*2.6. ompF and ompC Expression within Y. pseudotuberculosis Population*

Current experimental evidence suggests that bacterial cultures are constituted of heterogeneous subpopulations [55] with differential responses to environmental stresses [56,57]. For instance, Sánchez-Romero and Casadesús have observed the population heterogeneity of *S. enterica* cells in the expression level of *ompC* under kanamycin exposure [57]. Moreover, a low level of *ompC* correlates with high kanamycin resistance of the cell.

Our qRT-PCR results based on average Ct values did not reveal any significant transcriptional response of porins to long-term antibiotic exposure for most samples. They said nothing about the variability of *ompF* and *ompC* expression in the *Y. pseudotuberculosis* population or the amount of their gene products in the cell. To answer these questions, we first developed fluorescent reporter strains of *Y. pseudotuberculosis* 488. The GFP reporter systems were constructed by fusion of the *ompF*/*ompC* regulatory regions, including binding sites for transcription and translation factors (5- UTRs and signal sequences for sRNAs, C-terminal YQF amino acids for proteases), to the GFP reporter gene in a low copy number plasmid pACYC184 (Figure S1). Thus, the fluorescence of *Y. pseudotuberculosis* transformants potentially not only provided a high-resolution method for assaying *ompF*/*ompC* promoter activity but also indirectly indicated OmpF and OmpC amounts. We excluded chloramphenicol from the experiments due to the plasmid containing the gene for its resistance. However, we additionally verified that the presence of this antibiotic in the medium did not affect the level of sample fluorescence (Figure S2).

There was little evidence of any change in the mean fluorescence intensity, and hence OmpF and OmpC expression, after prolonged exposure to tetracycline and kanamycin (Figure 5), while qRT-PCR revealed the *ompF* upregulation in the presence of tetracycline at 27 ◦C (Figure 2). This difference indicates an additional post-transcriptional regulation of OmpF.

**Figure 5.** GFP fluorescence intensity in *Y. pseudotuberculosis* 488 transformed with *ompF*/*ompC* promoter-fused GFP reporter constructs following 16 h antibiotic treatment monitored by flow cytometry. (**A**) 27 ◦C incubation temperature; (**B**) 37 ◦C incubation temperature. All results are expressed as mean ± standard deviation between three experimental trials. An asterisk indicates *p*-value < 0.05 vs. respective control. Significance was calculated using one-way ANOVA. The histograms from a single representative experiment are shown. Control—without antibiotic treatment, Km—kanamycin, Tet—tetracycline, and Cb—carbenicillin.

After treatment with sub-MIC of carbenicillin at 27 ◦C, there was a 2.3-fold decrease in the *ompF* level and, on the contrary, a 2.5–3-fold increase in the *ompC* level. This is consistent with the previous observation that adaptation to β-lactams can involve increased expression of OmpC relative to OmpF since an increase in OmpC proportion hinders their penetration [37]. It is worth noting that carbenicillin inhibited the expression of *ompF* only at 27 ◦C. We observed that this transcript was prevailing compared to *ompC* in the untreated samples at both incubation temperatures; however, at 37 ◦C, its level decreased by more than 18 times (Figure S3). This initially low amount of *ompF* transcripts and additional nutrient deficiencies due to the stationary growth phase could prevent a further decrease in its expression at 37 ◦C under carbenicillin stress.

The distribution of GFP fluorescence within the sample cultures under kanamycin and tetracycline exposure was comparable to the untreated controls. Only carbenicillin presence led to the appearance of two *Y. pseudotuberculosis* subpopulations with different fluorescence intensities, most notably at 27 ◦C (Figure 6). F1 subpopulation responded to carbenicillin stress by 1.9-fold inhibition of *ompF* expression, while the cells from F2 increased it by 1.8 times. *ompC* was upregulated under carbenicillin exposure in both subpopulations; however, in C2, its level was significantly higher. Moreover, most of the cells formed subpopulations with a lower level of both *ompF* and *ompC* expression (71% and 78% of the total cell amount, respectively; Figure 6C).

We thus propose that modification of the bacterial envelope by reduced *ompF*/increased *ompC* expression confers adaptive fitness advantages to carbenicillin in a fraction of the bacterial population.

Previously it was shown that heterogenous responses improve the survival of bacterial cultures under the effect of stress factors [58]. Moreover, gene expression variability, including porin genes, is considered to underly adaptive resistance in phenotypically heterogeneous microbial populations [57,59].

**Figure 6.** Fluorescence heterogeneity of *Y. pseudotuberculosis* 488 transformed with *ompF/ompC* promoter-fused GFP reporter constructs following treatment with a sublethal dose of carbenicillin monitored by flow cytometry. (**A**) Fluorescence histograms; (**B**) fluorescent intensity of subpopulations; (**C**) the percentage of cells in each subpopulation. All results are expressed as mean ± standard deviation between three experimental trials. An asterisk indicates *p*-value < 0.05 between groups. Significance was calculated using one-way ANOVA. The histograms from a single representative experiment are shown. F1, F2, and C1, C2—subpopulations of *Y. pseudotuberculosis* 488 with low/high GFP fluorescence.

## **3. Materials and Methods**

#### *3.1. Bacterial Strain and Growth Conditions*

*Y. pseudotuberculosis* 488 strain (= 117 strain, O:1b serotype) isolated from a patient with FESLF in the Russian Far East region was used in the present study.

To study short-term porin response to antibiotic stress, bacterial cells were grown in LB broth overnight at 27 ◦C with shaking. The next day, overnight cultures were diluted at a ratio of 1:50 into fresh LB medium and incubated with agitation at 27 ◦C (optimal growth temperature) and 37 ◦C (host temperature) to mid-exponential growth phase (OD600 = 0.4–0.6). Then, the antibiotics kanamycin (Sigma-Aldrich, St. Louis, MA, USA), tetracycline (Sigma-Aldrich, St. Louis, MA, USA), chloramphenicol (Sigma-Aldrich, St. Louis, MA, USA), and carbenicillin (Sigma-Aldrich, St. Louis, MA, USA) were added to the media in sublethal concentrations (0.5 MIC). After a 1 h antibiotic treatment (180 rpm, 27 ◦C and 37 ◦C), the cells were mixed with RNAprotect Bacteria Reagent (Qiagen, Hilden, Germany) and centrifuged.

To study porin response to long-term antibiotic exposure, overnight cultures were diluted and inoculated into LB broth media to obtain a final concentration of 5 × 105 CFU/mL. Bacteria were incubated with sublethal concentrations of antibiotics (kanamycin, tetracycline, chloramphenicol, and carbenicillin) for 16 h at 27 ◦C and 37 ◦C. RNAprotect Bacteria Reagent (Qiagen, Hilden, Germany) was added to the culture samples.

Three independent bacterial cultures for each test or control condition were prepared as biological replicates for RNA isolation.
