A 91-year-old patient presenting with severe neuro-cognitive disorders entered the emergency department after suffering injuries from a fall during 2021. She was hospitalized for three weeks where she received a 2 day course of tazobactam. She had no known history of travel or acute infections or stay in palliative care. She was screened for CPE following a suspected outbreak of KPC on the floor where she is hospitalized. The result of her mCIM came back positive and no carbapenemase genes were found following testing for NDM, KPC, OXA-48-like, IMP, VIM, GES, NMC and IMI. The patient was moved to a residential and long-term care center and in June 2022 and tested positive for mCIM following routine screening practices. At this time PCR targets for FRI had been developed due to a separate FRI case reported in Canada [11]. The CPE isolate (N22-01531) was confirmed to be FRI-positive as were all retrospective isolates from this patient (data not shown).
Susceptibility testing showed resistance to aztreonam only. The isolate was susceptible to meropenem and ertapenem while intermediate to imipenem (Table 1). Similarly to another report [14] novel β-lactam/ β-lactamase inhibitor combinations were effective against our clinical FRI containing isolate. As previously observed for FRI-6 (Boyd et al. 2020), phenotypic detection of carbapenemase using mCIM and Rosco Neo-Rapid tests exhibited positive results while β CARBA was negative. Concerning FRI-8 [12], only mCIM detected FRI production. Similar discrepancies between phenotypic assays were reported concerning other FRI-variants; indeed, while β CARBA didn’t detect FRI-1 [15], mCIM was positive for multiple variants (FRI-8, FRI-4, FRI-2, FRI-5) [5, 7, 8, 16] and carba NP showed variable results (negative for FRI-8, inconsistent for FRI-11 [9] and positive for FRI-4, FRI-5, FRI-2, FRI-1 [4, 5, 7, 8]). Interestingly, mCIM was positive for all FRI-variants positive, however, this test may exhibit false positive results with E. cloacae due to AmpC over production [17, 18]. Indeed, further work needs to be done in FRI-type carbapenemase to avoid over testing E. cloacae non-carbapenemase producers using mCIM.
Data from WGS revealed the presence of an Enterobacter asburiae harbouring the novel FRI-variant blaFRI−12 (reference sequence number NG_081789.1). FRI-12 had the closest amino acid identity to FRI-5 differing by 13 amino acids (95.6% identity) (Supplementary Fig. 1).
The blaFRI−12 gene was found on a 104 Kb IncR plasmid (pN22-01531, Fig. 1). The closest hit to a known plasmid was pKeioCLMIC23 harbouring blaFRI−5 (accession number AP028423.1) which shared 44% coverage at 95% identity. Indeed, when all other FRI plasmids were BLAST against pN22-01531only 5–17% coverage was observed. Previous reports have shown all blaFRI−type genes are harboured on plasmids containing the replicon IncFll-type [10]. However, some of these (FLC-1, FRI-2, FRI-4, FRI-9 and FRI-11) are multi-replicon plasmids that also contain IncR [10]. Interestingly, the IncFII-type FRI-5 plasmid (pKeioCLMIC23), which shared the closest identity to our FRI-12 plasmid did not contain the IncR replicon as observed in other FRI-variants. The alignment of FRI surrounding regions in different variants highlights a diverse arrangement of coding sequences (Supplementary Fig. 2). The FRI-12 surrounding region was most similar to FRI-5 where we observed a common 7 Kb of shared sequence representing FRI-12, FriR (a LysR transcriptional regulator), SafC and SafB (components of a saf operon for pili functionality). No additional resistance genes were found on the FRI-12 IncR plasmid.
Similar to our previous reports,[11, 12] we were able to successfully transform pN22-01531 into Escherichia coli TOP10 (labeled pFRI-12TF), however, conjugation experiments with Escherichia coli J53AzR were not successful. Other than with FRI-6 [11] this phenomenon is not uncommon in FRI-type plasmids. We cloned blaFRI−12 into pCR2.1TOPO and subsequently transformed this plasmid into E. coli TOP10 (pblaFRI-12) to test its antimicrobial susceptibility (Table 1). The transformant showed higher MICs than the clinical strain to aztreonam, ceftriaxone, cefoxitin, ertapenem, meropenem, imipenem, ceftolozone/tazobactam and piperacillin/tazobactam. However, the cloned blaFRI−12 gene showed higher MIC to piperacillin/tazobactam only. Further work will need to be conducted on FRI-12 plasmid copy number and gene expression to establish if they affect MIC to selected antimicrobials.
To our knowledge, this is the first published report of blaFRI−12 identified globally. This study highlights the benefits of phenotypic screening of carbapenemases in Enterobacter sp. isolates exhibiting non-susceptible MICs to carbapenems despite susceptibility to cephalosporins.
Table 1
Antimicrobial susceptibilities (mg/L) for strains tested in this report.
Sensititre CAN1MSF | E. asburiae N22-01531 | | E.coli TOP10 (pFRI-12TF)a | | E.coli TOP10 (pblaFRI-12)b | | E.coli TOP 10 | |
amikacin | <=8 | S | <=8 | S | <=8 | S | <=8 | S |
aztreonam | 16 | R | > 16 | R | > 16 | R | <=1 | S |
cefepime | <=1 | S | <=1 | S | <=1 | S | <=1 | S |
ceftazidime | <=4 | S | <=4 | S | <=4 | S | <=4 | S |
ceftriaxone | <=1 | S | 4 | R | <=1 | S | <=1 | S |
ceftazidime/avibactam | <=4 | S | <=4 | S | <=4 | S | <=4 | S |
ciprofloxacin | <=0.06 | S | <=0.06 | S | <=0.06 | S | <=0.06 | S |
colistin | <=1 | I | <=1 | I | <=1 | I | <=1 | I |
doxycycline | <=4 | S | <=4 | S | <=4 | S | <=4 | S |
ertapenem | <=0.25 | S | > 2 | R | <=0.25 | S | <=0.25 | S |
ceftolozone/tazobactam | <=1 | S | 8 | R | <=1 | S | <=1 | S |
gentamicin | <=2 | S | <=2 | S | <=2 | S | <=2 | S |
imipenem/relebactam | 2 | S | <=1 | S | <=1 | S | <=1 | S |
levofloxacin | <=0.5 | S | <=0.5 | S | <=0.5 | S | <=0.5 | S |
meropenem | 0.25 | S | 4 | R | <=0.06 | S | <=0.06 | S |
meropenem/vaborbactam | <=1 | S | <=1 | S | <=1 | S | <=1 | S |
minocycline | <=4 | S | <=4 | S | <=4 | S | <=4 | S |
piperacillin/tazobactam | <=8 | S | > 64 | R | > 64 | R | <=8 | S |
plazomicin | <=1 | S | <=1 | S | <=1 | S | <=1 | S |
tigecycline | <=0.5 | S | <=0.5 | S | <=0.5 | S | <=0.5 | S |
tobramycin | <=2 | S | <=2 | S | <=2 | S | <=2 | S |
trimethoprim/ sulfamethoxazole | <=2 | S | <=2 | S | <=2 | S | <=2 | S |
Etest | | | | | | | | |
cefoxitin | 6 | S | 8 | S | 8 | S | < 0.25 | S |
Imipenem | 2 | I | 4 | R | 0.5 | S | 0.125 | S |
Bold formating indicates resistant phenotype according to CLSI (M100:2023). S-susceptible; I-intermediate; R-resistant
aE.coli TOP10 electrotransformed with pFRI-12
b bla FRI−12 and upstream region amplified from N22-01531 with primers FRI-12A (forward-GATTTCTCATTGTATACCAACC) and FRI-12B (reverse-TGGCGGACATTTCATGGCGC), cloned into pCR21.TOPO (ThermoFisher Scientific). pCR2.TOPO contains an ampicillin resistance gene (blaTEM−1) and kanamycin resistance gene (aphII).
Figure 1. FRI-12 plasmid generated with Proksee. Innermost two tracks represent forward and reverse open reading frames of pN22-01531. Third track represents BLASTn of FRI-12 plasmid to its closest related plasmid (pFRI-5). Outer two tracks are annotations generated by CARD RGI and mobileOG databases.
Supplementary Fig. 1. Amino Acid alignment of FRI-variants and phylogenetic tree showing relatedness among FRI-type sequences. Conserved Ambler class A regions; active site motifs 70SXXK73, 130SDN132, 166EXXXN170 and 234KTG236 and cysteine residues C69 and C238 are shown [6].
Supplementary Fig. 2. Schematic representation of FRI region among all FRI variants. Black arrows are hypothetical proteins. All other arrows are labelled.