A Homologue of the Recombination-Dependent Growth Gene, rdgC, Is Involved in Gonococcal Pilin Antigenic Variation

Corresponding author: H. Steven Seifert, Northwestern University Medical School, 303 E. Chicago Ave., Searle 6-458, Mailcode S213, Chicago, IL 60611., h-seifert{at}nwu.edu (E-mail)

Communicating editor: R. MAURER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Neisseria gonorrhoeae pilin undergoes high-frequency changes in primary amino acid sequence that aid in the avoidance of the host immune response and alter pilus expression. The pilin amino acid changes reflect nucleotide changes in the expressed gene, pilE, which result from nonreciprocal recombination reactions with numerous silent loci, pilS. A series of mini-transposon insertions affecting pilin antigenic variation were localized to three genes in one region of the Gc chromosome. Mutational analysis with complementation showed that a Gc gene with sequence similarity to the Escherichia coli rdgC gene is involved in pilus-dependent colony phase variation and in pilin antigenic variation. Furthermore, we show that the Gc rdgC homologue is transcriptionally linked in an operon with a gene encoding a predicted GTPase. The inability to disrupt expression of this gene suggests it is an essential gene (engA, essential neisserial GTPase). While some of the transposon mutations in rdgC and insertions in the 5'-untranslated portion of engA showed a growth defect, all transposon insertions investigated conferred an aberrant cellular morphology. Complementation analysis showed that the growth deficiencies are due to the interruption of RdgC expression and not that of EngA. The requirement of RdgC for efficient pilin variation suggests a role for this protein in specialized DNA recombination reactions.

THE obligate human pathogen Neisseria gonorrhoeae (the gonococcus, Gc) expresses type IV pili, which are filamentous, surface-exposed organelles essential for full infectivity (KELLOGG et al. 1963 ; SWANSON et al. 1987 ). Gc pili are required for both adherence to the urogenital epithelium (SWANSON 1973 ) and for full DNA transformation competence (SPARLING 1966 ; SEIFERT et al. 1990 ). The pilus is composed primarily of the pilE gene product, pilin (SWANSON et al. 1971 ). Gc pilus antigenic variation occurs when DNA sequences from one of several silent pilin gene copies (pilS) are transferred unidirectionally to replace variable sequences within the pilE gene (HAGBLOM et al. 1985 ; HAAS and MEYER 1986 ; SEGAL et al. 1986 ). The Sma/Cla DNA repeat, located at the 3' end of all pilin loci, is necessary for efficient pilin antigenic variation (WAINWRIGHT et al. 1994 ). The pilE sequence changes alter the amino acid sequence of the pilin protein and the antigenicity of the pilus (VIRJI et al. 1983 ; VIRJI and HECKELS 1984 ). Presumably, these antigenic changes aid in the avoidance of the host immune system (BRINTON et al. 1978 ; BOSLEGO et al. 1991 ) and alter pilus-mediated epithelial cell adherence (VIRJI et al. 1982 ; RUDEL et al. 1992 ; JONSSON et al. 1994 ; LONG et al. 1998 ). Antigenic variation of pilin can confer changes in pilus-dependent colony morphology (HAGBLOM et al. 1985 ; HAAS and MEYER 1986 ; SWANSON et al. 1986 ; LONG et al. 1998 ), which reflect a phase switch between a number of states of pilus expression (SWANSON and BARRERA 1983 ; HAAS and MEYER 1986 ; LONG et al. 1998 ), although colony morphology phase switches can also occur by means independent of antigenic variation (MEYER et al. 1984 ; SEGAL et al. 1985 ; HAAS et al. 1987 ; JONSSON et al. 1991 ).

In Gc, DNA recombination is required for natural DNA transformation, DNA repair, and pilus antigenic variation, and all the processes are mediated by RecA (KOOMEY et al. 1987 ; SEIFERT 1996 ; MEHR and SEIFERT 1997 ). In Escherichia coli, RecA promotes DNA renaturation and strand exchange during homologous recombination, and mediates recombinational repair and SOS induction (RADDING 1989 ; WEST et al. 1992 ; COX 1993 ; KOWALCZYKOWSKI et al. 1994 ). Gc and E. coli RecA share 65% identity and 81% similarity (FYFE and DAVIES 1990 ) and can cross-complement for some recombination functions (KOOMEY and FALKOW 1987 ; FYFE and DAVIES 1990 ). Since E. coli RecA acts in concert with other proteins, the necessity for Gc RecA in pilin antigenic variation suggests that other factors involved in homologous recombination also have a role in this process.

Although Gc pilin antigenic variation is mediated by RecA-dependent homologous recombination, several attributes of this system suggest that both general and pil-specific recombination factors are involved. First, Gc pilin antigenic variation is usually unidirectional, whereas homologous recombination often results in the reciprocal exchange of DNA sequences. Second, pilin antigenic variation occurs at a high frequency of >10^-2 pilE variants per total pilE (SERKIN and SEIFERT 1998 ). Third, 30–450 bp of pilS DNA sequence is transferred into the pilE gene during each recombination, which is postulated to be mediated by recombination events at limited regions of homology of 8–30 bp (HOWELL-ADAMS et al. 1996 ; B. S. HOWELL-ADAMS and H. S. SEIFERT, unpublished results). Lastly, the pilin-associated Sma/Cla repeat, which is required for efficient pilin antigenic variation (WAINWRIGHT et al. 1994 ), is the target for multiple site-specific binding proteins (WAINWRIGHT et al. 1994 , WAINWRIGHT et al. 1997 ). Thus, it is likely that both homologous recombination activities and pilin-specific factors combine to accomplish pilin antigenic variation.

Previously, we screened random Gc transposon-generated mutants to identify genes essential for pilin antigenic variation (MEHR and SEIFERT 1997 ). We identified 22 unique transposon mutations that conferred an antigenic variation deficient (Avd) phenotype through multiple backcrosses (MEHR and SEIFERT 1997 ). An Avd phenotype was defined as the inability to produce a pilin antigenic variation-specific 460-bp PCR product in a colony-based assay. All Avd mutants also showed a reduced frequency of colony morphology phase variation. Of the transposon mutations, 11 were located in two regions of the Gc chromosome-encoding homologues of the E. coli RecF-pathway genes recQ and recO, suggesting that a Gc RecF-like recombination pathway is involved in pilin antigenic variation (MEHR and SEIFERT 1998 ). Furthermore, mutation of genes in the Gc RecBCD pathway showed that they are not involved in pilin antigenic variation, but are involved in DNA transformation (MEHR and SEIFERT 1998 ). Mutations in either Gc recombination pathway altered the repair of DNA damage (MEHR and SEIFERT 1998 ). Some of these findings are in contrast to the results of CHAUSSEE et al. 1999 , who reported that mutations in the Gc recD gene did not alter DNA repair capabilities, showed a minor decrease in DNA transformation competence, and increased pilus-dependent colony variations. The basis for the descrepancy between these two reports is not presently known.

In this article, we show that the other 11 transposon insertions that produce an Avd phenotype are grouped in a region of the Gc chromosome that encodes three genes; orf2, engA, and rdgC. Some of the transposon-generated mutations produced a growth deficiency in addition to interfering with pilin antigenic variation, and selected transposon insertions resulted in abnormal cellular morphologies that were independent of the growth phenotype. engA was shown to be an essential gene that is transcriptionally linked to rdgC. Complementation analysis proves that interruption of RdgC expression is responsible for both pilin variation and growth-deficient phenotypes of the transposon mutants.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Bacterial strains and growth conditions:
N. gonorrhoeae human volunteer isolate FA1090 1-81-S2 (SEIFERT et al. 1994 ) was grown on Gc medium base (GCB; Difco, Detroit) plus Kellogg supplements (KELLOGG et al. 1963 ) at 37° in 5% CO₂. Antibiotic resistance of Gc strain FA1090 was selected at 2 µg/ml erythromycin (Erm), 50 µg/ml spectinomycin (Spc), 2 µg/ml chloramphenicol (Cam), and 0.2 µg/ml tetracycline (Tet). E. coli strain DH10B (GIBCO BRL, Gaithersburg, MD) was grown on Luria-Bertani medium (Difco) at 37°. Antibiotic resistance of E. coli was selected at 40 µg/ml kanamycin (Kan), 250 µg/ml Erm, 25 µg/ml Cam, and 150 µg/ml ampicillin (Amp).

Molecular methods:
Unless otherwise noted, all enzymes used in this study were supplied and used under the conditions recommended by New England Biolabs (NEB, Beverly, MA). Isolation of DNA from agarose gels was accomplished using GeneClean (Bio101, Vista, CA). DNA ends with 5' or 3' overhangs were blunted using T4 DNA polymerase. Blunted vector DNA ends were treated with shrimp alkaline phosphatase (SAP) (United States Biochemical, Cleveland) prior to ligation with blunted insert DNA. Ligations were performed and prepared for transformation as previously described (MEHR and SEIFERT 1997 ). E. coli were transformed by electroporation using the Gene Pulser II electroporation system (Bio-Rad, Richmond, CA) according to the manufacturer's specifications. Southern blots were performed as described (SAMBROOK et al. 1989 ). Filters were probed with [-³²P]dCTP-labeled PCR products as described previously (MEHR and SEIFERT 1997 ). PCR was carried out essentially as described previously (MEHR and SEIFERT 1997 ), except for increasing extension times from 1.5 min/kb to 2 min/kb for Pfu thermostable polymerase (Stratagene, La Jolla, CA). RNA was prepared for Northern blotting by the TRIzol method (GIBCO BRL). Northern blots were performed as described (SAMBROOK et al. 1989 ). cDNA for RT-PCR analysis was synthesized according to the method described by WAINWRIGHT et al. 1994 using the primer RDGC1 and total RNA isolated from FA1090 variant 1-81-S2 at the late exponential stage of growth. PCR reactions were done in duplicate on 3 µl of the resultant cDNA with primer pairs GTP7/GTP8 or GTP5/RDGC2, to show linkage between engA and rdgC.

Cloning, sequencing and mutation of orf2, engA, and rdgC:
The transposon mutations conferring the Avd phenotype were cloned from the Gc chromosome into pHSS6 (SEIFERT et al. 1986 ). The clones were named according to the Avd strain from which they were created. For example, clone pAvd-J7 contains the ClaI chromosomal fragment harboring the mTnEGNS insertion from Avd strain J7. DNA sequencing of these mutations was accomplished using an Applied Biosystems International (Foster City, CA) ABI373 sequencer, and Perkin-Elmer (Norwalk, CT) AmpliTaq DNA sequencing kit with primers listed in Table 1. Sequence similarities to known genes or proteins were found by searching GenBank at the National Center for Biotechnology Information. The Gapped-BLAST search function (ALTSCHUL et al. 1997 ) was used to determine percent similarity and identity to known proteins. Double-strand DNA sequences of orf2, engA, and rdgC were deposited in the DDBJ/EMBL/GenBank databases under accession no. AF058711.

To mutate engA, the 3-kb region of the chromosome containing orf2, engA, and the 5' half of rdgC was PCR amplified using HFLU1 and RDGC1 (Table 1). The PCR product was cloned into pCR-Blunt (Invitrogen) to create pENGA. There are three ClaI sites within engA, at ~160, 650, and 690 bp downstream of the start codon. pENGA was digested completely with ClaI, and the large vector-containing fragment was isolated, blunted, and ligated to the blunted NotI fragment from pJD1145 (J. P. DILLARD, unpublished results) that contains the nonpolar ermC* gene. pENGA/Erm clones contained an ~530-bp deletion of engA replaced by ermC* fragment. The engA::ermC* mutation was transferred to Gc by transformation. Southern analysis showed that all Gc engA::ermC* insertions were merodiploid maintaining both the mutant and wild-type (WT) copy of engA (data not shown).

To complement the mutations in engA and rdgC, the wild-type versions of the genes were amplified from the genome using Pfu polymerase without their promoters. The PCR products were cloned into pGCC6 (I. J. MEHR and H. S. SEIFERT, unpublished results), which contains a fragment of the gonococcal chromosome where a mini-transposon insertion is in an intragenic region of the chromosome with no detectable transcription (I. J. MEHR and H. S. SEIFERT, unpublished results). Within the mini-transposon is a multi-cloning site adjacent to the lac regulatory sites and the lac repressor. The construct allows controlled expression of genes in trans on the gonococcal chromosome. Oligonucleotides RDGC9 and RDGC10 were used to amplify the wild-type rdgC (Table 1). The PCR product was digested with PacI and NsiI, specific for the 5' ends of RDGC9 and RDGC10, respectively, and ligated to PacI, NsiI double-digested pGCC6, creating clone pGCC6/rdgC. Similarly, oligonucleotides ENGA-6 and ENGA-5 were used to amplify and clone engA (Table 1). engA was also PCR amplified with its own promoter using oligonucleotides ENGA-4 and ENGA-5. This PCR product was cloned into pGCC5, which has the mini-transposon insertion without the lac regulatory cassette (I. J. MEHR and H. S. SEIFERT, unpublished results). Presence and orientation of the PCR products within the GCC vectors were determined by restriction enzyme digestion. Transposon mutants engA3::mTnEGNS, rdgC1::mTnEGNS, and rdgC4::mTnEGNS were each transformed with either the GCC6 allele alone, or the complementation alleles in clones pGCC6/rdgC and pGCC6/engA. Presence of the GCC constructs within the transformants was determined by PCR and Southern analysis (data not shown). Complemented mutants were grown in the presence of 1 mM IPTG to provide transcription from the lac promoters in the mTnGCC6 construct. The GCC5/engA was not grown with IPTG, as engA was under the control of its own promoter.

Phenotypic analyses of mutations:
Phenotypic analyses of the mutations were performed as described previously (MEHR and SEIFERT 1997 ). Briefly, the colony-based PCR assay for pilin antigenic variation was performed using an oligonucleotide specific for the pilE ribosomal binding site, PILRBS, and an oligonucleotide specific for two different pilS copies, HV-1 (Table 1). Individual Gc colonies expressing Erm^R were collected with sterile filter paper, placed in a lysis solution, and boiled. A portion of each colony lysate was used as template in the PCR reaction. Avd mutants did not produce a 460-bp PCR product. Growth on solid media was assessed by picking and suspending Gc colonies in GCBL media, plating serial dilutions, and determining the mean colony-forming units (CFUs) per colony. DNA transformation efficiency was determined in liquid medium by standard techniques (MEHR and SEIFERT 1998 ). A total of 1 µg/ml of chromosomal DNA encoding Spc^R was incubated with ~10⁷ cfu/ml of Gc at 37°, 5% CO₂ for 15 min, followed by a 1:10 dilution and another 37°, 5% CO₂ incubation of 4–6 hr. Efficiency was expressed as the mean number of Spc^R transformants per colony-forming unit. UV resistance was assessed by exposing different amounts of Gc plated on solid media to 0, 2, 4, 6, or 8 J/m², and comparing the surviving colony-forming units at each dose to the unexposed colony-forming units.

Colony phase variation was assessed by suspending 5–10 Gc colonies exhibiting a highly piliated colony morphology in GCBL, and plating on solid media (WAINWRIGHT et al. 1994 ). Colony phase variation was expressed as the percent colonies exhibiting a nonpiliated colony morphology (SWANSON et al. 1971 ).

Resistance to ionizing radiation was determined by exposing Gc suspended in GCBL to 0, 9, 18, or 27 krad of -rays in a Nordion Gammacell40 irradiator (Kanata, Ontario, Canada). The Gc suspension was diluted and plated on GCB. Relative resistance at each dose was calculated as percent survival compared to the number of colony-forming units in the unirradiated sample.

Transmission electron microscopy:
Bacterial cell morphology was examined using transmission electron microscopy of thin sections and whole mounts of gonococci. For thin sections, Gc were grown overnight in GC liquid medium (1.5% proteose peptone no. 3 [Difco], 0.4% K₂HPO₄, 0.1% KH₂PO₄, 0.1% NaCl) with Kellogg supplements and 0.042% sodium bicarbonate (MORSE and BARTENSTEIN 1974 ) with or without 1 mM IPTG at 37° with rotation. After 16–18 hr of growth, cultures were diluted back into log phase, and grown in identical conditions for an additional 3–4 hr. Gc from 1.5 ml of each culture were pelleted, washed in ice-cold 100 mM NaPO₄ buffer (pH 5.5), and pelleted again. Gc were fixed with 4% glutaraldehyde in 50 mM NaPO₄ buffer for 3 hr at room temperature. Fixed samples were washed four times at room temperature with 50 mM NaPO₄ buffer, and were postfixed for 1 hr at room temperature with 1% osmium tetroxide in 100 mM NaPO₄ buffer. After washing the samples four times for 15 min each with 50 mM NaPO₄ buffer, samples were dehydrated in increasing concentrations of ethanol (30, 50, 70, 90, 95, and 100%) and finally in propylene oxide. The samples were infiltrated with EM bed-812 resin (Electron Microscopy Sciences, Fort Washington, PA), embedded in Beem capsules, and sectioned (80 nm, silver) on a Reichert-Jung Ultracut E microtome. The sections were poststained with 4% uranyl acetate for 15 min and with Reynolds lead for 2 min.

For preparation of whole mounts, Gc were grown on GCB plates with or without 1 mM IPTG for 18–20 hr. Poly-L-lysine (1 µg ml^-1)-treated, carbon-coated nickel grids (Ladd Research Industries, Inc., Williston, VT) were used to lift cells from colonies. Grids were then incubated in drops of the following solutions: 1% glutaraldehyde in 100 mM cacodylate buffer for 2 min, twice in sterile water for 3 sec, and 1% uranyl acetate for 30 sec, and then were air dried. All samples were examined with a Jeol (Peabody, MA) JEM-100 CX II transmission electron microscope at 60 or 80 kV.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Cloning and sequencing of targets of the Avd mutations:
Gc chromosomal ClaI fragments containing each of the 22 transposon-generated mutations that conferred an Avd phenotype (MEHR and SEIFERT 1997 ) were cloned into E. coli. The DNA sequences flanking the transposon insertions were determined using oligonucleotide probes specific for the mini-transposon ends and were analyzed by comparison with sequences from GenBank and the Gonococcal Genome Sequencing Project database (GGSP; B. A. ROE, S. P. LIN, L. SONG, X. YUAN, S. CLIFTON, and D. W. DYER, University of Oklahoma Advance Center for Genome Technology [http://dna1.chem.ou.edu/gono.html]). Of the transposon insertions, 11 were shown to affect expression of the Gc homologues of recQ or recO (MEHR and SEIFERT 1998 ). The other 11 transposon insertions were located in a single region of the Gc chromosome containing three putative orfs. A total of 3 transposon insertions (from mutants E1, M1, and E2) was located in the noncoding region in between the first and middle orfs, 2 transposon insertions (from mutants J6 and B1) were located in the 3' coding region of the middle orf, and 6 transposon insertions (from mutants J7, C1, L1, L2, L3, and M4) were located in the last orf (Figure 1).

View larger version (12K): In this window In a new window Download PPT slide   Figure 1. Map of the orf2, engA, and rdgC chromosome region. Orf2 is similar to a hypothetical protein from H. influenzae. EngA is similar to bacterial GTPases. RdgC is similar to a protein in E. coli involved in recombination-dependent growth (RYDER et al. 1996 ). Open arrows within the genes indicate the predicted direction of transcription. Sites of mTnEGNS insertions are indicated by thin lines connecting the open-head arrows to the genes. The open head of the arrow indicates the direction of transcription of the ermC gene in the mini-transposon. The names of the Avd mutants are indicated above the arrows. The boxed insertions were chosen for further study. The small shaded triangle above engA indicates the region of engA deleted between the ClaI sites that was replaced by the transcriptionally nonpolar ermC* fragment.

The most 5' orf, orf2, encodes a predicted product that exhibits 16% identity, 45% similarity and 23% identity, 53% similarity to putative trans-membrane proteins from E. coli (GenBank accession no. AE000337-gi1788860) and Haemophilus influenzae (SwissProt accession no. P43989), respectively (data not shown). The second orf, engA (see below), encodes a predicted protein that exhibits consensus GTP binding domains (Figure 2; BOURNE et al. 1991 ; MARCH 1992 ; PILLUTLA et al. 1995 ). Gc EngA is highly similar to hypothetical proteins found in every sequenced and most of the partially sequenced prokaryotic genomes (data not shown). Comparisons to the E. coli (GenBank accession no. AE000337-gi1788858) and H. influenzae (SwissProt accession no. P44536) homologues show 48% identity, 72% similarity and 43% identity, 66% similarity, respectively, to Gc EngA (Figure 2). The gene immediately downstream of engA encodes a predicted protein exhibiting 35% identity, 64% similarity to E. coli RdgC and 16% identity, 41% similarity to an H. influenzae ORF that is also likely to be an RdgC homologue (Figure 1 and Figure 3; RYDER et al. 1996 ). The E. coli rdgC gene was identified as a gene that conferred a growth defect when mutated in a recombination-deficient genetic background (RYDER et al. 1996 ). While RdgC was proposed to be involved in the reinitiation of chromosomal replication after a collapse of a replication fork (RYDER et al. 1996 ), the exact role of this gene product in replication, recombination, or growth is unknown.

View larger version (101K): In this window In a new window Download PPT slide   Figure 2. Sequence alignment of N. gonorrhoeae EngA with E. coli and H. influenzae homologues. Gc, N. gonorrhoeae; Ec, E. coli; and Hf, H. influenzae. Black shading indicates conserved residue, gray shading represents similar residues, and no shading shows no conservation. Consensus G1, G2, G3, and G4 GTPase domains are indicated with an overline (BOURNE et al. 1991 ). The G2 domain is the GTPase domain with the least sequence conservation.

View larger version (64K): In this window In a new window Download PPT slide   Figure 3. Sequence alignment of N. gonorrhoeae, E. coli, and H. influenzae RdgC homologues. Gc, Ec, and Hf and shading are the same as in Figure 2.

Many, but not all, of the transposon mutations affected both growth and antigenic variation (MEHR and SEIFERT 1997 ). To determine whether these phenotypes were due to the disrupted expression of engA or rdgC, three transposon mutations where chosen for further study. E2 (engA3::mTnEGNS, the mutation in mutant Avd-E2) was located in the 5'-noncoding region of engA (Figure 1), and exhibited both an Avd phenotype and a growth defect (MEHR and SEIFERT 1997 ). J7 (rdgC1::mTnEGNS, the mutation in Avd-J7) was located in the 5' portion of rdgC (Figure 1), and exhibited an Avd phenotype, but wild-type colony growth (MEHR and SEIFERT 1997 ). L3 (rdgC4::mTnEGNS, the mutation in Avd-L3) was located in the 3' portion of rdgC (Figure 1), and exhibited both an Avd phenotype and a growth defect (MEHR and SEIFERT 1997 ).

Expression of rdgC in trans complements Avd mutations for pilin phase and antigenic variation:
Only a few plasmid complementation systems have been developed for Neisseria, and for single-copy stable complementation, chromosomal insertions are preferred. Previously, we had identified a random mini-transposon insertion mutant (Cg8; MEHR and SEIFERT 1997 ) that did not alter any recombination-associated phenotypes, and was not in an expressed region of the chromosome. This mini-transposon insertion has been used to create a Gc chromosomal complementation system (NICS; I. J. MEHR and H. S. SEIFERT, unpublished results). The Gc engA or rdgC gene was introduced into the unexpressed site (carrying the Erm^R, mini-transposon insertion of mutant Cg8) in the chromosome of Gc strain FA1090, variant 1-81-S2 (SEIFERT et al. 1994 ) under the control of the lac regulatory system (SEIFERT 1997 ). Transformants were confirmed by Southern blot analysis, and were named GCC6/engA and GCC6/rdgC. The E2 (engA3), J7 (rdgC1), and L3 (rdgC4) transposon mutations were then moved by DNA transformation from a recA6 background to the wild-type recA background of the engA and rdgC complementation strains. Six phenotypes were tested: pilin antigenic variation as measured by the colony-based PCR assay, pilus-based colony morphology phase variation, growth, DNA transformation efficiency, UV resistance, and -ray resistance. None of the three transposon mutants carrying a wild-type recA gene exhibited a significant decrease in their resistance to DNA damage by UV light or -ray radiation, and all three transposon mutants were wild type for DNA transformation (data not shown).

Interestingly, none of the engA3 (E2), rdgC1 (J7), or rdgC4 (L3) transposon mutations in the wild-type recA background carrying an empty complementation locus (GCC6) produced the consistent Avd phenotype found by the colony-based PCR assay of the parental recA6 strains. Approximately one-fifth of all PCR reactions done on colony lysates of these mutants with a wild-type recA gene exhibited the 460-bp recombination-dependent products. This is in contrast to these same mutations in an induced recA6 background (RecA⁺), where none of the PCR reactions generated product (data not shown; MEHR and SEIFERT 1997 ). This suggested that the L3, E2, and J7 Avd mutations in a wild-type recA background were disabled only partially for pilin antigenic variation. Analysis of pilus-dependent colony morphology phase variation supported this conclusion since the E2 (engA3), J7 (rdgC1), and L3 (rdgC4) in the wild-type recA background decreased phase variation 2- to 5-fold (Figure 4), while these same mutations decreased colony phase variation >20-fold in the induced recA6 background (data not shown; MEHR and SEIFERT 1997 ). Since a fully induced recA6 gene does not produce wild-type levels of RecA (SEIFERT 1997 ), the differences in phase and antigenic variation phenotypes are likely to be due to different RecA levels. However, a role for the recA promoter (which is not used in the recA6 gene) cannot be ruled out. Expression of rdgC, but not engA, in trans complemented the phase variation deficiencies observed for all three transposon insertion mutations, showing that it was the interruption of RdgC activity that was responsible for the Avd phenotypes of all three mutants (Figure 4). The fact that expression of RdgC from the lac promoter resulted in a level of colony phase variation higher than the wild-type strain in mutant L3 (L3-6/rdgC, Figure 4) suggests that a truncated form of RdgC produced in mutant L3 can complement the wild-type proteins for increased frequencies of antigenic variation.

View larger version (23K): In this window In a new window Download PPT slide   Figure 4. Colony morphology phase variation of complemented rdgC and engA mutants. All strains are derived from FA1090 1-81-S2 (SEIFERT et al. 1994 ). L3 contains the rdgC4::mTnEGNS mutation. E2 contains the engA3::mTnEGNS mutation. J7 contains the rdgC1::mTnEGNS mutation. Complemented strains and controls are indicated by a -5 for the GCC5 construct, which is the same as GCC6 without the lac sequences, a -6 for the GCC6 construct, a -6/rdgC for the GCC6/rdgC construct, and a -6/engA for the GCC6/engA. All strains are recA+, except for recA9, which has recA disrupted by the ermC gene. The mean and standard error of six experiments are shown. *P < 0.05, **P < 0.001 relative to WT-5 by the Student's t-test.

Expression of rdgC in trans complements Avd mutations for growth:
The E2 (engA3) and L3 (rdgC4) mutations in the GCC6 background produced growth defects (Figure 5) similar to those observed in the recA6 background (MEHR and SEIFERT 1997 ). As expected, the J7 (rdgC1) transposon mutation in the GCC6 background did not confer a growth defect (Figure 5), consistent with its phenotype in the recA6 background (MEHR and SEIFERT 1997 ). The engA3 and rdgC4 growth defects were partially complemented by trans expression of rdgC, but not engA (Figure 5).

View larger version (26K): In this window In a new window Download PPT slide   Figure 5. Growth of complemented rdgC and engA mutants. Strains are the same as previously shown. The mean and standard error of six experiments are shown. *P < 0.05 compared to WT-5, recA9-5, and uncomplemented mutants by the Student's t-test.

engA and rdgC are in an operon:
The isolation of several Avd mutations in the 5'-noncoding region of engA (Figure 1) and the lack of promoter sequences directly upstream of the rdgC coding region (data not shown) suggested that these two genes were arranged in an operon. RT-PCR analysis confirmed the genetic evidence that engA and rdgC are transcriptionally linked (data not shown).

rdgC and engA mutations confer altered cellular morphology:
Because the E2 (engA3) and L3 (rdgC4) mutants exhibited a growth defect, we investigated whether these and the J7 (rdgC1) mutant showed changes in cellular morphology. Transmission electron microscopy of both thin sections and whole mounts of the mutant strains and of wild-type FA1090 demonstrated that all three mutant strains exhibited abnormal cellular morphologies (Figure 6). While FA1090 showed the normal diplococcal shape with one septum-like structure in the center of diplococci, the three mutants had a more rounded cell shape, and often contained more than one septum-like structure per cell (Figure 6, bottom). The thin sections confirmed the presence of multiple septal-like structures in some of the mutant cells (as indicated by arrowheads in Figure 6, top). Furthermore, the septum-like structures of some of the mutant cells were off center (as indicated by arrows in Figure 6, top). Finally, the mutant strains expressed an electron-dense layer surrounding each cell, whereas FA1090 only exhibited this electron-dense material between neighboring cells (Figure 6, bottom). These alterations in cellular morphology in mutants that are impaired in pilin antigenic variation suggest a link between these processes, but show that the growth phenotypes are not explained solely by abnormal cell division.

View larger version (108K): In this window In a new window Download PPT slide   Figure 6. Transmission electron microscopy of rdgC and engA mutants. FA1090 denotes wild-type 1-81-S2, and mutant strains are the same as previously shown. The top shows uranyl-acetate- and Reynolds-lead-stained thin sections of gonococci. Arrows indicate aberrantly placed septa. Arrowheads indicate cells with multiple septa. The bottom shows uranyl-acetate-stained whole mounts of gonococci. The bar in each panel represents 1 µm.

engA is an essential gene:
The E. coli Era GTPase is essential, shown by a conditionally lethal mutation (MARCH et al. 1988 ). The engA gene product shows sequence similarity to the consensus GTPase domains of Era (Figure 2), although EngA is not a close homologue to Era. Since no transposon mutations were found in the 5' portion of the coding region of engA (Figure 1) despite the isolation of several independent insertions upstream and downstream of the gene, it was possible that engA is an essential gene in Gc. To test this hypothesis, engA was cloned and disrupted with the ermC* gene (Figure 1). When this mutation was transformed into Gc, Erm^R transformants were acquired infrequently, at ~1 x 10^-7 Erm^R transformants/cfu, and all Erm^R transformants were merodiploid for engA in that each contained both the wild-type engA gene and an ermC* disrupted engA (data not shown). This result is consistent with the engA gene product being essential.

engA was also cloned into the GCC5 expression vector. GCC5 is in the same locus as GCC6 but does not carry lac regulatory sequences. This construct allowed expression of engA under the control of its own promoter (I. J. MEHR and H. S. SEIFERT, unpublished results; see MATERIALS and METHODS). The GCC5/engA construct was introduced into the chromosome of wild-type Gc (Figure 4). Gc GCC5/engA positive clones were then transformed with mutated versions of engA that was disrupted by ermC* (Figure 1). Transformants that contained an engA disruption at the wild-type chromosomal locus were acquired only in the presence of the GCC5/engA construct (data not shown), confirming that disruption of engA is lethal and confirming that a promoter sequence is present upstream of engA.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Cloning and sequence analysis identified that 11 of 22 transposon mutations conferring an Avd phenotype (MEHR and SEIFERT 1997 ) were located in a region of the gonococcal chromosome containing three putative orfs: orf2, engA, and rdgC (Figure 1). Database searches suggested that Orf2 might be a trans-membrane protein of unknown function, that EngA was likely to be a GTPase, and that RdgC was likely to be involved in growth linked to recombination. Analysis of three mutants, E2 (engA3::mTnEGNS, located in the 5'-noncoding region of engA), J7 (rdgC1::mTnEGNS, located in the 5' portion of rdgC), and L3 (rdgC4::mTnEGNS, located in the 3' portion of rdgC), revealed that both the growth and pilin antigenic variation phenotypes conferred by these mutations were complemented by the expression of rdgC in trans. Transmission electron microscopy demonstrated that an abnormal cellular morphology was also associated with each of these three transposon mutations, but that the abnormal cellular morphology did not correlate with the different growth phenotypes.

EngA is somewhat similar to the Era/ThdF family of essential GTPases, although closer homologues are found in most sequenced bacterial genomes. All of these EngA homologues are hypothetical proteins with no known function, but the conserved GTPase domains and ubiquitous presence in bacteria suggest a primary conserved function. Era has been studied extensively in E. coli, and is thought to be a membrane-localized (MARCH 1992 ; LIN et al. 1994 ), RNA binding protein (CHEN et al. 1999 ; MEIER et al. 1999 ) involved in cell division (MARCH et al. 1988 ). E. coli ThdF is involved in thiophene and furan oxidation (ALAM and CLARK 1991 ). Both proteins bind and hydrolyze GTP by similar motifs, although their roles in bacterial physiology are apparently quite divergent. There is significant sequence divergence between EngA, Era, and ThdF, except for the four highly conserved GTP binding domains, of which EngA has two sets (Figure 2). We show that EngA is essential to Gc growth, and assume that EngA is a GTPase, based on sequence similarity to E. coli Era and ThdF (BOURNE et al. 1991 ; MARCH 1992 ; PILLUTLA et al. 1995 ). It will be interesting to determine whether the close homologues in other bacterial species are also essential, and whether they might be a target for novel antimicrobials.

It is puzzling that growth defects were caused by transposon insertions in both the 5'-noncoding region of engA and the 3'-coding region of rdgC, but not the 5'-coding region of rdgC. Furthermore, growth defects caused by transposon insertions in either region can be complemented by rdgC expression in trans, suggesting that disruption of rdgC expression alone was responsible for the growth defects. One hypothesis for the variable growth phenotypes is that intermediate expression of rdgC is detrimental to the cell, but full or no expression of rdgC has no growth phenotype. This hypothesis assumes that the insertions in the 5'-noncoding region of engA reduce transcription of the operon where enough EngA is produced for cell survival, but not enough RdgC is produced to avoid a growth defect. The 3'-coding region insertions in rdgC would be predicted to produce a truncated version of RdgC, which could affect the growth phenotype by binding and inactivating a putative interacting protein. In support of this hypothesis is the observation that the growth phenotype became more severe as the mini-transposon insertions left more of the RdgC sequence intact (MEHR and SEIFERT 1997 ). Therefore, mutations in the 5'-coding region of RdgC would effectively block expression of the interfering domains of RdgC, providing a null phenotype and allowing for normal growth on solid media.

The behaviors of some rdgC mutations in Gc are in contrast to rdgC mutations in E. coli (RYDER et al. 1996 ). In E. coli, rdgC mutations only showed a growth phenotype when combined with other mutations in recombination genes, recBC, sbcBC, and recA (RYDER et al. 1996 ). Our analysis of the repair phenotypes of gonococcal recB recC and recD mutants suggested that gonococci are naturally sbcB deficient and resemble E. coli recBC sbcBC mutants (MEHR and SEIFERT 1998 ). It is clear from the present study that in the gonococcus RdgC contributes to pilin antigenic variation and is necessary for wild-type growth and normal cellular morphology. The phenotypes in the gonococcus are reminiscent of the E. coli phenotypes, but show some differences. First, only some of the gonococcal rdgC mutations show a growth phenotype, and the expression of these is independent of other mutations. Second, since gonococcal rdgC mutations are deficient in pilin antigenic variation, the RdgC protein plays a role in this specialized recombination system. Finally, no recombination phenotype was observed for E. coli rdgC mutants (RYDER et al. 1996 ), although E. coli does not carry a complex recombination system analogous to the pilin system.

Two other gonococcal genes have been isolated that link DNA recombination and growth or cell division: comL (FUSSENEGGER et al. 1996A ) and tpc (FUSSENEGGER et al. 1996B ). Both genes were identified by screening for reduced transformation competence. A comL mutant shows smaller bacterial cell size (FUSSENEGGER et al. 1996A ), while tpc mutants exhibit abnormal septation by growing in clusters of four instead of two (FUSSENEGGER et al. 1996B ). The occurrence of three genes that link cell division to DNA recombination may indicate a close relationship between recombination machinery and the cell envelope in this gram-negative, diplococcal organism. The presence of engA and rdgC in an operon, the possible role for RdgC in cell division, and the essential nature of EngA raise the possibility that these proteins cooperate during cell division. The inability to mutate engA will make determining whether a relationship exists between these proteins more difficult.

The data presented here demonstrate that in the gonococcus RdgC plays a role in pilin antigenic variation, but that rdgC mutants are only partially disabled in a wild-type recA background. The role that RdgC plays in pilin antigenic variation is less obvious than the role of RecA, RecO, or RecQ, which are likely involved in DNA strand exchange and homologous pairing (MEHR and SEIFERT 1998 ). All of our molecular models for the gene conversion events leading to pilin antigenic variation require two gonococcal chromosomes (HOWELL-ADAMS et al. 1996 ; B. S. HOWELL-ADAMS and H. S. SEIFERT, unpublished results), and are therefore compatible with a function that links pilin variation to replication. Clearly, further biochemical characterization of RdgC is required to elucidate RdgC's role in pilin antigenic variation, and possibly chromosome replication. It is interesting to note that RdgC homologues are found only in 2 of the 12 completely sequenced bacterial genomes, E. coli and H. influenza, and only a small subset of partially sequenced organisms besides Neisseria species, while engA homologues are found in most of the fully and partially sequenced organisms. The identification of a role for RdgC in the recombinations leading to pilin variation provides the first direct role for this protein in a recombination process.

	FOOTNOTES

¹ Present address: Laboratory Corporation of America, Research Triangle Park, NC 27709.

	ACKNOWLEDGMENTS

We thank Kimberly Kline for technical assistance and Seifert laboratory members for review of the manuscript. We also thank Maya Moody of the Northwestern University Medical School Cell Imaging Facility for electron microscopy assistance. This work was supported by Public Health Service grant R01 AI33493. I.J.M. was partially supported by Public Health Service training grant T32 AI07476. C.D.L. was partially supported by Public Health Service training grant T32 GM08061.

Manuscript received October 6, 1999; Accepted for publication October 11, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALAM, K. Y. and D. P. CLARK, 1991  Molecular cloning and sequence of the thdF gene, which is involved in thiophene and furan oxidation by Escherichia coli. J. Bacteriol. 173:6018-6024[Abstract/Free Full Text].

ALTSCHUL, S. F., T. L. MADDEN, A. A. SCHAFFER, J. ZHANG, and Z. ZHANG et al., 1997  Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25:3389-3402[Abstract/Free Full Text].

BOSLEGO, J. W., E. C. TRAMONT, R. C. CHUNG, D. G. MCCHESNEY, and J. CIAK et al., 1991  Efficacy trial of a parenteral gonococcal pilus vaccine in men. Vaccine 9:154-162[Medline].

BOURNE, H. R., D. A. SANDERS, and F. MCCORMICK, 1991  The GTPase superfamily: conserved structure and molecular mechanism. Nature 349:117-127[Medline].

BRINTON, C. C., J. BRYAN, J.-A. DILLON, N. GUERNIA, L. JACKOBSON et al., 1978 Uses of pili in gonorrhea control: role of bacterial pili in disease, purification, and properties of gonococcal pili and progress in the development of a gonococcal pilus vaccine for gonorrhea, pp. 155–178 in Immunobiology of Neisseria gonorrhoeae, edited by G. F. BROOKS, E. C. GOTSCHLICH, K. K. HOLMES, W. D. SAWYER and F. E. YOUNG. American Society for Microbiology, Washington, DC.

CHAUSSEE, M. S., J. WILSON, and S. A. HILL, 1999  Characterization of the recD gene of Neisseria gonorrhoeae MS11 and the effect of recD inactivation on pilin variation and DNA transformation. Microbiology 145:389-400[Medline].

CHEN, X., D. L. COURT, and X. JI, 1999  Crystal structure of ERA: a GTPase-dependent cell cycle regulator containing an RNA binding motif. Proc. Natl. Acad. Sci. USA 20:8396-8401.

COX, M. M., 1993  Relating biochemistry to biology: how the recombinational repair function of RecA protein is manifested in its molecular properties. Bioessays 15:617-623[Medline].

FUSSENEGGER, M., D. FACIUS, J. MEIER, and T. F. MEYER, 1996a  A novel peptidoglycan-linked lipoprotein (ComL) that functions in natural transformation competence of Neisseria gonorrhoeae.. Mol. Microbiol. 19:1095-1105[Medline].

FUSSENEGGER, M., A. F. KAHRS, D. FACIUS, and T. F. MEYER, 1996b  Tetrapac (tpc), a novel genotype of Neisseria gonorrhoeae affecting epithelial cell invasion, natural transformation competence and cell separation. Mol. Microbiol. 19:1357-1372[Medline].

FYFE, J. A. and J. K. DAVIES, 1990  Nucleotide sequence and expression in Escherichia coli of the recA gene of Neisseria gonorrhoeae.. Gene 93:151-156[Medline].

HAAS, R. and T. F. MEYER, 1986  The repertoire of silent pilus genes in Neisseria gonorrhoeae: evidence for gene conversion. Cell 44:107-115[Medline].

HAAS, R., H. SCHWARZ, and T. F. MEYER, 1987  Release of soluble pilin antigen coupled with gene conversion in Neisseria gonorrhoeae.. Proc. Natl. Acad. Sci. USA 84:9079-9083[Abstract/Free Full Text].

HAGBLÖM, P., E. SEGAL, E. BILLYARD, and M. SO, 1985  Intragenic recombination leads to pilus antigenic variation in Neisseria gonorrhoeae.. Nature 315:156-158[Medline].

HOWELL-ADAMS, B., L. A. WAINWRIGHT, and H. S. SEIFERT, 1996  The size and position of heterologous insertions in a silent locus differentially affect pilin recombination in Neisseria gonorrhoeae.. Mol. Microbiol. 22:509-522[Medline].

JONSSON, A. B., G. NYBERG, and S. NORMARK, 1991  Phase variation of gonococcal pili by frameshift mutation in pilC, a novel gene for pilus assembly. EMBO J. 10:477-488[Medline].

JONSSON, A. B., D. ILVER, P. FALK, J. PEPOSE, and S. NORMARK, 1994  Sequence changes in the pilus subunit lead to tropism variation of Neisseria gonorrhoeae to human tissue. Mol. Microbiol. 13:403-416[Medline].

KELLOGG, D. S., JR., W. L. PEACOCK, W. E. DEACON, L. BROWN, and C. I. PIRKLE, 1963  Neisseria gonorrhoeae. I. Virulence genetically linked to clonial variation. J. Bacteriol. 85:1274-1279[Abstract/Free Full Text].

KOOMEY, J. M. and S. FALKOW, 1987  Cloning of the recA gene of Neisseria gonorrhoeae and construction of gonococcal recA mutants. J. Bacteriol. 169:790-795[Abstract/Free Full Text].

KOOMEY, M., E. C. GOTSCHLICH, K. ROBBINS, S. BERGSTRÖM, and J. SWANSON, 1987  Effects of recA mutations on pilus antigenic variation and phase transitions in Neisseria gonorrhoeae.. Genetics 117:391-398[Abstract/Free Full Text].

KOWALCZYKOWSKI, S. C., D. A. DIXON, A. K. EGGLESTON, S. D. LAUDER, and W. M. REHRAUER, 1994  Biochemistry of homologous recombination in Escherichia coli.. Microbiol. Rev. 94:401-465.

LIN, Y. P., J. D. SHARER, and P. E. MARCH, 1994  GTPase-dependent signaling in bacteria: characterization of a membrane-binding site for era in Escherichia coli.. J. Bacteriol. 176:44-49[Abstract/Free Full Text].

LONG, C. D., R. N. MADRASWALA, and H. S. SEIFERT, 1998  Comparisons between colony phase variation of Neisseria gonorrhoeae FA1090 and pilus, pilin, and S-pilin expression. Infect. Immun. 66:1918-1927[Abstract/Free Full Text].

MARCH, P. E., 1992  Membrane-associated GTPases in bacteria. Mol. Microbiol. 6:1253-1257[Medline].

MARCH, P. E., C. G. LERNER, J. AHNN, X. CUI, and M. INOUYE, 1988  The Escherichia coli Ras-like protein (Era) has GTPase activity and is essential for cell growth. Oncogene 2:539-544[Medline].

MEHR, I. J. and H. S. SEIFERT, 1997  Random shuttle mutagenesis: gonococcal mutants deficient in pilin antigenic variation. Mol. Microbiol. 23:1121-1131[Medline].

MEHR, I. J. and H. S. SEIFERT, 1998  Differential roles of homologous recombination pathways in Neisseria gonorrhoeae pilin antigenic variation, DNA transformation, and DNA repair. Mol. Microbiol. 30:697-710[Medline].

MEIER, T. I., R. B. PEERY, S. R. JASKUNAS, and G. J. ZHAO, 1999  16S rRNA is bound to Era of Streptococcus pneumoniae.. J. Bacteriol. 181:5242-5249[Abstract/Free Full Text].

MEYER, T. F., E. BILLYARD, R. HAAS, S. STORZBACH, and M. SO, 1984  Pilus genes of Neisseria gonorrhoeae: chromosomal organization and DNA sequence. Proc. Natl. Acad. Sci. USA 81:6110-6114[Abstract/Free Full Text].

MORSE, S. A. and L. BARTENSTEIN, 1974  Factors affecting autolysis of Neisseria gonorrhoeae. Proc. Soc. Exp. Biol. Med. 145:1418-1421[Medline].

PILLUTLA, R. C., J. D. SHARER, P. S. GULATI, E. WU, and Y. YAMASHITA et al., 1995  Cross-species complementation of the indispensable Escherichia coli era gene highlights amino acid regions essential for activity. J. Bacteriol. 177:2194-2196[Abstract/Free Full Text].

RADDING, C. M., 1989  Helical RecA nucleoprotein filaments mediate homologous pairing and strand exchange. Biochim. Biophys. Acta 1008:131-145[Medline].

RUDEL, T., J. P. M. VAN PUTTEN, C. P. GIBBS, R. HAAS, and T. F. MEYER, 1992  Interaction of two variable proteins (PilE and PilC) required for pilus-mediated adherence of Neisseria gonorrhoeae to human epithelial cells. Mol. Microbiol. 6:3439-3450[Medline].

RYDER, L., G. J. SHARPLES, and R. G. LLOYD, 1996  Recombination-dependent growth in exonuclease-depleted recBC sbcBC strains of Escherichia coli K-12. Genetics 143:1101-1114[Abstract].

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SEGAL, E., E. BILLYARD, M. SO, S. STORZBACH, and T. F. MEYER, 1985  Role of chromosomal rearrangement in N. gonorrhoeae pilus phase variation. Cell 40:293-300[Medline].

SEGAL, E., P. HAGBLÖM, H. S. SEIFERT, and M. SO, 1986  Antigenic variation of gonococcal pilus involves assembly of separated silent gene segments. Proc. Natl. Acad. Sci. USA 83:2177-2181[Abstract/Free Full Text].

SEIFERT, H. S., 1996  Questions about gonococcal pilus phase- and antigenic variation. Mol. Microbiol. 21:433-440[Medline].

SEIFERT, H. S., 1997  Insertionally inactivated and inducible recA alleles for use in Neisseria.. Gene 188:215-220[Medline].

SEIFERT, H. S., E.-Y. CHEN, M. SO, and F. HEFFRON, 1986  Shuttle mutagenesis: a method of transposon mutagenesis for Saccharomyces cerevisiae.. Proc. Natl. Acad. Sci. USA 83:735-739[Abstract/Free Full Text].

SEIFERT, H. S., R. S. AJIOKA, D. PARUCHURI, F. HEFFRON, and M. SO, 1990  Shuttle mutagenesis of Neisseria gonorrhoeae: pilin null mutations lower DNA transformation competence. J. Bacteriol. 172:40-46[Abstract/Free Full Text].

SEIFERT, H. S., C. J. WRIGHT, A. E. JERSE, M. S. COHEN, and J. G. CANNON, 1994  Multiple gonococcal pilin antigenic variants are produced during experimental human infections. J. Clin. Invest. 93:2744-2749.

SERKIN, C. D. and H. S. SEIFERT, 1998  The frequency of pilin antigenic variation in Neisseria gonorrhoeae.. J. Bacteriol. 180:1955-1958[Abstract/Free Full Text].

SPARLING, P. F., 1966  Genetic transformation of Neisseria gonorrhoeae to streptomycin resistance. J. Bacteriol. 92:1364-1371[Abstract/Free Full Text].

SWANSON, J., 1973  Studies on gonococcus infection. IV. Pili: their role in attachment of gonococci to tissue culture cells. J. Exp. Med. 137:571-589[Abstract].

SWANSON, J. and O. BARRERA, 1983  Gonococcal pilus subunit size heterogeneity correlates with transitions in colony piliation phenotype, not with changes in colony opacity. J. Exp. Med. 158:1459-1472[Abstract/Free Full Text].

SWANSON, J., S. J. KRAUS, and E. C. GOTSCHLICH, 1971  Studies on gonococcus infection. I. Pili and zones of adhesion: their relation to gonococcal growth patterns. J. Exp. Med. 134:886-906[Abstract].

SWANSON, J., S. BERGSTRÖM, K. ROBBINS, O. BARRERA, and D. CORWIN et al., 1986  Gene conversion involving the pilin structural gene correlates with pilus+ in equilibrium with pilus- changes in Neisseria gonorrhoeae.. Cell 47:267-276[Medline].

SWANSON, J., S. BERGSTRÖM, J. BOSLEGO, and M. KOOMEY, 1987  Gene conversion accounts for pilin structural changes and for reversible piliation "phase" changes in gonococci. Antonie Van Leeuwenhoek 53:441-446[Medline].

VIRJI, M. and J. E. HECKELS, 1984  The role of common and type-specific pilus antigenic domains in adhesion and virulence of gonococci for human epithelial cells. J. Gen. Microbiol. 130:1089-1095[Medline].

VIRJI, M., J. S. EVERSON, and P. R. LAMBDEN, 1982  Effect of anti-pilus antisera on virulence of variants of Neisseria gonorrhoeae for cultured epithelial cells. J. Gen. Microbiol. 128:1095-1100[Medline].

VIRJI, M., J. E. HECKELS, and P. J. WATT, 1983  Monoclonal antibodies to gonococcal pili: studies on antigenic determinants on pili from variants of strain P9. J. Gen. Microbiol. 129:1965-1973[Medline].

WAINWRIGHT, L. A., K. H. PRITCHARD, and H. S. SEIFERT, 1994  A conserved DNA sequence is required for efficient gonococcal pilin antigenic variation. Mol. Microbiol. 13:75-87[Medline].

WAINWRIGHT, L. A., J. V. FRANGIPANE, and H. S. SEIFERT, 1997  Analysis of protein binding to the Sma/Cla DNA repeat in pathogenic Neisseriae.. Nucleic Acids Res. 25:1362-1368[Abstract/Free Full Text].

WEST, A. B., T. M. ROBERTS, and R. D. KOLODNER, 1992  Regulation of the reverse transcriptase of human immunodeficiency virus type 1 by dNTPs. Proc. Natl. Acad. Sci. USA 89:9720-9724[Abstract/Free Full Text].

WRIGHT, C. J., A. E. JERSE, M. S. COHEN, J. G. CANNON, and H. S. SEIFERT, 1994  Nonrepresentative PCR amplification of variable gene sequences in clinical specimens containing dilute, complex mixtures of microorganisms. J. Clin. Microbiol. 32:464-468[Abstract/Free Full Text].

This article has been cited by other articles:

				J. Hwang and M. Inouye RelA Functionally Suppresses the Growth Defect Caused by a Mutation in the G Domain of the Essential Der Protein J. Bacteriol., May 1, 2008; 190(9): 3236 - 3243. [Abstract] [Full Text] [PDF]

				P. L. Kohler, H. L. Hamilton, K. Cloud-Hansen, and J. P. Dillard AtlA Functions as a Peptidoglycan Lytic Transglycosylase in the Neisseria gonorrhoeae Type IV Secretion System J. Bacteriol., August 1, 2007; 189(15): 5421 - 5428. [Abstract] [Full Text] [PDF]