Departments of Infectious Diseases and Microbiology,1 Infectious Disease Epidemiology, Faculty of Medicine, Imperial College London, London,2 Department of Genitourinary Medicine, Royal Hallamshire Hospital, Sheffield, United Kingdom3
Received 15 May 2003/ Returned for modification 8 July 2003/ Accepted 1 September 2003
| ABSTRACT |
|---|
|
|
|---|
| INTRODUCTION |
|---|
|
|
|---|
In vivo during the course of infection with N. gonorrhoeae some of the organisms lyse, releasing free DNA that will be available for incorporation by recombination into the chromosome of other gonococci, contributing to the genetic variation. Intergenic recombination, recombination between different gonococci, might occur during mixed infections of N. gonorrhoeae, in which an individual is infected with more than one strain in the same anatomical site (3, 7, 9, 20, 21). Evidence for mixed gonococcal infectionswhich are expected to occur in highly sexually active individuals, generating new genotypic strainshas recently been demonstrated (15). Recombination between gonococci of the same genotype, intragenic recombination, has been shown to occur in vivo and in vitro (1, 6, 26), resulting in greater genetic diversity. By using highly discriminatory typing techniques, it is possible to monitor the transmission and genetic variation of strains of N. gonorrhoeae between individuals, discriminating between linked and nonlinked individuals. opa typing (20), which indexes the variation within the 11 gonococcal opa genes that encode outer membrane proteins, involved in adherence and invasion of epithelial cells, has been shown to be able to confirm sexual contacts (10, 24, 25) and is currently one of the most discriminatory genotypic method to differentiate between gonococcal strains (23).
This study aimed to monitor strain variation and transmission of gonorrhea within a city with a single clinic, the Genitourinary Medicine (GUM) clinic, over a 2-year period, by collecting isolates of N. gonorrhoeae from consecutive patients with gonorrhea and characterizing the strains by opa typing. The underlying mechanism of persistence of genotypes was also investigated.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Patient data. Contact tracing data from patients with gonorrhea were obtained by interviews with health advisers at the GUM clinic as previously described (25). Data including age, ethnicity, partial postal code, and information on sexual partnerships in the previous 3 months were sought. Ethics approval was obtained. Patients were offered the choice of notifying their own sexual partners of their infection or providing contact information in order for the health adviser to inform the sexual partners confidentially.
Intragenic recombination. Four isolates from each of the two opa type clusters (opa-1 and opa-4), two from women and two from men, one from each year, were subcultured daily using a sweep from multiple colonies on GC agar (Difco, Becton Dickinson, Oxford, United Kingdom) supplemented with 1% IsoVitaleX (BBL, Becton Dickinson, Oxford, United Kingdom). DNA lysates were prepared on days 1, 7, 14, 21, and 28 and used as the DNA source for opa typing (25).
Intergenic recombination. Transformation experiments were performed using chromosomal DNA from donor strains FA19 Rifr (rifampin-resistant gonococcal strain FA19) (13) or ciprofloxacin-resistant strains that had a single gyrA mutation (gyrA AA91 = Phe), or double gyrA mutation (gyrA AA91 = Phe, AA95 = Gly). DNA was extracted according to the manufacturer's recommended method using a Qiamp kit (QIAGEN, Crawley, West Sussex, United Kingdom), and the concentration of the chromosomal DNA was determined using a spectrophotometer (Shimadzu, Milton Keynes, Buckinghamshire, United Kingdom) with its DNA concentration program. The absorbance of the sample was measured at 260 and 280 nm; a 50-µg/ml concentration of double-stranded DNA gives an optical density (OD) of 1.0 at 260 nm, and the ratio of the OD readings provides an estimate of the purity of the DNA.
Transformation experiments were performed using a suspension of a pilated culture of the recipient isolate prepared in Proteose Peptone with an OD at 540 nm of 1.0 (11). Recipient isolates for the transformation were two isolates from each of the two opa type clusters, the opa-1 and opa-4 clusters, and a clinical isolate that had not been stored at -80°C; all of the isolates were sensitive to rifampin and ciprofloxacin. One hundred microliters of the recipient suspension was placed onto a GC agar plate to which 100 µl of 0.5-µg/ml chromosomal DNA from FA19 Rifr was added, and the plate was incubated at 36°C with 5% CO2 for 6 h. The growth was suspended in 1 ml of Proteose Peptone (Difco, Becton Dickinson), 100 µl was placed onto a GC agar plate containing rifampin (50 µg/ml) in duplicate, the plate was incubated for up to 5 days at 36°C with 5% CO2, and the number of transformants was determined. The total viable count was determined on GC agar. The experiments were repeated using the same recipients with donor DNA from the ciprofloxacin-resistant isolates. Selection of transformants was performed on GC agar plates containing ciprofloxacin at a concentration of 1.0 µg/ml for the DNA from the donor isolate with the double gyrA mutation (MIC of ciprofloxacin, 16.0 µg/ml) or 0.12 µg/ml for the DNA from the donor isolate with a single gyrA mutation (MIC of ciprofloxacin, 1.0 µg/ml). Negative controls to which no donor DNA was added were included in each experiment. Transformation experiments were repeated three times. A selection of transformants from each experiment were auxotyped, serotyped, and opa typed.
por gene sequencing. Eight isolates from each of the two opa type clusters that were isolated throughout the 2-year period were selected. Four control isolates with the same auxotype and serovar as the two clusters, isolated from patients attending the GUM clinic at St. Mary's Hospital, London, United Kingdom, were also sequenced with respect to the por gene. The por gene was amplified from DNA lysates, using primers spanning from the nonencoding region, based on the numbering of FA19 (14) (por1, 5'-72GAATTCAAGCTTATGAAAAAATCCCTGATTGCCC105-3') to the postencoding region (por2, 5'-1052GTCGACCTGCAGTTAGAATTTGTGGCGCAGA1064-3'). Each PCR contained a 200 µM concentration of each deoxynucleoside triphosphate (dATP, dCTP, dGTP, and dTTP) (Roche, Lewes, East Sussex, United Kingdom), 10 µl of PCR buffer (20 mM Tris-HCl [pH 8.4], 50 mM KCl) (Gibco BRL, Paisley, United Kingdom), 1.5 mM MgCl2, 50 pmol of each primer, and 5 µl of DNA lysate. The reaction mix was overlaid with a drop of mineral oil, the samples were heated to 95°C using a Hybaid Ominigene PCR machine for 5 min, and then 2.5 U of Taq DNA polymerase (Gibco BRL) was added. Amplification was performed using 30 cycles of 94°C for 1 min, 60°C for 1 min, and 72°C for 1 min, with a final extension of 72°C for 5 min, and then the mixture was held at 4°C.
The PCR product was purified by gel excision using a Geneclean kit (Bio 101, Stratagene, Amsterdam, The Netherlands). The sequencing reaction was performed using six internal primers (por3, 5'-495TTCAGGCTACCGGCGGATG475-3'; por4, 5'-638TTCAGCGGCAGCGTACAATAC658-3'; por5, 5'-883GACGTACAGGGCATTATTGTC863-3'; BDR 81, 5'-476ATCCGCGCCGGTAGCCTGAAC496-3'; BDR 82, 5'-658GTATTGTACGCTGCCGCTGAA637-3'; and BDR 83, 5'-860TACGACAATAATGCCCTGTAC880-3'). Each reaction mix contained 8 µl of terminator ready reaction mix (Perkin-Elmer, Cambridge, United Kingdom), 6 µl of purified PCR product, 3 pmol of primer, and water to a final volume of 20 µl. Amplification was performed on a Robocycler (Perkin-Elmer) at 96°C for 30 s, 50°C for 15 s, and 60°C for 4 min for 25 cycles, and then the mixture was cooled to 4°C. The sequencing reaction product was purified by ethanol precipitation. To each 20-µl volume, 2 µl of 3 M sodium acetate, pH 4.6, and 50 µl of 95% ethanol was added; the solution was then vortexed and placed on ice for 10 min. The solution was then centrifuged at 5,000 x g for 20 min before the supernatant was removed. The pellet was washed in 250 µl of 70% ethanol and centrifuged for 10 min at 5,000 x g, and the supernatant was removed. The pellet was dried on a hot block, prior to the addition of 25 µl of template suppression reagent (Perkin-Elmer). The sample was vortexed and centrifuged to form a pellet and was heated to 95°C for 2 min to denature. The sample was vortexed and centrifuged to form a pellet again and was loaded onto an ABI Prism 310 sequencer (Perkin-Elmer). Sequences were analyzed and aligned using Sequencher (Gene Codes Corporation, Ann Arbor, Mich.).
Statistical analysis. Chi-square tests, where appropriate, were performed in Microsoft Excel.
Nucleotide sequence accession numbers. All sequences have been submitted to GenBank and have the following accession numbers: AY297696 to AY297711.
| RESULTS |
|---|
|
|
|---|
opa typing. opa typing revealed 115 different opa profiles (Table 1) within the 314 isolates, of which 75 (23.9%) were unique, and two different types, previously designated the opa-1 and opa-4 clusters (25), accounted for 41.7% of the strains. The two large clusters of 62 (opa-1) and 69 (opa-4) isolates with indistinguishable opa profiles persisted throughout the 2-year study period (Fig. 1).
|
|
2 test]).
|
2 test]). The men within the opa-4 cluster were significantly more likely to be of a black ethnic group than were men within other clusters (P = 0.04 [
2 test]). The majority of individuals within the cluster had central Sheffield postcodes, which was significant compared to individuals not within the large clusters (P = 0.01 [
2 test]) (Table 2). Recombination. The opa profiles of the isolates chosen from the clusters that were subcultured in vitro for 4 weeks were indistinguishable at each time point. Intragenic recombination or mutation affecting the opa genes of the isolates was not detected.
All isolates tested were transformable and produced rifampin-resistant colonies. Isolates of the opa-4 cluster produced a significantly higher mean number of rifampin-resistant transformants per microgram of donor DNA (P < 0.001) than isolates of the opa-1 cluster (Table 3). No spontaneous rifampin-resistant mutants occurred. Transformation using ciprofloxacin-resistant DNA as the donor again produced transformants with isolates from both clusters, with a higher mean number of transformants when DNA with a single gyrA mutation was used than when DNA with a double gyrA mutation was used. The 80 selected ciprofloxacin- and rifampin-resistant transformants (10 ciprofloxacin and rifampin transformants of each of the two cluster isolates from opa-1 and opa-4) that were auxotyped, serotyped, and opa typed showed no variation between transformants from the same experiment. The phenotypes of the transformants were the same as those of the original isolate in both auxotype and serovar. The opa profiles of the transformants were the same as those of the original isolate, either opa-1 or opa-4.
|
| DISCUSSION |
|---|
|
|
|---|
N. gonorrhoeae is a nonclonal bacterium that exhibits a high level of genetic variation through recombination and mutation of surface exposed proteins, preventing the establishment of a protective immune response (13, 27). Not all gonococci have the same rate of genetic variation with more clonal lineages, such as AHU IA-2 strains, which diversify more slowly than other phenotypic strains (8). PAOU IB-2 and A IB-3 strains, the opa-1 and opa-4 phenotypes, are not known to be of a more clonal lineage; therefore, the persistence of the two genotypes was surprising. The persistence may have been due either to specific properties of the gonococci or to the fact that the individuals within the clusters had different sexual mixing patterns.
Intragenic recombination and mutation were not detectable in the opa genes of the selected isolates from either of the two clusters over the 4 weeks of in vitro subculture. It is possible that in vitro recombination did not occur and not that the gonococcal strains were incapable of intragenic recombination. In vivo it is possible that intragenic recombination and mutation are more frequent, but the relationship between the stability of opa genes in vivo and in vitro is not clear. There were some sexual contacts who had dates of attendance at the GUM clinic more than a month apart who were concordant in opa type and phenotype, perhaps indicating that in vivo variation of opa genes has not occurred over a long time, supporting the in vitro findings. Intergenic recombination of gonococci with an opa-1 or opa-4 genotype was detected, and their stability was not due to an inability to recombine. In vivo intergenic recombination between different gonococci may occur during a mixed infection (15), and the resulting incorporation of DNA into the chromosome is likely to change the genotype of the strain, depending on which gene(s) was affected by the exchange of DNA. The factors influencing which genes undergo recombination in vitro or in vivo are not known, but typically up to a few kilobases of DNA are replaced (16). The finding that the selected isolates within the two separate clusters had the same por sequence supports the opa typing results, which has been previously shown (10, 24) and which suggests circulation of a single strain within each cluster.
From these results we hypothesize that in vivo variation of opa genes and the resulting opa profile occur predominantly through intergenic recombination, rather than intragenic recombination and mutation. Therefore, the diversity of opa types is dependent on the number of different strains of N. gonorrhoeae recombining in vivo during mixed infections. If no mixed infections occur in vivo the diversification of opa genes is likely to be reduced. The risk of a mixed infection occurring in an individual depends on the sexual activity of that person, the person's position within a sexual network, and also the sexual activities of the person's sexual partners. Within a local population, mixed gonococcal infections within the same individual are more likely in those with higher rates of partner change when there are a variety of different circulating strains. Therefore, in highly sexually active populations a greater level of diversity of opa genes and opa types is expected to occur (23, 25). However, if there were a limited number of circulating strains, or a single strain within a highly sexually active group, the potential for opa gene variation, or that of any other surface exposed protein, would be reduced by the reduction in potential for a mixed gonococcal infection.
Within the clusters in Sheffield where the two clusters were supported over a long period, it appears that there was assortative or restricted sexual mixing, which reduced the likelihood of mixed infections occurring and therefore maintained the genotype. The sexual activities of individuals within the clusters must have been at a level where on average one sexual contact became infected with N. gonorrhoeae to allow the persistence of the two opa type strains. It is also conceivable that in vivo the cluster isolate strains were inhibited in some way from undergoing intergenic recombination on contact with DNA from gonococci of another genotype. This may be possible due to strain dominance, which has been suggested as a possible reason for the clonal nature of AHU IA-2 strains (8).
Investigation of opa gene variation in vitro and in relation to epidemiologically identified transmission links appears to indicate that variation in vivo occurs more frequently through intergenic recombination. In vivo intergenic recombination would occur during mixed gonococcal infections, which is related to the sexual activities of individuals and the prevalence of gonorrhea. Hence, the degree of genetic diversity exhibited by N. gonorrhoeae is a reflection of the level of sexual activity.
| ACKNOWLEDGMENTS |
|---|
This work was funded by The Wellcome Trust.
| FOOTNOTES |
|---|
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Antimicrob. Agents Chemother. | Clin. Microbiol. Rev. |
|---|---|
| Clin. Vaccine Immunol. | ALL ASM JOURNALS |
|---|