Putative Link between the Virulence-Associated fluA Gene and Fluoroquinolone Resistance in Uropathogenic Escherichia coli

Abstract

Resistance to fluoroquinolones is a major problem in the treatment of various infectious diseases, including urinary tract infections (UTI) caused by Escherichia coli. It has been recognized that fluoroquinolone-resistant E. coli strains generally harbor fewer virulence genes than their susceptible counterparts (2, 4, 5, 7, 8). We observed an increased prevalence of the virulence-associated fluA gene, which encodes the autotransporter protein Ag43a (9), in E. coli isolated in 2006 to 2008 compared to that in isolates collected prior to fluoroquinolone use (1983 to 1986). This increase was seen only in E. coli isolated from patients with urosepticemia, displaying a fluoroquinolone resistance rate of 26%, but not among isolates from patients with UTI, all of which were sensitive to these agents. Interestingly, 11 of the 12 resistant isolates carried fluA. We therefore chose to study a set of resistant and sensitive strains from patients with urosepticemia for a possible association between fluA and fluoroquinolone resistance. A total of 34 sensitive and 57 resistant E. coli isolates collected from patients with urosepticemia between 2000 and 2009 at the Karolinska University Hospital Solna, Stockholm, Sweden, were investigated. Fluoroquinolone resistance was determined according to guidelines of the Swedish Reference Group for Antibiotics (www.srga.org). Isolates sensitive to ciprofloxacin but resistant to nalidixic acid were considered resistant after confirmation of target alteration by sequencing (10). The presence of fluA was determined by PCR (9), and isolates were assigned to phylogenetic group A, B1, B2, or D by multiplex PCR (1). Prevalence data were compared by using Fisher's exact test or the chi-square test, as appropriate. Differences with P values of <0.05 were considered statistically significant. The frequency of phylogenetic groups B2 and D differed significantly between the resistant and sensitive populations (Table ​(Table1).1). While the sensitive isolates mostly belonged to group B2, the resistant ones were associated with group D. This distribution is in line with previous studies (4, 5, 7, 8). TABLE 1. Distribution of phylogenetic groups and fluA in relation to susceptibility to fluoroquinolones As indicated by our initial observation, the prevalence of fluA was significantly higher in fluoroquinolone-resistant isolates than in fluoroquinolone-sensitive isolates (42 of 57 [74%] and 16 of 34 [47%], respectively; Table ​Table1).1). Also, the prevalence of fluA differed depending on the phylogenetic origin (P < 0.001), with fluA significantly associated with group D (83% versus 45% in the remaining groups, P = 0.0002). Further analysis revealed equal distribution of fluA in sensitive and resistant isolates belonging to group D (Table ​(Table1).1). In contrast, fluA was associated with resistance in isolates of group B2. This again is in agreement with other studies suggesting a relationship between the presence or absence of certain virulence genes and antimicrobial resistance only in group B2 (3, 7, 8) but not in group D (6). In conclusion, we report here that the virulence-associated fluA gene is positively linked to fluoroquinolone resistance, particularly in E. coli belonging to phylogenetic group B2. To our knowledge, fluA has not been previously described in this connection and might provide another factor to understand the complex relationship between fluoroquinolone resistance and virulence in E. coli.