Porcine aminopeptidase N binds to F4+ enterotoxigenic Escherichia coli fimbriae

F4+ enterotoxigenic Escherichia coli (ETEC) strains cause diarrheal disease in neonatal and post-weaned piglets. Several different host receptors for F4 fimbriae have been described, with porcine aminopeptidase N (APN) reported most recently. The FaeG subunit is essential for the binding of the three F4 variants to host cells. Here we show in both yeast two-hybrid and pulldown assays that APN binds directly to FaeG, the major subunit of F4 fimbriae, from three serotypes of F4+ ETEC. Modulating APN gene expression in IPEC-J2 cells affected ETEC adherence. Antibodies raised against APN or F4 fimbriae both reduced ETEC adherence. Thus, APN mediates the attachment of F4+E. coli to intestinal epithelial cells.

F4 fimbriae are important ETEC virulence factors and exist as three antigenic variants, namely F4ab, F4ac, and F4ad [5]. These three F4 fimbriae are similar, but differ in the faeG gene, which encodes the major fimbrial subunit, resulting in different adhesive properties and specificities in attachment to the small intestine [6,7]. Strains in which faeG is deleted exhibit significantly reduced adherence to host cells [8]. Oral administration of F4 fimbriae or FaeG induces a protective mucosal immune response in F4 receptor positive piglets and FaeG mediates ETEC binding to host cells [4,6,7]. It seems likely that the major FaeG subunit is not only an essential component of F4 fimbriae but also directly mediates the binding of F4 + E. coli [9].
Various potential host receptors for F4 fimbriae have been described, including MUC4, MUC13, MUC20, ITGB5, and TFRC [10][11][12][13]. The polymorphic XbaI restriction enzyme site in intron 7 of the muc4 gene has been used as a biomarker to classify an important percentage of piglets as susceptible or resistant to F4 + ETEC infections [14][15][16]. Mucin 4 polymorphisms and their candidate glycoprotein receptors are highly associated with the MUC4-susceptible genotype [17]. However, MUC4 genotypes are not completely associated with F4 ETEC susceptibility and there are likely to be other F4 receptors [18,19]. Recently, porcine aminopeptidase N (APN) was reported to serve as a receptor protein for F4ac + ETEC [20]. APN, also known as ANPEP and PEPN, is a Zn 2+ membrane-bound exopeptidase that is highly expressed on the intestinal mucosa [21]. APN can promote intestinal epithelial cell endocytosis in F4Rs piglets and is involved in the induction of mucosal immunity [20]. Here we desired to characterize the interaction between APN and FaeG, to investigate whether modulating APN expression in IPEC-J2 cells could affect ETEC adherence, and to determine whether APN is directly involved in the adherence of F4 + ETEC to host cells.

Protein-protein interaction assays
Agglutination assays were conducted as described previously [28]. F4 + E. coli were cultured overnight at 37 °C, diluted with two volumes of PBS after centrifugation, and washed twice with PBS. Bacterial suspensions (10 µL) were applied to glass slides and mixed with APN protein.
Visible agglutination within 2 min incubation was considered as positive.
To investigate the role of glycans in the APN-FaeG interaction, in some cases, PDVF membranes were treated with 0-20 mM NaIO 4 (Sigma-Aldrich) in 50 mM sodium acetate, pH 4.5, at 37 °C in the dark for 30 min-2 h [20,31,32]. Membranes were thoroughly with TBST, blocked with a 2% BSA, and then used in Western blotting experiments as described above.

Western blotting
Proteins were harvested in RIPA buffer with PMSF and incubated overnight with polyclonal antibodies against APN. Blots were developed using enhanced chemiluminescence (ECL) (Pierce) reagents.

Adhesion and inhibition assays
In vitro adhesion assays were performed as previously described [25,38]. Bacteria (1 × 10 7 CFUs) were added to a monolayer of about 1 × 10 5 cells in each well of a 96-well culture plate (Corning, NY, USA) for 1 h at 37 °C (6% CO 2 ). Cell monolayer were washed gently three times with PBS and then 0.5% Triton X-100 was added for 20 min. Lysates were serially diluted and spread on LB agar to enumerate adherent bacteria. The experiments were repeated three times.
Both the anti-F4 fimbriae monoclonal antibody and the anti-APN polyclonal antiserum were used for in vitro inhibition assays. Anti-APN polyclonal antiserum at 1:1, 1:10, and 1:100 dilutions was co-incubated with a monolayer of about 1 × 10 5 IPEC-J2 cells in each well of a 96-well culture plate for 2 h at 37 °C before adding bacteria. The monoclonal antiserum against F4 fimbriae (1:100 dilution) was co-incubated with bacterial suspensions for 30 min at 37 °C (6% CO 2 ) with gentle agitation prior to their addition onto the IPEC-J2 cell monolayer. Lysates were serially diluted and spread on LB agar to enumerate adherent bacteria. The experiments were repeated three times.

Statistical analyses
All analyses were performed using SPSS 16.0 software (SPSS Inc., USA) using t tests. A p value of less than 0.05 was considered statistically significant.

APN interacts with both F4 + fimbriae and with FaeG
We first used agglutination assays to test for interactions between APN and F4 + E. coli. Recombinant strains expressing F4 fimbriae had the strongest agglutination, while ΔfaeG mutants exhibited weak agglutination with APN. Compared with F4ad bacteria, the groups of F4ab and F4ac have a more visible reaction but the difference among three serotypes are not significant (Table 1). Both yeast two-hybrid and pulldown assays were used to determine whether the APN protein binds directly to FaeG. The positive β-galactosidase activities from yeast two-hybrid experiments showed that APN interacted with FaeG when co-expressed in yeast ( Figure 1A) and the pulldown results with purified APN and FaeG also demonstrated that APN binds directly to FaeG in vitro ( Figure 1B). Treating PDVF membranes to which APN/ FaeG pulldown samples had been transferred with metaperiodate (NaIO 4 ) did not have significant impact to the results of the APN-FaeG pulldown ( Figure 1C).

F4 + binding to IPEC-J2 cells differing in APN expression
To evaluate the potential involvement of APN as an F4 + E. coli receptor, we knocked down APN expression in IPEC-J2 cells using pcDNA ™ 6.2-GW/miR-APN. We observed a substantial reduction in APN expression as determined by using RT-PCR ( Figure 2A) and Western blotting ( Figure 2B). The adhesion of F4 ETEC to IPEC-J2 cells transfected with pcDNA ™ 6.2-GW/miR-APN was significantly reduced ( Figure 3A).
The cell line, pEC129-APN-IPEC-J2, that overexpresses APN was also constructed and characterized by using RT-PCR ( Figure 2A) and Western blotting ( Figure 2B). The adhesion of F4 ETEC to the pEC129-APN-IPEC-J2 cells was substantially increased, as compared with adhesion to the original IPEC-J2 cells ( Figure 3A). The addition of both APN polyclonal antiserum and an anti-F4 fimbriae monoclonal antibody to IPEC-J2 cells also reduced ETEC adhesion ( Figure 3B).
While it is known that F4 fimbriae receptors on the gut epithelium determine susceptibility to F4 + ETEC, the identity of these receptors is still under active investigation [3,43,44]. Polymorphisms in intron 7 of the MUC4 gene have been used to classify an important percentage of piglets as susceptible or resistant to F4 ETEC [16,17]. Although Ren [10,11,13,18].
Several glycoproteins and glycolipids isolated from porcine intestinal cells have been studied for their potential to act as F4 receptors, such as GP74 (TF), IGLad (intestinal neutral glycosphingolipid), and IMTGP (intestinal mucin-type glycoprotein) [43,45,46]. However, the functions of these potential receptors are not well characterized.
Here we characterized a newly described receptor for F4 + fimbriae, APN, which also serves as a receptor for the transmissible gastroenteritis virus (TGEV), porcine epidemic diarrhea virus (PEDV), and coronavirus [20,47,48]. APN is particularly highly expressed in the intestinal mucosa and is also associated with the MUC4 susceptible genotype [17,21]. APN was recently described as a potential receptor for F4ac + fimbriae; variations in the α2-3,6,8 sialic acid binding site of APN result in reduced binding of F4 fimbriae and binding of F4 fimbriae to APN results in clathrin-mediated endocytosis of the fimbriae [20]. Goetstouwers et al. reported that there are no genetic polymorphisms or expression differences in the ANPEP gene that have been associated with F4 ETEC susceptibility and hypothesized that differences in F4 binding to ANPEP are due to modifications in carbohydrate moieties [20,49].
We found that IPEC-J2 cells express APN and that F4 E. coli was able to adhere to IPEC-J2 cells in an APNdependent manner. Pre-incubation with APN polyclonal antiserum and anti-F4 fimbriae monoclonal antibody both reduced ETEC adherence to IPEC-J2 cells. Results from Y2H and pulldown assays also showed that FaeG binds directly to APN. We did not find an impact on APN-FaeG binding after treating samples with metaperiodate (NaIO 4 ), suggesting that, at least under our in vitro conditions, APN glycosylation does not play a significant role in FaeG binding. However, the molecular   details regarding APN-FaeG interactions and their roles in ETEC adherence await further experimentation.