Emergence of mcr-9.1 in ESBL-producing Clinical Enterobacteriaceae in Pretoria, South Africa: Global Evolutionary Phylogenomics, Resistome and Mobilome.

Background. Extended-spectrum {beta}-lactamase (ESBL)-producing Enterobacteriaceae are critical-priority pathogens that cause substantial fatalities. With the emergence of mobile mcr genes mediating resistance to colistin in Enterobacteriaceae, clinicians are now left with little therapeutic options. Methods. Eleven clinical Enterobacteriaceae strains with resistance to cephems and/or colistin were genomically analysed to determine their resistome, mobilome, and evolutionary relationship to global strains. The global phylogenomics of mcr-9.1-bearing genomes were further analysed. Results & conclusion. Ten isolates were ESBL positive. The isolates were multidrug-resistant and phylogenetically related to global clones, but distant from local strains. Multiple resistance genes, including blaCTX-M-15 blaTEM-1 and mcr-9.1 were found in single isolates; ISEc9, IS19, and Tn3 transposons bracketed blaCTX-M-15 and blaTEM-1. Common plasmid types included IncF, IncH and ColRNAI. Genomes bearing mcr-9.1 clustered into six main phyletic groups (A-F), with those of this study belonging to clade B. Enterobacter sp. and Salmonella sp. are the main hosts of mcr-9.1 globally, albeit diverse promiscuous plasmids disseminate mcr-9.1 across different bacterial species. Emergence of mcr-9.1 in ESBL-producing Enterobacteriaceae in South Africa is worrying due to the restricted therapeutic options. Intensive One Health molecular surveillance might discover other mcr alleles and inform infection management and antibiotic choices.


Introduction 48
Enterobacteriaceae producing extended-spectrum β -lactamases (ESBLs) are categorised as 49 critical priority 1 pathogens requiring urgent attention with regards to the development of 50 8 distance of those genomes from each other in the distance trees ( Fig. 2B-2C). The 176 evolutionary epidemiology of the mcr-9.1-bearing plasmids or chromosomes is further shown 177 by Fig. 2B-2C, with different plasmid types and chromosomes mediating the spread of this 178 gene within and between species throughout the world. The various mcr-9.1-bearing genomes 179 (plasmids or chromosomes) clustered phylogenetically into six, herein labelled as A-F, with 180 Enterobacter sp. 18A13 plasmid pECC18A13-1 DNA from Japan seeming to be the earliest 181 ancestor and Enterobacter hormaechei strain S13 plasmid pSHV12-1301491 from the USA 182 seeming to be the most recent. All the mcr-9.1 contigs in this study fell within the B 183 phylogenetic cluster, although the genomes that aligned most closely with these contigs 184 changed their phylogenetic clustering between K006/K130 and K063 ( Fig. 2B-2C). 185 186 Several integrons bearing gene cassettes of antibiotic resistance genes were identified in 187 various contigs in the isolates (Table 3), with none being detected in the C. freundii strains. 188 PM005 (Providencia alcalifaciens) had a unique integron, In27, whilst K001 (K. variicola) 189 and four E. hormaechei strains harboured In191. K006, K063 and K130, originating from 190 different patients and having the mcr-9.1 gene, harboured the same integrons and very similar 191 gene cassettes. EC009 and Ec010 had the same class 1 integron and gene cassettes and 192 belonged to the same clone (Table 3). Most of the isolates had multiple plasmid replicons, 193 with ColRNAI being the commonest; ColRNAI was also found in the mcr-9.1 strains. The 194 isolates were further grouped using their pMLST (plasmid multi-locus sequence typing). 195 Except for CF004 and PM005, which had no pMLST but contained an A/C 2 replicon, 196 IncHI2  and IncF subtypes were the main plasmid types. Notably, all the mcr-9.1 197 strains had an IncHI2[ST-1] plasmid whilst two, K006 and K063, plasmids. Although the resistance genes within strains having the same plasmid types were 199 . CC-BY-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

247
The closest evolutionary relative of K001 was a single strain, WUSM_KV_44 [bla LEN-13 , 248 fosA, oqxA, oqxB] from the USA ( Figure 5) whilst PM005, which was initially identified as 249 . CC-BY-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint .

Discussion 253
We herein report on the first emergence of mcr-9.1 in both South Africa and Africa, in three 254 E. hormaechei strains with a rich repertoire of resistomes and mobilomes. The mcr-9.1 strains 255 were collected among clinical Enterobacteriaceae strains during a molecular surveillance 256 procedure to identify ESBL producers in a referral laboratory. It is a worrying observation to 257 have all the strains, sourced from different patients and wards within the same hospital, being 258 multi-drug resistant. Although the resistomes largely reflected the resistance phenotype, 259 resistance to antibiotics were observed without an underlying resistance gene being 260 identified. Specifically, the C. freundii strains expressed resistance to several antibiotics 261 although not more than three resistance genes were found in them. This could suggest the 262 presence of an unknown resistance determinant or the use of active efflux. Furthermore, 263 sensitivity to certain antibiotics such as colistin was not corroborated by the presence of mcr-264 9.1. Sensitivity to colistin in strains having the mcr gene has been reported previously and 265 chromosomal mutations have been shown to exert a higher resistance MIC than mcr genes 266 3,[25][26][27] . 267 Only a single strain with the mcr gene was colistin resistant and the identified mutations in 268 mgrB and pmrAB were also found in susceptible strains. Hence, the identified mutations 269 cannot be responsible for colistin resistance. Consequently, other factors identified for 270 colistin resistance or other unknown determinants could be responsible for the observed 271 colistin resistance 3,7,11,28 . The three strains bearing the mcr-9.1 gene were closely related, but 272 the contigs bearing the mcr-9.1 genes were not of the same sequence homology 273 (Supplemental data 2), with the mcr-9.1-contigs' distance trees showing that K006 and K130 274 . CC-BY-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint .

12
were closer (Fig. 2B). It is worth noting that K006 and K130 were phylogenetically distant, 275 but their mcr-9.1-contigs' were of closer nucleotide identity than K006 and K063, which 276 were of the same clone (Fig. 4). This observation, plus the close alignment of the mcr-9.1-277 contigs' with plasmid genomes on Genbank (Fig. 2B), strongly suggest that the mcr-9.1 gene 278 could have been horizontally, instead of vertically, acquired. 279

280
The global evolutionary trajectory of the mcr-9 gene and genomes shows six major 281 phylogenetic clades, herein labelled A to F. The strains and plasmids in clade A seem to be 282 the earliest ancestors of this gene with F being the last and most recent. Moreover, 283 Enterobacter spp., particularly E. hormaechei, and Salmonella enterica plasmids are the 284 commonest hosts of the mcr-9 gene (Fig. 2B). This is not surprising as mcr-9.1 was first 285 identified in most Salmonella Typhimurium 25 . The diversity of species and plasmids 286 involved in the dissemination of this gene explains its promiscuity and rapid spread around 287 the globe, further corroborating the need to restrict the use of colistin in both veterinary and 288 human medicine 5,29 . 289

290
The genetic support of the various resistance genes, particularly bla CTX-M-15 , bla TEM-1 , bla OXA 291 and mcr-9.1 identified in these isolates are not new. In particular, the co-occurrence of bla  and bla TEM-1 within Tn3 composite transposons, ISEc9 and IS19 on IncF type plasmids is 293 widely reported in both South Africa and worldwide 4,[20][21][22]30,31 . Further, the presence of the 294 mcr-9.1 gene on IncHI2 plasmids as well as the presence of a cupin fold metalloprotein, as 295 observed here, are common around mcr-9.1 genes 25 . The promiscuity of IncF plasmids and 296 the abundance of integrons, transposons and ISs in the genomes of these strains is worrying 297 as they might mobilise and facilitate the faster dissemination of these multiple resistance 298 genes to other species and clones 4,31,32 . 299 13 It is interesting to note that not all the closely related strains around the mcr-9.1 strains 301 harboured the mcr gene. Some of the global strains with very close phyletic clustering with 302 the non-mcr-positive strains contained mcr, and in some cases, carbapenemase (bla NDM , 303 bla OXA-48 , bla VIM and bla IMP ) genes. Such resistome differences between strains belonging to 304 the same clade suggest the gain and loss of resistance plasmids during the evolutionary and 305 epidemiological trajectory. This could be the case for the other E. hormaechei strains that had 306 no mcr gene. The 11 strains included in this study were phylogenetically distant from any 307 strain from South Africa or Africa, but of very close evolutionary distance to international 308 strains, some of which harboured very rich resistomes, including the co-occurrence of 309 carbapenemases and mcr genes. Thus, the possibility of these strains having been imported 310 cannot be ruled out, a situation that warrants constant surveillance and screening of medical 311 tourists, particularly those from carbapenemase-and mcr-endemic areas 6,33 . 312 313 Although carbapenemases were not found in these strains as have been reported elsewhere in 314 mcr-positive strains 34-38 , the co-occurrence of mcr and ESBL genes is worrying as they 315 restrict therapeutic options. To date, only mcr-1 genes have been reported in South Africa 316 6,27,39,40 , making the emergence of mcr-9.1 in strains as old as 2013 a worrying situation. This 317 suggests that other mcr variants could be present in South Africa and Africa, and intensive 318 clinical surveillance is necessary to unearth these and pre-empt further escalation. 319 320 Acknowledgements: none 321

Funding: none 322
Transparency declaration: authors declare no conflict of interest 323 . CC-BY-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint .
is the (which was not peer-reviewed) The copyright holder for this preprint . CC-BY-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint . . CC-BY-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint . CC-BY-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint .  It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the

(which was not peer-reviewed)
The copyright holder for this preprint

Table 2. Point mutations on the colistin chromosomal resistance genes of the E. hormaechei isolates from South Africa
*Reference E. hormaechei genome used was Enterobacter hormaechei strain:C15117 (PRJNA494598) .

CC-BY-ND 4.0 International license
It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the

(which was not peer-reviewed)
The copyright holder for this preprint  Table 3 , a r r 7

. Gene cassettes and integrons found in the isolates
, a a c 3 (  2  5  8  9  -3  3  9  8  ) , a r r 7 It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the

(which was not peer-reviewed)
The copyright holder for this preprint It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the

(which was not peer-reviewed)
The copyright holder for this preprint

CC-BY-ND 4.0 International license
It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the

(which was not peer-reviewed)
The copyright holder for this preprint   .

CC-BY-ND 4.0 International license
It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the

(which was not peer-reviewed)
The copyright holder for this preprint    .

CC-BY-ND 4.0 International license
It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the

(which was not peer-reviewed)
The copyright holder for this preprint  Figure S1. The evolutionary epidemiology of mcr-9.1 genes contained on different plasmids and genomes worldwide. A PDF version of Figure   2A that can be zoomed to clearly see the plasmids and genomes bearing the mcr-9.1 genes are shown for K006 (A), K063 (B) and K130 (C), which are coloured or highlighted as yellow.
Supplemental Table S1. Antimicrobial sensitivity results and resistome of the isolates. The sensitivity of the isolates to the various antibiotics tested using the MicroScan is shown, with those coloured as green being resistant according to the CLSI breakpoints. Those coloured as blue are resistant according to the EUCAST breakpoints. Those not coloured are susceptible. The various antibiotic classes to which the antibiotic agents belong are shown above each antibiotic in unique colours and the resistance genes per isolate are shown in the last column.
Supplemental dataset 1. General data of demographic, phenotypic and genomic results used for this study.

CC-BY-ND 4.0 International license
It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the

(which was not peer-reviewed)
The copyright holder for this preprint  x i a n g f a n g e n s i s s t r a i n 0 2 0 0 3 8 p l a s m i d p C T X M 9 _ 0 2 0 0 3 8 p h o s p h o e t h a n o l a m i n e -l i p i d A t r a n s f e r a s e M C R -9 . 2 ( m c r -9 ) g e n e , m c r -9 .   ia e s tr a in M 1 7 2 7 7 p la s m id p 1 7 2 7 7 A _ 4 7 7 , c o m p le te s e q u e n c e C it ro b ac te r fr eu n d ii st ra in 5 2 5 0 1 1 p la sm id p 5 2 5 0 1 1 -H I2 , co m p le te se q u en ce Cr on ob ac ter sa ka za ki i str ain GZ cs f-1 pl as m id pG W 1, co m pl ete se qu en ce Klebs iella aerog enes strain KA_P 10_L 5_03. 19 plasm id pIMP IncH1 2_334 kb, comp lete seque nce NODE_ 53_leng th_2661 _cov_38 .7403, whole genome shotgun sequenc e En ter ob ac ter clo ac ae str ain RJ 70 2 pla sm id pIM P2 6, co mp let e seq ue nc e En te ro ba ct er ho rm ae ch ei st ra in C 15 11 7 pl as m id pS PR C -E ch o1 , co m pl et e se qu en ce E nt er ob ac te r ho rm ae ch ei su bs p. xi an gf an ge ns is st ra in O S U V M C K P C 4-2 pl as m id pO S U E C _D , co m pl et e se qu en ce x i a n g f a n g e n s i s s t r a i n 0 2 0 0 3 8 p l a s m i d p C T X M 9 _ 0 2 0 0 3 8 p h o s p h o e t h a n o l a m i n e -l i p i d A t r a n s f e r a s e M C R -9 . 2 ( m c r -9 ) g e n e , m c r -9 .   En ter ob ac ter ho rm ae ch ei str ain S1 3 pla sm id pS HV 12 -13 01 49 1, co mp let e seq ue nc e E nt er ob ac te r ho rm ae ch ei st ra in S 6 pl as m id pI nc H I2 -1 50 22 64 , co m pl et e se qu en ce E s c h e r i c h i a c o l i s t r a i n 6 8 A p l a s m i d p h o s p h o e t h a n o l a m i n e t r a n s f e r a s e M C R -9 . 1 ( m c r -9 ) g e n e , m c r -9 . 1 a l l e l e , c o m p l e t e c d s E n te r o b a c te r s p . E 2 0 , c o m p le te g e n o m e E n t e r o b a c t e r h o r m a e c h e i s u b s p . x i a n g f a n g e n s i s s t r a i n 0 2 0 0 3 8 p l a s m i d p C T X M 9 _ 0 2 0 0 3 8 p h o s p h o e t h a n o l a m i n e -l i p i d A t r a n s f e r a s e M C R -9 . 2 ( m c r -9 ) g e n e , m c r -9 .   author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10. 1101/2019