Abstract
The genus Ochrobactrum currently comprises nine species, Ochrobactrum anthropi (Holmes et al., 1988), Ochrobactrum intermedium (Velasco et al., 1998), Ochrobactrum lupini (Trujillo et al., 2005), Ochrobactrum tritici and Ochrobactrum grignonense (Lebuhn et al., 2006), Ochrobactrum gallinifaecis (Kämpfer et al., 2003), Ochrobactrum oryzae (Tripathi et al., 2006), Ochrobactrum pseudintermedium (Teyssier et al., 2007) and Ochrobactrum cytisi (Zurdo-Piñeiro et al., 2007). Only three species, O. anthropi, O. intermedium and O. pseudintermedium, have been reported to occur in clinical material (Alnor et al., 1994; Appelbaum & Campbell, 1980; Barson et al., 1987; Braun et al., 1996; Brivet et al., 1993; Cieslak et al., 1992, 1996; Earhart et al., 1997; Greven & Nelson, 2001; Inoue et al., 1999; Jelveh & Cunha, 1999; Kettaneh et al., 2003; Mahmood et al., 2000; Perez-Blanco et al., 2005; Romero-Gomez et al., 2004; Vaidya et al., 2006; Wheen et al., 2002; Teyssier et al., 2007).
O. anthropi was described by Holmes et al. (1988) to accommodate oxidase-positive, Gram-negative, non-fermentative bacilli previously known as Achromobacter group Vd, occasionally infecting humans. Today O. anthropi is recognized as an emerging pathogen of low virulence, predominantly isolated from immunocompromised patients, but there are also 19 cases of it being isolated from immunocompetent hosts (Vaidya et al., 2006). Only two cases of O. intermedium infection have been reported. In 1999, O. intermedium was isolated from blood and a liver abscess in a patient after liver transplantation (Möller et al., 1999), and in 2005, a presumptive systemic O. intermedium infection in a patient with bladder cancer was reported (Apisarnthanarak et al., 2005). Recently a novel species, O. pseudintermedium, was described that was isolated from patients hospitalized in intensive-care units (Teyssier et al., 2007). The differentiation of Ochrobactrum species by physiological tests is quite difficult, because of their high phenotypic similarity (Table 1). O. anthropi is often the only Ochrobactrum species present in the databases of commercially available identification systems (including the API 20NE kit) (Teyssier et al., 2005a) and as a consequence O. intermedium may be misidentified as O. anthropi as occurred in a recently published case report describing O. anthropi as the causative agent of a pelvic abscess in an immunocompetent host (Vaidya et al., 2006). The initial identification of that isolate (RMA 16449) as O. anthropi (Vaidya et al., 2006) was performed with the API 20E kit (bioMérieux) and by partial 16S rRNA gene sequencing. Strain RMA 16449 was later included in comparative analysis to study the genetic diversity among 112 Ochrobactrum strains by applying a polyphasic approach using extended 16S rRNA gene analysis, recA gene sequencing, fatty acid analysis and DNA–DNA hybridization studies. Of the 16S rRNA gene, 1390 bp were sequenced as described by Kämpfer et al. (2003). In the BLASTN search using the sequence information from the entire 1390 bp fragment (GenBank accession no. AM409326), the highest similarity of 100 % was indeed detected with O. anthropi strain CCUG 39736 (GenBank accession no. AM114408). However, to add to the confusion, this reference strain, initially identified as O. anthropi, was recently transferred to O. intermedium (see the CCUG Homepage: ). An identity of 99 % was detected with the type strain of O. intermedium, CCUG 24694T (GenBank accession no. AM114411), and only 97.8 % to the type strain of O. anthropi, LMG 3331T (GenBank accession no. AM114398). Hence isolate RMA 16994 appeared to be more closely related to O. intermedium than to O. anthropi. Identification of isolate RMA 16994 as O. intermedium was confirmed by the presence of the O. intermedium-specific sequence motifs ggctaat and ggttagtggagacactatcc at positions nt 487–493 and nt 905–924 of the 1390 bp 16S rRNA gene fragment as described by Scholz et al. (2006). The second motif is also shared with Brucella spp. However, both motifs are absent in O. anthropi and are instead represented by the sequences gactttt and ggacacagagatgtgtct at these positions, respectively.