The Genus Ochrobactrum as Major Opportunistic Pathogens

Ochrobactrum species are non-enteric, Gram-negative organisms that are closely related to the genus Brucella. Since the designation of the genus in 1988, several distinct species have now been characterised and implicated as opportunistic pathogens in multiple outbreaks. Here, we examine the genus, its members, diagnostic tools used for identification, data from recent Ochrobactrum whole genome sequencing and the pathogenicity associated with reported Ochrobactrum infections. This review identified 128 instances of Ochrobactrum spp. infections that have been discussed in the literature. These findings indicate that infection review programs should consider investigation of possible Ochrobactrum spp. outbreaks if these bacteria are clinically isolated in more than one patient and that Ochrobactrum spp. are more important pathogens than previously thought.


Introduction
Gram-negative, non-fermenting bacteria are an emergent worry in medical situations and are becoming a growing cause of severe infections. Pathogens of this type are opportunistic and include many different bacterial species, such as Ralstonia spp., Pseudomonas aeruginosa, Sphingomonas paucimobilis and Brevundimonas spp. [1][2][3][4][5]. Gram-negative, non-fermenting bacteria can infect both patients undergoing treatments and individuals outside of a clinical setting with various underlying conditions or diseases. Another type of these bacteria are the members of the α-proteobacterial genus Ochrobactrum [6].
Ochrobactrum spp. are found in a wide variety of environments including water, aircraft water, soil, plants and animals [6][7][8][9][10][11][12]. Several Ochrobactrum spp. have been investigated for their potential to degrade xenobiotic pollutants and for heavy metal detoxification under a variety of environmental conditions [13][14][15][16]. Ochrobactrum spp. are very closely related to brucellae, and even though they are considered to be of low virulence, they have increasingly been found to cause infections (some serious including endocarditis and septicaemia) in immunocompetent hosts [17,18].
Investigation of the scientific/medical literature presented a wide variety of infections resultant from Ochrobactrum spp. and these were resistant to wide variety of antibiotics. Our data point to the genus being a more common pathogen than previously supposed, with many of the infections/conditions caused by Ochrobactrum spp. being aggressive and debilitating. The overall aim of this work is to present a summary of the types of Ochrobactrum spp. infections, any underlying disorders/illnesses in patients that accompany these infections and the potential treatments that can be used in the management of infections to support medical specialists.

Genus Ochrobactrum
The genus Ochrobactrum emerged from what was previously categorised as the CDC group VD1-2. The type species Ochrobactrum anthropi had previous been called Achromobacter VD based on the Special Bacteriology Section of the US Center for Disease Control [19]. Initial results indicate members of the group grew on MacConkey agar producing catalase, oxidase and urease; strains could be Gram-negative to variable [20]. However, the taxonomic position of Achromobacter became complicated and the name Achromobacter and related CDC group VD were no longer accepted by Bergeys Manual [19] leading to a new classification and the emergence of the genus Ochrobactrum [21]. Ochrobactrum spp. are phylogenetically related to members of the alpha-2 subdivision of Proteobacteria. They are catalogued on the Brucella rRNA branch of rRNA superfamily IV. Thus, from the previous CDC group Vd, a novel genus and a new species, Ochrobactrum anthropi, was proposed [21,22]. The type strain was Gram-negative, aerobic, rod shaped, non-pigmented and motile. It produced acid from a selection of carbohydrates and reduced both nitrate and nitrite and possessed a GC ratio between 56 to 59% [21]. Almost all 56 strains categorised as CDC GroupVD that were used to support the new genus Ochrobactrum came from various human clinical specimens. Since the initial description of O. anthropi, several other species have since been described (Table 1 and Figure 1). Certain Ochrobactrum spp. can be opportunistic pathogens especially in a hospital environment with the majority of reported cases due to hospital-acquired infections in patients with indwelling and invasive medical devices, including central venous catheters and drainage tubes [23]. In addition, the organism shows widespread resistance to penicillins and other antibiotics that cause clinical management issues with immunocompromised hosts [24,25]. The phylogenetic relationship between all described Ochrobactrum spp. can be seen in Figure 1.

Identification of Ochrobactrum spp.
Ochrobactrum species are Gram-negative and composed of short rods that are straight or slightly

Identification of Ochrobactrum spp.
Ochrobactrum species are Gram-negative and composed of short rods that are straight or slightly curved with one end flame shaped. They are generally motile and do not produce haemolysis on blood agar [43].

Biochemical Identification
Biochemical identification can be carried out using biochemical-testing kits such as the API 20NE or Vitek-2 (BioMèrieux, Las Balmas, France). When biochemical testing is carried out, it is normal to test isolates against Brucella agglutinating sera to prevent misdiagnosis with Brucella its close neighbour [44]. It has been shown that commercial kits are generally unsuitable for identification or differentiation amongst Ochrobactrum [45]. Analysis of 103 clinically relevant Ochrobactrum strains indicated that biochemical reaction profiles of the API and BD Phoenix™ 100 systems for identifying Ochrobactrum isolates can only be used at the genus level [46]. Care is required when identifying Ochrobactrum in clinical situations as misidentification has occurred with Brucella melitensis [47].
For identification of Ochrobactrum spp., it was proposed that the isolation of non-fastidious, non-fermenting, oxidase-positive, Gram-negative rods that are resistant to Beta-lactams (except imipenem) indicates the isolate is from the genus Ochrobactrum [43]. The API 20NE will confirm the identification to genus level for the majority of strains (Table 2). In addition, it has been proposed that urease activity, the mucoidy of the colonies and growth at 45 • C on tryptic soy agar coupled to susceptibility to colistin, tobramycin and netilmicin should be used as differentiating characteristics in the determination of O. anthropi and O. intermedium to the species level [43].
In many clinical situations, the Microscan Walkaway system is used for primary identification and any unusual non-fermentative bacteria are analysed via biochemical analysis methods such as the RapID NF Plus system. This strategy has been shown to generally perform very well [48]. There have been cases of misdiagnosis of Ochrobactrum anthropi (subsequently confirmed by VITEK) as Shewanella putrefaciens [48]. Of course, the opposite has also been reported where a Brucella suis bacteraemia was mistakenly identified as Ochrobactrum anthropi by the VITEK 2 system [49,50]. These studies underscore the difficulty encountered in identifying unusual Gram-negative, non-fermentative bacteria such as Ochrobactrum.

Fatty Acid Analysis
Use of fatty acid analysis as a differentiation marker using the Sherlock System and comparison with the Sherlock database provided the identification result for O. anthropi with an ID score of 0.556, indicating its poor utility for differentiation at the species level [45].

Molecular Identification
Molecular tools have long been applied to the typing of Ochrobactrum species. Early studies utilised pulsed-field gel electrophoresis and rep-PCR for the epidemiological analysis [52] followed by AFLP (Amplified Restriction Fragment Length Polymorphism) to confirm the relatedness of O. anthropi and O. intermedium with its Brucella relatives [53] using a limited number of isolates. The molecular diversity of a larger number of Ochrobactrum strains were investigated by comparing environmental isolates from soil and the rhizoplane and comparing these to a number of clinical isolates [12]. Rep-PCR using a combination of BOX and REP primers were used to profile the isolates. The isolates used in this study clustered according to their species designation [12] indicating that rep-PCR profiling offered a good tool for species delineation.
However, the differentiation of species is somewhat difficult because of their phenotypic similarity and indeed confusion amongst 16s rDNA sequences [45]. Errors still occur such as in the case of bacteraemia where the causative agent was recognised as Ralstonia paucula by the Microscan Walkaway system but later following DNA sequencing was identified as O. anthropi [54].
16s rDNA sequence similarity between O. anthropi and O. intermedium ranged from 97.9% to 98.7% depending on the strains compared [43] suggesting a higher genetic deviation in O. intermedium than is found in O. anthropi. The genetic structure of a collection of 65 isolates (37 clinical, 11 environmental and 17 from culture collections) illustrative of the known natural distribution of O. intermedium was analysed by MLSA (Multi-Locus Sequence Analysis) [53].
A recA-PCR RFLP (Restriction Fragment Length Polymorphism) assay was also developed to study interspecies variability within Ochrobactrum using recA sequences from known isolates including 38 O. anthropi strains and type strains of O. intermedium, O. tritici and O. lupini and comparing these with closely related Brucella strains [54]. It was concluded that recA-sequence analysis provided a reliable molecular subtyping tool for Ochrobactrum at both the inter-and intraspecies level. Subsequently, a sensitive recA gene-based multi-primer single-target PCR assay has been created to differentiate O. anthropi, O. intermedium and Brucella that had been reported to cause diagnostic difficulty (Table 3)  Ochrobactrum species (all confirmed as Ochrobactrum species by 16s rDNA sequencing) to examine comparative identification techniques ranging from commercial kits to biochemical and ribotyping [43].

MALDI-TOF MS
MALDI-TOF MS (Matrix-Assisted Laser Desorption/Ionisation-Time-of-Flight) was initially used to identify Ochrobactrum intermedium from a range of difficult to identify strains as an alternative to Vitek, API or 16s rDNA sequencing in a large validation screen with some 204 genera showing discordant results from different identification methods [58]. The method has since found utility for evaluation within the Ochrobactrum genus. The utility of automated rep-PCR (DiversiLab TM system, BioMèrieux, Las Balmas, France) and MALDI-TOF MS analysis was compared for typing of 23 O. anthropi clinical isolates (bacteraemias) [44]. MALDI-TOF MS evaluation clustered the 23 strains of O. anthropi into a single group containing four distinct subgroups at close distance, indicating a high similarity between the isolates but also its accuracy in identification [44]. The technique of MALDI-TOF MS is gaining widespread usage in clinical situations and is increasingly utilised for Ochrobactrum identification in the clinic [59].

Ochrobactrum spp. Virulence
Ochrobactrum spp. are considered to be of low virulence. A study carried out by Yagel et al. into the virulome of Ochrobactrum spp. looked at the genomes of 130 isolates [60]. These isolates were taken from clinical, environmental, animal and plant settings. The study identified a limited number of virulence factors in the majority of these isolates. They found lipid A biosynthesis genes in all genomes analysed. They also found other virulence-associated genes in the majority of isolates such as genes associated with fatty acid biosynthesis (fabZ), carbohydrate metabolism (pgm and cgs), cell wall biosynthesis (wbpL) and biofilm formation (ricA, 95%). Genes for other more widespread Gram-negative virulence-associated proteins were not found in these genomes [60].

Outbreak Identification
All obtainable publications (journal articles, case reports and conference proceedings) discussing Ochrobactrum spp. infections were recovered using the PubMed, Web of Knowledge and Google Scholar search databases from 1980 to April 2020. The terms "CDC group VD1-2", "Ochrobactrum", "Ochrobactrum spp.", "Ochrobactrum anthropi" and "Ochrobactrum intermedium" as well as all species names listed in Table 1 were searched. Any publications that discussed infection were set aside. These papers/abstracts were then read and the required information extracted from them. This information included year, geographic location, patient information (age, sex and any underlying medical conditions), antimicrobial testing, treatment and patient outcomes where available. The references cited from these publications were also checked for any publications/reports that may not have been found during the database searches.

Outbreak Analysis
The results of the investigations of the literature can be seen in Tables 4 and 5. The tables summarise year, geographic location, patient information (age, sex and any underlying medical conditions), antimicrobial testing, treatment and patient outcome. One hundred seventeen separate instances of Ochrobactrum anthropi infection (277 individual cases) were identified along with a further eleven instances (twelve cases) of Ochrobactrum intermedium, Ochrobactrum oryzae, Ochrobactrum pseudogrignonense, Ochrobactrum pseudintermedium and Ochrobactrum tritici infection. The major breakdown of O. anthropi related conditions were as follows: forty-six instances of bacteraemia (42%) from which three were described as "bloodstream infections" that were usually associated with catheters, fourteen instances of septicaemia/sepsis/septic shock (12%) and two further instances of biliary sepsis (2%), nine instances of endophthalmitis, eight instances of peritonitis, four instances of pneumonia (8%) and two instances each endocarditis (2%). Other infections included two cases of keratitis (2%), four of various types of abscess (neck, pelvic, pancreatic and retropharyngeal) (3%) and one instance each of "hand infection" and brain empyema (1%). There have also been multiple reported instances of Ochrobactrum spp. infection that have caused two or more conditions. These include bacteraemia and necrotising fasciitis, bacteraemia and pneumonia, septicaemia and peritonitis and two instances of septic shock and endocarditits. Ten cases of death associated to Ochrobactrum spp. infection (all O. anthropi) have also been reported in the literature, four with sepsis/septicaemia (one with endocarditis), two with peritonitis and one each with a bloodstream infection, pyrogenic infection, endocarditis and infection of transjugular intrahepatic portosystemic shunt.

Underlying Conditions/Illness
The bulk of Ochrobactrum related infections (Tables 4 and 5) had an associated underlying disorder or disease that increased patient susceptibility to infection. Multiple patients, who were afflicted with a variety of different cancers or those with kidney failure (caused by diabetes mellitus), contracted Ochrobactrum-related bacteraemia/septicaemia due to a catheter/undergoing dialysis. These demonstrate how Ochrobactrum acts as an opportunistic pathogen in immunocompromised individuals. Infections were both hospital and community acquired. This is of interest as opportunistic pathogens such as Ochrobactrum spp. are mostly contracted in clinical environments. It was also interesting that a high level of instances of infection, 23 separate instances, occurred where patients had no underlying health conditions.

Pseudo-Outbreaks
To date, six pseudo-outbreaks have been described with Ochrobactrum spp. (Tables 4 and 5). These may be challenging as they may lead to unessential/unneeded treatments such as needless courses of antibiotics or patient interventions (e.g., the removal of indwelling devices including various catheter types) and can waste both time and resources in both the clinical laboratory and treatment ward settings. Pseudo-outbreaks have many possible causes including contaminated water or materials used in the clinical testing laboratory or contaminated medical solutions such as saline. Montaña et al. described how O. anthropi was the reason for a pseudo-outbreak in a general treatment ward in an Argentinean hospital due to contaminated collection tubes [61]. No symptoms connected with bacterial infection were observed in any patients, even though O. anthropi was identified in microbiological testing. The recovered bacteria were carbapenem-resistant.

Treatment of Ochrobactrum spp. Infections
Treatment of Ochrobactrum spp. infections is often problematic, due to their resistance to different families of antibiotics such as β-lactams (penicillins, cephlasporins and emerging cases of carbapenem resistance). The antibiotic susceptibility profiles of some 103 typed strains of Ochrobactrum were analysed using the E-test™ for 19 clinically relevant antimicrobials [46]. In general, strains were highly resistant to β-lactam antibiotics, susceptible to ciprofloxacin, and 97.1% of the strains tested were susceptible to trimethoprim/sulfamethoxazole. This suggests that ciprofloxacin and/or trimethoprim/sulfamethoxazole in combination may be useful for empirical treatment of Ochrobactrum infections [46]. In the majority of outbreaks described in Table 4, aminoglycoside, fluoroquinolone, carbapenem or trimethoprim/sulfamethoxazole antibiotics were used in patient treatment. In the majority of cases, these treatments were successful in curing infections. However, as can be seen in Table 4, resistance was observed in various different outbreaks to all these antibiotics. An example of this is reported in a case of O. anthropi bacteraemia in a patient in Japan in 2013 where susceptibility testing showed the organism to be resistant to aztreonam, ceftazidime, cefepime, ciprofloxacin, gentamicin, levofloxacin, piperacillin, piperacillin-tazobactam and trimethoprim-sulfamethoxazole [51]. There have been no controlled trials of antibiotic therapies for Ochrobactrum spp. infections in humans therefore treatment should be based upon the results of in vitro susceptibility testing on the isolated clinical strains. Resistance to β-lactam antibiotics (cephalosporins, cephamycins and β-lactamase inhibitors) is due to a chromosomal gene (bla och ) that is similar to the Ambler class C β-lactamase gene. This gene encodes an AmpC-like enzyme that is called OCH [168]. In addition, a plasmid-borne bla oxa-181 gene has been found in some Ochrobactrum intermedium strains giving resistance to carbapenems [169]. Three Ochrobactrum spp. strains isolated from birds in Pakistan harboured aminoglycoside (aadB, aadA2, aac6-Ib and strA, strB) β-lactam (bla och2 and carb2), tetracycline (tetG), chloramphenicol (floR), sulphonamide (sulI) and trimethoprim (dfrA10) resistance genes [170].

Conclusions
Ochrobactrum spp. are not presently thought of as major pathogens. Nevertheless, as a result of our literature search, it can be seen that there have been 128 separate outbreaks of Ochrobactrum spp. infections reported. Thus, the consideration that they may be innocuous should in our opinion be reconsidered based on these findings. Although the genus is considered of low virulence and of lower risk compared to other non-fermenting Gram-negative bacteria such as Pseudomonas aeruginosa, we feel it must not be ignored as a potential cause of infections (nosocomial or otherwise) and should be included in routine screening programs in hospitals.