Overexpression of an outer membrane protein associated with decreased susceptibility to carbapenems in Proteus mirabilis

Abstract

Proteus mirabilis isolates commonly have decreased susceptibility to imipenem. Previously, we found P. mirabilis hfq mutant was more resistant to imipenem and an outer membrane protein (OMP) could be involved. Therefore, we investigated the role of this OMP in carbapenem susceptibility. By SDS-PAGE we found this OMP (named ImpR) was increased in hfq mutant and LC-MS/MS revealed it to be the homologue of Salmonella YbfM, which is a porin for chitobiose and subject to MicM (a small RNA) regulation. We demonstrated that ImpR overexpression resulted in increased carbapenem MICs in the laboratory strain and clinical isolates. Chitobiose induced expression of chb (a chitobiose utilization operon). Real-time RT-PCR and SDS-PAGE were performed to elucidate the relationship of hfq, impR, chb and MicM in P. mirabilis. We found MicM RNA was decreased in hfq mutant and chbBC-intergenic region (chbBC-IGR) overexpression strain (chbIGRov), while impR mRNA was increased in hfq mutant, micM mutant and chbIGRov strain. In addition, mutation of hfq or micM and overexpression of chbBC-IGR increased ImpR protein level. Accordingly, chitobiose made wild-type have higher levels of ImpR protein and are more resistant to carbapenems. Hfq- and MicM-complemented strains restored wild-type MICs. Mutation of both impR and hfq eliminated the increase in carbapenem MICs observed in hfq mutant and ImpR-complementation of hfq/impR double mutant resulted in MICs as hfq mutant, indicating that the ImpR-dependent decreased carbapenem susceptibility of hfq mutant. These indicate MicM was antisense to impR mRNA and was negatively-regulated by chbBC-IGR. Together, overexpression of ImpR contributed to the decreased carbapenem susceptibility in P. mirabilis.

Fig 1. The SDS-PAGE profile of OMPs from overnight cultures of wild-type, its derivatives and wild-type treated with chitobiose.

The representative result from three independent experiments is shown. The arrow indicates the band of ImpR. M, marker; wt, wild-type; hfq, hfq mutant; micM, micM mutant; ImpRov, ImpR-overexpressing strain; chbIGRov, chbBC-IGR-overexpressing strain; impR, impR mutant.

Fig 2. The RNA levels of impR (A) and MicM (B) in the wild-type P. mirabilis and its derived strains.

The relative RNA levels of impR in the wild-type, mutants of hfq and micM, and chbIGRov were quantified by real-time RT-PCR. The RNA was prepared using overnight bacterial cultures. The relative RNA levels of MicM in the wild-type, mutants of hfq, impR and hfq/impR, and chbIGRov were also determined in the same way. The expression level for the wild-type cells was set at 1. The data represent the averages of three independent experiments with standard deviations. Significant difference from the wild-type was indicated with the asterisk (*, P<0.05; **, P<0.01; ***, P<0.001 by Student’s t-test analysis). wt, wild-type; hfq, hfq mutant; impR, impR mutant; hfq/impR, hfq and impR double mutant; chbIGRov, chbBC-IGR overexpressing strain; micM, micM mutant.

Fig 3. Genomic location of impR (A) and micM (B) in P. mirabilis.

The number indicates the start nucleotide number of each gene in the genome and the size (bp) of each gene is indicated.

Fig 4. The expression of MicM in the wild-type and hfq mutant.

The RNA amount of MicM in the wild-type or hfq mutant was quantified by real-time RT-PCR. Overnight bacterial cultures were diluted to OD₆₀₀ of 0.1 and incubated for 4 h, 8 h and 16 h before total RNA was prepared. The expression level for the wild-type cells at 4 h was set at 1. The data represent the averages of three independent experiments with standard deviations. Significant difference from the wild-type at different time point was indicated with the asterisk (*, P<0.01 by Student’s t-test analysis). wt, wild-type; hfq, hfq mutant.

Fig 5. (A) Chitobiose induced chb promoter activity.

The xylE activity in the chb-xylE reporter plasmid-transformed wild-type P. mirabilis in the presence or absence of 2 mM chitobiose was measured after induction for 4 h and 8 h. (B) Chitobiose induced expression of chb mRNA. The RNA of the wild-type was prepared using overnight bacterial cultures and relative chb mRNA amount was quantified by real-time RT-PCR using primers for chbBC-IGR. The expression level of control cells without chitobiose (nil) was set at 1. The data represent the averages of three independent experiments with standard deviations. Significant difference from the no chitobiose control at different time point was indicated with the asterisk (*, P<0.01 by Student’s t-test analysis). wt, wild-type.

Fig 6. Model for ImpR regulation by MicM sRNA.

(A) In the absence of chitobiose, the expression of the ImpR porin is silenced at the post-transcriptional level by pairing of MicM sRNA with the 5’ UTR of impR mRNA, promoting cleavage of the impR mRNA by a ribonuclease (RNase E?). At the same time the chb operon is transcriptionally repressed. (B) In the presence of chitobiose, an inducer to activate transcription of the chb operon, processing of the chb transcript releases chbBC IGR RNA. This RNA base-pairs with MicM making it susceptible to the action of a ribonuclease (RNase E?). The drop in MicM levels relieves impR repression leading to a burst of ImpR translation. ImpR assembles in the outer membrane resulting in increased carbapenem MICs. MicM is an Hfq-dependent and constitutively-expressed sRNA.