Prevalence and characterization of plasmid-mediated quinolone resistance genes in Proteus species isolated from different patients

Proteus spp. are widely distributed opportunistic pathogens that can cause various human infections. A total of 361 clinical specimens were obtained from patients who were attending to different hospitals in El-Minia governorate, Egypt. Approximately 23 % of the samples belong to Proteus spp. isolates which were obtained from various clinical sources. After biochemical identification, 42.1 % of isolates were found to belong to Proteus vulgaris and 57.8 % to P. mirabilis . The urine samples collected from catheterized patients represented 32.6 % of all the clinical specimens, and the majority of the recorded isolates were Proteus spp. The antibacterial sensitivity of the Proteus spp. was examined using 16 different antibiotics from various families. The most effective antibiotics were Amikacin; Levofloxacin, and Meropenem, recording 68.6 %, 66.2 %, and 62.2 % of the isolates sensitivity to each of these antibiotics, respectively. Using the ureR -based PCR, 48 % of the isolates were identified as P. mirabilis . Moreover, the Qnr genes ( i.e ., qnrA , qnrB , qnrS , qnrD , and qnrC ) and the aac (6')- Ib-cr gene had been identified in 40 % of P. mirabilis isolates. The aims of the study were to investigate the prevalence of Proteus spp. in El-Minia, Egypt; determine the antibacterial susceptibility pattern of these isolates, and characterize the PMQR genes in Proteus spp. Quinolone resistance in P. mirabilis isolates might have been brought on by mechanisms other than qnr and aac (6')-Ib genes. Finally, since Proteus spp. are widespread in the environment; healthcare facilities must uphold stringent sanitation standards to reduce the incidence of the nosocomial infections.


Introduction
In 1885, Hauser identified the genus Proteus for the first time (Sadiq and Abd Alhadee, 2009). Proteus spp. are Gram-negative rods that are members of the Enterobacteriaceae family (Knirel et al., 2011). These species are parts of the humans and animals' intestinal tract normal bacterial flora, and due to fecal pollution; these species are prevalent in the soil and water (O'Hara et al., 2000).
Currently, the identified species of the genus Proteus, include P. mirabilis; P. vulgaris, P. penneri, P. cibarius, P. terrae, P. hauseri, and three unnamed genospecies 4, 5, and 6 (Girlich et al., 2020). All species of this genus; with the exception of P. cibarius and P. terrae, have been isolated from clinical specimens sampled from humans (Hamilton et al., 2018). However, P. mirabilis and P. vulgaris account for the vast majority of the clinical isolates (Ioannou and Vougiouklakis, 2020).
After Escherichia coli and Klebsiella pneumonia; P. mirabilis is the third most prevalent cause of urinary tract infections (UTI) and complicating UTIs in those patients undergoing long-term catheterization (Filipiak et al., 2020). It can also cause infections of the eye; ear, nose, and skin, as well as infections of the wounds; respiratory system, and osteomyelitis (Schaffer and Pearson, 2017).
Proteus' infectious nature is typically correlated with their antibiotic resistance, which helps their survival in various healthcare facilities. Proteus has a number of virulence features, including the ability to adhere to surfaces; penetration into the host's body, and dissemination throughout the body. Swimming; swarming, and twitching are three kinds of movement recognized in the genus Proteus (Hola et al., 2012). The swarming phenomenon is the most characteristic feature that distinguishes Proteus rods from the other members of its family (Rózalski et al., 1997). It has many other virulence factors, including adhesion by five common fimbriae type; flagella, toxins, several enzymes [i.e., hemolysin, Proteus toxic agglutinin (Pta), and urease], quorum sensing, and immune evasion (Hayder et al., 2020). The ability of Proteus to form a biofilm is considered as an important virulence factor, which helps this genus in its survival and antibacterial resistance (El-Kazzaz, 2021). UreR is the only known regulatory gene observed in P. mirabilis, which can only be detected in urea-inducible gene clusters (Zhang et al., 2013).
The resistance of P. mirabilis to the routinely used treatments is increasing by time (Allawi and Motaweq, 2019). Two mechanisms of quinolone resistance can be found: Chromosomal-mediated resistance and plasmid-mediated resistance. The chromosomalmediated resistance includes mutations in the quinolones' primary target molecules "Quinolone resistance determining region (QRDR)" in DNA gyrase and topoisomerase IV. The GyrA and GyrB encode for the tetrameric enzyme DNA gyrase, which is composed of two subunits. The ParC and ParE encode for the topoisomerase IV enzyme also (Pathirana et al., 2018).
Three mechanisms are involved in the plasmidmediated quinolone resistance (PMQR): Qnr genes, which include the qnrA; qnrB, qnrC, qnrD, and qnrS genes that mediate target protection, by interacting with DNA gyrase and topoisomerase IV. Moreover, they confer resistance to the quinolones by inhibiting quinolone entry into the enzyme-DNA cleavage complexes (Mirzaei et al., 2019;Yassine et al., 2019). The second mechanism is mediated by a type of an aminoglycoside-modifying enzyme aac(6')Ib; called aac(6')Ib-cr, which acts by inactivating the fluoroquinolones via an unprotected amino nitrogen on the piperazine ring. Finally, PMQR resistance that is mediated through plasmid-mediated quinolone efflux Novel Research in Microbiology Journal, 2023 pumps, such as those encoded by the QepA and OqxAB genes (Vila et al., 2011).
Because of the significance of Proteus species as opportunistic pathogens that can cause a range of human diseases, the objectives of this study were isolation and detection of Proteus spp. from different clinical sources; molecular characterization of P. mirabilis using polymerase chain reaction (PCR) by detection of the urease enzyme gene; ureR, which is regarded as a diagnostic feature of these bacteria, and studying the effect of different antibacterial agents against these tested Proteus isolates.

Bacterial isolation
From November 2017 to July 2018; approximately 361 clinical urine samples were collected from patients undergoing treatments at El-Minia governorate hospitals, including El-Minia General Hospital; El-Minia University Main Hospital, and El-Minia University Hospital for Kidney and Urology. The specimens were collected from urinary tract infection patients (70 specimens); catheterized patients (49 specimen), otitis media (78 specimen), wound and burn infections (153 specimens), and swabs from sore throat (11 specimens). For initial bacterial isolation and diagnosis, all specimens were cultured individually on different culture media, such as Macconkey agar. Subsequent identification was based on the cultural features and performing different biochemical assays, in reference to Kamel and Al-Yasseen, (2009).

Identification of Proteus isolates
Proteus spp. had been identified using different biochemical assays, such as Indole test. The bacterial isolates were inoculated individually in Sulfideindole-motility (SIM) medium, incubated for 24-48 h at 37 °C, and then 0.5 ml of Kovac's reagent was added. The tubes were gently shaken and allowed to stand for a few minutes. The formation of a cherry red ring layer on the medium's surface indicated that the isolate belong to P. vulgaris, whereas P. mirabilis isolates yielded negative results (Ahmed, 2015).
Meanwhile, all isolates were tested for chloramphenicol (30 μg) resistance except for the urine isolates. The diameter of each antibacterial disc's inhibition zone (mm) was measured using a calibrated ruler. Results were interpreted according to the tables of performance standards for antimicrobial discs susceptibility tests that were outlined by CLSI. (Mirzaei et al., 2019).

Molecular studies
Identification of P. mirabilis was carried out by amplification of ureR gene, and characterization of the PMQR genes in the Proteus spp. was performed using PCR-based technique.

Bacterial DNA extraction
The bacterial DNA was extracted using the boiling procedure. A loopful of bacteria that was cultivated on nutrient agar (NA) was inoculated into 500 µl of sterile dist. water, and mixed for 30 s using a vortex. After 10 min. of boiling at 100°C, the resulting suspension was chilled for an additional 10 min, and centrifuged (8000 ×g) for 5 min. The supernatant was Novel Research in Microbiology Journal, 2023 collected and used as a template DNA (Cattoir et al., 2007).

Detection of P. mirabilis isolates using PCR technique
In this study, in order to identify P. mirabilis as being the most recovered Proteus isolates, ureR-based PCR was used, which amplifies a specific region of the ureR gene found in P. mirabilis. The forward and reverse primers; ureRF1: 5′-GGTGAGATTTGTATTAATGG-3′, and ureRR1: 5′-ATAATCTGGAAGATGACGAG-3′; respectively, were used to amplify the 225 bp DNA product of P. mirabilis. The employed reaction conditions were as following; 4 min. of denaturation at 94 °C, 40 sec of denaturation at 94 °C for 30 cycles, 1 min. of annealing at 58 °C, 20 sec of extension at 72 °C, and 10 min. of extension at 72 °C. To determine the size of the bacterial DNA, the PCR products were electrophoresed using a 1% agarose gel, stained with ethidium bromide, and visualized under ultraviolet light (Zhang et al., 2013).

Detection of the PMQR genes by multiplex PCR assay
To detect the QNR genes (i.e., qnrA; qnrB, qnrS, qnrD, and qnrC) and aac (6')-Ib-cr gene, which confer Proteus' resistance to the quinolones, multiplex PCR based technique was used. More than one target sequence was amplified using more than one pair of primers in a single reaction tube, which was carried out employing the primers provided in Table (1). The used processing conditions were; Initial denaturation for 2 min. at 95 °C, followed by 35 cycles at 95 °C for 50 sec; at 53 °C for 40 sec, and at 72 °C for 80 sec, and a final extension at 72 °C for 5 min. (Dasgupta et al., 2015;Park et al., 2006).

Statistical analysis
The analysis of data was carried out using the IBM SPSS version 25 statistical package software. Data were expressed as number and percentage for qualitative data. Analyses were done to compare categorical variables using Fisher's exact test. Kappa test was used to assess the agreement between two variables. p-value less than 0.05 was considered statistically significant.

Bacterial isolation
Approximately 83 isolates were found to be belonging to Proteus spp. as a result of growth of these isolates on MacConkey agar and recording positive results in the different biochemical assays. Out of the 361 samples collected from patients with various infections, 16 isolates were isolated from catheterized patients' urine samples; 18 isolates from urine samples collected from hospitalized patients with urinary tract infections, 27 isolates from infected wounds specimens, and 22 isolates were recovered from otitis media, as illustrated in Table (2).

Identification of Proteus isolates
Using an Indole test, about 48 P. mirabilis isolates and 35 P. vulgaris isolates were identified. The P. mirabilis isolates were recovered from urine samples collected from urinary tract infected hospitalized patients; urine samples from catheterized patients (9), infected wounds (12), and from otitis media (16). On the other hand, P. vulgaris isolates were obtained from urine samples from urinary tract infected hospitalized patients (7), urine samples from catheterized patients (7), infected wounds (15), and from otitis media(6), as presented in Table (3).

Confirmation of the identity of P. mirabilis isolates using PCR
Using ureR-based PCR method; 40/ 83 isolates were confirmed as P. mirabilis (48.0 %). The PCR product band was detected at 225 bp, as shown in Fig.  (1).

PMQR characterization via multiplex PCR assay
Using multiplex PCR, about 16/ 40 (40 %) Proteus spp. were observed to harbor some of the Qnr genes and aac(6')Ib gene, as illustrated in Table (5). The qnrA gene was detected in 6 (15 %) of P. mirabilis isolates at 580 bp, as shown in Fig. (2). The qnrB gene was observed in 6 (15 %) isolates at 496 bp (Fig. 3). Meanwhile, four (10 %) isolates were recorded to harbor the qnrD gene that was detected at 582 bp (Fig. 4), and 2 (5%) isolates had the qnrS gene at 428 bp (Fig. 5). However, the qnrC gene could not be detected in any isolate. About 4 isolates (10 %) were recorded to have the aac(6')Ib gene, which was observed at 482 bp, as shown in Fig. (6     The percentage (%) of P. mirabilis isolates that had qnr and aac(6')Ib genes were correlated to the total number of confirmed P. mirabilis isolates

Relationship between antibacterial sensitivity and the occurrence of PMQR genes
The relation between the antibacterial patterns of ciprofloxacin (10 μg), levofloxacin (5 μg), and the presence of PMQR genes and aac(6')Ib gene, is demonstrated in Table (6). The results showed that only 4/6 isolates that had qnrA gene were both ciprofloxacin and levofloxacin resistant, 4/6 isolates that bear the qnrB gene were ciprofloxacin resistant, and 2/6 were levofloxacin resistant. All isolates that expressed the qnrD gene were sensitive to both ciprofloxacin and levofloxacin; however, all isolates that harbor the qnrS gene were sensitive to both ciprofloxacin and levofloxacin. Meanwhile, all isolates that had aac(6')Ib gene were resistant to both of ciprofloxacin and levofloxacin. There was a significant correlation between the occurrence of aac(6')Ib gene and levofloxacin resistance (p value < 0.05).

Discussion
Overall, in both of the community and hospital environments; Proteus spp. are regarded as significant infectious pathogenic bacteria. Approximately 361 clinical samples were collected in this study to investigate the presence of Proteus spp. Proteus bacteria were detected in 23 % of the tested samples. This result is compatible with that observed by Parajuli et al., (2021), while a similar study reported by Bahashwan and El Shafey, (2013) revealed that only 3 % of all isolated bacteria were identified as Proteus spp.
The various kinds of virulence factors of Proteus bacteria have a significant role in its invasiveness and therefore the increased percentage of its isolates. In addition, the improper use of antibacterial agents increases the infections caused by Proteus spp., the use of infected urinary catheters and\or other Novel Research in Microbiology Journal, 2023 Where; * Represent the significant statistical differences that were detected using Fisher's exact test (p value < 0.05) indwelling equipment in an unsanitary environment in some hospitals, may also have a role in this problem.
P. mirabilis and P. vulgaris were the two Proteus spp. isolated in the present study. P. mirabilis accounted for 57.8 % of the isolates while P. vulgaris accounted for only 42.1 %. These results are consistent with those presented by Ahmed, (2015) study, which reported that the rate of isolation for P. mirabilis was (66.6 %), while for P. vulgaris it was 33.3 %. Our study revealed that samples of urine collected from catheterized patients have the highest prevalence rate of Proteus spp. (32.6 %), followed by otitis media (28.2 %). In a previous study carried out in Baghdad at 2015, Proteus spp. isolated from UTI represented 60 %; those isolated from wounds accounted for 23.3%, and 16.6 % were isolated from burns (Ahmed, 2015). In contrast to this result, a study performed in Ghana at 2010 revealed that wound isolates were the most common (64.5 %), followed by ear swabs (Patrick et al., 2010). Another study conducted in India reported that 80.2 % of the Proteus spp. were isolated from pus, while 8.9 % were recovered from urine (Pal et al., 2016).
The widespread antibacterial resistance of Proteus spp. isolated from various clinical samples has been reported in several previous studies reported by Kumburu et al., (2017); Allawi and Motaweq, (2019). In the current study, about 82 Proteus isolates (98.7 %) showed resistance to Ampicillin, where the widespread usage of antibiotics in the treatment of numerous infectious disorders may account for this result. Meanwhile, the combination of amoxicillin and clavulanic acid was effective for only 5 isolates, recording 92.7 % of resistance. Similar results were observed in several previous studies reported by Sanches et al., (2019); Thabit et al., (2020); Parajuli et al., (2021 Kumburu et al., (2017) reported that Proteus isolates had 53.6 % resistance to ceftriaxone. Meanwhile, at 2019 in Saudi Arabia, the recorded sensitivity pattern of Proteus isolates were 78.2 % , 21.8 %, 23.2 % for cefoxitin; ceftriaxone and cefepime, respectively (Bandy and Tantry, 2021). The resistance to carbapenem antibiotics (i.e., Imipenem and Meropenem) was also observed in this study, where meropenem was more effective than imipenem, as 52 Proteus isolates were sensitive to meropenem (62.6 %), while only 44 Proteus isolates were sensitive to imipenem (53 %). Similarly, in India at 2008, the susceptibility of Proteus isolates to imipenem and merpenem antibiotics was 96.8 % and 100 %, respectively (Sonavane et al., 2008). In Iran at 2010, the Proteus isolates susceptibility to imipenem was 100 % (Khorshidi and Sharif, 2010). In another study conducted in Egypt by Bahashwan and El Shafey, (2013), imipenem had the highest sensitivity percentage (91 %) against Proteus isolates. While in Sudan at 2020, the Proteus isolates showed complete sensitivity to imipeneme (100 %) and 94.4 % sensitivity to meropenem (Hamid et al., 2020). Studying the antibacterial resistance to aminoglycosides (i.e., Gentamicin and Amikacin) showed that amikacin was more effective than gentamicin, as 57 Proteus isolates were sensitive to amikacin (68.6 %) and 48 isolates only were sensitive to gentamicin (57.8 %), in consistent with the findings of the previous study reported by Kadhim, (2017), where the Proteus isolates demonstrated different sensitivities of 62.7 % and 58.8 % to amikacin and gentamicin, respectively. In a later study, Han et al., (2020) recorded that the resistance to amikacin and gentamicin was 20 % and 46 %, respectively. Currently, the recorded quinolone resistance of Proteus spp. (i.e., Ciprofloxacin and Levofloxacin) was 46.9 % and 33.7%, respectively. In a similar study carried out in India by Rao et al., (2014), the observed antibacterial resistance to ciprofloxacin and levofloxacin was 37 %, 13 %, respectively. While in Tanzania, the ciprofloxacin resistance was 39.3 % (Kumburu et al., 2017). Another similar study conducted by Uzoamaka et al., (2017) showed that the Proteus isolates had sensitivity percentages of 68.4 %, 78.9 % to ciprofloxacin and levofloxacin; respectively. In a previous study, Proteus isolates showed high antibacterial resistance to tetracycline (84.3 %) (Nemati, 2013). Currently, we observed that the chloramphenicol resistance was (87.2 %), while in another study carried out in Tanzania; it was 67.9 % (Kumburu et al., 2017). In this study, the resistance to nitrofurantoin was 73.5 %, in agreement with the previous study reported by Sonavane et al., (2008) that only 22.4 % of the Proteus isolates were susceptible to nitrofurantoin.
Using ureR-based PCR, the results of the present study revealed that 48.0 % of Proteus isolates were confirmed as P. mirabilis. Several similar studies used ureR-based molecular method for identification of P. mirabilis (Alatrash and Alyasseen, 2017; Kamil and Jarjes, 2021). Additionally, the existences of PMQR and aac(6')Ib genes were examined, which can encourage the development of higher levels of quinolone resistance in the Gram-negative bacteria. Currently, only 16 P. mirabilis isolates showed the existence of PMQR and aac(6')-Ib genes by 35 % and 10 %, respectively Among all the recovered isolates, qnrA and qnrB genes were the most common (15 %), followed by qnrD and aac(6′)Ib (10 %), and qnrS (5 %). However, in this study no Proteus isolate was recorded to harbor the qnrC gene. However in Canada, only 14 % of Proteus isolates had been reported to have qnrB gene, while 26 % were harboring aac(6')-Ib gene (Han et al., 2020). All Proteus isolates that had qnrB and aac(6')-Ib genes were resistant at least to either norfloxacin or ciprofloxacin. However, another study reported by Mirzaei et al., (2019) that 4.5 % and 0.9 % of Proteus isolates had aac(6ʹ)-Ib-cr and qnrA genes, respectively. Kotb et al., (2019) study showed that qnrB gene had been detected in 62.9 % of the resistant isolates, which represented the most Novel Research in Microbiology Journal, 2023 common gene among the quinolone resistant Proteus isolates. The present study showed that Proteus spp. had higher ciprofloxacin resistance than that of levofloxacin. The relationship between the presence of qnr genes, aac(6')-Ib gene and the resistance of ciprofloxacin and levofloxacin was analyzed. The obtained results recorded a significant correlation between levofloxacin resistance and the existence of aac(6')-Ib gene (p < 0.05).

Conclusion
Since Proteus spp. are so common in the environment; strict sanitation standards must be maintained within the healthcare facilities to decrease the rate of nosocomial infections. Moreover, the use of catheters should be avoided whenever possible, and if it is necessary; it must be applied under clean precautions and for restricted intermittent periods. Furthermore, the improper use of antibiotics is a contributing factor in the Proteus spp. growing resistance to the most studied antibiotics. Therefore, understanding the local bacterial etiology and susceptibility patterns is necessary to identify any potential changes that may happen. In P. mirabilis isolates, the quinolone resistance may have been caused by other mechanisms in addition to qnr genes and the aac(6')-Ib gene; hence, more future researches are required to examine these possible resistance mechanisms.