A plasmid-encoded nicotinamidase (PncA) is essential for infectivity of Borrelia burgdorferi in a mammalian host: Plasmid-encoded nicotinamidase of B. burgdorferi

A plasmid-encoded nicotinamidase (PncA) is essential for infectivity of Borrelia burgdorferi in a mammalian host: Plasmid-encoded nicotinamidase of B. burgdorferi
of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Molecular Microbiology (2003) 48 (3), 753–764 © 2003 Blackwell Publishing Ltd Blackwell Science, LtdOxford, UKMMIMolecular Microbiology0950-382XBlackwell Publishing Ltd, 200348 3 Original Article J.E.Purser et al.Plasmid-encoded nicotinamidase of B.burgdorferi Accepted 24 December, 2002. *For correspondence at theDepartment of Pathology and Laboratory Medicine. E-mailSteven.J.Norris@uth.tmc.edu; Tel. ( + 1) 713 500 5338; Fax( + 1) 713 500 0738. A plasmid-encoded nicotinamidase (PncA) is essential for infectivity of Borrelia burgdorferi   in a mammalian host Joye E. Purser, 1  Matthew B. Lawrenz, 1,2  Melissa J. Caimano, 3,4  Jerrilyn K. Howell, 1  Justin D. Radolf 3,5  and Steven J. Norris 1,2 * 1 Graduate School of Biomedical Sciences and Department of Pathology and Laboratory Medicine, and 2 Department of Microbiology and Molecular Genetics, Medical School, University of Texas Health Science Center at Houston, MSB 2.278, 6431 Fannin St., Houston, TX 77225, USA. 3 Center for Microbial Pathogenesis, 4 Department of Pathology and 5 Department of Medicine and Department of Genetics and Developmental Biology, University of Connecticut Health Center, Farmington, CT 06030, USA. Summary Borrelia burgdorferi  , a spirochaete that causes Lymeborreliosis, contains 21 linear and circular plasmidsthought to be important for survival in mammals orticks. Our results demonstrate that the gene BBE22encoding a nicotinamidase is capable of replacing therequirement for the 25 kb linear plasmid lp25 duringmammalian infection. Transformation of B. burgdor- feri   lacking lp25 with a shuttle vector containing thelp25 gene BBE22 (pBBE22) restored infectivity inmice to a level comparable to that of wild-type Borre- lia  . This complementation also restored the growthand host adaptation of lp25 –   B. burgdorferi   in dialysismembrane chambers (DMCs) implanted in rats. A sin-gle Cys to Ala conversion at the putative active siteof BBE22 abrogated the ability of pBBE22 to re-estab-lish infectivity or growth in DMCs. Additional Salmo- nella typhimurium   complementation studies andenzymatic analysis demonstrated that the BBE22gene product has nicotinamidase activity and is mostprobably required for the biosynthesis of NAD. Theseresults indicate that some plasmid-encoded productsfulfil physiological functions required in the enzooticcycle of pathogenic Borrelia  .Introduction Lyme borreliosis is a chronic, systemic illness caused byinfection with Borrelia burgdorferi   and the closely relatedspirochaetes Borrelia garinii   and Borrelia afzelii  . In 2000,17 730 cases of Lyme disease were reported in the UnitedStates (Centers for Disease Control and Prevention,2002). Transmission occurs with the bite of an infected Ixodes   tick, resulting in a localized lesion called erythemamigrans. B. burgdorferi   then disseminate to distant sites,including the nervous system, heart and joints, where theycontribute to a constellation of manifestations includingneuroborreliosis, cardiac block and arthritis (Steere,2001). B. burgdorferi   is an obligate parasite of its mam-malian and arthropod hosts and is not naturally transmit-ted by other means. Therefore, continued existence of thebacterium is reliant upon both long-term survival within,and successful transmission between, these radically dif-ferent host environments.The genetic mechanisms involved in the infection pro-cess are poorly understood. The B. burgdorferi   B31genome was sequenced recently (Fraser et al  ., 1997;Casjens et al  ., 2000) and consists of a 911 kb linear chro-mosome and 21 linear and circular plasmids ranging insize from 9 kb to 56 kb. The consistent presence of thislarge group of plasmids in Lyme disease Borrelia   sug-gests that they play important roles in the survival andpathogenesis of these organisms (Casjens et al  ., 2000;Palmer et al  ., 2000). Sequence analysis of the plasmidsrevealed the presence of 175 multigene families (most ofwhich are unique to Borrelia   species) and relatively fewgenes with predicted functions (Fraser et al  ., 1997;Casjens et al  ., 2000). Many of the plasmid-encodedgenes undergo dramatic changes in expression levelsduring the transition between mammal and tick environ-ments. For example, the lipoproteins outer surface proteinA (OspA) and OspB are expressed at high levels in thetick midgut but not in the tick salivary glands or within themammalian host, whereas OspC is dramatically upregu-lated during tick feeding and transmission of organisms tomice (Indest et al  ., 2001a; Schwan et al. , 1995; Schwan et al. , 2000; Seshu and Skare, 2001). Temperature, pH and as yet unidentified host factors contribute to this tran-sition (Schwan et al  ., 1995; Tilly et al  ., 1997; 2001; Akins et al  ., 1998; Cassatt et al  ., 1998; Carroll et al  ., 1999; Yang  754 J. E. Purser et al.  © 2003 Blackwell Publishing Ltd, Molecular Microbiology  , 48 , 753–764 et al  ., 1999; 2000; Schwan and Piesman, 2000; 2001;Gilmore et al  ., 2001; Hefty et al  ., 2001; 2002; Indest et al  .,2001b; Revel et al  ., 2002). The sigma factors RpoS andRpoN have been shown to play a central role in the reg-ulation of OspC and decorin-binding protein A (DbpA)(Hübner et al  ., 2001). One of the linear plasmids, lp28-1,contains the VMP-like sequence ( vls  ) locus involved inantigenic variation of the surface-exposed lipoprotein,VlsE (Zhang et al  ., 1997; Eicken et al  ., 2002). The cp26genes guaA  and guaB   encode proteins homologous topurine biosynthesis enzymes, and B. burgdorferi guaA was capable of complementing an Escherichia coli   GMPsynthetase mutant (Margolis et al  ., 1994), providing evi-dence that plasmid-encoded genes can also providephysiological functions.Reduced infectivity of B. burgdorferi   has long beenassociated with repeated in vitro   passage and the result-ing loss of plasmids (Johnson et al  ., 1984; Barbour, 1988;Simpson et al  ., 1990; Norris et al  ., 1995). Recent studiesindicated that absence of lp25 is associated with a loss ofinfectivity in C3H mice and that absence of lp28-1 resultsin reduced infectivity (Xu et al  ., 1996; Purser and Norris,2000; Labandeira-Rey et al  ., 2001). The consistenthybridization of probes specific for lp25 (14/15 strains)and vls   sequences from lp28-1 (15/15 strains) with apanel of North American B. burgdorferi   isolates furthersuggests that these plasmids are needed for the survivalof Lyme disease Borrelia   (Palmer et al  ., 2000). However,the requirement for specific B. burgdorferi   plasmids orgenes at any stage of the infectious cycle has not beendemonstrated conclusively, in large part because of thedifficulties encountered with transformation and gene dis-ruption in low-passage, infectious Borrelia   (Hübner et al  .,2001; Eggers et al  ., 2002; Elias et al  ., 2002; Lawrenz et al  ., 2002).In this study, we determined that the lp25 gene BBE22encodes a protein with nicotinamidase activity and iscapable of restoring infectivity to a B. burgdorferi   clonelacking lp25. These findings demonstrate that the plas-mids of Lyme disease Borrelia   fulfil important physiologi-cal as well as pathogenic functions, and represent the firstfulfilment of the so-called molecular Koch’s postulates(Salyers and Whitt, 2001) for a Borrelia   gene. Results Lp25 provides a required physiological function  In a previous study, well-characterized B. burgdorferi   B31low-passage clones were examined for infectivity, andclones lacking lp25 were shown to be non-infectious inimmunocompetent C3H/HeN mice at a dosage of 10 5 organisms (Purser and Norris, 2000). To determine theimportance of the immune response in this phenotype, theinfectivities of B. burgdorferi   B31 clones containing allplasmids (5A4) or lacking lp25 (5A13) or lp28-1 (5A8)were compared in immunocompetent C3H/HeNHsd andC3H/Smn.ClcrHsd- scid   mice, which have a severe com-bined immunodeficiency (SCID) phenotype. The lp25-deficient clone 5A13 could not be cultured from eitherSCID or wild-type mice 2 weeks after inoculation(Table 1). B. burgdorferi   5A13 were not recovered frominfected mice of either strain even if the inoculum wasincreased to 10 6  or 10 7  organisms (data not shown). Incontrast, a B. burgdorferi   B31 clone lacking lp28-1 (5A8)has decreased infectivity in immunocompetent mice(Purser and Norris, 2000), but was cultured from all tis-sues examined in SCID mice (Table 1). The growth of B.burgdorferi   clones 5A4, 5A8 and 5A13 was also examinedin the dialysis membrane chamber (DMC) model, in whichborrelia are cultivated in chambers of dialysis tubingimplanted within the peritoneal cavities of rats for 12–14 days (Akins et al  ., 1998). These ‘host-adapted’ organ-isms are exposed to a tissue-like environment, but areprotected from antibody and cellular responses by the8000 Da dialysis membrane barrier. In DMCs, B. burg-  Table 1. Infectivity of B. burgdorferi   B31 clones lacking plasmids lp28-1 or lp25 in immunocompetent or severe combined immunodeficiency(SCID) mice. a Mouse strain B. burgdorferi B31 cloneNo. of cultures positive/totalNo. of mice positive/totalBladderHeartJointEarAll sitesC3H/HeNHsd5A4 (all plasmids)5/5 b 6/66/66/623/236/6C3H/HeNHsd5A13 (lp25 – )0/60/60/60/60/240/6C3H/HeNHsd5A8 (lp28-1 – )0/60/63/61/64/244/6C3H/Smn.ClcrHsd-scid5A4 (all plasmids)6/66/66/66/624/246/6C3H/Smn.ClcrHsd-scid5A13 (lp25 – )0/60/60/60/60/240/6C3H/Smn.ClcrHsd-scid5A8 (lp28-1 – )6/66/66/66/624/246/6 a. Groups of six immunocompetent C3H/HeN.Hsd or C3H/Smn.ClcrHsd (SCID) mice were inoculated intradermally with 10 5  of the indicated B.burgdorferi   B31 clone. Two weeks after inoculation, the indicated tissues were cultured in BSK-II medium and examined for the growth of B.burgdorferi  . b. One culture was contaminated.  Plasmid-encoded nicotinamidase of   B. burgdorferi755  © 2003 Blackwell Publishing Ltd, Molecular Microbiology  , 48 , 753–764 dorferi   clones 5A4 (containing all plasmids) and 5A8(lp28-1 – ) increased from 10 3  ml - 1  to ª 10 7  ml - 1 ; however,clone 5A13 (lp25 – ) was neither visible by darkfield micros-copy nor could it be rescued by culturing DMC fluid in BSKmedium after explantation. In some experiments, 5A13was implanted at a higher concentration (10 6  organismsml - 1 ) to permit visualization by darkfield microscopy; thesestudies revealed that organisms were still present in sim-ilar numbers after explantation, but were non-motilereflecting loss of viability. Taken together, these resultssuggested that gene(s) of lp25 fulfil a physiological rolerather than providing an immunologically related function,as is apparently the case with lp28-1. BBE22 confers infectivity to non-infectious, low-passage   B. burgdorferiA survey of the sequence of lp25 (Fraser et al  ., 1997)revealed the presence of one open reading frame (ORF),BBE22, with a predicted metabolic function. The putativeproduct of BBE22 has 33.7% identity and 45.5% similarityto PncA, a well-characterized nicotinamidase of Salmo- nella typhimurium   that catalyses the deamination of nico-tinamide to nicotinic acid. BBE22 also exhibited significanthomology to functionally defined or putative nicotinami-dases/pyrazinamidases in a wide variety of eubacteria,including E. coli   and Yersinia  , Mycobacterium  , Brucella  , Ralstonia  , Aquifex   and Pyrococcus   species. BBE22 isalso a member of pfam00857, a large family of hydrolaseenzymes including several nicotinamidase/pyrazinami-dases (Conserved Domain Database, http:// www.ncbi.nlm.nih.gov). BBE22 was therefore selected forfurther analysis. A segment of DNA from lp25 containing BBE22 wasamplified and inserted into the shuttle vector pBSV2(Stewart et al  ., 2001), resulting in the recombinant plas-mid pBBE22 (Fig. 1). The insert sequence matched thatof lp25 co-ordinates 14 571–16 628 and thus included the537 bp BBE22 ORF plus ª 1 kb of sequence 5 ¢  to the geneto include potential transcriptional and regulatory regions,as well as ª 0.5 kb of downstream sequence. BBE22 wasthe only large ORF contained within the insert; however,the 183 bp gene BBE23, encoding a putative hypotheticalprotein, was also included to ensure that adequateupstream sequence was present for expression of BBE22.The resulting construct, pBBE22, was used to transformlp25-deficient B. burgdorferi   5A13. Infectivity of 5A13/ pBBE22 and the 5A13/pBSV2 vector control was exam-ined in C3H/HeN mice. When 10 5  5A13/pBBE22 wereused as the inoculum, viable B. burgdorferi   were recov-ered from all mice in each of the four tissues examined(Experiments 1 and 2 in Table 2), demonstrating thatinfectivity had been restored at this dosage; identicalresults were obtained with the 5A4 positive control, whichcontains lp25 and hence BBE22. In contrast, the 5A13/ pBSV2 vector-only control could not be recovered fromthe 11 mice in these two experiments. [An additionalmouse from this group in Experiment 1 was culture posi-tive at all sites, but it was determined by polymerase chainreaction (PCR) analysis of all cultures that this mouse hadbeen inadvertently inoculated with 5A13/pBBE22.] Toexamine the infectivity of 5A13/pBBE22 relative to thewild-type 5A4 further, a median infectious dose (ID 50 )determination was performed (Experiment 3 in Table 2).In this experiment, the ID 50  determined for 5A13/pBBE22was 178 organisms, whereas the ID 50  for the wild-typeclone 5A4 was 18 organisms. No cultures were positivefrom any of the mice infected with up to 10 7  5A13/pBSV2(Experiment 3 and additional experiments), so the ID 50 was > 10 7  for these organisms.Histological examination of the tibiotarsal joints of ani-mals infected with 10 5  5A13/pBBE22 for 2 weeks demon-strated that pathologic changes (tendonitis and periostitis)were present in four out of six mice, compared with sixout of six mice infected with clone 5A4 (Fig. 2). The pathol-ogy scores were also lower overall (0, 0, 4, 7, 12 and 15out of 25) than those of the 5A4-inoculated mice (10, 13,13, 16, 17 and 18 out of 25), and were significantly differ-ent by the Student’s two-tailed t  -test ( P    =  0.022). No his-topathological abnormalities were observed in mice Fig. 1. Construction of pBBE22. BBE22, and ª 1 kb and 0.5 kb of sequence 5 ¢  and 3 ¢  of the gene, respectively, were amplified by PCR with primers containing Kpn  I sites. The resulting PCR product was then digested with Kpn  I and ligated at the corresponding restriction site in the shuttle plasmid pBSV2, which contains the replication locus of the B. burgdorferi   circular plasmid cp9 (Stewart et al  ., 2001). BBE23, a predicted ORF encoding a 61-amino-acid hypothetical pro-tein, was included in the construct to ensure that the regulatory region of BBE22 was not truncated. pBBE22M is identical to pBBE22 except for a Cys Æ Ala substitution at codon 120 of BBE22. The drawing of pBSV2 is from Stewart et al  . (2001); used with permission.  756 J. E. Purser et al.  © 2003 Blackwell Publishing Ltd, Molecular Microbiology  , 48 , 753–764 inoculated with 5A13 or 5A13/pBSV2 controls. Thus,transformation with a BBE22-containing shuttle vectorrestored infectivity of 5A13 to nearly wild-type levels,although the transformant had a somewhat higher ID 50 and resulted in decreased joint histopathology relative tothe wild-type 5A4 control. Borrelia burgdorferi   strain HP-J is a high-passage, non-infectious clone of B. burgdorferi   that is missing 14 plas-mids (Lawrenz et al  ., 2002), but can still be cultivated in vitro  . HP-J was transformed with pBBE22 and tested forinfectivity in SCID mice and for growth in DMCs. No organ-isms were recovered from tissues of the inoculated mice,nor from the DMCs, indicating that complementation withpBBE22 was not sufficient to restore infectivity to the HP-J strain (data not shown). Mutation of BBE22 putative active site abolishes the ability to restore infectivity  The protein sequence of BBE22 was compared with thoseof other related nicotinamidases. Protein sequence align-ments reveal highly conserved domains among all othermembers of the hydrolase family (pfam00857), and PncAproteins from E. coli  , S. typhimurium   and Mycobacterium tuberculosis   have confirmed nicotinamidase/pyrazinami-dase function (Frothingham et al  ., 1996). Based on muta-tional analysis of members of this family, the proposedmechanism of hydrolase activity and a high degree ofconservation, the cysteine located at position 120 in theBBE22 product was postulated to be a required aminoacid within the active site of BBE22. Therefore, PCRmutagenesis was used to create construct pBBE22M,which was identical to pBBE22 except for a Cys-120Alaamino acid change. B. burgdorferi   5A13 was transformedwith pBBE22M and assayed for infectivity phenotype inC3H/HeN mice. No organisms were recovered from miceinfected with 5A13/pBBE22M, whereas mice inoculatedwith controls 5A4 or 5A13/pBBE22 were consistentlyinfected (Table 3). Therefore, complementation with themutated form of BBE22 failed to restore infectivity. Table 2. Complementation of infectivity of B. burgdorferi   5A13 with plasmid pBBE22 in C3H/HeN mice. a B. burgdorferi   B31 clone and inoculum b ID 50c  (no. oforganisms)No. of cultures positive/total No. of mice positive/totalBladderHeartJointEarAll sitesExperiment 15A4 (all plasmids)6/66/66/66/624/246/65A13 (lp25 – )0/60/60/60/60/240/65A13/pBBE22 (BBE22 + )6/66/66/66/624/246/65A13/pBSV2 (BBE22–)0/50/50/50/50/200/5Experiment 25A4 (all plasmids)6/66/66/66/624/246/65A13 (lp25 – )0/60/60/60/60/240/65A13/pBBE22 (BBE22 + )6/66/66/66/624/246/65A13/pBSV2 (BBE22–)0/60/60/60/60/240/6Experiment 3 (ID 50  determination)5A41810 5 3/33/33/33/312/123/310 4 3/33/32/2 d 3/311/113/310 3 3/33/33/33/312/123/310 2 3/33/32/2 d 3/311/113/310 1 1/31/31/31/34/121/35A13/pBSV2 (–BBE22)  > 10 7e 10 5 0/30/30/30/30/120/310 4 0/30/30/30/30/120/310 3 0/30/30/30/30/120/310 2 0/30/30/30/30/120/310 1 0/30/30/30/30/120/35A13/pBBE22 ( + BBE22)17810 5 3/33/33/33/312/123/310 4 3/33/33/33/312/123/310 3 3/33/33/33/312/123/310 2 1/31/31/30/33/121/310 1 0/30/30/30/30/120/3 a. Experimental procedure as described in Table 1. b. Inoculum =  10 5  in Experiments 1 and 2. c. Calculated by the method of Reed and Muench (1938). d. One culture had bacterial contamination. e. 5A13/pBSV2 was also non-infectious at doses of 10 6  or 10 7  organisms, as determined in a separate experiment.  Plasmid-encoded nicotinamidase of   B. burgdorferi757  © 2003 Blackwell Publishing Ltd, Molecular Microbiology  , 48 , 753–764 In vivo expression of BBE22 restores growth of lp25  –  B. burgdorferi in DMCs Borrelia burgdorferi   5A4, 5A13/pBBE22, 5A13/pBBE22Mand 5A13/pBSV2 were implanted into rat peritoneal cavi-ties within DMCs as described previously (Akins et al  .,1998). Complemented B. burgdorferi   5A13/pBBE22 notonly grew to the same density as wild-type 5A4 organismswithin the chambers, but also exhibited the same host-adapted protein expression pattern when examined bySDS-PAGE and silver staining (data not shown). In con-trast, neither B. burgdorferi   control strain 5A13/pBSV2 nor5A13/pBBE22M (with a mutated form of BBE22) survivedin DMCs. Complementation of a Salmonella pncA – nadB –  mutant with Borrelia  BBE22 Borrelia burgdorferi   construct pBBE22 was transformedinto an S. typhimurium pncA – nadB  –  double mutant todetermine whether the B. burgdorferi   gene productexpression would complement the pncA –  mutational defi-ciency and allow growth of the mutant on minimal mediausing nicotinamide as a sole substrate for NAD pro-duction. The S. typhimurium pncA – nadB  –  mutant used inthis study is deficient in both pathways of NAD synthesis(see Discussion  ): one using exogenously acquired nicoti-namide as a precursor and one involving de novo   synthe-sis from aspartate (Zhu et al  ., 1991). This mutant will notgrow in M9 minimal medium unless it is supplementedwith nicotinic acid, the end-product of PncA activity. Todetermine whether BBE22 could complement the pncA mutation, the S. typhimurium pncA – nadB  –  mutant wastransformed with pBBE22, pBBE22M or pBSV2; wild-type S. typhimurium   was included as a positive control (Fig. 3).pBBE22 conferred on S. typhimurium pncA – nadB  –  theability to grow in minimal medium +  nicotinamide(Fig. 3E), whereas transformation with pBBE22M orpBSV2 did not permit growth under these conditions. Todetermine whether a promoter known to be expressedefficiently in E. coli   would improve the complementationof the S. typhimurium   mutation, BBE22 was cloned intothe E. coli   expression vector pQE30 to producepQE30:BBE22. Expression of BBE22 from the T5 pro-moter/  lac   operator of pQE30 further increased the growthrate of the S. typhimurium   mutant so that it closely resem-bled that of the wild-type strain (Fig. 3B). All strains wereable to grow in the presence of nicotinic acid (Fig. 3C andF), indicating that the conversion of nicotinamide to nico-tinic acid is the critical activity provided by BBE22. The BBE22 product has nicotinamidase activity  The Salmonella pncA – nadB  –  mutant transformed withpQE30:BBE22 was assayed for nicotinamidase activityby quantifying the release of ammonia from theconversion of nicotinamide to nicotinic acid. In thepresence of nicotinamide, only Salmonella   complementedwith pQE30:BBE22 liberated amounts of ammonia Fig. 2. Occurrence of joint inflammation in mice infected with B. burg- dorferi   clone 5A13 complemented with pBBE22. C3H/HeN mice were injected intradermally with 10 5   B. burgdorferi   5A4 (wild type), 5A13 (lp25 – ) or 5A13/pBBE22 (5A13 complemented in trans   with BBE22). Two weeks after infection, tibiotarsal joints were isolated, decalcified, sectioned, stained with haematoxylin and eosin and analysed for histopathology. Tissues from mice inoculated with (A) 5A4, (B) 5A13 and (C) 5A13/pBBE22. Pathological changes, including mixed inflam-matory infiltrates in the tendon (T), tendon sheath (TS) and tendon sheath space (Sp), were noted in 6/6 5A4-infected mice, 0/6 5A13-infected mice and 4/6 5A13/pBBE22-infected mice (see text).
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks