ABSTRACT
Testing 1,666 fecal or intestinal samples from healthy and diarrheic pigs, we obtained hemolytic Escherichia coli isolates from 593 samples. Focusing on hemolytic E. coli isolates without virulence-associated genes (VAGs) typical for enteropathogens, we found that such isolates carried a broad variety of VAGs typical for extraintestinal pathogenic E. coli.
TEXT
Hemolytic Escherichia coli strains are common in the intestine of clinically healthy and diseased pigs (2, 9–12) and often do not possess virulence-associated genes typical of intestinal pathogenic isolates (iVAGs) like est-1a, est-2, eltB-Ip, and faeG, which are frequently observed among enterotoxigenic E. coli, or stx2e and fedA, which are characteristic of edema disease-causing E. coli (9, 10). As such isolates were not the main focus of recent clinical diagnosis, knowledge of their virulence gene profile is poorly defined. Based on our initial observations (10), we assumed that such E. coli strains harbor high numbers of VAGs typical for extraintestinal pathogenic E. coli (ExPEC; eVAGs), which cause urinary tract infections, septicemia, and meningitis in both humans and animals (5). The present study assessed the frequency of hemolytic E. coli isolates lacking iVAGs in swine and provides a first insight into virulence gene features as an initial step toward understanding this E. coli group.
As depicted in Fig. 1, E. coli cells were obtained from feces or intestinal contents of clinically healthy and diarrheic pigs. Alpha-hemolysis of E. coli was tested as previously described (8). Hemolytic E. coli isolates from clinically healthy pigs were obtained from 132 pigs on 24 farms between 2002 and 2009 (for details, see Table S1 in the supplemental material). During a longitudinal study, hemolytic E. coli isolates from healthy pigs were studied over 19 months at another farm. Here, 127 randomly chosen samples from 73 pigs were collected. Samples of intestinal E. coli isolates from diseased pigs with enteritis or edema disease were obtained between 2007 and 2009 during routine laboratory diagnosis (Institut f�r Hygiene und Infektionskrankheiten der Tiere, Justus-Liebig-Universit�t, Giessen, Germany). The pigs (n = 1,407) were from 530 farms with unknown production status.
Flowchart of sampling and characterization of E. coli isolates in this study. Clinically healthy pigs were from production units in Berlin, Brandenburg, Lower Saxony, Saxony, Thuringia, and Schleswig-Holstein, Germany. Diseased pigs were from production units in North Rhine-Westphalia, Hesse, Schleswig-Holstein, Bavaria, Lower Saxony, and Baden-Wuerttemberg, Germany, and Styria and Lower Austria, Austria. Asterisks indicate that the remaining 59 (out of 121) hemolytic E. coli isolates were not available as cultures and therefore could not be included in further analysis.
Hemolytic isolates were tested for the presence of iVAGs, including faeG, fanA, fasA, fedA, and fimF41a (coding for subunits of F4, F5, F6, F18, and F41 fimbriae, respectively) and stx2e, est-1a, est-2, eltB-Ip (coding for toxins) by PCR as previously described (1, 6). All isolates carrying at least one of these iVAGs were excluded from further investigations. In cases in which more than one hemolytic E. coli isolate was isolated from a single farm, duplicate isolates, as determined by macrorestriction analysis (pulsed-field gel electrophoresis [PFGE]) (8), were excluded from further analyses (Fig. 1). Nonduplicate hemolytic E. coli isolates without iVAGs were then tested for eVAGs (Table 1) and classified into phylogenetic groups according to the EcoR system by PCRs previously described (3, 5, 10). Statistical analyses for comparison of the occurrence of genes between isolates from animal groups were performed using two-tailed Fisher's exact test. Analysis of the occurrence of eVAGs according to an EcoR group used one-way analysis of variance (ANOVA) with a Bonferroni post hoc test.
Occurrence of eVAGs in porcine intestinal hemolytic E. coli and comparison to data from an already published study on porcine intestinal nonhemolytic E. colia,b
Hemolytic E. coli isolates were isolated from 62 clinically healthy pigs (47.0% of all samples). These 62 isolates (1 isolate per pig) belonged to 37 PFGE types, of which 32 did not carry iVAGs. In the longitudinal study, hemolytic E. coli isolates were isolated from 47 randomly chosen intestinal samples (37% of all samples). These 47 hemolytic isolates (1 isolate per sample) were divided into eight PFGE types, 5 of which did not carry iVAGs. One PFGE type with a unique and stable PFGE pattern and without iVAGs was detected over the 19-month sampling period in 20 different animals. Hemolytic E. coli isolates were isolated from 484 samples from diseased pigs (34.4% of all samples). A total of 121 isolates did not carry iVAGs. A total of 62 randomly chosen isolates—each from a different farm—were included in further analysis (Fig. 1).
The 37 isolates with a unique PFGE type from clinically healthy pigs, including the five isolates from the longitudinal study and 62 isolates from diseased pigs, were tested for eVAGs (Table 1; Fig. 2). All isolates carried the gene ompA. Two genes that were absent from all isolates were iha and sat. Only tsh and traT were found significantly more often in isolates from diseased piglets (P < 0.05). Genes hra, chuA, fyuA, iroN, irp2, astA, and ibeA were present significantly more often in isolates from healthy piglets (P < 0.05).
Presence or absences of eVAGs in 99 porcine intestinal hemolytic E. coli isolates in this study. n, number of isolates per pattern. Solid squares indicate the presence of the respective gene, and open squares indicate its absence.
Hemolytic isolates belonged to EcoR groups A (59 isolates), B1 (14 isolates), B2 (18 isolates), and D (8 isolates). Isolates belonging to EcoR group B2 carried significantly more eVAGs (median, 18; minimum, 15; maximum, 22) than group D (median, 11; minimum, 7; maximum, 16), B1 (median, 9.5; minimum, 6; maximum, 18), and A (median, 7; minimum, 5; maximum, 15) members (P < 0.001) (see Table S2 in the supplemental material). Genes sit (chromosomal), neuC, ibeA, gimB, and tia were exclusively found in EcoR group B2 isolates. Almost all EcoR group B2 members carried genes hra, papC, fyuA, malX, iroN, kpsMTII, sfa/focCD, irp2, sit (chromosomal), traT, astA, vat, hlyA, and cnf1/2 (Table S2).
As hypothesized, a considerable proportion of porcine intestinal hemolytic E. coli isolates lacking iVAGs carried a broad variety of eVAGs. We identified a number of eVAGs that have not been reported for porcine E. coli isolates so far. Among these were genes coding for adhesins (hra, tsh, and mat), iron acquisition (ireA), protectins (neuC), and an invasion-related factor (gimB). Genes iha and sat, which are common in extraintestinal pathogenic E. coli (ExPEC) isolates were not detected, and only iha was previously reported to be present in intestinal E. coli from pigs (2, 13). In conclusion, almost all eVAGs can also be found in porcine intestinal E. coli.
A comparison between porcine intestinal hemolytic isolates without iVAGs of the present study with previously published intestinal porcine nonhemolytic isolates (10) showed that a broad variety of eVAGs were present in both groups (Table 1), but hemolytic isolates carried higher numbers of such VAGs. In line with other studies, E. coli isolates affiliated with EcoR group B2 carried more eVAGs, in particular iron acquisition genes, than EcoR A, B1, or D isolates (5, 7).
In general, we detected more hemolytic isolates without iVAGs (82.2%) than with iVAGs (17.8%) in clinically healthy pigs. In contrast, a lower number of hemolytic isolates without iVAGs (25.0%) than with iVAGs (75.0%) were isolated from diarrheic pigs. Further studies should prove the role of hemolytic isolates without iVAGs in healthy and diseased pigs.
ACKNOWLEDGMENTS
This work was supported by grants InnoProfile 03 IP 611 and FUGATO “E. coli chick,” both funded by the Bundesministerium f�r Bildung und Forschung (Germany).
FOOTNOTES
- Received 27 April 2011.
- Accepted 24 September 2011.
- Accepted manuscript posted online 30 September 2011.
↵† Supplemental material for this article may be found at http://dx.doi.org/10.1128/AEM.05289-11.
- Copyright � 2011, American Society for Microbiology. All Rights Reserved.