Characterizing the Type 6 Secretion System (T6SS) and its role in the virulence of avian pathogenic Escherichia coli strain APECO18

Bacterial strains, plasmids and growth conditions

Strains and plasmids are shown in Table 1. The WT APEC strain was isolated from the pericardium of a chicken with signs of colisepticemia, and the sequenced genome is available at CP006830.1 (Nicholson et al., 2016). The O serogroup was identified at the E. coli reference center at Pennsylvania State University.

E. coli DH5α was used for cloning. All E. coli strains were grown in Lysogeny (LB) broth (BD Difco™, Franklin Lakes, NJ) at 37 °C with shaking, unless otherwise specified. The medium was supplemented with ampicillin (Amp 100 µg/mL), chloramphenicol (Cm 10–25 µg/mL) or L-arabinose (6.5 mM) as necessary.

DNA extraction

Bacterial DNA was obtained from whole organisms using the boil prep method as previously described (De Oliveira et al., 2020). Specifically, cultures and cell lysate were centrifuged at 16,700 g for 3 min.

Detection of T6SS genes by PCR

The presence of T6SS genes hcp, evpB, impK, vasK and icmF was analyzed in a collection of 106 avian fecal E. coli (AFEC) recovered from the feces of healthy production birds, 102 poultry litter E. coli isolated from the litter of poultry barns and 454 APEC isolates recovered from the lesions of production birds diagnosed with colibacillosis by polymerase chain reaction (PCR) amplification (see Table S1). APECO18 (CP006830.1) was used as template for primer design. Primers were obtained from Sigma (St. Louis, MO). Reactions were performed in a 25 µl volume as previously described (De Oliveira et al., 2020). The PCR conditions were as follows: 94 °C for 5 min; 30 cycles of 94 °C for 30 s, 60 °C for 30 s, 68 °C for 3 min, and a final extension step of 72 °C for 10 min. Primers used are listed in Table 2. APEC O18 was used as a positive control for primer check and sterile water in place of DNA for negative control check.

PCR products were subjected to horizontal gel electrophoresis in a 1.5% agarose gel (LE Agarose, Lonza, Alpharetta, GA) at 200 V for 70 min. A Hi-Lo molecular weight marker (50–10,000bp; Minnesota Molecular, Minneapolis, MN) and negative (sterile water) and positive controls from our lab collections were included as necessary. After electrophoresis, the gel was stained in 0.25% ethidium bromide solution (Sigma Aldrich. St. Louis, MO) for 20 min and viewed under UV light using an Omega Lum G imager (Aplegen, San Francisco, CA).

Construction of mutants and complemented strains

Isogenic mutants were constructed for the T6SS genes hcp, evpB and impK from T6SS cluster 1 (T6SS1) and icmF from T6SS cluster 2 (T6SS2) using the Lambda-Red recombination system (Datsenko & Wanner, 2000) using the APECO18 as parental strain. Briefly, oligonucleotides specific to the chloramphenicol cassette flanked by 50 nt extensions homologous to 5′- and 3′-ends of the gene to be deleted were used to amplify the chloramphenicol resistance cassette from plasmid pKD3 (ATCC^®, Manassas, VA). The PCR products were run on a 1.5% agarose gel, and gel extraction of the specific fragment was performed using QIAquick Gel Extraction Kit (Qiagen, Germantown, MD). The extracted fragments were electroporated into APECO18 containing the lambda-red expression plasmid pKD46 (ATCC^®, Manassas, VA). After electroporation, the cells were grown in SOC (super optimal broth with catabolite repression) for 90 min and plated on LB agar containing 25 µg/mL chloramphenicol. Colonies were screened by PCR to identify deletion mutants. The chloramphenicol resistance cassette was cured by transforming the helper plasmid pCP20 (ATCC^®, Manassas, VA) into the mutants and screening for chloramphenicol sensitive colonies. In trans complementation was performed by cloning PCR-amplified genes into the XbaI and HindIII restriction sites of plasmid pBAD24 (ATCC^®, Manassas, VA), and transforming the construct into their mutant counterparts. Primers used are listed in Table 2.

Growth curve analysis

The growth of WT APECO18 and mutant strains was analyzed in minimal medium M9 broth. Briefly, strains were incubated overnight in LB broth containing chloramphenicol (10 µg/mL) at 37 °C. Next, OD₆₀₀ of the cultures was measured and cultures were diluted to an OD₆₀₀ of 0.05 in M9 supplemented or not with arabinose at a final concentration of 0.2%. Cultures were incubated at 37  °C with shaking at 220 rpm, and OD ₆₀₀ measurements were obtained every 30 min. To measure OD₆₀₀, 300 µL of the growing culture was dispensed in a well of a 96 well plate, and the absorbance was read using an ELX 808 Ultra microplate reader (Bio-Tek Instruments, Winooski, VT). Growth curves were performed in biological duplicates. Absorbance of replicates was averaged, and data was plotted against time to build the growth curves. Growth curves were carried out for a total of 7–8 h.

Analysis of Expression of T6SS genes

qRT-PCR analysis

DNase-treated RNA was reverse transcribed using the First-Strand cDNA synthesis Kit from APExBio (Boston, MA). Briefly, 1 µg of DNAase-treated RNA was mixed with 1 µL of Random primers (50 µM), 1 µL of 10mM dNTP mixture, and adjusted to 10 µL with RNase-free water. The mixture was heated at 65 °C for 5 min and chilled on ice for 2 min for denaturation. The mixture was then centrifuged, and the cDNA synthesis mix was prepared by adding 4 µL of 5x first-strand buffer, 1 µL of RNase inhibitor, 1 µL of Reverse Transcriptase and RNase-free water to a total volume of 20 µL. The cDNA synthesis reaction was set as follows: 2 min at 25 °C, 50 min at 42 °C, and 15 min at 75 °C. The resultant cDNA was diluted 1:20 before use as template for qRT-PCR and stored at −20  °C until use.

Quantitative real-time RT-PCR (qRTPCR) was performed in a qTower³ G qPCR System (Analytik Jena, Jena, Germany) and analyzed using qPCRsoft v 4.1 software. Primers for qRT-PCR were obtained from Sigma Aldrich (St. Louis, MO). Reactions were performed using either qPCR Master Mix with Sybr^® Green (Goldbio, St. Louis, MO) or Platinum™ SYBR™ Green qPCR SuperMix-UDG (Invitrogen, Carlsbad, CA). Each reaction was performed in a final volume of 20 µl containing 25 ng cDNA (1 µL diluted cDNA), 1 µL each primer (10 µM), 10 µL 2x qPCR Master Mix Sybr^® Green (Goldbio, St. Louis, MO) or 10 µL Platinum™SYBR™ Green qPCR SuperMix-UDG (Invitrogen, Carlsbad, CA) and nuclease-free water to a final volume of 20 µL. When using Platinum™ SYBR™ Green qPCR SuperMix-UDG was used 1 µL of (1:10) ROX Reference Dye was added to each reaction. PCR conditions were as follows: 50  °C for 2 min hold, 95 °C for 2 min hold, and 40 cycles of 95 °C for 15 s and 60  °C for 30 s.

Threshold fluorescence was established within the geometric phase of exponential amplification, and the cycle threshold (CT) was determined for each RNA sample. Experiments were performed in biological and technical duplicates. The CT from each replicate was averaged. Expression levels were normalized using the housekeeping gene 16S rRNA as endogenous control. Melting curve analyses were performed after each reaction to ensure amplification specificity. Differences (n-fold) in transcripts were calculated using the relative comparison method (Schmittgen et al., 2000).

Adherence and invasion assays

Cell adherence and invasion assays using chicken fibroblasts DF-1 cells were performed as previously described (Barbieri et al., 2017) with slight modifications. The cells were cultured in 75 cm ²cell culture flasks (Corning^®, NY) in DMEM-F12 (ATCC^®, Manassas, VA) containing 10% fetal bovine serum (FBS; Sigma, MO) with incubation at 37 °C and 5% CO₂. Cells were transferred to sterile 24-well-plates at a concentration of 1 × 10⁵ cells/well 48 h prior each experiment. Wild-type APECO18, mutants and complemented strains were grown statically in LB broth overnight. Next, cultures were diluted 1:50 in fresh LB broth and grown at 37 °C with shaking at 220 rpm for 2 h. Cultures were then induced with 0.2% L-arabinose for 2 h. After induction, cultures were washed with PBS and adjusted to 1 ×10⁸cells/mL. 10 µL of bacterial suspension was used for infection. DF-1 cells were washed once with PBS and then exposed to bacteria at a multiplicity of infection (MOI) of 10. The 24-well-plates were centrifuged at 500 ×g for 5 min and incubated for 1 h for adherence and 4 h for invasion. The following steps were performed as described by (Barbieri et al., 2017)

The input dilution of bacteria was also plated and counted to determine the CFU for each inoculum used in the assay. Each experiment was performed in technical triplicates. The data presented here represents the average of two different experiments.

To visualize the association between bacteria and DF-1, 1 ×10 ⁵ cells per well were plated on glass coverslips in 24-well-plates and infected with a MOI of 10 CFU/cell as described above. After 1 h or 4 h post-infection, cells were fixed with 3.7% formaldehyde in PBS for 10 min at room temperature and stained with Giemsa for 20 min at room temperature. Samples were then washed with water and observed under a light microscope at 1,000x and photographed.

Plaque assay

The plaque assay was performed as previously described (Hasselbring, Patel & Schell, 2011) with slight modifications. Overnight cultures of wild type APECO18, mutant and complemented strains were diluted 1:50 and grown to exponential phase at 37 °C with shaking. Then 0.2% L-arabinose was added to the cultures of the complemented strains for induction of gene expression, with incubation for an additional 2 h. Bacteria were then pelleted by centrifugation at 5,700 ×g for 7 min, washed with SorC (16.7 mM Na₂H/KH₂PO₄/50µM CaCl₂, pH 6.0) and re-suspended in SorC to 5 ×10⁷cFU/mL.

D. discoideum AX3 cells were cultured in HL5 broth and collected by centrifugation at 500 ×g for 5 min, washed once with SorC, re-suspended in SorC to 1 × 10⁶ cells/mL, and serially diluted in 10-fold increments in SorC. D. discoideum dilutions were mixed 1:1 with bacterial suspensions to generate bacterium-to-amoeba ratios ranging from 5 ×10¹:1 to 5 ×10⁵:1. Aliquots of 10 µL of the mixed suspensions were spotted onto SM/5 agar plates and allowed to dry in a laminar hood under a sterile airflow. The plates were incubated at 22 ° C, examined after 3 and 5 d of incubation for plaque formation by D. discoideum, and photographed. Strains were then classified as sensitive or resistant to predation by D. discoideum based on the presence or absence of “predation plaques” formed by the amoeba. Experiments were done in duplicate.

Statistical analysis

For the analysis of the prevalence of T6SS genes harbored by strains from different collections examined in the study, the number of genes were treated as quantitative variables and the data was analyzed using non-parametric tests due to asymmetry in the distribution of the genes. Non-parametric analysis was also applied to compare the phenotypes presented by the mutants in comparison to the WT strains. Direct comparisons (where possible) between two groups were made using the Mann–Whitney U test. All statistical analysis was performed using GraphPad Prism (Version 7.0d) for Windows (GraphPad, La Jolla, CA). Statistical significance was accepted when p < 0.05. For adherence and invasion assays, bacterial groups were compared using Student’s t- test and differences were considered significant when p < 0.05.

Prevalence of T6SS genes is significantly higher in APEC than in poultry litter E. coli and AFEC isolates

T6SS clusters are common in the genomes of E. coli pathotypes and are divided into three groups based on size, homology and structural analysis of clusters (Ma et al., 2014; Ma et al., 2013). Here, genomic analysis found that APECO18 harbors two putative T6SS clusters, T6SS1 and T6SS2. The T6SS1 cluster is 30.2 kb in length, with a GC content of 52.2%, and is flanked by tRNA genes. The T6SS2 cluster is 27.9 kb in length, with a 52% GC content and also flanked by tRNA genes. Most of the genes analyzed in this study, namely evpB, impK and hcp belong to T6SS1 cluster. Additionally, we analyzed the icmF gene from cluster 2. Schematic representation of these clusters is shown in Fig. 1. To gain insight into the relationship between the presence of T6SS genes and the pathogenesis of APEC, we used PCR to screen a collection 454 APEC, 102 poultry litter E. coli and 106 AFEC isolates.

The prevalence evpB, impK and hcp was significantly higher (p < 0.05) in APEC than in poultry litter E. coli and AFEC isolates; the prevalence of impK and hcp was also significantly higher in AFEC than in poultry litter E. coli isolates. The prevalence of vasK was significantly higher (p < 0.05) in APEC than in poultry litter E. coli isolates. IcmF was not detected in AFEC isolates, and the prevalence did not significantly differ between poultry litter E. coli and APEC isolates (Fig. 2). These findings support our hypothesis that T6SS contributes to the virulence of APEC.

Deletion of T6SS genes did not affect the growth of APECO18 in minimal medium

To further investigate the role of T6SS genes in the pathogenesis of APEC strains, isogenic deletion mutants for evpB, impK, hcp and icmF were generated using APECO18 as the parental strain (Fig. S1). Prototypic strain APECO18, which belongs to O18 serotype and B2 phylogenetic group, was used as template for designing the primers. To exclude that the mutations have an influence on the growth rate of the mutants, APECO18 wild-type, mutants and complemented strains were assessed for their ability to grow in minimal medium M9. No differences were found in the growth rates of hcp, evpB, impK or icmF mutants compared to the wild-type strain in minimal medium (Fig. S2).

APECO18 expresses genes encoding a T6SS

Since T6SS secretion has been shown to be temperature dependent in other organisms (Bladergroen, Badelt & Spaink, 2003), the expression of T6SS genes by APECO18 grown in LB and DMEM media at 37 °C or 42 °C was analyzed. Genes encoding 16S rRNA were used as endogenous controls for data normalization. As shown in Fig. 3, APECO18 expressed T6SS genes in both LB and DMEM with significant differences (p < 0.05) observed according to the media or temperature used. For instance, evpB expression was significantly higher (p < 0.05) in LB regardless of the temperature; vasK expression was significantly higher (p < 0.05) in DMEM at 42 °C compared to DMEM 37 °C and compared to LB at 42 °C.

Because expression of the T6SS has been shown to be upregulated in the presence of the host (De Pace et al., 2010), we also assessed whether the expression of T6SS genes by APEC O18 was upregulated in the presence of chicken fibroblasts DF-1. Expression of evpB was significantly (p <  0.05) enhanced at 3 h post-infection, and expression of icmF was significantly enhanced (p < 0.05) at 1 h post-infection and enhanced further at 3 h post- infection (Fig. 4).

Role of T6SS in the adherence of APECO18 to chicken fibroblasts DF-1

Because T6SS genes icmF, hcp, and clpV have been previously recognized as being involved in the binding of APEC Sept362 to HeLa cells (De Pace et al., 2011; De Pace et al., 2010), we assessed whether hcp and other T6SS genes (evpB, impK, and icmF) are involved in the adherence of APECO18 to chicken fibroblasts DF-1.

We did not observe a negative effect of the mutation of T6SS genes hcp, evpB, impK, and icmF in the adherence capability of APECO18. An unexpected significant increase (p < 0.05) of 90% in the adherence APECO18 ΔimpK in relation to the WT strain was observed. Complemented strains APECO18 ΔimpKc, APECO18 ΔevpBc and APECO18 ΔicmFc showed an adherence rate that was 2.27, 16.30 and 6.47 times higher than the wild-type respectively (p < 0.05) (Fig. 5A). These results suggest that the T6SS genes analyzed in this study are not involved in the adhesion of APECO18 to chicken fibroblasts DF-1.

Role of T6SS in the invasion of chicken fibroblasts DF-1 by APECO18

Deletion of hcp and icmF has been shown to affect the invasion of HeLa cells by APEC Sept362 (De Pace et al., 2011; De Pace et al., 2010). To investigate whether hcp, evpB, impK and icmF are involved in invasion of chicken fibroblasts DF-1 by APECO18, we assessed the invasion capability of APECO18 WT, mutants and complemented strains.

Similar to what was found for adherence capability, we did not observe a negative effect of the mutation of T6SS genes on the invasion capability of the mutants compared to the wild-type strain. A decrease in invasion rate of 61%, 78% and 78% was observed for APECO18 Δhcp, APECO18 ΔevpB, and APECO18 ΔvasK respectively in relation to the wild-type strain, however these differences were not statistically significant (p>0.05) (Fig. 5B). An unexpected increase of 20% in the invasion rate of DF-1 by APECO18 ΔimpK compared to the WT strain was observed, but it was not significant (p >  0.05). Complemented strain APECO18 ΔevpBc showed an invasion that was 3.13 times the invasion rate presented by the wild-type APECO18 (p < 0.05).

Deletion of hcp increases sensitivity of APECO18 to predation by D. discoideum

We assessed the ability of APECO18 WT, mutants and complemented strains to resist to predation by D. discoideum. Deletion of impK and icmF did not affect resistance of APECO18 to predation by D. discoideum. An increased sensitivity was observed in the ΔevpB strain but no restoration was observed in the complemented strain indicating a possible polar effect of the evpB deletion on this phenotype (Fig. S3). However, deletion of hcp affected the strain’s sensitivity to predation by D. discoideum. At the bacteria to amoeba ratio of 5 ×10¹:1, the bacterial lawn was completely consumed and replaced by D. discoideum fruiting bodies in the WT, mutant, and complemented strain. At the ratio 5 ×10²:1 the amoeba consumed most of the bacteria, but some of the bacterial lawn was still observed at the edges of the spot in the WT, mutant, and complemented strains. The difference between the WT strain and Δhcp was observed at the bacteria-amoeba ratio of 5 ×10³:1. At this ratio, the WT strain was resistant to predation by the amoeba, with no plaques observed. However, the mutant strain became sensitive to predation with a large predation plaque observed on the bacterial lawn. The phenotype was restored in the complemented strain. In summary, 500 amoeba cells were necessary to form a plaque on the WT bacterial lawn, while only 50 amoeba cells were able to form a plaque on the Δhcp strain. This result indicates that Hcp is involved in the ability of APECO18 to resist predation by D. discoideum (Fig. 6).