Functional insight from the tetratricopeptide repeat-like motifs of the type III secretion chaperone SicA in Salmonella enterica serovar Typhimurium



SicA functions both as a class II chaperone for SipB and SipC of the type III secretion system (T3SS)-1 and as a transcriptional cofactor for the AraC-type transcription factor InvF in Salmonella enterica subsp. enterica serovar Typhimurium. Bioinformatic analysis has predicted that SicA possesses three tetratricopeptide repeat (TPR)-like motifs, which are important for protein–protein interactions and serve as multiprotein complex mediators. To investigate whether the TPR-like motifs in SicA are critical for its transcriptional cofactor function, the canonical residues in these motifs were mutated to glutamate (SicAA44E, SicAA78E, and SicAG112E). None of these mutants except SicAA44E were able to activate the expression of the sipB and sigD genes. SicAA44E still has a capacity to interact with InvF in vitro, and despite its instability in cell, it could activate the sigDE operon. This suggests that TPR motifs are important for the transcriptional cofactor function of the SicA chaperone.


Salmonella species are the causative agents of several diseases, ranging from self-limiting diarrhea to life-threatening typhoid fever (Ruby et al., 2012). The virulence genes important for Salmonella pathogenesis are clustered in several Salmonella pathogenicity islands (SPIs). SPI-1 and SPI-2 each encode a type III secretion system (T3SS), which are employed to transport effector proteins directly into the host cytosol. The T3SS are composed of more than 20 proteins: structural proteins, secreted proteins, chaperones, and regulators (Galan & Wolf-Watz, 2006; McGhie et al., 2009).

T3SS chaperones are normally small, acidic, and helical structures (Parsot et al., 2003). Class I chaperones bind to one or more effector proteins, whereas class II chaperones generally bind at least two translocator proteins. Finally, class III chaperones act on flagellar proteins. Class II chaperones are characterized by the presence of three tetratricopeptide repeat (TPR) motifs, which are important for protein–protein interactions (Pallen et al., 2003; Buttner et al., 2008; Lunelli et al., 2009; Izore et al., 2011). Each TPR motif is composed of an antiparallel (helix–turn–helix) tandem array of 34 amino acids with small and/or hydrophobic residues at positions 8 (A or G), 20 (A), and 27 (A); these residues are essential for the binding capabilities of the TPR motif (D'Andrea & Regan, 2003; Magliery & Regan, 2004). The peptide-binding groove formed by multiple helical structures is predicted to be the interaction domain for two translocator proteins (Edqvist et al., 2006; Job et al., 2010).

SicA, a class II chaperone, contains eight alpha-helical regions (H1–H8) and has three putative TPR motifs (H2/H3, H4/H5, and H6/H7), with two additional helices at the N-terminus (H1) and another helix at the C-terminus (H8; Pallen et al., 2003). SicA directly interacts with the domain between N-terminal 80–100 amino acids of translocator SipB to prevent its premature association with SipC and their subsequent degradation (Tucker & Galan, 2000; Kim et al., 2007). SipB and SipC are involved in epithelial cell invasion, contact hemolysis, and macrophage cell death (Kaniga et al., 1995; Tucker & Galan, 2000; McGhie et al., 2009). After SipB and SipC are secreted by the SPI-1 T3SS, SicA is free to bind InvF, which is an AraC/XylS family activator of the SPI-1 T3SS, to activate the transcription of the sicAsipBCDAiacPorfXsicPsptP operon and other effector genes, including sigDE and sopE (Darwin & Miller, 2000, 2001). However, it is still unknown which region or specific conformation of SicA is responsible for this transcriptional activation.

In this study, we substituted the canonical residues of the three TPR motifs in SicA and tested whether these mutations could rescue the low transcription of sipB and sigDE observed in a sicA mutant. We demonstrate that the mutation of any of the TPR motifs of SicA greatly affects the transcription of these two genes except SicAA44E, which retain the ability to activate sigDE operon. Furthermore, SicAA44E could bind InvF regardless of dramatic changes in its surface charge.

Materials and methods

Bacterial strains and growth conditions

The S. Typhimurium strains used in this study are listed in Table 1 and were grown at 37 °C in LB broth containing 0.3 M NaCl (SPI-1-inducing conditions). When required, antibiotics were added at the following concentrations: ampicillin, 100 μg mL−1; kanamycin, 50 μg mL−1; and chloramphenicol, 30 μg mL−1. For the induction of genes from plasmids, isopropyl 1-thio-β-d-galactopyranoside (IPTG) and l-arabinose were used at 0.1 mM.

Table 1. Bacterial strains and plasmids used in this study
UK-1Salmonella enterica serovar Typhimurium, wild typeMoreno et al. (2000)
YKJ005UK-1, ΔsicAThis study
YKJ042UK-1, sigD::3 × FLAG; KmrKim et al. (2011)
YKJ043UK-1, ΔsicA sigD::3 × FLAG; KmrThis study
pKD46pSC101. PBAD-gam bet exo, oriTS; AprDatsenko & Wanner (2000)
pKD3FRT-cat-FRT, oriR6K; Apr, CmrDatsenko & Wanner (2000)
pCP20pSC101. cI857 λPR flp oriTS; AprDatsenko & Wanner (2000)
pBAD24Expression plasmid containing PBAD; AprGuzman et al. (1995)
pSicApBAD24, PBAD-sicA::FLAG; AprThis study
pSicAA44EpBAD24, PBAD-sicAA44E::FLAG; AprThis study
pSicAA78EpBAD24, PBAD-sicAA78E::FLAG; AprThis study
pSicAG112EpBAD24, PBAD-sicAG112E::FLAG; AprThis study
pGEX-KGGST fusion plasmid; AprGuan & Dixon (1991)
pGST-InvFpGEX-KG, Ptac-GST::invF; AprThis study
pET-28a(+)His fusion plasmid; KmrNovagen
pHis-SicApET28-a(+), Ptac-(His)6::sicA; KmrThis study
pHis-SicAA44EpET28-a(+), Ptac-(His)6::sicAA44E; KmrThis study
pHis-SicAA78EpET28-a(+), Ptac-(His)6::sicAA78E; KmrThis study
pHis-SicAG112EpET28-a(+), Ptac-(His)6::sicAG112E; KmrThis study

Construction of S. Typhimurium mutant strains

All S. Typhimurium strains and primers used in this study are listed in Tables 1 and 2, respectively. To disrupt sicA, the Cmr gene was PCR-amplified from the pKD3 plasmid with the primer pairs mutsicA-L/mutsicA-R, and then, the resulting PCR product was electroporated into the S. Typhimurium strain UK-1 carrying pKD46. Finally, the integrated Cmr gene was removed using the pCP20 plasmid to generate the ΔsicA mutant. The identities of the constructs were verified by colony PCR. The ΔsicA sigD::3 × FLAG strain was generated by P22-mediated transduction of the sigD::3 × FLAG allele from YKJ042 into the ΔsicA mutant.

Table 2. Primers used in this study
PrimersOligonucleotide sequences 5′–3′
Mutant strain construction
 Flip recognition target (FRT) sequences are italicized
Plasmid construction
 Underlined sequences represent restriction enzyme recognition sites

Plasmid construction

The plasmid and primer sequences are listed in Tables 1 and 2, respectively. The pSicA-FLAG plasmid was generated by cloning a PCR-amplified fragment of the full-length sicA open reading frame from S. Typhimurium into the EcoRI/SacI sites of the pBAD24 plasmid. Amino acid substitutions were introduced into the pSicA-FLAG vector using a QuikChange site-directed mutagenesis kit (Stratagene) according to the manufacturer's instructions to generate three mutations: A44E, A78E, and G112E. The resulting PCR-amplified plasmids were treated with DpnI and transformed into Escherichia coli DH5α. For the in vitro binding assay, plasmids encoding hexahistidine-tagged wild-type and mutated SicA were constructed by cloning PCR-amplified fragments into the pET-28a vector. The GST-InvF plasmid was generated by cloning the invF gene into the pGEX-KG vector. The sequences and expected mutations of the resulting plasmids were verified by DNA sequencing.

Total RNA isolation, cDNA synthesis, and qRT-PCR

Total RNA was isolated from S. Typhimurium grown under SPI-1-inducing conditions using the RNeasy plus mini kit (Qiagen) following the manufacturer's protocol. First-strand cDNA synthesis (from 1 μg RNA) was performed using random hexamer primers and M-MLV reverse transcriptase (Promega). The absence of contaminating genomic DNA in the reactions was verified by conducting the cDNA synthesis reaction without the reverse transcriptase enzyme. The qRT-PCR analyses were performed using the LightCycler 480 system (Roche Applied Science, Mannheim, Germany) with SYBR green (SYBR green Master Mix; Roche), each of the primers listed in Table 2, and sample cDNA. The PCR conditions were as follows: 95 °C for 5 min, followed by 45 cycles of 95 °C for 10 s, 60 °C for 10 s, and 72 °C for 10 s. Data were analyzed with LightCycler 480 software (Roche), and relative expression was calculated using the ‘delta–delta CP’ method.

Analysis of whole-cell lysates and secreted proteins

Bacterial strains were grown under SPI-1-inducing conditions at 37 °C to an OD600 nm of 1.1 for SipB or 1.4 for SigD analysis. Cells were pelleted by centrifugation, and the resulting pellets were resuspended in SDS sample buffer and boiled for 10 min. The supernatant was also collected and filter-sterilized through a 0.22-μm pore filter, precipitated with 10% trichloroacetic acid at 4 °C, washed with cold acetone, and resuspended in SDS sample buffer. Whole-cell lysates and supernatant samples were subjected to SDS-PAGE and analyzed by Western blotting with anti-SipB (Kim et al., 2007), anti-SigD (Santa Cruz Biotechnology), anti-FLAG (Sigma Aldrich), or anti-DnaK (Enzo Life sciences) antibodies. Blots were incubated with HRP-conjugated goat anti-rabbit or goat anti-mouse IgG (both from Bio-Rad) secondary antibodies and were developed using the BM chemiluminescence blotting substrate (Roche).

Native gel electrophoresis

For native gel electrophoresis, purified SicAWT and SicAA44E proteins were mixed with a native sample buffer with or without Coomassie Brilliant Blue G250 and were subsequently run on a 4–15% gel at 100 V for 1 h in Tris-glycine buffer (Bio-Rad). Gels were stained with Coomassie Blue.

In vitro binding assay

Plasmids encoding GST-tagged constructs and His-tagged constructs were transformed into E. coli BL21 (DE3) pLysS, and expression of the fusion proteins was induced by 0.1 mM IPTG for 3 h at 25 °C. The cells were washed with RIPA buffer (50 mM Tris-HCl, pH 7.4, 1% NP-40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM EDTA, and protease inhibitor cocktail; Roche) and disrupted by sonication. To increase protein solubility, 1% sarcosine was added to the RIPA buffer before sonication. The lysates containing GST or GST-tagged proteins were mixed with glutathione–Sepharose beads for 1 h at 4 °C. The beads were washed with RIPA buffer and incubated with His-tagged SicA proteins for 3 h at 4 °C with gentle rocking. The protein complexes were pulled down with the beads, washed, and resuspended in SDS sample buffer. Samples were subjected to SDS-PAGE and Western blotting with anti-GST and anti-His antibodies (both from Santa Cruz Biotechnology).

Statistical analysis

Data were analyzed by one-way anova followed by Tukey's post hoc test using spss software to determine the statistical significance (< 0.05).


TPR motifs are important for SicA transcriptional cofactor function

The class II chaperone SicA has three putative TPR-like motifs similar to other class II chaperones, including SycD (LcrH), PcrH, and IpgC in Yersinia, Pseudomonas, and Shigella, respectively (as determined by a sequence alignment; Pallen et al., 2003; Broms et al., 2006; Edqvist et al., 2006; Buttner et al., 2008; Izore et al., 2011). Upon mutation of the canonical residues in these TPR-like motifs, SycD and PcrH became unstable and failed to protect their cognate substrates from degradation (Broms et al., 2006; Edqvist et al., 2006). To test whether similar point mutations in the putative TPR motif of SicA could result in the same consequences, we mutated the eighth residue of each TPR motif to glutamate (A44E, A78E, and G112E) as described in Fig. 1a. A mutant of SycD with point mutations in the TPR motif had poor stability when expressed with its native promoter (Edqvist et al., 2006); similarly, all mutated forms of SicA under the control of its native promoter were present at substantially lower levels than wild-type SicA (data not shown). Therefore, we constructed new plasmids in which the wild-type sicA and the sicA mutants expressing mutated SicA proteins were placed under the control of the arabinose-inducible promoter (PBAD).

Figure 1.

TPR motifs are important for SicA transcriptional cofactor function. (a) Secondary and tertiary structures of SicA constructed with Phyre2. Asterisks indicate the canonical TPR sites, which are point-mutated to glutamate in this study. (b) The wild-type, ΔsicA mutant, and the isogenic strains expressing the wild-type (SicAWT) or point-mutated SicA proteins (SicAA44E, SicAA78E, and SicAG112E) were grown under SPI-1-inducing conditions, and SicA expression was induced by the addition of arabinose for 1 h. Total RNA was isolated from the bacteria, and the levels of sipB and sigD mRNAs were determined by qRT-PCR analysis. The graph represents the relative mRNA levels for the sipB and sigD. (c) To detect the levels of SipB and SigD, the plasmids were introduced into Salmonella Typhimurium strains expressing chromosomally tagged SigD::3 × FLAG. Bacterial strains were grown for 3 h under SPI-1-inducing conditions. Whole-cell lysates and culture supernatants were analyzed by immunoblotting with anti-SipB, anti-FLAG, and anti-DnaK antibodies. The cytoplasmic protein DnaK was used as a control for cytoplasmic contamination of the culture supernatants.

Because SicA acts as a transcriptional cofactor on genes within SPI-1, we investigated whether the mutant SicA could activate the transcription of the sicAsipBCDAiacPorfXsicPsptP and sigDE operon. Thus, we isolated mRNA from S. Typhimurium stains grown under SPI-1-inducing conditions and examined the relative amounts of sipB and sigD expression by qRT-PCR analysis. As noted in earlier studies, the relative expressions of the sipB and sigD genes were down-regulated in a sicA null-mutant strain, indicating that SicA is required for efficient transcriptions of the sicAsipBCDA and sigDE operons (Fig. 1b; Darwin & Miller, 2000; Tucker & Galan, 2000). qRT-PCR analysis showed that the mutation of the TPR motifs of SicA was no more able to complement the defect of the sicA mutant to transcribe the sipB gene contrary to the wild-type SicA protein. Moreover, similar results were obtained for the sigD gene except for SicAA44E, albeit at little lower levels than wild-type SicA (Fig. 1b). These results suggest that point mutation in the canonical residues of the TPR motifs of SicA greatly affects the role of SicA as a transcriptional cofactor.

We next examined the protein level of SipB and SigD using strains expressing a C-terminal 3 × FLAG-tagged SigD. As shown in Fig. 1c, SipB was not detected in whole-cell lysates or culture supernatants of S. Typhimurium expressing the mutant SicA. In addition, the SigD-3 × FLAG protein was expressed and secreted in the presence of SicAA44E, whereas SicAA78E and SicAG112E were unable to induce its expression, because no SigD-3 × FLAG proteins are detected in the supernatant of the strain expressing mutated SicA proteins SicAA78E and SicAA112E (Fig. 1c). These findings are consistent with qRT-PCR results shown in Fig. 1b and demonstrate that substitution of the canonical residue in the first TPR motif of SicA does not affect the activation of the sigDE operon. Collectively, these results indicate that TPR motifs of SicA play an important role for the transcriptional activation of T3SS-related genes by InvF even if SicAA44E is able to function as a co-activator of transcription as like wild-type SicA.

TPR mutants of SicA interact with InvF

The transcription of the sigDE operon requires both InvF and SicA, possibly through direct InvF–SicA binding (Darwin & Miller, 2001). Because SicAA44E initiated sigD transcription, we hypothesized that SicAA44E interacted with InvF as efficiently as wild-type SicA. In an in vitro binding assay, both hexahistidine-tagged SicAWT and SicAA44E were coprecipitated with GST-InvF, whereas they were unable to interact with GST alone (Fig. 2a). Interestingly, even SicAA78E and SicAG112E mutants could interact with InvF, but it was weaker than the wild-type SicA (Fig. 2b).

Figure 2.

TPR mutants of SicA interact with InvF. Escherichia coli BL21 (DE3) pLysS strains harboring plasmids encoding GST, GST-InvF, and His-SicA (wild-type or point-mutated SicA) were incubated until the OD600 nm reached 0.6, when expression of target proteins was induced with 0.1 mM IPTG for 3 h at 25 °C. To solubilize GST-InvF, sarcosine (1%, v/v) was added to RIPA buffer before cell lysis. Sepharose 4B GST beads were used to pull down GST or GST-InvF, and then, the lysates of His-SicA-expressing bacteria were added. Proteins that were pulled down by the GST beads were analyzed by immunoblotting with anti-GST (upper) and anti-His (lower) antibodies. ‘In’ indicates the input of the bacterial lysate containing His-SicAA44E (a) or His-SicAA78E and His-SicAG112E (b).

SicAA44E has different surface charges and is less stable than wild-type SicA

To compare the protein stability of SicAWT-FLAG and SicAA44E-FLAG, S. Typhimurium strains containing the wild-type and mutated sicA open reading frame were grown under SPI-1-inducing conditions until mid-log-phase, and SicA expression was induced by adding l-arabinose to these cultures for 30 min. After adding chloramphenicol to inhibit bacterial protein synthesis, the intracellular levels of SicA were monitored. As shown in Fig. 3a, the expression levels of mutant and wild-type SicA induced by arabinose were similar. Wild-type SicA (SicAWT-FLAG) was detectable up to 60 min, whereas mutant SicA (SicAA44E-FLAG) levels dramatically decreased within 10 min after chloramphenicol treatment (Fig. 3a). The instability of the mutant SicA is consistent with the previous reports about SycD from Yersinia and PcrH from Pseudomonas (Broms et al., 2006; Edqvist et al., 2006).

Figure 3.

Protein stability and surface charge of SicAA44E. (a) The ΔsicA mutant strains harboring the plasmids expressing the wild-type (SicAWT) or point-mutated SicA (SicAA44E) were grown under SPI-1-inducing conditions. Protein expression was induced by the addition of arabinose for 30 min, and chloramphenicol was added to the cultures to inhibit de novo protein synthesis. Samples were collected at the indicated time points, and whole-cell lysates were subjected to Western blotting with anti-FLAG or anti-DnaK antibodies. DnaK was used as a loading control. (b) SDS and non-SDS gel electrophoresis of SicAWT and SicAA44E. For SDS-PAGE, purified His-tagged SicAWT and SicAA44E were mixed with SDS sample buffer, boiled, and subjected to 12% SDS-PAGE. For native PAGE, purified proteins were mixed with equal volumes of native sample buffer with or without Coomassie Brilliant Blue G250 and were run on a 4–15% gradient gel under nondenaturing conditions. Gels were stained with Coomassie Blue. Arrows indicate the position of SicAWT or SicAA44E.

It has been reported that TPR-containing proteins can form homo-oligomers and that this oligomerization is crucial for proper protein function in the cell. Therefore, we examined the oligomerization status of wild-type and mutant SicA on a native PAGE gel. Interestingly, SicAA44E had a similar electrophoretic mobility as the wild-type protein did when both were loaded with Coomassie Blue G250, which equalizes a surface negative charge of each protein without inactivating the proteins (Fig. 3b). However, without Coomassie Blue G250, only the SicAWT protein ran on the native gel. This finding may indicate that the net surface charge of SicA was dramatically changed by the point mutation introduced into its TPR region. This result showed that the substitution of a canonical residue in the first TPR motif of SicA changed the net charge of protein, but did not affect its interaction with InvF.


We have shown here that SicA, a class II chaperone, has three canonical TPR motifs that are critical for the transcription of sipB and sipC genes encoding the T3SS-1 translocases. To determine the implication of TPR motif in protein function, the three mutations were introduced into the canonical residue of each TPR motif of SicA (at positions 44, 78, and 112). Their replacement by the large glutamic acid could prevent the tight packing of the TPR helices leading to misfolding of SicA (Edqvist et al., 2006). S. Typhimurium strains expressing SicA with mutations in the TPR-like motifs (A44E, A78E, and G112E) could not restore the transcription of sipB and sipD genes. Also, we showed that SicA TPR motifs were not necessary for binding to InvF. Among the three mutated forms of SicA studied here, only SicAA44E was able to activate the transcription of the late effector protein SigD. Collectively, these results suggest that the TPR motif in SicA serves two important but disparate functions.

Because the TPR motif was originally identified in yeast cell cycle proteins as a protein–protein interaction domain, it has been found in organisms ranging from bacteria to humans (Blatch & Lassle, 1999). TPR motif-containing proteins act as scaffolds for the assembly of different multiprotein complexes (Zeytuni & Zarivach, 2012). In pathogenic Gram-negative bacteria containing T3SSs, it has been suggested that class II chaperones including SicA stabilize early secretion substrates, such as translocators that polymerize or oligomerize to form a functional T3SS (Izore et al., 2011). In addition to its chaperone function, SicA acts as a cofactor for the InvF-dependent transcriptional activation of the sicAsipBCDAiacPorfXsicPsptP and the sigDE operons by binding directly to InvF (Darwin & Miller, 2000, 2001). Prior to this study, it was unknown whether the TPR motifs in SicA were required for the transcription of these two operons and its interaction with InvF. Here, using recombinant mutant proteins and native PAGE, we showed that mutations in conserved residues within the TPR-1 motif caused dramatic changes in the net surface charge of SicA. Given that SicAA44E ran differently from the wild-type SicA in a native PAGE gel when the charges were not neutralized, it is likely that SicAA44E has a different structure than SicAWT. Although this change makes SicAA44E protein to lose its ability to regulate SipB production, SicAA44E was still able to activate the sigD transcription, suggesting SicAA44E–InvF complex is compatible to the functional transcription units. However, in case of SicAA78E and SicAG112E mutants, they also could form complex with InvF, but could not activate the sigD transcription. A44E, A78E, and G112E are mutants to the H2/H3, H4/H5, and H6/H7 regions, respectively. From the result, we can speculate that the specific structure by H4/H5 and H6/H7 is indispensable for the InvF activation to induce the sigDE operon, but the first TPR motif (H2/H3) is dispensable. Although the recently elucidated crystal structure of SicA suggests that it adopts a different conformational stoichiometry compared to other class II chaperones, including IpgC in Shigella and SycD in Yersinia (Priyadarshi & Tang, 2010), more work must be carried out to determine how this structure is related to the functions of SicA.

SicA has dual functions in activating the transcription of two T3SS effector operons; it may execute these functions by modulating its structure. Therefore, further studies of the factors that control these structural changes are critical. Our data may support the functional complexity of the chaperone SicA in the regulation of the T3SS (Tucker & Galan, 2000) through insight from its TPR motifs. Furthermore, defining additional functions of SicA will aid in our understanding of the survival mechanisms of Gram-negative intracellular pathogens. Finally, because the substrate-binding surface of class II chaperones and the transcriptional activation of virulent effectors are distinct from substrate-binding TPRs in eukaryotes, identifying a ligand that disrupts the specific interactions with SipB or InvF could contribute to the development of specific and potent antibacterial agents.


This research was supported by a grant from the Korea Healthcare Technology R&D Project, Ministry for Health, Welfare & Family Affairs, Republic of Korea (A090891). The authors declare that there are no conflicts of interest.

Authors’ contribution

J.S.K. and B.-H.K. contributed equally to this work.