Identification and Molecular Cloning of a Unique Hyaluronan Synthase from Pasteurella multocida

Paul L. DeAngelis, Jing Wei, Richard R. Drake, Ann Mary Achyuthan

Journal of Biological Chemistry · 1998 · 156 citations · 24 references

DOIFull text

Open access

Concepts

Abstract

Type A Pasteurella multocida, a prevalent animal pathogen, employs a hyaluronan [HA] polysaccharide capsule to avoid host defenses. We utilized transposon insertional mutagenesis to identify the P. multocida HA synthase, the enzyme that polymerizes HA. A DNA fragment from a wild-type genomic library could direct HA production in vivo in Escherichia coli, a bacterium that normally does not produce HA. Analysis of truncated plasmids derived from the original clone indicated that an open reading frame encoding a 972-residue protein was responsible for HA polymerization. This identification was confirmed by expression cloning in E. coli; we observed HA capsule formation in vivo and detected activity in membrane preparations in vitro. The polypeptide size was verified by photoaffinity labeling of the native P. multocida HA synthase with azido-UDP sugar analogs. Overall, the P. multocida sequence is not very similar to the other known HA synthases from streptococci, PBCV-1 virus, or vertebrates. Instead, a portion of the central region of the new enzyme is more homologous to the amino termini of other bacterial glycosyltransferases that produce different capsular polysaccharides or lipopolysaccharides. In summary, we have discovered a unique HA synthase that differs in sequence and predicted topology from the other known enzymes. Type A Pasteurella multocida, a prevalent animal pathogen, employs a hyaluronan [HA] polysaccharide capsule to avoid host defenses. We utilized transposon insertional mutagenesis to identify the P. multocida HA synthase, the enzyme that polymerizes HA. A DNA fragment from a wild-type genomic library could direct HA production in vivo in Escherichia coli, a bacterium that normally does not produce HA. Analysis of truncated plasmids derived from the original clone indicated that an open reading frame encoding a 972-residue protein was responsible for HA polymerization. This identification was confirmed by expression cloning in E. coli; we observed HA capsule formation in vivo and detected activity in membrane preparations in vitro. The polypeptide size was verified by photoaffinity labeling of the native P. multocida HA synthase with azido-UDP sugar analogs. Overall, the P. multocida sequence is not very similar to the other known HA synthases from streptococci, PBCV-1 virus, or vertebrates. Instead, a portion of the central region of the new enzyme is more homologous to the amino termini of other bacterial glycosyltransferases that produce different capsular polysaccharides or lipopolysaccharides. In summary, we have discovered a unique HA synthase that differs in sequence and predicted topology from the other known enzymes. Certain bacterial pathogens, namely Gram-negative Carter type APasteurella multocida and Gram-positive Groups A and C streptococci, produce extracellular capsules composed of HA 1The abbreviations used are: HA, hyaluronan, hyaluronate or hyaluronic acid; HAS, HA synthase; Tn, transposon; GlcA, glucuronic acid; GlcNAc, N-acetylglucosamine; Glc, glucose; GalNAc, N-acetylgalactosamine; ORF, open reading frame; kb, kilobase(s); PCR, polymerase chain reaction. (1Carter G.R. Annau E. Am. J. Vet. Res. 1953; 14: 475-478PubMed Google Scholar, 2Kass E.H. Seastone C.V. J. Exp. Med. 1944; 79: 319-330Crossref PubMed Scopus (76) Google Scholar). HA, a linear polysaccharide consisting of alternating β1,4-GlcA and β1,3-GlcNAc, is also a component of many vertebrate tissues (3Laurent T.C. Fraser J.R.E. FASEB J. 1992; 6: 2397-2404Crossref PubMed Scopus (2078) Google Scholar). The HA capsules of these bacteria are virulence factors that allow the microbes to be more successful pathogens by rendering host defenses, including complement-mediated killing and phagocytosis, less effective (4Harmon B.G. Glisson J.R. Latimer K.S. Steffens W.L. Nunnally J.C. Am. J. Vet. Res. 1991; 52: 1507-1511PubMed Google Scholar, 5Wessels M.R. Moses A.E. Goldberg J.B. DiCesare T.J. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 8317-8321Crossref PubMed Scopus (276) Google Scholar). Furthermore, HA is not immunogenic because the polymer is a normal component of the host body (6Quinn A.W. Sing K.P. Proc. Soc. Exp. Biol. Med. 1957; 95: 290-294Crossref PubMed Scopus (16) Google Scholar). The capsules of other bacteria that are composed of different polysaccharides, however, are usually major targets of the immune response. The antibodies generated against capsular polymers are often responsible for clearance of microorganisms and long term immunity. Therefore, the HA capsule serves P. multocida and Groups A and C streptococci as molecular camouflage. Various HAS enzymes have been described in the last 5 years (reviewed in Ref. 7Weigel P.H. Hascall V.C. Tammi M. J. Biol. Chem. 1997; 272: 13997-14000Abstract Full Text Full Text PDF PubMed Scopus (624) Google Scholar). HasA from Streptococcus pyogenes was the first HAS to be definitively identified (8DeAngelis P.L. Papaconstantinou J. Weigel P.H. J. Biol. Chem. 1993; 268: 19181-19184Abstract Full Text PDF PubMed Google Scholar). This integral membrane protein utilizes intracellular UDP-GlcA and UDP-GlcNAc as substrates. The nascent HA chain is extruded through the membrane to form the extracellular capsule. DG42, a Xenopus protein with a previously unknown function that had similarity to HasA (9DeAngelis P.L. Yang N. Weigel P.H. Biochem. Biophys. Res. Commun. 1994; 199: 1-10Crossref PubMed Scopus (35) Google Scholar), was shown to be a HA synthase (10Meyer M.F. Kreil G. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 4543-4547Crossref PubMed Scopus (88) Google Scholar, 11DeAngelis P.L. Achyuthan A.M. J. Biol. Chem. 1996; 271: 23657-23660Abstract Full Text Full Text PDF PubMed Scopus (50) Google Scholar). Several groups identified human and murine homologs of DG42, named HAS1, HAS2, and HAS3, that directed HA synthesis (7Weigel P.H. Hascall V.C. Tammi M. J. Biol. Chem. 1997; 272: 13997-14000Abstract Full Text Full Text PDF PubMed Scopus (624) Google Scholar); these mammalian enzymes are similar at the amino acid level, but they reside on different chromosomes. Recently a viral HAS, an ORF called A98R, with 28–30% identity to the streptococcal and vertebrate enzymes was discovered (12DeAngelis P.L. Jing W. Graves M.V. Burbank D.E. van Etten J.L. Science. 1997; 278: 1800-1804Crossref PubMed Scopus (111) Google Scholar). PBCV-1 (P aramecium bursariaChlorella virus) produces authentic HA polysaccharide shortly after infection of its Chlorella-like green algae host (12DeAngelis P.L. Jing W. Graves M.V. Burbank D.E. van Etten J.L. Science. 1997; 278: 1800-1804Crossref PubMed Scopus (111) Google Scholar). A98R is the first virally encoded enzyme reported to produce a carbohydrate polymer. This finding is also the initial discovery of HA outside of animals and their bacterial pathogens. Carter type A P. multocida, the causative agent of fowl cholera, is responsible for great economic losses in the U. S. poultry industry. A capsule mutant did not thrive in the bloodstream of turkeys after intravenous injection in comparison with the encapsulated parental strain (13Snipes K.P. Ghazikhanian G.Y. Hirsh D.C. Avian Dis. 1986; 31: 254-259Crossref Scopus (23) Google Scholar). This spontaneously arising mutant strain was also 105-fold less virulent than wild type, but the nature of the genetic defect was not known (13Snipes K.P. Ghazikhanian G.Y. Hirsh D.C. Avian Dis. 1986; 31: 254-259Crossref Scopus (23) Google Scholar). Wild-type cells treated with hyaluronidase became complement-sensitive (14Hansen L.M. Hirsh D.C. Vet. Microbiol. 1989; 21: 177-184Crossref PubMed Scopus (32) Google Scholar) and were more readily phagocytosed (4Harmon B.G. Glisson J.R. Latimer K.S. Steffens W.L. Nunnally J.C. Am. J. Vet. Res. 1991; 52: 1507-1511PubMed Google Scholar) in comparison with untreated microbes. Here we report the identification of a unique HAS from P. multocida with a primary structure that does not strongly resemble the enzymes fromStreptococcus, PBCV-1 virus, or higher animals. Tn916 was used to disrupt and to tag the HA biosynthesis locus. The Tn element on a nonreplicating plasmid, pAM150 (provided by D. B. Clewell), was introduced into the wild-type encapsulated P. multocida strain P-1059 (ATCC number 15742) by electroporation (15DeAngelis P.L. Microb. Pathog. 1998; 24 (in press)Crossref PubMed Scopus (6) Google Scholar). Altered colony morphology was initially screened by visual examination with oblique lighting. Wild-type P-1059 forms large mucoid (“wet” appearance) colonies that appear iridescent (red and green coloration). Smaller, “drier” colonies lacking iridescence were chosen and streaked out. India ink staining and light microscopy were used as a secondary screen to assess the state of encapsulation. The position of the Tn elements in the mutant chromosome was mapped by Southern analysis (15DeAngelis P.L. Microb. Pathog. 1998; 24 (in press)Crossref PubMed Scopus (6) Google Scholar, 16Ausubel F.M. Brent R. Kingston R.E. Moore D.D. Seidman J.G. Smith J.A. Struhl K. Short Protocols in Molecular Biology. 3rd. Ed. John Wiley & Sons, Inc., New York1995Google Scholar). The DNA sequences at the Tn-disrupted sites from several independently selected mutants were obtained by direct dideoxy sequence analysis of tagged chromosomal DNA (15DeAngelis P.L. Microb. Pathog. 1998; 24 (in press)Crossref PubMed Scopus (6) Google Scholar). Briefly, a chimeric DNA fragment consisting of a 12-kb portion of the Tn916 element and a short region of the P. multocida DNA generated by HhaI digestion of mutant chromosomal DNA was purified by agarose gel electrophoresis (all of the wild-type P-1059HhaI genomic fragments are ≤ 7 kb). The chimeric fragment served as the template in cycle sequencing reactions using33P terminators (Thermosequenase® system, Amersham) and a Tn916 right arm terminus primer (5′-GACCTTGATAAAGTGTGATAAGTCC-3′). The sequence data were used to design PCR primers. Gel-purified PCR products were labeled with digoxigenin utilizing the High Prime system (Boehringer Mannheim). A λ library of Sau3A partially digested P-1059 DNA was made usingBamHI-cleaved λZap Express™ vector system (Stratagene). The plaque lifts were screened by hybridization (16Ausubel F.M. Brent R. Kingston R.E. Moore D.D. Seidman J.G. Smith J.A. Struhl K. Short Protocols in Molecular Biology. 3rd. Ed. John Wiley & Sons, Inc., New York1995Google Scholar) with digoxigenin-labeled PCR product. Escherichia coli XLI-Blue MRF′ was co-infected with individual purified positive λ clones and ExAssist helper phage to yield phagemids. The resulting phagemids were transfected into E. coli XLOLR cells to recover the plasmids. The plasmids were transformed into a host more suitable for HA polysaccharide production, E. coli K5 (strain Bi8337–41; provided by I. Orskov and F. Orskov). This strain produces UDP-GlcA, a required substrate for HA biosynthesis that is not found at significant levels in most laboratory strains. Additionally, K5 possesses many other genes essential for capsular polysaccharide transport in E. coli (17Roberts I.S. Mountford R. Hodge R. Jann K.B. Boulnois G. J. Bacteriol. 1988; 170: 1305-1310Crossref PubMed Google Scholar). Another host employed for expression studies was E. coli EV5 (provided by R. P. Silver; Ref. 18Vimr E.R. Aaronson W. Silver R.P. J. Bacteriol. 1989; 171: 1106-1117Crossref PubMed Google Scholar), an acapsular derivative of a K1 strain (which produces a polysialic acid capsule) that possesses all the same general capsular polysaccharide transport machinery as K5 but not high levels of UDP-Glc dehydrogenase. Cultures of the E. coli transformants with the candidate plasmids grown in completely defined medium were tested for HA polysaccharide production as described previously (19DeAngelis P.L. Weigel P.H. Diagn. Microbiol. Infect. Dis. 1994; 20: 77-80Crossref PubMed Scopus (5) Google Scholar) except that the cell pellets were extracted with 8 m urea, 0.01% SDS at 95 °C for 2 min. The HA test assay (Pharmacia Biotech Inc.) employs a specific HA-binding protein to detect HA at concentrations of ≥0.1 μg/ml (20Tengblad A. Biochem. J. 1980; 185: 101-105Crossref PubMed Scopus (134) Google Scholar). Multiple determinations of HA levels were averaged. The HA concentration in bacterial cultures was normalized for differences in cell number by measuring the A600 value and presenting the data as μg HA/ml /A600 of bacteria. One plasmid, pPm7A, with a 5.8-kb insert conferred E. coli K5 with the ability to produce HA; no HA was produced by cells with vector plasmid alone. A truncated derivative of pPm7A containing a ∼3.3-kb insert, called pPmΔ6e, could direct the biosynthesis of HA when transformed into E. coli K5. Therefore, the sequence of both strands of the pPm7A plasmid corresponding to the pPmΔ6e DNA insert was determined. A single complete 972-residue ORF, which we called PmHAS, was found. The PmHAS ORF in the pPm7A insert was amplified by 13 cycles of PCR (16Ausubel F.M. Brent R. Kingston R.E. Moore D.D. Seidman J.G. Smith J.A. Struhl K. Short Protocols in Molecular Biology. 3rd. Ed. John Wiley & Sons, Inc., New York1995Google Scholar) with Taq polymerase (Fisher) and primers corresponding to the sequence near the deduced amino and carboxyl termini (codons in capital letters: sense, 5′-gcgaattcaaaggacagaaaATGAAcACATTATCACAAG-3′, and antisense, 5′-gggaattctgcag ttaTAGAGTTATACTATTAATAATGAAC-3′; start and stop codons, respectively, in bold). Codon 2 (T → C) was altered (italic lowercase letter) to increase protein production in E. coli. The primers also contained EcoRI and PstI restriction sites (underlined letters) to facilitate cloning into the expression plasmid pKK223-3 (tac promoter; Pharmacia). The resulting recombinant construct, pPmHAS, was transformed into E. coli SURE cells (Stratagene), and this strain was used as the source of membrane preparations for in vitro HAS assays (21DeAngelis P.L. Weigel P.H. Biochem. 1994; 33: 9033-9039Crossref PubMed Scopus (97) Google Scholar). Log phase cultures (LB broth, 30 °C) were induced with 0.5 mm isopropylthiogalactoside for 3 h before harvest. The plasmid was also transformed into E. coli K5; the resulting strain was examined for the presence of capsule by light microscopy and buoyant density centrifugation (19DeAngelis P.L. Weigel P.H. Diagn. Microbiol. Infect. Dis. 1994; 20: 77-80Crossref PubMed Scopus (5) Google Scholar). The K5 bacterial cultures were not induced routinely because did not increase HA levels in or defined medium The sugar and were and purified as described A. 1994; PubMed Scopus Google and P. I. A. van I. Biochem. 1998; (in Google Scholar, preparations from P. P.L. 1996; PubMed Scopus Google Scholar) in mm were with for 30 on before with light The were with acid before gel electrophoresis analysis (16Ausubel F.M. Brent R. Kingston R.E. Moore D.D. Seidman J.G. Smith J.A. Struhl K. Short Protocols in Molecular Biology. 3rd. Ed. John Wiley & Sons, Inc., New York1995Google Scholar, A. 1994; PubMed Scopus Google Scholar). was the was a of the normal sugar was with the and Ref. A. 1994; PubMed Scopus Google Scholar) was also used as 8 transformants produced by several of mutagenesis were screened for differences in colony light microscopy with India the cells from colonies no capsule the cells from iridescent colonies to have a capsule of of the of the wild type of the acapsular named and which had Tn elements that mapped to the same genomic fragments not to wild-type colony morphology at of The Tn element in had from the original position and at new as by Southern on the other all acapsular the Tn element at the original not significant HAS activity was detected in membrane preparations from mutant HAS activity was obtained from wild-type P-1059 cells that the Tn elements in mutants and were responsible for the HA biosynthesis locus. the the Tn sites of acapsular PCR the mutant chromosomal DNA template was with a primer derived from sequence at the mutant and a primer corresponding to the terminus of A specific PCR was on the other no was of or the Tn916 right arm primer were The PCR was used as a hybridization to a of the P. multocida were found after and these phage were into plasmids. One plasmid, pPm7A, was found that could direct E. coli K5 to produce HA in vivo μg of E. coli K5 with plasmids did not produce HA μg E. coli XLOLR coli EV5 cells (which UDP-Glc not produce HA they the pPm7A plasmid μg This genetic that the insert of pPm7A does not a UDP-Glc A truncated derivative of the pPm7A plasmid with the insert of HA biosynthesis μg of K5 pPmΔ6e, contained a single complete ORF encoding a protein plasmids with of this ORF the ability to direct HA polymer production in E. coli K5. We that this protein was the HA synthase and called The predicted and the is We used PmHAS as the in W. W. J. Biol. PubMed Scopus Google Scholar) of the protein sequence data The central portion of PmHAS is most homologous to bacterial glycosyltransferases from a of including and that form or the carbohydrate of 2 The most sequence are the and was no significant similarity of PmHAS to the or vertebrate short region of streptococcal HasA with PmHAS in a 2 A of the first of PmHAS are also similar to of the mammalian an enzyme that of 2 The ORF of PmHAS is very similar to the amino terminus of several UDP-Glc from bacteria including E. coli, and Streptococcus The in the original pPm7A clone be to in complete of The other ORF of PmHAS is very homologous to the E. coli K5 protein a protein in the transport of capsular polysaccharide of the The predicted size of for PmHAS was confirmed by photoaffinity labeling of membrane preparations from P. multocida and into a protein in an with the corresponding sugar the of In an of the normal HA did not this protein not Furthermore, derived from Tn mutants had no or very of into this protein 3 derived from E. coli SURE cells containing the plasmid, from cells with the vector pKK223-3 HA in vitro when with both UDP-GlcA and UDP-GlcNAc of was observed UDP-GlcNAc was or were with The HAS activity derived from recombinant HAS was similar to the enzyme obtained from wild-type P. multocida because at more activity than P.L. 1996; PubMed Scopus Google Scholar). Cultures of recombinant E. coli were also tested for the presence of HA polysaccharide with a assay utilizing labeled HA-binding E. coli K5 with produced μg K5 cells with pKK223-3 vector did not produce HA μg wild-type P. multocida P-1059 grown in the same produced μg E. coli K5 with produced high levels of HA that the cells became encapsulated The of the capsule of the recombinant strain was a bacterial cell of 0.5 This capsule could be by with hyaluronidase HA the native K5 host strain transformants containing pKK223-3 vector a readily capsule as by light K5 cells with were also encapsulated by buoyant density The recombinant cells on of the the vector cells or recombinant cells through the We found that the PmHAS protein polymerizes authentic HA The carbohydrate from recombinant E. by the HA-binding protein and was to HA these are as specific for HA polysaccharide (20Tengblad A. Biochem. J. 1980; 185: 101-105Crossref PubMed Scopus (134) Google Scholar, Biophys. PubMed Scopus Google Scholar). UDP-GlcA and UDP-GlcNAc were required for HA synthesis in vitro. Furthermore, and but not into in the of streptococcal HasA (21DeAngelis P.L. Weigel P.H. Biochem. 1994; 33: 9033-9039Crossref PubMed Scopus (97) Google Scholar) and Xenopus P.L. Achyuthan A.M. J. Biol. Chem. 1996; 271: 23657-23660Abstract Full Text Full Text PDF PubMed Scopus (50) Google Scholar), that polypeptide PmHAS, sugar groups to the nascent HA encapsulated Gram-negative including E. and of genes responsible for capsule biosynthesis in (17Roberts I.S. Mountford R. Hodge R. Jann K.B. Boulnois G. J. Bacteriol. 1988; 170: 1305-1310Crossref PubMed Google Scholar, 18Vimr E.R. Aaronson W. Silver R.P. J. Bacteriol. 1989; 171: 1106-1117Crossref PubMed Google Scholar, B. E.R. Microbiol. PubMed Scopus Google Scholar). often genes encoding enzymes required for sugar glycosyltransferases for the and in polysaccharide we have found that the Type A P. capsule a a HA synthase, and a UDP-Glc dehydrogenase. The Tn916 elements in the P. multocida acapsular mutants and were not in the HAS but were in the The resulting a resulting in no expression of the genes encoding the HAS or the UDP-Glc dehydrogenase. sites these but were no the S. pyogenes HAS is by an UDP-Glc and an UDP-Glc in this however, the element is at the of van I. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar). the amino acid level, PmHAS is not as similar to the other as initially short at and at of PmHAS, are in Another similar is found at of The of the first and the are in all the other several and found in all previously are from PmHAS (7Weigel P.H. Hascall V.C. Tammi M. J. Biol. Chem. 1997; 272: 13997-14000Abstract Full Text Full Text PDF PubMed Scopus (624) Google Scholar, P.L. Jing W. Graves M.V. Burbank D.E. van Etten J.L. Science. 1997; 278: 1800-1804Crossref PubMed Scopus (111) Google Scholar). Instead, a of bacterial glycosyltransferases more with the sequence in the central portion of the P. multocida HAS which have been shown or predicted to GlcNAc, or are the size of the P. multocida is to that their amino termini sequences with the of the PmHAS polypeptide of the first of PmHAS similarity to of the mammalian be a of a structure The last of this new HAS are not very similar to other in the sequence data Therefore, the P. multocida HAS be unique or be the of a new of PmHAS is the size of the or vertebrate Ref. 7Weigel P.H. Hascall V.C. Tammi M. J. Biol. Chem. 1997; 272: 13997-14000Abstract Full Text Full Text PDF PubMed Scopus (624) Google Scholar). Furthermore, the of PmHAS and the other known are the K. W. Biol. Chem. 1993; Scholar), PmHAS is predicted to have candidate on and and both termini of the protein be in the this that of the P. multocida polypeptide is outside of the the other a different topology is predicted for the other of (7Weigel P.H. Hascall V.C. Tammi M. J. Biol. Chem. 1997; 272: 13997-14000Abstract Full Text Full Text PDF PubMed Scopus (624) Google Scholar, P.L. Yang N. Weigel P.H. Biochem. Biophys. Res. Commun. 1994; 199: 1-10Crossref PubMed Scopus (35) Google Scholar). enzyme analysis of streptococcal HasA with these the to of near the amino a by at the carboxyl of the protein P.L. Weigel P.H. 1997; Scholar). The are short the of the polypeptide chain is In the be to these of the same or different to a with the and We J. E. and as as and of for for of the microscopy and Weigel for on the

References

24