Success of Escherichia coli O25b:H4 Sequence Type 131 Clade C Associated with a Decrease in Virulence

Escherichia coli O25b:H4 sequence type 131 (ST131), which is resistant to fluoroquinolones and which is a producer of CTX-M-15, is globally one of the major extraintestinal pathogenic E. coli (ExPEC) lineages. Phylogenetic analyses showed that multidrug-resistant ST131 strains belong to clade C, which recently emerged from clade B by stepwise evolution. It has been hypothesized that features other than multidrug resistance could contribute to this dissemination since other major global ExPEC lineages (ST73 and ST95) are mostly antibiotic susceptible.

2 adults (n=7). The 21 fimH30 strains had been isolated during the same period in Spain (n=3) and in Ile de France (n=18) from the blood (n=9) and the urinary tract (n=1) of adults patients and the feces of health adults (n=8) and children (n=3). Four of the 18 fimH22 strains were resistant to nalidixic acid whereas 17 of the 21 fimH30 strains were resistant to both nalidixic acid and ciprofloxacin according to the agar disk diffusion method interpreted following the EUCASTrecommendations (www.eucast.org). Three of the 18 fimH22 strains and six of the 21 fimH30 strains were extended-spectrum β-lactamase (ESBL) producers according to the double disk synergy test (1). The 39 strains were stored in 20% glycerol at -80°C until use. E. coli CFT073 and E. coli K-12 MG1655 strains were used as positive and negative controls, respectively, in the sepsis mouse model, and E. coli UTI89 and E. coli UBA83972 strains as positive and negative controls, respectively, in the yeast agglutination assays.

Genome sequencing and analysis
Whole genome sequencing (WGS) of the 39 strains was performed. Total DNA was extracted using NucleoMag Tissue (Macherey-Nagel, Düren, Germany). Libraries were prepared using Nextera technology and sequenced with an Illumina HiSeq (IntegraGen Genomics, Evry, France) using a 2x100 base pair (bp) paired-end strategy. Reads were assembled with SPAdes (2). We investigated how the different subclades B and C of ST131 was represented in our collection and how they evolved in frequency over time in a larger collection from established by Kallonen et al. (3). To that end, we complemented the studied 39 genomes with 218+21 genomes from two published studies [218 from Kallonen et al. (3), 21 from Ben Zakour et al. (4)]. We annotated the genomes using Prokka (5), identified the core genome using Roary (6), and inferred the phylogenetic history of all these strains after removing the identified recombination using Gubbins (7) and the phylogenetic software RAxML (8). Outlier sequences were removed. The tree was then viewed and modified using iTOL (9). From this phylogeny, we identified sequences belonging to different subclades B0-B5 and C0-C2 of O25b:H4 ST131, as defined in Ben Zakour et al. (4).

Gene deletion and complementation
Primers and plasmids used for gene deletion and complementation are listed in Table S3  MT Divergent was put in competition with its ΔfimB::kan, ΔibeA::kan and ΔibeART::kan mutants in order to be able to assess the number of CFU/ml for each strain. In the double MT Divergent ΔfimB::FRT ΔibeART::kan mutant, the kanamycin cassette was removed from the ΔfimB gene but conserved in the ΔibeART operon. The same process was used for Recombinant when it was put in competition with its ΔfimB::kan mutant.

Kinetics of early biofilm formation
The primary step of biofilm formation was measured in the 39 strains by using BioFilm Ring Test ® (BioFilm Control, Saint-Beauzire, France) according to the manufacturer's recommendations and as previously described (11). Briefly, each strain was sub-cultured twice on brain heart infusion (BHI) agar (Benton Dickinson, Le Pont-de-Claix, France) at 37°C for 24 h. Three colonies of the second subculture were suspended in BHI broth (Biofilm Control) and density at 600 nm (OD600) was measured (Ultrospec10: Biochrom, Cambridge, UK). An adequate volume of this suspension was then added to a mixture (1% vol/vol) of experiments were repeated at least three times for each strain. BHI broth was used as negative control, and strains S250 and 39, previously described with this method as early and never biofilm producers, respectively, as control strains (12).

Expression of type 1 fimbriae
Expression of type 1 fimbriae was assessed in the 39 strains and mutants by using the yeast cell (Saccharomyces cerevisiae) agglutination assay as previously described (13), Briefly, for each strain, five single colonies were inoculated into 3 ml of LB and incubated for 24 h in shaking and static conditions. Agglutination was performed by mixing 50 µL of yeast cells (diluted to 5% in phosphate buffer saline (PBS) and 50 µL of bacterial culture on a glass slide.
To study the early expression of type 1 fimbriae, the same protocol was applied to cultures incubated for 2 and 5 h in shaking conditions. To this end, 50 µL of yeast cells were mixed with the 50 µL-pellet (obtained after centrifugation: 3000g for10 min at 4°) of the 2 h-shaking LB culture and 50 µL of the 5 h-shaking LB culture. Bacterial-yeast aggregation was macroscopically observed and a bacterial strain was considered negative for type 1 fimbriae expression when no visible aggregation was produced after 3 min.

Maximum growth rate
Fitness assay was performed for the 39 strains as previously described in LB (14). The plate was incubated in an automatic spectrophotometer (Tecan Infinite F200 Pro) that measures the OD600 in each well every 5 min over a period of 24 h. The experiment was repeated three times. Growth curves were then analyzed and maximum growth rates was calculated and expressed in h -1 . We decided to explore only the maximum growth rate, because contrary to the lag time (i.e. time required to reach the exponential growth phase) and 6 the final optical density, the maximum growth rate does not depend on the initial number of bacteria and on the well position, and is less affected by the form of the growth curve (15).

Mouse models
Mono infection assay and competition assay in the sepsis model Female mice OF1 of 14-16 g (4-week-old) from Charles River ® were used to assess the individual strain ability and, in competition assays, the strains' relative ability to cause sepsis, as previously described (16). Briefly, 200 µl of a suspension of 10 9 E. coli/mL in physiological serum was inoculated by subcutaneous injection in the neck, either alone or mixed at a ratio of 1:1. Time to death was monitored during the following seven days, and all mice were frozen at -20°C at time of death. Mice surviving after 7 days were considered cured and were sacrificed. From ten to twenty mice were used for the strains tested alone and from five to ten for the strains tested in competitions. CFT073 strain, known to quickly kill any inoculated mouse, was used as positive control, and E. coli K-12 MG1655 strain, unable to kill any inoculated mouse as negative control (17)

Competition assay in the intestinal colonization model
Six-week-old female mice CD-1 from Charles River® (L'Arbresle, France) pre-treated with streptomycin before inoculation of challenging strains were used to assess strains' relative ability to colonize the mouse intestine, as previously described (14). Briefly, mice received water supplemented with streptomycin sulfate (5 g/liter) for five days in order to eliminate E. coli from gut before studied strain inoculation, and were fed from then with sterile food until the end of the experiment. Streptomycin treatment was stopped five days before studied strain inoculation to allow the subsequent colonization of the mouse intestine by streptomycinsensitive strains. At inoculation day, about 10 6 E. coli mixed at a 1:1 ratio were administrated by force-feeding in 200 µL of physiological water to mice free of coliform flora (controlled by plating the feces on Drigalski plates). At days 1, 4 and 7, the intestinal population of E. coli was estimated by plating dilutions of weighed fresh feces on LB agar and LB agar with appropriate antibiotics (kanamycin 50 mg/L or ciprofloxacin 1 mg/L). At least five mice were used for each competition assay. CIs were obtained as described above Competition assay in the urinary tract infection model CBA female mice of 8-22g (8-week-old) from Janvier ® (Le Genest-Saint-Isle, France) were used to assess strains' relative ability to cause an ascending unobstructed urinary tract infection, as previously described (18). Briefly, following a 12 h hydric restriction, anesthetized mice (using a xylazine/ketamine mixture) were infected with about 10 8 E. coli mixed at a 1:1 ratio in 50 µL of physiological water, via the transurethral route into the bladder. Once the anesthetic effect achieved, hydric restriction was stopped. Mice were sacrificed 48 h after the bacterial inoculation. The bladder and the kidneys were aseptically collected, weighed and pounded in 1 mL of physiological water. The populations of E. coli were estimated by plating dilutions of the bladder and kidneys suspensions on LB agar and/or LB agar with appropriate antibiotics (ciprofloxacin 1 mg/L or ampicillin 100 mg/L). Ten mice were used for each competition assay. Mice were considered infected if at least one colony grew on the kidneys' LB agar plates, which corresponds to approximately 60-70 CFU/g of the kidneys. CIs were obtained as described above. In the absence of colony on LB agar with antibiotics, CI was