Susceptibilities of 23 Desulfovibrio Isolates from Humans

Abstract

Antimicrobial susceptibilities of 23 strains of Desulfovibrio spp. were tested by Etest. Generally, Desulfovibrio spp. were highly susceptible to sulbactam-ampicillin, meropenem, clindamycin, metronidazole, and chloramphenicol: MIC₉₀s of 6, 4, 0.19, 0.25, and 8 μg/ml, respectively. In addition, these strains generally showed high MICs to piperacillin and piperacillin-tazobactam. Desulfovibrio fairfieldensis (eight strains) was the species least susceptible to most agents, especially β-lactams, and was the only species resistant to fluoroquinolones. Desulfovibrio desulfuricans strain Essex 6 isolates were less susceptible to β-lactams than D. desulfuricans strain MB isolates.

Desulfovibrio spp. are gram-negative anaerobes and a type of dissimilatory sulfate-reducing bacteria. Most established species of Desulfovibrio are distributed in the environment, but some Desulfovibrio spp. reside in oral cavities and intestinal tracts of animals, including humans (1, 17). In 1996, Tee et al. first reported the isolation of Desulfovibrio species from a blood culture of a patient with cholecystitis and suggested that Desulfovibrio species might act as an opportunistic pathogen (16). Since then, several case reports that suggested Desulfovibrio may be the causative organism have been published (5, 7, 8, 10, 11, 14, 15). In the published case reports, most Desulfovibrio strains were isolated from the blood of patients who suffered from a brain abscess, appendicitis, intra-abdominal abscess, or abdominal wall abscess, while some were isolated from peritoneal fluid or the pus from abscesses of various origins. Gibson et al. have also reported that Desulfovibrio spp. might be associated with ulcerative colitis (4). Furthermore, Langendijk et al. found that some Desulfovibrio spp. may be associated with the early onset of periodontitis, rapidly progressive periodontitis, adult periodontitis, and refractory periodontitis (6).

Presently, four Desulfovibrio spp. (D. fairfieldensis, D. desulfuricans, D. piger, and D. vulgaris) are recognized to be associated with humans (5, 8). They are slow growers and relatively difficult to isolate from clinical specimens by a conventional approach. Identification to the species level without molecular techniques is considerably difficult. Lozniewski et al. tested the susceptibility of 16 clinical isolates of Desulfovibrio spp. and showed the broad MIC range of some antimicrobial agents (9). However, they did not identify their isolates to the species level. Furthermore, Warren et al. tested 18 clinical Desulfovibrio isolates, which were identified to the species level; however, only scattered strains of D. desulfuricans were included in their study (18). Therefore, the information available regarding the antibiogram of human Desulfovibrio isolates is extremely limited and incomplete.

Considering the present situation, it is important and necessary to obtain additional information concerning the antibiogram of Desulfovibrio spp. for the empirical treatment of anaerobic infections in which Desulfovibrio spp. might be involved. Our laboratory has collected Desulfovibrio strains from various human specimens, both clinical and nonclinical, over the past several years. After molecular identification by 16S rRNA sequencing, 13 isolates of D. desulfuricans strains Essex 6 and MB were included in our collection. Therefore, in this study, we performed the Etest to obtain additional knowledge regarding the antimicrobial susceptibilities of Desulfovibrio species.

Twenty-three strains of Desulfovibrio spp. were tested. These strains were isolated from human specimens and identified by classical phenotypic and molecular methods such as 16S rRNA gene sequencing. Specimens from which isolates were obtained and the number of isolates are as follows: mucosal swab of a patient with pouchitis, 6; stool specimens,7; tongue coating, 6; appendicitis, 3; and blood culture, 1. The following four Desulfovibrio strains were used as reference strains: D. fairfieldensis ATCC 700045, D. desulfuricans Essex 6 (ATCC 29577^T), D. desulfuricans MB (ATCC 27774), and D. piger ATCC 29098^T. Bacteroides fragilis ATCC 25285^T was used as the quality control strain. A special agar medium for Desulfovibrio, which was formulated and named “Desulfovibrio agar” (DA) by the author, was also used as a susceptibility test medium. The ingredients of DA were as follows: polypeptone, 15 g; soya-peptone, 7.5 g; yeast extract, 7.5 g; beef extract, 7.5 g; l-cysteine HCl, 0.75 g; ferric ammonium citrate, 0.75 g; dextran sodium sulfate, 10 g; and agar powder, 15 g per liter of distilled water (pH 7.0). In our preliminary experiment, it was confirmed that Desulfovibrio spp. grew more rapidly on the surface of DA than on the surface of Brucella blood agar (BBA) and formed a blackish or black halo around the colonies on this medium due to H₂S production followed by FeS formation. DA was also used for a pour plate method as described below.

Nine antianaerobic antimicrobial agents were used: sulbactam-ampicillin, piperacillin, piperacillin-tazobactam, cefoxitin, cefotaxime, meropenem, clindamycin, chloramphenicol, and metronidazole. Eleven non-antianaerobic antimicrobial agents were also included in this study: ampicillin, cefoperazone, sulbactam-cefoperazone, cephalothin, ceftazidime, cefpirome, erythromycin, minocycline, ciprofloxacin, levofloxacin, and sulfamethoxazole-trimethoprim.

The Etest was first performed according to a commonly used protocol. Namely, the bacterial suspension was inoculated on the surface of BBA by swabbing. The bacterial suspension was adjusted to McFarland no. 1 standard solution. However, this protocol is not always appropriate for testing Desulfovibrio species because achieving confluent growth of the inoculated organisms is difficult on the surface of BBA. Then we used the DA medium. Although Desulfovibrio spp. grew better on the surface of this medium, it was still difficult to achieve confluent growth by swab inoculation. Finally, we used the pour plate method described by Wilkins et al. (19). Briefly, 100 μl of bacterial suspension (McFarland no. 1 standard) was mixed with 20 ml DA medium, which was maintained at 50°C and then poured into petri dishes. When this pour plate method was used, a clear transparent elliptical zone of growth inhibition was observed within 48 h of incubation, with another zone of blackened medium surrounding this clear zone due to H₂S formation. After 48 h of anaerobic incubation (atmosphere of 80 to 85% N₂, 5 to 10% CO₂, and 10% H₂), the MIC was read as the concentration at which the border of the elliptical zone of growth inhibition intersected the scale in the Etest strips. Results for reference strains using the DA agar method were comparable to those of the standard BBA method (data not shown).

β-Lactamase production was tested using the nitrocefin hydrolysis test (Cefinase; Becton-Dickinson Co., Ltd.) according to the manufacturer's instructions.

Regardless of the species, all of the Desulfovibrio spp. tested in this study were susceptible to five antianaerobic agents with low MIC₉₀s, i.e., sulbactam-ampicillin (6 μg/ml), clindamycin (0.19 μg/ml), meropenem (4 μg/ml), metronidazole (0.75 μg/ml), and chloramphenicol (8 μg/ml). On the other hand, these strains showed high MIC₉₀s toward the other antianaerobic agents: piperacillin (>256 μg/ml), piperacillin-tazobactam (>256 μg/ml), cefoxitin (>256 μg/ml), and cefotaxime (>256 μg/ml) (Table 1). The high susceptibilities of Desulfovibrio spp. to clindamycin, metronidazole, chloramphenicol, and imipenem have been noted in previously published reports (9, 18). In this study, one D. fairfieldensis strain that was intermediately resistant to meropenem was recognized. With regard to the susceptibility of Desulfovibrio spp. to carbapenems, previous reports have demonstrated their uniform susceptibility to imipenem (9, 18). However, another report demonstrated that D. fairfieldensis strains were less susceptible to ertapenem, with fairly high MIC₉₀ (>32 μg/ml) and MIC₅₀ (>32 μg/ml) values (18). Further studies are required to determine the susceptibilities of D. fairfieldensis strains to an array of carbapenems, which are now very often chosen as empirical treatment for serious infections.

TABLE 1.

Susceptibilities of Desulfovibrio isolates from humans to 20 antimicrobial agents

Species and antimicrobial agent (no. of isolates)	MIC (μg/ml)a
	Range	50%	90%
Desulfovibrio spp. (n = 23)
Antianaerobic
Sulbactam-ampicillin	0.064-6	0.38	6
Piperacillin	32->256	>256	>256
Piperacillin-tazobactam	16->256	>256	>256
Cefoxitin	16->256	>256	>256
Cefotaxime	0.047->256	1.5	>256
Meropenem	0.023-12	0.19	4
Clindamycin	0.016-0.25	0.125	0.19
Chlorampenicol	1.5-12	4	8
Metronidazole	<0.016-1.5	0.125	0.25
Others
Ampicillin	0.19-24	0.75	8
Cefoperazone	12->256	>256	>256
Sulbactam-cefoperazone	1->256	4	>256
Cefalothin	16->256	>256	>256
Ceftazidime	2->256	6	>256
Cefpirome	0.5->256	4	>256
Erythromycin	0.064-4	0.5	1.5
Minocycline	0.19-24	6	24
Ciprofloxacin	0.125->32	0.5	>32
Levofloxacin	0.094->32	0.75	>32
Sulfamethoxazole-trimethoprim	>32	>32	>32
D. fairfieldensis (n = 8)b
Antianaerobic
Sulbactam-ampicillin	4-6
Piperacillin	>256
Piperacillin-tazobactam	>256
Cefoxitin	64->256
Cefotaxime	0.047->256
Meropenem	2-12
Clindamycin	0.032-0.19
Chloramphenicol	3-6
Metronidazole	<0.016-0.38
Others
Ampicillin	4-8
Cefoperazone	>256
Sulbactam-cefoperazone	24->256
Cefalothin	>256
Ceftazidime	>256
Cefpirome	>256
Erythromycin	0.5-4
Minocycline	4-24
Ciprofloxacin	>32
Levofloxacin	>32
Sulfamethoxazole-trimethoprim	>32
D. desulfuricans Essex 6 (n = 7)c
Antianaerobic
Sulbactam-ampicillin	0.125-1.5
Piperacillin	>256
Piperacillin-tazobactam	16->256
Cefoxitin	>256
Cefotaxime	0.047-6
Meropenem	0.064-0.25
Clindamycin	0.094-0.25
Chloramphenicol	1.5-12
Metronidazole	0.064-1.5
Others
Ampicillin	0.19-2
Cefoperazone	16->256
Sulbactam-cefoperazone	3->256
Cefalothin	3->256
Ceftazidime	3->256
Cefpirome	0.5-32
Erythromycin	0.38-0.75
Minocycline	2-12
Ciprofloxacin	0.19-2
Levofloxacin	0.25-3
Sulfamethoxazole-trimethoprim	>32
D. desulfuricans MB (n = 6)d
Antianaerobic
Sulbactam-ampicillin	0.064-0.25
Piperacillin	32->256
Piperacillin-tazobactam	32-96
Cefoxitin	16->256
Cefotaxime	0.5-1.5
Meropenem	0.023-0.064
Clindamycin	0.094-0.19
Chloramphenicol	1.5-4
Metronidazole	0.016-0.38
Others
Ampicillin	0.19-0.38
Cefoperazone	12-24
Sulbactam-cefoperazone	1-3
Cefalothin	2-6
Ceftazidime	2-4
Cefpirome	0.5-2
Erythromycin	0.38-1
Minocycline	3-8
Ciprofloxacin	0.125-0.38
Levofloxacin	0.25-0.5
Sulfamethoxazole-trimethoprim	>32
D. piger (n = 2)e
Antianaerobic
Sulbactam-ampicillin	0.5-1
Piperacillin	>256
Piperacillin-tazobactam	>256
Cefoxitin	16
Cefotaxime	1-1
Meropenem	0.032-0.19
Clindamycin	0.016-0.016
Chloramphenicol	1.5-3
Metronidazole	0.032-0.5
Others
Ampicillin	12-32
Cefoperazone	12-32
Sulbactam-cefoperazone	1-1.5
Cefalothin	96->256
Ceftazidime	3-3
Cefpirome	0.75-4
Erythromycin	0.016-0.125
Minocycline	0.19-1
Ciprofloxacin	0.125-0.19
Levofloxacin	0.095-0.125
Sulfamethoxazole-trimethoprim	>32

^a50% and 90%, MIC₅₀ and MIC₉₀, respectively.

^bOne isolate from pouchitis, five from stool specimens, and two from tongue coating.

^cOne isolate from appendicitis, two from stool specimens, and four from tongue coating.

^dOne isolate from blood, three from pouchitis, and two from appendicitis.

^eBoth isolates from pouchitis.

Fifteen Desulfovibrio strains, including 8 D. fairfieldensis strains and 7 D. desulfuricans Essex 6 isolates, were highly resistant to both piperacillin and cefoxitin: MICs of >256 μg/ml for piperacillin and 64 to >256 μg/ml for cefoxitin, respectively. D. desulfuricans MB isolates were susceptible or intermediately susceptible to piperacillin and cefoxitin, except for one resistant isolate. Most of the D. fairfieldensis strains were cefotaxime resistant (MIC range, >265 μg/ml), but the other species showed low MICs to cefotaxime (MIC range, 0.5 to 6 μg/ml).

All of the Desulfovibrio strains tested showed relatively low MIC₉₀s to erythromycin (MIC₉₀, 1.5 μg/ml), slightly higher MIC₉₀s to ampicillin (MIC₉₀, 8 μg/ml) and minocycline (MIC₉₀, 24 μg/ml), and consistently higher MIC₉₀s to sulfamethoxazole-trimethoprim (MIC₉₀, >32 μg/ml). Their susceptibilities to the remaining seven non-antianaerobic agents were strain dependent with broad MIC ranges.

Eight D. fairfieldensis strains were obviously highly resistant to four cephems (cephalothin, cefoperazone, ceftazidime, and cefpirome), sulbactam-cefoperazone, and two fluoroquinolones (ciprofloxacin and levofloxacin). This study, together with the previous report, demonstrated that among three species of the genus Desulfovibrio, D. fairfieldensis was the species least susceptible to antimicrobial agents. D. fairfieldensis strains from our collections were highly resistant to narrow- and broad-spectrum cephems (cephalothin, cefoxitin, cefoperazone, cefotaxime, ceftazidime, and cefpirome) with no detectable β-lactamase by the nitrocefin tests. The mechanism underlying the trend of resistance of this species to β-lactams remains to be elucidated. D. desulfuricans Essex 6 isolates showed relatively lower MICs to fluoroquinolones (0.19 to 3 μg/ml). D. desulfuricans Essex 6 isolates were more resistant to β-lactam antibiotics than were D. desulfuricans MB isolates. Morin et al. have demonstrated that some D. desulfuricans strains possess a β-lactamase gene (bla_DES-1) and suggested that the gene is associated with a certain subtype of D. desulfuricans, although they did not discriminate D. desulfuricans Essex 6 from D. desulfuricans MB isolates in their study (12). In our study, it is notable that all four β-lactamase producers from among the D. desulfuricans strains were the D. desulfuricans Essex 6 isolates. Therefore, the existence of bla_DES-1 in our D. desulfuricans strains should be investigated.

The conventional Etest method is simple and easy to perform and may be suitable for all anaerobes, including some slow growers (2, 3). Desulfovibrio spp. have a tendency to form tiny, viscous, and pitting colonies on the agar surface: some modification was required to obtain a good performance of the Etest. We proposed to use the DA agar medium and to incorporate the inocula in the pour plate instead of swabbing them. In these bacterium-impregnated agar plates, after only 2 days of incubation, all of the Desulfovibrio strains grew homogeneously and formed clear, black-edged, elliptical zones of inhibition. The reproducibility of the MIC determinations by this method was well acceptable, since the MICs were identical or varied within a twofold dilution range for several strains tested, including for the reference strains. The MICs obtained for the quality control strains (B. fragilis ATCC 25285) were within the ranges of the CLSI reference values (13; data not shown).

It is essential to increase research efforts to isolate Desulfovibrio strains from clinical specimens in order to establish a more useful antibiogram of the Desulfovibrio species. However, in the present situation, in addition to the published information concerning the antimicrobial susceptibilities of Desulfovibrio, the findings of this study may also be useful for clinicians in treating patients with Desulfovibrio bacteremia and patients with mixed infections associated with endogenous Desulfovibrio spp.