Introduction

Post-weaning diarrhea (PWD) causes severe economic losses to swine production due to high morbidity and mortality (Fairbrother et al. 2005; Rausch et al. 2017). Enterotoxigenic Escherichia coli (ETEC) strains expressing F18ac fimbriae (F18ac+ ETEC) or K88 fimbriae (K88+ ETEC) are the most common bacterial pathogens causing PWD (de la Fé Rodríguez et al. 2011; Frydendahl 2002; Zhang et al. 2007). Adhesins and enterotoxins [heat-labile enterotoxin (LT) and heat-stable enterotoxin (ST)] of ETEC are known as virulence determinants in PWD (Fairbrother et al. 2005; Zhang et al. 2007). Adhesins expressed by F18ac+ ETEC initially mediate adherence to pig small intestinal epithelial cells. Adhered bacteria then produce and release enterotoxins into the gut lumen, leading to fluid homeostasis and PWD. Thus, it is believed that initial adherence mediated by classes of adhesins is critical for F18ac+ ETEC infection. Adhesins involved in the F18ac+ ETEC adherence process include F18ac fimbriae and flagella (Frydendahl 2002; Duan et al. 2013). F18ac fimbriae mediate F18ac+ ETEC adherence to specific host cell receptors by their adhesive subunit (fedF). Flagella are suggested to contribute to F18ac+ ETEC adherence in two ways: (i) promoting bacteria attachment to host tissue by their motility property; (ii) directly increasing the adherence mechanism as adhesins. Adhesins involved in diffuse adherence (AIDA-I) are expressed by F18ac+ ETEC. Additionally, these contribute to colonization and biofilm formation for porcine diarrheagenic E. coli strain PD20 (Ravi et al. 2007).

Type I fimbriae are the most commonly expressed pili on the surface of E. coli. They are encoded by the fim gene cluster that includes nine genes (fimA–fimH) and are required for fimbrial biosynthesis (Orndorff and Falkow 1984; Klemm and Christiansen 1987). The major structural fimbrial subunit is encoded by the fimA gene, whereas the minor and adhesive subunit, which is located at the tip of the fimbriae, is encoded by the fimH gene (Klemm 1984; Krogfelt et al. 1990). Type I fimbriae mediate bacterial adherence by binding its fimA adhesin subunit to mannose-containing glycoprotein receptors on host cells. It is also known that type I fimbriae are important adhesins for many pathogens. In uropathogenic E. coli (UPEC) causing urinary tract infections (UTIs), type I fimbriae are implicated as virulence factors for UPEC bacterial initial adherence and subsequent invasion to urinary epithelial cells and biofilm formation (Anderson et al. 2003; Wiles et al. 2008; Bien et al. 2012). Type I fimbriae mediate adhesion and invasion to human brain microvascular endothelial cells (HBMEC) to develop neonatal meningitis (Murphy et al. 2013). Additionally, type I fimbriae of E. coli strain LF82 mediate adherence to host cells in Crohn’s disease (Boudeau et al. 2001). Also, the type I fimbriae are commonly expressed in F18ac+ ETEC strains. However, their role(s) in adherence to pig epithelial cells remains poorly understood.

In this study, we constructed mutants with deletion of the major subunit (F18acΔfimA) or the minor subunit (F18acΔfimH) and examined the ability of the mutant strains to adhere to piglet epithelial IPEC-1 and IPEC-J2 cell lines. In addition, we explored the cross-regulation of type I fimbriae and other adhesins of F18ac+ ETEC.

Materials and methods

Bacterial strains, plasmids, and cell lines

Bacterial strains and plasmids used in this study are listed in Table 1. Antibiotic-resistant strains were cultured in LB containing 100 μg/mL of ampicillin (Amp) or 34 μg/ml of chloramphenicol (Cm). For determining bacterial growth rates, wild-type and mutants strains were grown in LB broth, and complemented strains in LB broth containing 100 μg/mL of Amp, for 6 h at 37 °C in a shaking incubator at 220 rpm. Afterwards, the optical density was measured hourly for each cultured strain at 600 nm (OD600).

Table 1 Bacterial strains and plasmids used in this study
Full size table

Newborn porcine small intestinal epithelial cell lines IPEC-J2 and IPEC-1 were cultured in Dulbecco’s minimal Eagle medium (DMEM) (Gibco) supplemented with heat-inactivated fetal bovine serum (FBS; Gibco) in 5% CO2 at 37 °C.

Mutant strains and complementary strains construction

The F18ac+ ETEC fimA (GenBank: CP019558.1) and fimH (GenBank: JF289169.1) isogenic gene mutants were constructed using the phage λ-red-mediated recombination system as previously described (Datsenko and Wanner 2000). Briefly, the Cm resistance-encoding gene sequence was amplified from the chloramphenicol cassette of plasmid pKD3 in polymerase chain reaction (PCR) amplification with ΔfimA-F and ΔfimA-R primers (Table 2). Amplified PCR products were purified from agarose gels and transferred into F18ac+ ETEC-competent cells containing pKD46 plasmid. Positive colonies were selected on LB agar plates containing Amp (100 μg/mL) and Cm (34 μg/mL). Allelic replacement of fimA by the Cm cassette was verified by PCR screening using fimA-specific primers (pBR-fimA-F, pBR-fimA-R; Table 2). Finally, Flp recombinase-expressing vector pCP20 was introduced to delete the Cm cassette. Similarly, the ΔfimH mutant was generated using ΔfimH-F and ΔfimH-R primers (Table 2). Deletion of fimA or fimH in the mutant strains was confirmed by PCR screening, DNA sequencing, and hemagglutination assays.

Table 2 Oligonucleotides used in this study
Full size table

The full-length fimA and fimH genes were PCR-amplified from F18ac+ ETEC genomic DNA using primers pBR-fimA-F/pBR-fimA-R and pBR-fimH-F/pBR-fimH-R, respectively (Table 2). Plasmids pBR322/fimA or pBR322/fimH were introduced into F18acΔfimA or F18acΔfimH to generate complementary strains. Restoration of type I fimbriae expression in the complementary strains was verified in yeast cell agglutination and hemagglutination assays.

Yeast cell agglutination assay and mannose-sensitive hemagglutination (MSHA) test

The yeast cell agglutination assay was performed as previously described (Müller et al. 2009). Briefly, bacteria that were subcultured four times by serial two-fold dilution in LB broth at 37 °C under static conditions and Pichia pastoris GS115 were washed and resuspended to OD600 = 0.5 using phosphate buffered saline (PBS, pH7.4) with or without 0.5% D-mannose. Suspensions were mixed at a 1:1 (v/v) ratio. The slide was gently rotated for 5 min at room temperature (RT), while monitoring for visible HA was conducted. In the MSHA assay, 25 μL of 5% pig erythrocyte suspension with or without D-mannose were mixed with an equal volume of a bacterial suspension on a glass slide. The agglutination was evaluated after being gently rotated for 5 min at RT.

Adherence assays and adherence inhibition assays

Bacterial adherence assays were performed as previously described (Jouve et al. 1997). Briefly, in each well of a 48-well tissue culture plate (Corning), 1 × 105 IPEC-1or IPEC-J2 cells were challenged with E. coli at 2 × 106 colony-forming units (CFUs) for 1 h at 37 °C. Cells were washed three times with PBS to remove non-adherent bacteria, then lysed with 0.5% Triton X-100. Cell lysates were serially diluted with PBS and plated on LB agar plates overnight at 37 °C.

The adherence inhibition assay was performed by coincubation of the F18ac+ ETEC parent strains and cells in the presence of 0.5% D-mannose. The number of adherent bacteria was determined as described above.

Quantification of biofilm formation

Biofilm formation in a 96-well microtiter plates assay was carried out as previously described (Duan et al. 2013). Briefly, overnight-grown bacterial cultures were adjusted to OD600 = 1.0 using fresh biofilm-inducing medium (Hossain and Tsuyumu 2006). 150-μL culture suspensions were added to each well of the 96-well microtiter plates and incubated for 24 h at 30 °C statically. Non-adherent bacteria in the wells were removed by washing with ddH2O. Biofilm was stained with 200 μL of 2% crystal violet (CV) at RT for 15 min and solubilized by 95% ethanol for 30 min at RT. OD600 values were measured by a spectrophotometer (BioTek, USA).

RNA isolation and real-time quantitative polymerase chain reaction (RT-PCR)

Total RNA from bacterial samples was extracted using a TRIzol as previously described (Duan et al. 2013). RNA quality and yield were verified with agarose gel analysis. The PrimeScript® RT Reagent Kit with gDNA Eraser (Takara) was used to synthesize the cDNA according to the manufacturer’s protocol. Primers were designed to amplify fliC, AIDA-I, and fedF genes to target regions of unique sequences within each gene (Table 2). PCRs were performed in triplicate with the ABI7500 instrument. All data were normalized with the endogenous reference gene gapA and analyzed by the 2−ΔΔCT method.

Statistical analysis

Each experiment was performed in triplicate and repeated at least three times to ensure reproducibility. Data were analyzed using Student’s t-test for independent samples. Differences were considered significant if p ≤ 0.05.

Results

Characterization of ΔfimA and ΔfimH mutants and complementary strains

Construction of F18ac+ ETEC mutants and the complemented strains (F18acΔfimA/pfimA and F18acΔfimH/pfimH) were verified by DNA sequencing and agglutination reactions. Unlike the parent strain, the F18acΔfimA and F18acΔfimH mutant strains lost their ability to agglutinate with yeast cells and pig erythrocytes. In contrast, the complementary strains restored their ability to agglutinate with yeast cells and pig erythrocytes in a mannose-sensitive manner like the parent strains. All strains had similar growth rate after culturing in LB broth in a shaking incubator for 6 h at 37 °C (complemented strains in LB broth with Amp) (Fig. 1).

Fig. 1
figure 1

Bacterial growth curves of F18+ enterotoxigenic Escherichia coli (ETEC) wild-type (WT), mutant strains, and complementary strains. WT F18ac+ ETEC, F18acΔfimA, F18acΔfimA/pfimA, F18acΔfimH, and F18acΔfimH/pfimH were grown in LB broth at 37 °C for 6 h. The optical density at 600 nm (OD600) was measured hourly. The data presented are the means of four independent measurements

Full size image

Type I fimbriae mediate F18ac+ ETEC adherence to IPEC-1 and IPEC-J2 cells

F18ac+ ETEC strains bind to IPEC-1 and IPEC-J2 porcine cell lines that are derived from the small intestine of newborn piglets (Duan et al. 2013). It was found that both F18acΔfimA and F18acΔfimH mutants showed reduced adherence ability to either the IPEC-1 or IPEC-J2 cell lines. Cell viable counts revealed that the F18acΔfimA mutant exhibited about 38% reduction in adherence compared to that of the parent strain, and the F18acΔfimH mutant showed about 43% reduction of adherence (Fig. 2a). In a similar way, F18acΔfimA and F18acΔfimH mutants showed 62% and 64% adherence to IPEC-J2 cells (Fig. 2b). In contrast, F18acΔfimA/pfimA and F18acΔfimH/pfimH complemented strains showed similar adherence to both cell lines as the parent strains.

Fig. 2
figure 2

Adherence of WT F18ac+ ETEC, isogenic mutants, and complementary strains to IPEC-1 and IPEC-J2 cells. a Adherence to IPEC-1 cells by WT F18ac+ ETEC, F18acΔfimA, F18acΔfimA/pfimA, F18acΔfimH, and F18acΔfimH/pfimH strains. b Adherence to IPEC-J2 cells by WT F18ac+ ETEC, F18acΔfimA, F18acΔfimA/pfimA, F18acΔfimH, and F18acΔfimH/pfimH strains. The adherence index of WT strains was set as 100%. Each assay was conducted in triplicate and independently repeated at least three times. *Indicates statistically significant differences to the F18ac+ ETEC WT strain (p < 0.05)

Full size image

The adherence inhibition assay showed that the adherence of wild-type F18ac+ ETEC strains to IPEC-1 and IPEC-J2 cells was blocked 51% and 73%, respectively, by 0.5% D-mannose compared with wild-type strains. The magnitude inhibition of 8% D-mannose was not significantly different compared to that from 0.5% D-mannose (data not shown). The results indicated that type I fimbriae were involved in F18ac+ ETEC adherence to both IPEC-1 and IPEC-J2 cells in vitro.

Type I fimbriae enhance F18ac+ ETEC biofilm formation

F18ac+ ETEC strains formed biofilm in the 96-well microplates in vitro. F18acΔfimA and F18acΔfimH mutants exhibited about 41% and 69% absorbance, respectively, compared to the parent strain absorbance (Fig. 3, p < 0.01). In contrast, the biofilm formation ability of both complemented strains was restored.

Fig. 3
figure 3

Assessment of biofilm formation of WT F18ac+ ETEC, isogenic mutants, and complementary strains. Biofilm biomass was spectrophotometrically measured at OD600 after CV staining. The experiment was performed at least three times and six replicates were used in each single test. **Indicates statistically significant differences compared to the F18ac+ ETEC WT strain (p < 0.01)

Full size image

Type I fimbriae affect the expression of other adhesins

To study whether deletion of the fimA or fimH genes affected the expression of other adhesins for F18ac+ ETEC, quantitative PCR (qPCR) was used to quantify the transcription of fedF, fliC, and AIDA-I genes using gapA as a normalizing internal standard. The expression of AIDA-I was upregulated 2.84- and 3.37-fold in F18acΔfimA and F18acΔfimH mutants, respectively (Table 3). The deletion of fimA resulted in 1.32- and 1.21-fold decrease in fliC and fedF expression, respectively. Deletion of the fimH caused 1.54- and 1.17-fold decrease in fliC and fedF expression, respectively (Table 3). In contrast to AIDA-I, these differences are not statistically significant.

Table 3 Transcript expression of the other potential adhesin genes was measured using quantitative PCR (qPCR)
Full size table

Discussion

PWD is one of the most common diseases in piglets. Piglets develop PWD 3–10 days after weaning, which causes weight loss, slow growth, and acute death (Svensmark et al. 1989; Verdonck et al. 2002; Vu-Khac et al. 2007). To date, there are no effective prevention strategies against PWD. Therefore, gaining further insight to the pathogenesis of the disease can be helpful to develop prevention strategies. The most important pathogens causing PWD are ETEC, E. coli strains producing enterotoxins. ETEC associated with PWD produce fimbriae including K88 and F18 and enterotoxins including LT and ST (Frydendahl 2002; Zhang et al. 2007). Initial attachment and adherence to the host’s mucosal surface mediated by bacterial fimbriae is a prerequisite for PWD. Bacterial adherence is a complex and multifactorial process, which is often mediated by different adhesins (Nandre et al. 2016). F18ac fimbriae and flagella are the known adhesins involved in mediating F18ac+ ETEC adherence. It was demonstrated that the absence of F18ac fimbriae or flagella highly impairs the adherence ability of F18ac+ ETEC (Frydendahl 2002; Duan et al. 2013). Data from the current study suggest that type I fimbriae play roles in the adherence of F18ac+ ETEC strains to piglet epithelial cells.

Type I fimbriae are expressed by most E. coli strains and mediate mannose-sensitive (MS) adherence to host epithelial cells (Iida et al. 2001). This study showed that the adherence of F18ac+ ETEC to IPEC-1 and IPEC-J2 was highly inhibited with exogenous 0.5% D-mannose. However, an increase in D-mannose concentration did not lead to additional inhibition of adherence. Data from this study showed that the F18acΔfimA and F18acΔfimH mutants had significantly lower adherence rates to IPEC-1 and IPEC-J2 cells compared to those of the wild-type strain or the complemented strains. These results indicated that adherence roles are evident for F18+ ETEC type I fimbriae. However, the adherence to porcine cell lines was not completely abrogated in the type I fimbriae mutants. This may suggest that other adhesins could be involved in mediating F18ac+ ETEC adherence.

Adherence ability was not significantly different between the F18acΔfimA and F18acΔfimH mutants. FimH is the lectin and adhesin of type I fimbriae. FimH binds to receptors containing terminal mannose residues and mediates agglutination of yeast and red blood cells (Schembri et al. 2000; Stahlhut et al. 2009); therefore, FimH-mediated binding can be inhibited by exogenous D-mannose. Both F18acΔfimA and F18acΔfimH mutants lost their ability to agglutinate with yeast cells and pig erythrocytes, indicating deficient expression of fimH in these two mutant strains. However, the mechanism by which FimA deletion impairs FimH expression is still unknown. Nevertheless, the results from this study indicated that adhesin FimH of type I fimbriae also mediate F18ac+ ETEC attachment to piglet intestinal epithelium cells besides F18 fimbriae and flagella.

Biofilm formation facilitates firm and irreversible adhesion to a surface, which is highly advantageous for bacteria survival (Koczan et al. 2011). Initial bacterial adherence to cell surfaces is in favor of biofilm formation. Increasing evidences suggest that type I fimbriae play attractive roles in the enhancement of biofilm formation (Nandre et al. 2012; Grzymajło et al. 2013; Kuźmińska-Bajor et al. 2015). The current study found that the F18acΔfimA and F18acΔfimH mutant strains were less effective in biofilm formation compared to the F18ac+ ETEC parent strains. This study also observed that the introduction of 0.5% D-mannose in the biofilm-inducing medium significantly reduced the biofilm formation levels of the F18ac+ ETEC parent strain. After incubation with 0.5% D-mannose, the absorbance of F18ac+ ETEC parent strains was inhibited by 69% as compared with the parent strains. These results demonstrate that type I fimbriae may play a role in enhancing F18ac+ ETEC biofilm formation in vitro. But whether type I fimbriae are directly involved in the biofilm formation process or indirectly down-regulate expression of other genes warrants further studies.

Expression of surface structures in E. coli is a complex and cross-regulated process involving type I fimbriae and other surface organelles, including P fimbriae, K88 fimbriae, and flagella (Holden et al. 2006; Lane et al. 2007; Zhou et al. 2013). Using RT-PCR, an up-regulation of the non-fimbrial adhesion AIDA-I in the in vitro cultures of the above fimA and fimH deletion mutants was detected, which was in contrast to the gene regulations for F18ac and flagellin expressions. AIDA-I is an auto-transporter protein expressed on the bacterial outer membrane and appears to mediate bacterial initial adherence and biofilm formation in vitro (Ravi et al. 2007). In order to save energy and survive in vivo, bacteria predominantly produce specific type fimbriae during the infection process (Croxen and Finlay 2010). For F18ac+ ETEC, specific F18ac fimbriae are the predominant adhesins. This may explain why type I fimbriae play a minor role in the F18ac+ ETEC adherence in vivo. It was found that AIDA-I and F18ac fimbriae are governed on the same plasmid (Mainil et al. 2002), but they could be regulated differently in the mutant strains. The exact mechanism of cross-regulation between type I fimbriae and other adhesion factors needs to be further investigated.

In conclusion, the present study demonstrated that type I fimbriae contribute to the adherence of F18ac+ ETEC to porcine intestinal epithelial cells and the early phases of biofilm formation in vitro. Additionally, there may be regulatory cross-links among F18ac+ ETEC adhesins, including type I fimbriae and AIDA-I adhesin. As the adherence of F18ac+ ETEC to piglet epithelial cells is considered a crucial and initial step of pathogenesis, the results from this study provide insights into the pathogenic mechanisms of the PWD-causing F18ac+ ETEC and, perhaps, helpful information for the development of new strategies against F18ac+ ETEC infections. Notwithstanding some viewpoints which considered that type I fimbriae contribute little, or even negatively, to pathogenicity in vivo (May et al. 1993; Chen et al. 2003), the significance of type I fimbriae of F18ac+ ETEC in the detailed pathogenesis of PWD remains to be elucidated in future studies.