The Influence of Outer Membrane Protein on Ampicillin Resistance of Vibrio parahaemolyticus

The antibiotic resistance of the food-borne pathogen Vibrio parahaemolyticus has attracted researchers' attention in recent years, but its molecular mechanism remains poorly understood. In this study, 7 genes encoding outer membrane proteins (OMPs) were individually deleted in V. parahaemolyticus ATCC33846, and the resistance of these 7 mutants to 14 antibiotics was investigated. The results revealed that the resistance of the 7 mutants to ampicillin was significantly increased. Further exploration of 20-gene transcription changes by real time-qPCR (RT-qPCR) demonstrated that the higher ampicillin resistance might be attributed to the expression of β-lactamase and reduced peptidoglycan (PG) synthesis activity through reduced transcription of penicillin-binding proteins (PBPs), increased transcription of l,d-transpeptidases, downregulated d,d-carboxypeptidase, and alanine deficiency. This study provides a new perspective on ampicillin resistance in OMP mutants with respect to PG synthesis.


Introduction
Vibrio parahaemolyticus is a Gram-negative bacterium causing intestinal infections. It is commonly found in seafood, such as blood clams and shrimp [1], causing signifcant economic losses as an aquatic pathogen [2]. Due to the overuse of antibiotics, multidrug-resistant V. parahaemolyticus has been isolated in recent years. Te strains have demonstrated resistance to ampicillin, cefazolin, penicillin, and so on [3].
Studies investigating the mechanism of ampicillin resistance mainly focus on three points: drug permeation, peptidoglycan (PG) synthesis, and β-lactamase. In Neisseria meningitidis, a single point mutation in the porin PorB can strongly afect the binding and permeation of β-lactam antibiotics [4]. Afnity binding studies of four transformants revealed decreased afnity of PBP4 for ampicillin [5]. In Enterococcus faecium, L,D-transpeptidase-mediated resistance may emerge in various pathogens [6]. In V. parahaemolyticus, a novel class A carbenicillin-hydrolyzing β-lactamase, bla (CARB-17), was responsible for the intrinsic penicillin resistance [7].
Te role of outer membrane protein (OMP) in antibiotic resistance is usually related to permeation [8] as antibiotic susceptibility is related to OMP channel size [9]. Ampicillin enters E. coli through OmpF [10]. In carbapenem-resistantEnterobacter aerogenes, the expression of OMPs is defcient due to overexpressed sRNA or decreased due to single-point mutation [11]. In addition, performing gene knockout in V. parahaemolyticus is challenging, so few studies have reported antibiotic resistance in OMP mutants.
In this work, VP_RS22195, VP_RS23020, VP_RS16800, VP_RS16465, VP_RS20840, VP_RS03765, and VP_RS11205 were knocked out by ATCC33846. Subsequently, the growth, outer membrane (OM) permeability, and minimum inhibitory concentration (MIC) of OMP deletion mutants were evaluated. In total, 14 antibiotics targeting diferent sites in cells were selected, including β-lactams (ampicillin), aminoglycosides (streptomycin, kanamycin, gentamicin, tobramycin), quinolones (ciprofoxacin, nalidixic acid), furans (nitrofurantoin), cationic antimicrobial peptides (polymyxin B), rifampicin, clarithromycin, tetracycline, chloramphenicol, and novobiocin. After ampicillin treatment, the transcriptional changes of 20 genes in each mutant were further detected. Tis study gives a better understanding of the role of OMPs in V. parahaemolyticus OM and the cell wall. Table 1 lists all strains and plasmids used in this study. V. parahaemolyticus ATCC33846 was used to construct OMP deletion strains. Te bacteria were grown in Luria-Bertani (LB) broth medium containing 10 g/L NaCl, 10 g/L trypsin (OXOID), and 5 g/L yeast extract (OXOID) at 37°C and 200 rpm for liquid culture.

Construction of Deletion Plasmid pOTC-SB.
Te deletion plasmid pOTC-SB was composed of genes CmR, p15A, traJ, oriT, and cre. Te CmR and p15A gene fragment (1) was amplifed from template pACYC184 with primers Cm-p15A-F and Cm-p15A-R, and then digested with FastDigest enzymes SpeI and PstI (Termo Scientifc). Te traJ and oriT gene fragment (2) was amplifed from template pDS132 with primers traJ-oriT-F and traJ-oriT-R and then digested with FastDigest enzymes SalI and SpeI (Termo Scientifc). Te cre gene fragment (3) was amplifed from template pDTW109 by primers Ptac-cre-F and Ptac-cre-R. Te sacB gene fragment was amplifed from template pDS132 by primers sacB-F and sacB-R. Finally, the ligation product of (1) and (2) was connected to (3) by using one-step cloning (ClonExpress II One Step Cloning Kit, Vazyme). Te resulting plasmid was named pOTC as it contained oriT and cre. Linearized pOTC by BstZ17I and sacB were combined by one-step cloning to construct pOTC-SB.

Construction of OMP Deletion Mutants.
Upstream and downstream homology arms were amplifed from the chromosomal DNA of V. parahaemolyticus ATCC33846 by the corresponding U/D-(target gene)-F/R primers, as displayed in Table 2. Te loxL-Gm-loxR fragment was amplifed from pWJW101 [15] by primers Gm-R and Gm-F, and the homology arms and loxL-Gm-loxR were integrated by fusion PCR (4). Ten, the pDS132 plasmid and (4) were separately digested with the FastDigest enzymes XbaI or SalI and ligated by T4 ligase (VP_RS23020 was connected to pDS132 by one-step cloning). E. coli CC118 was transformed with products and cultured on LB plates containing 30 μg·ml −1 gentamicin, and the plasmid extracted from E. coli CC118 was electrotransformed into E. coliS17-l (λpir).
Following knockout, Gm was removed by Cre in pOTC-SB. Te deletion mutants were conjugated with S17-1 containing pOTC-SB and then spread-plated with 30 μg·ml −1 chloramphenicol. Agarose gel electrophoresis revealed three results. Te frst result indicated a single band with the length of homology arms plus Gm; the second result showed a single band with the length of homology arms plus 100 bp; the third result displayed a combination of the above two types of bands. To ensure thorough Cre, the strains of the frst and third colonies were cultured again in LB with chloramphenicol until the second result appeared. Ten, negative selection was performed on an LB plate with gentamicin to ensure the removal of Gm from the genome by Cre. Te pOTC-SB was removed by culturing in LB with 10% sucrose without antibiotics. Te negative selection was performed on an LB plate with chloramphenicol to verify the removal of pOTC-SB, which has chloramphenicol resistance. Te verifcation of OMP deletion mutants by agarose gel electrophoresis is shown in Figure 1.
After OMP deletion mutants were constructed, mutants and wild-type strains were streaked and purifed on the LB plate 5 times, then cultured in LB liquid overnight and stored at −70°C. Before each assay, strains were streaked and activated on the LB plate for 25 h, and then a single colony was cultured in LB liquid for 14 h. A bacterial solution was then used in each assay.

Growth Curve.
Te overnight cultured strains were added to 50 ml of LB broth medium with an initial OD 600 of 0.02. Te mixture was cultured at 37°C and 200 rpm. All assays were performed in triplicate.

Permeation Assay.
Te OD 600 of overnight-cultured strains was adjusted to 0.5. Te sediments were washed twice and suspended in 10 mmol·L −1 pH 7.4 PBS. Subsequently, 48 μl NPN solution was added to 1152 μl of the bacteria solution before being observed with the fuorescence spectrophotometer. Te detection was carried out under excitation light at 350 nm, emission light at 428 nm, and a 2.5 mm slit width. All assays were carried out in triplicates, and three parallels were performed in each group.

Te Minimum Inhibitory Concentration (MIC) Assay.
All antibiotics except ampicillin were diluted by a two-fold dilution method to adjust the concentration to 0.0078125-256 μg·ml −1 in 96-well plates. As the 2-fold dilution of high-concentration solutions would result in large intervals, another concentration setting was used for ampicillin. For solutions with a concentration under 100 μg·ml −1 , a 2-fold dilution method was used to achieve a concentration range of 0.98-125 μg·ml −1 . For solutions with a concentration range of 100-1000 μg·ml −1 , the resulting concentration was set to 100, 200, 300, . . ., 1000 μg·ml −1 . In the concentration range between 1000 and Wild-typeVibrio parahaemolyticus ATCC CC118(λpir) λ-pir lysogen of CC118 (D(ara-leu) araD DlacX 74 galE galK phoA20 thi-1rpsE rpoB argE(Am) recA1) [12] S17-1(λpir) λ-pir lysogen of S17-1 (thi pro hsdR-hsdM + recA RP4 2-Tc::Mu-Km::Tn7 (Cm r )) [ Derived from pDS132 by adding VP_RS22195 homologous arms Tis study pΔVP_RS23020 Derived from pDS132 by adding VP_RS23020 homologous arms Tis study pΔVP_RS16800 Derived from pDS132 by adding VP_RS16800 homologous arms Tis study pΔVP_RS16465 Derived from pDS132 by adding VP_RS16465 homologous arms Tis study pΔVP_RS20840 Derived from pDS132 by adding VP_RS20840 homologous arms Tis study pΔVP_RS03765 Derived from pDS132 by adding VP_RS03765 homologous arms Tis study pΔVP_RS11205 Derived from pDS132 by adding VP_RS11205 homologous arms Tis study Table 2: Primers used in this study. Names , the concentration points were set as 1000, 1250, 1500, and 1750 μg·ml −1 . In addition, for concentrations above 2000 μg·ml −1 , fnal concentrations of 2000, 2500, and 3000 μg·ml −1 were set up. Te strains cultured overnight were then transferred to 5 ml LB test tubes with an initial OD 600 of 0.02 for 4-5 h. Ten, the OD 600 of freshly cultured strains was adjusted to 0.001. Ten, 100 μl of diluted bacteria solution was added to each well, and the culture was incubated at 37°C for 18 h. OD 600 was measured by BioTek Cytation 5. All assays were carried out in duplicate and performed three times in parallel in each group.

Real-Time Polymerase Chain Reaction (RT-qPCR).
Te overnight cultured strains were transferred into 5 ml LB test tubes with an initial OD 600 of 0.02 for 4-5 h. Half of the Names Te restriction enzyme sites are in bold.
bacterial solution was taken out into a sterilized empty test tube. Ampicillin was added to the experimental groups at the fnal concentration of 16 μg·ml −1 , and sterilized water with the same volume of ampicillin was added to tubes as the control. After incubating at 37°C and 200 rpm for 0.5 h, the sediment was used for RNA extraction. Te methods used for RNA extraction, reverse transcription, and RT-qPCR were the same as those described in [17]. All assays were performed in triplicates.

Deletion of OMP Afects Cell Growth, OM Permeation, and
Antibiotic Resistance. Bacterial growth has four phases, including the lag phase, the exponential phase, the stationary phase, and the decline phase. In the lag phase, bacteria produce new enzymes to digest, build biomass, and prepare for cell division [18]. ATCC33846 started to grow after a lag phase of 2 hours following the inoculum. Compared to ATCC33846, some OMP mutants demonstrated a shorter lag phase, including ΔVP_RS22195, ΔVP_RS16800, ΔVP_RS16465, and ΔVP_RS20840 ( Figure 2). Te shorter lag phase indicated that the OMPs were the proteins prepared in the lag phase. Deletion of OMPs lightened the burden of preparation for cells in the lagged phase. In the exponential phase, ΔVP_RS16800, ΔVP_RS16465, ΔVP_RS20840, and ΔVP_RS03765 (Figures 2(c)-2(f )) had a higher growth rate than ATCC33846. Furthermore, ΔVP_RS22195, ΔVP_RS16800, ΔVP_RS16465, ΔVP_RS20840, and ΔVP_RS03765 entered the decline phase directly, without an obvious stationary phase. In contrast, ΔVP_RS23020 and ΔVP_RS11205 showed poor growth (Figures 2(b) and 2(g)).
Te N-Phenyl-1-naphthylamine (NPN) probe was used to test the OM permeability of OMP deletion mutants. NPN can be excited to form green fuorescence in the bacterial inner membrane, a hydrophobic environment [19]. A decreased OM permeability was observed in ΔVP_RS16465 ( Figure 3).
Fourteen antibiotics were selected for the MIC assay ( Figure 4). In ΔVP_RS23020, increased resistance to antibiotics (ciprofoxacin, nalidixic acid, and novobiocin) inhibiting DNA synthesis was observed. Rifampicin inhibits RNA synthesis, and a 4-fold increase in the rifampicin MIC was observed in ΔVP_RS22195. Resistance to antibiotics (clarithromycin, chloramphenicol, tetracycline, streptomycin, kanamycin, gentamicin, and tobramycin) inhibiting protein synthesis was all increased in ΔVP_RS22195 and ΔVP_RS20840. Among protein synthesis-inhibiting antibiotics, resistance to 3 aminoglycosides (streptomycin, kanamycin, and gentamicin) was increased, whereas the fold change of tobramycin MIC was not obvious. Notably, ΔVP_RS11205 was more resistant to aminoglycosides than other mutants, and the tetracycline susceptibility of ΔVP_RS23020 was increased 2-fold. Nitrofurantoin inhibits carbohydrate metabolism enzymes and interferes with cell wall synthesis [20]. Resistance to nitrofurantoin was not decreased after OMP deletion. In addition, resistance to β-lactams, which act on peptidoglycan (PG) synthesis, showed a general increase. Polymyxin B damages Gramnegative bacterial OM and resistance to polymyxin B was increased 2.8-fold in ΔVP_RS20840. All OMP mutants showed at least a 64-fold increase in ampicillin resistance.

Changes in the Transcription of 20 Genes in OMP Deletion
Mutants under the Absence and Stimulation of Ampicillin. RT-qPCR was performed to detect the transcriptional changes of 20 genes to further study the causes of increased ampicillin resistance in OMP deletion mutants (Table 3). Tese genes can be divided into fve types: OMP genes, β-lactamase genes, PG synthesis-related genes, stress-regulation-related genes, and lipid A synthesis genes. VP_RS17515 expresses β-lactamase. VP_RS13510, mrcB, mrdA, VP_RS02165, VP_RS03450, VP_RS22785, dacB, VP_RS22200, VP_RS09310, VP_RS15980 were selected as PG synthesisrelated genes through BLASTp in NCBI, and the proteins were homologous with PBP1A, PBP1B, PBP2, PBP3, PBP5, PBP4, LdtA, LdtD, and LdtF, respectively, in E. coli. In order to directly observe gene regulation under membrane stress, uhpA and VP_RS14060, which are homologous to the response regulator proteins rcsB and cpxR in E. coli, respectively, were selected. lpxA is a lipid A synthesis gene. β-lactams inhibit bacterial growth by binding to penicillinbinding proteins (PBPs) and interfering with PG synthesis [27]. According to the growth curve, bacteria in the initial exponential phase were used as the experiment sample. Te ampicillin MIC value of ATCC33846 was 3.90625-15.625 μg·mL −1 , while the MIC values of OMP mutants were above 600 μg·mL −1 . Terefore, 16 μg·mL −1 ampicillin, which would not kill ATCC33846 and have a better efect on mutants, was set as the pretreatment. Figure 5 shows the transcription changes of 20 genes in mutants compared to ATCC33846. In ΔVP_RS23020 ( Figure 5(b)), all genes exhibited transcriptional upregulation except for VP_RS11205. In ΔVP_RS16465 and ΔVP_RS20840 (Figures 5(d) and 5(e)), VP_RS23020 was upregulated over 9fold. Te transcription of VP_RS17515 was also upregulated over 9-fold in ΔVP_RS23020 but downregulated in ΔVP_RS22195, ΔVP_RS16800, ΔVP_RS20840, ΔVP_RS03765,and ΔVP_RS11205 ( Figures 5(a), 5(c), 5(e)-5(g)). Most PG synthesis genes were downregulated in mutants except for ΔVP_RS23020. However, in ΔVP_RS23020, the transcription of VP_RS22200, VP_RS09310, and VP_RS15980 (Ldts) were all upregulated. Among PG synthesis genes, the transcriptional changes of VP_RS22785 and VP_RS09310 showed an interesting phenomenon. Tese two genes were downregulated over 2-fold in ΔVP_RS22195, ΔVP_RS16800, ΔVP_RS16465, and ΔVP_RS20840. In contrast, they were upregulated over 2-fold in ΔVP_RS23030 and ΔVP_RS11205, VP_RS22785, and VP_RS09310. In ΔVP_RS03765, these two genes showed no obvious downregulation compared to other genes. Moreover, VP_RS20840 showed the same trend as VP_RS22785 in all mutants. Moreover, the transcription of lpxA was downregulated in all mutants except for ΔVP_RS23020. Figure 6 displays the transcription changes of OMP genes of ATCC33846 and mutants under ampicillin  stimulation compared to the untreated group. All OMP genes were downregulated in ΔVP_RS22195, ΔVP_RS23020, and ΔVP_RS20840 (Figures 6(b) 6(c), and 6(f )). Furthermore, VP_RS16800 was downregulated over 2-fold in ATCC33846, ΔVP_RS23020, ΔVP_RS20840, and ΔVP_RS03765 (Figures 6(a), 6(c), 6(f ), and 6(g)), and over 4fold in ΔVP_RS22195 and ΔVP_RS11205 (Figures 6(b) and 6(h)). VP_RS17515 demonstrated over 2-fold upregulation in ATCC33846, ΔVP_RS16800, and ΔVP_RS16465 (Figures 6(a), 6(d), and 6(e)). In addition, transcription of PG synthesis genes showed signifcant changes in OMP deletion mutants but not after ampicillin stimulation. Tese fndings implied that OMP deletion caused internal resistance to ampicillin instead of inducing a stress response with stimulation of ampicillin.

Discussion
Te MIC change folds showed two major increases in resistance to aminoglycosides and ampicillin (Figure 4), and the MIC value is shown in Table S1. Te MIC values of 3 kinds of aminoglycosides (streptomycin, kanamycin, and gentamicin) increased in all 7 OMP mutant strains, indicating that the mechanism of aminoglycoside resistance could be related to OM. Aminoglycosides act on bacterial ribosomes and inhibit translation [28]. Moreover, aminoglycosides are polycationic at physiological pH and can replace divalent cations on lipopolysaccharide, thereby increasing membrane permeability [29]. However, the results of the MIC fold change of polymyxin B, which could also damage OM by replacing divalent cations on lipopolysaccharides, showed that not all the OM of mutants was more resistant to cationic antibiotics. In addition, after the deletion of OMP genes, the fold change of tobramycin's MIC was lower than that of other aminoglycosides, and ΔVP_RS03765 was sensitive to tobramycin. Terefore, the aminoglycoside resistance mechanism remains unclear in OMP deletion mutants.
A signifcant increase in ampicillin resistance was observed, with ATCC33846 showing a small MIC value (fuctuating 3.90625-15.625 μg·mL −1 ). However, 10.6% of V. parahaemolyticus isolates from the coast of Korea were sensitive to ampicillin, while 87.2% were resistant [30]. It is speculated that the low ampicillin MIC of ATCC33846 was caused by repeated subculturing through diferent generations without antibiotics in the laboratory and an ATCC33846 sample at the early exponential phase. In addition, diferent CARB β-lactamases in various V. parahaemolyticus strains exhibit diferent ampicillin hydrolysis rates [31].
Ampicillin resistance in diferent OMP mutants may result from reduced PG synthesis activity and expression of β-lactamase. PG synthesis can be afected by reduced transcription of PBPs, increased transcription of Ldts, downregulated D,D-carboxypeptidase, and alanine defciency.
Reduced transcription of PBPs was inferred to be one cause of ampicillin resistance. In ΔVP_RS22195, ΔVP_RS03765,and ΔVP_RS11205, PG synthesis-related genes were downregulated (Figures 5(a), 5(f ), and 5(g)). VP_RS22195 is located adjacent to VP_RS22200 in the genome and has 40.45% protein similarity to the murein lipoprotein Lpp in E. coli (Table 3). Furthermore, the Nterminus of lipoprotein LpoB is required for the activation of PBP1B in E. coli [32]. Lipoprotein NlpI is a part of PG biosynthetic multienzyme complexes and acts as an adaptor [33]. In addition, transcription of uhpA in ΔVP_RS22195 was downregulated over 15-fold ( Figure 5(a)). Te lack of lipoprotein increases the periplasmic distance, and OM stress signals cannot be transmitted by the Rcs system [34]. PG synthesis was afected after the deletion of VP_RS22195, resulting in a slower division rate in the exponential phase than ATCC33846 (Figure 2(a)). Another cause might be the upregulated Ldts in ΔVP_RS23020. In a β-lactam-resistant mutant of Enterococcus faecium, Ldt fm was found to account for β-lactam resistance by using a diferent substrate from D,D-transpeptidase [6].
ΔVP_RS16465 and ΔVP_RS20840 might exhibit ampicillin resistance due to the downregulated transcription of VP_RS22785 (Figures 5(d) and 5(e)), which encodes a D,Dcarboxypeptidase cleaving pentapeptides into tetrapeptides [23]. Te V. choleraeD,D-endopeptidase ShyA could recognize but not cleave dimers containing pentapeptides [35]. Terefore, PG was protected from being cleaved by D,D-endopeptidases, and existed PG was maintained, reducing the activity of PG synthesis to resist ampicillin. In addition, with lower tetrapeptides, the transcription of VP_RS09310 was downregulated, which is homologous to L,D-transpeptidase LdtD cleaving tetrapeptide to form mDAP 3 -mDAP 3 crosslinks (Table 3). However, the deletion of PBP5 in E. coli increased the sensitivity to β-lactams [36]. It is speculated that 3 homologs of PBP5 in V. parahaemolyticus and one of the downregulated homologs do not exert a signifcant impact on maintaining the normal cell shape. In addition, the OMP gene VP_RS20840 showed the same trend in transcriptional change as VP_RS22785 in all OMP mutants ( Figure 5). Tis fnding implies an unknown relationship between VP_RS22785 and VP_RS20840. In addition, alanine defciency may be involved in increased ampicillin resistance. Previous research reported lysis in an E. coli strain lacking alanine racemase in the absence of D-ala, which is mainly caused by defects in PG synthesis [37]. OmpA was upregulated after the addition of alanine through analysis of proteomics and RT-qPCR [38]. It is speculated that lower levels of alanine entered the cell following OmpA deletion, which might further afect the transcription of PBP genes. VP_RS16465, VP_RS20840, and VP_RS03765 belong to the OmpA protein family (Table 3), while the diferent transcription trends of VP_RS22785 inferred diferent functions of OmpA. Te existence and expression of β-lactamase were one of the causes of ampicillin resistance in V. parahaemolyticus. After ampicillin treatment, VP_RS17515 (β-lactamase) was upregulated in other strains except for ΔVP_RS23020 ( Figure 6(c)). Te same conclusion was obtained in V. parahaemolyticus V110 [7]. Although the transcription of VP_RS17515 was slightly downregulated in ΔVP_RS23020 under ampicillin stimulation (Figure 6(c)), it was upregulated more than 4-fold in ΔVP_RS23020 ( Figure 5(b)). VP_RS23020 encodes maltoporin. In addition, the maltose metabolism pathway was potentially involved in the resistance to antibiotics that target cell wall biosynthesis. In a Lactococcus lactis strain resistant to lactococcin 972, which is a bacteriocin that inhibits cell wall biosynthesis by binding to lipid II, maltose metabolic genes were deleted. However, this strain showed no lactococcin 972 sensitivity in the maltose medium [39]. VP_RS13510 PBP1A family penicillin-binding protein Peptidoglycan synthase which is essential for cell elongation [22] mrcB Penicillin-binding protein 1B Peptidoglycan synthase which is essential for cell division [22] mrdA Penicillin-binding protein 2 Transpeptidase essential for cell elongation [23] VP_RS02165 Penicillin-binding protein 3 Transpeptidase essential for cell division [23] VP_RS22785 D-alanyl-D-alanine carboxypeptidase D,D-carboxypeptidase which cleaves terminal D-ala in the peptidoglycan [24] dacB Serine-typeD-Ala-D-Ala carboxypeptidase D,D-carboxypeptidase which cleaves terminal D-ala in the peptidoglycan [23] VP_RS22200 L,D-transpeptidase family protein Attaching Lpp to mDAP 3 of peptidoglycan [23] VP_RS09310, VP_RS15980 L,D-transpeptidase family protein Synthesizing mDAP 3 -mDAP 3 cross-links in the peptidoglycan [23] uhpA Transcriptional regulator, 29.06% similar to RcsB in E. coli RcsB is response regulator of Rcs sensing outer membrane stress [25] VP_RS14060 Response regulator, 62.45% similar to CpxR in E. coli CpxR is response regulator of Cpx sensing inner membrane stress [26] lpxA Acyl-ACP-UDP-N-acetylglucosamine O-acyltransferase Essential for the biosynthesis of lipid A Lipoprotein potentially afected OM biosynthesis through phospholipids. Transcription of lpxA was downregulated over 3.7-fold in ΔVP_RS22195 (Figure 5(a)), indicating that the lack of VP_RS22195 had an efect on OM synthesis. Te maturity of lipoprotein is associated with phosphatidylglycerol [40]. Moreover, crosstalk between phospholipids and lipopolysaccharide synthesis was observed. LpxK catalyzes the synthesis of lipid IV A from lipid A disaccharide, which depends on the concentration of unsaturated fatty acids [41]. Furthermore, transcription of lpxA was downregulated over 2-fold in ΔVP_RS16800, ΔVP_RS20840, ΔVP_RS03765, and ΔVP_RS11205. Nevertheless, the relationship between these OMPs and OM synthesis remains unknown.  VP_RS22195  VP_RS23020  VP_RS16800  VP_RS16465  VP_RS20840  VP_RS03765  VP_RS11205  VP_RS13510  VP_RS02165  VP_RS22785  VP_RS22200  VP_RS09310  VP_RS15980  VP_RS14060  mcrB  mrdA  dacB  VP_RS17515  uhpA  IpxA   ΔVP_RS11205   ×   VP_RS22195  VP_RS23020  VP_RS16800  VP_RS16465  VP_RS20840  VP_RS03765  VP_RS11205  VP_RS13510  VP_RS02165  VP_RS22785  VP_RS22200  VP_RS09310  VP_RS15980  VP_RS14060  mcrB  mrdA  dacB  VP_RS17515  uhpA  IpxA  - (f) (g) Figure 5: Comparison of transcription of 20 genes in the 7 V. parahaemolyticus OMP deletion mutants, using wild-type ATCC33846 as a control. Five kinds of genes are divided by dotted line: OMP genes, β-lactamase genes, PG synthesis-related genes, stress-regulation-related genes, and lipid A synthesis genes. VP_RS17515 expresses β-lactamase. Te transcription change fold is calculated by 2 −ΔΔCt . Each bar represents the mean value of three biological replicates, and error bars represent standard deviations calculated from three biological replicates. -ΔΔCt  Figure 6: Comparison of transcription of 20 genes in the 7 V. parahaemolyticus OMP deletion mutants under stimulation of ampicillin (AMP), using wild-type ATCC33846 as a control. Five kinds of genes are divided by dotted line: OMP genes, β-lactamase genes, PG synthesis-related genes, stress-regulation-related genes, and lipid A synthesis genes. Te transcription change fold is calculated by 2 −ΔΔCt . Each bar represents the mean value of three biological replicates, and the error bars represent standard deviations calculated from three biological replicates.

Conclusions
Deletion of OMP afects growth and OM permeation, and MIC and OMP mutants demonstrated signifcantly increased ampicillin resistance. Further RT-qPCR analysis showed several possible causes of ampicillin resistance in OMP mutants, including the expression of β-lactamase, the reduction of PG synthesis activity due to reduced transcription of PBPs, increased transcription of Ldts, downregulated D,D-carboxypeptidase, and alanine defciency.
Tis study provides a new perspective on ampicillin resistance in OMP mutants with respect to PG synthesis. Future work will focus on the role of OMPs in the synthesis of OM and PG.

Data Availability
All the data generated or analysed during this study are included within the article.

Conflicts of Interest
Te authors declare that they have no conficts of interest.