CDC Group O-3: Phenotypic Characteristics, Fatty Acid Composition, Isoprenoid Quinone Content, and In Vitro Antimicrobic Susceptibilities of an Unusual Gram-Negative Bacterium Isolated from Clinical Specimens

Abstract

Between 1983 and 1994, 13 phenotypically similar unidentified clinical isolates were received by the Special Bacteriology Reference Laboratory, Centers for Disease Control and Prevention (CDC). Sources included blood (four strains), lung (three strains), knee fluid and duodenal tissue (one strain each), bone, and lymph node tissue (two strains each). All were aerobic glucose-oxidizing, slender, long, curved gram-negative rods that utilized xylose, sucrose, and maltose; did not grow on MacConkey agar in 1 to 2 days; were oxidase positive; hydrolyzed esculin; and grew on Campylobacter selective medium. All were negative for urease, indole, nitrate reduction, and gelatin hydrolysis. All were motile by means of a single polar flagellum with a noticeably short wavelength; however, motility was sometimes difficult to demonstrate. The cellular fatty acid compositions of these strains, as analyzed by gas-liquid chromatography, were unique, characterized by relatively large amounts of 16:1ω7c, 16:0, and 18:1ω7c with smaller amounts of 12:0, 3-OH-12:1, 14:0, 15:0, 18:0, Br-19:1, and 19:0cyc^11–12. High-performance liquid chromatography and mass spectrometry of the quinone extracts of three representative strains showed ubiquinone-10 as the major component. Based on the breakpoints for the family Enterobacteriaceae, all the strains were susceptible in vitro to aminoglycosides, sulfamethoxazole-trimethoprim, and chloramphenicol but were resistant to most beta-lactams except imipenem. The MICs of amoxicillin-clavulanate and ciprofloxacin for these strains clustered around the breakpoints, which makes it difficult to predict the strains’ response in vivo to these agents. This group has been designated CDC oxidizer group 3 (O-3).

In recent years the evolution of human disease, in conjunction with the continuing improvement in detection and identification of bacterial pathogens, has yielded numerous previously unrecognized agents. Some of these agents include new Bartonella and Bordetella species; the new genera Afipia, Roseomonas, and Balneatrix; and other currently unclassified isolates (1–3, 8, 11–17). In 1983, the Special Bacteriology Reference Laboratory of the Centers for Disease Control and Prevention (CDC) received the first of 13 clinical isolates of unusual gram-negative, glucose-oxidizing bacilli that shared a unique phenotypic and cellular fatty acid (CFA) profile. These isolates were from male and female patients and from a variety of sources, including blood, bone, knee fluid, pulmonary, and duodenal and lymph node tissues.

Presented herein is a polyphasic study of this group, to which the provisional name CDC oxidizer group 3 (O-3) has been assigned. Included within this study are morphologic, biochemical, CFA, isoprenoid quinone, and in vitro antimicrobial susceptibility determinations.

MATERIALS AND METHODS

Bacterial strains.

The O-3 isolates studied, along with their sources, geographic origins, and submission dates, are presented in Table 1. All strains were stored as suspensions in defibrinated rabbit blood in liquid nitrogen. Unless otherwise indicated, the strains were cultured on heart infusion agar supplemented with 5% rabbit blood (RBA) (BBL prepared media; BBL Microbiology Systems, Cockeysville, Md.) and incubated at 35°C in a candle jar.

TABLE 1.

Sources and clinical information for CDC group O-3 strains in this study

Strain no.a	Source	Sender locationb	Date received (mo-day-yr)	Sexc, age of patient (yr)	Clinical history of patient
F4739	Pulmonary nodule	NE	8-24-83	F	NAd
F6758	Endotracheal tube aspirate	MD	5-15-85	M	Pneumonia, AIDS
G185	Tissue and bone	CA	6-8-87	F, 92	NA
G1067	Duodenal tissue	MD	3-3-88	M	Duodenal ulcer
G1747	Knee aspirate	TX	9-12-88	M, 1	Joint sepsis
G7248	Blood	GA	3-6-92	M, 58	NA
G7521	Blood	GA	5-22-92	F	NA
G7522	Blood	GA	5-22-92	F	NA
G8743e	Lymph node	NJ	10-12-93	F, 21	NA
G8767e	Lymph tissue	NJ	11-1-93	F, 34	NA
G8822e	Sternum	NJ	12-2-93	M, 41	Mediastinitis
G9220	Blood	GA	9-15-94	F, 58	NA
G9273	Lung	AL	10-18-94	F, 55	NA

^aCDC Special Bacteriology Reference Laboratory number. 

^bNE, Nebraska; MD, Maryland; CA, California; TX, Texas; GA, Georgia; NJ, New Jersey; AL, Alabama. 

^cF, female; M, male. 

^dNA, not available. 

^eIsolated from patients in the same hospital. 

Phenotypic tests.

Biochemical testing was done by the methods of the CDC Special Bacteriology Reference Laboratory (17). The flagellum staining was performed by the Ryu stain method of Kodaka et al. (5), using a commercially available reagent (Carr-Scarborough Microbiologicals, Decatur, Ga.). With the exception of those for oxidase, catalase, growth temperature, and gelatin digestion, all biochemical tests were performed at 35°C in an aerobic incubator. The oxidase, catalase, and growth temperature tests were determined at 1 day and gelatin digestion was determined at 7 and 14 days of incubation. Five isolates were tested for growth on Trypticase soy agar with 5% sheep blood (BBL). Twelve isolates were tested for growth on Campy selective CVA media (BBL) at 35°C in an atmosphere of 5% hydrogen, 10% CO₂, and 85% nitrogen. The growth of these isolates was compared to their growth on RBA at 35°C in a candle jar atmosphere.

CFA analysis.

Cells were saponified, and the liberated fatty acids were methylated and analyzed by capillary gas-liquid chromatography (GLC) (17). The amide-linked hydroxy acids that were not totally released by this saponification procedure were completely released by a subsequent acid hydrolysis of the methanolic aqueous layer after the methylation step (17). The identification of fatty acids and the determination of double bond positions in monounsaturated acids were accomplished by GLC and GLC-mass spectrometry (GLC-MS). The confirmation of hydroxy acids was accomplished by both acetylation and GLC-MS analysis, as described previously (17). Isoprenoid quinones were extracted from 100 mg of lyophilized cells and were analyzed by reverse-phase high-performance liquid chromatography and MS (6, 7).

In vitro antimicrobial susceptibility tests.

Antimicrobial susceptibility profiles were determined by the broth microdilution method described by the National Committee for Clinical Laboratory Standards (NCCLS) (9), except that results were read after a 24-h incubation. The strains were streaked onto Trypticase soy agar with 5% sheep blood (BBL) and incubated for 18 to 24 h at 35°C in ambient air. Growth was taken from the plate with a sterile cotton-tipped swab and suspended in a tube of unsupplemented cation-adjusted Mueller-Hinton broth (Difco, Detroit, Mich.) to a density equivalent to a McFarland standard of 1.0. Eight-milliliter aliquots of the broth suspensions were added to 32 ml of sterile distilled water and inoculated into MIC plates with the MIC 2000 (Dynatech Laboratories, Chantilly, Va.). Microtiter MIC plates were prepared at the CDC according to the recommendations of the NCCLS (9) (final volume per well, 100 μl). The plates were incubated in ambient air for 24 h at 35°C, with a final inoculum concentration of approximately 5 × 10⁵ CFU/ml. The following organisms were used as controls: Escherichia coli ATCC 25922, E. coli ATCC 35218, Pseudomonas aeruginosa ATCC 27853, and Enterococcus faecalis ATCC 29212.

RESULTS

As indicated in Table 1, all O-3 strains were received from laboratories within the United States. The first was received in 1983, and most of the others were received after 1991. Isolates were obtained from a variety of sources, including blood, respiratory, bone, joint, and lymphatic tissues. Patients from whom these isolates were obtained ranged in age from 1 to 92 years, and no significant preference for either gender was noted. Clinical information was available on five patients, and no common underlying syndrome was identified.

The phenotypic characteristics of the O-3 strains in this study are presented in Table 2. Cells grown on heart infusion agar at 35°C for 18 to 24 h were thin, medium to slightly long curved rods with tapered ends (sickle-like) that sometimes formed rosettes (Fig. 1). All the strains grew slightly to moderately well on RBA that was incubated either aerobically or in a candle jar atmosphere for 18 to 24 h. Isolated colonies were circular, entire, translucent, and very punctate. Frequently, no hemolytic reaction was observed on RBA after overnight incubation at 35°C; however, five strains produced a faint green discoloration. Five strains were also tested for growth on Trypticase soy agar supplemented with 5% sheep blood, and all grew at rates and with morphologies similar to those obtained on RBA. All 13 strains were aerobic glucose oxidizers. All the strains produced acid from d-xylose, sucrose, and maltose; were oxidase positive; and hydrolyzed esculin. When examined at 2 days of incubation, all the strains were negative for growth on MacConkey agar; however, growth was detected at 3 to 7 days for five strains. All were negative for urease, indole, nitrate reduction, and gelatin hydrolysis. Four strains reduced 0.01% nitrite without gas formation. All the strains were motile by means of a single polar flagellum with a noticeably short wavelength (Fig. 2); however, motility was sometimes difficult to demonstrate.

TABLE 2.

Phenotypic characteristics of CDC group O-3 and similar taxa^a

Test performed	% Positiveb
	CDC oxidizer group			Agrobacterium yellow group (2 strains)	Sphingomonas speciesc (134 strains)
	O-3 (13 strains)	O-1 (62 strains)	O-2 (66 strains)
Motility	100	100	100	100	100
Acid from:
d-Glucose	100	69 (31)	73 (11)	(100)	93 (7)
d-Xylose	100	0	2	50 (50)	96 (4)
d-Mannitol	0	0	2	0	0
Lactose	0	0	2	(100)	93 (7)
Sucrose	100	0	64 (36)	(100)	93 (7)
Maltose	100	0	71 (27)	100	97 (3)
Catalase	23; 62w	98	91	100	95
Oxidase	100	77	97	100	75
Growth on MacConkey agar	(38)	6 (40)	5 (5)	(50)	10 (13)
Simmons citrate	0	0	0	0	6 (1)
Urea (christensen’s)	0	2	12	(100)	6 (3)
Nitrate reduction	8	0	15	0	0
Gas from nitrate	0	0	0	0	0
Indole	0	0	0	0	0
Gelatin hydrolysis	0	25	38	0	2
Yellow growth pigment	0	100	100	100	95
Growth in nutrient broth with:
0% NaCl	100	34	92	50 (50)	93
6% NaCl	0	0	22	0	0
Growth temperature
25°C	92	90	89	100	100
35°C	100	100	100	100	100
42°C	40	24	31	0	24
Esculin hydrolysis	92 (8)	93 (2)	64	50 (50)	91
Lysine decarboxylase	0	0	0	0	0
Arginine dihydrolase	0	4	22	0	8
Ornithine decarboxylase	0	0	6	0	0
3-Ketolactonate	ND	ND	ND	100	0

^aData from reference 11. 

^bUnless otherwise indicated, data are given as percent positive at 48 h of incubation; numbers in parentheses are percent positive at 3 to 7 days of incubation. Abbreviations: w, weak reaction; ND, not determined. Catalase and growth temperature were read at 24 h of incubation; gelatin hydrolysis was read after 14 days of incubation; 3-ketolactonate reactions were developed with 7-day cultures. 

^cPhenotypically similar to S. paucimobilis and S. parapaucimobilis. 

FIG. 1.

CDC group O-3 strain G8822 with Gram stain, grown in heart infusion agar at 35°C for 24 h. Magnification, ×3,000.

FIG. 2.

CDC group O-3 strain G8822 with flagellum stain, grown in tryptone glucose yeast extract medium at 25°C for 24 h. Magnification, ×3,000.

Table 2 also contains phenotypic results for four similar glucose-oxidizing gram-negative species and CDC groups. The “Agrobacterium yellow group” was first described in 1985, and CDC groups O-1 and O-2 were first described in 1996 (12, 17). The strains designated Sphingomonas species represent isolates that phenotypically resemble Sphingomonas paucimobilis and Sphingomonas parapaucimobilis (17). O-3 is the only group in which a yellow growth pigment is not produced and the only group of predominately curved rods. Other tests useful in differentiating these taxa include those for oxidation of d-xylose, lactose, sucrose, and maltose; Christensen’s urea hydrolysis; and 3-ketolactonate production.

To determine the growth ability of the O-3 group on Campylobacter selective medium, 12 isolates were tested for growth on Campy CVA. The results were then compared to their growth on RBA. All the isolates grew as well or, in most cases, better on the Campy CVA plates under microaerophilic conditions.

The CFA composition of the O-3 group is presented in Table 3. All 13 strains shared a unique profile characterized by relatively large amounts of 16:1ω7c (8 to 22%), 16:0 (15 to 32%), and 18:1ω7c (32 to 50%) with smaller amounts (1 to 10%) of 12:0, 3-OH-12:1, 14:0, 15:0, 18:0, Br-19:1, and 19:0cyc^11–12. In addition, 0 to 2% amounts of 3-OH-12:0, i-15:0, 2-OH-14:0, i-17:0, 17:1ω8c, 17:1ω6c, 17:0, 18:2, 18:1ω9c, and 20:4 were observed in the CFA profile of the O-3 group. Upon acid hydrolysis of two representative strains, additional 1% amounts each of 3-OH-12:0 and 2-OH-14:0 were released, indicating that these acids are both ester and amide linked. No additional hydroxy acids were released upon acid hydrolysis. High-performance liquid chromatographic and MS data of the quinone extracts of three representative strains confirmed ubiquinone-10 (Q-10) as the major component, constituting approximately 90 to 100% of the total ubiquinone content. The quinone content of the O-3 group is most similar to that of Afipia species in that Q-10 is essentially the only quinone present, as studied by our methods (8). Therefore, the presence of Q-10, in combination with the described CFA profile, provides a rapid chemical identification of the O-3 isolates.

TABLE 3.

CFA composition of CDC group O-3

Fatty acida	Amtb
12:0	1–6
3-OH-12:1	1–5
3-OH-12:0	0–1
14:0	1–5
i-15:0	0–1
15:0	1–5
2-OH-14:0	0–1
16:1ω7c	8–22
16:0	15–32
i-17:0	0–T
17:1ω8c	0–1
17:1ω6c	0–1
17:0	0–1
18:2	0–2
18:1ω9c	0–2
18:1ω7c	32–50
18:0	1–4
Br-19:1	2–10
19:0cyc	1–6
20:4	0–T

^aThe number before the colon indicates the number of carbons; the number after the colon is the number of double bonds; ω, the position of the double bond counting from the hydrocarbon end of the carbon chain; c, cis isomer; OH, a hydroxy group at the 2(α)- or 3(β)- position from the carboxyl end; i, iso; cyc, a cyclopropane ring structure at the 11-12 position; Br, branched; and T, 0.4 to 0.8%. 

^bValues are percentage of total fatty acids and are arithmetic ranges. 

The results of antimicrobial susceptibility testing of all O-3 isolates against 19 antimicrobial agents are shown in Table 4. Based on NCCLS interpretative standards for the family Enterobacteriaceae (10), all the isolates tested were susceptible to the aminoglycosides (amikacin, gentamicin, and tobramycin), trimethoprim-sulfamethozaxole, and imipenem. Twelve of the 13 isolates were susceptible to chloramphenicol, with one isolate (G8743) yielding an intermediate MIC of 16 μg/ml. All the isolates, with the exception of G185, were resistant to most of the beta-lactams, including ampicillin, cefazolin, cefoxitin, cefotaxime, ceftriaxone, cefotetan, ceftazidime, and aztreonam, although clavulanic acid did appear to inhibit this activity. Resistance to tetracycline was variable. The ciprofloxacin and amoxicillin-clavulanate MICs clustered around the breakpoints, which makes the in vivo responses of these organisms to these antimicrobial agents difficult to predict.

TABLE 4.

Broth microdilution MICs for 13 isolates of CDC group O-3

Antimicrobial agent	MIC (μg/ml)
	Range	MIC50a	MIC90a
Amikacin	≤1	≤1	≤1
Amoxicillin-clavulanateb	4/2–64/32	16/8	16/8
Ampicillin	16–>64	>64	>64
Aztreonam	64–>64	>64	>64
Cefazolin	>32	>32	>32
Cefotaxime	4–>64	64	>64
Cefotetan	>64	>64	>64
Cefoxitin	32–>32	32	32
Cefpodoxime	4–>16	>16	>16
Ceftazidime	32–128	64	128
Ceftriaxone	8–>64	>64	>64
Chloramphenicol	≤1–16	≤1	4
Ciprofloxacin	1–>8	4	8
Gentamicin	≤0.25	≤0.25	≤0.25
Imipenem	≤1–2	2	2
Piperacillin	16–>128	64	128
Tetracycline	≤1–32	4	8
Tobramycin	≤0.25	≤0.25	≤0.25
Trimethoprim-sulfamethoxazoleb	≤0.12/2.3–2/38	≤0.12/2.4	0.5/9.6

^aMIC₅₀ and MIC₉₀, MICs at which 50 and 90%, respectively, of the isolates are inhibited. 

^bShown as MIC of first drug/MIC of second drug. 

DISCUSSION

CDC group O-3 represents 13 similar clinical isolates received by the Special Bacteriology Reference Laboratory between 1983 and 1994. The patient information submitted with these isolates indicates that this group may be isolated from males and females of all ages. Although these isolates often were obtained from sites associated with invasive disease (e.g., blood and lymph nodes), the limited amount of clinical information received makes it difficult to estimate their clinical significance. Additional studies are needed to better understand their pathogenic potential.

The duodenal tissue isolate is particularly interesting in that it was obtained from a patient with a diagnosis of duodenal ulcer and was submitted to CDC as a Campylobacter species. This suggests a potential for misidentification of the O-3 organism as a member of the genus Campylobacter in the clinical laboratory, especially since it grows well in Campylobacter selective medium.

O-3 strains are characterized by distinctive biochemical and CFA profiles, as well as cellular and flagellar morphology. However, the O-3 biochemical profile shares some similarities with other unclassified groups, including CDC groups O-1 and O-2, and the Agrobacterium yellow group (12, 17). Tests useful in differentiating among these groups include those for cellular morphology; oxidation of lactose, d-xylose, sucrose, and maltose; urea hydrolysis; 3-ketolactonate production; flagellar morphology; and presence of a yellow growth pigment. CFA analysis is also useful in differentiating the three CDC oxidizer groups and Sphingomonas. The utility of CFA analysis in identifying Agrobacterium yellow group strains will be better understood when more strains become available.

The antimicrobial susceptibility profiles of these isolates show a high level of in vitro resistance to many of the commonly used broad spectrum agents, including most of the beta-lactam drugs. The results for ciprofloxacin cluster around the breakpoint. In light of these findings, the clinical significance of O-3 isolates should be carefully assessed in determining appropriate antimicrobial therapy. At this time there is very little information available on the natural pathogenicity of these organisms.

The ultimate taxonomic classification of the O-3 group will require the use of molecular techniques, such as 16S rRNA sequencing and DNA-DNA hybridization analysis. At this time, we do not know whether the O-3 group represents one or more species. Some Special Bacteriology Reference Laboratory groups, such as DF-2 (now Capnocytophaga canimorsus) (2) and NO-2 (now Bordetella holmesii) (15), have been shown to be genetically homogeneous, whereas other groups, such as WO-1 and Pink Coccoid I through IV, have represented multiple species (4, 11, 14). We plan to initiate molecular taxonomic studies of these strains in the near future; however, the findings of this study should assist clinical microbiologists in the identification of this group, which will ultimately provide an increased understanding of its true clinical significance.