Role of Low-Molecular-Mass Penicillin-Binding Proteins, NagZ and AmpR in AmpC β-lactamase Regulation of Yersinia enterocolitica

Yersinia enterocolitica encodes a chromosomal AmpC β-lactamase under the regulation of the classical ampR-ampC system. To obtain a further understanding to the role of low-molecular-mass penicillin-binding proteins (LMM PBPs) including PBP4, PBP5, PBP6, and PBP7, as well as NagZ and AmpR in ampC regulation of Y. enterocolitica, series of single/multiple mutant strains were systematically constructed and the ampC expression levels were determined by luxCDABE reporter system, reverse transcription-PCR (RT-PCR) and β-lactamase activity test. Sequential deletion of PBP5 and other LMM PBPs result in a continuously growing of ampC expression level, the β-lactamse activity of quadruple deletion strain YEΔ4Δ5Δ6Δ7 (pbp4, pbp5, pbp6, and pbp7 inactivated) is approached to the YEΔD123 (ampD1, ampD2, and ampD3 inactivated). Deletion of nagZ gene caused two completely different results in YEΔD123 and YEΔ4Δ5Δ6Δ7, NagZ is indispensable for YEΔ4Δ5Δ6Δ7 ampC derepression phenotype but dispensable for YEΔD123. AmpR is essential for ampC hyperproduction in these two types of strains, inactivation of AmpR notable reduced the ampC expression level in both YEΔD123 and YEΔ4Δ5Δ6Δ7.


INTRODUCTION
Yersinia enterocolitica, a member of Enterobacteriaceae, is a zoonotic pathogen widely distributed in nature (Wang et al., 2011;Liang et al., 2012). Most Y. enterocolitica exhibits intrinsic resistance to β-lactm antibiotics by the production of chromosomally encoded β-lactamases called BlaA (a class A enzyme showing constitutive expression) and BlaB (an inducible AmpC-type β-lactamase), respectively (Cornelis and Abraham, 1975;Bent and Young, 2010).
Since the effects of the above-mentioned genes in Y. enterocolitica were seldom reported, we elucidated the role of low-molecular-mass penicillin-binding proteins (LMM PBPs) (PBP4, PBP5, PBP6, and PBP7), NagZ and AmpR in the Y. enterocolitica ampC regulation. Firstly, we investigated the effects of each LMM PBP on the expression of AmpC β-lactamase by monitoring the ampC promoter activity from a series of LMM PBPs mutant strains and confirmed by quantitative reverse transcription-PCR (qRT-PCR). Secondly, nagZ gene was deleted in two ampC derepressed strains YE D123 and YE 4 6 5 7 to determine the role for ampC expression.

Bacterial Strains, Plasmids, Primers, and Growth Conditions
Strains and plasmids used in this study were listed in Table 1. Individual genes were deleted initially from Y. enterocolitica subsp. palearctica 105.5R(r) (Wang et al., 2011). Luria-Bertani (LB) agar plates and broth were used as culture media for Y. enterocolitica (28 • C) and Escherichia coli (37 • C). For induction assay, cefoxitin was used according to the references (Guerin et al., 2015;Liu et al., 2016).

Measurement of the ampC Promoter Activity
The method of measuring the ampC promoter activity with the luxCDABE reporter system was reported previously . The reporter plasmid pLUXampC was transferred into the tested strains, and the luminescence was measured by using an Infinite M200 Pro spectrophotometer. The value of luminescence/OD600 was used to assess the ampC promoter activity.

Determination of β-Lactamase Activity and Antibiotic Susceptibility Testing
Specific β-lactamase activities were spectrophotometrically determined with nitrocefin (Oxoid) as a substrate as previously described . One unit of β-lactamase activity (U/mg) was defined as the number of nanomoles of nitrocefin hydrolyzed per minute per milligram of protein. Antibiotic susceptibility was determined using the standard 2-fold serial broth microdilution method according to the Guidelines of the Clinical Laboratory Standards Institute (CLSI, 2015).

N-Acetyl-β-Glucosaminidase Activity Assay
The N-acetyl-glucosaminidase activity of the whole cell lysates of wild-type strain 105.5R(r) and YE Z were measured using 4-nitrophenyl N-acetyl-β-D-glucosaminide as a chromogenic substrate (Sigma). The presence of p-nitrophenol were detected by monitoring the optical density at 405 nm by 10 h continuously.

Complementation Assay
The ORF of nagZ was amplified and cloned into the broad-hostrange expression vector pSRKTc to construct plasmid pNagZ. Transformants were selected on 10 µg/ml tetracycline Yersinia selective LB plates, acquisition of the appropriate plasmid was confirmed by colony PCR.

Role of LMM PBPs in the Expression of AmpC β-Lactamase
After a series of LMM PBPs mutant strains were constructed, reporter plasmid pLUXampC was used to monitor the ampC expression level . As shown in Figure 1, deletion pbp5 caused a visible increase in the ampC promoter activity under both basal and induced conditions; but deletion of pbp4, pbp6, and pbp7 did not affect the AmpC expression obviously. In the group of double and triple mutant strains, ampC derepression only appeared in pbp5 background, the ampC promoter activity of YE 4 5, YE 5 6, and YE 5 7 exhibited a marked rise compared with YE 4 6, YE 4 7, or YE 6 7. The level of ampC expression keep increasing in triple mutant strains YE 4 5 6, YE 4 5 7, and YE 5 6 7, but not in YE 4 6 7. Finally, the quadruple deletion strain YE 4 5 6 7 displayed the highest level of ampC promoter activity. These results suggested that PBP5 plays the most important roles in Y. enterocolitica ampC regulation. The qRT-PCR assay reconfirmed the results observed from ampC promoter activity assay ( Table 2).

Role of NagZ in AmpC Derepression of Y. enterocolitica
In agreement with our previous data , AmpD deletion strain YE D123 exhibit a derepression phenotype, and the β-lactamase activity of YE D123 is slightly higher than YE 4 5 6 7 (Figure 2). To evaluate the role of NagZ in AmpC derepression, nagZ gene was deleted in both derepression strains to construct YE D123 Z and YE 4 5 6 7 Z. As shown in Figure 2, nagZ was indispensable for ampC over expression of YE 4 5 6 7, the β-lactamase activity of nagZ deletion strain YE 4 5 6 7 Z was decreased significantly, closed to the wild-type strain level. In complementation assay, YE 4 5 6 7 Z (pNagZ) restored the β-lactamase activity to the level of YE 4 5 6 7. However, NagZ was dispensable in YE D123, the β-lactamase activity of nagZ deletion strain YE D123 Z was nearly as high as YE D123 (Figure 2). These results suggested that NagZ was needed in PBPsdriven AmpC derepression, but did not perform its expected function in AmpD mutation strains. Antibiotic susceptibility test was also performed, as shown in Table 3, the MIC values of YE 4 5 6 7 Z were slightly below the wild-type strain 105.5R(r), far from its parent strain YE 4 5 6 7 for almost all tested β-lactams; but only a marginal distinction between YE D123 and YE D123 Z was found. These results illustrated that AmpD/PBPs regulate AmpC expression through NagZ dispensable/indispensable ways in Y. enterocolitica.

N-Acetyl-β-Glucosaminidase Activity Assay
The nagZ mutation strain YE Z was constructed, and determined by the enzyme activity of the both wild-type strain and YE Z for 10 h using N-acetyl-β-D-glucosaminide as substrate. As shown in Figure 3, YE Z abolished the N-acetyl-βglucosaminidase activity completely, it was suggested that NagZ is the only enzyme that with N-acetyl-β-glucosaminidase activity in Y. enterocolitica.

Role of AmpR in ampC Expression of in Y. enterocolitica
In the paradigm of the ampR-ampC system, the ampR gene is located immediately adjacent to ampC, and AmpR plays a pivotal role in the regulation of AmpC (Seoane et al., 1992). To assess the role of AmpR in Y. enterocolitica, we compared the β-lactamase activity of YE D123 R, YE 4 5 6 7 R with their parent strains YE D123, YE 4 5 6 7, respectively. As a result, ampR inactivation dramatically reduced the β-lactamase  activity of both YE D123 R and YE 4 5 6 7 R, regardless of adding cefoxitin or not (Figure 2).

DISCUSSION
The ampR-ampC system from Citrobacter freundii and Enterobacter cloacae has been well studied in the early 1990s Peter et al., 1988). However, newly discovered ampC regulators such as, PBP4 (DacB) or NagZ in Enterobacteriaceae was not yet understood. A deep study in Y. enterocolitica ampR-ampC system would be helpful to improve the comprehensive understanding of Enterobacteriaceae ampC regulation.
PBPs are a group of enzymes involved in cell-wall recycling and the processes of AmpC β-lactamases regulation. In E. coli model, deletion of three or four PBPs and the concomitant inhibition of PBP 1a, 1b, and/or 2 results in an increased level of β-lactamase induction (Pfeifle et al., 2000). However, since E. coli lacks the chromosomal ampR gene, the result may be inconsistent with other members of the Gram-negative bacteria which have a chromosome encoding the ampR-ampC system. In 2009, Moya et al. demonstrated the inactivation of DacB (PBP4), a nonessential low-molecular mass PBPs is the principal reason for one-step high-level ampC expression in clinical strains of P. aeruginosa (Moya et al., 2009). Interestingly, inactivation of PBP4 in E. cloacae triggered a significant increase of βlactams resistance, but without an obvious upregulation of ampC gene, it may be suggested that PBP4 regulates AmpC at a posttranscriptional level (Guerin et al., 2015). In this study, we found deletion of pbp4 did not elevate the ampC expression level, this result is accordance with E. cloacae. After that, we deleted all four LMM PBPs one after another, and found that PBP5 is the most effective PBP involved in the regulation of ampC in Y. enterocolitica. Of the single-mutation strains, only the pbp5 deletion strain YE 5 showed an obvious rise in ampC expression level. Likewise, for multi-mutation strains, the function of PBP4, PBP6, and PBP7 in ampC regulation were detected only if in pbp5 background. According to the results shown in Figure 1 and Table 2, we deduced the hierarchy of the role of PBPs genes in ampC derepression: PBP5 > PBP4 > PBP7 > PBP6. Although DacB may regulates AmpC at a post-transcriptional level (Guerin   , but no trace of post-transcriptional mechanism has been found in Y. enterocolitica. Along with the popular research of ampC regulation, there is growing evidence that some bacteria may regulate the expression of ampC through at least two different ways, one of which was NagZ-dependent, while the other worked without the participation of NagZ (Huang et al., 2012;Guerin et al., 2015). In the study on P. aeruginosa, nagZ inactivation was shown to attenuate ampC expression and was critical for basallevel ampC derepression in both PA D (ampD inactivation) and PA dB (pbp4 inactivation) mutants (Asgarali et al., 2009;Zamorano et al., 2010). However, nagZ had little effect on the cefoxitin-induced ampC expression level in both PA D and PA dB, which indicated that an unidentified non-NagZ product at work in this induction process. Furthermore, two different regulation ways of β-lactamase have been found in S. maltophilia, on one hand NagZ was essential for KJ DI (ampD inactivation) ampC overexpression, on the other hand, nagZ inactivation hardly influenced the ampC expression level of KJ mrcA (pbp1a inactivation; Huang et al., 2012). In this study, we also found two different ampC regulation ways exist in Y. enterocolitica, the patterns of which were just the reverse of that in S. maltophilia (Huang et al., 2012). The β-lactamase activity of YE D123 was not affected by the inactivation of the nagZ gene, whereas the introduction of nagZ into the PBP mutation strain YE 4 5 6 7 dramatically reduced the β-lactamase activities at both the basal and induced level (Figure 2). As shown in Table 3, the antibiotic resistance of YE 4 5 6 7 and YE D123 were marked improved compare with wild-type strain, the MIC value of these two strains in TZP, PIP, CFZ, CAZ, CRO, and ATM is rising sharply. While after inactivation of nagZ gene simultaneously, only a marginal distinction between YE D123 and YE D123 Z was found, but the MIC values of YE 4 5 6 7 Z has shifted down significantly, far from its parent strain YE 4 5 6 7 for almost all tested β-lactams. To further confirm the function of NagZ, we constructed a nagZ deletion strain YE Z, and detected the N-acetyl-β-glucosaminidase activity of it to compare with the wild-type strain Y. enterocolitica 105.5R(r), the results showed that the ability of hydrolysis chromogenic substrate was completely lost in nagZ mutation strain YE Z (Figure 3), suggesting that NagZ (YE105_RS06670) was the only enzyme that possessed N-acetyl-β-glucosaminidase activity in Y. enterocolitica 105.5R(r). However, even though there is no readable N-acetyl-β-glucosaminidase activity in YE Z, we also did the bioinformatic search to look for possible NagZ homologs in genome to find the protein worked in YE D123 Z. According to the gene function annotation of 105.5R(r), we considered the YE105_RS13000 may have similar function with NagZ, but it was not clear if this protein participated the ampC regulation or not. Therefore, further studies needed to performed to elucidate the function of YE105_RS13000 in Y. enterocolitica ampC regulation.
In Y. enterocolitica, the function of AmpR was roughly the same as other members of Enterobacteriaceae or P. aeruginosa. The introduction of ampR into the AmpC hyperproduction strains YE D123 and YE 4 5 6 7 resulted in a sharp decline in the ampC expression (Figure 2). The inducibility of YE D123 R and YE 4 5 6 7 R also disappeared completely (Lindberg et al., 1985;.
In conclusion, in terms of AmpC β-lactamase regulation, Y. enterocolitica shared some common characteristics with P. aerugiosa and other members of Enterobacteriaceae, but it also had its own features. This was the first investigation to the characterization of Y. enterocolitica ampC regulation. It provided a more comprehensive understanding of the AmpC β-lactamase regulation in Gram-negative bacteria.

AUTHOR CONTRIBUTIONS
CL, CCL, SS, HJ, and XW designed the experiment together. YC and HH performed data analysis. JL and RD participated in the manuscript translation. ZG, JZ, and ZZ contributed to finish the work. All authors contributed to writing of the manuscript.