Design

A method for amplification and characterization of the porin protein (Por), the major outer membrane protein of N. gonorrhoeae, which uses PCR amplification and hybridization with oligonucleotide probes to identify sequences of the variable regions (VR), was adapted for use with non-culture-based clinical samples.  

Participants

Reference bacterial strains included four N. gonorrhoeae strains used as por VR typing controls and 13 other Neisseria species. Clinical samples from two studies were tested. Matched cervical swab and wick specimens from 20 women who were culture positive for N. gonorrhoeae presenting to Baltimore sexually transmitted disease clinics were tested. Primary cultures were available for 17 of these participants. The cervical swab and wick specimens were stored in phosphate buffered saline at –70^oC.Paired urine specimens, collected about 2 months apart, from 72 female sex workers participating in a study of STI control strategies in Madagascar, which were analyzed by ligase chain reaction (LCR) for N. gonorrhoeae, were tested. Thirty-one additional LCR-positive urine specimens from this study were tested. The urine specimens were stored uncentrifuged at –70^oC.

Description of Tests and Diagnostic Standard

Nested PCR primers (GCPorBFouter/GCPorBRouter and PIB Fpr/PIB Rpr) were designed to amplify both the PIA and PIB alleles of the N. gonorrhoeae porB gene. The gonococcal porB biotin labeled oligonucleotide probes matched coding regions for the variable regions of por B, including PIA VRs 1, 2, 3, 6, and 7, and PIB VRs 1, 3, 5, 6, and 7.At least 2 and up to 9 probes were used for each VR. The probes were highly specific for the sequences found in the variable regions of porB and did not hybridize to sequences that differed by more than 2 bp.Hybridization of amplicon and probe was performed on nylon membranes using a vacuum apparatus. Binding was visualized using avidin-peroxidase and a chemiluminescent substrate.  

To validate the typing method, urine from healthy volunteers was spiked with various dilutions of reference bacterial suspensions, including mixed suspensions of PIA and PIB control strains at ratios of PIA cfu to PIB cfu of 1:1, 10:1, 100:1, 1:100, and 1:10.Some of the urine specimens, both whole urine and the cell pellet, were subjected to a freeze-thaw cycle.Prior to PCR amplification, the experimental and clinical specimens were centrifuged and DNA was isolated using the High Pure DNA kit (Roche Diagnostics).  

Main Outcome Measures

The hybridization results for the individual VRs, referred to as the “VR type”, and the results for all VRs tested, the por VR type, for the experimental and clinical specimens were reported. A mixed por type was defined as hybridization of dissimilar probes at each of two or more individual VRs.  

Main Results

The sensitivity of the method was 1 to 10 copies of template DNA. The effect of a freeze-thaw cycle on the detection of N. gonorrhoeae spiked into urine specimens was to decrease the limit of detection by 10-fold for whole urine samples but not for urine pellets. Typing of either the PIA or PIB reference strain was achieved with 10 to 100 copies of template DNA in the presence of 100-fold excess of the other strain. The probes did not hybridize to porin DNA amplified from Neisseria spp. other than N. gonorrhoeae.

Typing results were obtained for all 20 N. gonorrhoeae culture positive cervical swab and wick specimens and the 17 N. gonorrhoeae isolates and included 3 PIA types and 15 PIB types. Distinctly different probes for the same VR region hybridized to porB DNA amplicons from the same patient for 8 of the 20 women, indicating that DNA from more than one N. gonorrhoeae strain was present in the sample. Multiple N. gonorrhoeae por types were found in the swab specimens but not wick specimens from 4 women, in the wick specimens but not the swab specimens of 4 women, and different types were found in the swab and wick specimens of 2 women. Of the 8 specimens with multiple types identified, 4 specimens had both a PIA and a PIB type and 4 specimens had more than one PIB type. Multiple por types were not detected in any of the culture isolates.   

Typing results were obtained for 60 of 126 N. gonorrhoeae LCR positive and 3 of 49 LCR negative urine specimens and included 10 PIA types and 10 PIB types. Multiple N. gonorrhoeae por types were found in 13 of the 63 typed specimens; 8 specimens had both a PIA and a PIB type, 1 specimen had a PIA and 2 different PIB types, and 4 samples had two PIA types.   

Authors’ Conclusions

The method described in this paper was useful for characterizing the gonococcal outer membrane Por protein from non-culture-based clinical samples. Infections with multiple por types (found in 40% and 22% of the women in the Baltimore and Madagascar studies, respectively) were easily identified in direct clinical specimens. Some por types and some VR types were common among the geographically diverse populations. The por typing method can be used in conjunction with NAAT to characterize strains in individuals who have been identified as having gonococcal infections. Genetic typing of the porin protein of N. gonorrhoeae may contribute to studies of host immunity, gonococcal epidemiology, and pathogenesis.    

Source of funding: A University of North Carolina Sexually Transmitted Diseases Cooperative Research Center grant and a National Institutes of Health grant

For correspondence: Margaret C. Bash, HFM-428, Division of Bacterial, Allergenic, and Parasitic Products, Office of Vaccines Research and Review, Center for Biologics Evaluation and Research, 1401 Rockville Pike, Rockville, MD 20852.E-mail address: bash@cber.fda.gov