INTERNATIONALJOURNAL OF SYSTEMATICBACTERIOLOGY, July 1993, p. 490-499 Vol. 43, No. 3 0020-7713/93/030490-10$02.00/0 Copyright 0 1993, International Union of Microbiological Societies

Phylogeny of Species in the Family Isolated from Human Dental Plaque and Description of orale sp. nov.

FLOYD E. DEWHIRST,l* C.-K. CASEY CHEN,,? BRUCE J. PASTER,' AND JOSEPH J. -ZAMBON2 Department of Molecular Genetics, Forsyth Dental Center, 140 Fenway, Boston, Massachusetts 021 15, and Department of Oral Biology and Penodontology, School of Dental Medicine, State University of New York at Buflalo, Buflalo, New York 14214,

Fourteen human periodontal isolates recovered from a purported -selective medium containing 1 pg of clindamycin per ml displayed biochemical traits which differed from those described for E. corrodens. These organisms were gram-negative rods which corroded agar. The isolates were oxidase positive and urease, indole, and esculin negative. They differed from E. corrodens in catalase, nitrate reduction, lysine decarboxylase, and ornithine decarboxylase activities. One isolate, strain UB-294, was presumptively identified as Kingella denitnjkans. A second isolate, strain UB-204, differed from E. corrodens by being catalase positive and nitrate reduction negative. Twelve isolates, including strain UB-3ST (T = type strain), were phenotypically similar to except that they did not produce acid from maltose and were not beta-hemolytic. Essentially complete (1,480-base) 16s rRNA sequences were determined for strains UB-3ST, UB-204, and UB-294 and the type strains of animalis, Neisseria canis, Neisseria denitr@cans, , Neisseria JEavescens, Neisseria macaca, and Neisseriu polysaccharea. These sequences were compared with the previously published sequences of six other species belonging to the family Neisseriaceae. On the basis of the results of the comparative sequence analysis, UB-294 was confirmed as a K. denitrificans strain, UB-204 was identified as a member of a new species which may belong in the genus Eikenella, and UB-38T was identified as a member of a new species of the genus Kingelk, for which we propose the name Kingella orale. Since strain UB-204 was the only representative of a new species, it was not named. DNA probes for identification of E. corrodens, K. denitrificans, and K. orale based on 16s rRNA sequence information are described.

In a previous study examining the prevalence of Eikenella MATERIALS AND METHODS corrodens in the human oral cavity, we frequently encoun- tered gram-negative phenotypically similar to, but Bacterial strains. Isolation of strains, characterization of distinguishable from, Eikenella corrodens (2). These non- the periodontal status of patients, and partial biochemical Eikenella corrodens isolates grew well on selective medium characterization of the 14 non-Eikenella corrodens, agar- containing 1 kg of clindamycin per ml, corroded agar, and corroding strains have been described previously (2, 3). The did not produce acid from sugars as determined by standard Neisseria species used for 16s rRNA sequencing were broth tests. In contrast to Eikenella corrodens, these isolates obtained from the American Type Culture Collection. The lacked one or more of the following biochemical traits: bacterial strains used, their sources, and their accession nitrate reduction activity, lysine decarboxylase activity, and numbers are shown in Table 1. ornithine decarboxylase activity. In a subsequent study of Biochemical characterization. Partial biochemical charac- six of these isolates, we reported that they exhibited less terization of six of the strains was described previously (3). than 33% DNA homology with Eikenella cowodens refer- Bacteria were grown on Trypticase soy agar plates supple- ence strains (3). Four of these strains exhibited more than mented with 5% sheep blood, 5 mg of hemin per liter, and 0.5 70% DNA homology with one another, indicating that they mg of menadione per liter at 37°C in a humidified anaerobic belonged to a single species. In this study, we further chamber containing 5% CO,. Catalase activity, esculin hy- characterized these six isolates, as well as eight additional drolysis, indole production, reduction of nitrate to nitrite, strains. We found that the 14 isolates fell into three groups and urease activity were determined by using the methods on the basis of biochemical characteristics and protein described in the Anaerobe Laboratory Manual (11). Oxidase profiles determined by sodium dodecyl sulfate (SDS)-poly- activity was determined by using a 1% aqueous solution acrylamide gel electrophoresis (PAGE). 16s rRNA se- of N,Nfl,"-tetramethyl-p-phenylenediamine dihydrochlo- quences were determined for representative isolates belong- ride (Becton Dickinson Reagent Droppers). Acid production ing to each of these groups and for a number of Neissena from glucose, xylose, mannitol, lactose, sucrose, and mal- species. The 16s rRNA sequence information was used to tose was determined by using broth cultures as described in determine the phylogenetic position of each of these groups the Manual of Clinical Microbiology (25). Strains UB-38= of non-Eikenella corrodens organisms within the family (T = type strain), UB-204, and UB-294 and the type strains Neisseriaceae and to design DNA probes. of Kingella and Neissen'a species were further characterized by using RapID/NH System (Innovative Diagnostics Sys- tems, Inc., Atlanta, Ga.) and Microscan HIND (Baxter * Corresponding author. Diagnostics Inc., Deerfield, Ill.) panels. For these tests, the t Present address: Department of Periodontology, School of strains were grown on brain heart infusion-blood-agar plates Dentistry, University of Southern California, Los Angeles, CA aerobically in an atmosphere containing 76% N,, 9% H,, 9% 90089. CO,, and 6% 02.The RapID/NH System and Microscan

490 VOL.43, 1993 KINGELLA ORALE SP. NOV. 491

TABLE 1. Strains, sources, and accession numbers Strain or species" Site of oral sample' Clinical diagnosis' Culture collection designation(s)d GenBank accession no.= Eikenella corrodens strains ATCC 23834T Sputum ATCC 23834T M22512 UB-344 Sub-GP UP Eikenella-like strain UB-204 Sub-GP AP CCUG 28283 LO6169 Kingella denitri cans strains ATCC 33394 zti ATCC 33394T M22516 UB-294 Supra-GP LJP CCUG 28284 L06166f Kingella orale strains UB-13 Sub-GP PH UB-25 Sub-GP PH UB-38T Supra-GP AP ATCC 51147T, CCUG 30450T LO6166 UB-75 Sub-GP AP UB-107 Sub-GP PH UB-149 Supra-GP AP UB-180 Supra-GP AP UB-209 Supra-GP AP UB-220 Supra-GP PH UB-2369 Sub-GP PH UB-245 Sub-GP PH UB-295 Supra-GP LJP Kingella kingae ATCC 23330T M22517 Neisseria gonowhoeae NCTC 8375T X07714 ATCC 43768T LO6167 Neisseria jiavescens ATCC 13120T M6168f Neisseria macaca ATCC 33926T LO6169 Neisseria canis ATCC 14687T LO6176 Neisseria elongata ATCC 25295= LO617lf Neissek animalis ATCC 19573T LO6172 Neisseria denitnjicans ATCC 14686T L06173f Wtreoscilla stercoraria ATCC 15218T LO6176 UB, State University of New York at Buffalo School of Dental Medicine, Buffalo, N.Y.; ATCC, American Type Culture Collection, Rockville, Md. 'Sub-GP, subgingival plaque; supra-GP, supragingival plaque. ' Clinical diagnosis of patient from which the strain was isolated. PH, periodontally healthy; AP, adult periodontitis; UP, localized juvenile periodontitis. The criteria used for diagnosis have been described previously (2). Culture collections in which strains have been deposited or from which strains were obtained. ATCC, American Type Culture Collection, Rockville, Md.; CCUG, Culture Collection of the University of Goteborg, Goteborg, Sweden; NCTC, National Collection of Type Cultures, London, United Kingdom. 16s rRNA sequences are available for electronic retrieval from GenBank under the accession numbers shown. Through cross-distribution, these sequences should also be available from European (EMBL) and Japanese (DDBJ) data bases. Sequence determined in this study and deposited by us. This strain did not react with the K orak DNA probe.

HIND panels contain 10 and 18 tests, respectively, including blood-agar plates were gently suspended in 10 mM Tris tests for alkaline phosphatase, ornithine decarboxylase, pro- buffer (pH 7.4) at a concentration of approximately lo8 cells line aminopeptidase, nitrate and nitrite reduction, resazurin per ml. Samples were negatively stained with 1% (wtkol) reduction, and production of acid from glucose. These phosphotungstic acid (pH 6.5) for 15 to 20 s. Specimens were panels were incubated in air at 37°C for 4 h. Strains were examined with a JEOL model JEM-1200EX transmission examined for growth on MacConkey agar (Difco) by incu- electron microscope operating at 100 kV. bation at 37°C in a humidified anaerobic chamber containing 16s rRNA sequencing. rRNA was isolated and partially 5% CO, for 7 days. purified by a modification of the procedure of Pace et al. (19) Protein profiles. Bacteria grown on blood-agar plates for 2 as previously described (20). rRNA sequences were deter- days were harvested into sterile saline and pelleted by mined by using a modification of the standard Sanger centrifugation. The cells were resuspended in distilled water, dideoxy chain termination technique (17). Seven primers and the protein concentration was determined by the method of Lowry et al. (18). The cell suspension was then diluted complementary to conserved regions of the 16s rRNA with an equal volume of 2 x sample buffer (0.125 M Tris-HC1 sequence were elongated by using avian myeloblastosis [pH 6.81, 8% SDS, 20% glycerol, 10% 2-mercaptoethanol, virus reverse transcriptase (9). 0.006% bromophenol blue) and boiled for 10 min before use. 16s rRNA data analysis. A program set for data entry, SDS-PAGE was performed by using a Laemmli gel system editing, sequence alignment, secondary structure compari- (16) with a 3.8% stacking gel and a 12% separating gel. A son, similarity matrix generation, and dendrogram construc- 20-kg portion of protein was loaded per lane. The protein tion for 16s rRNA data was written in Microsoft Quick- bands were visualized by Coomassie brilliant blue (Sigma) BASIC for use on IBM PC-AT and compatible computers. staining. The molecular weights of protein bands were RNA sequences were entered and aligned as previously estimated by reference to molecular weight standards (low- described (20). Similarity matrices were constructed from molecular-weight calibration kit; Pharmacia LKB Biotech- the aligned sequences by using only those sequence posi- nology, Inc., Piscataway, N.J.) run in the same gel. tions for which 90% of the strains had data. The similarity Electron microscopy. Cells grown on brain heart infusion- matrices were corrected for multiple base changes by the 492 DEWHIRST ET AL. INT. J. SYST.BACTERIOL.

TABLE 2. Differential characteristics of K orale and phenotypically similar speciesa

N. N. animalk N. elongata K. denitn'ficans Strain Eikenella Characteristic K. oraleb K. kingae N. canis denitn'ficans UB-204 corrodens

Cell shape Rods Cocci Cocci Rods Rods Rods Rods Beta-hemolysis + ------Catalase - + Acid production from: Glucose Maltose Fructose Sucrose Nitrate reduction Nitrite reduction Gas produced from nitrite Lysine decarboxylase Ornithine decarboxylase - - Alkaline phosphatase' + ------Resazurin reductione V + + + + + [-I Prohe aminoDeDtidase" - + - + + + + G+C content bf'DNA (mol%) 56-58 47 50 56 ND 53 54-57 60 56-58

~~ ~ Data from this study and reference 26. +, more than 90% of the strains are positive; [+I, more than 75% of the strains are positive; V, variable reactions occur; [-I, less than 25% of the strains are positive; -, less than 10% of the strains are positive; ND, not done or not known. Data are for the following 11 strains: UB-13, UB-25, UB-38=, UB-75, UB-107, UB-149, UB-180, UB-209, UB-220, UB-245, and UB-295. The reactions shown are the reactions for all 11 strains except as noted below. Strain UB-75 is catalase positive. Weakly positive as determined by the RapID/NH System (Innovative Diagnostics) and negative as determined by the standard fermentation broth test. Reactions were determined for the type strains. Data were also obtained from the RapID/NH System differential chart. method of Jukes and Cantor (12). Phylogenetic trees were GenBank accession numbers. The GenBank accession constructed by the neighbor-joining method (21). numbers for the organisms examined in this study are shown DNA probe design. The 16s rRNA sequence from a target in Table 1. organism was aligned with all sequences available from the same family, as well as a selection of sequences from more RESULTS AND DISCUSSION distantly related taxa. The nontarget sequences were con- verted to a format in which differences from the target Our previous reports on non-Eikenella cowodens oral sequence were shown. Regions of approximately 24 to 30 isolates suggested that these strains could be placed into bases where the sequences of all species differed from the three groups on the basis of phenotypic and DNA-DNA target sequence by at least 4 bases were selected as probe homology data (2, 3). The first group contained 12 strains targets. Oligonucleotide DNA probes were synthesized by representing a new Kingella species, which we describe Midland Certified Reagent, Midland, Tex. below and for which we propose the name Kingella orale. DNA probe labeling. Probe oligonucleotides were end We refer to this group hereafter as K. orale. DNA-DNA labeled with [32P]ATPby using the T4 polynucleotide kinase hybridization of strain UB-38, the proposed type strain of K. method (22). The labeled oligonucleotide probes were puri- orale, and type strain ATCC 23834 and representative strain fied by chromatography on a Sep-Pack C,, column (23). UB-344 of Eikenella corrodens gave homology values of 1 to Dot blot hybridization. A loopful of cells from each of the 7% (3). The second group, containing one strain, strain strains to be examined, including the reference control UB-294, is identified below as Kingella denitri#cans. The strains, was scraped from a 2-day blood-agar plate and third group also contained one strain, strain UB-204, and placed on 9-cm-diameter Whatman no. 541 filter paper in an was identified as a probable new species closely related to 8-by-8 array of 5-mm circles on 8-mm centers. The cells were Eikenella corrodens (levels of homology with ATCC 23834T, lysed by floating the filter on 50 ml of lysis buffer (5% Igepal, 22 to 31%) (3). In this study additional strains and methods 5% SDS, 10 mM EDTA, 10 mM Tris-HC1 [pH 7.81) for 1h at were used to further characterize these three groups of oral 60°C. The filter was washed with lx SSC (lx SSC is 0.15 M isolates. NaCl plus 0.015 M sodium citrate) for 10 min at room Biochemical characterization. All strains grew aerobically temperature. The cells were fixed by floating the filter on 50 and anaerobically, were oxidase positive, and were negative ml of 0.5% glutaraldehyde in 1X SSC at 5°C for 30 min. The for indole production, esculin hydrolysis, and production of filter was washed with lx SSC for 10 min at room temper- acid from glucose, lactose, maltose, mannitol, sucrose, and ature. The filter was floated on 50% ethanol-50% 1x SSC for xylose as determined by standard broth tests. The strains 10 min at room temperature and then air dried. The filter was which we examined could be separated into three groups on blocked by floating it on 50 ml of 0.1 M triethanolamine (pH the basis of catalase, nitrate reduction, and lysine (or orni- 8.0) for 10 min at room temperature. Acetic anhydride (125 thine) decarboxylase activities (Table 2). These results are in pl) was added, and the filter was incubated for 10 min. The agreement with our previous data (3). Alkaline phosphatase filter was washed twice with IXSSC and air dried. Hybrid- and proline aminopeptidase activities were also useful for ization of DNA probes with immobilized cells was per- differentiating K. orale from Eikenella con-odens, K deni- formed as previously described (27). Localization and quan- trijicans, and strain UB-204. K. orale is alkaline phosphatase tification of bound radioactivity were determined with a positive and proline aminopeptidase negative, while the Betascope 603 blot analyzer (Betagen, Waltham, Mass.). other organisms have the opposite reactions. The 16s rRNA VOL.43, 1993 KINGELLA ORALE SP. NOV. 493

FIG. 1. Protein profiles of Eikenella corrodens-like strains. Whole-cell proteins were separated by SDS-12% PAGE and stained with Coomassie blue. sequence information described below indicated that these weight of 87,000 and a second having a slightly variable organisms are members of the family Neisseriaceae. Since it molecular weight of about 36,000. Several strains also pos- is known that Neissena and Kingella species require special sessed a third major cellular protein having an apparent media for testing acid production from carbohydrates (13), molecular weight of 58,000. Thus, protein profile information strains UB-38T,UB-204, and UB-294 and the type strains of supports the presence of the three groups identified by the Neisseria, Kingella, and Eikenella species listed in Table biochemical characteristics. 2 were examined by using RapID/NH System and Micro- Colony and cell morphology. The colony morphology of Scan HIND panels. As determined with these panels, UB- each of the 14 non-Eikenella corrodens strains was typical of 3gT and UB-294 produced low levels of acid from glucose. the colony morphology of Eikenella corrodens. The colonies Our previous inability to identify UB-294 as a K denitrifi- were 2 to 3 mm in diameter after 3 days of growth on blood- cans strain was due to our perception of the strain as agar and were round with slightly irregular borders, and each anaerobic and asaccharolytic (2, 3). colony possessed an umbonate elevation, as previously Protein profiles. Figure 1 shows the SDS-PAGEprofiles of described (3). Corrosion or pitting of the agar surface was total cellular proteins for the 14 strains and two Eikenella observed around the granular periphery of each colony. corrodens reference strains. The 12 phenotypically similar After 5 to 6 days of incubation, well-isolated colonies were K orale strains, including UB-38T,had recognizably similar more than 6 mm in diameter. The ability to form spreading protein profiles which differed from the protein profiles of colonies suggests that these microorganisms may have the two Eikenella corrodens reference strains, as well as the twitching motility, as observed in other Eikenella, Kingella, protein profiles of phenotypically unique strain UB-204 and and Neisseria species. K denitri3cans UB-294. These 12 K. orale strains had two The ultrastructures of K. orale UB-3gT, strain UB-204, major cellular proteins, one having an apparent molecular and K. denitrificans UB-294 were examined by electron 494 DEWHIRST ET AL. INT. J. SYST.BACTERIOL.

FIG. 2. Electron micrographs of strains UB-38T (top) (bar = 400 nm) and UB-204 (bottom) (bar = 500 nm).

microscopy. The cells of K. orale UB-38=were 0.6 to 0.7 by possessed unipolar fimbriae (Fig. 2, top). The fimbriae were 1 to 3 pm, regular to irregular, gram-negative rods (Fig. 2, about 5 nm in diameter and ranged from 1 to 10 pm long. top) and were often observed in pairs or short chains. The Fimbriae were not observed on broth-grown cells. The cells cells did not possess flagella, but cells grown on blood-agar of UB-204 (Fig. 2, bottom) were 0.5 to 0.6 by 3 to 8 pm, and VOL. 43, 1993 KINGELLA ORALE SP. NOV. 495

KO .MCAU~AGAGUWGAUCCUGGCUCAGAWGMCGCUGGCGG~UG~A~CAUGCMWCGMCGGCAGCAC~GAGCUUGCUCU~-GWGGCGAWGGCGMCGGWGAW~AKO 204 .MCAUMGA~GAUCCUGGCUCAGAWGMCGCUGGCGG~UG~ACACAUGCMWC~CGGCAGCGGGgGAWGCUUG~CUU~WCCGGCGAWGGC~CGGWGA~M204 294 ..ACAUnAGACUWGAUCCUGGCUUGAUUCMCGCUGGCGG~UG~ACA~UGCMWCGGACGG~GCGG--AGWGCUUGCACW---GCCGGCGAWGGCGMCGGWGAWM 294 Ec MAWGMGACWUGAUCAUGGCUUGAUUCMCCCUGCGG~GGCCUMCACAUGCMWCGMCGWM~GGMGMGCUUGCUUC~-GCUGACCAWGGCGCACGGWGA~EC 10 20 30 40 50 60 70 80 90 100 110 120

KO UGCAUCGGMCWACCGAWMUGGGGGAUMCUAUCCGAMGGCKO 204 UAUAUCGGMCWACCGAGCAWGGGGGAUMCCMUC~GAWGGCUMUACCGCAUACACUCUGAGGAG~G~GGGGACC~GGC~GCGCU~GAGCGGCC~C204 2% CAUAWGGMUWACCGAWACUGGGGGAUMCCAAUCGMACC 294 Ec UWCUGGGAMCUGCCUGAUGGAGGGGGA~~CUG~CG~G~~CCG~UMCWCGCMGAC~~GGGG~CCUUCGGGCCU~GC~UCG~U~GCC~~~Ec 130 140 150 160 170 180 190 200 210 220 230 240

KO KO 204 204 294 294 Ec Ec

KO KO 204 204 294 294 Ec EC

KO CACGGUACCUnAAGAAUMGCACCGG~~CWGCCaGCaGCCgCGG~~CWAGGWGC~GC~G~WaCUGGGCW~GCGAGCG~GACG~~G~KO 204 GACGWACCWAAGAAUMGCACCGGCUMCUACGUGCCACCAGCCGCGG~UAC~GGWGCGA~~~CGGM~CUGGGC~GCGAGCG~GACG~G~G204 294 GACGGUACCUaMGMUMGCACCGGCUM~CWGCAG294 EC GACGWACCCGCAGMGMGCACCGGCUM~CWGCCAGCAEC 490 500 510 520 530 540 550 560 570 580 590 600

KO KO 204 204 294 2% EC EC

KO GAUGGCGAAGGCAGCCCCcUGGGAUMUACUGACGUUCAUGCUC~GCWGGWAG~~G~A~UACCCUG~WC~CGCCCU~CGAUWCM~GCU~GGGCMKO 204 GAUGGCGMGGCAGCCCCcUGGGAUMUACUWCGUUCAUCC204 294 GAUGGCGMGGCAGCCCCCUGGGAUMUA~~C~CAUGC294 Ec GWGGCGMGGCGGCCCCCUGGACGMGACUCACGCU~GWGCCAMGEC 730 740 750 760 770 700 790 800 810 820 830 840

KO CAUgAWGCWAWAGCWAGCUMCGCW~WGACCGCCUGGGGAWACG~CGCMGA~CU~G~GACGGGGACCCGCACMGCGWGGAU~U~~KO 204 CUugAWGCWAWACCWAGCUMCGCW~UCGACCGCC204 294 CWGAWGCWAWAGCCUAGCUMCGCW~UCGACCGC~GGGA~CG~GCM~~~GGMWGACGGGGACCCG~CMGCGWGGAU~UWGGA~U294 EC CWGAGGCWGGCWCCGGAGCUMCGCGUUMGUCGACCGCCEC 850 860 870 880 890 900 910 920 930 940 950 960

KO UCGAUGCMCGCGAAGMCCUUACCUGWCUUGACAUWACGGMUCUUC~GAGACG~GA~CCUUCGGGAGCC~CACAGWG~GCAUGGCUWCWCAGCUCWWCW~KO 204 UCGAUGCMCGCGMGMCCUUACCUGWCWGACAUWACGGMC~CCAGAGACGGMGG~GC~GG~CC~M~CAGWGCUG~UGG~WCWCAGCUCWWCWGA204 294 UCGAUGCMCGCGAACAACCWACCUGWCWGACAUWACGGMGM~CAGAGACGA~WGCCUUCGGGAGCC~~CAGWGCUG~UGGCUWCWCAGCUCWWCWGA294 EC UCGAUGCMCGCGMGMCCWACCUGW~CAUCCACG~~CA~GAUGAGMUWGCCWCGG~CCWGAGACAGWGCUGCAUGGCUWCW~GCUCW~WGAEc 970 980 990 1000 1010 1020 1030 1040 1050 lob0 1070 1osO

KO GAUWUGGWUAAGUCCCGCMCGAGCGCMCCCWWCAWAGUUGCCACCA-WCAGUUGGGCACUCUMUGA~CUGCCGWGAC~CCGGAG~GWGGGGAUGACW~WCKO 206 GAUGWGGGWMGUCCCGCMCCACCCCMCCCWWCAWAGUUGCCAUCA-UlllAGUUGGGCACUCUMU~ACACUGCCGWGACMACCGGAGGMGWGGGGAUGACW~WC 204 294 GAUGWGGGWMGUCCCGCMCCAGCGCMCCCUVGUCAWA~GCCAUCA-WlllAGUUGGGCACUCUMUGGACUGCCGWGAC~CCGGAGGMGWGGGGAUGACW~WC294 Ec MUGUUGGWUAAWCCCGCMCGAGCGCMCCCWAUCCUlllGUUGCCAGCGWCCGGCCGG~CU~GGAGA~GC~WGAU~CUGGAGGMGWGGGGAUGACWCMWCEc 1090 1100 1110 1120 1130 1140 1150 1160 1170 1180 1190 1200

KO CUCAUGGCCCWAUGACCAGGGCWCACACWCAUACMUGWCGWACAGAGGWAGCCMGCCGCGAGWGGAGCCMUCUCA~CCGAUCWAWCCGGAWGCA~CUGCMCKO 204 CUCAUGGCCCUUAUGACCAGGGCWCACACWCAUACMUGWCGWACAGAGGWAGCCMGCCGCGAGWGGAGCCMUCCCA~CCGAUC~WCCGGAWGCACU~GCMC 204 294 CUCAUGGCCCWAUGACCACGGCUUCACACWCAUACMUGWCGWACAGAGGWAGCCMGCCGCGAGWGGAGCC~.UCC~~CCGAUCWAWCCGGAWGCACUCUGCMC294 EC AUCAUGGCCCUUACGACCAGGGCUACACACWGCUACMUGGCGCAUACMAGAGMGCGACCUCGCGAGAGCMGCG~CCUCAUMAWGC~CGUAWCCGGAWGGAWCUGCMC EC 1210 1220 1230 1240 1250 1260 1270 1280 1290 1300 1310 1320

KO UCGAGUGCAUGAAGUCGGMUCGCUAGUMUCGCAGGUCAGCAUACUGCGWWAUACGUUCCCGGGUCWWACACACCGCCCGUCACACCAUGGGAWGGGGGAUACCAGMWAGGUKO 204 UCGAGUGCAUGAAGUCGGAAUCGCUAWMUCGCAGGUCACCAUACUGCGW~UACGWCCCGGGUCWGUACACACCGCCCWCACACCAUGGGAGUGGGGGAUACCAGMGCAGW 204 294 UCGACUGCAUGAAGUCGGAAUCGCUAGUMUCGCAGWCAGCAUACUGCGGUgAAUACCWCCCGGWCWWACACACCGCCCWCACACCAUGGGAGUGGGGCAUACCAGMWAGW 294 Ec UCGACUCCAUGAAWCGGAAUCGCUAGUMUCWGGAUCAGMUGCCACGWGAAUAC~CCCGGGCCWWACACACCGCCCWCACACCAUGGGAGUGGGWGCAAAAGMWAGW Ec 1330 1340 1350 1360 1370 1380 1390 1400 1410 1420 1430 1440

KO AGGCUAACCGCAAGGAGGCCGCUUACCACGGUAUGUWCAUGACUGGGG...... K i nge 1 la orate UB-38 204 ACGAUAACCGCAAGGGWCCGCWGCCACGGUAUGCUUCAUCACUGGGG...... Strain UB-204 294 AGGGUAACCGCAAGGAGCCCGCWACCACGGUAUGCUUCAUGACUGGGG...... Strain UB - 294 Ec AGCUUAACCWCGGGAGGGCGCWACCACUWGUCAWCAUGACUGGGGUGAAGUCGUMCMGGUAACCGUAGGGGAACCUGCGGWGGAUCACCUCCWA 1450 1460 1470 1480 1490 1500 1510 1520 1530 1540 FIG. 3. Aligned 16s rRNA sequences. The numbering system is the Escherichia coli numbering system (1).The sequences are presented by using the International Union of Biochemistry single-letter codes for nucleotide bases and ambiguities. Lowercase letters indicate some uncertainty in base identity. Dashes indicate gaps that were inserted for sequence alignment, and dots indicate regions at each end that were not sequenced. KO, K orale UB-38T; 204, strain UB-204; 294, strain UB-294; Ec, Escherichia coli. 496 DEWHIRST ET AL. INT. J. SYST. BACTERIOL.

Probe sequence Target sequence Kingella orale UB-38T Strain UB-204 Eikenella corrodens Kingel 1a kingae Kingella denitrificans ATCC 33394T Kingella denitrificans UB-294 Neisseria canis Neisseria polysaccharea Neisseria macaca Neisseria elongata Neisseria animalis Neisseria denitrificans Simonsiell a muell eri Escherichia coli Actinobacillus actinomycetemcomitans Porphyromonas gingivalis FIG. 4. DNA probe design for strain UB-38=. The target sequence corresponds to bases 462 to 485 of the Escherichia coli sequence. Matches between the target sequence and the sequences of other species are indicated by dots. Mismatches are indicated by the International Union of Biochemistry single-letter codes for nucleotide bases and ambiguities. Dashes indicate gaps inserted for sequence alignment.

the cells of K. denitrificans UB-294 were 0.6 to 0.7 by 2 to 4 An examination of strain UB-236 by 16s rRNA sequence pm. The cells of both strain UB-204 and strain UB-294 had analysis is planned to resolve the phylogenetic position of fimbriae similar to those of strain UB-38T cells. this organism relative to other K. orale strains and to refine 16s rRNA sequencing. Essentially complete 16s rRNA or modify the K. orale probe if necessary. In conjunction sequences (1,480 bases) were determined for strains repre- with phenotypic analysis, the K. orale probe should be senting each of the three groups (UB-38T, UB-204, and useful for identifying additional strains of this species. The UB-294). The 16s rRNA sequences of several Kingella and two Eikenella cowodens reference strains and UB-294 bound Neisseria species have been determined previously in our the Eikenella corrodens probe, but not the UB-38T probe. laboratory (5, 6, 8). The 16s rRNA sequences of Neisseria The reaction of our previously designed Eikenella corrodens animalis, Neisseriu canis, Neisseria denitrificans, Neisseria probe (Fig. 5)with UB-294 cells was expected on the basis of elongata, Neisseria jlavescens, Neissen'a macaca , and subsequent sequence information which showed that these Neisseriapotysaccharea are reported for the first time in this organisms have identical sequences at the target site. The paper. The sequences of UB-38*, UB-204, and UB-294 Eikenella corrodens probe has only two G-U mismatches aligned with the Escherichia coli sequence are shown in Fig. with the type strain of K. denitnBcans, and, although not 3. The sequences of all three strains are unique, but, as tested, would be expected to show some cross-reactivity described below, the sequence of UB-294 differs by only with this organism. A comparison of the sequence of three bases from the sequence of the type strain of K. Eikenella corrodens with the sequences of all human oral denitrificans, thus identifying this strain as a K. denitrificans members of the family Neisseriaceae indicates that there is strain. no single 24-base target which differs by two mismatches 16s rRNA-based DNA probes, A 24-base region, starting at from all other species. Using 16s rRNA targets requires a Escherichia coli position 462, was chosen as the target site minimum of two probes for correct identification of for a K orale DNA probe. As shown in Fig. 4, the target Eikenella corrodens. The four 16s rRNA probes for sequence in K. orale UB-3gT differs by 4 to 10 bases from Eikenella corrodens described by other investigators should, sequences in other species. A probe designed for Eikenella on the basis of the results of a sequence analysis, cross-react cowodens before the sequences of K. denitnficans and with one species, or one probe should cross-react with most strains UB-38T, UB-204, and UB-294 were obtained is oral Neisseria and Kingella species (lo). Twenty-four-base shown in Fig. 5. Deficiencies in this probe are discussed sites beginning at Escherichia coli positions 208 and 470 may below. Cells of the 14 isolates and 2 Eikenellu corrodens be useful as probe targets for specific identification of reference strains were hybridized with these probes. All 12 UB-204. Twenty-four-base sites beginning at Escherichia strains were similar biochemically, as determined by SDS- coli positions 84, 220, and 1013 may be useful as probe PAGE protein profiles, and, except for UB-236, bound the targets for identification of K. denitrificans. K. orale probe but not the Eikenella cowodens probe. Phylogenetic analysis based on 16s rRNA sequence compar- UB-236 and UB-204 cells did not bind either probe. Figure 1 isons. A similarity matrix based on a comparison of the 16 shows that strain UB-236 lacks some fainter bands in the 40- sequences listed in Table 1 is shown in Table 3. A phyloge- to 70-kDa region that are present in other K. orale strains. netic tree derived from the distance matrix by using the VOL.43, 1993 KINGELLA ORALE SP. NOV. 497

Probe sequence Target sequence Eikenella corrodens Kingella denitrificans UB-294 Kingella denitrificans ATCC 33394T Strain UB-204 Neisseria elongata Neisseria animalis Neisseria denitrificans Simonsiella muelleri Neisseria canis Kingella orale strain UB-38T Kingella kingae Neisseria gonorrhoeae Neisseria polysaccharea Neisseria flavescens Neisseria macaca Escherichia coli A. actinomycetemcomitans Porphyromonas gingivalis FIG. 5. DNA probe design for Eikenella corrodens. The target sequence of the probe corresponds to bases 462 to 488 of the Escherichia coli sequence. For further explanation see the legend to Fig. 4.

neighbor-joining method is shown in Fig. 6. The tree con- 16s rRNA sequence analysis and phenotypic analysis. The tains three clusters. The first cluster contains K. orale, following traits qualify this group for inclusion in the genus Kingella kingae, and N. canis. Grouping of K. orale with K. Kingella: gram-negative rods, oxidase positive, catalase kingae, but not N. canis, has occurred in many different negative, ornithine and lysine decarboxylase negative, in- analyses in which different outgroups or bootstrapping was dole negative, and urease negative. The genus Kingella used (5). The second cluster contains Eikenella corrodens, currently contains three species, K. kingae (the type spe- strain UB-204, K denitrificans ATCC 33394* and UB-294, cies), K. denitrificans, and K orale. Kingella indologenes N. elongata, N. denitrificans, and Simonsiella muelleri’.The was previously transferred to the genus Suttonella as Sut- sequence of K denitrificans UB-294 is 99.8% similar to the tonella indologenes (8). Below we provide an emended sequence of the type strain, and these organisms form a tight description of the genus Kingella which is consistent with cluster on the tree. The sequence of UB-204 is 97.0% similar exclusion of K indologenes and inclusion of K orale. Traits to the Eikenella corrodens sequence, and this organism is the which define K orale and differentiate it from phenotypi- nearest neighbor of Eikenella corrodens on the tree. The cally similar species are shown in Table 2. third cluster contains Neisseria gonon-hoeae, N. polysac- The natural habitat of K. orale appears to be human dental charea, N. flavescens, and N. macaca. A more detailed plaque. It was recovered from a majority of patients and, in analysis of the phylogenetic structure of the family Neisser- some patients, was the predominant cultivable organism (3). iaceae, including Centers for Disease Control groups M5, It was recovered less frequently from saliva and mucosal M6, and EF4 and Simonsiella species, is in preparation (7). sites, such as buccal mucosa, tongues, and tonsils, than from Phylogenetic significance of EikeneUa corrodens-like groups. dental plaque (4). The ability of K orale to produce infection The results of biochemical, protein profile, DNA-DNA (3), or disease is unknown. It was recovered from both peri- and 16s rRNA sequence analyses suggest that at least four odontally healthy and diseased subjects (4). However, its species of agar-corroding bacteria can be isolated from closest relative, K kingae, is an opportunistic pathogen that dental plaque on Eikenella con-odens-selective plates con- occasionally causes osteomyelitis (15). Because Kingella taining 1 pg of clindamycin per ml incubated anaerobically. and Eikenella species are known to exchange antibiotic Sequence differences of three bases between strain UB-294 resistance plasmids with N. gononhoeae and Neisseria and the type strain indicate that there is some genetic meningitidis (14), K. orale could possibly act as a plasmid heterogeneity among strains of K. denitn’cans. Strain UB- reservoir. The recognition that there are a number of rod- 204 represents a new species, probably in the genus shaped members of the family Neisseriaceae, some which Eikenella, on the basis of its phenotypic and 16s rRNA are recovered under anaerobic or microaerophilic condi- sequence similarity to Eikenella corrodens. Since we have tions, such as Eikenella corrodens, may facilitate identifica- identified only one isolate of this species, it is not named in tion of new members of this family. this paper. Strains of K orale cluster together, as deter- Emended description of the genus Kingella. As described in mined by biochemical analysis, SDS-PAGE protein profiles, Bergey’s Manual of Systematic Bacteriology (24), with the DNA-DNA hybridization (strains UB-3ST, UB-209, UB-220, following modifications. Straight rods 0.6 to 1 pm in diame- and UB-295) (3), and DNA probe reactivity (except UB- ter by 1 to 3 pm long. Indole and urease negative. Chemo- 236). They are closely related to K. kingae as determined by organotrophic. Acid is produced from glucose and a limited 498 DEWHIRST ET AL,. INT. J. SYST. BACTERIOL.

TABLE 3. Similarity matrix based on 16s rRNA sequence comparisons

% of similarity or % of difference"

Taxon

K orale UB-38T 96.2 96.8 95.5 95.0 95.0 95.0 95.7 95.8 94.2 94.4 95.4 96.0 94.8 94.0 K kingae 3.9 95.6 94.6 94.2 95.5 95.4 95.1 95.2 94.2 94.0 94.8 95.4 94.4 92.9 N. canis 3.3 4.6 95.0 95.5 95.6 95.5 95.9 95.9 94.2 95.6 96.3 96.3 95.9 94.3 Eikenella corrodem 4.7 5.6 5.2 97.1 97.0 97.1 96.8 96.8 95.6 93.9 94.8 95.4 95.2 93.4 Eikenella sp. strain UB-204 5.1 6.1 4.6 3.0 96.5 96.3 96.5 95.4 93.7 93.8 94.7 95.2 95.8 92.7 K denitn'jicans 5.2 4.6 4.6 3.1 3.6 99.8 97.0 96.9 94.9 94.7 95.4 95.5 96.0 92.6 K denitnpcans UB-294 5.1 4.7 4.6 3.0 3.7 0.2 97.0 97.0 95.0 94.7 95.6 95.6 96.0 92.5 N. elongata 4.4 5.0 4.2 3.3 3.6 3.0 3.1 96.9 94.6 94.8 95.4 96.3 96.4 93.3 N. denitn'jicans 4.3 5.0 4.2 3.2 4.8 3.2 3.1 3.2 96.2 94.8 95.5 96.1 95.3 94.0 Simonsiella muelleri 6.1 6.0 6.0 4.5 6.5 5.3 5.2 5.6 3.9 93.0 94.1 94.2 93.3 92.8 N. gonorrhoeae 5.8 6.2 4.6 6.3 6.5 5.5 5.5 5.4 5.4 7.3 98.5 97.4 96.3 93.0 N. polysaccharea 4.8 5.4 3.8 5.4 5.5 4.7 4.6 4.7 4.6 6.2 1.5 97.9 97.1 93.9 N. flavescens 4.1 4.8 3.8 4.8 4.9 4.7 4.5 3.8 4.1 6.1 2.6 2.1 97.3 94.3 N. macaca 5.3 5.9 4.2 5.0 4.3 4.2 4.2 3.7 4.9 7.0 3.8 2.9 2.8 92.8 K stercoraria 6.3 7.5 5.9 6.9 7.7 7.8 7.9 7.0 6.3 7.6 7.3 6.4 6.0 7.5 " The values above the diagonal are uncorrected percentages of similarity based on a comparison of 1,438 base positions. The values below the diagonal are percentages of difference corrected for multiple base changes by the method of Jukes and Cantor (12). number of other carbohydrates. The G+C content of the Growth is also supported by Todd-Hewitt broth supple- DNA is 47 to 58 mol%. mented with 2 mg of KNO, per ml, 50 pg of L-cysteine Description of Kingeh orale sp. nov. Kingella orale hydrochloride per ml, and 5 pg of hemin per ml, but the yield (o.ra'le. L. adj. orale, oral, pertaining to the mouth). The is low. Does not grow on MacConkey agar. Colonies are description below is based in part on our previous data (2,3). round with slightly irregular borders and flat to umbonate, Rods or coccobacilli approximately 0.6 to 0.7 pm in diameter and each colony has a granular periphery. Colonies appear by 1 to 3 pm long with rounded ends. Cells can form pairs or to corrode the agar surface. Oxidase positive as determined chains. Cells have monopolar fimbriae up to 10 pm long. by using tetramethyl-p-phenylenediamine.Catalase nega- Gram negative, but there is a tendency to resist Gram tive. Produces acid from glucose weakly. Does not produce decolorization. Not motile by means of flagella, but cells acid from lactose, maltose, mannitol, sucrose, and xylose. form spreading colonies. Aerobic or facultatively anaerobic. Does not reduce nitrate or nitrite. Does not produce lysine or The specific growth requirements of K orale are unknown. ornithine decarboxylase, indole, or urease or hydrolyze Growth is supported by 5% sheep blood agar supplemented esculin. The habitat of K orale appears to be human dental with 5 mg of hemin per liter and 0.5 pg of menadione per ml plaque. The G+C content of the DNA is 56 to 58 mol%. The in both anaerobic and aerobic environments with C02. type strain is UB-38. Description of type strain UB-38. The phenotypic charac- teristics of strain UB-38T are the same as those described (% Difference) above for the species. This strain was isolated as the predominant cultivable microorganism from a supragingival Kingella orale U0-30 T Kingella kingae ATCC 23330 T plaque sample from a patient with adult periodontitis. The Neisseria ranis ATCC 14607 T G+C content of the DNA is 56 mol%. The type strain has Eikenella corrodens ATCC 23834 T been deposited in the American Type Culture Collection, Eikenella sp U0-204 Rockville, Md., as strain ATCC and in the Culture Kingella deni tri firans ATCC 33394 T 51147 - Kingella deni tri firans U0-294 Collection of the University of Goteborg, Goteborg, Swe- - Neisseria elongata ATCC 25295 T den, as strain CCUG 30450. Neisseria denitrificans ATCC 14606 T - Simonsiella muelleri ATCC 29543 T ACKNOWLEDGMENTS Neisseria gonorrhoeae NCTC 0375 T Neisseria polysaccharea ATCC 43760 T This work was supported by Public Health Service grants DE- Neisseria flavesrens ATCC 13120 T 04881, DE-04898, and DE-08303 from the National Institute of Neisseria maraca ATCC 33926 T Dental Research. Vitreosrilla stercoraria ATCC 15210 T FIG. 6. Phylogenetic tree based on 16s rRNA sequence compar- REFERENCES isons. Bar = 3% difference in nucleotide sequences. The level of 1. Brosius, J., M. L. Palmer, P. J. Kennedy, and H. F. Noller. 1978. difference between two species is determined by measuring the Complete nucleotide sequence of a 16s ribosomal RNA gene lengths of the horizontal lines connecting the species. Vertical from Escherichia coli. Proc. Natl. Acad. Sci. USA 754801- distance has no meaning. 4808. VOL.43, 1993 KINGELLA ORALE SP. NOV. 499

2. Chen, C.-K. C., R G. Dunford, H. S. Reynolds, and J. J. and Eikenella corrodens. Antimicrob. Agents Chemother. 32: Zambon. 1989. Eikenella corrodens in the human oral cavity. J. 765-767. Periodon tol. 60: 611-616. 15. Lacor, M., M. Duarte, A. Beutler, R. Auckenthaler, and S. 3. Chen, C.-K. C., T. V. Potts, and M. E. Wilson. 1990. DNA Suter. 1991. Osteoarticular infections due to Kingella kingae in homologies shared among E. corrodens isolates and other children. Eur. J. Pediatr. 150612-618. corroding bacilli from the oral cavity. J. Periodontal Res. 16. Laemmli, U. K. 1970. Cleavage of structural proteins during the 25: 106-112. assembly of the head of bacteriophage T4. Nature (London) 4. Chen, C.-K. C., and J. J. Zambon. 1992. Unpublished data. 227:680-685. Dewhirst, F. E., and B. J. Paster. 1992. Unpublished data. 5. 17. Lane, D. J., B. Pace, G. J. Olsen, D. A. Stahl, M. L. Sogin, and 6. Dewhirst, F. E., B. J. Paster, and P. L. Bright. 1989. Chro- N. R. Pace. 1985. Rapid determination of 16s ribosomal RNA mobacterium, Eikenella, Kingella, Neisseria, Simonsiella, and sequences for phylogenetic analyses. Proc. Natl. Acad. Sci. l4treoscilla species comprise a major branch of the beta group USA 82:6955-6959. by 16s ribosomal ribonucleic acid sequence comparison: transfer of Eikenella and Simonsiella to the family 18. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R J. Randall. Nekseriaceae (emend.) Int. J. Syst. Bacteriol. 39:258-266. 1951. Protein measurement with the Folin phenol reagent. J. 7. Dewhirst, F. E., B. J. Paster, G. J. Fraser, and S. Barns. 1992. Biol. Chem. 193:265-275. Unpublished data. 19. Pace, B., E. A. Matthews, K. D. Johnson, C. R. Cantor, and 8. Dewhirst, F. E., B. J. Paster, S. La Fontaine, and J. I. Rood. N. R. Pace. 1982. Conserved 5s rRNA complement to tRNA is 1990. Transfer of Kingella indologenes (Snell and Lapage 1976) not required for protein synthesis. Proc. Natl. Acad. Sci. USA to the genus Suttonella gen. nov. as Suttonella indologenes 79: 36-40. comb. nov.; transfer of Bacteroides nodosus (Beveridge 1941) 20. Paster, B. J., and F. E. Dewhirst. 1988. Phylogeny of campylo- to the genus Dichelobacter gen. nov. as Dichelobacter nodosus bacters, wolinellas, Bacteroides gracilis, and Bacteroides ure- comb. nov.; and assignment of the genera Cardiobacterium, olyticus by 16s ribosomal ribonucleic acid sequencing. Int. J. Dichelobacter, and Suttonella to Cardiobacteriaceae fam. nov. Syst. Bacteriol. 3M6-62. in the gamma division of Proteobacteria on the basis of 16s 21. Saitou, N., and M. Nei. 1987. The neighbor-joining method: a rRNA sequence comparisons. Int. J. Syst. Bacteriol. 40:426- new method for reconstructing phylogenetic trees. Mol. Biol. 433. Evol. k406-425. 9. Dewhirst, F. E., B. J. Paster, I. Olsen, and G. J. Fraser. 1992. 22. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular Phylogeny of 54 representative strains of species in the family cloning: a laboratory manual, 2nd ed., p. 11.31. Cold Spring Pasteurellaceae as determined by comparison of 16s rRNA Harbor Laboratory Press, Cold Spring Harbor, N.Y. sequences. J. Bacteriol. 174:2002-2013. 23. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular 10. Dix, K., S. M. Watanabe, S. McArdle, D. I. Lee, C. Randolph, B. cloning: a laboratory manual, 2nd ed., p. 11.39. Cold Spring Moncla, and D. E. Schwartz. 1990. Species-specific oligonucle- Harbor Laboratory Press, Cold Spring Harbor, N.Y. otide probes for the identification of periodontal bacteria. J. 24. Snell, J. J. 1984. Genus IV. Kingella Henriksen and Bsvre 1976, Clin. Microbiol. 28319-323. 449AL,p. 307-309. In N. R. Krieg and J. G. Holt (ed.), Bergey’s 11. Holdeman, L. V., E. P. Cato, and W. E. C. Moore. 1977. manual of systematic bacteriology, vol. 1. The Williams & Anaerobe laboratory manual, 4th ed., p. 125-128. Anaerobe Wilkins Co., Baltimore. Laboratory, Virginia Polytechnic Institute and State University, 25. Vera, H. D., and D. A. Power. 1980. Culture media, p. 965-999. Blacksburg. In E. H. Lennette, A. Balows, W. J. Hausler, Jr., and J. P. 12. Jukes, T. H., and C. R. Cantor. 1969. Evolution of protein Truant (ed.), Manual of clinical microbiology, 3rd ed. American molecules, p. 21-132. In H. N. Munro (ed.), Mammalian protein Society for Microbiology, Washington, D.C. metabolism, vol. 3. Academic Press, Inc., New York. 26. Verdos, N. A. 1984. Genus I. Neisseria Trevisan 1885, 105AL,p. 13. Knapp, J. S., and K. K. Holmes. 1983. Modified oxidation- 290-296. In N. R. Krieg and J. G. Holt (ed.), Bergey’s manual fermentation medium for detection of acid production from of systematic bacteriology, vol. 1. The Williams & Wilkins Co., carbohydrates by Neisseria spp. and Branhamella catarrhalis. Baltimore. J. Clin. Microbiol. 18:56-62. 27. Wesley, I. V., R. D. Wesley, M. Cardella, F. E. Dewhirst, and 14. Knapp, J. S., S. R Johnson, J. M. Zenilman, M. C. Roberts, and B. J. Paster. 1991. Oligonucleotide probes for Campylobacter S. A. Morse. 1988. High-level tetracycline resistance resulting fetus and Camnlobacter hyointestinalk based on 16s rRNA from TetM in strains of Neisseria spp., Kingella denitrificans, sequences. J. Clin. Microbiol. 2918121817.