Screening of mcr-1 among Gram-Negative Bacteria from Different Clinical Samples from ICU Patients in Alexandria, Egypt: One-Year Study

Resistance to antibacterial agents represents a global threat. Infections caused by multi-drug resistant (MDR) and extensively drug-resistant (XDR) bacteria are growing, representing a major therapeutic challenge. MDR bacteria are defined as those resistant to at least one agent in at least three distinct categories of antibacterial agents. In comparison, XDR bacteria are defined as those resistant to at least one agent in all categories except for two or fewer categories of antibacterial agents (Magiorakos et al. 2012) Currently, colistin remains one of the last resort treatments against these infections (El-Sayed Ahmed et al. 2020; Wang et al. 2020b). Colistin was first used in the 1950s to treat infections caused by Gram-negative bacteria. Then, in the 1970s, it was replaced by other newly discovered antimicrobial agents, which did not have the toxic effects caused by colistin (El-Sayed Ahmed et al. 2020). However, due to the increased resistance to all available antibacterial agents, colistin has resurfaced, in the 1990s, as the last line of defense against infections caused by MDR and XDR Gram-negative bacteria, although its safety profile has not changed and its dosing problem. There is a lack of a universal synchronization of colistin dose units, which leads to suboptimal dosing potentially contributing to the resistance problem (Lim et al. 2010; Ahern and Schnoor 2012; Kaye et al. 2016). In addition to MDR and XDR infections, colistin can be combined with other antibacterial agents to manage pan-resistant Gram-negative bacteria (Sayyahfar et al. 2021).

Interestingly, some Gram-negative bacteria are intrinsically resistant to colistin as Morganella spp., Serratia spp., Providencia spp., and Proteus spp. Colistin resistance was thought to be only chromosomally mediated; however, in 2015, the plasmid-mediated resistance to colistin was reported for the first time. Since then, different mcr-1 alleles have been described; furthermore, ten different mcr family genes (mcr-1-family gene to mcr-10-family gene) have been reported worldwide (Hussein et al. 2021). The emergence of horizontally acquired resistance hampers colistin as a last resort against MDR Gram-negative bacteria (Liu et al. 2016; Sun et al. 2018; El-Sayed Ahmed et al. 2020; Xu et al. 2021). Interestingly, colistin susceptible Escherichia coli isolates harboring either mcr-1 or other mcr family genes have been reported (Wang et al. 2017; Chen et al. 2019). The aim of our present study was to investigate the presence of mcr-1 among different Gram-negative bacteria including Enterobacteriaceae (except intrinsically resistant to colistin) and Pseudomonas aeruginosa.

A total of 480 Gram-negative isolates were collected from June 2019 to June 2020. Most of the collected isolates were K. pneumoniae (62.71%), followed by E. coli (22.71 %), P. aeruginosa (11.46%), Enterobacter cloacae complex, and Enterobacter asburiae represented (2.92%) and (0.20 %) of the isolates, respectively. These isolates were collected from different types of clinical samples, including blood cultures (26%), urinary tract infections (17%), aspirates and swabs from surgical site infections (fluid and tissues) (31%), sterile body fluids, including (CSF, pleural fluid and perineal fluid) (2%), and respiratory tract infections (24%). Regarding the susceptibility testing results, resistance to the third and fourth generation cephalosporins was 82.29% and 78.33%, respectively, and they were distributed as follows: K. pneumoniae 92.36% and 87.38%, E. coli 78.9% and 77.06%, E. cloacae complex 64.29% and 71.43%, and P. aeruginosa 40% and 34.5%, respectively. Moreover, 49.79% of the isolates were resistant to MEM. They were distributed as follows: K. pneumoniae 68.44%, E. coli 6.42%, E. cloacae complex 21.43%, and P. aeruginosa 41.82%. Most of the Gram-negative organisms were resistant to CAZ and FEP, and almost half of the isolates were resistant to carbapenems and gentamicin. Among 480 isolates, 338 (70.4%) were MDR Gram-negative bacteria; of these seven isolates belong to E. cloacae complex, 25 were P. aeruginosa, 60 were E. coli, while 246 of them were K. pneumoniae isolates. The susceptibility testing results are shown in (Table I).

Among 480 Gram-negative bacterial isolates, only six (1.25%) harbored the mcr-1 gene, while 474 (98.75%) did not harbor that gene.

The mcr-1 gene was sequenced, then the BLASTN (https://blast.ncbi.nlm.nih.gov) tool for was used for confirmation. All isolates were found to harbor mcr-1.1. The obtained sequences were deposited in GenBank (accession numbers: MZ820395, MZ820396, MZ820398, MZ820399, MZ820400, MZ820401).

Then, to confirm the mcr-1 gene subtype, we amplified and sequenced the whole mcr-1 gene of all isolates; however, it could not be amplified in one K. pneumoniae isolate using this pair of primers. The obtained sequences were deposited in GenBank (accession number: MZ820389, MZ820390, MZ820391, MZ820393, MZ820394).

Remarkably, two isolates out of six isolates harboring mcr-1.1 were susceptible to colistin. Moreover, one of them could not be amplified using this pair of primers. However, it was successfully amplified and sequenced using the first primers set, and it was found to harbor mcr-1.1. Its sequence was submitted to Gen-Bank, as mentioned earlier.

Then, all K. pneumoniae isolates harboring mcr-1.1 and the isolate “PCMKP-01” were capsule typed using the wzi gene sequencing; three were K-58, one K-9, and one K-45. A phylogenetic tree was constructed, and it is shown in Fig. 1.

Fig. 1

The characteristics of the six isolates that harbor mcr-1.1 is shown in Table II. Three of the six isolates harbored bla_CTX-M. Five of the isolates were resistant to carbapenems. These five isolates harbored the bla_OXA-48 gene, bla_NDM-1 was present in three of them, while bla_VIM and bla_KPC were absent in these isolates. The distribution of the different beta-lactamase genes among the six isolates is shown in Table III.

Colistin remains our last resort against MDR and XDR Gram-negative bacteria. Moreover, it has been combined with other antibacterial agents to manage infections caused by pan-resistant Gram-negative bacteria (Sayyahfar et al. 2021). Different mechanisms of resistance contribute to the reduced susceptibility to colistin. However, the most worrying mechanism is plasmid-borne MCR-mediated resistance due to its ability to horizontally transfer between different species and the speed with which it is evolving. Several mcr family genes have been detected since they were first described in 2015 (Feng 2018; Xu et al. 2021).

Some studies reported the presence of mcr-1 among E. coli isolates susceptible to colistin, which would further complicate the situation in health care settings. Patients infected with organisms harboring the mcr-1 gene represent a potential threat for mcr-1 transmission because these organisms could escape being detected by conventional phenotypic methods. Furthermore, other resistance genes (beta lactamases and non-beta lactamases genes) could also be transmitted along with the mcr-1 gene. (Yuan et al. 2021) This study aimed to investigate the presence of mcr-1 among different Gram-negative bacteria, including Enterobacteriaceae (except intrinsically resistant to colistin) and P. aeruginosa.

Using SYBR Green-based real-time PCR, we screened for mcr-1 among the 480 Gram-negative bacterial isolates. Only 6 (1.25%) isolates harbored this gene, which was confirmed by sequencing of the amplicon obtained. Four of these six isolates were K. pneumoniae, one isolate was E. coli, and the remaining one was P. aeruginosa.

The first report of mcr-1 from a clinical isolate in Egypt was provided in 2016 by Elnahriry et al. (2016). Then, different studies reported the presence of mcr-1 among E. coli clinical isolates in Egypt, including Anan et al. (2021), who found only four (7.5%) E. coli harboring mcr-1 among colistin-resistant isolates, and El-Mokhtar et al. (2021) who reported that all their studied E. coli isolates resistant to colistin carried mcr-1. Moreover, another study reported the presence of mcr-1 in one E. coli and one K. pneumoniae isolate among their 450 enterobacterial isolates (Zafer et al. 2019).

Furthermore, Elmonir et al. (2021) reported that all colistin-resistant K. pneumoniae isolates harbored the mcr-1 gene. Abd El-Baky et al. (2020) reported the presence of mcr-1 among their P. aeruginosa isolates. Additionally, Yanat et al. (2016) reported the presence of mcr-1 in E. coli clinical isolate in Algeria, Alghoribi et al. (2019) also found mcr-1 gene in uropathogenic E. coli in Saudia Arabia, and mcr-1 was also reported from E. coli clinical isolate in Lebanon (Al-Bayssari et al. 2021). However, some studies could not detect mcr-1 among colistin-resistant bacterial isolates, Ramadan et al. (2020) did not find any of the eight mcr family genes (mcr-1 to mcr-8) among 65 Gram-negative bacterial isolates (Soliman et al. 2020b). In a study in Tunisia, they could not find any of the mcr genes, from mcr-1 to mcr-5 (Jaidane et al. 2018). Sadek et al. (2020b) reported only one E. coli isolate harboring mcr-1 among 128 colistin resistant E. coli strains isolated from meat and meat product samples in Egypt.

In this study, six isolates that harbored mcr-1.1 were tested for colistin susceptibility; four were found to be resistant to colistin, while two isolates (one K. pneumoniae and one P. aeruginosa) were found to be susceptible to colistin (≤ 0.5 µg/ml). To confirm the mcr-1 subtype, we attempted to amplify and sequence the whole mcr-1 gene among the six isolates, using another set of primers. Five isolates harbored mcr-1.1 and the remaining one (K. pneumoniae) could not be amplified using this set of primers. Interestingly, mcr-1.1 was reported in Egypt before; in one uropathogenic E. coli (UPEC) (Zakaria et al. 2021) and five E. coli isolated from chicken (Soliman et al. 2021). Moreover, mcr-1.1 was reported in E. coli isolate obtained from wound drainage (Eltai et al. 2020) from Qatar. Girardello et al. (2021) reported the presence of mcr-1.1 in E. coli clinical isolate in Sao Paulo, and mcr-1.1 was found in two K. pneumoniae clinical isolates by Rocha et al. (2020).

In this study, the K. pneumoniae isolate, whose mcr-1 gene could not be amplified using the second set of primers (used to amplify the whole mcr-1 gene), was susceptible to colistin. Another colistin susceptible isolate was P. aeruginosa, which was found to harbor the mcr-1.1 gene.

Previously, some studies have reported Gram-negative isolates susceptible to colistin and harboring the mcr-1 gene. Wang et al. (2017) described two colistin-susceptible E. coli isolates possessing the mcr-1 gene. Terveer et al. (2017) also reported a colistin-susceptible E. coli harboring mcr-1, which was not functioning. Using WGS, they found that the gene was rendered not functional by a transposon (IS10R) insertion (Terveer et al. 2017). Zhou et al. (2018) also described a fluoroquinolone-resistant but colistin-susceptible E. coli carrying mcr-1, which was also non-functional by inserting a 1.7-Kb IS1294b element. Chen et al. (2019) also reported a colistin-susceptible E. coli harboring mcr-1, which was non-functional because of the insertion of another gene. Jiang et al. (2020) described a colistin susceptible E. coli harboring the mcr-1 gene, which had mutations in the mcr-1 promotor sequence.

However, to the best of our knowledge, this is the first study reporting the presence of mcr-1 in colistin susceptible K. pneumoniae and P. aeruginosa in Egypt. The whole mcr-1 gene in the susceptible K. pneumoniae could not be amplified using the second set of primers and it was found to be susceptible to colistin; it may be due to an insertion sequence that rendered the gene non-functional. However, in our case, WGS was not feasible due to its high cost and the fact that this study was not funded. Besides the mcr-1 gene, other mcr family genes have also been described in colistin-susceptible bacteria. Ragupathi et al. (2020) reported the presence of the mcr-3.30 gene in colistin-susceptible Aeromonas veronii. The mcr-3.30 gene was disrupted due to the insertion of ISAs18 (Ragupathi et al. 2020). The mcr-9 is another mcr family gene that was reported in other studies, including a study conducted by Soliman et al. (2020a), who reported the presence of this gene in colistin susceptible Enterobacter hormaechei clinical isolate from Egypt. Other colistin susceptible E. hormaechei harboring mcr-9 were also reported in Egypt in pets with respiratory diseases (Khalifa et al. 2020a) and from food of animal origin (Sadek et al. 2020a). Carroll et al. (2019) reported the presence of mcr-9 in the MDR Salmonella enterica subsp. enterica serotype Typhimurium isolate, which was susceptible to colistin. Also, another study conducted by Kananizadeh et al. (2020) reported the presence of mcr-9 in E. cloacae complex in Japan. Khalifa et al. (2020b) found mcr-9 in colistin-susceptible foodborne K. pneumoniae. The mcr-9 subtype was also reported by Marchetti et al. (2021) who reported mcr-9.2 in a colistin susceptible E. cloacae. The mcr-10 was described by Wang et al. (2020a) in Enterobacter roggenkampii that was susceptible to colistin.

In this study, four K. pneumoniae isolates that harbored the mcr-1.1 gene and “PCMKP-01” were capsule typed using the wzi gene sequence analysis. The most common K-type found among these isolates was K-58. It was found in three isolates. K-58 has not been associated with virulence (Turton et al. 2010). These three isolates were MDR strains of K. pneumoniae resistant to colistin that harbor the mcr-1.1 gene. The two remaining K-types were K-9 and K-45. K-45 type K. pneumoniae was considered the positive control isolate “PCMKP-01”. It was the MDR colistin-resistant isolate, while K-45 type isolate was colistin susceptible K. pneumoniae that harbored mcr-1.1.

Three of the isolates co-harbored mcr-1.1 and bla_CTX-M. Sadek et al. (2021) reported that nine of the isolates co-harbored the mcr-1 gene and ESBL genes. In the present study, five isolates harbored mcr-1.1 and were resistant to carbapenems. These five isolates had bla_OXA-48, and three of them had bla_NDM-1 (one P. aeuruginsa isolate and two K. pneumoniae isolates). bla_VIM and bla_KPC were not found among these five isolates. Singh et al. (2021) reported that all the isolates harboring mcr-1 in their study co-harbored bla_OXA-48. Han et al. (2020) reported the identification of the XDR E. coli clinical isolate co-harboring mcr-1 and bla_NDM-1. Al-Bayssari et al. (2021) reported the co-existence of bla_NDM-4 and mcr-1 among E. coli clinical isolates.

To our knowledge, it is the first time to report colistin susceptible P. aeruginosa and K. pneumoniae harboring the mcr-1.1 gene in Egypt. The mcr-1.1 gene was fully sequenced in P. aeruginosa, while in K. pneumoniae it could not be fully sequenced, which indicated some abnormality in this gene. The most frequently found K-type was K-58. Five of the isolates were resistant to carbapenems and co-harbored bla_OXA-48 and mcr-1, and three of them co-harbored mcr-1.1, bla_OXA-48, and bla_NDM-1. Co-existence of these genes together is a clear therapeutic challenge. Further studies are still needed to investigate the presence of the plasmid-borne mcr genes among colistin susceptible isolates to shed more light on its significance as a potential threat.