Occurrence and Characteristics of Mobile Colistin Resistance (mcr) Gene-Containing Isolates from the Environment: A Review.

The emergence and spread of mobile colistin (COL) resistance (mcr) genes jeopardize the efficacy of COL, a last resort antibiotic for treating deadly infections. COL has been used in livestock for decades globally. Bacteria have mobilized mcr genes (mcr-1 to mcr-9). Mcr-gene-containing bacteria (MGCB) have disseminated by horizontal/lateral transfer into diverse ecosystems, including aquatic, soil, botanical, wildlife, animal environment, and public places. The mcr-1, mcr-2, mcr-3, mcr-5, mcr-7, and mcr-8 have been detected in isolates from and/or directly in environmental samples. These genes are harboured by Escherichia coli, Enterobacter, Klebsiella, Proteus, Salmonella, Citrobacter, Pseudomonas, Acinetobacter, Kluyvera, Aeromonas, Providencia, and Raulotella isolates. Different conjugative and non-conjugative plasmids form the backbones for mcr in these isolates, but mcr have also been integrated into the chromosome of some strains. Insertion sequences (IS) (especially ISApl1) located upstream or downstream of mcr, class 1-3 integrons, and transposons are other drivers of mcr in the environment. Genes encoding multi-/extensive-drug resistance and virulence are often co-located with mcr on plasmids in environmental isolates. Transmission of mcr to/among environmental strains is clonally unrestricted. Contact with the mcr-containing reservoirs, consumption of contaminated animal-/plant-based foods or water, international animal-/plant-based food trades and travel, are routes for transmission of MGCB.

1. Introduction

The emergence and spread of plasmid-mediated movable colistin (COL) resistance (mcr) genes jeopardize the efficacy of COL considered a last resort drugs for treating deadly infections caused by multi- and extensively drug-resistant Gram-negative bacilli (GNB) [1,2]. All the while, COL resistance was thought to be due to chromosomal mutations (such as widely dispersed two-component system prmAB and PhoPQ, and mgBr in Klebsiella), which are not transferable [3]. However, in 2015, mcr-1 gene was detected in Escherichia coli isolates from humans, food animals and the environment in China [4]. Indeed, this heralded the emergence of pandrug-resistant bacteria (superbugs) [5]. Since, after the discovery of mcr-1 gene, eight other mcr gene types (mcr-2 to mcr-9) with their very many variants have been detected in isolates from humans, animals, and environment in six of the seven continents (except for Antarctica) [6,7,8,9]. The mcr genes encode MCR, which are cytoplasmic transmembrane proteins found in GNB [10]. These proteins are phosphoethanolamine (pEtN) transferases conferring resistance to COL by attaching a pEtN moiety to the lipid A of lipopolysaccharide in bacterial cell membrane thereby abolishing the negative charges to which cationic colistin/polymyxins have affinity [3,10,11,12]. Because COL was barely used in human medicine from 1970 to 1994 [13] but has been habitually used for livestock production for more than 50 years, it is clear that COL selection pressure evolved due to veterinary use [14,15]. There is an avalanche of publications in the literature showing that worldwide, humans and animals (especially livestock) are colonized by commensal and potentially pathogenic colistin-resistant organisms (COLROS)/mcr gene-containing bacteria (MGCB) [16,17,18]. The rapid global spread of MGCB has left the world with a very limited therapeutic option such as tigecycline [19]; but unfortunately, plasmid-mediated mobile genes encoding high tigecycline level resistance was recently detected [20]. By this development, the global public health is gravely threatened, and the spread of these superbugs continues. Worryingly, the O’Neill’s projection that, if the cause of antimicrobial resistance (AMR) is not tackled now, 10 million human lives could be lost worldwide annually by 2050 due to antimicrobial-resistant infections [21] might become a reality.

It is largely recognized that the environment plays a huge role in disseminating clinically relevant antimicrobial resistance genes (ARGs) [22,23,24,25]. Accordingly, the United Nations Environment Programme (UNEP) recently identified environmental AMR as the top six emerging issues of concern [25,26]. In One Health model, surveillance of ARGs in humans, animals, soil, plant, aquatic and aquaculture environment is crucial in determining the extent of dissemination of a targeted ARG [2,27]. Potentially pathogenic COLROS emanating from any of these ecosystems can move with relative ease to another without barriers [2,25]. Experts agree that the control of ARGs’ spread cannot be achieved without tackling this problem from the environment [2,22,23,24,25,26]. It is established that antimicrobial selection pressure in the environment results due to discharges from humans and animals (containing un-metabolized antimicrobials in faeces/urine into sewage), hospitals, industries, and run-offs from refuse dumps (often contains anthropogenic wastes) and farmlands especially where biocides are used [11,12,23,24,25,28]. Putative sources of MGCB in the environment include human and animal excretions, aquaculture water, sewages/wastewaters from hospitals and laboratories [11,12,23,28]. The hospital environment often contains numerous antibiotics which perhaps stimulate COL resistance [29]. Thus hospital effluents constitute a putative source of COLROS in water and soil ecosystem [30]. The most worrisome situation about COL resistance is the presence of other antimicrobial agents (including disinfectants and metals) in an environment, could stimulate COL selection pressure, and mcr gene is often co-located with multiple AMR (most troubling being carbapenemases, extended-spectrum β-lactamases [ESBL], and plasmid-mediated quinolone resistance [PMQR] encoding genes) and virulence genes on highly promiscuous plasmids [31]. There have been extensive discourse on acquired resistance [32,33,34,35] and virulence genes [36], as well as the drivers (plasmids, integrons, transposons, insertion sequences, etc.) facilitating the horizontal/lateral spread (by conjugation, transformation or transduction) of these genes among bacterial population [37,38,39,40,41,42]. Of the drivers of resistance, plasmids are the major vehicle involved in the transmission of resistance/virulence genes [39,40,41,42]. Plasmids are small circular/linear double-stranded DNA unit which is chromosome independent and capable of self-replication, and they belong to many families (with many incompatibilities and replicons) with broad or narrow host ranges [37,41,43]. Resistance and virulence plasmids carry genes encoding antimicrobial resistance and virulence factors, respectively, thus enhancing bacterial survival fitness [38,43].

The use of antimicrobials as biocides on plants potentially results in colonization of plants and soil by antimicrobial-resistant organisms [43]. Use of animal manure and insufficiently-treated/untreated sewage sludge as fertilizer in farmlands and aquaculture are other potential routes of introduction of COLROS into the soil and aquaculture [43,44]. Plants (fruits, vegetables, and grasses) could get contaminated with COLROS emanating from soil [45,46]. Unfortunately, these fruits and crops are often consumed raw or undercooked by humans and animals thus posing a risk to public health [46]. Run-offs from farmlands primarily where insufficiently-treated wastewater, sewage sludge or animal manure was used as fertilizer, and aquaculture water discharged or used for irrigation, may introduce COLROS/mcr genes and un-metabolized antimicrobials into aquatic systems (surface/groundwater bodies and aquaculture) in which bacteriophages (by transduction) and integrons could facilitate the spread of mcr gene [11,23,24,29,30]. In fact, evidence has shown that mcr and non-mobile COL resistance (nmcr) genes were mobilized from perhaps aquatic environment [2,27,47,48,49]. Feral aquatic animals (such as fish, birds, mammals, reptiles, amphibians, molluscs, etc) could get colonized by COLROS, thereby contaminating the food chain [23,50].

The role of wildlife (especially those that migrate for long distances) in the dissemination of clinically-relevant ARGs and the public health impact is well recognized [50,51,52,53,54]. Although antimicrobials are not used in wildlife, contact with human and domesticated animal excretions (especially for terrestrial animals) in the environment, could result in infection of wildlife by COLROS [52,53]. Consumers of undercooked wildlife meat products are at huge risk of acquisition of COLROS [50,55]. Carnivores could also get colonized by MGCB through feeding on wild animals, and then further disseminate mcr genes in the environment. Moreover, wild animals could disseminate mcr genes to places (such as playing grounds, farmlands, parks, markets, open slaughterhouses) frequented by people [50,52,55]. Syanthropic flies that feed on garbage, carrion, human and animal wastes could get colonized by COLROS. Often the colonization of fly occurs on their body surfaces or intestines where the horizontal exchange of ARGs takes place. Thereby serving as reservoirs or vectors of transmission of these organisms to other ecological niches when they perch on animals, plants, foods and contact surfaces in households and public places, and/or when they serve as feed for fish and birds [56,57,58]. Wild birds (especially migratory birds) could deposit MGCB (through defecation, grooming or drinking) into surface waters and onto fruits/vegetables [45,59]. In some parts of the world (especially in rural areas in developing countries), surface water is used for bathing, laundering, recreation, fishing, and as drinking water for human and animals [60,61,62]. Such activities could facilitate the exchange of COLROS from the environment to humans and animals [60,62]. Integrated farms facilitate the exchange of COLROS between food animals and aquaculture since excretions from livestock that may contain COLROS/MGCB and un-metabolized COL, serve as food to fish which receives little or no supplementation [23,24,28,48]. Farmworkers’ paraphernalia (such as gloves, boots, wheel-barrow, vehicles, etc.) are potential vehicles of transmission of COLROS from farm-yards to other ecological niches [63].

Presence of mcr genes in the natural environment could complicate the transmission dynamics of COLROS, thus impacting the epidemiology and increasing the rate of evolution of MGCB [64,65]. Understanding the occurrence and magnitude of COL resistance in the environment creates the needed impetus to tackle the problem [66]. Information on the occurrence, phenotypic and genotypic characteristics of MGCB isolated from the environment is crucial for an understanding of the epidemiology, genetic environment and mechanism of acquisition of mcr genes by environmental isolates. Such information would be necessary in designing and prioritizing surveillance programs that may generate essential data for performing risk assessment, implementation of effective antimicrobial stewardship plans, developing effective strategies for control of COLROS, and reducing the risk to public health. In this review, the objective is to report the findings of studies on plasmid-mediated COL resistance among isolates from the environment.

2. Literature Search Strategy and Data Extraction

Studies that assessed the presence of mcr genes in isolates from environmental sources (water, sewages, aquaculture, aquatic-based foods, wildlife including flies and reptiles) worldwide were included in this review. Peer-reviewed works of literature were identified by searching databases such as Pubmed, MEDLINE, EMBASE, Scopus, and Web of knowledge. Google search engine was also used to retrieve grey literature. The following search terms and/or text words were used for the search: “mobile colistin resistance gene”, “plasmid-mediated colistin resistance gene”, “plasmid-borne COL resistance”, “mobile COL resistance”, “movable COL resistance genes”, “enterobacteria”, “Gram-negative bacilli”, “bacterial isolates”, “environment”, “wildlife”, “wild animals”, “migratory animals”, soil”, “water”, sewages”, “fish”, “aquatic-based foods”, and names of specific countries in the world. References of identified publications were reviewed for additional pertinent articles. Information extracted from included studies includes the first author’s surname, year of publication, and country where the study was conducted. Other information extracted were isolation/study period, type of mcr gene assayed, sample processed, number of samples, number of isolates subjected to mcr assay. We also extracted the number and type of organism positive for mcr gene, mcr gene variant detected; sequence type, virulence genes, plasmid type, associated insertion sequence, and additional resistance factors identified in test isolate (Table 1, Table 2, Table 3, Table 4, Table 5, Table 6, Table 7 and Table 8).

Table 1

Studies reporting plasmid-mediated colistin resistance in isolates from environmental contact surfaces.

Country	Source of Isolate	Date of Isolation (mcr Gene Assayed)	Number of Isolates Tested for mcr	Identified Gene/Variant (Number of Organism)	Sequence Type and/or Phylogroup (Virulence Genes)	Plasmid (Associated Insertion Sequence)	Additional Resistance Traits	Key Points/Conclusion	Reference
Italy	Hospital surfaces	2016–2017 (mcr-1)	300	mcr-1 (4 Acinetobacter iwoffi and E. coli, Citrobacter freundii (1), Pseudomonas aeruginosa (1) and Proteus putida (1), Enterobacter (E.) cloacae (3) and E. agglomerans (3), 6 K. pneumoniae and 2 K. oxytoca)	-	-	-	- mcr-1-carrying GNB contaminate surfaces in hospitals representing a reservoir of hard-to-treat nosocomial pathogens	[29]
China	Contact surfaces at public transportation routes	2016–2017 (mcr-1)	mcr was detected in samples	mcr-1 (23 E. coli and 3 K. pneumoniae)	E. coli (ST2253, ST58, ST48, ST1249, ST7122, ST744, ST189, ST101, and ST10 complex); K. pneumoniae (ST37)	IncI2, IncX4 and IncH2.	-	- First report of mcr-1-carrying organisms contaminating contact surfaces at public transportation routes in China	[67]

mcr: mobile colistin resistance gene; -: no data; Additional resistance traits: Additional resistance traits: resistance factors identified in one mcr-positive isolate or pooled factors in more than one mcr-positive isolate; Sequence type: comprise all sequence types of mcr gene-positive isolates; Plasmid: plasmid types identified in a/pooled mcr gene-positive isolate; Inc.: incompatibility; IS: insertion sequence.

Table 2

Studies reporting plasmid-mediated colistin resistance in sewages/wastewaters.

Country	Source of Isolate	Date of Isolation (mcr Gene Assayed)	Number of Isolates Tested for mcr	Identified Gene/Variant (Number of Organism)	Sequence Type and/or Phylogroup (Virulence Genes)	Plasmid (Associated Insertion Sequence)	Additional Resistance Traits	Key Points/Conclusion	Reference
China	Hospital, wastewater	2013–2016	-	mcr-1 (8 E. coli)	-	-	-	- mcr-1originated from agriculture and spread to human and environmental ecosystems by transposon, ISApl1 and plasmids	[2]
China	Hospital sewage	2017(mcr-1)	1	mcr-1.1 and mcr-3.5 (1 E. coli)	ST410	IncHI2 and IncN	aac(3)-IVa, aac(3)-IId, aac(6′)-Ib-cr, aadA22, aadA16, aph(3′)-Ia, aph(3′)-Ib, aph(6′)-Id, aph(4)-Ia, rmtB, strA, blaCTX-M-65, blaCMY-2, blaNDM-5, blaTEM-1B, fosA, lnu(F), mphA, floR, arr-3, aac(6′)-Ib-cr, oqxA, oqxB, sul1, sul2, tet(A), tet(M), dfrA14 and dfrA27.	- Environmental strains can transfer mcr-1, carbapenem and high-level aminoglycoside resistance simultaneously-Carbapenem and colistin-resistant strains in the environment are diverse in clonal backgrounds	[9]
Spain	Sewage water	2013 (mcr-1)	90	mcr-1 (29 E. coli,1 K. pneumoniae)	E. coli (ST1196/ST632); K. pneumoniae (ST526)	Inc2	blaCTX-M-55 and blaTEM-1	- mcr-1 can act as silent gene thereby evading phenotypic detection and favouring its dissemination-Sewages are huge reservoirs of mcr-1 from where it is disseminated to other ecosystems	[81]
Thailand	Wastewaters	2015 (mcr-1 and mcr-2)	65	mcr-1 (2 E. coli)	ST5951and ST6624	IncHI2 and IncX4	bla CTX-M-14	- Diverse carbapenem- and colistin-resistant E. coli clones are shared between humans and their immediate environment	[83]
Bangladesh	Urban sludge	-	48	mcr-1 (1 E. coli)	-	-	-	- First report of mcr-1-positive isolate in Bangladeshi environment	[84]
China	Sewage	2015 (mcr-1)	9	mcr-1 (9 E. coli and 2 Kluyvera ascobarta), novel mcr-1.7 and mcr-1.4 (2. E. coli)	ST10, ST34, ST48, ST1196 and novel ST7086 and ST7087 (dnaG, int, parA, nikB, traB, ydfA, ydgA, yejO and yfjP)	IncX4, IncI2, IncHI2, IncN and IncP (ISApI1); for chromosome: IS911∆,IS30∆, Ec23∆ and IS1294)	blaCTX-M-14, blaTEM-1b, blaCTX-M-55, blaCMY-2, blaCTX-M-125, strA, strB, aac(3)-IId, aadA2, aadA1, aadA22, aph(30)-Ia, aac(3)-Iva), ph(4)-Ia, oqxB, oqxA, qnRS1, fosA, lnu(F), mph(A), met(B), cmlA1 and floR genes, sul1, sul2, sul3, tet(M), tet(B), tet(A), dfrA14 and dfrA12	– First report of mcr-1.4 and mcr-1.7- mcr-1 gene variants are carried on chromosome and plasmids in diverse unrelated species	[87]
China	Hospital sewage	2015 (mcr-1)	1	mcr-1 (K. pneumoniae)	ST313 (tra, trb and trfA)	IncP-1 (ISApl1 and IS26)	aac(3)-Iva, aadA2, aph(3′)-Ia, aph(4)-I, blaTEM-135 and new blaSHV195 variant, floR, fosA, oqxA, oqxB, sul2, dfrA12, tet(A) and ORFble	- mcr-1-poitive K. pneumoniae ST313 can spread through sewage- First report of ORFble gene with undetermined function in environmental isolate	[88]
China	Hospital sewage	2016 (mcr-1 and mcr-2)	25	mcr-1 (9 E. coli)	ST7122, ST10, ST410, ST2016, ST349, ST6756 and ST101	IncX4	blaCTX-M-15, blaCTX-M-3-like, blaCTX-M-14 and blaTEM-1	- First report of mcr-1-positive E. coli and blaKPC-carrying Enterobacter cloacae in sewage	[89]
China	Hospital sewage	2015 (mcr-1)	1	mcr-1 (1 Kluyvera ascobarta)	traC	IncI2 (ISApI1)	blaCTX-M-185 and fosA	- First report of mcr-1-positive Klyuvera ascorbata in environmental sample	[90]
Lebanon	Sewer water	mcr-1 to mcr-8	8	mcr-1 (8 Proteus mirabilis)	-	-	blaTEM and IntI1	- First report of mcr-1 in Proteus mirabilis in sewers highlighting the need for improved sanitation in war refugee camps in Lebanon	[95]

mcr: mobile colistin resistance gene; -: no data; Additional resistance traits: resistance factors identified in one mcr-positive isolate or pooled factors in more than one mcr-positive isolate; Sequence type: all sequence types of mcr gene-positive isolates; Plasmid: plasmid types identified in one or pooled mcr gene-positive isolates; Inc.: incompatibility; ∆: truncated; IS: insertion sequence.

Table 3

Studies reporting plasmid-mediated colistin resistance in isolates from fresh and seawaters.

Country	Source of Isolate	Date of Isolation (mcr Gene Assayed)	Number of Isolates Tested for mcr	Identified Gene/Variant (Number of Organism)	Sequence Type and/or Phylogroup (Virulence Genes)	Plasmid (Associated Insertion Sequence)	Additional Resistance Traits	Key Points/Conclusion	Reference
China	Well water	2015 (mcr-1)	10	mcr-1 (2 E. coli)	ST48 and ST10	-	blaCTX-M-14 and blaCTX-M-65	-mcr-1-positive ESBL-producing E. coli of ST10 complex and carbapenem-resistant organisms have disseminated into ground water in China highlighting that these waters are sources of resistance to last resort antibiotics	[62]
Brazil	Coastal water at public beaches	2016 (mcr-1)	3	mcr-1 (3 E. coli)	ST1638, ST46 and ST10, and B1 (iss, gad and mchF)	IncI1, ColRNAI, IncX4, IncFIB, IncQ1, IncX4 IncHI2 and IncN.	blaCTX-8, blaCTX-M-1, qnrB19, aadA2, strA, strB, sul2, blaTEM-1B, catA1, aadA1, sul1, tetA, dfrA1, dfrA8 and tetB	-Anthropogenic activities results in dissemination of mcr-1-carrying E. coli to coastal waters in Brazil presenting health danger to users of public beaches	[103]
Italy	Well/Stream	2014–2015 (mcr-1)	264	mcr-1.2 (1 E. coli)	ST10 (pap2)	IncX4 and IncX3	blaCTX−M−1 and blaSHV−1	-First report on mcr-1.2 in ST10 E. coli from ground water highlighting the need for improved water treatment protocols and surveillance of resistance genes-Environmental water is a reservoir of antimicrobial resistance genes	[100]
Norway	Water from public beach	2010 (mcr-1)	82	mcr-1 (2 E. coli)	ST10 (A and B1)	IncFII, IncI1, IncFIB, Col156, IncX4 and ColRNAI	aadA5, blaCTX-M-1, blaTEM-1B, dfrA17, strA, strB, sul2 and tet(B)	-mcr- 1 have spread to areas without selection pressure-mcr-1-carrying Enterobacteriaceae in Norwegian surface waters emanate from anthropogenic sources	[101]
Malaysia	Water	2014	1	mcr-1(1 E. coli)	nikB	IncI	-	-mcr-1 have been in existence more than expected but ISApl1facilitated its spread among bacteria	[105]
Malysia	Pond water	2013	1	mcr-1 (E. coli)	ST410	IncFII, Incl2, ColRNAI, and IncFIB	aph(3)-Ia, aadA1, aadA2, strA, strB, blaTEM-1B, fosA, catA2, cmlA1, sul3, tet(A), tet(34) and drfA14	-mcr-1-positive E. coli strains are non-clonal and they often carry multiple resistance genes	[106]
Lebanon	Irrigation water	2018 (mcr-1 to mcr-8)	22	mcr-1 (22 E. coli)	-	-	Intl1, blaTEM, blaCTX-M, blaSHV, blaNDM-1 and blaOXA-48	- mcr-1 can persist in water matrix-Antimicrobial-resistant organisms in Lebanese irrigation can spread into the Mediterranean Basin	[108]
Lebanon	Sewer/domestic waters	mcr-1	36	mcr-1 (36 E. coli)	-	IncI2 and IncX4	Intl1, blaTEM, blaCTX-M, blaSHV, blaOXA-48 and blaKPC	-Unsanitary conditions in Lebanon and Syrian war refugee camps facilitate the spread of mcr-1	[109]
China	River, lake and fountain water	2017 (mcr-1 to mcr-7)	5	mcr-3.14 (2 Aeromonas caviae); mcr-3-like4 (1 A. veronii, 1 A. caviae and 2 A. bivalvium); mcr-3.18 (1 A. caviae)	hylA and dgkA in A. caviae	-	Except in A. bivalvium:blaMOX-7, aadA1, tet(A), sul1, blaMOX-4 and tet(E)	-First report of mcr-3.13 to mcr-3.18-Aeromonas which is common in aquatic environment represent important reservoir of mcr-3group gene	[110]
China	River/lake/fountain water	(mcr-1, and mcr-2)	mcr-1 (detected in samples)	mcr-1 (16 E. coli, 1 Enterobacter cloacae, Citrobacter freundii (2), Citrobacter braakii (2) and Kebsiella oxytoca (2)	ST10, ST43, ST101, ST206, ST1638 and ST181	-	qnrS, oqxA, oqxB, qnrB, blaCTX-M-15, blaCTX-M-55 and blaTEM-1	-Enrichment broth culture increases the rate of detection of mcr-1-carrying organism in water-Responsible use of polymyxins in agriculture and clinical settings curbs the spread of mcr-1into the environment	[113]
China	River	2017 (mcr-1 to mcr-5)	1500	mcr-1 (17 E. coli and 1 Enterobacter cloacae) and mcr-3 (2 Aeromonas veronii and 4 A. hydrophila)	-	-	blaSHV, blaTEM, blaCTX−M−9, sul1, sul2, tetM, tetA, qnrB, qnrS, oqxAB, aac(6′)-Ib-cr, rmtA, rmtB, floR and fosA.	-Sewage is a source of antibiotic resistance genes in urban rivers	[114]
Algeria	Sea water	2016 (mcr-1)	246	mcr-1.1 (1 E. coli), mcr-1.5 (1 E. coli)	ST23 and ST115	IncI2 and IncHI2A	aac(3)-IId, aadA1, aadA2, aph(3′)-Ia, aph(3)-Ib, aph(6)-Id, blaTEM-1B, mph(A), cml, sul1, sul3, tet(A), dfrA, and dfrA14.	-First report of mcr-1-carrying E. coli ST23 and ST115 in Algerian coast highlighting the global spread of the gene and the need for improved waste water/sewage treatment protocols	[102,115]

mcr: mobile colistin resistance gene; -: no data; Additional resistance traits: resistance factors identified in one or pooled mcr-positive isolate; Sequence type: all sequence types of mcr-positive isolates; Plasmid: plasmid types in one orpooled mcr-positive isolate; Inc.: incompatibility; IS: insertion sequence.

Table 4

Studies reporting plasmid-mediated colistin resistance in aquaculture environment and aquatic-based foods.

Country	Source of Isolate	Date of Isolation (mcr Gene Assayed)	Number of Isolates Tested for mcr	Identified Gene/Variant (Number of Organism)	Sequence Type and/or Phylogroup (Virulence Genes)	Plasmid (Associated Insertion Sequence)	Additional Resistance Traits	Key Points/Conclusion	Reference
China	Integrated fish-duck farm	(mcr-1 to mcr-8)	59	mcr-1.1 (7 E. coli) and mcr-3 (1 Aeromonas)	ST48,ST93, ST156, ST162 and ST648 (usp, cdtA, set1A and hylEclyA)	IncHI2, IncI2, IncX4, IncP and Incp0111 (ISApI1)	bla TEM-1	-Aquaculture is a route for transfer of mcr-1between humans and animals-First report of mcr-8 in aquaculture	[28]
Germany	Fishes	2005–2011 (mcr-1 to mcr-3)	479	mcr-3.6 (1 Aeromonas allosaccharophila), mcr-3.8 (1 A. jandei and 1 A. hydrophila) and mcr-3.9 (1 A. hydrophila)	dgkA	ISEcp1, ∆ISAeme4, ∆ISAeme5, IS50R, IS3, IS5 and IS30)	blaCEPH-A3, blaFOX-5, blaOXA-12, tet(E), ampH, cphA2, tet(D) and tet(M)	-First report of mcr-3.6 to mcr-3.9 highlighting that mcr-3 group gene have been present for the past 12 years in Europe	[118]
Norway	Imported sea food	2015 (mcr-1)	1	mcr-1 (1 E. coli)	ST48	IncHI2, IncN and IncX3	gyrA mutation, blaTEM-1B, dfrA12, dfrA15, mph(A), sul1, sul3, tet(A), cmlA1, strA, strB, aadA2, aadA1, qnrS1 and qepA.	-Aquatic-based food trade is a route for transfer of mcr-1	[117]
Spain	Mussels	2012–2016	19	mcr-1 (1 Salmonella enterica Serovar Risen)	ST469 (invA)	-	aac(6′)-Iaa, aadA1, aadA2, blaTEM-1B, cmlA1, sul1, sul3, tet(A), and dfrA1	-First report of mcr-1-positive Salmonella in ready-to-eat mussels highlighting the risk of transmission of mcr-1 from Galicia to European markets	[120]
China	Grass carp fish	2016–2017 (mcr-1 to mcr-5)	190	mcr-1 (7 E. coli)	ST ST48, ST7508, ST2040 ST7013 and ST156	IncI2, IncP and IncX4	aadA2, aadA1, aph(3′)-Ic, aac(3)-IId, strA, strB, blaTEM-1A, blaCTX-M-14, blaCTX-M-55, blaTEM-176, blaOXA-10, blaCARB-2, qnrS1, oqxB, oqxA, cmlA1, floR, arr-2, sul2, sul3, tet(A), tet(M), dfrA14, dfrA16, fosA3, erm(B), mph(A), tet(A), and dfrA17	-IncP plasmid may facilitate the dissemination of mcr-1 gene across various hosts just like Inc2 and IncX4 plasmids	[123]

mcr: mobile colistin resistance gene; -: no data; Additional resistance traits: resistance factors identified in one or pooled mcr-positive isolate; Sequence type: all sequence types of mcr gene-positive isolates; Plasmid: plasmid types identified in one or pooled mcr-positive isolates; Inc.: incompatibility; ∆: truncated; IS: insertion sequence.

Table 5

Studies reporting plasmid-mediated colistin resistance in isolates from soil/manure ecosystem.

Country	Source of Isolate	Date of Isolation (mcr Gene Assayed)	Number of Isolates Tested for mcr	Identified Gene/Variant (Number of Organism)	Sequence Type and/or Phylogroup (Virulence Genes)	Plasmid (Associated Insertion Sequence)	Additional Resistance Traits	Key Points/Conclusion	Reference
Estonia	Pig slurry	2011–2014 (mcr-1)	141	mcr-1 (3 E. coli)	taxA, taxB, taxC, trbM, pap2, topC, hicA, hicB, parA, pilX operon	IncX4 (IS26)	blaCTX-M-1, blaTEM-1 and blaAmpC-1	- Highly mobile plasmids carry mcr-1 in genetically diverse strains even from the same environment	[126]
China	Farming soil	2016 (mcr-1 to mcr-5)	53	mcr-1 (6 E. coli)	ST2060, ST3014, ST6756 and ST1560 (pap2)	InFIB, IncX1, IncHI1, IncFII, IncFIA, pO111 and IncHI2 (ISApl1)	aph(4)-Ia, aacA4, aac(3)-IVa, aph(3′)-Ia, aac(6′)-Ib-cr, aadA2, blaTEM-1B, blaCTX-M-55, fosA, mph(A), cmlA1, floR, catA1, catB3, AAR-3, sul1, sul2, sul3, tet(B), tet(M), dfrA1, dfrA12, blaCTX-M-27, QnrS1, oqXB, oqxA, mef(B), aac(3)-IId, aacA1, aacA2, aph(3′)-IIa, aadA5, dfrA17, aph(4)-Ia, blaCMY-2 and blaTEM-18.	- First report of mcr-1-carrying ESBL-producing E. coli in soil highlighting the influence of animal manure in transmission of colistin resistance	[127]

mcr: mobile colistin resistance gene; -: no data; Additional resistance traits: resistance factors identified in one mcr-positive isolate or pooled factors in more than one mcr-positive isolate; Sequence type: all sequence types of mcr-positive isolates; Plasmid: plasmid types identified in one or pooled mcr-positive isolates; Inc.: incompatibility; IS: insertion sequence.

Table 6

Studies reporting plasmid-mediated colistin resistance in botanical ecosystems.

Country	Sample (Number) Source of Isolate	Date of Isolation (mcr Gene Assayed)	Number of Isolates Tested for mcr	Identified Gene/Variant (Number of Organism)	Sequence type and/or Phylogroup (Virulence Genes)	Plasmid (Associated Insertion Sequence)	Additional Resistance Traits	Key Points/Conclusion	Reference
Portugal	Vegetables	2013–2014 (mcr-1)	138	mcr-1 (1 E. coli)	ST1716 (gad)	IncHI2, IncI1, IncI2, IncQ, IncP and IncY	Class 2 integrons, aadA1, aph(4)-Ia, blaTEM-1, estX-12, floR, sat2, strA, strB, sul2and tetA	-Organic and conventional vegetables are reservoirs of resistance genes, integrons and transposons-First report of class 3 integrons in isolate from vegetable	[133]
China	Vegetables	2015- 2016 (mcr-1 and mcr-2)	244	mcr-1 (7 E. coli and 2 Raulotella ornitholytica)	ST48, ST2505, ST156, ST795 and ST69	Incx4, IncI2 and IncHI2	floR, blaCTX-M-14, blaCTX-M-55, fosA and oqxAB	-Vegetables are source of mcr-1-carrying Enterobacteriaceae-Proper animal manure disposal and improved irrigation water warranted	[134]
China	Fruit surfaces	2016 (mcr-1 to mcr-8)	mcr was detected in samples	mcr-1 (1 E. coli and 1 K. pneumoniae)	E. coli (ST189) and K. pneumoniae (ST422) (astA)	E. coli (IncHI1 and IncFIA); K. pneumoniae (IncHI1 and IncFIB)	E. coli:aadA2, aadA1, floR, cmlA1, sul2, sul3, tetA, tetM, dfrA12, mdfA;K. pneumoniae:blaSHV-110, qnrS1, oqxA, oqxB, fosA6, sul1, tetA and dfrA1	-Fruits are direct source of mcr-1-carrying Enterobacteriaceae to humans	[138]
China	Vegetables	2017–2018 (mcr-1 to mcr-8)	mcr-1 detected in vegetable samples	mcr-1 (23 E. coli and 1 Enterobacter)		IncX4, IncI2 and IncHI2	aph(3′)-Ia, aac(3)-IV, aph(4)-Ia, aadA2, aadA1, cmlA, floR, tet(M), fosA3, blaCTX-M-14, sul2, sul3 and mph(A)	-Dissemination of mcr-1 among Enterobactericeae in vegetables is mainly mediated by IncX4 and IncI2 plasmids-mcr-1 is often co-carried with blaCTX-M	[135]
China	Vegetables	2017–2018 (mcr-1 to mcr-8)	-	mcr-1 (2 E. coli)	ST156 and ST2847	IncX4 and IncI2	blaNDM-5 and blaNDM-9	-First report of concomitant occurrence carriage of blaNDM-5/9 and mcr-1 in isolate from fresh vegetables-Plasmids IncX4 and IncI2 have spread globally	[136]

mcr: mobile colistin resistance gene; -: no data; Additional resistance traits: resistance factors identified in one or pooled factors mcr-positive isolates; Sequence type: all sequence types of mcr-positive isolates; Plasmid: plasmid types identified in one or pooled mcr-positive isolates; Inc.: incompatibility; IS: insertion sequence.

Table 7

Studies reporting plasmid-mediated colistin resistance in wildlife (mammals, birds and flies).

Country	Source of Isolate	Date of Isolation (mcr Gene Assayed)	Number of Organism Isolates Tested for mcr	Identified Gene/Variant Detected (Number of Organisms)	Sequence Type and/or Phylogroup (Virulence Genes)	Plasmid (Associated Insertion Sequence)	Additional Resistance Traits	Key Points/Conclusion	Reference
Lithuania	Rectal swabs/faeces of European herring gull (Larus argentatus)	2016 (mcr-1)	117	mcr-1 (1 E. coli)	-	IncI	-	-First report of mcr-1-positive bacteria in migratory bird highlighting the capacity of these birds to spread the gene to various continents	[140]
Spain	Cloacal swabs of a wild bird (black vulture)	2015- 2016 (mcr-1 and mcr-2)	94	mcr-1 (1 E. coli)	ST162	-	bla CIT	-Eating habits of migratory wild birds facilitates acquisition of multiresistant organisms from human environment and global dissemination of these organisms	[141]
Brazil	Pododermatitic Magellanic penguins (Spheniscus magenallicus)	June 2013 (mcr-1)	1	mcr-1 (1 E. coli)	ST10 and A (gad)	IncFIB, IncN, IncHI2, IncHI2A, IncI1 and IncX4	blaCTX-M-1, aadA1, sul2, tetA and tetB	- First report of mcr-1 in penguins highlighting the role of IncX4 plasmid in global dissemination of mcr-1 even to wildlife	[143]
Argentina	Faeces of Kelp gulls (Larus dominicanus) (50)	2012	5	mcr-1 (5 E. coli)	ST744 (4 strains), ST101(1 strain)	IncI2 (ISApI1)	blaCTX-M-14 and blaCTX-M-2	-First report of mcr-1 in Kelp gulls highlighting the capacity of these migratory water birds to disseminate the gene globally	[144]
Algeria	Stool of a Barbary macaques monkey (Macaca sylvanus) (86)	2016 (mcr-1	1	mcr-1 (1 E. coli)	ST405 (fyuA)	-	blaCTX-M-15, blaTEM-1 and qnrB19	-First report of mcr-1 in monkeys-Wild animals can disseminate multidrug-resistant organisms in places frequented by people and in other ecosystems	[55]
China	Flies/faecal swabs/nest swabs of wild birds	2014–2015 (mcr-1)	245	mcr-1 (37 Enterobacteriaceae); mcr-1 detected in samples	-	-	-	-Carbapenem- and colistin-resistant Enterobacteriaceae have disseminated into diverse environment in China	[86]
Pakistan	Cloacal swab of an Eurasian coot (Fulica tra)	2014 (mcr-1)	1	mcr-1 (E. coli)	ST345	IncI2	bla CTX-M-15	-First report of mcr-1 in wildlife in Asia highlighting the presence of this gene in environment without selective pressure-Migratory wild bird can disseminate mcr-1 from Asia to other continents	[145]
China	Flies in animal farms	2016 (mcr-1)	52	mcr-1 (E. cloacae) and mcr-1.3 (1 Raulotella planticola)	-	IncI2 and IncX4 (ISApI1)	ESBL genes	-First report of mcr-1-positive Raulotella planticola-Uncommon species can transmit mcr-1 to common enterobacterial species in which COL resistance is enhanced	[149]
Thailand	Blow flies (Chrysomya megacephala)	2019 (mcr-1)	-	mcr-1 (34 E. coli and 17 K. pneumoniae); mcr-8 (17 K. pneumoniae)	E. coli (ST10, ST648, ST549, ST58, ST181, ST218, ST201, ST162, ST457, ST1244, ST2345, ST2705 and ST5487); K. pneumoniae: (ST34); hicAB, ehAB, eivA,C,E,F,G,I,J, entA,B,C,E,F,S, epaO,P,Q,S,H,I,JkfuABC, iroE, ureABCDEFG, ycfM g, and many others	IncX4, IncHI1B, IncHI1A and IncHI1A-HI1B	E. coli: aac(3′)-IId, aph (3′)-Ib, aph(6)-Id, aph(3′)-Ia, aadA2, aadA1, aadA17, blaTEM-1b, blaCTXM-14, blaCTXM-15, mdfA, InuF, mefB, tetA, tetB, tetM, cmlA, catA2, floR, qnrS1, dfrA12, sul2 and sul3; K. pneumoniae: qnrS1, blaTEM-1b, tetA, blaSHV-40 and fosA	- Flies can disseminate highly virulent multidrug-resistant organisms especially in low- and middle-income countries with unsanitary conditions	[152]

mcr: mobile colistin resistance gene; -: no data; Additional resistance traits: resistance factors identified in one or pooled mcr-positive isolate; Sequence type: comprise all sequence types of mcr gene-positive isolates; Plasmid: plasmid types identified in one or pooled mcr-positive isolate; Inc.: incompatibility; IS: insertion sequence.

Table 8

Studies reporting plasmid-mediated colistin resistance in isolates from multiple environmental ecosystems.

Country	Source of Isolate	Date of Isolation (mcr Gene Assayed)	Number of Isolates Tested for mcr	Identified Gene/Variant (Number of Organism)	Sequence Type and/or Phylogroup (Virulence Genes)	Plasmid (Associated Insertion Sequence)	Additional Resistance Traits	Key Points/Conclusion	Reference
Germany	Farm boot, manure, barn flies and barn dog faeces	2011–2012 (mcr-1)	25, 3, 1 and 5 from boot swabs, flies, barn dog and manure, respectively	mcr-1 (2, 3, 1 and 1 from boot swabs, manure, stable fly, and barn dog, respectively	ST10, ST1140, ST5281, ST1011 and ST342	IncX4, IncI1,IncFII, IncQ1,	strA-like, strB-like, blaTEM-1B-likeblaCTX-M-1, tet(A), aadA1, aadA2, aad5, aph3-likeI, aph(30)-Ia-like, blaTEM1B, cmlA1-like, dfrA1, dfrA14-like, sul2, sul3 and ul2-like	-First report of mcr-1 in environment close to pig farms-mcr-1can be transmitted by diverse routes, including flies feeding on animal manure	[65]
China	Fences and water from swine farm	1999- 2015 (mcr-1)	mcr-1 detected in samples	mcr-1(2 E. coli from fence and 1 E. coli from water)	-	-	-	-Development of real-time PCR conjugated probes for detection of mcr-1 in samples-mcr-1 is prevalent among human and animal population in China highlighting first report on detection of mcr-1 in pets	[68]
Switzerland	Lake/river water and vegetables	2012 -2014 (mcr-1)	74 (water), 60 (vegetable)	mcr-1 (1 E. coli from water, 2 E. coli from vegetable)	ST359/B1, ST167/A, and ST4683/B1	-	blaCTX-M-55 and blaCTX-M-65	-Plant-based food trade is a route for dissemination of colistin-resistant Enterobacteriaceae	[99]
Bangladesh	Flies (Musca domestica) and pond water	2018 (mcr-3)	52	mcr-3 (1 E. coli from water, and 4 E. coli from flies)	-	-	-	-First report of mcr-3-positive carbapenem-resistant E. coli in Bangladesh	[107]
China	River water and egret faeces	2015 (mcr-1)	6	mcr-1 (1 E. coli from water and 2 E. coli from egret faeces)	-	-	tetL, tetO, aac6ib, aadA, blaCMY-2, blaCTX-M-14, IntI and tnpA-1	-First report of simultaneous detection of mcr-1in egrets and their habitat highlighting the role of migratory water birds in contamination of the environment	[111]
Algeria	Soil/manure and irrigation water	2016–2018 (mcr-1 to mcr-5)	103	mcr-1 (6 E. coli) and mcr-3 (2 E. coli)	ST405, ST10, ST345 and ST155	-	bla TEM-12	First report of mcr-3 in environmental isolates in AlgiersAnimal manure is a source for transfer of mcr to soil and irrigation water	[131]

mcr: mobile colistin resistance gene; -: no data; Additional resistance traits: resistance factors identified in one or pooled mcr-positive isolates; Sequence type: all sequence types of mcr-positive isolates; Plasmid: plasmid types identified in one or pooled mcr-positive isolates; Inc.: incompatibility; IS: insertion sequence.

3. Plasmid-Mediated Colistin Resistance in Isolates from Environmental Ecosystems

3.1. Environmental Contact Surfaces

The presence of antimicrobial-resistant organisms on environmental contact surfaces is concerning because these surfaces play a role in the epidemics of COLROS [67]. Seven publications investigated on plasmid-mediated colistin resistance in a total of 775 isolates from environmental contact surfaces. Two of the studies investigated the presence of mcr gene directly in the samples. Fifty-four isolates (33 E. coli, 8 Klebsiellae, 4 Acinetobacter iwofii, 6 Enterobacter, and 1 each for Citrobacter freundii, Pseudomonas aeruginosa, and Proteus putida) were reported to harbour mcr-1 among the tested isolates.

In Asia, contact surfaces in livestock farms were reported to be reservoirs of MGCB. Two mcr-1-carrying nonpathogenic strains were detected among 9 E. coli isolates from fences at a pig farm in China [68], suggesting that animals potentially contaminate their immediate environment with COLROS. Animal dejections (urine and faeces), wild animals (rodents, insects), contaminated water and feed, and/or hands of persons visiting/working in the farm are potential sources/routes of contamination of contact surfaces in a farm environment. In the mcr-1-positive E. coli strains, mcr-1 was located on plasmids of various sizes (<60–150 kb) suggesting that the gene could be transferred between plasmids resulting in its rapid spread among bacterial population [68]. In another study from the same country, 26 mcr-1-carrying enterobacterial strains (23 E. coli and 3 K. pneumoniae) were isolated from contact surfaces (hand rails, vending machine and so on) at public transportation routes [67], suggesting that these surfaces are sources of colonization of travelers by COLROS and that travelers potentially disseminate these organisms from one location to another thereby posing public health risks [69]. Interestingly, most of the mcr-1-positive strains were recovered from samples collected from areas with a high density of hospitals or traffic indicating that these isolates could be of nosocomial origin thus highlighting the need for hand hygiene to prevent transmission of MGCB capable of causing diseases with pandemic potential. The mcr-1 was located on different plasmids (IncI2, IncX4 and IncH2) and the mcr-1-positive E. coli isolates belonged to phylogroups B1 and A, suggesting that diverse promiscuous plasmids facilitate the spread of mcr-1 among commensal E. coli strains in the environment. There were 4 resistance genes (including ESBL gene) in 2 different antibiotic families in the mcr-1-positive E. coli isolates and the K. pneumoniae isolates, indicating that diverse multidrug-resistant organisms (MDROS) could be acquired during travel and subsequently disseminated to other locations thus posing serious public health risks [69]. The mcr-1-positive E. coli isolates were extensively diverse (belonging to 9 different STs) with pandemic high-risk (HiR) international zoonotic clones ST10 (the most dominant ST complex of human strains) and ST101 complexes [70,71] being predominant while all the K. pneumoniae isolates were of ST37 (Table 1). Thus, suggesting that hands of travelers are routes for disseminating hospital-acquired colistin-resistant enterobacterial clones into the public. This could result in the outbreak of hard-to-treat pandemic infections that can easily spread among individuals in densely-populated areas as often seen at transportation routes. Presence of MGCB on environmental contact surfaces in other Asian countries is yet to be reported. However, it is worth noting that in Bangladesh, mcr-1 and mcr-2 were detected in E. coli isolates from poultry birds/street foods but not in samples from poultry farm environment [72], suggesting that COLROS may not survive harsh environmental conditions particularly where high biosecurity measures (such as disinfection of farm environment) are put in place.

Farm worker’s paraphernalia were reported to be vehicles for transmission of MGCB in Europe. Two strains carrying mcr-1 on IncX4 plasmid were detected among 25 E. coli isolates (8%) from boots used 50–150 m distance from pig farms in Germany [65], suggesting that farmers’ wears (often in contact with animal dejections) are potential vehicles for transmission of MGCB from livestock farms to other locations. Thus contact with materials used in animal farms is a potential route for the acquisition of COLROS. In the strains, mcr-1 was carried on plasmids of varied sizes, further suggesting that mcr-1 is transferred between plasmids thereby facilitating its rapid spread among bacterial population. The mcr-1-positive E. coli strains were of ST1140 and HiR ExPEC clone ST10 that causes intestinal and several extraintestinal diseases worldwide [70,71,73], and harboured 10 other resistance genes (including ESBL gene) in 5 different antimicrobial families (Table 8) suggesting they are multidrug-resistant organisms (MDROS) capable of causing hard-to-treat infections thereby posing a serious challenge to public health. It was also reported that the ST10 isolate and another mcr-1-IncX4-positive faecal E. coli isolate from a barn dog were related differing only by 7 single nucleotide polymorphisms (SNPs) suggesting that IncX4 is a major driver of mcr-1 among E. coli strains colonizing different ecological niches [65]. In the same country, a retrospective survey detected 43 mcr-1-carrying strains among 436 pooled faecal/boot swab E. coli isolates (9.9%) from 15 farms [74], further indicating that farmers or their wears are vehicles for transmission of COLROS. Contact surfaces in European hospitals were also reported to be reservoirs of MGCB. Twenty-five mcr-1-carrying multidrug-resistant (MDR) strains (4 each of Acinetobacter iwoffi, and E. coli, 1 each of Citrobacter freundii, Pseudomonas aeruginosa and Proteus putida, 3 each of Enterobacter (E.) cloacae and E. agglomerans, 6 K. pneumoniae, and 2 K. oxytoca) were detected among 300 isolates (8.3%) from hospital surfaces in Italy [29], suggesting that these surfaces represent huge reservoir of diverse hard-to-treat nosocomial pathogens that can easily diffuse into the public thereby portending danger to public health. Intriguingly, a Proteus putida strain harboured mcr-1 thereby suggesting that the ability of Proteus species to acquire/transfer mcr have been underestimated all this while that this organism was regarded as being intrinsically-resistant to COL hence a non-mcr-carrier [75,76,77].

3.2. Sewages/Wastewaters

Wastewaters and sewages contain nutrients that support the growth of bacteria, thus are increasingly being recognized as a source of new emerging pathogens and antibiotic resistance [78]. The more concern about wastewater treatment plants (WWTP) is the hygienic quality of receiving waters and for water reuse [78]. Fourteen publications investigated the mcr gene in a total of 185 isolates from sewages/wastewaters. Sixty-three isolates (57 E. coli, 3 Kluyvera, 2 Klebsiella pneumoniae and 1 unspecified isolate) were reported to harbour mcr-1 among the isolates tested. Four of the studies investigated mcr gene directly in sewage/wastewater samples.

In Europe, wastewaters and sewages are noted as reservoirs of mcr-1. In a German study, mcr-1 was detected in influent and effluent waters of 3 out of 7 WWTPs [79], indicating that WWTPs are reservoirs of COLROS. Remarkably, one of the WWTPs containing mcr-1 had never received wastewater from hospitals, intensive animal farms or food industry, thus suggesting that domestic wastewater (from toilets and bathrooms) or wildlife (rodents, insects, birds, amphibians, worms, and so on) might be the sources of the gene in those plants. Apart from mcr-1, other resistance genes (including ESBL and AmpC genes) in 3 different antibiotic classes were detected in the wastewaters (Table 2), suggesting that WWTPs are reservoirs of multiple resistance genes originating from anthropogenic sources. These WWTPs are sources from which mcr/other ARGs are spreading to other ecosystems. Although mcr-1-carrying bacteria were not recovered from the wastewater samples, gene markers for A. baumanii, E. coli and K. pneumoniae were detected, thus incriminating these organisms as reservoirs of mcr-1 in wastewaters. In a recent study from the same country, a relatively lower copy number of mcr-1 compared to mcr-3, mcr-4, mcr-5 and mcr-7 was detected in 14 municipal wastewater samples [80], suggesting that wastewaters from anthropogenic sources are cocktails of mcr genes and potential sources of dissemination of these determinants to other ecosystems. Therefore surveillance of only an mcr gene-type in an environmental matrix could result in underestimation of the magnitude of this global problem [80]. Thus, in order to adequately determine the extent of plasmid-mediated colistin resistance in an ecological niche, there is the need for assaying different mcr gene-types thus warranting the development of affordable rapid kits that could detect all the existing mcr gene-types and the ones that are yet to emerge.

In a study from Spain, 30 strains (29 E. coli and one K. pneumoniae) carrying mcr-1 on IncI2 plasmid were detected among 90 isolates (33.3%) from sewage [81], suggesting that IncI2 plasmid is one of the common drivers of mcr-1 in the sewage environment. There was diversity among the isolates with the ESBL-producing MPEC strains belonging to ST1196 and ST224 while the K. pneumoniae strain belonged to ST526 (Table 2). This means that diverse multidrug-resistant (MDR) COLROS are present in sewage in Spain. Though the ST224 strain harboured mcr-1, it exhibited COL susceptibility which might be caused by the cell wall structure or the copy number of mcr-1-IncI2 plasmid [81]. This suggests that mcr-1 could act as a silent gene evading phenotypic detection, thereby favouring its dissemination [81]. Transformation of mcr-1 from the mcr-1-positive E. coli strains into a recipient organism was positive, indicating that these strains can rapidly spread the gene to other organisms by HGT [82]. In another Spanish study, mcr-1 was detected in filtered pellets from untreated and treated wastewater of domestic, hospital, industrial and agricultural origins [66], further indicating that these are sources of COLROS in WWTPs. Expectedly, the untreated wastewater contained a significantly higher mcr-1 gene copy numbers than treated wastewater meaning that treating wastewater decreases the number of COLROS but does not entirely remove the mcr-1. Thus, even after treatment, wastewater may still be a source of mcr-1 into receiving water bodies [66]. This warrants the need for improvement of protocols used in sewage/wastewater treatment.

The presence of MGCB in wastewater/sewages/sludges in Asian countries has been reported. Two pathogenic strains carrying mcr-1 on IncHI2 and IncX4 plasmids were detected among 65 ESBL-producing E. coli isolates (3.1%) from canal water in Thailand [83], suggesting that diverse plasmids are spreading genes conferring COL resistance and resistance against last-resort antibiotics in the study area. In a Bangladeshi study, an mcr-1-carrying E. coli strain was detected among 48 isolates from sludge samples [84], suggesting that mcr-1-positive E. coli has dispersed into the Bangladesh environment possibly due to improper disposal of anthropogenic/agricultural wastes.

In China, several reports indicated that sewages/wastewaters constitute reservoirs of COLROS/mcr genes. Eighteen of 24 samples of wastewater at different stages of treatment (75%), including water to be released to the sea post-treatment were observed to contain mcr-1-carrying E. coli and Klebsiella strains [85], further indicating that COLROS can survive during wastewater treatment thereby ending up in water bodies. This highlights the need for effective sewage/wastewater treatment methods. A surveillance study on dissemination of mcr and carbapenem resistance in Chinese poultry production sector detected mcr-1 in sewage sample from a poultry farm, and an enterobacterial isolate carrying mcr-1 on IncI plasmid was obtained from the sample [86], further suggesting that IncI plasmid is a major driver of COL resistance in livestock environment and that animal farm wastes constitute a reservoir for MGCB. In another Chinese study, 11 mcr-1carrying strains (9 E. coli and 2 Kluyvera) were detected in sewage samples [87]. Two of the mcr-1-positive E. coli strains harboured new mcr-1 variants, named mcr-1.7 on IncI2 plasmid and mcr-1.4 on IncX4 plasmid; further indicating that IncI2 and IncX4 plasmids are the commonest drivers of mcr in uncommon and common species from the environmental ecosystem in China. This may mirror what is happening in human/animal ecosystems in the country since the sewage samples originated from these settings. Apart from the fact that the mcr-1-positive E. coli strains transferred mcr-1 genes at a high frequency (10−5 to 10−7 cells per recipient cell) to a recipient organism, the presence of ISApI1 upstream of mcr-1 in the strains imply they could rapidly transfer the genes to diverse organisms. However, mcr-1 was chromosomally-located in 2 of the mcr-1-positive E. coli isolates indicating that these organisms maintain the gene through the vertical transmission to their progenies hence making elimination of the gene from the environment practically impossible.

Furthermore, there were 31 other resistance genes (including ESBL and pAmpC genes) in 8 different antimicrobial families in the isolates (Table 2), further indicating that the mcr-1-positive E. coli strains were multidrug-resistant organisms (MDROS) thus posing challenge to public health. Nevertheless, the strains exhibited susceptibility to some antimicrobial agents, including imipenem and tigecycline. It is however troubling that the MPEC strains were extensively diverse (5 STs) with some isolates belonging to zoonotic HiR ExPEC clones (ST10, ST34, ST48). This further suggests that transmission of mcr-1 among environmental E. coli strains is nonclonal.

Several Chinese studies noted that hospital sewage (which may contain clinical specimens such as faeces, fluid, blood, decomposed tissues and so on) is a potential source/route of spread of COLROS. In a study on the global spread of mcr-1, 8 mcr-1-carrying E. coli strains were recovered from influents/effluents of tertiary care teaching hospitals in China [2]. In another study from China, a multidrug-resistant ST313 K. pneumoniae strain carrying mcr-1on IncP-1 plasmid (a broad host-range incompatibility group) was isolated from hospital sewage [88], suggesting that sewages are sources for the dissemination of mcr-1 plasmids transferrable to diverse organisms. There was ISApl1 downstream of mcr-1 suggesting that the gene was acquired, and 9 other resistance genes in 9 different antimicrobial families as well as a novel ORFble gene with yet unknown function (Table 2), further suggesting that sewages in China are potential sinks for novel antimicrobial determinants [88]. In another Chinese study, 9 mcr-1-carrying E. coli strains were detected among 25 Enterobacteriaceae (isolated from hospital sewage) [89], suggesting that E. coli is the most common COLROS in sewage ecosystem. In the mcr-1-positive E. coli isolates, mcr-1 including 4 genes encoding β-lactam resistance was located on ~33 kb IncX4 plasmid (Table 2). There was extensive diversity among the strains (7STs) some of which were zoonotic HiR ExPEC lineages such as ST10 and ST101 [70,71]. These E. coli clones might eventually be released into farmlands and water bodies. Interestingly, an mcr-1-IncX4 plasmid of the ST10 and ST7122 isolates conjugated with the genome of a recipient organism at very high-frequency rates of 2.79 × 107–7.60 × 108, further indicating that IncX4 is a major driver of COL resistance in environmental strains.

From the same country, a Kluyvera ascobarta strain carrying mcr-1 on IncI2 plasmid was recovered from hospital sewage [90], further indicating that diverse uncommon species are spreading COL and multiple resistance genes in the Chinese environment. The Kluyvera strain can rapidly spread COL resistance having transferred mcr-1-IncI2 plasmid to a recipient organism at a very high frequency (10−5 cells per recipient). There were 2 other resistance genes (including a novel ESBL gene) conferring resistance to 2 different antimicrobial classes (multi-resistance) in the strain with extensive susceptibility to 14 antimicrobials, including imipenem, implying that mcr-1 is not necessarily carried by extensively drug-resistant strains [90]. In a recent Chinese study, an E. coli ST410/phylogroup A strain carrying a rare combination of mcr-1.1, mcr-3.5 and blaNDM-5 on a 265.5 kb IncHI2-ST3 plasmid (containing IncN, IncP and IncX3 plasmids) was isolated from influx mainstream of hospital sewage [9], further indicating that HiR zoonotic ExPEC clones [91] are circulating mcr-1 on diverse plasmids in the Chinese environment. It is worrisome that the organism harboured 31 other resistance genes in 10 different antimicrobial families meaning it is extensively drug-resistant; treatment of infections caused by extensively drug-resistant strains are often difficult. However, the organisms were susceptible to aztreonam-avibactam and tigecycline. A similar amikacin-susceptible clinical human ST206 E. coli strain also harbouring mcr-1.1, mcr-3.5 and blaNDM-5 was also detected in China [92], suggesting that diverse E. coli clones containing COL and carbapenem resistance genes, are circulating in human and environmental ecosystems in the country. However, the sewage and human isolates were not the same since the human strain did not contain rmt gene which codified high-level amikacin resistance; nevertheless, there is the possibility of acquisition of this gene from the sewage environment. Unfortunately, E. coli clones containing both carbapenem and COL resistance genes have disseminated into the environmental ecosystem; these determinants can rapidly spread to diverse clones [9,92,93,94]. Worse of it, is that the mcr-1.1-, mcr-3.5-, blaNDM-5- and rmt-plasmids, were transferred to another organism at a very high (10−3–10−4) frequency. In fact, these 4 genes together were transferred at an outrageously high (10−10) frequency, suggesting that multiple drug resistance genes, even when not located on the same plasmid can be spread at a very high rate [9].

The spread of MGCB into aquatic ecosystem due to poor environmental sanitation has been reported in Asia. Eight mcr-1-carrying multidrug-resistant Proteus mirabilis strains were isolated from sewer waters in the Syrian war refugee camps in Lebanon [95], further suggesting that Proteus species are potentially mcr-harbouring organisms which should no longer be neglected. Retrospective studies to investigate mcr in COL-R Proteus isolates from various ecosystems are warranted. There was a β-lactam determinant as well as class 1 integrons in the strains, indicating that diverse MGEs facilitate the spread of mcr-1 in aquatic ecosystems. Interestingly, the Proteus strains expressed biofilm meaning they can potentially persist in the camp environment making it practically impossible to eliminate mcr-1. This finding of mcr-1-positive Proteus in refugee camps calls for urgent intervention because these organisms can cause highly-resistant clinical conditions (such as bacteraemia, urinary, respiratory, eye, and wound infections) mostly affecting immunocompromised persons (such as children, women and diseased individuals) predominating in the camps [95].

Although colistin-resistance have been detected in sewage/wastewaters in South America (Figure 1), the presence of mcr in these ecosystems in the continent is yet to be reported. Nevertheless, it is worthy to note that no mcr-1-carrying strain was detected among 7 COL-R enterobacterial isolates from sewage samples in Brazil [96]. Similarly, no strain among 35 enterobacterial isolates from sewage in Venezuela harboured mcr-1 [97].

Figure 1

Countries in which mobile colistin resistance (mcr) gene mcr-1, including other mcr gene types indicated, have been detected in environmental (water, aquaculture, soil, plant, sewage, wastewater, and wildlife) reservoirs.

3.3. Water (Fresh and Sea), Aquaculture and Aquatic-Based Foods

3.3.1. Freshwater and Seawater Ecosystem

Water bodies (both surface and underground) are receptacles of anthropogenic wastes (from human and animal sources); thus they are reservoirs of ARGs [98]. Eighteen publications reported on plasmid-mediated colistin resistance in a total of 2107 isolates from fresh and seawater. These studies investigated mcr genes in a total of 1652 isolates and reported mcr-1 gene in 68 isolates (60 E. coli, 2 Enterobacter, 4 Citrobacter and 2 Klebsiella) and mcr-3 gene-type variants in 3 E. coli and 11 Aeromonas isolates, respectively.

Dissemination of MGCB into the surface and underground waters have been reported in Europe. An mcr-1-carrying multidrug-resistant ST359 E. coli strain was detected among 74 ESBL-producing Enterobacteriaceae (1.35%) isolated from rivers and lakes in Switzerland [99], suggesting contamination of Swiss surface waters by COLROS capable of causing difficult-to-treat infections. In a study from Italy, an ST10 E. coli strain carrying mcr-1.2 together with β-lactam resistance genes (including ESBL gene) on IncX4 plasmid, was detected among 264 isolates (4.2%) from well/stream water [100], suggesting that underground waters in Italy are reservoirs of HiR zoonotic COL-R ExPEC [70,73]. The mcr-1-positive E. coli strain transferred COL resistance to a recipient organism at a very high rate of ~10−2 transconjugants per recipient, meaning it can rapidly spread the resistance genes to other organisms in the aquatic systems thereby posing public health risks.

Similarly, 2 ESBL-producing ST10 E. coli strains carrying mcr-1 on diverse plasmids (IncFII, IncI1, IncFIB, Col156, IncX4 and ColRNAI) were detected among 83 coliform isolates (2.4%) from seawater at a public beach in Norway [101], suggesting that coastal seawater is a potential source for acquisition of HiR pandemic COL-R ExPEC clones [70,73]. Since ST10 is a zoonotic clone, its presence in seawater suggests it may have originated from humans bathing in the water, contamination from boat toilets, farm animals, fertilizers (manure) used in agriculture or migrating birds [101]. Presence of mcr-1-positive E. coli strains in coastal waters/public beaches is worrisome because individuals visiting these beaches for recreation and other purposes could get infected with organisms that potentially cause difficult-to-treat infections. Even worse was the fact that the mcr-1-positive E. coli strains harboured 8 other resistance genes in 5 different antimicrobial families (Table 3), further suggesting that seawater might be spreading a cocktail of multiple resistances, posing a significant threat to public health [102].

South American studies also reported the presence of MGCB in surface waters. Three multidrug-resistant E. coli isolates of ST10/phylogroup B1, ST46 and ST1638 carrying mcr-1 on ~33 kb IncX4 plasmid, were recovered from coastal public beach waters in Brazil [103], suggesting that Brazilian beaches are potential reservoirs of colistin-resistant HiR pandemic ExPEC clones thus posing serious public health risks, especially to those (such as residents, beach workers, tourists, and wildlife) directly exposed to this infectious threat from water exposure, contact with sand or through food consumption on the beach [70,73,103]. There were 15 extra resistance genes in 7 different antimicrobial families in the strains (Table 3), further suggesting that sea water is a reservoir for multiple resistance genes possibly emanating from anthropogenic/agricultural sources. However, the strains were susceptible to carbapenems. Of note, in another Brazilian study, no mcr-1-carrying strain was detected among 5 colistin-resistant enterobacterial isolates from lakes and rivers [96]. In another Brazilian study, no mcr-1- or mcr-2-carrying strain was detected among 40 Enterobacteriaceae isolated from urban lakes contaminated with human wastes [104]. However, these lakes contained ESBL-producing ST131 E. coli, which is a zoonotic pandemic HiR international ExPEC clone [70,73], thus highlighting the risk of poor environmental sanitation in Brazil.

MGCB has been detected in surface/underground waters in Asia. An E. coli strain carrying mcr-1 on IncI2 plasmid was isolated from water in Malaysia [105], suggesting that IncI2 may be a common plasmid spreading colistin-resistant resistance in Malaysian aquatic systems. From the same country, a retrospective study on 900 ST410 E. coli isolates collected in 2009 from pond water, detected one mcr-1-carrying strain (0.1%) [106], suggesting that MGCB has been circulating in Malaysia for the past 10 years. There were 13 resistance genes in 7 antimicrobial families in the mcr-1-positive E. coli strain (Table 3), further indicating that accumulated environmental surface waters are potential sinks for multiple resistance genes possibly originating from human/animal settings. In a Bangladeshi study, one mcr-3-carrying strain was detected among 12 carbapenem-resistant E. coli isolates (8.3%) from pond water [107], suggesting that surface water systems in Bangladesh are potential reservoirs of colistin-resistant as well as carbapenem-resistant organisms, thus posing a worrisome threat to public health. These COLROS possibly emanated from human/animal settings and they could spread from these ponds to other ecosystems by rainfall run-offs. Also, animals depending on these ponds for sustenance, could get infected with COLROS and consequently disseminate them to other ecological niches.

The presence of MGCB in irrigation water was reported in Asia. Twenty-two mcr-1-carrying E. coli strains were isolated from irrigation water in Lebanon [108], suggesting that anthropogenic/agricultural wastes have contaminated Lebanese irrigation system. This finding highlights the need for improved waste management in Lebanon, a country whose infrastructural facilities are being overstretched due to the prolonged Syrian war. Interestingly, the mcr-1-positive E. coli isolates survived in water for 45 days without losing mcr-1 implying that the gene persists in water matrix [108]. There were 5 genes encoding β-lactam resistance (including against carbapenem) and class 1 integrons in the strains (Table 3), indicating that diverse MGEs drive the spread of genes conferring resistance against last-resort antibiotics in the region. The presence of mcr-1-positive E. coli strains in the Lebanese irrigation system requires urgent attention because these superbugs could spread into the Mediterranean Basin contaminating plants and sea thereby spreading (by water current) to other parts of the world. In another Lebanese study, 32 mcr-1-carrying E. coli strains were isolated from domestic (drinking/well waters) and sewer-generated water samples collected from Syrian war refugee camps [109], suggesting that underground and drinking waters are a potential source/route for the acquisition of COL resistance. Possible sources of contamination of domestic water include sippage from sewers into underground water or human carriers. Class one integrons and β-lactam resistance genes (including carbapenem genes) were also present in the strains (Table 3), further indicating that diverse MGEs are involved in the spread of COL resistance in aquatic and other ecosystems in Lebanon. This report highlights the need for increased surveillance of domestic waters for MGCB, and improvement in the disposal of wastes as well as water treatment in the refugee camps. Survival of the strains in water for 30 days without the loss of mcr-1, further indicates persistence of the gene in water.

Several investigators from China reported the presence of mcr genes in the water ecosystems. In one study, the mcr-1 was detected on 33 kb IncX4 and 60 kb IncI2 plasmids, as well as on non-conjugative plasmids in E. coli isolates from seawater samples [85], further indicating that seawater is a route for spreading (by sea current, fishing or ships) COL resistance into aquatic ecosystems worldwide. It also further suggested that IncX4 and IncI2 plasmids are the major drivers of COL resistance in seawater ecosystems. In another study, 2 Aeromonas caviae strains carrying novel mcr-3 genes named mcr-3.13 and mcr-3.18 as well as mcr-3-like4 gene, were isolated from river/lakes/fountain water samples [110], suggesting that mcr-3 genes have disseminated into environmental waters in China. Two mcr-3.14- and mcr-3-like4-carrying A. bivalvium strains were also isolated from the same water samples. These strains contained 6 additional resistance genes in 4 antimicrobial families (Table 3) but exhibited susceptibility to some antibiotics, including polymyxin B, carbapenems and COL suggesting that Aeromonas species (common inhabitants of aquatic systems) are reservoirs of mcr-3 and multiple resistance genes thus posing a challenge to public health.

In a further study, 2 mcr-1-carrying strains were detected among 6 colistin-resistant E. coli isolates (33.3%) from river samples [111], further suggesting the spread of COLROS in surface waters in China. There were transposon and class 1 integrons, including 6 resistance genes in 3 different antibiotic families in the organisms (Table 8), further indicating that diverse MGEs drive COL resistance in the Chinese environment. Similarly, 2 mcr-1-carrying strains were detected among 10 ESBL-producing E. coli isolates (20%) from well water collected from sites where chicken manure has been applied in the soil for a long time [62], suggesting that animal manure is a potential source of COLROS in underground water. Human carriers, sippage from septic tanks, as well as dejections from wildlife (such as reptiles and overflying birds) particularly if the well was not covered, might also be the sources of the mcr-1-positive E. coli strains. Nonetheless, it is worrisome that the mcr-1-positive E. coli strains were ESBL-producers and belonged to HiR zoonotic pandemic ExPEC ST10 and ST48 clones (both of ST10 complex) [70,73], thus posing health challenge to the public. These findings strongly suggest that in China, COLROS are circulating within the animal-human-environmental ecosystem because isolates in ST10 complex are the most common faecal E. coli strains detected in humans in China [62,112].

In a different study, 23 mcr-1-carrying Enterobacteriaceae (16 E. coli, 1 Enterobacter cloacae and 2 each of Citrobacter freundii, Citrobacter braakii, and K. oxytoca) were isolated from mcr-1-positive water samples [113], suggesting that although E. coli is the most commonly detected mcr-1-harbouring organism, transmission of this gene is not restricted to any species but in diverse organisms in aquatic environment thus posing public health risk. Seven other genes encoding β-lactams (including ESBL) resistance and PMQR were found in the mcr-1-positive E. coli isolates which were extensively diversified belonging to 6 different clones dominated by HiR pandemic clone ST10 [71,73], thus portending grave danger to public health. In another study, 25 mcr gene-positive isolates comprising 18 mcr-1-carrying isolates (17 E. coli, and 1 Enterobacter cloacae) and 7 mcr-3-carrying Aeromonas strains (2 Aeromonas veronii and 4 A. hydrophila) were detected among 1500 isolates (1.67%)from water samples collected from different points of a river [114], suggesting that COLROS are widely distributed in rivers in China. It further shows that diverse organisms are spreading COL resistance in aquatic systems, while Aeromonas seems to be a major reservoir of mcr-3. In addition, there was high concentration (2.0–2.7 log10 GC/mL) of mcr-1 and other resistance genes in the river and its surrounding environment, further indicating that contaminants in Chinese surface waters are the majority of anthropogenic/agricultural origins.

Although there is a paucity of information from Africa, there are evidence that COLROS has disseminated into water ecosystem in the continent. Two tigecycline-resistant E. coli strains of ST23 and ST115 carrying mcr-1.5 on IncI2 plasmid and mcr-1.1 on IncHI2 A plasmid, respectively, were detected among 246 colistin-resistant isolates (0.8%) from seawater polluted with domestic, hospital, agricultural and industrial wastes [102,115], further indicating that anthropogenic/agricultural wastes (human and animal ecosystems in Algiers contain MGCB) are sources of pandemic HiR international COL-R ExPEC clones in seawaters [70,73]. This finding calls for serious concern because tigecycline is one of the last antibiotics used for managing deadly infections. There is need for rapid surveillance of COL and tigecycline resistance in African countries because in most of these nations, the use of antimicrobials in human/animal/environmental settings is not strictly controlled and sanitary facilities are lacking resulting in poor management of wastes which most often are disposed into the environment (eventually carried to surface waters by runoffs) and surface waters. Unfortunately coastal and rural dwellers in Africa use these waters for a variety of purposes (recreation, drinking, fishing, laundering, bathing) potentially exposing them to infection by superbugs. In another Algerian study, an mcr-1-carrying ST345 E. coli strain was isolated from 10 irrigation water samples [116], further indicating that water bodies in the country have been contaminated by diverse colistin-resistant E. coli clones originating from anthropogenic/agricultural wastes.

3.3.2. Aquaculture Environment and Aquatic-Based Foods

The heavy use of antimicrobials in fresh and saltwater aquaculture underline why aquacultural environments/aquatic based-foods is now increasingly recognized as major reservoirs and source of dissemination of antimicrobial resistance [12,28,47]. Eight publications assessed plasmid-mediated colistin resistance in a total of 838 isolates from aquaculture/aquatic-based foods. Seventy-two isolates (62 E. coli, 3 Aeromonas carried mcr-3 gene variants, 2 Klebsiella pneumoniae and 1 Salmonella enterica serovar Rissen) were reported to harbour mcr gene. One of the publications detected mcr-1 gene in seafood samples by direct sample testing.

European studies reported MGCB in aquaculture/aquatic-based foods. A quinolone-resistant ST48/phylogroup A E. coli strain carrying mcr-1 on IncHI2, IncN and IncX3 plasmids was isolated from seafood (scampi) imported into Norway from Bangladesh [117], suggesting that seafood trade is a potential route for intercontinental dissemination of COLROS. There were 16 resistance factors in 11 different antimicrobial families, including heavy metals in the strain (Table 4), suggesting that contact or consumption of seafoods are sources for acquisition of multiple resistance genes thus posing a worrisome threat to public health. In a retrospective German study, 3 Aeromonas strains (1 A. allosaccharophila, 1 A. jandei, and 1 A. hydrophila) isolated from fishes in 2005–2008, were observed to carry novel chromosomally-encoded mcr-3.6, -3.8 and -3.9, respectively [118], suggesting that Aeromonas species potentially maintain diverse mcr-3 genes in aquatic environment. There were 8 other resistance genes in 2 different antibiotic families in the strains (Table 4), further suggesting that fishes are a potential reservoir of multi-resistance genes. Interestingly, the mcr-3 gene variants in the isolates were acquired and not part of the indigenous genomes suggesting that mcr-3 genes circulating in the aquaculture environment may not necessarily have originated from the aquatic environment. The Aeromonas strains/mcr-3 genes probably disseminated into the aquatic environment from human/animal settings since plasmid- and chromosomally-encoded mcr-3 genes have also been from humans/animals Aeromonas strains in the study region [110,119]. Detection of mcr-3 in a strain isolated in 2005 which is at least 10 years older than all the mcr-3-positive isolates from China suggests that mcr-3 gene group has been present for at least 12 years in Europe [118].

Recently in Galicia Spain, an mcr-1-carrying ST469 Salmonella enterica Serovar Risen strain was detected among 19 Salmonella isolates (0.3%) from 5907 mussels collected from processing facilities [120], suggesting that these bivalve molluscs were captured in water bodies contaminated by anthropogenic/agricultural wastes such as sewage discharges/outfall, combined sewer overflows, rainwater, aquaculture and wildlife discharges since ST469 Salmonella Rissen clone have been reported in livestock [121]. There were 9 other resistance genes in 6 different antimicrobial families in the strain (Table 4), suggesting that MDR Salmonella could be acquired from mussels thus posing a worrisome threat to the handlers and consumers of seafood. This finding of mcr-positive Salmonella strain in Galician mussels calls for concern because Galicia is the third-largest producer of cultured mussel worldwide and is considered the leading supplier of mussels to the European market [120]. It is also worthy to note that in Portugal, none of mcr-1 to mcr-5 was detected in isolates from aquaculture environment [122].

In the recent past, intensive aquaculture in China was characterized by heavy use of antimicrobials, including COL [12,47]. Some Chinese investigators reported the presence of mcr genes/MGCB in the aquaculture environment/aquatic-based foods. The mcr-1 was detected in 9 of 63 (14%) and 2 of 12 (17%) seafood samples sourced locally and from overseas, respectively [85], further suggesting that local and international trade of aquatic-based foods are potential routes for the dissemination of MGCB. Multidrug-resistant E. coli, K. pneumoniae and A. veronii strains susceptible to some antibiotics, including meropenem and tigecycline were isolated from the mcr-1-positive samples, further indicating that diverse multidrug resistance organisms could be acquired from sea-based foods. In another Chinese study, 7 strains carrying mcr-1 on different plasmids (IncP, IncX4 and IncI2) and on the chromosome were detected among 190 E. coli isolates (3.7%) from cultured grass carp fish [123], further suggesting that COL resistance is circulating by horizontal and vertical transfer among isolates from aquatic ecosystems. There was diversity among the isolates (ST7508, ST2040, ST156 and ST48), suggesting that promiscuous plasmid types resulted in a diverse range of mcr-1-carrying clones in aquaculture environment. In addition, conjugation was positive at a very high frequency, and there were 19 other resistance genes in 9 different antimicrobial families plus a composite transposon in the strains, thus suggesting that the organisms could rapidly transfer COL and multiple resistances, thus posing a serious challenge to public health.

Integrated livestock-fish farms have also been reported to be reservoirs of COL resistance in China. Recently, 54 E. coli and 2 K. pneumoniae strains carrying mcr-1.1 on different plasmids (IncHI2, IncI2, IncX4, IncP, and Incp0111) and 4 mcr-3-carrying Aeromonas strains, were detected among 143 colistin-resistant isolates from samples collected from duck-fish integrated fishery [28]. There was ISApI1 upstream of mcr-1 in the mcr-1- positive E. coli isolates which were extensively diverse belonging to 5 ST with the zoonotic ExPEC clone ST93 being the predominant [71,73] (Table 4). This further indicates that diverse MGEs transfer COL resistance in diverse clones of E. coli in livestock/aquatic ecosystems in China. It also further indicates that Aeromonas is a potential reservoir of mcr-3 in livestock/aquatic ecosystems. Remarkable, 2 ST156/phylogroup B1/cluster C2 mcr-1-positive E. coli strains from a duck and a fish, had zero SNPs between them indicating that integrated farming is a route for transferring mcr genes between livestock and aquaculture [28]. Additionally, the mcr-1-positive E. coli isolates from the integrated farms were genetically related to those from humans in the farm region, thus suggesting that MGCB transfer between animals and humans via the aquatic food chain [28]. There is a dearth of information on plasmid-mediated colistin resistance in aquaculture ecosystem in other Asian countries. However, it is worth noting that no mcr-1- or mcr-3-carrying strain was detected among 6 colistin-resistant Klebsiella isolates from fish in India [124]. Contaminated human carriers, improper disposal of human and animal sewage to the aquatic environment, feeding of fish with contaminated feed or animal manure especially in integrated farms, dejections of wild animals, and/or the emergence of mcr genes from aquatic organisms, are possible sources of MGCB in aquatic/aquaculture environment [56,123]. Therefore, to curb the spread of COL resistance through the trade of aquatic products, increased surveillance of aquaculture is warranted [123].

4. Soil/Manure Ecosystem

There is a growing interest about the role of animal manure and other resistance reservoirs (such as sewages/wastewaters/sludges, aquaculture and wildlife) in the transmission of resistance genes to the soil which constitutes a source for dissemination of these genes to botanical, aquatic and wildlife ecosystems [22,23,24,25]. Nine publications investigated on plasmid-mediated colistin resistance in a total of 276 reported isolates from soil/manure/slurry/sediment. Thirteen isolates (14 E. coli and 2 Enterobacter) were reported to harbour mcr-1 gene. Three of the study quantified mcr-1 gene in soil/manure samples.

Studies from South America reported the presence of mcr gene in soil. In a Brazilian study, mcr-1 was detected in soils from vegetable production areas that received non-composted poultry litter as organic fertilizer as well as in native vegetation areas without livestock manure [125], suggesting that the COL resistance in the soil does not necessarily occur only when livestock manure has been used. Anthropogenic wastes (especially from industries, hospitals, and laboratories) that may contain antibiotics, aquaculture, sewages/wastewaters and wildlife discharges, are putative sources of COL selection pressure in the soil where livestock manure has never been applied. Although not surprising, the mcr-1 was more abundant in the fertilized vegetable production area than in the native vegetation area, suggesting that non-composted poultry litter is a potential source of mcr-1/MGCB in the soil.

European studies also reported that manure/soil is reservoirs of mcr genes/MGCB. Two E. coli strains of ST5281 and ST1011 carrying mcr-1 on diverse plasmids (IncI1, IncFII, IncFIB, IncX1 and IncQ1) were isolated from manure collected close to pig farms in Germany [65], suggesting that livestock manure is a reservoir of promiscuous plasmids resulting in diverse range of mcr-1-positive E. coli clones in manure/soil environment. These organisms also harboured 17 resistance genes in 5 different antimicrobial families (Table 8) indicating they are multidrug-resistant thus posing a serious risk to public health, especially to the handlers of the manure, crop farmers, and consumers of plant products from farms fertilized with these manures (Table 8). Interestingly, there was close-relatedness between the ST1011 isolate to another E. coli isolate from a stable fly (Musca domestica) collected from a close distance to another pig farm in the same country, thus suggesting that flies are potential vectors of MGCB transferring these organisms from livestock farms to other ecological niches. Since these garbage flies usually feed on and breed in animal manure [57], there is a need for increased environmental sanitation especially regarding the disposal of animal manure. In an Estonian study, 3 E. coli strains carried mcr-1 on IncX4 plasmid were detected among 141 ESBL-producing isolates (2.6%) from pig slurry originating from a farm [126], further indicating that animal farm wastes are potential sources for the spread of COL resistance to the soil and other ecosystems. There were IS26 upstream of mcr-1 and 3 genes (including ESBL and pAmpC gene) in the mcr-1-positive E. coli strains, suggesting that different MGEs (insertion sequences and plasmids) drive COL resistance in manure/soil environment.

In Asia, investigators documented mcr genes/MGCB in the soil/manure ecosystem. In a recent Lebanese study, 3 mcr-1-carrying E. coli strains belonging to different clones were isolated from 41 poultry litter/faecal samples [116], further suggesting that animal (poultry) manure is a source of dissemination of COL resistance into the environment. In a study from China, 6 strains carrying mcr-1 on IncH12 plasmid and chromosome were detected among 10 colistin-resistant E. coli isolates (60%) from farming soil where there is intensive livestock farming [127], suggesting that IncH12 is a major driver of COL resistance in isolates from soil/manure ecosystems in the country. There was chromosomal integration of mcr-1 in some of the strains implying vertical transmission of mcr-1in E. coli strains from soil/manure. Thirty-five other resistance genes in 10 different antimicrobial families were harboured by the mcr-1-positive E. coli strains (Table 5), indicating that soil especially where livestock farming is practiced, is a huge reservoir of multiple resistance genes thus posing a worrisome threat to public health especially to individuals directly in contact with contaminated soil. In another Chinese study, high concentration (1~105) of mcr-1 was detected in faecal/soil samples from a chicken farm [128], further suggesting that soil constitute a huge reservoir for COLROS/mcr genes. A similar study detected a higher concentration (1.87 × 107–1.82 × 109 copies/g dry weight) of mcr-1 in 16 of 51 manure samples (1.4%) from farms in China [129], suggesting that mcr-1 is widely distributed in animal manures in the country. Encouragingly, there was significant (>90) reduction in the quantity of mcr-1 after 15 days of composting suggesting that by composting, mcr-1 can efficiently be eliminated from livestock manure thereby preventing its spread to the environment [129]. Some other Chinese investigators showed that banning the use of COL resulted in a reduced concentration of the antibiotic and mcr-1 in feed and fresh manure [130], thus suggesting that COL in manure exerts direct selection pressure for the accumulation of mcr-1 and that banning non-therapeutic COL use may curb the spread of MGCB/mcr genes. The same Chinese group also showed that anaerobic digestion of animal manure reduces the number of MGCB better than natural drying. Evidence that soil/animal manure are reservoirs for MGCB has also emerged from Africa. Very recent, 5 mcr-1-carrying and 2 mcr-3 carrying E. coli strains were detected among 28 isolates (17.9%) from soil/animal manure collected from agricultural sites/farms in Algeria [131], further indicating that animal manure is a source of spread of different mcr gene-types in Algeria. The strains were multidrug resistance phenotype and they were diverse belonging to HiR pandemic zoonotic ExPEC clones (ST405, ST10, and ST155) [70,73], thus posing a threat to public health, especially to the farmers and when water run-offs transport the pathogens to other ecosystems. This highlights the need for composting/anaerobic digestion of animal manure before disposal into the environment. There was ESBL gene in the ST405 isolates, a known clone for globally dissemination ESBL genes [55]. This clone was associated with diseases in humans and wildlife in Algeria [55,132], suggesting it has disseminated into diverse ecosystems in the country. Unfortunately, conjugation was positive in some of the strains suggesting that the strains could transfer COL resistance to organisms in various ecosystems.

5. Botanical Ecosystem

The importance of soil as a source of antimicrobial-resistant organisms contaminating plants (vegetables and fruits) is well recognized [22,23,24,25]. Plant-based foods are often consumed raw or undercooked hence potential route for dissemination of antimicrobial-resistant organisms as well as emerging pathogens thus posing a grave danger to public health. Eight publications investigated on plasmid-mediated colistin resistance in a total of 746 isolates from plant (fruits and vegetables) samples. One of the studies probed mcr-1 gene directly in the samples prior to isolation. Thirty-one isolates (28 E. coli, 2 Raulotella ornitholytica and 1 K. pneumoniae) were reported to harbour mcr-1 among the isolates tested.

Studies from European countries reported the presence of MGCB in vegetables. Two mcr-1-carrying E. coli strains of ST167 (ExPEC clone) and ST4683, were detected among 60 ESBL-producing Enterobacteriaceae (3.3%) isolated from vegetables imported into Switzerland from different countries [99], suggesting that international plant-based foods trade is a potential route for global dissemination of diverse virulent clones of mcr-1-positive E. coli [71,73], thus posing serious threat to public health. In a Portuguese study, one pathogenic mcr-1-carrying ST1716 E. coli strain was detected among 138 isolates (0.7%) from conventional fresh vegetables [133], further indicating that these vegetables are a source of COLROS in Portugal. This poses serious risks to the health of handlers and consumers especially those that consume vegetables raw or undercooked. Possible sources of the mcr-1-positive E. coli include soil, irrigation water or wildlife ejections. There were class 2 integrons, 6 prophage regions, and 8 other resistance genes in 6 different families of antimicrobials in the strains suggesting that the genes were horizontally acquired by transduction (since bacteriophages often attack organisms colonizing plants) and/or other means. It also suggested that integrons are common drivers of COL and multiple resistance in botanical ecosystems. The occurrence of MGCB in vegetables has also been reported in Asia. In a study from China, mcr-1-carrying multi-drug resistance E. coli strains were detected in 18 out of 271 vegetable samples (7%) [85], further suggesting that vegetables are reservoirs of MGCB in the country. These mcr-1-positive E. coli strains possessed similar traits, as described above. In another study from China, 9 strains (7 E. coli and 2 Raoultella ornithinolytica) carrying mcr-1 on chromosome and diverse plasmids (IncX4, IncI2 and IncHI2/ST3) of varying sizes, were detected among 270 ESBL-producing isolates (3.3%) from vegetables [134], suggesting that these plasmids are drivers of COL resistance in diverse organisms from plants and that mcr-1 could persist in plants ecosystems having integrated in chromosome of these organisms. There was extensive diversity among the mcr-1-positive E. coli isolates that were in 6 STs with some of them belonging to the ExPEC clones (ST156, ST69 and ST48) [70,73]; these isolates also possessed ISApI1 upstream of mcr-1 and harboured 5 other resistance genes in 4 different antimicrobial families indicating that diverse MGEs drive the spread of COL resistance in botanical environment just like in other ecosystems and that handlers/consumers of these vegetables could be exposed to virulent COLROS capable of causing difficult-to-treat diseases.

In a recent Chinese study, 25 mcr-1-carrying enterobacterial strains (24 E. coli and 1 Enterobacter) were isolated from 19 vegetable samples [135], further supporting that diverse species of COLROS colonize vegetables. There was no clonal restriction in the acquisition of mcr-1 among the mcr-1-positive E. coli isolates and they were extensively diverse belonging to 16 STs dominated by ST744 and ST224. Transposons and insertion sequences, as well as 14 resistance genes in 7 antimicrobial families, were present in the strains (Table 6). These findings suggest that diverse MGEs drive COL resistance in various E. coli clones in China. In another new Chinese study, 2 carbapenem/fosfomycin-resistant E. coli isolates carrying mcr-1 on ~33 kb IncX4 plasmid (in ST156 isolate) and ~60 kb IncI2 plasmid (in ST2847 isolate), were isolated from vegetables [136], suggesting that resistance against last resort antibiotics could be acquired by having contact with/consuming vegetables, thus posing serious risks to public health. However, the strains exhibited susceptibility to amikacin and tigecycline similar to E. coli strains from humans in China that also harboured both carbapenem and COL resistance genes [137], suggesting that MGCB are circulating from the human setting (through sewages/wastewaters and irrigation waters) to plant ecosystem in the country.

Fruits were reported as the source for dissemination of plasmid-mediated colistin resistance in China. An ST189 E. coli and an ST442 K. pneumoniae strain carrying mcr-1 on IncFIA and IncHI1 plasmids, respectively, were detected among isolates from marketed fruits [138], suggesting that fruits are potential reservoirs of diverse COLROS thus fruit trade is a potential route for dissemination of these organisms. There were 18 resistance genes in 9 different antimicrobial families in the strains (Table 6), suggesting they are MDR thus posing a troubling threat to public health. Worse of it is that the ST442 K. pneumoniae is a progenitor of the widely spread multidrug-R ST258 K. pneumoniae strain while the E. coli contained astA gene encoding heat-stable enterotoxin-1 associated with diarrhoea in humans [138]. It is worth mentioning that no mcr-1 or mcr-3-carrying strain was detected among 17 colistin-resistant Klebsiella isolates from vegetables/fruits in India [124]. However, chromosomal mgrB alterations was detected in some of the strains, suggesting that diverse mechanisms mediate COL resistance in environmental isolates [75,76,77].

There is no report on MGCB colonizing plants in the North and South America and Africa. Therefore, surveillance of plasmid-mediated colistin resistance in botanical ecosystem is warranted in these continents. Nevertheless, it is worth noting that no mcr-1 or mcr-2-positive strain was detected among 240 shigatoxin-producing E. coli isolates from vegetables/fruits in the US [139], and no mcr-1-carrying strain was detected among COL-resistant enterobacterial isolates from vegetables in Brazil [96].

It is evident that vegetables and fruits are reservoirs and source of spread of MGCB; therefore, proper disposal of animal excrement before use as fertilizer and improvement of irrigation water need to be taken [134].

6. Wildlife (Birds, Mammals, Reptiles, and Flies)

Because antimicrobials are not used in the wildlife, the presence of any acquired ARG in the absence of selection pressure often indicates transfer from other ecosystems [50,51,52,53,54,55]. Seventeen publications investigated on plasmid-mediated colistin resistance in a total of 1073 isolates from wildlife (birds, flies and mammals). Four of the studies probed mcr gene directly in the samples. A total of 113 isolates (67 E. coli, 4 Pseudomonas, 7 Enterobacter and 27 K. pneumoniae) harboured mcr gene. Fifty-six E. coli, 6 Enterobacter and 2 K. pneumoniae isolates harboured mcr-3 gene, respectively. The mcr-1 gene was observed in one Enterobacter, 56 E. coli, 4 Pseudomonas and 17 K. pneumoniae, respectively while mcr-8 gene was detected in 17 K. pneumoniae. Two flies were reported to carry the mcr-2 gene.

6.1. Wild Birds

Studies from Europe reported wild birds as potential reservoirs of mcr-1-positive E. coli. A mcr-1-carrying ESBL-producing strain was detected among 177 E. coli isolates (0.6%) from European herring gulls (Larus argentatus) in Lithuania [140], suggesting that these migratory birds can potentially spread genes conferring resistance against last resort antibiotics from Europe to other places, especially south where the birds move to during winter [140]. The birds could potentially spread COL resistance to water ecosystems since they migrate through the Baltic sea. In a Spanish study, an mcr-1-carrying pAmpC-producing ST162 E. coli strain was isolated from a black vulture [141], suggesting that scavenging animals could acquire COLROS from habitats contaminated by anthropogenic/animal wastes and then disseminate them to other locations. Scavenging on carrions, slaughterhouse wastes and flies potentially exposes vultures to colonization by antimicrobial-resistant organisms. The investigators detected HiR pandemic international ST131 ExPEC strains that harboured genes encoding ESBL, pAmpC and carbapenemases, indicating wide distribution of multidrug-resistant genes in the Spanish environment [141]. Therefore, wild birds can disseminate these genes encoding resistance to last-resort antibiotics to diverse ecological niches posing a serious threat to public health. It is also worth noting that no strain among 19 isolates from urban birds (yellow-legged gulls, pigeons and chickens) in France harboured mcr-1 to mcr-5 [142].

Migratory wild aquatic birds in South America were reported to harbour mcr-1-carrying E. coli. An ESBL-producing ST10/phylogroup A E. coli strain carrying mcr-1 on 33 kb IncX4 plasmid was isolated from a Magellanic penguin (Spheniscus magellanicus) suffering from pododermatitis [143]. In the strain, diverse plasmids carried 5 other resistance genes in 5 different antimicrobial families (Table 7), thus suggesting that multidrug-resistant HiR zoonotic ExPEC clones have acquired resistances from various sources and have spread widely in the environment [71,73]. Similarly, 5 ESBL-producing E. coli strains (of ST744 and ST1010) carrying mcr-1 on ~57 kb IncI2 plasmid were isolated from Kelp gulls (Larus dominicanus) in Argentina [144], suggesting that wild migratory birds could potentially disseminate diverse mcr-1-carrying E. coli clones to other parts of the globe since these gulls fly across continents [144]. There was ISApI1 upstream of mcr-1 and conjugation was at a very high frequency of ~2 × 10−6, further suggesting that diverse MGEs facilitate rapid spread of COL resistance in wild aquatic habitats.

Asian studies also reported wild migratory birds as potential carriers of mcr-1-carrying E. coli. An ESBL-producing ST354 E. coli carrying mcr-1 on 63 kb IncI2 plasmid was isolated from a long-range wild migratory waterbird Eurasian coot (Fulica atra) in Pakistan [64,145], further suggesting that IncI2 is one of the major plasmids driving the spread of mcr-1 in wildlife ecosystems. Since Eurasian coots migrate from Europe to Asia, they could potentially disseminate MGCB into water bodies and other ecosystems to and from these continents, thus posing risks to the health of aquatic animals and the public [64,145]. In a Chinese study, one mcr-1-carrying multidrug-resistant strain was detected among 6 colistin-resistant E. coli isolates (16.7%) from faecal samples collected from egret habitat [111], further supporting that aquatic wild birds disseminate COL resistance to various ecosystems. It is worth noting that another Chinese study focusing on ESBL-producing E. coli, observed 20% phenotypic COL resistance among strains from wild birds [146].

Wild birds are potential reservoirs of COLROS. These birds might have picked the MGCB by contact with contaminated water, or consumption of food material or drinking water contaminated with anthropogenic/agricultural wastes or from other ecosystems.

6.2. Wild Mammals

So far, only a study reported MGCB in a wild mammalian species. Bachiri et al. [55] isolated an mcr-1-carrying ST405/phylogroup D E. coli strain from a Barbary macaques monkeys (Macaca sylvanus) in Algeria, suggesting that HiR pandemic zoonotic ExPEC clone is circulating COL resistance in the wild. There were genes encoding ESBL and PMQR in the strain, further suggesting that wildlife potentially disseminate genes conferring resistance to last-resort antibiotics [50,51,52,53]; it also supports that ST405 E. coli clone disseminates ESBL worldwide [55]. ExPEC ST405 was associated with infertility and urinary tract infection in humans in Algeria [132,147] and it has been found in livestock and environment in the country [131], thus suggesting that this zoonotic pandemic clone is circulating in human/animal/environment ecosystems in Algiers. Direct contact with human/animal dejections, eating vegetables/fruits or drinking contaminated water is possible routes by which the monkey acquired the ST405 ExPEC clone. Conjugation was negative suggesting that mcr-1 was located on the chromosome in the strain, hence can be maintained/persist in the wild. It is worth noting that no mcr-1-carrying strain was detected among 54 E. coli isolates from rodents in China [68]. Also, no mcr-1-harbouring strain was detected among 269 faecal E. coli isolates from wild mammals/birds in the UK [148].

6.3. Flies

The role of flies as vectors for transmission of antimicrobial-resistant organisms is well recognized [57,58]. Their synanthropic nature linking human, animal and environmental ecosystems allows them to be colonized by diverse organisms and to deposit these organisms in various ecological niches [57].

Flies were reported as carriers of MGCB in Europe. An ST1011 E. coli carrying mcr-1 on IncX4 plasmid was isolated from a barn stable fly (Musca domestica) captured from a 50–150 m distance to a pig farm [65], suggesting that flies potentially transmit COLROS from one spot to another even to far distances. It also suggested that IncX4 is spreading mcr to and from diverse ecosystems. The strain harboured 9 other resistance genes in 5 different antimicrobial families on diverse plasmids (Table 8), suggesting that flies transmit multiple resistance genes together, thus posing serious public health risk. It is worth mentioning that in a retrospective German study, none among 24 ESBL-producing E. coli isolates from 21 flies were observed to harbour mcr-1-like to mcr-8-like genes [74].

Asian studies also reported flies as potential reservoirs of MGCB. Four enterobacterial strains carrying mcr-1 on IncI2 plasmid with transposons (ISEcp1 and ISSen6), were isolated from flies in China [86], further supporting that diverse MGEs are driving COL resistance in wildlife, including in flies. These strains have the capacity to rapidly transfer mcr-1 to other organisms since having transferred the gene to a recipient organism at very high frequencies of 9.0 × 10−10 to 5.0 × 10−3 cells per donor cell. Interesting, direct sample testing detected the presence of mcr-1 and carbapenem genes in various environmental niches (wild bird nests, dog kennel and poultry farms) some of which were regarded negative for the genes by isolation method, thus indicating that environmental niches constitute huge reservoirs of genes conferring resistance to last-resort antibiotics (phantom resistomes) which might be underestimated [86]. Therefore, direct sample testing followed by isolation could be the best approach for surveillance of resistance to last-resort antibiotics in environmental niches. In another Chinese study, an Enterobacter cloacae carrying mcr-1 on IncX4 plasmid and a Raulotella planticola harbouring mcr-1.3 were isolated from flies captured in poultry farms [149], suggesting that IncX4 is a common plasmid spreading mcr in the wild and that flies potentially transfer diverse organisms, including uncommon COLROS, from livestock farms to other ecosystems and vice versa. The feeding habit of flies allows them to be colonized by diverse organisms present in various ecological niches [57]. There was ISApl1 flanking mcr-1.3 in the R. planticola, further supporting that diverse MGEs are involved in the acquisition/spread of genes conferring resistance against last resort antibiotics in the environment.

In a recent Chinese study, mcr-1, mcr-2 and mcr-3 was detected in 109 flies (86 M. domestica and 23 P. terraenovae) captured in public places and near dumpsites [150], suggesting that flies potentially harbour organisms carrying mcr gene-types possibly originating from other ecosystems and that flies transfer these to human habitations posing serious public health risk such as in food, water and wound contamination. Subsequent culture of the mcr-positive samples yielded only mcr-1-containing strains (4 E. coli, 2 P. stuartii, 2 P. alcalifaciens, and one E. cloacae), further suggesting that by isolation method, some COLROS might be missed resulting in underestimation of the magnitude of COL resistance in an ecological niche. Similarly, in a Bangladeshi study, 4 mcr-3-carrying strains were detected among 40 carbapenem-resistant E. coli isolates (10%) from 60 flies (M. domestica) [107], further suggesting that mcr-3 has spread in diverse ecosystems in Bangladesh.

In a study from Thailand, mcr-1 and mcr-3 were detected in 16 flies collected from urban areas and livestock farms [151], further indicating that flies are colonized by COLROS in diverse ecological niches in the country. However, only mcr-3-carrying strains (11 E. coli, 2 Enterobacter aerogenes, 4 Enterobacter cloacae and 2 K. pneumoniae) some of which also contained ESBL and pAmpC genes were isolated from the samples. This further supports that direct sample testing before isolation is a better approach for antimicrobial resistance surveillance. In another Thai study, 48 enterobacteria (17 ST43 K. pneumoniae and 31 E. coli) carrying mcr-1 on IncX4 plasmid were isolated from 300 blowflies (Chrysomya megacephala) (16%) collected from a local market in an urban community, a rural area and a city suburb [152], suggesting that MGCB is widely distributed in human environment in Thailand and that IncX4 plasmid is a major driver of mcr-1 in strains from flies. Unfortunately, mcr-8 was also present in all the K. pneumoniae strains, suggesting that flies potentially transmit virulent K. pneumoniae clones. The ST43 K. pneumoniae has been associated with abdominal infections, bacteraemia and intensive care unit infections [152]. Equally worrisome is that the mcr-1-positive E. coli strains were extensively diversified with HiR ExPEC clone ST10 dominating among 12 STs (Table 7). More of 20 resistance genes in 9 different antimicrobial families were found in the ST10 isolates, further indicating that flies can potentially spread virulent COLROS carrying multiple resistance genes, thus portending threat to public health. Furthermore, the mcr-1-positive E. coli strains transferred COL resistance to a recipient organism at a higher frequency than by K. pneumoniae strains, suggesting that E. coli transfers mcr-1 more rapidly than Klebsiella. Moreover, of the plasmids, the IncX4 plasmid was transferred at a higher mean frequency than the other plasmids (IncHI1A, IncHI1B, and IncHI1A-IncHI1B) further supporting that IncX4 is a major driver of mcr-1. This plasmid was further proven to potentially reduce the virulence of MGCB by injecting an ~1 × 105 CFU of ST34 K. pneumoniae strains into Galleria mellonella larvae and observing significantly lowered mortality. It is worth mentioning that none of the studies on COL resistance among isolates from reptiles such as in turtles in Brazil, [96], reptiles in the US [139], and snake in Taiwan [153], detected any mcr-carrying strain. This warrants further surveillance since reptiles make contact with aquatic and terrestrial ecosystems.

7. Concluding Remarks

This review showed that diversity of MGCB such as E. coli, Enterobacter, Klebsiella, Proteus, Salmonella, Citrobacter, Pseudomonas, Acinetobacter, Kluyvera, Aeromonas, Providencia, and Raulotella have disseminated into environmental reservoirs, including contact surfaces in hospitals, public transportation routes and livestock farms, soil/manure/sludge, plants (vegetables and fruits), aquatic (aquaculture, seawater, ground and surface waters, sewage and wastewaters), and wildlife. These reservoirs are potential sources for further dissemination of mcr genes. Anthropogenic activities such as defecation in open environment/water, bathing/swimming in water bodies, improper disposal of the slaughterhouse, home, hospital and laboratory wastes, inappropriate use of antimicrobial agents in humans, animals/aquaculture and plants, are the major causes of dissemination of mcr genes into the environment.

Environmental isolates harbour mcr genes together with many virulence and resistance genes, including those conferring resistance against last resort antimicrobials. These organisms are superbugs capable of causing untreatable infections with pandemic potential. If unchecked, these organisms may diffuse into the human and animal ecosystems and present a challenge to control AMR [154]. Some environmental isolates have acquired megaplasmid with numerous ARGs (some harbour ≥10 genes). A further transmission of MGCB harbouring megaplasmid from the environment to human and animal ecosystems may result in the actualization of the O’Neill’s projection of 10 million AMR infection-associated deaths per 2050 [21]. Carbapenems and tigecycline, as well as some other commonly used antimicrobial agents, seem to be effective against most isolates in this review. The implementation of antibiotic stewardship programmes should preserve the efficacy of these last resort agents which could be used in treating cases associated with MGCB.

Drivers of plasmid-mediated colistin resistance facilitating horizontal/lateral transfer of mcr genes in the environment are diverse genetic elements, including conjugative plasmids of different replicons and incompatibility, class 1–3 integrons, transposons, complete, and truncated insertion sequences. IncHI2, IncI and IncX4 plasmids seem to be the predominant plasmid types harboured by isolates from different environmental reservoirs worldwide. The mcr-carrying environmental strains have the potential to spread worldwide since they transferred mcr gene-bearing plasmids to recipient strains at a very high frequency [15]. Nevertheless, mcr gene has integrated into chromosomal DNA and/or non-conjugative plasmids of environmental strains enabling the transfer of these genes to their progenies by vertical transmission thus ensuring the persistence of mcr genes among clonal lineages [155]. Transmission of mcr gene among environmental strains is clonally unrestricted and diverse highly virulent zoonotic pandemic and epidemic clones of E. coli and Klebsiella pneumoniae are circulating in environmental ecosystems worldwide.

Colonization of wildlife by MGCB implies that COL plasmid is maintained in bacterial populations regardless of antimicrobial selective pressure [114]. Since some mcr-1-linked plasmids like IncI2, IncX4 and IncHI2 plasmids (which are predominant in environmental isolates) could persist and increase the fitness of their host cells, MGCB in an environment such as wildlife without antibiotic pressure, may have an advantage [41,156,157].

Global production and trade of fresh plant produce and aquatic-based foods constitute potential routes of dissemination of MGCB. Integrated farms are sources of transfer of mcr genes into aquaculture which in itself have been associated with a high rate of human colonization by MGCB [28,158]. However, since livestock-fish integrated farming are considered economical and efficient farming modes in most developing countries, there is a compelling need for assessment and supervision of antimicrobial use and spread of ARGs within the aquaculture industry [28].

As demonstrated, banning of the use of COL other than therapy in livestock will curb the spread of MGCB from animal to human and environmental ecosystems. Commendably, some countries in the European Union as well as others like China, Brazil and Argentina, has taken the lead in enforcing the ban on the non-therapeutic use of COL [158].

Some of the isolates considered negative in various studies might harbour mcr gene types other than those assayed. This warrants an urgent need for affordable methods that can detect all the currently known mcr gene-types (mcr-1 to mcr-9, and the ones that are yet to emerge) for rapid and adequate surveillance of plasmid-mediated colistin resistance. Subjecting mcr-carrying isolates to high throughput analysis such as next-generation sequencing would help to provide information about the genetic context of the gene, elucidation of mcr genes that could be missed by other molecular techniques as well as the phylogenetic relationship of the isolates [114]. This information is crucial for understanding the epidemiology of COLROS and devising effective control strategies to reduce public health risks.

Since COL determinant emerging from any part of the globe can rapidly spread worldwide by international travel (even short distance travel) and food trade, there is a need for increased surveillance of mcr genes in environmental reservoirs, especially in Africa where the use of COL is largely uncontrolled, and sanitation is poor, and South America where public and environmental sanitation is also considered suboptimal [152]. Indeed, it is evident that by horizontal/lateral and vertical transfer, mcr genes (mcr-1, mcr-2, mcr-3, mcr-5, mcr-7, and mcr-8) have spread widely into diverse environmental niches (Figure 1). Thus, these ecosystems constitute underestimated vast reservoirs (‘phantom resistome’) of these mcr genes. This further underlines the need for One Health approach.