Discordant Carbapenem Susceptibility in Methylobacterium Species and Its Application as a Method for Phenotypic Identification

We encountered a meropenem-resistant, imipenem-susceptible clinical Methylobacterium isolate and sought to determine if discordant carbapenem susceptibility (meropenem resistance/imipenem susceptibility) was common among different Methylobacterium species. We concurrently investigated whether this phenotype was expressed in the genus Roseomonas, the other recognized group of pink-pigmented nonfermenting gram-negative bacilli.

The genus Methylobacterium is comprised of pink-pigmented nonfermenting gram-negative bacilli. These organisms are widely distributed in nature and have been isolated from chlorinated potable water supplies (6). Their presence and role as opportunistic pathogens and presence in the nosocomial setting are well documented (1, 4, 5, 7, 9, 10, 14).

Isolates can be cultivated on a variety of solid media, but small colonies may only be detected after incubation for 4 to 5 days (15). Optimal growth occurs at 25 to 30°C; most strains fail to grow at 37°C. Biochemical characterization allows definitive identification and differentiation from other pink-pigmented nonfermenters (15). 16S rRNA-based molecular studies have allowed phylogenetic analysis and rapid identification of Methylobacterium species (6, 13).

Current interpretive standards for antibiotic susceptibility testing do not apply to Methylobacterium species (12). Two large Methylobacterium studies assessing in vitro antibiotic susceptibility have demonstrated broad resistance to beta-lactam antibiotics (2, 6). To date, isolates tested have been imipenem susceptible (1, 7, 9). No data with regard to meropenem susceptibility has been published to our knowledge.

Four clinical isolates from our laboratory were identified as Methylobacterium species, by morphologic and biochemical characteristics, at a reference laboratory (Laboratoire de Santé Publique du Québec). We also acquired three Methylobacterium species and three Roseomonas species from the American Type Culture Collection (ATCC; Table 1).

TABLE 1.

In vitro carbapenem susceptibility of Methylobacterium and Roseomonas spp.^a

Isolate	Medium	Imipenem		Meropenem
		Mean zone inhibition (mm)	Mean MIC (μg/ml)	Mean zone inhibition (mm)	Mean MIC (μg/ml)
Methylobacterium spp.
Isolate no. 1	SDA	44	0.75	6	>32
	NA	52	0.25	6	>32
Isolate no. 2	SDA	48	0.5	6	>32
	NA	52	0.25	6	>32
Isolate no. 3	SDA	48	1	6	>32
	NA	50	0.38	6	>32
Isolate no. 4	SDA	50	0.75	6	>32
	NA	48	0.38	6	>32
Isolate no. 5 (ATCC 29983)	SDA	58	0.25	6	>32
	NA	64	0.064	6	>32
Isolate no. 6 (ATCC 43645)	SDA	41	0.5	6	>32
	NA	52	0.125	6	>32
Isolate no. 7 (ATCC 27886)	SDA	46	0.38	6	>32
	NA	50	0.125	6	>32
Roseomonas spp.
Roseomonas genomospecies 5 (ATCC 49960)	SDA	63	0.032	68	0.006
	NA	64	0.012	69	0.004
R. cervicalis (ATCC 49957)	SDA	39	0.75	50	0.023
	NA	44	0.094	54	0.012
R. gilardi (ATCC 49956)	SDA	55	0.047	47	0.047
	NA	58	0.032	54	0.032

^a Susceptibility was determined by disk diffusion (mean zone inhibition) and E-test (mean MIC) methods on SDA and NA at 25°C for 72 h. Mean MICs from duplicate experiments are rounded up to the nearest E-test scale value. 

Susceptibility testing was performed by E-test (imipenem and meropenem; AB BIODISK) and disk diffusion (imipenem and meropenem, 10-μg disks; OXOID) on Sabouraud dextrose agar (SDA) and nutrient agar (NA). E-test and disk diffusion methods were carried out as described in the manufacturer's instructions and published protocol (11), respectively.

Incubation of plates at 25°C for 72 h allowed all isolates to attain optimal growth for susceptibility determination. All Methylobacterium species isolates were extremely susceptible to imipenem but highly resistant to meropenem (Table 1). All Roseomonas isolates were extremely susceptible to both carbapenems.

In vitro, meropenem is more active than imipenem against members of the family Enterobacteriaceae and most oxidase-positive and/or glucose-nonfermenting gram-negative bacilli (3, 8). The marked discordance between meropenem and imipenem susceptibility that we describe has not to our knowledge been observed in other gram-negative bacilli.

We assessed three distinct Methylobacterium reference strains. 16S rRNA-based phylogenetic analysis has previously placed these species within two major subclusters (subcluster I, M. extorquens and M. organophilum; subcluster II, M. mesophilicum) (6). Given the interspecies consistency of the susceptibility phenotype we describe, it seems plausible that the mechanisms contributing to this phenotype are conserved at the genus level. Further in vitro study is needed to characterize the underlying mechanism for this discordant susceptibility.

Our findings could contribute to more rapid identification of clinical Methylobacterium isolates and conceivably could be exploited for screening potable water supplies and environmental samples. Additionally, our data suggests that discordant carbapenem susceptibility may have the potential to differentiate between Methylobacterium and Roseomonas species. However, our observations include a small number of isolates. Further studies are necessary to confirm our findings and to elucidate the responsible mechanism.