β-lactam resistance associated with β-lactamase production and porin alteration in clinical isolates of E. coli and K. pneumoniae

β-lactam resistance represents a worldwide problem and a serious challenge for antimicrobial treatment. Hence this research was conducted to recognize several mechanisms mediating β-lactam resistance in E. coli and K. pneumoniae clinical isolates collected from Mansoura University hospitals, Egypt. A total of 80 isolates, 45 E. coli and 35 K. pneumoniae isolates, were collected and their antibiotic susceptibility was determined by the Disc diffusion method followed by phenotypic and genotypic detection of extended-spectrum β-lactamases (ESBLs), AmpC β-lactamase, carbapenemase enzymes. The outer membrane protein porins of all isolates were analyzed and their genes were examined using gene amplification and sequencing. Also, the resistance to complement-mediated serum killing was estimated. A significant percentage of isolates (93.8%) were multidrug resistance and showed an elevated resistance to β-lactam antibiotics. The presence of either ESBL or AmpC enzymes was high among isolates (83.75%). Also, 60% of the isolated strains were carbapenemase producers. The most frequently detected gene of ESBL among all tested isolates was blaCTX-M-15 (86.3%) followed by blaTEM-1 (81.3%) and blaSHV-1 (35%) while the Amp-C gene was present in 83.75%. For carbapenemase-producing isolates, blaNDM1 was the most common (60%) followed by blaVIM-1 (35%) and blaOXA-48 (13.8%). Besides, 73.3% and 40% of E. coli and K. pneumoniae isolates respectively were serum resistant. Outer membrane protein analysis showed that 93.3% of E. coli and 95.7% of K. pneumoniae isolates lost their porins or showed modified porins. Furthermore, sequence analysis of tested porin genes in some isolates revealed the presence of frameshift mutations that produced truncated proteins of smaller size. β-lactam resistance in K. pneumoniae and E. coli isolates in our hospitals is due to a combination of β-lactamase activity and porin loss/alteration. Hence more restrictions should be applied on β-lactams usage to decrease the emergence of resistant strains.


Introduction
In the 21 st century, the emergence and dissemination of resistant bacteria to antimicrobial agents is considered a challenging and threat to global public health as antibiotic resistance leads to higher medical costs, longer hospital stays, and increased mortality [1,2]. β-lactam antibiotics are among the most commonly prescribed antibiotics due to their minimal side effects and broad antibacterial spectrum. However, various mechanisms are responsible for resistance to β-lactam compounds, such as the production of degrading enzymes, alteration of the drug target (modification of penicillin-binding proteins), decreased membrane permeability, and drug efflux pump [3,4].
β-lactam compounds resistance is a rising problem and the production of different β-lactamases was documented to be the main cause of β-lactam resistance especially among gramnegative bacilli. Carbapenems group is often used as β-lactam of last resort in treating infections caused by multi-drug resistant bacteria especially those belonging to Enterobacteriaceae. However, in this family reduced susceptibility to carbapenem group can be frequently associated with carbapenemases production, membrane impermeability coupled with elevated expression of other β-lactamases, or a combination of these mechanisms [5][6][7].
The gram-negative bacteria outer membrane permits the passive diffusion of small hydrophilic solutes and important antibiotics (β-lactams and fluoroquinolones) by channel-forming proteins [8]. Porins are considered one class of these proteins that are present in gamma-proteobacteria members such as E. coli, Salmonella, Shigella, and others [9][10][11][12]. Porins also maintain the envelope integrity of the cells and act as receptors for bacteriocins and bacteriophages. Additionally, they participate in bacterial pathogenesis such as invasion, serum resistance, and adherence [13]. In gram-negative bacteria, different porins types have been identified and classified according to their functional structure, their activity, and their regulation and expression [10].
From the gram-negative bacteria, E. coli and K. pneumoniae which are responsible for the majority of nosocomial and community-acquired infections. The three major porins that have been identified in E. coli include OmpF, OmpC, and PhoE which differ from one another according to charge and size of solutes [10,14,15]. Also, K. pneumoniae contains two main porins, Ompk35, and Ompk36, through which hydrophilic solutes gain access to bacterial-cell [10]. Loss of membrane permeability may be due to mutation in porin that renders it nonfunctional or alteration in expression level or may be due to complete loss of the porin proteins [16]. Many studies illustrated that loss of Ompk35 and/or Ompk36 in K. pneumoniae, contribute to their resistance to cephalosporins and carbapenems [17][18][19]. Also, mutations in the OmpC or OmpF genes that result in subsequent porin loss have been reported in the resistance of Enterobacter spp., E. coli, and S. marcescens, for carbapenems [20,21]. This work aimed to investigate the prevalence of porin alteration mediated resistance to β-lactam antibiotics in βlactamase producing and non-producing strains of multidrug-resistant clinical isolates of E. coli and K. pneumoniae .

Bacterial isolates
A total of 80 clinical isolates were separated from urine and sputum samples which were collected from microbiological laboratories in Mansoura University hospitals, Dakahlia governorate, Egypt. These isolates were identified as 45 isolates of E. coli and 35 isolates of K. pneumoniae using biochemical standard assay methods [22,23]. This work was done after approval of the administrative authorities (Research Ethics Committee) in the faculty of pharmacy. Mansoura University, Egypt.

Phenotypic screening for ESBLs enzymes production
The production of ESBL by clinical isolates was tested by Modified Double Disc Synergy Test (MDDST) using amoxicillin-clavulanic acid disc (20/10 μg) along with three third-generation cephalosporins discs {ceftazidime(30μg), ceftriaxone (30 μg) and cefotaxime (30 μg)} and one fourth-generation cephalosporin disc (cefepime 30 μg) [26]. The amoxicillin-clavulanate disc was centered on Mueller-Hinton agar plate lawned with test organism suspension equivalent to 0.5 McFarland standards. The discs of cephalosporins were placed around the amoxicillinclavulanate disc with a distance of 15mm (for third-generation cephalosporins discs) and 20mm (for fourth-generation cephalosporin disc), and then the plates were incubated at 37˚C. Positive results for ESBL production are considered by any increase in the inhibition zone around any of these cephalosporin discs towards the disc of amoxicillin-clavulanate.

Phenotypic screening for AmpC enzyme production
AmpC enzyme production for all isolated strains was determined by three-dimensional enzyme extract method [27]. Mueller Hinton agar plates were inoculated by E. coli DH5α and cefoxitin discs (30 μg) were centered on them. The extraction of crude enzymes from all isolates was done as mentioned by Livermore et al., 1984 [28]. A circular well and a linear slit were made in the agar plates as described before [29]. The enzyme extracts were added to the wells then the plates were kept upright for 5-10 min and incubated overnight at 37˚C. Clear distortion of cefoxitin disc inhibition zone was shown in isolates which were AmpC β-lactamase producers while AmpC β-lactamase non-producers gave no distortion.
Green PCR Master Mix (2x) (Thermo Fisher Scientific Inc, USA), 1μl of total DNA, and 9.5 μl nuclease-free water. The PCR program was done as follows: initial denaturation at 94˚C for 10 min; 35 cycles of DNA denaturation at 94˚C for 30 s, annealing at a temperature specified for each primer as listed in Table 1 for 30 S and extension at 72˚C for 1 min followed by final elongation step at 72˚C for 7 min. The PCR products were visualized by electrophoresis in 1% agarose gels stained with ethidium bromide.

Isolation and characterization of the outer-membrane proteins
The outer-membrane proteins (OMPs) of all E.coli and K. pneumoniae isolates were analyzed by SDS-PAGE [31] and compared to standard E.coli ATCC 25922 and K. pneumoniae ATCC 33495. Barwa and Shaaban procedure was used for the preparation of OMPs from all tested isolates [32]. Then OMPs were analyzed by SDS-PAGE (10%) and gels were visualized by staining with Coomassie blue.

Serum bactericidal assay
The resistance of all isolates to complement-mediated serum killing was estimated; bacteria were cultivated in nutrient broth until mid-logarithmic phase, centrifuged, washed twice with barbital buffer saline (BBS), and resuspended to a final concentration of 10 6 CFU mL -1 in BBS. 40% normal human serum (NHS) in BBS was inoculated with the bacteria for 2 hrs at 37˚C with shaking. Heat-inactivated NHS at 56˚C for 30 min (HI-NHS) was considered as control. Samples which were taken at 0 and 2 hrs were serially diluted and then cell survival was determined by plating them onto nutrient agar plates followed by overnight incubation at 37˚C.

Sequencing analysis of OMP genes
OmpF gene of three E. coli isolates (E7, E27, and E36), ompC gene of one E. coli isolate (E38) and ompK35 gene of two K. pneumoniae isolates (K5 and K 23) were sequenced as previously described [34]. In brief, the tested outer membrane protein genes were amplified using PCR, each reaction mixture (50ul) contained 1 ng of DNA, 0.5 μm of each primer, 200 μm of dNTPs, 2mM MgCl 2 ,10 μl of 5x Q5 buffer and 1 U of Q5 High-Fidelity DNA polymerase (NEB, UK). The QIA quick Gel Extraction kit (Qiagen, USA) was employed to purify PCR products according to the supplier's protocol. The purified amplicons were subsequently sequenced in both directions with the Sanger method using the BigDye 1 Terminator v3.1 Cycle Sequencing Kit and Applied Biosystems Genetic Analyzer 3500 (Thermo Fisher Scientific Inc, USA). The obtained sequence for each strain was analyzed and aligned with ompF, ompC of E. coli K12 reference gene sequences, and ompK35 reference gene sequence (GenBank Accession no JX310553.1) using Blast program of NCBI. To identify non-synonymous point mutations, DNA sequences were translated to amino acids using the BLAST software.

Statistical analysis
Data were fed to the computer and analyzed using IBM SPSS Corp. Released 2013. IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp. Qualitative data were described using number and percent. Significance of the obtained results was judged at the (0.05) level. Graph pad prism version 6.01 was used for figure design. Chi-Square test for comparison of 2 or more groups.

Identification and antimicrobial susceptibility testing of bacterial strains
In this study, 80 clinical isolates of Enterobacteriaceae were collected from several microbiological laboratories in Mansoura University Hospitals. Of these, 45 isolates were identified as E. coli and 35 isolates were identified as K. pneumoniae using the standard biotyping method. Twenty-six E. coli isolates and 16 K. pneumoniae isolates were obtained from urine samples collected from the Urology and Nephrology Center whereas 5 and 6 isolates of K. pneumoniae were separated from urine samples collected from Emergency Hospital and Children Hospital respectively. Additionally, 19 E. coli isolates and 8 isolates of K. pneumoniae were obtained from sputum samples which were collected from Emergency Hospital and Children Hospital. According to the breakpoints which were indicated in CLSI guidelines [25], the antibiotic susceptibility patterns for all isolates were analyzed. Resistance was observed to amoxicillinclavulanic acid in 93.3% of E. coli isolates followed by 91.1% were resistant to ceftazidime, ceftriaxone, cefotaxime, ofloxacin, and trimethoprim-sulfamethoxazole in E. coli isolates. Also, E. coli strains showed a high resistance level to piperacillin-tazobactam (88.9%), cefepime (68.9), and imipenem (60%). In contrast, resistance to nitrofurantoin was seen only in 35.6% of E. coli isolates, and susceptibility to amikacin was retained by most of E. coli isolates (86.7%).

Analysis of ESBLs and AmpC enzymes production
ESBLs and AmpC enzymes were both phenotypically (Fig 1) and genetically detected (Fig 2). The Determination of ESBL production by the MDDST method showed that 38 isolates of E. coli (84.4%) and 29 isolates of K. pneumoniae (82.9%) were considered positive for ESBL production (Figs 1A and 3). On the other hand, AmpC production which was detected by the three-dimensional extract method indicated that 38 E. coli isolates and 29 K. pneumoniae isolates were AmpC producers (Figs 1B and 3B). Moreover, phenotypic tests showed that 3 E. coli isolates and 6 K. pneumoniae isolates were considered ESBL and AmpC non-producers besides, 4 E. coli isolates produced only ESBL, and 4 E. coli isolates had AmpC activity only among all tested isolates while the remainder 34 E. coli produced both ESBL and AmpC enzymes.
Considering PCR analysis, 95.6% of E. coli isolates (43/45) and 85.7% of K. pneumoniae (30/35) had at least one of the tested ESBL genes. The distribution of tested ESBL genes among the clinical isolates revealed that bla TEM-1 and bla CTX-M-15 were predominant among E. coli isolates (87% and 89% respectively). In contrast, bla SHV-1 was detected in 27% of E. coli isolates. Moreover, bla SHV-1 , bla TEM-1, and bla CTX-M-15 genes were predominantly carried by K. pneumoniae isolates whereas they were detected in 66%, 74%, and 86% (Fig 4). Also, AmpC gene detection by PCR revealed that it was harbored by 38 E. coli isolates and 29 K. pneumoniae isolates (Fig 4). The two E. coli isolates and also two K. pneumoniae which found to be sensitive to all tested antimicrobials showed that they did not harbor any of the tested β-lactamase genes.

Analysis of carbapenemases presence
The presence of carbapenemases was studied through the MHT test whereas 27 isolates of E. coli (60%) and 21 K. pneumoniae isolates (60%) had a positive Hodge test hence, they produced a distorted or clover-leaf shaped inhibition zone ( Fig 1C). Also, it was found that 25 E. coli isolates and 21 K. pneumoniae isolates were positive for both ESBL and carbapenemase.
Also, PCR indicated that 29 E. coli strains and 19 K. pneumoniae strains harbored the bla NDM-1 gene, and 12 E. coli isolates and 16 K. pneumoniae isolates contained the bla VIM-1 gene ( Fig 5). However, bla OXA-48 was present in only 7 isolates of E. coli and 4 isolates of K. pneumoniae (Fig 4). The results showed that the coexistence of the ESBL encoding gene and carbapenemase encoding gene in the same strain was found among all tested E. coli isolates. Also, 21 K. pneumoniae isolates were found to carry any of the tested carbapenemase genes but in association with ESBL and AmpC presence.

Analysis of outer-membrane proteins (porins)
SDS-PAGE analysis of porins showed that 93.3% of E. coli isolates (42/45) and 85.7% of K. pneumoniae isolates (30/35) lost their porins or showed a modification in their electrophoretic migration pattern by comparison to the standard strain (Fig 6). From the results of porins analysis, it was found that 34 E. coli isolates that exhibited a loss in their porins had β-lactamase producing capability except 14 isolates had no carbapenemase-producing capability, 4 isolates did not produce AmpC enzyme and 4 isolates had no ESBL producing capability. Also, E. coli isolates with altered porin pattern had β-lactamase producing capability except for 6 isolates that could not produce carbapenemase enzymes, and 4 isolates were negative for ESBL enzymes but harbored one or more of the tested β-lactamase genes.

PLOS ONE
β-lactam resistance associated with β-lactamase production and porin alteration in clinical isolates

PLOS ONE
β-lactam resistance associated with β-lactamase production and porin alteration in clinical isolates Taking into account porin genes PCR analysis, among 45 E. coli isolates, 35 isolates exhibited the OmpC gene and 32 isolates had the OmpF gene. They had β-lactamase producing capability except 4 isolates did not produce AmpC or ESBL, 10 isolates did not produce carbapenemase, and 2 sensitive isolates which had no β-lactamase producing capability. On the other hand, 19 K. pneumoniae isolates harbored the OmpK35 gene, as 14 of them produced ESBL and AmpC enzyme and 9 of them are considered also carbapenemase producers.

PLOS ONE
β-lactam resistance associated with β-lactamase production and porin alteration in clinical isolates

Serum bactericidal assay
The survival of bacteria in 40% NHS was measured for a period of 2 hrs. For E. coli isolates, 73.3% (33/45) were serum resistant and 90.9% (30/33) of these isolates were capable of producing at least one type of the tested β-lactamases (ESBL, AmpC and carbapenemases). On the other hand, 40% (14/35) of K. pneumoniae isolates were serum resistant from which 92.8% (13/14) were positive for all 3 tested β-lactamases. Furthermore, 28.5% (6/21) of serum sensitive K. pneumoniae isolates did not produce any type of the tested β-lactamases. Interestingly, we observed that 90.9% (30/33) of serum resistant E. coli isolates and 100% (14/14) of serum resistant K. pneumoniae isolates showed either loss or alteration in their porin pattern.

Sequencing analysis of OMP genes
The sequence analysis of ompF, ompC, and ompK35 genes of representative group of isolates (nos. 38E, 7E, 27E, 36E, 5K and 23K) demonstrated that most of these tested genes contained a frameshift mutation and all of them had a stop codon at different positions (Table 2). These mutations resulted in the production of truncated proteins of a smaller size than the wild type

PLOS ONE
β-lactam resistance associated with β-lactamase production and porin alteration in clinical isolates

PLOS ONE
β-lactam resistance associated with β-lactamase production and porin alteration in clinical isolates proteins. Hence, SDS-PAGE analysis of these isolates showed that they exhibited a modified porin pattern (Fig 7). E. coli isolate no.38 had a frameshift mutation and a stop codon at amino acid position 298 in OmpC porin and accordingly, a band of lower molecular mass (33 Kda) was observed (Fig 7)

Discussion
The widespread use of antibiotics leads to the emergence of many resistant microorganisms which is increasing worldwide and becoming one of the major serious health care problems [35,36]. Our results showed a greater prevalence of resistance to common antibiotics and this agreed with that of Hassan et al. [37]. Indeed, the antibiotic susceptibility pattern of either isolated E. coli strains or K. pneumoniae strains revealed a high resistance rate (greater than 80%) which was recorded with ceftazidime, ceftriaxone, cefotaxime, amoxicillin/ clavulanic acid, trimethoprim/ sulfamethoxazole, piperacillin/ tazobactam, and ofloxacin, and these results are in agreement with other reports that showed higher resistance to sulfamethoxazole /trimethoprim (85.7%) and ciprofloxacin (72%) [  mechanism of β-lactam resistance is the production of β-lactamase especially in Enterobacteriaceae [46]. Their continuous mutation leads to various β-lactamases enzymes [47]. Among them, ESBLs are of great interest and have been reported worldwide especially in Enterobacteriaceae [26,48]. At present, more than 300 different ESBL variants have been identified [24] but the most common types which were reported in Enterobacteriaceae in several areas of the world are CTX-M, SHV, and TEM [49,50]. Also, AmpC β-lactamases have been found all over the world and contribute to the problem of third generation cephalosporin resistance [51,52]. Egypt is considered among the countries with the highest rate of ESBL production among Enterobacteriaceae [53], whereas previous studies in Egypt showed an elevated rate of ESBL producing isolates [39,44,54,55]. This coincides with our results as ESBL production was detected in 38 isolates of E. coli (84.4%) and 29 isolates of K. pneumoniae (82.9%), while a low rate of ESBL production was observed by Abdelmegeed et al., in E. coli isolates [56]. As well as higher proportions were found in other countries including India (˃80%) and China (˃60%) [57]. PCR was used for confirmation of phenotypic outcomes and discrimination of different types of genes encoding ESBL enzymes among tested isolates. Many studies in Egypt and other countries showed that the most commonly detected β-lactamase-encoding gene among the ESBL-producing Enterobacteriaceae was bla CTX-M [55,58] whereas, the bla CTX-M gene was able to be horizontally transferred by several mobile genetic elements [59]. Our data showed that bla CTX-M-15 was the most frequently observed ESBL-encoding gene (found among 89% of E. coli isolates and 86% of K. pneumoniae isolates) and this finding was consistent with the results of other studies performed in Egypt which indicated that bla CTX-M-15 was the most common ESBL encoding gene among ESBL-producing Enterobacteriaceae [55,60,61]. Besides, the next most frequently detected gene was bla TEM-1 which was found in 87% of E. coli isolates and 74% of K. pneumoniae isolates. Additionally, SHV was detected in 66% of K. pneumoniae isolates and a small percentage of E. coli isolates (27%) and these results are in agreement with a study of Abdallah et al., who indicated that CTX-M was the predominant detected ESBL gene followed by TEM then by SHV [55]. In contrast to our results, Yazdi et al., and Bajpai et al., reported that the most prevalent β-lactamase-encoding gene was bla TEM followed by bla SHV [62,63] and in Zaki et al., study SHV was the most commonly detected gene (61.22%) [39]. The wider distribution of ESBL enzymes may be due to the mobilization of their genes on a genetic element as was reported by Shahid et al. [64].
Regarding AmpC enzyme production and the prevalence of its encoding gene, it was found in a higher percentage of E. coli and K. pneumoniae isolates (84.4% and 82.9% respectively). So the resistance to amoxicillin/clavulanate was high in tested isolates as AmpC enzymes are poorly inhibited by clavulanic acid [65]. In contrast, previous studies from Egypt showed a lower prevalence of the AmpC enzyme among Enterobacteriaceae [30, 56,66]. The possible reason for the high prevalence of ESBL and AmpC producing organisms may be due to excessive usage of extended-spectrum cephalosporin in the treatment of gram-negative infection [67]. An important observation in this study is that many AmpC positive isolates were also ESBL-producer (67/80) and this result was equivocal with a study carried by Park et al., who indicated that the organism which was AmpC-producer is frequently associated with a high co-presence rate of ESBLs [67].
Carbapenems are the preferred therapy for ESBL producing Enterobacteriaceae [68]. However, the development of carbapenemases producing Enterobacteriaceae becomes a public health threat and leaves few therapeutic choices [69]. Many reports that have been published worldwide including Middle East countries on carbapenemases production among Enterobacteriaceae spp. showed a high presence of gram-negative bacteria which exhibited carbapenem resistance in different areas in Egypt [43, 70,71]. Similarly in the current study, 60% of E. coli isolates and 60% of K. pneumoniae isolates were found to be carbapenemase producers and harbored carbapenemase encoding genes; bla NDM-1 , bla VIM-1, and bla OXA-48 . This result is consistent with the high rate of carbapenemase production (69.8%) which was reported in Mahmoud et al. study [72] which may be due to widespread abuse of carbapenem in Egyptian hospitals [54,73]. In contrast, Amjad et al. illustrated that 38% of E. coli isolates and 17% of K. pneumoniae were carbapenemase production positive [74], and Zaki et al., reported that 34.1% of E. coli isolates were carbapenemase producers [39]. Among the many types of carbapenemase encoding genes, the widespread of NDM type among Enterobacteriaceae has been reported in many countries [75][76][77]. The same was observed in our results which revealed that bla NDM-1 gene was the dominant carbapenemase encoding gene in tested isolates since it was present in all carbapenemase-producing isolates except E25 and E29 which did not contain any of the tested genes and K14 and K27 which contained bla VIM-1 and bla OXA-48 only respectively. These findings confirm previous reports which indicated that bla NDM-1 gene was common in Egypt and the Middle East.
[43]. However, a low prevalence of bla NDM-1 gene was found in previous studies [66,78] and no NDM-1 producing isolates were found in a study in China [79]. Besides this, bla VIM-1 was detected in 35% of the isolated strains while bla OXA-48 was detected in 13.8% of the tested isolates, whereas OXA-48 producing Enterobacteriaceae recently detected in Egypt [32,39,80]. Moreover, OXA-48 was commonly identified among Enterobacteriaceae in other countries including Saudi Arabia [81,82]. North Africa, and Turkey [83]. Results of various studies on K. pneumoniae in Egypt are coincident with the current study as they reported that the NDM gene was the most predominant detectable carbapenemase gene followed by VIM [32,84]. On the other hand, El kholy et al found that bla OXA-48 dominated (40.6%) followed by bla NDM1 (23.7%) among K. pneumoniae isolates and a small percentage of E. coli harbored bla NDM-1 [61]. It was reported that the NDM1 gene is harbored by diverse plasmids that also carry multiple resistance genes to macrolide, rifampin, carbapenem, cephalosporin, and Sulfamethoxazole and few treatment options are available for those strains which carried this gene [85]. So it was found in our study that the strains which harbored NDM-1 showed elevated variability resistance to various tested antibiotics. The current study showed excessive coexistence of various resistance genes in tested isolate and this leads to elevated variability in resistance and this agreed with that of Martin and Bachman [86].
The β-lactam resistance in gram-negative bacteria has been attributed not only to the presence of the hydrolyzing enzymes but also to the modification in the permeability of the outer membrane as well as to upregulation of multidrug efflux pump and alteration of antibiotic target proteins [11,87]. Porin loss or mutation of the porin-coding sequence decreases membrane permeability and leads to impairment of antibiotic entry [88], hence loss of porins have been associated with a carbapenem and extended-spectrum cephalosporin resistance [89]. The mutational loss or alteration of porins in Enterobacteriaceae has been reported to be responsible for decreased susceptibility to cephalosporin and carbapenem [90,91]. The increased resistance of ESBL producer E. coli and K. pneumoniae to cefotaxime and oxyimino β-lactam due to loss of porins was reported in many studies [92,93]. Also, Domenech-Sanchez et al., found that the combination of decreased outer membrane permeability and the presence of SHV-1 and TEM-1 β-lactamase elevated cefotaxime MIC [94]. Therefore, we examined the porins as an additional resistance mechanism. Our results demonstrated that all the studied isolates showed porin loss or alteration in porin structure except three isolates of E. coli and 5 isolates of K. pneumoniae. And these results are in agreement with that obtained by Kitchel and coauthors who illustrated that the absence of OmpK35 was in 80% of K. pneumoniae isolates [95]. As well as it was observed that all of them harbored one ESBL and/or AmpC enzyme and this agreed with that of Wozniak et al. [88], additionally, this finding is consistent with that ESBL producing K. pneumoniae isolates don't express OmpK35 and both OmpK35 and Ompk36 porins were expressed in K. pneumoniae isolate lacking ESBL which was reported in other studies [94,96]. The loss of porins participates in the antimicrobial resistance of ESBL producing bacteria [16]. Besides, previous studies indicated that the absence or low expression of OmpK35in K. pneumoniae and OmpC and OmpF porins in E. cloacae has been associated with carbapenem resistance [34,97]. The present data confirm this conclusion as almost all imipenem resistant isolates showed loss or alteration in tested porins and also the carbapenemase production was detected in them except 2 out of 23 strains of K. pneumoniae and 9 out of 27 strains of E. coli did not produce carbapenemase and this finding was concordant with the results of Barwa and Shaaban who found that all tested carbapenem-resistant isolates lost OmpK35 except one strain [32]. Besides that, it was reported that porin loss increases the resistance of ESBL-producing organisms to other nonβ-lactam drugs such as fluoroquinolones [98,99] and this explains the reason for the decreased susceptibility to ofloxacin by the tested organisms.
Since serum resistance is an important virulence trait for extraintestinal pathogenic E. coli, we reported that 73.3% of E. coli isolates collected from urine or sputum were serum resistant. However, the ability of serum sensitive isolates (26.7%) to cause extraintestinal infection may be due to the presence of antibodies that inhibited complement-mediated killing as indicated by Coggon et al. [100] In the current study, the association between serum resistance and production of different β-lactamases was clear among E. coli and K. pneumoniae isolates where the possession of β-lactamases was beneficial for some strains regarding serum survival and competitive fitness as reported elsewhere [101][102][103][104]. Several studies reported that certain porins in E. coli and K. pneumonia play a role in complement system activation as they promote the deposition of C1q and thus induce antibody-dependent classical pathway bactericidal activity. Therefore, loss or mutation of these immunogenic porins results in elevated resistance to complement-mediated killing [105][106][107]. This agreed with our results where 93.6% (44/47) of serum resistant E. coli and K. pneumoniae isolates showed either porin loss or alteration in their porin electrophoretic migration pattern.
Mutations in ompK35 gene of tested K. pneumoniae isolates K5 and K23, ompF gene of E. coli isolates E7, E27, and E36 and ompC gene of E. coli isolate E38 involved a stop codon that formed truncated porins lacking the C-terminal phenylalanine residue which is essential for membrane anchoring [108] thus producing nonfunctional protein and porin cannot be inserted in the outer membrane. The same results were observed by Wozniak et al., who found a frameshift mutation and a stop codon at position 144 in the tested ompK35 genes [88]. In addition, Wozniak et al., reported that OmpK36 was the preferred porin and was thus conserved by K. pneumonia because of its smaller pore than OmpK35 and was, therefore, more restrictive for antibiotic entrance. Hence, in our study, the band observed in the SDS-PAGE analysis of 5K and 23K (Fig 7) corresponds most likely to OmpK36.
The exceeding emergence of β-lactam resistance in E. coli and K. pneumoniae strains in Mansoura, Egypt is disturbing. Moreover, our results demonstrated that β-lactam resistance in E. coli and K. pneumoniae isolates were mediated by β-lactamases plus porin loss or mutation of the porin-coding sequence. Hence, the recommendation must be taken during the administration of β-lactam and carbapenem in our hospitals to decrease the spread of β -lactam resistant isolates.