Characterization of translocon proteins in the type III secretion system of Lawsonia intracellularis

Abstract

Lawsonia intracellularis, the etiologic agent of proliferative enteropathy (PE), is an obligate intracellular Gram-negative bacterium possessing a type III secretion system (T3SS), which enables the pathogen to translocate effector proteins into targeted host cells to modulate their functions. T3SS is a syringe-like apparatus consisting of a base, an extracellular needle, a tip, and a translocon. The translocon proteins assembled by two hydrophobic membrane proteins can form pores in the host-cell membrane, and therefore play an essential role in the function of T3SS. To date, little is known about the T3SS and translocon proteins of L. intracellularis. In this study, we first analyzed the conservation of the T3S apparatus between L. intracellularis and Yersinia, and characterized the putative T3S hydrophobic major translocon protein LI1158 and minor translocon protein LI1159 in the L. intracellularis genome. Then, by using Yersinia pseudotuberculosis as a surrogate system, we found that the full-length LI1158 and LI1159 proteins, but not the putative class II chaperone LI1157, were secreted in a − Ca²⁺ and T3SS-dependent manner and the secretion signal was located at the N terminus (aa 1–40). Furthermore, yeast-two hybrid experiments revealed that LI1158 and LI1159 could self-interact, and LI1159 could interact with LI1157. However, unlike CPn0809 and YopB, which are the major hydrophobic translocon proteins of the T3SS of C. pneumoniae and Yersinia, respectively, full-length LI1158 was non-toxic to both yeast and Escherichia coli cells, but full-length LI1159 showed certain toxicity to E. coli cells. Taken together, despite some differences from the findings in other bacteria, our results demonstrate that LI1158 and LI1159 may be the translocon proteins of L. intracellularis T3SS, and probably play important roles in the translocation of effector proteins at the early pathogen infection stage.

Introduction

Lawsonia intracellularis is an obligate intracellular Gram-negative bacterium isolated in 1993 and the etiologic agent of proliferative enteropathy (PE), an infectious disease macroscopically characterized by thickening of the intestinal epithelium due to enterocyte proliferation [1, 2]. After an initial close association with the cell membrane of enterocytes, L. intracellularis is endocytosed into host cells and inhibits their maturation, and then the infected host cells will continue to undergo mitosis and proliferation and finally form hyperplastic crypts [2, 3]. Mild to severe diarrhea are the major clinical signs observed in infected animals [3]. Despite decades of research on this pathogen, current understanding of its molecular pathogenesis is still rather limited due to its obligate intracellular nature, genetic intractability, and lack of an in vitro bacterial model [4, 5].

The type III secretion system (T3SS), a well-characterized system, plays significant roles in the pathogenicity of many Gram-negative bacterial pathogens [6]. Currently, several enteropathogenic pathogenic bacteria, such as Yersinia, Salmonella spp., Shigella spp., and enteropathogenic and enterohemorrhagic Escherichia coli (EPEC/EHEC), have been well characterized and are known to utilize T3SS for infection of their hosts to cause a variety of infectious diseases [6]. The Ysc-Yop T3SS from Yersinia was the first characterized T3SS and is often regarded as the archetype [7, 8]. The T3S components and its substrates, namely the Yop effector proteins (YopH, YopO/YpkA, YopP/YopJ, YopE, YopM, and YopT), are all encoded by a 70-kb virulence plasmid called pYV in Y. enterocolitica [9]. In vitro, the need for contact with eukaryotic cells can be circumvented by chelating Ca²⁺ ions with agents such as EGTA or oxalate [10]. Since the T3SS has a conserved structural framework across different animal bacterial pathogens, Yersinia has been successfully employed as a surrogate system to identify novel T3SS substrates for genetically intractable strains [11].

T3SS is a syringe-like apparatus comprising 20–25 proteins, which compose a base, an extracellular needle, a tip complex, and a translocator or translocon (used throughout this article). The tip protein (LcrV of the Ysc family) complex serves as a platform for the assembly of the translocon, which forms a translocon pore in the host cell membrane [6, 7, 12]. The translocon pore is assembled by two hydrophobic membrane proteins, a major translocon protein and a minor translocon protein determined according to their relative sizes compared with each other [12]. Each pair of major and minor translocon proteins share a common class II chaperone, whose coding genes are localized next to the genes encoding the translocons and the tip proteins, and are expressed from one operon [12]. For the Ysc family in Yersinia, the needle protein YscF and tip protein LcrV are indispensable for the translocation of effector proteins into host cells [13, 14]. LcrH/SycD, an archetypal class II chaperone, binds to the translocon proteins (YopB and YopD, the major and minor translocon protein) and prevents their premature folding and homo- or hetero-oligomerization in the cytosol of bacterial cells [15].

Most proteins composing the T3S apparatus are conserved among species possessing a T3SS. Genes encoding the T3S apparatus have been identified in the complete genome of L. intracellularis strain PHE/MN1-00, and the expression of some components has been verified by RT-PCR and serological response during infection [2, 16]. Moreover, an operon genetically related to the Salmonella pathogenicity island 2 (SPI2) and composed of three genes, including a chaperone (LI1157, PcrH), an effector protein (LI1158, SseC), and a hypothetical protein LI1159 (referred to as SseD), was found to be highly expressed in the cytoplasm of porcine enterocytes infected with L. intracellularis [17]. Combined with findings in other similar bacteria, it can be inferred that this T3SS may be crucial for intracellular development and probably plays a significant role in the virulence of this unusual pathogen. However, the experimental identification of T3S components and substrates of L. intracellularis is largely hampered by the difficulty in culturing of the bacteria under axenic (cell-free) conditions, and the best method for the identification may be the use of a T3S surrogate system.

In this study, we first bioinformatically and experimentally characterized LI1158 and LI1159 as putative hydrophobic translocon proteins. Furthermore, by using the yeast two-hybrid system, we found that LI1158 and LI1159 can self-interact, and LI1159 can interact with LI1157. These results are consistent with the assumption that LI1158 and LI1159 are the major and minor translocon proteins of L. intracellularis T3SS, respectively, and both play certain roles in the target-cell attachment and invasion of L. intracellularis. However, we also obtained some different results compared with the findings in other bacteria. Our yeast two-hybrid (Y2H) assay showed that LI1158 had no interaction with LI1159 and LI1157. The expression of LI1158 resulted in no strong inhibitory effect on the growth of yeast cells and E. coli cells, but LI1159 expression obviously inhibited the growth of E. coli cells. These results improve our understanding of the T3SS translocon proteins in L. intracellularis and will facilitate the development of more specific serological diagnostic assays and effective vaccines to control PE outbreak.

Materials and methods

Strains, cell lines, and media

E. coli DH5α (TransGen Biotech, Beijing, China) was used for plasmid amplification, serial dilution patch tests, and growth curve assay. The strain was propagated in Luria–Bertani (LB) supplemented with 100 μg mL⁻¹ ampicillin at 37°C with agitation or on LB agar plates. The Y. enterocolitica T3S-proficient MRS40_ΔyopHOPEM and T3S-deficient MRS40_ΔyscF were used for T3S secretion assays, and MRS40_ΔyopHOPEM was also used for serial dilution patch tests and growth curve assay. Yersinia was routinely grown in brain heart infusion (BHI; TransGen Biotech) with 200 µg mL⁻¹ ampicillin to select the expression vectors at 30°C with agitation or on BHI agar plates.

The Saccharomyces cerevisiae strain Y2H gold (MATa, trp1-901, leu2-3, 112, ura3-52, his3-200, gal4Δ, gal80Δ, LYS2::GAL1_UAS–Gal1_TATA–His3 GAL2_UAS–Gal2_TATA–Ade2, URA3::MEL1_UAS–Mel1_TATA-AUR1-C MEL1) (Clontech, Takara Bio USA, Inc., Mountain View, CA, USA) was used for Y2H experiment and W303-1A (MATa ade2-1 ura3-1 his3-11,15 trp1-1, leu2-3,112 can-100) (Clontech) was use for growth inhibition assay. Both strains were grown at 30°C in yeast extract peptone dextrose (YPD) or synthetic dropout medium (SCM minus).

The human embryonic kidney (HEK) 293 T cell line (Thermo Fisher Scientific, San Jose, CA, USA) was cultured in Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen Tech, Carlsbad, CA, USA) supplemented with 10% (v/v) fetal bovine serum (FBS, Gibco, Waltham, MA, USA), 1 mmol L⁻¹ glutamine, and 100 U mL⁻¹ penicillin and streptomycin.

Plasmid construction

The primers and plasmids used in this study are listed in Tables 1 and 2, respectively. The DNA fragments encoding full-length LI1157-LI1159 amplified from genomic DNA extracted from L. intracellularis-positive porcine ileal mucosa with a C-terminal hemagglutinin (HA) epitope tag fragment were separately cloned into the EcoR I/Hind III sites of the pBAD24 vector. The first 40 amino acids (aa) of LI1158 and 1159 were cloned into frame with the C-terminal amino acids of YopE (YopE_Δ15) in the previously constructed pBAD24-YopE-HA plasmid, respectively [18]. Plasmids were introduced into Y. enterocolitica by electroporation.

Table 1 Primers used in this study.

Table 2 Plasmids construction in this study.

The DNA fragment encoding full-length LI1158 was cloned into the galactose-inducible yeast expression vector pRS416-GAL1, Y2H bait expression vector pGADT7 (Clontech), and prey expression vector pGBKT7 (Clontech), respectively. The DNA fragment encoding full-length LI1159 was cloned into the pGADT7 and pGBKT7, respectively. The DNA fragment encoding full-length LI1157 was cloned into the pGADT7. Yeast transformation was performed using the lithium acetate method and yeast whole cell lysates were prepared from yeast cells grown to log phase in selective medium and resolved by SDS­PAGE.

The DNA fragment encoding full-length LI1158 was cloned into the mammalian expression vector psin-SFB (psin-EF1 plasmid, tagged with S Protein-Flag-streptavidin-Binding peptide). Plasmid DNA was harvested from E. coli using an endotoxin-free DNA isolation kit (Qiagen, Valencia, CA, USA) and transfected into HEK293T cell monolayers by using Lipofectamine^® 2000 reagent (Invitrogen, Carlsbad, CA, USA). At 48 h after transfection, cells were lysed in lysis buffer (phosphate 20 mM, NaCl 0.5 M, imidazole 20 mM, Triton X-100 2%, pH 7.4) at 4 °C for 1 h, and the purified lysates were resolved by SDS–PAGE.

Bioinformatic analysis

BLAST algorithms were used to identify sequence similarity between proteins. TMHMM 2.0 and InterPro were used to locate Transmembrane domain (TM) and coiled-coil (CC) domain in full-length LI1158/LI1159 proteins.

Y. enterocolitica T3S secretion assays

Secretion assays in Y. enterocolitica were performed as described previously [19]. The supernatant and remaining pellet were separated by centrifugation and analyzed by SDS–PAGE. Proteins were transferred to a polyvinylidene difluoride (PVDF) membranes, which were probed with rat monoclonal anti-HA antibody (CST; 1:2000). Immunoblot detection was performed with secondary antibodies directed against rabbit antibodies and conjugated to horseradish peroxidase (Thermo Fisher) and enhanced chemiluminescence (ECL) chemiluminescent substrate (Proteintech, Wuhan, China).

Yeast growth assays

Yeast growth assays were performed as described previously [20]. Briefly, overnight cultures of recombinant strains grown in SD-Ura medium were washed, and absorbance values were normalized to an optical density of 1.0 at OD₆₀₀. Each strain was serially tenfold diluted for four times and spotted (4 µL) onto repression (2% glucose) or induction (2% galactose) solid selective medium. Yeast cultures were then incubated at 30°C for 2–4 days.

Preparation of yeast extracts and immunoblot analysis

The yeast extract preparation and immunoblot analysis were performed as described previously [20]. Mouse monoclonal anti-Flag antibodies (Proteintech) were used for immunoblotting. Immunoblot detection was performed with secondary antibodies directed against mouse antibodies and conjugated to horseradish peroxidase (Thermo Fisher) and ECL chemiluminescent substrate (Proteintech).

Extract preparation and immunoblot analysis in mammalian cells

The mammalian cell extract preparation and immunoblot analysis were performed as described previously [20]. Briefly, asynchronously growing HEK293T cells were seeded at 2.5 × 10⁵ cells/well in a 6-well tissue culture plate and serum-starved overnight. Two micrograms of plasmid DNA was transfected into HEK293T cells with LipofectamineTM 2000 (Invitrogen) according to the manufacturer’s protocol. Cells were washed with ice-cold PBS containing 1 mM Na₃VO₄ and 10 mM NaF and were then lysed with 300 µL of RIPA buffer containing protease inhibitors. Equal volumes of samples were subjected to SDS-PAGE. Proteins were transferred to PVDF membranes, which were probed with mouse monoclonal anti-Flag antibodies (Proteintech). Immunoblot detection was performed with secondary antibodies directed against mouse antibodies and conjugated to horseradish peroxidase (Thermo Fisher) and ECL chemiluminescent substrate (Proteintech).

Yeast-two-hybrid assays

Combinations of plasmids (bait and prey) were transformed into S. cerevisiae strain Y2H Gold, and the interaction was tested by serial dilution patch tests on the selective medium (Leu-, Trp-; growth control) and low-stringency medium (Leu-, Trp-, His-). The plasmids pGBKT7-p53 (human protein p53) and pGADT7-T (large T antigen of SV40) as the positive control and pGBKT7-Lam (Lamin C) and pGADT7-T as the negative control were co-transformed into yeast cells.

Results

The core T3S apparatus is conserved in L. intracellularis

Genes encoding the T3S apparatus have been identified and named in the complete genome of strain PHE/MN1-00 (accession number AM_180252), which are scattered throughout L. intracellularis genome (Table 3). In order to better analyze the conserved nature of T3SS, we blasted the sequence of T3S components in L. intracellularis PHE/MN1-00 and Y. enterocolitica pYVe8081 (accession number NC_008791.1) with the blast analysis tool at the National Center for Biotechnology Information (NCBI). Although approximately 25 proteins are required to build the T3S apparatus, only genes encoding the core of T3S apparatus proteins conserved in different bacteria and partial orthologous proteins with low sequence conservation have been annotated in the complete genome of L. intracellularis strain PHE/MN1-00 (Table 3) [7]. BLASTP analysis of these proteins revealed that their sequence similarity to the orthologous proteins in Y. enterocolitica T3SS was 24–66%, which was comparable to the conservation among different T3SS-containing pathogens (Table 3). Hence, it can be speculated that the T3SS of L. intracellularis may be crucial for its intracellular development and probably plays a significant role in the virulence of this unusual pathogen.

Table 3 Orthologous proteins between L. intracellularis and Y. enterocolitica in type 3 secretion system.

LI1158 and LI1159 possess the characteristics of T3S major and minor translocon proteins

The assembly of T3S apparatus requires the tip complex and the translocon, which are also T3SS substrates, and the translocon proteins form pores in the plasma membrane of the target host cells [12]. As shown in Figure 1, the translocon operons of different bacterial species generally encode two hydrophobic translocon proteins and a hydrophilic tip protein, which are co-expressed with a class II chaperone and, in some cases, with other regulatory proteins. Similar to the case of these translocon operons, LI1158 (the putative major translocon protein) was co-expressed with LI1157 (the putative chaperone of translocon proteins) and LI1159 (the putative minor translocon protein) from one operon in L. intracellularis genome. However, different from other bacteria, both LI1156 and LI1160, the LI1157 and LI1159 neighboring protein, did not encode the tip protein. Instead, LI1156 encoded a hypothetical protein without conservation with LcrV, while LI1160 encoded YscC, an outer-membrane protein ring subunit of T3SS. Given that LcrV is needed for the insertion of a translocon into the target cell membrane, it is surprising that genome analysis did not annotate the LcrV orthologue in L. intracellularis, which may be attributed to the low sequence homology of L. intracellularis LcrV with its homologous gene. In future studies, it would be interesting to analyze the LcrV protein and its interaction with LI1158 and LI1159 in more detail.

Figure 1

Putative components of the T3SS encoded by the LI1157–1159 operon in L. intracellularis identified by comparison with other bacteria. The arrangement of T3SS translocon operons is highly conserved in pathogenic Gram-negative bacteria, with several well characterized archetypal translocon operons serving as a basis for comparison. LI1157–LI1159 have the characteristics of the typical T3SS translocon operon structure found in a variety of pathogenic bacteria. The role of the putative proteins is indicated using a color code, with arrows denoting gene transcription orientation.

Sequence analysis revealed that LI1158 shared 27% identity with the YopB of Yersinia spp. (Table 3). Further bioinformatic analysis suggested that LI1158 harbors two TM (aa 179–200, 206–223) in its central region, which are flanked by two CC domains (aa 125–170, 334–354) (Figure 2). In addition, an “SseC (SPI-2 effector of Salmonella)-like family” domain was identified in the C-terminal portion of LI1158 (aa 120–308) (30% identity; 52% similarity with a 38% coverage). SseC domains were found in the major hydrophobic translocon proteins of all bacterial T3SSs considered in this study, while they were absent in most minor hydrophobic translocon proteins (Figure 1) [21, 22]. In general, the SseC domains of the translocon proteins displayed some common features, including two or more hydrophobic domains (HD) combined with one or more CC domains. In most cases, the SseC and hydrophobic domains encompassed the C-terminal end of the protein (Figure 2). Furthermore, the PxLxxP motif, which has been identified in translocon proteins and is essential for binding LcrH_1 and a chaperone in C. pneumoniae, was also found in LI1158 (aa 105–110) [22, 23]. Thus, LI1158 exhibited structural similarities to the T3SS major hydrophobic translocon protein.

Figure 2

Domain maps of the minor and major translocon proteins. Comparison of L. intracellularis LI1158/LI1159 with the major/minor translocon proteins from other Gram-negative bacteria. Numbers indicate amino acid positions. The distribution of predicted TM, CC domains and “SseC-like family” within LI1158/LI1159 and the other major/minor translocon proteins is shown schematically. The TM and CC domains of LI1158/LI1159 were predicted by TMHMM 2.0 and Marcoil, respectively. The SseC domain of LI1158/LI1159 was identified by Scansite 4.0 searches. The TM and CC domains of other Gram-negative bacteria were modified from reference [12].

Sequence analysis demonstrated that LI1159 shared 24% identity with the YopD of Yersinia spp. (Table 3). Further bioinformatic analysis suggested that LI1159 harbors one CC domain (aa 204–247) at the C terminus (Figure 2). However, different from other minor translocon proteins which were predicted to contain one TM, LI1159 had no TM. Given that the TM is essential for the insertion of minor translocon proteins into the target cell membrane, further research is needed to determine which amino acids are involved in the insertion of LI1159 into the cell membrane.

Taken together, these bioinformatics analysis results suggested that despite the difference between L. intracellularis and other Gram-negative bacteria, LI1158 and LI1159 may function as the major and minor translocon proteins of T3SS in L. intracellularis.

LI1158 and LI1159 are secreted in a Y. enterocolitica surrogate system with the secretion signal located at the N terminus

Currently, it is urgent but almost impossible to elucidate the function T3SS in L. intracellularis through genetic manipulation due to its obligate intracellular nature. In our previous study, we have constructed an in vitro secretion assay by using the heterologous Yersinia T3SS as an efficacious alternative to evaluate potential T3SS substrates [18]. In this study, we used the same method to determine whether LI1157–LI1159 are T3SS substrates. As described previously, LI1157–LI1159 were expressed in T3S-competent (WT) or -null (∆yscF) Y. enterocolitica as an HA-tagged protein under the control of an arabinose-inducible promoter, respectively. RplJ (an endogenous ribosomal protein) and YopE (an endogenous T3SS substrate) were used as the negative and positive controls, respectively. Cultured cells expressing the recombinant plasmid were centrifuged and fractionated into cell-free culture supernatants and Yersinia-containing whole-cell pellets. Immunoblot analysis of WT cultures demonstrated almost equal specific signals of LI1157–LI1159 in + Ca²⁺ (repressive) and − Ca²⁺ (inductive) cell pellet fractions (Figure 3). However, LI1158 and LI1159 were found in the supernatant under T3SS-inductive conditions but not under T3SS-repressive conditions, which is similar to the case of the positive control YopE. This result was not due to bacterial lysis, because the Yersinia cytoplasmic protein RplJ was detected in whole-cell pellet samples but not in the supernatant fractions (Figure 3). LI1158, LI1159, and the positive control YopE were not detected in the supernatant from the ∆yscF strain, which also validates that the secretion of LI1158 and LI1159 depends on a functional T3SS in Yersinia (Figure 3). Consistent with the known behavior of most T3S chaperones, LI1157 was detected in whole-cell pellet samples but not in the supernatant fractions.

Figure 3

LI1159 and LI1159 are the substrates of the Yersinia type 3 secretion system. T3SS-competent Y. enterocolitica MRS40_∆YopHOPEM and T3SS-null Y. enterocolitica MRS40_∆yscF expressing various HA-tagged proteins (LI1157, LI1158, LI1159, LI1158_N40-YopE_∆N15, LI1159_N40-YopE_∆N15, RplJ, YopE) were cultivated in BHI media under either T3SS-repressive (+ Ca²⁺) or -inductive (− Ca²⁺) conditions. The bacteria were induced to express the T3SS and transgenes by adding 0.2% L-arabinose and incubated at 37 °C for 6 h. Equal amounts of bacteria cultures were separated into cell-free culture supernatants (S) and whole-cell pellets (P), and samples corresponding to 0.10 OD₆₂₀ per mL for S and 0.02 OD₆₂₀ per mL for P were analyzed on 12% (w/v) polyacrylamide gels. Specific proteins were detected by immunoblot with anti-HA followed by visualization with alkaline phosphatase-conjugated secondary antibodies and development with ECL.

To further identify the T3S signal of the LI1158 and LI1159 protein, we analyzed the secretion of the fusion protein consisting of the first 40 aa of LI1158 or LI1159 and YopE_ΔN15 (LI1158_N40-YopE_ΔN15 or LI1159_N40-YopE_ΔN15). As a result, LI1158_N40-YopE_ΔN15 and LI1159_N40-YopE_ΔN15 were detected in the supernatant from WT cultures grown under T3SS-inductive but not under T3SS-repressive conditions (Figure 3). LI1158_N40-YopE_Δ15 and LI1159_N40-YopE_ΔN15 were not detected in the supernatant from the ∆yscF strain, which also confirms that the secretion of LI1158_N40-YopE_Δ15 and LI1159_N40-YopE_ΔN15 is dependent on a functional T3SS in Yersinia (Figure 3). Taken together, these results demonstrated that LI1158 and LI1159, but not LI1157, were exported by the Yersinia T3SS, with the secretion signal being located at the N terminus (aa 1–40).

LI1158 is non-toxic to yeast cells and exists as a stable oligomer/monomer in yeast and mammalian cells

The bioinformatic data and Yersinia secretion assay suggested that LI1158 might be the major hydrophobic translocon protein of L. intracellularis T3SS. Expression of T3SS major translocon proteins such as CPn0809 from Chlamydia pneumoniae was reported to be harmful to yeast cells [22]. To determine whether LI1158 also inhibits yeast growth, flag-tagged LI1158 was induced to be expressed in S. cerevisiae strain W303-1A. Unexpectedly, LI1158 did not cause growth inhibition, which was different from the positive control RipI, a virulent phytopathogenic effector protein [24] (Figure 4A).

Figure 4

LI1158 exists as oligomers/monomers in yeast and mammalian cells, respectively. A Expression of recombinant LI1158 is non-toxic in yeast. W303-1A yeast strains carrying the yeast expression vector pRS416-GAL1, expressing either empty vector or LI1158 with an N-terminal 3 × Flag, were grown overnight in repressing medium (2% glucose). Cultures were then normalized to OD₆₀₀ 1.0, and serial tenfold dilutions were grown at 30°C for 2 and 3 days in repressing and inducing medium (2% galactose and 1% raffinose), respectively. W303-1A yeast carrying the yeast expression vector pYES2/NT-RipI was used as the positive control. B LI1158 exists as oligomers in yeast cells. Induction of 3 × Flag-LI1158 expressions at different time points were verified by Western blotting using an anti-Flag antibody. C LI1158 exists as monomers in mammalian cells. Immunoblots of extracts of HEK293T cells transfected with psin-EF1-puro-SFB or pSIN-EF1-puro-SFB-LI1158 for 48 h. Cell lysates were probed with the anti-Flag tag antibody. Asterisks denote some unknown modification bands.

To determine the expression level of LI1158, yeast cell lysates were prepared and immunoblotted with an anti-FLAG antibody. The results showed that when galactose was not added (t = 0), no band was detected and the cross reaction with the anti-flag antibody could be excluded. With the extension of expression, an individual monomer and DTT- and SDS-resistant oligomers (dimer or tetramer, the molecular weight of the 3 × flag tag fusion LI1158 protein monomer is 47 kDa) appeared (Figure 4B). Surprisingly, two bands of 53–55 kDa occurred in SDS-PAGE, indicating that LI1158 underwent modification when artificially expressed in yeast cells.

However, different from the expression pattern in yeast, when we transfected HEK293T cells with the psin-LI1158-SFB plasmid and subjected the eluted protein to SDS-PAGE/Western blot, LI1158 existed as a monomer (the molecular weight of the S-protein-2 × flag-SBP tag fusion LI1158 protein monomer is 56 kDa) (Figure 4C). Two bands of 70–75 kDa occurred in SDS-PAGE of HEK293T cell lysates, indicating that the monomer of LI1158 also underwent modification when artificially expressed in both mammalian cells.

LI1158 and LI1159 self-interact and LI1159 interacts with LI1157

Previous research has indicated that translocon proteins interact with each other and each pair of major and minor translocon proteins share a common chaperone [7, 12]. Our bioinformatic analysis results also indicated that both LI1158 and LI1159 harbor the CC domains, which could be involved in the structural mechanisms of oligomerization. To further determine whether LI1158 and LI1159 have similar functional features to T3SS translocon proteins from other species, such as oligomerization, chaperone binding, and mutual interaction, we performed Y2H assays.

As shown in Figures 5A and B, Y2H gold yeast strains co-transformed with pGADT7-LI1158/pGBKT7-LI1158 or pGADT7-LI1159/pGBKT7-LI1159 plasmids grew on both plasmid selective medium and low-stringency selective medium, indicating that their interaction activates the expression of the Y2H reporter gene, though the activation was not as strong as that of the positive control (pGBKT7-p53/pGADT7-T). Surprisingly, the yeast cells co-transformed with pGADT7-LI1158/pGBKT7-LI1159 plasmids could not grow on the selective medium, suggesting that LI1158 and LI1159 have no interaction (Figure 5C). Yeast cells co-transformed with pGADT7-LI1157/pGBKT7-LI1159 could grow on the selective medium, but those co-transformed with pGADT7-LI1157/pGBKT7-LI1158 plasmid could not, indicating that LI1157 interacts with LI1159, but not with LI1158 (Figure 5C). Taken together, these results demonstrated that LI1158 or LI1159 is able to interact with itself and LI1159 can interact with the putative T3S-related class II chaperone LI1157.

Figure 5

LI1158 and LI1159 self-interact and LI1159 interacts with LI1157. A–C Different combinations of plasmids (bait and prey) were transformed into yeast strain Y2H Gold, and then streaked on selective medium (Leu-, Trp-; growth control) and low stringency medium (Leu-, Trp-, His-), respectively. The plasmids pGBKT7-p53 (human protein p53) and pGADT7-T (large Tantigen of SV40) were co-transformed into yeast cells as positive control. Then, self-activation was carried out to overrule the interaction with vectors.

Expression of recombinant LI1159, but not LI1158, is toxic to E. coli

Expression of T3SS major translocon proteins, such as CPn0809 from C. pneumoniae and YopB from Yersinia, was reported to be toxic to E. coli [15, 22]. To determine whether LI1158 and LI1159 also have a growth-inhibiting effect on E. coli, we utilized the LI1158-HA and LI1159-HA plasmid and tested them in comparison to the empty vector in a time-course experiment. Under non-inducing conditions, the growth rates of the three expression cultures were almost identical (Figure 6A). Upon induction of protein expression, the culture expressing either the empty vector or LI1158 protein continued to grow. In contrast, the growth of cultures expressing LI1159 protein was slow (4 h) or even stopped, indicating that LI1159 expression completely inhibited the growth of E. coli cells (Figure 6A). The LI1158-HA and LI1159-HA expression could be confirmed by using Western blot experiments (Figure 6B). To clarify whether LI1159 expression causes the growth arrest or actually kills the cells, we spotted cells from cultures after 1 h of protein expression and non-induced control cultures at different dilutions on plates (non-inducing medium). Cells from the three non-induced cultures showed an identical regular growth pattern on the medium. In contrast, E. coli cells induced to express LI1159 showed slower growth on the medium than cells expressing LI1158 and the empty vector (Figure 6C). Taken together, our findings demonstrated that LI1159 has a cytotoxic effect when expressed in E. coli cells, but LI1158 does not.

Figure 6

Expression of LI1159 is toxic to E. coli. A Growth curves of E. coli harboring different plasmid constructs under inducing or non-inducing conditions. Expression of LI1158-HA and LI1159-HA was induced by the addition of l-arabinose (+ l-arabinose, red mark). The mean of three independent replicates was used to create the growth curves by GraphPad Prism 7.0. B LI1158/LI1159-HA was expressed upon induction in E. coli cells. Induction of LI1158/LI1159-HA expressions were verified by Western blotting using an anti-HA antibody. C The toxic effect of LI1159 expression in E. coli was confirmed in serial dilution patch tests of E. coli transformants shown in A. Liquid cultures were grown for equal time periods and expression of proteins was induced by the addition of l-arabinose for 1 h. Ten microgram samples were taken and diluted ranging from 10^–1 to 10^–4. Dilutions were spotted onto solid LB-media with 1% glucose (repressing condition) and incubated overnight at 37°C.

Expression of recombinant LI1158 and LI1159 is non-toxic to Y. enterocolitica cells

Although LI1159, but not LI1158 was found to have a cytotoxic effect on E. coli cells, our Yersinia T3SS secretion assay indicated that LI1158/LI1159 is non-toxic to Y. enterocolitica cells (Figure 3), which prompted us to perform the time-course experiment and the serial dilution patch test in Y. enterocolitica. As expected, under both non-induced and induced conditions, the growth rates of the three expression cultures were almost identical (Figure 7A). Our western blot experiments also confirmed the expression of LI1158-HA and LI1159-HA in these growth curve experiments (Figure 7B). Cells from the three non-induced cultures and three induced cultures showed an identical regular growth pattern on plates (Figure 7C). Taken together, these results demonstrated that the expression of recombinant LI1158 and LI1159 is non-toxic to Y. enterocolitica cells.

Figure 7

Expression of recombinant LI1158/LI1159 is non-toxic for Y. enterocolitica cells. The growth curves (A), Western blot experiment (B) and serial dilution patch test (C) in Y. enterocolitica cells were obtained in the same way as in E. coli cells.

Discussion

Many Gram-negative bacteria pathogenic to plants and animals deploy T3SS to inject effector proteins into their hosts [7]. The finding that L. intracellularis may encode T3S components has stimulated particular interests in elucidating when and how L. intracellularis uses the T3S mechanism to deploy proteins to modulate host cell pathways like other Gram-negative pathogens [16]. However, it has been challenging to verify whether this secretion pathway is functional in L. intracellularis due to its obligate intracellular nature and genetic intractability. T3SS translocon is assembled by two hydrophobic membrane proteins, namely the major and minor translocon proteins at the tip of the T3SS needle, which form a pore in the plasma membrane of target cells. Currently, little information is available on the T3S components and translocon proteins of L. intracellularis.

A phylogenetic analysis of four conserved T3S apparatus proteins has revealed that the T3S apparatus of Desulfovibrio vulgaris, one of the three family members in Desulfovibrionaceae, belongs to the Ysc T3S apparatus of Yersinia spp., which is regarded as the archetype of T3SS on the plasmid [7]. L. intracellularis, another family member in Desulfovibrionaceae, possesses T3S components scattered throughout its genome (Table 3). Our BLASTP analysis revealed the sequence similarity between the orthologous proteins in the T3SS of Y. enterocolitica and L. intracellularis, further confirming that T3SS plays an important role in the intracellular pathogenesis of L. intracellularis.

In this study, we characterized the putative translocon proteins LI1158 and LI1159 in L. intracellularis, which are highly expressed in the enterocyte cytoplasm during infection [17]. Our bioinformatics analysis suggested the presence of common structural features of major translocon proteins, such as hydrophobic domains surrounded by CC structures, the SseC-like family domain, and the PxLxxP motif, in L. intracellularis LI1158 (Figure 2). However, LI1159, a putative minor translocon protein, only harbors the CC domain at its C terminus and has no TM. Considering its functions in the formation of a pore in the plasma membrane of target cells, further research is needed to detect and verify whether LI1159 contains hydrophobic domains.

Since translocon proteins are the substrate of T3SS, we employed the Yersinia T3SS as a heterogenous system to determine whether L. intracellularis LI1157–LI1159 are the substrate of T3SS. We found that LI1158 and LI1159, but not LI1157, are two novel substrates that have the potential to be secreted into culture supernatants, whose specific secretion signals located at the N terminus are recognized by Y. enterocolitica (Figure 3). Further research can be conducted to determine whether LI1158 and LI1159 have the capability of insertion into the target mammalian cell membrane by using Y. enterocolitica as a heterologous system.

Expression of T3SS major translocon proteins, such as CPn0809 from C. pneumoniae, was reported to be toxic to yeast cells [22]. Surprisingly, LI1158 showed no inhibitory effect on the growth of yeast cells in this study. Major translocon proteins are capable of forming homo- and/or hetero-oligomers [6]. In our Yersinia secretion assays, LI1158 did not form oligomers in the supernatant and the pellets. However, when heterogeneously expressed in S. cerevisiae, LI1158 existed as oligomers. When heterogeneously expressed in mammalian cells, LI1158 existed as monomers and underwent an unknown modification as in budding yeast. Further investigation is required to uncover the mechanism for the modification of LI1158 in yeast and mammalian cells.

A Y2H assay was performed to further verify whether LI1158 and LI1159 share common functional features with T3SS translocon proteins from other pathogenic species. We found that LI1158 and LI1159 are able to self-interact and LI1159 can interact with LI1157. However, LI1158 showed no interaction with LI1159 and LI1157. Given the importance of such interactions and their universality in other pathogens, these interactions need to be detected and verified in future experiments.

Finally, we analyzed the growth-inhibiting effect of LI1158 and LI159 on E. coli and Y. enterocolitica cells. Surprisingly, it was the minor translocon protein (LI1159) instead of the major translocon protein (LI1158) that led to E. coli cell death [15, 21]. These results also implied certain differences of L. intracellularis from other bacteria. Moreover, LI1159 did not inhibit the cell growth when expressed in Yersinia, which is totally different from its effect on E. coli cells.

Overall, despite certain differences from other bacteria, our results support the assumption that LI1158 and LI1159 are respectively the major and minor translocon proteins of L. intracellularis T3SS. Our findings improve understanding of the T3S components and will facilitate the development of more specific serological diagnostic assays and effective vaccines to control PE outbreak.