highlight the possible structural requirement for the design of a stronger inhibitor of bacterial hyaluronate lyase.1f9g l-ascorbic acid known as vitamin C hyaluronan S. pneumoniae hyaluronate lyase l-Ascorbic acid, also known as vitamin C (Vc),1 is synthesized in plants and in almost all animals except primates, guinea pig, Indian fruit bat, and some insects. Vc is necessary in the diet of these animals and usually exists in significantly large concentrations in their tissues (1Khurana S. Powers D.B. Anderson S. Blaber M. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 6768-6773Crossref PubMed Scopus (51) Google Scholar). Prolonged lack of Vc in the diet of humans results in scurvy, which is characterized by skin lesions, blood vessel fragility, and poor wound healing. Less severe deficiency of Vc produces alterations in connective tissue structure and may also cause decreased resistance to some infections (2Levine K. N. Engl. J. Med. 1986; 314: 892-902Crossref PubMed Scopus (502) Google Scholar). Vc is a multifunctional molecule in tissues. It usually acts as an antioxidant (3Hwang J. Peterson H. Hodis H.N. Choi B. Sevanian A. Atherosclerosis. 2000; 150: 275-284Abstract Full Text Full Text PDF PubMed Scopus (31) Google Scholar), free radical scavenger (4Kalka K. Mukhtar H. Turowski-Warke A. Merk H. Skin Pharmacol. Appl. Skin Physiol. 2000; 13: 143-149Crossref PubMed Scopus (41) Google Scholar), neuroprotectant, and neuromodulator (5Rice M.E. Trends Neurosci. 2000; 23: 209-216Abstract Full Text Full Text PDF PubMed Scopus (696) Google Scholar). It also plays an important physiological function in activating peptide hormones (2Levine K. N. Engl. J. Med. 1986; 314: 892-902Crossref PubMed Scopus (502) Google Scholar) and regulating cell division and growth (6Smirnoff N. Curr. Opin. Plant Biol. 2000; 3: 229-235Crossref PubMed Google Scholar). Vc is the single synthetic chemical that is manufactured and consumed in the greatest quantity in the world (1Khurana S. Powers D.B. Anderson S. Blaber M. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 6768-6773Crossref PubMed Scopus (51) Google Scholar). Although the importance of Vc in the normal function of animal tissues has long been known, the detailed molecular basis of Vc action, especially the mechanisms of its interactions with proteins and enzymes, is still largely unknown. Protein-ligand interaction is an important aspect of modern biochemistry. It provides information for the understanding of the essence of the molecular interactions, enzyme action mechanism, protein activity control, and drug design. In regards to the important function of Vc in so many life processes, the protein-Vc interfaces have not been fully characterized to date. The structural basis of the protein-Vc interface and the possible influence of Vc on enzyme activities are certainly issues of significant interest for the understanding of the functions and principles of Vc action. The protein-Vc interaction was first seen in the crystal structure ofd-xylose isomerase (Protein Data Bank accession number1xid) in which Vc was present between two tryptophan residues (7Carrell H.L. Glusker J.P. Burger V. Manfre F. Tritsch D. Proc. Natl. Acad. Sci. U. S. A. 1998; 86: 4440-4444Crossref Scopus (197) Google Scholar). It clearly emphasizes the importance of hydrophobic interactions in the protein-Vc interface. Here we present the crystal structure ofStreptococcus pneumoniae hyaluronate lyase (SpnHL) co-crystallized with Vc, which provides another opportunity to examine the protein-Vc interactions at the atomic level. S. pneumoniae is a Gram-positive bacterial pathogen that causes pneumonia, bacteremia, meningitis, sinusitis, and otitis media in humans worldwide, especially in neonates and children, and often leads to significant rates of mortality and morbidity. It secretes hyaluronate lyase to catalyze the degradation of hyaluronan (HA), one of the main components of connective tissues in animals, to expose tissue cells to bacterial toxins. Therefore, SpnHL is also being called the “spreading factor” (8Li S. Kelly S.J. Lamani E. Ferraroni M. Jedrzejas M.J. EMBO J. 2000; 19: 1128-1140Google Scholar, 9Jedrzejas M.J. Crit. Rev. Biochem. Mol. Biol. 2000; 35: 221-251Crossref PubMed Scopus (72) Google Scholar). The inhibition of hyaluronate lyase is expected to reduce the spreading of this pathogen in the most early stages of pneumococcal invasion. The action mechanism of hyaluronan degradation by SpnHL, which was revealed recently, provides a unique opportunity to target hyaluronate lyase in the prevention of the pneumococcal invasion. The Vc-complexed crystal structure of SpnHL is an attempt on this line of research. The enzymatic activity of SpnHL in the presence of various concentrations of Vc was measured. The crystal structure of this enzyme co-crystallized with Vc was determined at 2.0 Å resolution. The structural basis of the inhibitory effect of Vc on SpnHL enzymatic activity was established. l-Ascorbic acid (Sigma) was co-crystallized with the S. pneumoniaehyaluronate lyase at conditions similar to the native SpnHL crystallization condition with an additional 10–100 mmascorbic acid. A set of data at 2.0 Å resolution was collected at 100 K using one crystal at 50 mm ascorbic acid, synchrotron radiation, and a cryocooling technique at conditions similar to the native crystal (10Jedrzejas M.J. Mewbourne R.B. Chantalat L. McPherson D.T. Protein Expression Purif. 1998; 13: 83-89Crossref PubMed Scopus (47) Google Scholar, 11Jedrzejas M.J. Chantalat L. Mewbourne R.B. J. Struct. Biol. 1998; 121: 73-75Crossref PubMed Scopus (26) Google Scholar). Diffraction data were processed and scaled using the HKL program package (12Otwinowski Z. Minor W. Methods Enzymol. 1997; 276: 307-326Crossref PubMed Scopus (38446) Google Scholar). The crystal belongs to the orthorhombic space group P212121with cell parameters of a = 84.264 Å,b = 102.666 Å, and c = 103.253 Å. The data set is 90.1% complete (60% in the last shell) with Rsym = 0.087. Coordinates and B-factors of the protein part of the native SpnHL crystal structure (8Li S. Kelly S.J. Lamani E. Ferraroni M. Jedrzejas M.J. EMBO J. 2000; 19: 1128-1140Google Scholar) were used directly as a primary model in the SpnHL·Vc complex structure refinement. All waters and solution molecules were omitted from the model. The X-plor package (13Brunger A.T. Rice L.M. Methods Enzymol. 1997; 277B: 243-269Crossref Scopus (114) Google Scholar) was used to refine the structure against 54,217 reflections at a 2.0–45 Å resolution range (87.8% completeness), and 1% reflections were used inRfree calculation to monitor the refinement progress and the model improvement (14Brunger A.T. Methods Enzymol. 1997; 272B: 366-396Crossref Scopus (272) Google Scholar). Rigid body, position, and simulated annealing (3000 K) protocols were employed. The model was manually fitted into the electron density maps on graphics using program O (15Jones T.A. Zhou J.Y. Cowan S.W. Kjeldgaard M. Acta Crystallogr. Sect. A Foundations Crystallogr. 1991; 47: 110-119Crossref PubMed Scopus (13006) Google Scholar) in between each round of refinement calculations. The electron densities for Vc were observed from the beginning, but the Vc structure was only included into the structural refinement until theR factor dropped below 25%, and water molecules were incorporated thereafter. Only after this point, the B-factor refinement was introduced. The SpnHL enzymatic activity was measured using a modification of previously described protocols using either cetylpyridinium chloride (16Li M.W. Yudin A.I. VandeVoot C.A. Sabeur K. Primakoff P. Overstreet J.W. Biol. Reprod. 1997; 56: 1383-1389Crossref PubMed Scopus (55) Google Scholar) or cetrimide (17Berry A.M. Lock R.A. Thomas S.M. Rajan D.P. Hansman D. Paton J.C. Infect. Immun. 1994; 62: 1101-1108Crossref PubMed Google Scholar). The compounds assayed for inhibitory effects on the SpnHL enzyme activity were ascorbic acid, epinephrine, apigenin, salicylic acid, and histamine (Sigma). Volumes of 90 μl of inhibitor solutions at concentrations of 1, 2, 4, 6, 8, 10, 12, 14, 16, 18, and 20 mm containing 50 mm sodium acetate and 10 mm calcium chloride at pH 6.0 (apigenin was dissolved in 10% Me2SO water solution) were added along one row of a 96-well microtiter plate, leaving the first row as a blank control. 10 μl of the wild-type SpnHL at a concentration of 3.3 μg/ml was added to each well and incubated at room temperature for 1 h. Reactions using 40 units of bovine hyaluronidase (Sigma) at the same conditions were run in parallel. 25 μl of 1 mg/ml HA were added to each well to start the reaction, which was performed at room temperature for 15 min. Undigested HA was precipitated using freshly made 50 μl of 10% (w/v) aqueous cetylpyridinium chloride with an additional 2.0% NaOH to stop the enzyme activity. The absorbance was measured at 630 nm for cetylpyridinium chloride using an automated microplate reader (EL808, Bio-Tek Instruments, Inc., Winooski, VT). A higher absorbance reading corresponds to a higher remaining substrate HA concentration and, therefore, lowers enzyme activity. TheA595 absorbance readings were then converted to percent inhibition by subtracting theA595 value for the pure enzyme without inhibitor from the reading with inhibitor divided by the difference between the 100% inhibition (average values at 12.8, 14.4, and 16.0 mmascorbic acid) and 0% inhibition (A595 reading for the pure enzyme). To determine whether this inhibition was reversible, a sample of SpnHL enzyme activity was measured and designated as 100% activity. After dialyzing the sample against 20 mm Vc in buffer, the enzyme activity dropped to 0%. The sample was then dialyzed against the reaction buffer to remove the Vc, and 94% of the enzyme activity was recovered. Therefore, this inhibition was reversible. The slight drop in activity was caused by the volume changes during dialysis. The reaction initial velocity was measured in quadruplicate using a modified microplate assay (16Li M.W. Yudin A.I. VandeVoot C.A. Sabeur K. Primakoff P. Overstreet J.W. Biol. Reprod. 1997; 56: 1383-1389Crossref PubMed Scopus (55) Google Scholar). Initial hyaluronan concentrations were 0.2, and mg/ml in 50 mm acetate buffer, pH 10 mm calcium and at 100 μl of each of the concentrations of the were added to the microplate in quadruplicate and at room temperature for an to of SpnHL with 1, and 10 in 10 mm calcium chloride and 50 mm acetate buffer, pH were at for 1 and 100 μl of each of the enzyme were added to the first of each of the buffer was added to the remaining of the microplate as control. The were then incubated at for h. The enzyme were and each well was with the well was then with 100 μl of aqueous cetrimide and incubated at for h. The absorbance was measured at nm using an automated microplate reader and of SpnHL·Vc complex have been in the Protein Data Bank with accession The crystal structure of S. pneumoniae hyaluronate lyase in complex with Vc was determined at 2.0 Å resolution. In of residues were and waters were One Vc molecule was clearly seen in the electron density The factor was and = The protein part of the complex structure an and a by a The active site of this enzyme in the of the molecule in which a cleft was between these two structural The cleft was 10 Å in to units of the hyaluronan substrate chain which were HA2, and from the to the of the hyaluronan chain. and M. J. for The active site is at one of the to the of the hyaluronan and is of two an aromatic patch responsible for the cleavage site selection on the substrate chain and a catalytic group responsible for the cleavage of the between and units in the hyaluronan chain (8Li S. Kelly S.J. Lamani E. Ferraroni M. Jedrzejas M.J. EMBO J. 2000; 19: 1128-1140Google Scholar, K. Jedrzejas M.J. J. Mol. Biol. 2000; PubMed Scopus Google Vc binding in the SpnHL The of the and are One Vc molecule is to the cleft between two structural The protein part of the SpnHL·Vc complex structure has only slight changes with the native SpnHL crystal The is only Å for all protein In the cleft all in the protein-Vc complex structure are at the leaving the cleft Å is in the native SpnHL structure (8Li S. Kelly S.J. Lamani E. Ferraroni M. Jedrzejas M.J. EMBO J. 2000; 19: 1128-1140Google Scholar). The of Vc to the active site residues is shown in The aromatic patch of the enzyme active is of aromatic Trp-292, and (8Li S. Kelly S.J. Lamani E. Ferraroni M. Jedrzejas M.J. EMBO J. 2000; 19: 1128-1140Google Scholar) The side of and hydrophobic interactions with the hydrophobic on the hyaluronan chain. this the cleavage sites are with a and into catalytic In the hyaluronan and degradation the enzyme the degradation of the between and and Vc in the SpnHL·Vc complex structure was found to bind to group and the The Vc forms 25 interactions with residues of the enzyme These residues are also shown to with the substrate (8Li S. Kelly S.J. Lamani E. Ferraroni M. Jedrzejas M.J. EMBO J. 2000; 19: 1128-1140Google Scholar). The of these residues to the Vc are shown in of the interface Trp-292, and have been in and structural studies and have been shown to important in the normal function of this enzyme (8Li S. Kelly S.J. Lamani E. Ferraroni M. Jedrzejas M.J. EMBO J. 2000; 19: 1128-1140Google K. Jedrzejas M.J. J. Mol. Biol. 2000; PubMed Scopus Google Scholar). J. K. B. S. and M. J. to for Vc of Å were lyase was with vitamin to for Vc Å were in a lyase was with vitamin forms interactions with Vc, which for the most interactions all these residues in the protein-Vc interface. The group of is in to the ring of Vc 3 This structural provides the main hydrophobic interaction that the Vc molecule the is one of the aromatic patch residues and is responsible for the selection of the cleavage sites on the substrate chain. Therefore, the binding of Vc probably with the binding of hyaluronan substrate at the position, which is at the of the is one of the catalytic residues in the hyaluronan It one to the to the between and units (8Li S. Kelly S.J. Lamani E. Ferraroni M. Jedrzejas M.J. EMBO J. 2000; 19: 1128-1140Google Scholar). In the complex forms one with the Vc to for Vc Therefore, the aromatic patch and the catalytic group of the active of SpnHL are in the binding of The binding of Vc the aromatic patch and the catalytic and the part the is the only from the that is in the and the SpnHL·Vc interface. The causes a in the enzyme activity the cleft for substrate (8Li S. Kelly S.J. Lamani E. Ferraroni M. Jedrzejas M.J. EMBO J. 2000; 19: 1128-1140Google SpnHL·Vc complex structure showed that Vc is also in with and The cleft is highly charged positively by the of and residues the are residues present in the of are in the interaction with and are in Vc binding forms one with the Vc carboxyl the substrate is into the with and units and with the of the which is to the to The the enzyme activity by (8Li S. Kelly S.J. Lamani E. Ferraroni M. Jedrzejas M.J. EMBO J. 2000; 19: 1128-1140Google Scholar). In we that plays an important in the substrate after the initial is M.J. Crit. Rev. Biochem. Mol. Biol. 2000; 35: 221-251Crossref PubMed Scopus (72) Google The in with is the to The interaction with that the Vc binding also inhibits substrate binding at It was that one is to in two Vc binding sites in the complex structure 2000; PubMed Scopus (47) Google Scholar). The presence of one or residues is probably one of the in the protein-Vc interface. The 25 interactions in the SpnHL·Vc interface into two hydrophobic and to the hydrophobic interactions with and and with the In with the protein-Vc interface observed in isomerase in which hydrophobic interactions the hydrophobic and interactions are almost important in the SpnHL·Vc interface. Therefore, the SpnHL·Vc interface a of protein-Vc interface. The protein-Vc interactions cause structural changes at of the interface. In with the native SpnHL the side chain of these interface residues is the Vc has significant structural changes with its native crystal structure J. Acta Crystallogr. Sect. Struct. Sci. PubMed Scopus Google Scholar) used for and refinement. The carboxyl group is to the of the Vc ring The negative charges caused by the carboxyl group at physiological condition are important in the Vc molecule into the highly positively charged The binding that the ring of the Vc molecule provides the most hydrophobic binding the carboxyl group with the which is not in of the Vc and have been to have inhibitory effects on the enzyme activity of 1998; PubMed Scopus Google Scholar), which are a group of by for the hyaluronan usually hyaluronate to The for the of bacterial hyaluronate was from these activity clearly showed that Vc the hyaluronan degradation by Vc is similar to one of the units of the main substrate of is of of the One of the main components of acid, is also the in the Vc Therefore, Vc as a substrate of hyaluronate lyase. The effects of these compounds on the enzyme activity of SpnHL and bovine hyaluronidase were and measured by the microplate enzyme activity The results showed that of these compounds significant influence on the activity of bovine hyaluronidase at condition not Vc the SpnHL activity the of this inhibition was The initial velocity of the degradation at various concentrations of hyaluronan was measured in the presence of concentrations of Vc and in its The results were fitted to the model and the model by The with the model in an high value for the which is not present in the model. this that the concentration of Vc is for the binding to the is from this that Vc only to the free enzyme = not to the complex in the described in this Therefore, the substrate for the binding of Vc, and this inhibition is The degradation of hyaluronan in the connective tissues is an important in the pneumococcal invasion. The bacterial that hyaluronate lyase were shown to hyaluronate lyase K. V. B. J. PubMed Google Scholar). S. pneumoniae with hyaluronate lyase and cell showed significant in A.M. Paton J.C. Infect. Immun. 2000; PubMed Scopus (197) Google Scholar). Therefore, the inhibition of hyaluronate lyase activity is probably important in the of the pneumococcal invasion. animals usually to pneumococcal hyaluronate lyase a target for a Vc is the first chemical shown to have inhibition effect on the activity of this pneumococcal enzyme. The human requirement for Vc is Vc exists in human tissues at a level of to and has an with P. J. K. M. J. Biol. Full Text PDF PubMed Google Scholar, J.W. Sci. 2000; Google Scholar). concentrations of Vc were in the and the aqueous of the of of tissue J.W. Sci. 2000; Google Scholar), which corresponds to mm In the human Vc concentrations as high as mm were vitamin was at physiological concentrations P. M. J. Biol. Full Text PDF PubMed Google Scholar). Therefore, the inhibitory effect of Vc on the SpnHL enzymatic activity may have a physiological It has long been known that the deficiency of Vc may cause decreased resistance to some bacterial One to this is that animal as to using Vc to and the P. M. J. Biol. Full Text PDF PubMed Google Scholar). The inhibitory effect of Vc on SpnHL activity provides an additional possible to the Vc function as an The large concentration of Vc in human tissues the tissue to the pneumococcal a low level of natural resistance to bacterial invasion. The infections and caused by are significantly Therefore, Vc is probably a natural of the system against the pneumococcal in the tissues. usually in tissues with low concentrations of The normal Vc concentration in is mm and 10 higher in the and P. J. C in and Inc., Scholar), which is 10 the Vc concentrations in human tissues. The SpnHL·Vc complex structure also provides some for the design of a hyaluronate lyase on the interface expected that a stronger hyaluronate lyase inhibitor have a ring system to the hydrophobic binding to the one negative a carboxyl for is in the inhibitor structure to the negative charges to lead the inhibitor into the cleft In a stronger inhibitor expected to have an of hydrophobic interactions and to have charged as carboxyl The molecular of Vc are to its structural The free radical scavenger and antioxidant of Vc are directly to the active chemical of this studies the of these structural of Vc, a to This structural on Vc the of the main of connective from being by bacterial hyaluronate of or hyaluronate may by the of the large of Vc in tissues. The Vc structural to its interaction with proteins revealed from the SpnHL·Vc complex and the importance of hydrophobic and in the protein-Vc interface lead to the of the structure and function of Vc may or bacterial by directly the bacterial spreading as hyaluronate binding to the enzyme active site and with the binding of the hyaluronan All protein interface residues with Vc are all known bacterial hyaluronate (8Li S. Kelly S.J. Lamani E. Ferraroni M. Jedrzejas M.J. EMBO J. 2000; 19: 1128-1140Google Scholar). The studies on the SpnHL·Vc interface are significantly to all these bacterial hyaluronate hyaluronate lyase crystal structure was determined M.J. Chantalat L. Acta Crystallogr. Sect. Biol. Crystallogr. 2000; 56: PubMed Scopus Google active and is the same as is in Therefore, the results shown may to Streptococcus hyaluronate which that Vc also the with the to the S. to a