J. clin. Path., 1976, 29, 537-542

Isolation of oxidase-positive Gram-negative coccinot belonging to the genus Neisseria from theurogenital tractD. J. PLATT AND J. J. S. SNELL

From the Department of Medical Microbiology, Charing Cross Hospital, Fulham Palace Road,London W6 8RF and the National Collection of Type Cultures, Central Public Health Laboratory,Colindale NW9 SHT

SYNOPSIS In a 12-month period, oxidase-positive, Gram-negative cocci showing similar character-istics in biochemical tests have been isolated from the urogenital tract of 39 male and female patients.Although these organisms superficially resemble Neisseria gonorrhoeae, biochemical characteriza-tion and the results of DNA base composition analysis indicate that they do not belong to the genusNeisseria. The relationship of these organisms to the genera Neisseria, Achromobacter, and Pseudo-monas is discussed.

The isolation of Neisseria spp other than Neisseriagonorrhoeae from the urogenital tract has beenreported (Wilkinson, 1952). More recently, Jephcottand Morton (1972) have reported the isolation ofN. lactamicus from a genital site, and N. meningitidishas been similarly found in both symptomatic(Beck et al, 1974) and asymptomatic patients (Volkand Kraus, 1973).The isolation of these organisms from a small but

significant proportion of the patients attendingspecial clinics can lead to difficulty in the diagnosisof gonorrhoea.

Further complications may arise from the observa-tion, in direct smears, of organisms morphologicallyresembling gonococci, which on culture prove to becoccoid forms of Gram-negative bacilli.

This communication records the isolation ofoxidase-positive, Gram-negative cocci, showingsimilar characteristics in biochemical tests, from theurogenital tract of 39 patients attending the Depart-ment of Genitourinary Medicine at this hospitalover a period of 12 months. These organisms grewprofusely on nutrient agar and formed mucoidcolonies which developed a pink pigment. Furtherstudy showed that these organisms differ in otherfundamental characteristics from members of thegenus Neisseria.

Material and Methods

COLLECTION OF CLINICAL MATERIALSwabs of material from the urethra, cervix, andReceived for publication 28 October 1975

rectum were inoculated on to a modified ThayerMartin medium (Riddell and Buck, 1970) and in-cubated in candle extinction jars at 37°C for 48hours.

BACTERIOLOGICAL PROCEDURESOxidase-positive colonies were examined in thinfilms by Gram stain and also with a fluoresceinconjugated antigonococcal globulin (Difco)routinely used in this department for the confirma-tion of the identity of N. gonorrhoeae as previouslydescribed by Beck et al (1974). The ability to pro-duce acid from glucose, maltose, sucrose, fructose,and lactose was examined using serum agar sugars.Antibiotic sensitivities were determined by the disctechnique using Direct Sensitivity Agar (Oxoid)containing 5%% lysed horse blood.

Biochemical tests were performed using themethods and media described by Cowan and Steel(1965), Snell et al (1972), and Owen and Snell (1973).Pigment was extracted from a 72-hour culture on

nutrient agar using a mixture of diethyl ether:methanol (9:1). The absorption maxima of theextract were measured using a Unicam SP800 re-cording spectrophotometer.

Inclusions of ferrous or ferric iron were soughtusing the Tirmann-Schmelzer's Turnbull, and Perlreactions respectively (Culling, 1974).

DEOXYRIBONUCLEIC ACID BASECOMPOSITIONDNA was extracted by the method of Marmur(1961), and base compositions were estimated by the

537

on March 22, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.29.6.537 on 1 June 1976. Dow

nloaded from

D. J. Platt and J. J. S. Snell

melting temperature (Tm) method of Marmur andDoty (1962). Tm determinations were made in 0-1standard saline citrate buffer. The equation relatingTm to the percentage of guanine and cytosine in thetotal DNA (% GC) was derived from the generalequation of Owen et al (1969) which for 0-1 SSCreduces to: GC = 2-087 (Tm - 175.95) + 260-34.

ELECTRON MICROSCOPYTwenty-four and 48-hour cultures were examined bynegative staining (with 2% phosphotungstic acid atpH 6 8) and in section (following glutaraldehydefixation, post-fixation with osmium tetroxide, andstaining with uranyl acetate/lead citrate) using aPhillips EM200.

Results

DISTRIBUTION OF ISOLATESTable I shows the distribution among the 39 patients,by sex and site, of the isolates.

Site of Isolation Male Female Totaln = 19 n = 20 n = 39

Urethra 14 10 24Cervix only - 5 5Urethra and cervix - 5 5Rectum 5 - 5Total 19 20 39

Table I Distribution of isolates

Presenting symptoms among the 39 patients were:

5 patients with symptoms attributable to tricho-moniasis or candidiasis (confirmed by culture);10 patients from whom no established pathogen wasisolated (clinically non-specific genital infection);24 asymptomatic patients (attending either as sexualcontacts of patients with established clinical condi-tions or for routine/post-treatment check-ups). N.gonorrhoeae had been isolated from nine of these24 patients less than one month before detection ofthis organism.

This organism was recovered on two occasionsfrom five of the 39 patients, and on three separateoccasions from one of the patients, the longestperiod observed between isolations being 7j months.Isolations were obtained from 10 of the 39 patientsafter their first visit to this clinic. Direct microscopyshowed the presence of extracellular Gram-negativediplococci on two occasions.

CHARACTERISTICS OF ORGANISMSThe results of the characterization tests on 10 of thestrains studied in detail are shown in table II. Gramstained preparation of cultures showed the cells tobe Gram-negative cocci. The average diameter ofthe cells was 1 /t, although larger forms (up to 3 ,u)were often present. The cells were arranged singly, inpairs, and occasionally in clusters of up to ten.

After overnight incubation at 37'C on nutrient

Tests PositiveGrowth nutrient agarGrowth at 22°CGrowth at 30'CGrowth at 37'CCatalase productionOxidase productionGrowth on MacConkey's agar ()Anaerobic growth (blood agar) (±)

Tests NegativeCitrate utilization (Koser's)Indole productionMethyl red testAcetylmethyl carbinol productionNitrite reductionGelatin liquefactionSerum liquefactionO/F test (alkaline reaction)Acid production from' glucose, lactose, sucrose, maltoseGluconate oxidationMalonate utilizationArginine desimidase

Tests VariableHydrogen sulphide productionGrowth at 45'CTween 20 hydrolysisCitrate utilization (Christensen's)Tween 40 hydrolysis

Urease productionPhenylalanine deamination2Poly-beta-hydroxybutyrate

inclusion granulesGrowth on 2 sodium chlorideGrowth on 4 sodium chloride

Ornithine decarboxylaseLysine decarboxylaseGrowth at S°CLecithinase productionAesculin hydrolysisStarch hydrolysisCasein hydrolysisPhosphatase productionDeoxyribonuclease productionTween 80 hydrolysisGrowth on 6 '/. sodium chlorideNitrate reduction

I (±) strain positiveI (±) strain positive7 strains positive3 (i) strains positive1 strain positive

Table II Biochemical characters of 10 strains'(i) indicates poor growth or a weak positive reaction.'Phenylalanine deamination: method of Snell and Davey (1971).3Carbohydrates tested in serum agar sugars, peptone water sugars, and ammonium salt sugars. Glucose also tested in Hugh and Leifson's mediumand lactose in 10% lactose agar.

538

on March 22, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.29.6.537 on 1 June 1976. Dow

nloaded from

Isolation of oxidase-positive Gram-negative cocci not belonging to the genus Neisseria

agar the strains formed convex, translucent, greyishmucoid colonies, 0-5-2 mm in diameter. The coloniesdeveloped a pale coral pink pigmentation which was

scarcely noticeable after 24 hours' incubation butwhich increased in intensity after exposure to lightfor several days. No haemolysis was observed on 5 %horse blood agar. All isolates were capsulated.Motility was not demonstrated.The isolates gave variable and poorly reproducible

results on staining with fluorescein-labelled anti-gonococcal globulin. Eleven of the 45 strains ex-

hibited an intensity of fluorescence at isolation asbright as N. gonorrhoeae. Sixteen isolates were ex-amined in triplicate by this technique. Twelve gavestrong fluorescence on at least one occasion, but noisolate gave strong fluorescence on three occasions.These results suggest non-specific fluorescence al-though we are unable to explain the degree ofvariability observed. Triplicate controls of N.gonorrhoeae and N. perflava showed strong and zero

fluorescence respectively on each occasion.Of 12 isolates, all were sensitive to streptomycin

(25 ,ug/disc), kanamycin (30 ,ug/disc), gentamicin(5 ,Lg/disc), and tetracycline (10 ,ug/disc); all were

resistant to sulphonamide (500 ,ug/disc), co-trimoxa-zole (25 ,ug/disc), penicillin (1 5 ,ug/disc), ampicillin(2 ,ug/disc), cephaloridine (25 ,ug/disc), colistin (10

pg/disc), and nitrofurantoin (200 ,ug/disc). Ninestrains were sensitive and three resistant to nalidixicacid (30 ,ug/disc).

DNA BASE COMPOSITIONSThe Tms of the two strains examined were 83 9°Cand 84-1°C. These correspond to 68-2 and 68-7%GC.

ELECTRON MICROSCOPYElectron microscopy confirmed the coccoid shape ofthe cells (fig 1). Figure 2 shows a group of severalcells enclosed in what appears to be a commoncapsule. Figure 3 suggests the presence of a triplelayered cell wall structure which agrees with theGram-negative staining reaction of the cells.

PIGMENT CHARACTERIZATIONExtracts of the pigments showed absorption maximaat 465, 490, and 525 nm. This suggests a carotenoidtype of pigment (Pfennig, 1969).

STAIN FOR IRON INCLUSION GRANULESIron inclusion granules were not conclusivelydemonstrated. In no case was a stronger reactionobtained than with an Escherichia coli negativecontrol.

Fig 1 Coccal shapeof the cellsdemonstrated bynegative staining(x 4600).

539

on March 22, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.29.6.537 on 1 June 1976. Dow

nloaded from

Fig 2 Groups of cellsapparently enclosed in acommon capsule (x 100 000.)

Fig 3 Ultrathinsection showing detailof the cell wall,confirming the Gram-negative nature of thisorganism (x 20000).

x 0:!::.

on March 22, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.29.6.537 on 1 June 1976. Dow

nloaded from

Isolation of oxidase-positive Gram-negative cocci not belonging to the genus Neisseria

Discussion

The possibility cannot be excluded that the isolationof these organisms represents a cross infectionepisode. Attempts to recover this organism frommaterials used in the examination of the patients andthe examination room environment were repeatedlyunsuccessful. The recovery of strains from 10patients on their first attendance and the observa-tion of extracellular Gram-negative cocci in thedirect smear of one of these weigh against thishypothesis.We are unable definitely to assign these strains to a

genus or species. Initially we considered that thesestrains might be classified in the genus Micrococcus.However, we rejected this possibility on the basis ofwhat we consider to be the Gram-negative nature ofour strains.These were always Gram-negative with no

evidence of retention of crystal violet even in theearly stages of growth. Although some strains ofmicrococci may appear Gram-variable after pro-longed incubation, we have never observed completeremoval of crystal violet from all the cells.Although the evidence from the electron micro-

scope studies is not unequivocal, the fine structureappears more characteristic of a Gram-negative thana Gram-positive cell wall. Further evidence for theGram-negative nature of our strains was obtainedby comparing representative strains with somecollection strains of Micrococcus roseus, includingone highly mucoid strain. The findings can besummarized as follows:The strains of M. roseus showed slight or no lysis

with 0-5 % sodium dodecyl sulphate, whereas ourstrains were rapidly and completely lysed. Growthof the M. roseus strains was inhibited by lysozyme,whereas our strains were resistant to 1 mg oflysozyme. Strains of M. roseus showed little lysisafter prolonged sonication, whereas our strains were

completely lysed after only 30 seconds' sonication.A relationship to the genera Acinetobacter or

Neisseria is excluded by the DNA base compositionof these isolates (69 % GC compared with 39-47 %GC for Acinetobacter and 47-52% GC for Neisseria).Our strains appear similar to those described byCourtois et al (1954), which they named Neisseriacapsulata. However, the description of this species isnot sufficiently detailed to allow conclusive com-parison, and as far as we are aware the systematicposition of this species has not subsequently beendetermined.The arrangement of cells in what appear to be

common capsules (fig 2) suggested Siderocapsa as apossible genus. However, the description of thisgenus in the 8th edition of Bergey's manual

(Buchanan and Gibbons, 1974) is incomplete, and wewere unable to obtain any authentic strains of thisgenus for comparison. Our failure to demonstrateiron inclusion granules weighs against assignmentto this genus.The biochemical test results, DNA base composi-

tions, and antibiotic sensitivity patterns of thesestrains suggest either Pseudomonas or Achromobacteras suitable genera for their inclusion. Our strainsdiffer from the generic definition of Pseudomonas inbeing coccoid and non motile. Hendrie et al (1974)have recently proposed the rejection of the genericname Achromobacter. The genus Alcaligenes may bea suitable genus for some species previously classi-fied as Achromobacter. However, many of the speciesin both these genera are poorly defined. Hendrie etal (1974) further considered that absence of motilityshould exclude strains from these genera. Wetherefore conclude that the systematic position ofthese strains should be left open until other workershave had the opportunity to examine our strains.

There is no evidence to suggest that this organismwas playing a pathogenic role in any of the casesdescribed here, and the significance of isolation ofthis organism from the urogenital tract lies in thesuperficial resemblance to the gonococcus, con-stituting a possible source of error in the diagnosisof this disease. Tests useful for the differentiation ofthis organism from N. gonorrhoeae are shown intable III. For practical purposes, profuse growths onnutrient agar and at 22°C form useful differentialdiagnostic tests.

Described N. gonorrhoeaeStrains

Growth on nutrient agar +Growth at 22°C +Growth on MacConkey agar +Urease production +Acid produced from glucose - +Pink pigment +Accumulation of poly-B-hydroxy-

butyrate granules +

Table III Tests differentiating between the describedstrains and N. gonorrhoeae

Three of these strains have been deposited in theNational Collection of Type Cultures and allottedthe following catalogue numbers-NCITC 10991,NCTC 10992, and NCTC 10993.

We should like to thank Dr J. L. Fluker for access tothe clinical reports. DJP is in receipt of a grant fromthe Clinical Research Committee of the CharingCross Hospital.

541

on March 22, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.29.6.537 on 1 June 1976. Dow

nloaded from

D. J. Platt and J. J. S. Snell

References

Beck, A., Fluker, J. L., and Platt, D. J. (1974). Neisseriameningitidis in urogenital infection. Brit. J. vener. Dis., 50,367-369.

Buchanan, R. E. and Gibbons, N. E., Eds. (1974). Bergey'sManual of Determinative Bacteriology, 8th edition.Williams and Wilkins, Baltimore.

Courtois, G., Thys, A., and Verselder, R. (1954). Unenouvelle Neisseria, agent causal de meningite. Ann. Soc.belge Med. trop., 34, 13-20.

Cowan, S. T. and Steel, K. J. (1965). Manualfor the Identifica-tion of Medical Bacteria. University Press, Cambridge.

Culling, C. F. A. (1974). Handbook of Histopathological andHistochemical Techniques, 3rd edition. Butterworth,London.

Hendrie, M. S., Holding, A. J., and Shewan, J. M. (1974).Emended descriptions of the genus Alcaligenes and ofAlcaligenes faecalis and proposal that the generic nameAchromobacter be rejected: status of the named species ofAlcaligenes and Achromobacter: request for an opinion.Int. J. system. Bact., 24, 534-550.

Jephcott, A. E. and Morton, R. S. (1972). Isolation ofNeisseria lactamicus from a genital site. Lancet, 2, 739-740.

Marmur, J. (1961). A procedure for the isolation of deoxy-ribonucleic acid from micro-organisms. J. molec. Biol., 3,208-218.

Marmur, J. and Doty, P. (1962). Determination of the basecomposition of deoxyribonucleic acid from its thermaldenaturation temperature. J. molec. Biol., 5, 109-118.

Owen, R. J., Hill, L. R., and Lapage, S. P. (1969). Deter-mination of DNA base compositions from melting profilesin dilute buffers. Biopolymers, 7, 503-516.

Owen, R. J. and Snell, J. J. S. (1973). Comparison of groupIlf with Flavobacterium and Moraxella. Antonie v

Leeuwenhoek, 39, 473-480.Pfennig, N. (1969). Rhodopseudomonas acidophila, sp.n.,

a new species of the budding purple nonsulfur bacteria.J. Bact., 99, 597-602.

Riddell, R. H. and Buck, A. C. (1970). Trimethoprim as anadditional selective agent in media for the isolation ofN. gonorrhoeae. J. clin. Path., 23, 481-483.

Snell, J. J. S. and Davey, P. (1971). A comparison of methodsfor the detection of phenylalanine deamination by Morax-ella species. J. gen. Microbiol., 66, 371-373.

Snell, J. J. S., Hill, L. R., and Lapage, S. P. (1972). Identifica-tion and characterisation of Moraxella phenylpyruvica. J.clin. Path., 25, 959-965.

Volk, J. and Kraus, S. J. (1973). Asymptomatic meningo-coccal meningitis. Brit. J. vener. Dis., 49, 511-513.

Wilkinson, A. E. (1952). Occurrence of Neisseria other thanthe gonococcus in the genital tract. Brit. J. vener. Dis., 28,24-27.

542

on March 22, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.29.6.537 on 1 June 1976. Dow

nloaded from