Genitourin Med 1995;71:145-149

Resolution of clonal subgroups among Neisseriagonorrhoeae IB-2 and IB-6 serovars by pulsed-fieldgel electrophoresis

C L Poh, G K Loh, J W Tapsall

AbstractObjective-Analysis of macrorestrictionpatterns by PFGE to resolve the related-ness of clonal subgroups amongst Ngon-orrhoeae IB-2 and IB-6 serovar strains.Materials and methods-Nineteen IB-2and eight IB-6 serovar strains that dif-fered in either auxotype or penicillin sen-sitivity were isolated over a two and ahalf-year period from patients attendingseveral STD clinics in Sydney. Duringthis period, a major clone, Wt/IB-2 (FS),established on epidemiological grounds,was circulating amongst homosexualmales. The genetic relation of this majorclone to the other strains present in thecommunity was determined by pulsed-field gel electrophoretic (PFGE) analysisof DNA restriction fragments. GenomicDNA from the 27 isolates were prepared,digested with SpeI and BgIII and therestriction patterns were analysed bycontour-clamped homogeneous electricfield electrophoresis (CHEF) in a CHEFDRIII equipment.Results-Phenotypic characterisation ofthe 27 isolates by the combined use ofauxotype, serological characterisationand penicillin sensitivity indicated thepresence of subgroups within each of thetwo serovars. In the present study, PFGEanalysis of SPeI and BglII-generatedgenomic DNA restriction patterns fromsix ofthe ten Wt/IB-2 (FS) correlated wellwith phenotypic characterisation of thismajor clone. Four ofthe ten Wt/IB-2 (FS)were found to be clonally-derived vari-ants of this major clone as minor genomevariations (less than 3 DNA fragments)were observed. Distinct clones were rep-resented by three Wt/IB-2 (LS) isolates asthe DNA fingerprints generated fromthese were unrelated to the major clone.Analysis of PFGE patterns of 6 Pro/IB-2isolates showed that one was genotypi-cally identical to the major clone, twowere clonal variants and three had signif-icantly different patterns to indicate thatthey were genotypically unrelated.Wt/IB-6 isolates had heterogenous PFGEpatterns that were clearly unrelated tothe Wt/IB-2 serovar strains. Within theIB-6 serovar, there were three isolateswith the Wt/IB-6 (FS) phenotype thatcould be considered as clonal variantswhilst the rest were genotypically dis-tinct.Conclusions-PFGE analysis of

macrorestriction patterns generatedfrom SpeI- and BglII-cleavage ofgenomic DNA has enabled the establish-ment of clonal origins of strains presentin the Sydney community during theperiod of study. The delineation ofstrains belonging to major AIS groups byPFGE analysis presents a clearer epi-demiological picture than phenotypiccharacterisation alone.

(Genitourin Med 1995;71:145-149)

Keywords: Neisseria gonorrhoeae; clones; electro-phoresis

IntroductionGonorrhea is a major sexually transmitted dis-ease and is prevalent in both developed anddeveloping countries. Precise characterisationof N gonorrhoeae can provide valuable infor-mation concerning gonococcal strain popula-tions in any community, the temporal changesas well as the emergence and spread of anti-biotic resistant strains.' In the absence of avaccine, data such as these will be useful fordevising effective control measures for gono-coccal infections.

Currently, the most widely employedmethod for differentiation of N gonorrhoeaestrains is one based on auxotyping and sero-logical (A/S) characterisation2 with serotypingusing monoclonal antibodies to protein Iproviding far better discrimination thanauxotyping.3' However, characterisation ofstrains by serotyping can encounter problemssuch as non-typability of strains, availability,batch to batch variation of monoclonal anti-body reagents and the reproducibility of coag-glutination reactions, especially when usingmonoclonal antibodies such as 2D6, 2G2 and6D9.5We have recently developed pulsed-field

gel electrophoresis (PFGE) for subtypingN gonorrhoeae strains expressing identicalauxotype/serovar characteristics and this tech-nique appears promising for epidemiologicalinvestigations.6 In Sydney, Australia there hasbeen an outbreak of gonorrhoea in homosex-ual men with isolates of the IB-2 serovar andof the non-requiring (Wt) auxotype. An addi-tional phenotypic marker, full sensitivity topenicillin (FS), unusual in Sydney isolates,was also used to characterise these strains inan investigation of this outbreak.7 Otherisolates of the IB-2 serovar which differed inauxotype and/or penicillin sensitivity werealso isolated over the same period.

Department ofMicrobiology, FacultyofMedicine, NationalUniversity ofSingapore, KentRidge, Singapore 0511C L PohG K LohDepartment ofMicrobiology, PrinceofWales Hospital,Randwick, Sydney,Australia 2031JW TapsallCorrespondence to:Dr C L Poh.Accepted for publication19 December 1994

145

on February 29, 2020 by guest. P

rotected by copyright.http://sti.bm

j.com/

Genitourin M

ed: first published as 10.1136/sti.71.3.145 on 1 June 1995. Dow

nloaded from

Poh, Loh, Tapsall

Additionally, strains of the IB-6 serovar, alsoWt and FS were isolated. The serovars IB-6and IB-2 are differentiated solely on differentreactions with the 2D6 monoclonal antibody.

This study assessed the ability of the PFGEmethod to discriminate the phenotype Wt/IB-2 (FS) originally identified on epidemiologicalgrounds, its capacity to differentiate subtypeswithin the IB-2 serovar and to distinguish thephenotype Wt/IB-2/FS from Wt/IB-6/FS.

Materials and methodsBacterial strains and culture conditionsTwenty-seven clinical isolates of N gonor-rhoeae with their respective auxotype/serovarcharacteristics and penicillin susceptibilityphenotypes were listed in tables 1 and 2. Theisolates comprised ten selected Wt/IB-2 (FS)strains identified as being epidemiologicallylinked and nine other strains of the IB-2serovar (table 1) and 8 of the IB-6 serovar(table 2) with a different penicillin sensitivityor auxotype or both. Strains were grown onmodified Thayer-Martin agar (BBL, BectonDickinson Microbiology Systems, Cockeys-ville, USA) and incubated at 37°C in the pres-ence of 5% CO2 for 20 h.

Serological characterisationStrains were serotyped using monoclonalsmanufactured by Syva (Palo Alto, Calif.,USA) according to the nomenclature ofKnapp et a1 with six MAbs specific for pro-tein IA (PrIA) and six specific for protein IB(Pr IB).

Antimicrobial susceptibilityThe MIC of penicillin for the isolates weredetermined by the agar dilution method.9Isolates were classified as fully sensitive (FS)when the MIC was less than 003 mg/l, lesssensitive (LS) when the MIC ranged from0-06 to 0.5 mg/l and resistant (R) when theMIC was found to be greater than 1 mg/l.

Genomic DNA preparation and digestionGenomic DNA was prepared as describedpreviously.6 Plugs containing embedded DNA

Table 1 Phenotypic and genotypic characteristics ofN.gonorrhoeae IB-2 serovar strains

PFGE patternPenicillin

Strain No. Auxotype sensitivity SpeI Bg/II

92/049 WTr FS 1 192/133 WrT FS 1 292/213 WRT FS 1 192/215 WT FS 1 392/336 WT FS 1 292/038 W.T FS 1 292/239 WTr FS 1 192/386 WTr ES 1 192/413 WTr FS 1 192/320 WJT FS 1 191/238 WIT LS 2 492/290 Wrr LS 3 592/411 Wrr LS 3 692/142 PRO LS 3 792/279 PRO R 4 892/221 PRO R 5 892/053 PRO LS 1 191/302 PRO FS 1 292/421 PRO FS 1 2

were digested overnight with 20 U of SpeI orBglII in buffers as recommended by the manu-facturer (New England Biolabs, Beverly,Massachusetts, USA).

Pulsed-field gel electrophoresisPlugs containing digested DNA were loadedinto a 1% agarose gel and electrophoresed in acontour-clamped homogeneous electric field(CHEF DRIII) apparatus with a hexagonalelectrode array (BioRad, Richmond,California, USA) at 14°C. Pulse time wasramped from 1 to 15 s for 8 h and then 15 to25 s for 16 h following SpeI digestion and 1 sto 15 s for 22 h following BglII digestion. Gelswere stained with ethidium bromide and pho-tographed under UV transillumination.Rhodobacter sphaeroides 2.4.1. genomic DNAdigested by AseI was used as the molecularweight standard.'0

ResultsAuxotype/serovar and penicillin susceptibilitypatternsSerological characterisation of the 27 isolatesestablished two homogeneous groups.Nineteen of the isolates were found to belongto the IB-2 serovar and eight were of IB-6.The 19 IB-2 serovar strains were classifiedinto 2 auxotypes (wild-type and proline-requiring) and exhibited 3 different patternsof penicillin susceptibility. Twelve of the 19strains were fully sensitive to penicillin, 5 wereless sensitive and 2 were resistant (table 1).Similarly, two auxotypes were evidentamongst 8 IB-6 serovar strains and 4 werefully sensitive to penicillin, 3 less sensitive andone was resistant (table 2).

Genome fingerprints ofIB-2 and IB-6 serovarsWhen SpeI-digested preparations of genomicDNA from the 19 IB-2 serovar strains wereexamined, 12-17 fragments ranging from 2 to410 kb were observed. BglII digestion of thesame isolates yielded 17-24 fragments ofbetween 2 to 260 kb. Five different SpeI pat-terns were found among the 19 IB-2 strains(table 1) whilst digestion with BglII producedeight different patterns (table 2). Strains withSpeI type 3, 4 and 5 patterns were found todiffer from strains with type 1 pattern by up tonine fragments. Two of the four strains withidentical SpeI restriction patterns (fig 1, lanesC & E) could be further distinguished by BglIIdigestion (fig 2, lanes C & E). The group of

Table 2 Phenotypic and genotypic characteristics ofN.gonorrhoeae IB-6 serovar strains

PFGEpatternPenicillin

Strain No. Auxotype sensitivity SpeI BglII

90/125 WTr LS 6 991/347 WTr LS 7 1092/208 WT FS 8 1191/227 WJT FS 9 1291/236 WT FS 9 1391/384 WTr FS 10 1291/275 WvT1 R 11 1491/378 PRO LS 12 15

146 on F

ebruary 29, 2020 by guest. Protected by copyright.

http://sti.bmj.com

/G

enitourin Med: first published as 10.1136/sti.71.3.145 on 1 June 1995. D

ownloaded from

Resolution of clonal subgroups among Neisseria gonorrhoeae IB-2 and IB-6 serovars by pulsed-field gel electrophoresis

Figure 1 SpeIgenomemacrorestriction fragmentsgeneratedfrom N.gonorrhoeae Wt/IB-2serovar strains (lanenumbers B-J correspond tostrains 049, 133, 213,215, 238, 290, 142, 279and 221). Lane A: mol.wt marker (R. sphaeroides2.4.1. digested with AseI).

KB

910-410--

3328=--

275-244-214-

110-97-73--63-

31--18-

Figure 2 Bg/IIgenomemacrorestriction fragmentsgeneratedfrom N.gonorrhoeae WtIIB-2serovar strains (lanenumbers B-J correspond ofstrains 049, 133, 213,215, 238, 290, 142, 279and 221). Lane A: mol.wt marker (R. sphaeroides2.4.1. digested with AseI).

910410

275244214

1107363

31

Figure 3 SpeIgenomemacrorestriction fragmentsgeneratedfrom N.gonorrhoeae Wt/IB-6serovar strains (lanenumbers B-J correspond tostrains 125, 347, 208,227, 236, 336, 384, 275and 378). Lane A: mol.wt marker (R. sphaeroides2.4.1. digested with AseI).

A RB c n F F H

KB

910-

410-

360-

275-244--214-

110-'97-

73--63-

19 IB-2 serovar strains were found with SpeItype 1 pattern. Digestion with BglII furtherdifferentiated the 19 IB-2 strains into sevenwith type 1, five with type 2, one each withtype 3, 4, 5, 6 and 7, and two with type 8restriction patterns (table 1). The group offive strains with SpeI type 1 BglII type 2 pat-tern (S1B2) (for example, strain 133; fig 2,lane C) differed from the group consisting ofseven strains with SpeI type 1 BglII type 1 pat-tern (SIBl) (for example, strain 049, fig 2,lane B) in lacking a large DNA fragment ataround 259 kb, and the presence of an addi-tional DNA fragment of 195 kb (fig 2, com-pare lanes B and C). Strain 215 which wasassigned with SpeI type 1 BglII type 3 pattern(S1B3) was found to differ from the group ofseven strains with S1B 1 pattern by theabsence of a single DNA fragment of 229 kb(fig 2, lane E) and it also presented an addi-tional fragment of 195 kb which was similarlydetected in the group with S1B2 pattern.Strains with BglII type 4, 5, 6, 7 and 8 pat-terns showed significant DNA fragment dif-ferences when compared with strains showingtype 1, 2 or 3 patterns. There were only twoto three band differences amongst strains witheither BglII type 1, 2 or 3 patterns but up to 7band differences were observed between thiscluster of strains and strains displaying eithertype 4, 5, 6, 7 or 8 BglII restriction pattern.

Digestion of genomic DNAs of the eightIB-6 strains by either SpeI or BglII yieldedseven different patterns. The eight IB-6strains had SpeI digestion patterns which wereclearly different from the 19 IB-2 serovarstrains (fig 3). BglII digestion again allowedfurther differentiation of the pair (strains 227and 236) with identical SpeI type 9 pattern(fig 3, E & F) into two closely related BglIIpatterns which differed by 3 band differences(data not shown).

Correlation ofPFGE data with auxotype andpenicillin sensitivityIn the 10 Wt/IB-2 (FS) isolates examined,three different PFGE patterns were found. Sixof the ten isolates shared the same SiB1 pat-tern, three with S1B2 and one had a S1B3pattern. Three isolates with the non-requiringauxotype (Wt) were less sensitive to penicillinand the PFGE pattern of each isolate was dis-tinct and different from the 10 Wt/IB-2 (FS)isolates. The group of 6 Pro/IB-2 (LS) isolatesexhibited three different patterns of penicillinsensitivity. A single Pro/IB-2 (LS) isolate withits SiB1 profile was identical in genomemacrorestriction pattern to 6 Wt/IB-2 (FS)isolates (table 1). Two Pro/IB-2 (FS) isolateswith S1B2 profiles were found to be identicalto three of the Wt/IB-2 (FS) strains (table 1).The remaining three Pro/IB-2 isolates, two ofwhich were resistant to penicillin (92/279 and92/221) (fig 2, lanes I and J) and one had theLS penicillin phenotype (92/142) (fig 2, laneH), had three different PFGE patterns whichwere clearly unrelated to any of the Wt/IB-2(FS) strains (fig 2, lanes B, C and E) orWt/IB-2 (LS) strains (fig 2, lanes F and G).The 8 IB-6 isolates had heterogeneous

147 on F

ebruary 29, 2020 by guest. Protected by copyright.

http://sti.bmj.com

/G

enitourin Med: first published as 10.1136/sti.71.3.145 on 1 June 1995. D

ownloaded from

Poh, Loh, Tapsall

patterns that were clearly unrelated to anyof the IB-2 serovar isolates. Of the twoWt/IB-6 (IS) isolates, one (90/125) was iso-lated in 1990 and the other (91/347) in 1991.Both were characterised by two differentPFGE patterns (fig 3, lanes B and C).Phenotypically, the major group of IB-6 iso-lates was comprised of four that exhibited thenon-requiring auxotype (Wt) and were fullysensitive to penicillin. However, with theexception of isolates 227 and 236, PFGE pat-terns of the group indicated non-homogeneityin genomic composition (fig 3, lanes D to G).Isolate 92/208 (fig 3, lane D) which wasWt/FS had a PFGE pattern that was uniqueand different from all the Wt/IB-6 (FS) iso-lates. One of the Wt/IB-6 isolates examinedwas resistant to penicillin and examination ofthe PFGE pattern of this isolate (91/275) bySpeI and BglII digestion clearly establishedthat this strain was very different from the restof the Wt/IB-6 (FS) isolates. Another isolatethat had a PFGE pattern that was unique toitself and clearly different from all the Wt/IB-6isolates was isolate 91/378 (fig 3, lane J).Besides having a very different PFGE S12B 15profile, it also had a proline auxotype thatdifferentiated this isolate from the rest of theIB-6 serovars.

DiscussionPrecise characterisation of N gonorrhoeaestrains is essential for monitoring and control-ling spread of gonococcal infections. Currenttyping methods adopted in most referencelaboratories are based on a combination ofphenotypic characteristics such as the auxo-types and serotypes of strains. In strains thatcarry plasmids, differentiation of strainsbelonging to the same auxotype-serovar (A/S)groups could be achieved. However, plasmidprofiling is of limited value when a commonplasmid or a common combination of plas-mids are present." In some instances, differ-entiation of strains present within a commonA/S group could be enhanced by including anadditional phenotypic trait. As was shownrecently by Rowbottom et al,7 an additionalphenotypic marker such as sensitivity to peni-cillin was found to be useful for subtypingWt/IB-2 isolates and established the presenceof a common epidemic clone, Wt/IB-2 thatwas fully sensitive (FS) to penicillin in Sydneyover a period of 2j years.7 The isolation of theWt/IB-2 serovar strains that were less sensitive(LS) to penicillin and Pro/IB-2 strains thatwere either FS, LS or R during the sameperiod indicated that there is a shortcoming inthe use of the A/S scheme alone to subtypestrains. However, sensitivity or resistance topenicillin is a phenotypic trait and such alter-ations in antimicrobial susceptibility withouta genetic basis would not be effective indefining clonal subgroups present within apopulation.PFGE of genomic DNA of microorganisms

following cleavage with restriction enzymesyields DNA fingerprints of every strain. Thegenomic fingerprints readily provide an

assessment of inter-strain relationships.Strains arising from different clonal originswill produce unrelated genome fingerprints.In microorganisms such as Pseudomonas aerug-inosa, it has been found that clonally derivedvariants could diverge from a major commonrestriction pattern by up to six DNA frag-ments.'2 It was also established in the samestudy that 56% of changes in antimicrobialsusceptibilities of sequential isolates of thesame strain were due to variations in chromo-some structure and 25% were occurring in theabsence of any detectable change in genomicfingerprints. For organisms that can undergofrequent genome rearrangements either invivo or be induced by externally introducedgenetic elements, clonal variants that differedby a few DNA fragments in genome macro-restriction patterns can be expected.

In the present study, PFGE analysis ofmacrorestriction patterns resulting from SpeIand BglII digestion was found to correlatewith phenotypic characterisation of the pre-dominant Wt/IB-2 (FS) clone by the A/Sclassification scheme and antimicrobial sus-ceptibility pattern analysis. In addition, it wasable to establish the presence of clonal sub-groups within the IB-2 serovar. Thus, threeclonal variants with S1B2 and one with S1B3PFGE pattern could have arisen throughsome minor genomic mutations or re-arrange-ments of the Wt/IB-2 (FS) major clone with-out alteration of nutritional genetic markersand antibiotic susceptibility patterns. Themutational sites altered in these clonal vari-ants would most likely be the recognition sitesof the restriction endonucleases. However,two other Pro/IB-2 (FS) clonal variants withsimilar S1B2 and one Pro/IB-2 (LS) withSiBi PFGE patterns were found. These vari-ants may have arisen through mutations orgenome rearrangements affecting only theproline biosynthetic pathway genes or the fi-lactamase gene locus. Alternatively, mutantsthat were less sensitive to penicillin could alsoarise from other mechanisms such asdecreased drug uptake, alteration of outermembrane permeability and modified gyraseactivity.'3 The three Wt/IB-2 (LS) strains hadPFGE patterns differing by more than sixbands from the common major clone andwere concluded to be unrelated in genomiccomposition to the major clone. ThreePro/IB-2 strains (two were resistant to peni-cillin and one was less sensitive) with PFGEpatterns that were significantly different fromthe major clone would have been of differentclonal origins.The genomic contents of strains belonging

to IB-6 serovar as revealed by PFGE analysiswere very different from the group of IB-2serovar strains. Within the IB-6 serovar, therewere three strains with the Wt/IB-6 (FS)phenotype that could be considered as vari-ants derived from the same clone, one had asignificantly different PFGE pattern and wasconsidered to be of different clonal origin.Three other Wt/IB-6 strains with differingpenicillin sensitivity and one Pro/IB-6 (LS)strain also had PFGE patterns that were

148

on February 29, 2020 by guest. P

rotected by copyright.http://sti.bm

j.com/

Genitourin M

ed: first published as 10.1136/sti.71.3.145 on 1 June 1995. Dow

nloaded from

Resolution of clbnal subgroups among Neisseria gonorrhoeae IB-2 and IB-6 serovars by pulsed-f eld gel elec"rophoresis

unique and unrelated to the major group ofclonal variants.

In conclusion, PFGE analysis of macro-restriction patterns generated from cleavage ofgenomic DNAs ofN gonorrhoeae strains withSpeI and BglII correlated well with groupingof strains into either IB-2 or IB-6 serovarswhich was based on reaction differenceobserved with the 2D6 monoclonal antibody.The establishment of clonal variants and theidentification of distinct clones within each ofthe two serovars by PFGE analysis wouldfacilitate better monitoring of gonococcaltransmission in a community. Additionalprospective investigations of isolates derivedfrom cohorts and serial isolates from the samepatient will provide further insight intogenome variability and the changes in drugresistance observed.

This work was supported by NUS research grant RP 910459to C L Poh. The authors gratefully acknowledge the supply ofmonoclonal antibody reagents by Dr J R L Forsyth of theUniversity ofMelbourne.

1 Sarafian SK, Knapp JS. Molecular epidemiology of gonor-rhoea. Clin Microbiol Rev 1989;2(s):49-55.

2 Kohl PK, Ison CA, Danielsson D, et al. Current status ofserotyping of Neisseria gonorrhoeae. Eur Epidemtol1990;6:91-5.

3 Knapp JS, Sandstrom EG, Holmes KK. Overview of epi-demiological and clinical applications of auxotype/serovar classification of Neisseria gonorrhoeae. In:Schoolnik GK, Brooks GF, Falkow S, Frasch CE,Knapp JS, McCutchan JA, Morse SA, eds. ThePathogenic Neisseria, 1985, 6-12 Washington, D.C., AmSoc Microbiol Press.

4 Poh CL, Ocampo JC, Sng EH, Bygdeman SM.Characterisation of PPNG and non-PPNG Neissenagonorrhoeae isolates from Singapore. Genitourin Med1991;67;389-93.

5 Gill MJ. Serotyping Neisseria gonorrhoeae: a report of thefourth international workshop. Genitourin Med 1991;67:53-7.

6 Poh CL, Lau QC. Subtyping of Neisseria gonorrhoeae auxo-type-serovar groups by pulsed-field gel electrophoresis.JMedMicrobiol 1993;38:366-70.

7 Rowbottom JH, Tapsall JW, Plummer DC, et al. An out-break of a penicillin-sensitive strain of gonorrhoea inSydney men. Genitourin Med 1994;70:196-9.

8 Knapp JS, Tam MR, Nowinski RC, et al. Serologicalclassification of Neisseria gonorrhoeae with use of mono-clonal antibodies to gonococcal outer membrane proteinI. J Infect Dis 1984;150:44-8.

9 Tapsall JW, Phillips EA, Cossins YM, et al. Penicillin sen-sitivity of gonococci isolated in Australia: developmentof the Australian Gonococcal Surveillance Programme.BrJ Venereal Dis 1984;60:226-30.

10 Suwanto A, Kaplan S. Physical and genetic mapping of theRhodobacter sphaeroides 2.4.1. genome: genome size,fragment identification, and gene localization. J Bacteriol1989;17:5840-9.

11 Dasi MA, Nogueira JM, Camarena JJ, et al. Genomic fin-gerprinting of penicillinase-producing strains of Neisseriagonorrhoeae in Valencia, Spain. Genitourin Med 1992;68:170-3.

12 Struelens MJ, Schwam V, Deplano A, Baran D. Genomemacrorestriction analysis of diversity and variability ofPseudomons aeruginosa strains infecting cystic fibrosispatients. J Clin Microbiol 1993;31:2320-6.

13 Bryan LE. Two forms of antimicrobial resistance: bacterialpersistence and positive function resistance. JAntimicrobChemother 1989;23:817-23.

149

on February 29, 2020 by guest. P

rotected by copyright.http://sti.bm

j.com/

Genitourin M

ed: first published as 10.1136/sti.71.3.145 on 1 June 1995. Dow

nloaded from