Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Nov 28.
Published in final edited form as: Arch Microbiol. 2018 May 8;200(7):1075–1086. doi: 10.1007/s00203-018-1519-x

The transmembrane domain of the Staphylococcus aureus ESAT-6 component EssB mediates interaction with the integral membrane protein EsaA, facilitating partially regulated secretion in a heterologous host

Manar M Ahmed 1, Khaled M Aboshanab 2, Yasser M Ragab 3, Dominique M Missiakas 4, Khaled A Aly 1
PMCID: PMC9704473  NIHMSID: NIHMS1850213  PMID: 29737367

Abstract

The ESAT-6-like secretion system (ESS) of Staphylococcus aureus plays a significant role in persistent infections. EssB is a highly conserved bitopic ESS protein comprising a cytosolic N-terminus, single transmembrane helix and a C-terminus located on the trans-side of the membrane. Six systematic truncations covering various domains of EssB were constructed, followed by bacterial two-hybrid screening of their interaction with EsaA, another conserved integral membrane component of the ESS pathway. Results show that the transmembrane domain of EssB is critical for heterodimerization with EsaA. In vivo crosslinking followed by Western blot analysis revealed high molecular weight species when wild-type EssB and EsaA were crosslinked, but this band was not detected in the absence of the transmembrane domain of EssB. Heterologous overproduction of EssB, EsaA and five other components of the ESS pathway in Escherichia coli BL21(DE3), followed by fractionation experiments led to a remarkable increase in the periplasmic protein content, suggesting the assembly of partially regulated secretion mechanism. These data identify the transmembrane domain of EssB as indispensable for interaction with EsaA, thereby facilitating protein secretion across bacterial membranes in a fashion that requires other components of the ESS pathway.

Keywords: EssB–EsaA interaction, Bacterial two-hybrid, T7SS, ESAT-6

Introduction

Staphylococcus aureus poses a serious threat to the global healthcare system due to its significant pathogenic potential (Thammavongsa et al. 2015; Ventola 2015). Naturally, S. aureus colonizes one-third of the global population as a commensal, mostly inhabiting their skins and nares (van Belkum et al. 2009). Quite often however, this commensal lifestyle transitions into a pathogenic one, resulting in skin and soft tissue infections, but also in more invasive attacks that can be life threatening, such as bacteremia, sepsis and pneumonia (David and Damn 2010; Thammavongsa et al. 2015). Staphylococcal invasiveness is accompanied by the disruption of both, innate and adaptive immune responses, as well as the induction of several virulence mechanisms, such as the ESAT-6-like Secretion System (ESS) (Jongerius et al. 2012; Falugi et al. 2013; Anderson et al. 2017).

The ESS pathway is encoded by a cluster of at least 12 genes or more that produce an assortment of protein products, ranging from solely cytosolic, to membrane-associated and secreted effectors (Fig. 1) (Warne et al. 2016). The ESS system is often termed type 7b secretion system (T7bSS) and is predominantly found in S. aureus and other firmicutes, such as Bacillus subtilis, B. anthracis, Listeria monocytogenes, and Streptococcus agalactiae (Garufi et al. 2008; Anderson et al. 2013; Sysoeva et al. 2014; Cao et al. 2016). In contrast, the distant evolutionary relative, T7aSS, is found in actinobacteria, the prototypical system initially characterized in Mycobacterium tuberculosis (Abdallah et al. 2007). The first five genes of the ESS cluster code for a secreted effector, EsxA, two membrane proteins, EsaA and EssA, a cytosolic protein, EsaB and lastly, another membrane protein, EssB. These five genes reside within a conserved module that is present in many staphylococcal isolates and other firmicutes, such as Listeria monocytogenes and Streptococcus agalactiae (Fig. 1) (Abdallah et al. 2007; Warne et al. 2016). The seven remaining genes of the ESS cluster code for a membrane integrated ATPase (EssC), three additional effectors (EsxC, EsxB and EsxD), as well as a membrane platform protein (EssE), a nuclease effector (EssD) and its cognate inhibitor (EssI).

Fig. 1.

Fig. 1

Open in a new tab

Genetic arrangement of the ESS cluster in Gram-positive microorganisms. The ESS cluster from four S. aureus clinical isolates, S. agalactiae and L. monocytogenes (Abdallah et al. 2007; Warne et al. 2016). The conserved ESS module shared by all these isolates is indicated by the arrow (Warne et al. 2016). The proposed role of every gene is outlined in a color-coded fashion as indicated.

The secreted effector EsxA belongs to the WXG superfamily of proteins that is characterized by the presence of a tryptophan-X-glycine motif roughly in the middle of the protein sequence (Poulsen et al. 2014). Members of the WXG superfamily have been identified and characterized in both, T7a and T7b secretion systems, and have the natural tendency to engage in homodimeric interaction as well as heterodimeric interactions with other WXG proteins (Champion et al. 2009; Anderson et al. 2013). The crystal structure of EsxA reveals two α-helices folded into a hairpin bend around the WXG motif (Sundaramoorthy et al. 2008). In S. aureus, EsxA interacts with itself and with EsxC, another effector of the ESS pathway (Fig. 1) (Sundaramoorthy et al. 2008) (Anderson et al. 2013). In addition, two other effectors, EsxB and EsxD, strongly interact with each other (Anderson et al. 2013). While both EsxA and EsxB contain the WXG motif and thus represent canonical members of the WXG superfamily, EsxC and EsxD, in contrast, lack such motif and appear to be species- or strain-specific (Anderson et al. 2013). There is growing evidence supporting a seminal role played by various T7SS substrates during microbial infections. In M. tuberculosis, WXG effectors have been shown to orchestrate phagosome rupture, and to coordinate macrophage internalization, thereby directly contributing to virulence (Zhang et al. 2016). Similarly, T7b substrates contribute to the formation of persistent abscess lesions accompanied by increased bacterial load in animal infection experiments (Burts et al. 2008; Anderson et al. 2011).

Less is known about the translocation mechanism of ESS effectors into host cells across the staphylococcal envelope, but reports suggest that EssE, encoded by the ninth gene of the cluster is a critical membrane platform protein that coordinates the assembly of a functional ESS complex (Anderson et al. 2017). The ESS system from S. aureus also includes one membrane-associated SpoIIIE-FtsK-like ATPase, termed EssC (Fig. 1). SpoIIIE-FtsK-like ATPases are present in various T7SSs, and have been proposed to catalyze the transport of effector proteins across the ESS machinery (Zoltner et al. 2016). EssC may interact with the membrane platform component EssE, likely contributing to the maturation of a functional ESS system in the staphylococcal envelope (Zoltner et al. 2016; Anderson et al. 2017).

In this report, we focus on the dynamic interplay between two membrane-associated ESS proteins, EssB and EsaA, and the impact of their interaction on the ESS process. Both, EssB and EsaA are encoded by genes located within a conserved module of the ESS cluster (Fig. 1). EssB is a 52 kDa bitopic protein, with its sole 3 kDa transmembrane (TM) domain approximately located in the middle of the protein (Chen et al. 2012). Deletion of essB abolishes effector secretion, and expression of truncated EssB variants confers a dominant negative phenotype for substrate secretion, suggesting a major role for EssB during the ESS process (Chen et al. 2012). EssB interacts with itself through its C-terminal domain (~ 22 kDa) located on the trans-side of the plasma membrane (Fig. 2b). The N-terminal domain of EssB is located in the cytoplasm and has been proposed to mediate interactions with other components of the ESS pathway (Zoltner et al. 2013a, b). Very little is known about EsaA. Topology analysis using TMHMM server reveals the presence of six transmembrane (TM) domains spread throughout the protein length (Fig. 4a) (Krogh et al. 2001). EsaA is the second largest component of the ESS pathway, and carries a large soluble stretch that is likely exposed outside of the bacterial cell.

Fig. 2.

Fig. 2

Open in a new tab

The transmembrane domain of EssB is indispensable for interaction with EsaA. a Protein interactions assessed by the bacterial two-hybrid system. E. coli DHM1 cells harboring plasmid pairs T18-leucine zipper/T25-leucine zipper (Zip–Zip) as positive control, T18-essB/T25-esaA (EssB–EsaA) or T18-empty vector/T25-empty vector (T18–T25) as negative control were spotted either on (i) LB agar plates supplemented with the chromogenic substrate X-Gal, and IPTG, (ii) M63 plates supplemented with 0.2% maltose, and IPTG or (iii) MacConkey agar plates supplemented with 1% maltose, and IPTG. Blue and red colors in panels (i) and (iii), respectively indicate positive interaction and white color indicates negative interaction. Successful bacterial growth in panel (ii) indicates positive interaction whereas bacterial cell death is regarded as negative interaction. b Membrane topology of EssB, outlining the transmembrane domain (~ 3 kDa). The double arrow points to the C-terminal homodimerization domain of EssB, and the single arrow points to the transmembrane domain of EssB (Zoltner et al. 2013a, b). c Schematic presentation of wild-type and truncated essB constructs that were cloned for the purpose of bacterial two-hybrid analysis with EsaA. N N-terminal domain, TM transmembrane domain, C C-terminal domain. d Various restriction endonuclease digests loaded on agarose gel electrophoresis, showing the size of DNA inserts cloned into T18 vector. e Bacterial two-hybrid screening of the interaction output of various EssB domains with EsaA. E. coli DHM1 producing Zip–Zip, EssB–EsaA, NTM-EsaA, N-EsaA, TMC-EsaA, C-EsaA, TM-EsaA or T18–T25 as negative control were spotted on LB agar plates supplemented with the chromogenic substrate X-Gal, and IPTG. Blue color indicates positive interaction and white color indicates negative interaction.

Fig. 4.

Fig. 4

Open in a new tab

EssB does not interact with the soluble domain of EsaA. a Membrane topology of the 1009 amino acid long EsaA, outlining its 6 transmembrane domains (numbered 1 through 6), and the location of the ~ 90 kDa soluble domain (SD) residing between the first and second transmembrane domains, according to the TMHMM web-based topology prediction interactive tool. b Bacterial two-hybrid screening to full-length EssB interaction with a 70 kDa fragment of soluble domain of EsaA, EsaA(SD). E. coli DHM1 producing Zip–Zip, EssB–EsaA, EssB–EsaA(SD) or T18–T25 as negative control were spotted on LB plates supplemented with the chromogenic substrate X-Gal, and IPTG. Blue color indicates positive interaction and white color indicates negative interaction. c Visualization of the EsaA(SD). IPTG was added for 3 h to cultures of E. coli BL21(DE3) harboring pET15b empty vector (vector) or pET15b-esaA(SD) grown to OD600 of 0.6. Samples were collected and processed for SDS-PAGE analysis. An image of the Coomassie stained gel is shown. The arrow points to a new band that corresponds to the ~ 70 kDa soluble domain of EsaA. Sizes of protein markers are shown to the left of the panel.

Despite ample structural information on EssB, the molecular events that coordinate EssB interaction with other ESS proteins and subsequent contribution to the ESS process remain poorly understood. In this study, we identify the TM domain of EssB as an EsaA interaction domain. We subsequently show that EssB–EsaA interaction, together with five other components of the ESS pathway, facilitates partially regulated secretion in the Escherichia coli BL21(DE3) heterologous host.

Materials and methods

Bacterial strains and growth conditions

All strains described in this study are highlighted in Table 1. Cultures of Escherichia coli DH5α for cloning experiments (Meselson and Yuan 1968), and DHM1 (Euromedex, France) for the bacterial two-hybrid screening were grown at 37 and 30 °C, respectively. E. coli was grown in Luria-Bertani (LB) medium (peptone 1%, yeast extract 0.5%, NaCl 1%), and DNA manipulations followed standard procedures (Maniatis et al. 1982). The growth medium was supplemented with antibiotics (ampicillin 100 μg/ml; kanamycin 50 μg/ml) for plasmid propagation and isopropyl-β-D-thiogalactopyranoside (IPTG) was used at 1 mM final concentration when necessary. For bacterial two hybrid studies, E. coli DHM1 cultures were grown overnight at 37 °C in LB medium supplemented with ampicillin and kanamycin for vector propagation. Next day, cultures were diluted (1:100) in fresh broth and grown at 37 °C until the optical density at 600 nm (OD600) reached a value of 0.5. Next, 50-μl aliquots of cultures were spotted on LB agar plates containing ampicillin, kanamycin, IPTG (to induce gene expression) and 40 μg/ml of the chromogenic substrate 5-Bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal). Plates were incubated at 30 °C for 16–24 h (Anderson et al. 2013). Color change from white to blue was monitored as an indicator of positive interaction, or retention of white pigmentation as an indicator of negative interaction. Protein interaction was also monitored by spotting bacterial cultures (30 μl) on M63 minimal medium [(NH4)2SO4; 0.2%, KH2PO4; 1.36%, 0.5% FeSO4·7H2O, maltose; 0.2%, agar; 1.5%] scoring for viable bacteria, or on MacConkey agar medium (MacConkey agar 4%, maltose 1%) scoring for a change in color of the medium from white to red (reflecting negative and positive interactions, respectively). Both M63 and MacConkey agar media were supplemented with ampicillin, kanamycin and IPTG.

Table 1.

Primers, strains and plasmids

Primer, strain or plasmid (description)  Sequence Source/reference
Primers
essB (BTH, F) (Wild-type EssB for BTH screening) 5′-CGCAGTCTAGACATGGTTAAAAATCATAACCCTAAAAAT-3′ Aly et al. (2017)
essB (BTH, R) (Wild-type EssB for BTH screening) 5′-CGCAGGGTACCCTATTTTTTTCTTTCAGCTTCTTGGCGT-3′ Aly et al. (2017)
esaA (BTH, R) (Wild-type EsaA for BTH screening) 5′-CGCAGTCTAGACATGAAAAAGAAAAATTGGATTTATGCA-3′ Aly et al. (2017)
esaA (BTH, R) (Wild-type EsaA for BTH screening) 5′-CGCAGGGTACCTTAGATTAATCTCTCTTTCTTAAAGTGT-3′ Aly et al. (2017)
essB(1- 251F)-XbaI (EssB C-terminal Deletion; NTM) 5′-GCGCAGTCTAGACATGGTTAAAAATCATAACCCTA-3′ This study
essB(1-251R)-KpnI (EssB C-terminal Deletion; NTM) 5′-GCGCAGGGTACCCTATGAAAAATATAAAAAGGCTA-3′ This study
essB(1-228R)-KpnI (EssB C-terminal and TM deletion; N) 5′-GCGCAGGGTACCCTATTTGAAAACGGTATGTCCTA-3′ This study
essB(229-444F)-XbaI (EssB N-terminal deletion; TMC) 5′-GCGCAGTCTAGACGTTGCTATCGGTATGACAACGT-3′ This study
essB(229-444F)-XbaI (EssB N-terminal deletion; TMC) 5′-GCGCAGGGTACCCTATTTTTTTCTTTCAGCTTCTT-3′ This study
essB(252-444F)-XbaI (EssB N-terminal and TM deletion; C) 5′-GCGCAGTCTAGACATGAAGCATAATGAGCGCATTG-3′ This study
essB(252-444F)-XbaI (EssB N-terminal and TM deletion; C) 5′-GCGCAGGGTACCCTATTTTTTTCTTTCAGCTTCTT-3′ This study
esaA(F) (Soluble domain) 5′-CGCAGCATATGACTTTAATTGAAAAACAAAATTCATTAT-3′ This study
esaA(R) (Soluble domain) 5′-CGCAGGGATCCTTATAAAATCACCATTAAGATGAATTTC-3′ This study
essE(F) on pACYCDuet-1-5′ 5′-CGCAGCCATGGTTAAAGATGTTAAGCGAATAGATTATTT-3′ This study
essE(R) on pACYCDuet-1-3′ 5′-CGCAGAAGCTTTTACTCCTCTGCTTTATTAATATGATTT-3′ This study
esxA-essB(F) on pRSFDuet-1 5′-CGCAGAGATCTCATGGCAATGATTAAGATGAGTCCAGAGGAA-3′ This study
esxA-essB(R) on pRSFDuet-1 5′-CGCAGGGTACCCTATTTTTTTCTTTCAGCTTCTTGGCGT-3′ This study
essC(F) on pET5b 5′-CGCAGCTCGAGATGCATAAATTGATTATAAAATATAACA-3′ This study
essC(R) on pET5b 3′-CGCAGGGATCCCTATTTAAACCATCTAATCTTTTGATAA-3′ This study
Strains
E. coli DH5α F, endA1, glnV44, thi1, recA1, relA1, gyrA96, deoR, nupG, purB20, φp80dlacZΔM15, Δ(lacZYA-argF)U169, hsdR17(rK − mK + ), λ − Meselson and Yuan (1968)
E. coli DHM1 F-, cya-854, recA1, endA1, gyrA96 (Nal r), thi1, hsdR17, spoT1, rfbD1, glnV44(AS) Euromedex, France
E. coli BL21(DE3) fhuA2 [lon] ompT gal (λ DE3) [dcm] ΔhsdS λ DE3 = λ sBamHIo ΔEcoRI-B int::(lacI::PlacUV5::T7 gene1) i21 Δnin5 NEB
Plasmids
pKT25 (p25) Kanr, encoding the N-terminal 224 aa of the Bordetella pertussis adenylate cyclase Euromedex, France
pUT18C (p18) Ampr, encoding the C-terminal (225–399 aa) of the B. pertussis adenylate cyclase Euromedex, France
pET15b Ampr, T7 RNA polymerase-based expression vector Novagen
pRSFDuet-1 Km1, T7 RNA polymerase-based expression vector Novagen
pACYCDuet-1 Cmr, T7 RNA polymerase-based expression vector Novagen
Open in a new tab

Restriction endonuclease cleavage sites are underlined

Cloning

All plasmids described in this study are listed in Table 1. All genes were amplified by polymerase chain reaction (PCR) using the genomic DNA of strain USA300 (LAC) prepared with QIAamp® DNA Mini Kit (Qiagen, Hilden, Germany), and oligonucleotides listed in Table 1. Cloning for the bacterial two-hybrid analysis was performed using E. coli DH5α and vectors pUT18C and pKT25 (Euromedex, France). In this study, these two vectors are referred as p18 and p25. Target DNA sequences amplified using genomic DNA were purified from agarose gels and subjected to restriction endonuclease treatment using XbaI and KpnI, followed by fusion into p18 or p25 pre-cut with the same restriction endonucleases. The resulting plasmids were co-transformed in pairs (p18/p25) into electro-competent E. coli DHM1 for the bacterial two-hybrid screening.

For the heterologous production of ess genes in E. coli, three compatible expression vectors from Novagen were used: pRSFDuet-1 (Kanamycin resistant, Kmr), pACY-CDuet-1 (Chloramphenicol resistant, Cmr), and pET15b (Ampicillin resistant, Ampr). For the experiment shown in Fig. 5b, vector pRSFDuet-1 was used to clone the conserved ESS module (esxA-esaA-essA-esaB-essB) and vectors pACYCDuet-1 and pET15b were used to clone essE and essC, respectively (Table 1).

Fig. 5.

Fig. 5

Open in a new tab

Extensive release of proteins in the periplasm of E. coli when seven ESS proteins are co-overproduced. a Schematic depiction of strains S1, S2 and S3 carrying the indicated plasmids and genes. Both S2 and S3 carry genes esxA, esaA, essA, essB and essE. In addition strain S3 also carries the essC gene on a third plasmid. b Sub-cellular fractionations of strains S2 and S3 and control strain carrying vectors alone (S1). Bacterial cultures grown to OD600 of 0.6 were incubated with IPTG for 16 h at 28 °C. Cells were collected and fractionated into cytosolic, membrane and periplasmic fractions (as indicated). Samples were separated on SDS-PAGE and gels stained with Coomassie. Black arrowheads point to overproduced proteins otherwise absent in the vector alone extracts (S1). The new band in lane S2 middle panel (membrane fraction) corresponds to EssC. Molecular weight markers were loaded in lanes labeled M

Subcellular fractionation and periplasmic preparations by osmotic shock

Overnight cultures of strains expressing various genes of the ess cluster were diluted in fresh LB medium supplemented with antibiotics, and grown at 37 °C until OD600 of 0.6. Following addition of IPTG, cultures were transferred to 28 °C, and incubated for another 5 h. Next, bacterial cells were collected by centrifugation at 4000×g for 15 min. Cell pellets were subjected to sonication, and cell debris pelleted at 3000×g for 10 min. Total cell lysates were subjected to ultracentrifugation at 100,000×g for 2 h, at 4 °C, followed by the collection of supernatant (cytosol) and pellet (membrane) fractions.

Periplasmic fractions of culture samples were obtained largely as described (Neu and Heppel 1965). Briefly, equal amounts of cells were pelleted at 4000×g, 4 °C for 15 min. Next, cells were re-suspended in 20% sucrose, 1 mM ethyl-enediaminetetraacetic acid (EDTA), 30 mM Tris-HCl (pH. 8) and subjected to gentle shaking at room temperature for 10 min before centrifugation at 14,000×g for 15 min at 4 °C. Pellets were rapidly suspended in sterile cold water and incubated with gentle shaking for another 10 min at 4 °C. Periplasmic fractions (supernatants) were recovered following centrifugation of samples at 14,000×g for 15 min at 4 °C.

In vivo chemical crosslinking and sample analyses

E. coli DHM1 cells harboring p18essB or p18essB(NC) as well as p25esaA were grown overnight at 37 °C in LB medium supplemented with ampicillin and kanamycin. Cultures were diluted in fresh medium in the presence of antibiotics, grown to OD600 of 0.05 and incubated at 37 °C until the cultures reached OD600 of 0.6. IPTG was added for another 3 h. Cultures were normalized to OD600 of 0.5, and cells were washed twice in phosphate buffer saline (PBS). 1% of the homobifunctional crosslinker formaldehyde was added, followed by incubation for 1 h at room temperature. Next, 1M Tris-HCl, pH 6.8, was added to stop the reaction, followed by further incubation for 5 min at room temperature. Samples were washed twice with PBS. Equal volume of protein sample buffer was added, followed by the passage of each sample 7 times through 26 gauge needle, incubation at 37 °C for 10 min, and directly loading onto SDS-PAGE according to Schägger et al. (1988). Proteins in the gel were transferred onto a polyvinylidene difluoride (PVDF) membrane, incubated with primary antibody (anti-EssB), followed by horse radish peroxidase (HRP)-conjugated secondary antibody (Abcam). Immune reactive complexes were revealed by chemiluminescence using ImageQuant™ LAS 500 (GE Healthcare).

Membrane topology predictions

Membrane topology of EsaA was predicted using the TMHMM v. 2.0 online server, Center for Biological Sequence Analysis, Technical University of Denmark, following the submission of EsaA sequence to the server in a FASTA format as explained elsewhere (Krogh et al. 2001).

Results

EssB interaction with EsaA is validated under three different growth conditions

Both, essB and esaA reside in a highly-conserved module of the ESS cluster (Fig. 1). Topology predictions and high-resolution structural studies confirm the presence of an ~ 3 kDa transmembrane domain in the middle of EssB (Chen et al. 2012; Zoltner et al. 2013a, b). A multicomponent ESS complex from the membranes of S. aureus has been characterized, and shown to contain both EssB and EsaA (Aly et al. 2017). However, this study does not reveal how EssB and EsaA interact in this complex and the biological significance of the EssB–EsaA interaction within the ESS system has not been defined. To further validate EssB–EsaA using several experimental conditions, full-length EssB and EsaA were produced from two different vectors in E. coli DHM1 as translational hybrids, each carrying one-half of the Bordetella pertussis adenylate cyclase (cya). Bacteria were grown on plates in the presence of antibiotics for plasmid selection, and IPTG for the induced production of the translational hybrids. The chromogenic substrate X-Gal was included for colorimetric indication of positive interactions between hybrids (Fig. 2a). In the bacterial two-hybrid system, positive interaction between protein hybrids leads to bacterial color change from white to blue, stemming from the reconstitution of full-length Cya in vivo. Cya reconstitution in control experiments was demonstrated using the Zip–Zip pair that monitors homodimerization of the leucine zipper, a well characterized eukaryotic transcriptional regulator (Maxon et al. 1990; Garza and Christie 2013; Cukier 2014; Qin et al. 2015). The productive Zip–Zip interaction results in the catalytic breakdown of ATP into cAMP by Cya, thereby activating a cAMP-CAP promoter that controls the production of β-galactosidase. Subsequently, β-galactosidase production breaks down X-Gal, which turns blue (Fig. 2a(i)) (Karimova et al. 1998). In contrast, lack of interaction between the two Cya halves as observed when bacteria carry the two vectors without inserts (T18–T25) results in no color change and colonies appear white in the presence of intact X-Gal (Fig. 2a(i)). Similar to the Zip–Zip positive control, blue pigmentation was observed for E. coli DHM1 producing the Cya translational hybrids fused to EssB and EsaA (EssB–EsaA) confirming their strong interaction in the absence of any other ESS proteins as previously reported (Aly et al. 2017).

In this study, EssB–EsaA interaction was further validated using two additional growth conditions. DHM1 cells were grown on M63 minimal medium, where the output of positive protein–protein interactions is demonstrated in the form of bacterial colony survival (Fig. 2a(ii)). This is in contrast to the negative control where Cya reconstitution does not occur and results in cell death due to the failure to utilize maltose as a sole carbon source (Karimova et al. 1998). EssB–EsaA interaction was further demonstrated on MacConkey agar, where positive EssB–EsaA interaction results in colony color change from white to red, whereas negative interaction as in T18–T25 retains white pigmentation of cells (Fig. 2a(iii)). Taken together, these results confirm that EssB and EsaA interact strongly and independently of other ESS proteins.

The transmembrane domain of EssB is required for interaction with EsaA

Previous analyses on EssB led to a model whereby the ~ 22 kDa C-terminal domain of the protein forms a dimer that resides on the trans side of the bacterial membrane whereas the N-terminal ~ 25 kDa domain resides on the cis side of the membrane (Chen et al. 2012; Zoltner et al. 2013a, b). However, interacting partners were not revealed in these earlier studies (Chen et al. 2012; Zoltner et al. 2013a, b). When produced in E. coli BL21(DE3), EssB was also found to fractionate with the membrane in a manner mediated by the 3 kDa transmembrane domain located in the middle of the protein sequence (Fig. 2b) (Chen et al. 2012). Here, we attempt to identify for the first time the domain of EssB that mediates interaction with EsaA. To this end, systematic truncations were created in EssB, resulting in the construction of five different variants (Fig. 2c). In the first construct, the C-terminal domain was removed yielding variant NTM whereby “N” and “TM” denote the N-terminal and transmembrane domains of EssB, respectively. The variant named N lacked both the TM and C-terminal domains of EssB, while, conversely, the variant named C contains only the C-terminal domain and variant referred as TM encompasses the transmembrane domain solely. Variant TMC was constructed by removing the N-terminal domain of EssB (Fig. 2c). All these variants were cloned in vector T18 (Fig. 2d), and the resulting plasmids were transformed in E. coli DHM1 carrying T25 vector encoding esaA. The resulting strains were examined for the production and oligomerization of Cya on plates containing X-gal (Fig. 2e). The T18 variants producing NTM, TMC and TM retained strong interactions with EsaA, while variants producing N or C alone did not support interaction with EsaA resulting in no Cya activity and thus no change in color was observed; colonies remained white on X-gal containing plates (Fig. 2e). These results suggest that removal of the TM domain from EssB abolishes the interaction with EsaA, and that the presence of the TM, whether alone or fused to the N- or the C-terminus of EssB retained strong interaction with EsaA. These data show that the TM domain of EssB is necessary and sufficient to establish an interaction with EsaA. To confirm this finding, an additional construct encompassing the N and C-terminal domains but lacking the transmembrane segment, EssB(NC), was generated. The two-hybrid analysis indicates that removal of the TM domain from EssB leads to loss of interaction with EsaA, further validating the importance of the TM domain in mediating contact between the two proteins (Fig. 3a). Next, DHM1 cells producing wild-type EsaA fused to the C-terminal domain of B. pertussis Cya and either wild-type EssB or EssB(NC) fused to the N-terminal half of Cya were treated with the crosslinking agent formaldehyde in an attempt to capture complexes in vivo (Fig. 3b; + signs). In a control experiment, the crosslinker was omitted from the reaction (Fig. 3b; – signs). Western blot analysis using anti-EssB polyclonal antibodies reveals distinct high molecular weight band (marked by black arrowhead) in extracts containing wild-type EsaA and EssB that were not observed in extracts of bacteria carrying vector controls T18 and T25 that produce the Cya domains alone (Fig. 3b). When EssB(NC) was provided on plasmid T18, the higher molecular weight crosslinked band was not observed (Fig. 3b). Western blotting is a known sensitive approach that is suitable for the detection of limited protein quantities, and the lack of crosslinking in that regard was not due to notably reduced production or instability of the EssB(NC)-Cya hybrid as a band corresponding to this species migrated at the expected mobility on SDS-PAGE (Fig. 3b; black arrow). Collectively, these data convey that the EssB transmembrane domain mediates interaction with EsaA.

Fig. 3.

Fig. 3

Open in a new tab

Removal of the transmembrane domain from EssB abolishes interaction with EsaA. a Bacterial two-hybrid screening was used to assess the requirement of the transmembrane domain of EssB for interaction with EsaA using a new variant, EssB(NC), lacking the transmembrane domain. E. coli DHM1 variants producing Zip–Zip, EssB–EsaA, EssB(NC)–EsaA or T18–T25 as negative control were spotted on LB plates supplemented with the chromogenic substrate X-Gal, and IPTG. Blue color indicates positive interaction and white color indicates negative interaction. EssB(NC) denotes an EssB variant lacking only the transmembrane domain. b In vivo crosslinking and Western blot analysis of EssB interaction with EsaA. E. coli DHM1 variants producing T18–T25, EssB–EsaA, or EssB(NC)–EsaA were grown in LB liquid medium. Cultures were normalized to OD600 of 0.5, and washed twice prior to incubation with (+) or without (−) formaldehyde. Following quenching of the reactions, samples were separated by SDS-PAGE, transferred to a PVDF membrane, and subjected to immunoblot analysis with a polyclonal antibody against EssB (anti-EssB). The black arrow points to the monomeric EssB-Cya fusion variant. Black arrowhead points to crosslinked product.

EssB interacts with a hydrophobic region in EsaA

Next, we wondered how EsaA establishes interaction with EssB. Topology prediction supported the notion that EsaA is an integral membrane component with six TM domains and a large soluble stretch between TM1 and TM2 (Krogh et al. 2001; Zoltner et al. 2013, 2016). This domain referred as soluble domain of EsaA, EsaA(SD), is predicted to reside on the trans side of the bacterial membrane (Fig. 4a). To examine whether full-length EssB interacts with EsaA(SD), a new construct was generated for two-hybrid screening. However, this hybrid combination did not yield active adenylate cyclase (Fig. 4b). To evaluate whether the absence of interaction could be the result of an unstable EsaA(SD) polypeptide, the protein was also produced in E. coli BL21(DE3). Following induction with IPTG, extracts of cells carrying plasmid-encoded esaA(SD) but not the vector control alone, produced a new species that migrated with a mobility of approximately 70 kDa on SDS-PAGE (Fig. 4c). Together, these observations suggest that EssB does not interact with the soluble domain of EsaA but rather, EssB must interact with some hydrophobic segment of EsaA within the bacterial membrane.

EssB–EsaA interaction in the presence of five other ESS components facilitates partially regulated protein secretion in a heterologous host

To gain better insights into the biological significance of EssB–EsaA interaction, we took a closer look at the location of the two genes encoding EssB and EsaA. Both genes reside within a conserved module that must mediate important physiological feature(s) (Fig. 1). We decided to overproduce the entire conserved module containing EssB and EsaA, along with EssE, whether separately or together with EssC in E. coli BL21(DE3) cells (Fig. 5a). Our choice of including EssE and EssC in this overproduction attempt stems from previous work suggesting that EssE may function as a membrane platform protein that promotes substrate secretion across bacterial membranes (Anderson et al. 2017). In addition, we also chose EssC as the only known ATPase found in the staphylococcal ESS system, which may potentially energize ESS assembly and/or function (Zoltner et al. 2016).

The genes coding for the proteins described above were cloned in three compatible vectors that all carry a T7 promoter. In this manner, the transcription of plasmid-encoded genes can be simultaneously induced with IPTG when using E. coli BL21(DE3) as the host. Three strains were constructed (Fig. 5a). The first strain (S1) carried vectors lacking inserts and served as a control. Strain S2 produced the conserved ESS module (EsxA-EsaA-EssA-EsaB-EssB) encoded on vector pRSFDuet-1, along with EssE encoded on vector pACYCDuet-1. Strain S3 carried a third plasmid for production of EssC (pET15b encoded) (Fig. 5a). Fractionation experiments, followed by SDS-PAGE reveal significant differences in the cytosolic contents of strains S2 and S3 when compared with control strain S1 as indicated by black arrowheads (Fig. 5b, left panel). This result suggests the successful overproduction and cytosolic fractionation of various soluble components of ESS proteins. In addition, EssC, a ~ 165 kDa membrane-associated ATPase with two transmembrane domains was detected in the membrane fraction of extracts from strain S3 as indicated by a black arrowhead (Fig. 5b, middle panel). Importantly, the periplasmic extract of strain S3 contains an abnormal amount of proteins that are otherwise not observed in extracts processed from strains S1 and S2 (Fig. 5b, right panel). This finding likely represents leakage of otherwise cytosolic E. coli proteins, upon the assembly of a non-selective secretion mechanism in the heterologous E. coli BL21(DE3) host. This leakage is partially regulated in strain S3, since cells were viable. We conclude from this experiment that EssB and EsaA, together with other components of the conserved module, as well as EssE and EssC, collectively provide a raw secretion platform that supports partially regulated secretion across the cytoplasmic membrane in a heterologous host.

Discussion

The contribution of T7SS/ESS for virulence has been extensively examined both for firmicutes and actinobacteria (Abdallah et al. 2007; Anderson et al. 2017; Unnikrishnan et al. 2017). For example, three of the five T7SS paralogs in M. tuberculosis orchestrate bacterial virulence through the secretion of arsenals of Pro-Glu (PE) and Pro-Pro-Glu (PPE) pathogenic effectors (Groschel et al. 2016). In S. aureus, the ESS pathway contributes to the persistence of pathological lesions referred as abscesses; caused by wild-type bacteria but not ess mutant variants, and bacteria continue to replicate over months while shielded from phagocytes, antibodies and antibiotics. Disruption of mature lesions leads to the release of wild-type bacteria and fatal bloodstream infections in humans and animals (Burts et al. 2005; Anderson et al. 2011, 2017). How exactly proteins are secreted by the ESS pathway and how they favor bacterial replication in abscess lesions is unknown. Here, we examine the process leading to the assembly of a partially regulated secretion mechanism and examine the contributions of EsaA and EssB. These proteins are encoded by genes found in a highly-conserved module of the ESS pathway (Fig. 1) (Warne et al. 2016). EssB is a bitopic ESS component, with its sole transmembrane domain residing roughly in the middle of the polypeptide sequence (Chen et al. 2012). Structure of EssB has been studied in detail, and shows that EssB interacts with itself, and its ~ 22 kDa trans-side C-terminus is a self-association domain (Zoltner et al. 2013a, b). In addition, the N-terminus of EssB resides in the cytosolic compartment, and structurally resembles Ser/Thr protein kinases (Zoltner et al. 2013). Despite numerous structural insights into the nature of EssB, key questions remain unanswered. For example, which domains of EssB are engaged in interaction with other components of the ESS cluster? how are such interactions established? and what is the biological significance of these interactions?

In this study, we document several important findings. First, the TM domain of EssB mediates interaction with EsaA, another conserved and integral membrane component of the ESS system. Little is known about EsaA, but a homolog of EsaA has been found to function as a phage receptor in B. subtilis (Baptista et al. 2008). EsaA is embedded in the bacterial membrane through 6 TM helices that are spread throughout the protein length. We show that the TM domain of EssB interacts with EsaA, and our findings are based on an unbiased construction of EssB truncations that cover various possibilities, followed by in vivo examination of the impact of these truncations on the interaction with EsaA. The data reveal that the presence of the TM domain of EssB, whether alone or together with N- or C-terminal domains of EssB, always results in a positive interaction with EsaA. In contrast, when the TM domain of EssB is removed, the interaction with EsaA is abrogated.

Our findings also shed new insights into the nature of EsaA interaction with EssB. We find that the sole soluble domain of EsaA does not interact with EssB (Fig. 4b). This suggests that EssB–EsaA interaction may take place within the hydrophobic vicinity of the bacterial membrane, possibly through EssB interaction with one of the five TM domains located in the second half of EsaA or the first TM domain located before the soluble stretch. It is worth noting that the entire molecular mass of the soluble domain of EsaA is about 90 kDa, and we tested an ~ 70 kDa segment of that domain that was stable when produced in E.coli BL21(DE3) cells (Fig. 4c). TM–TM interactions have been identified in several biological systems, and TM peptides have been used to perturb TM–TM interactions, with potential therapeutic applications (Gerber and Shai 2001; Gerber et al. 2004; Fink et al. 2012). This will be potentially explored in future interaction studies of the ESS pathway.

A notable question in T7SS/ESS research is the nature of the secretion mechanism through which various effectors pass into the cell exterior. Since essB and esaA are encoded within a highly conserved ESS module, it is justifiable to assume that they must play a fundamental contribution to both, ESS assembly and function. Our finding that the transmembrane domain of EssB is required for interaction with EsaA, and indirect observations suggesting that such an interaction takes place within the bacterial membrane leads us to hypothesize that the two proteins may contribute to substrate passage through the bacterial membrane. During our investigations using E. coli as a heterologous host for the production and study of a putative secretion mechanism, we discovered that protein combinations including EssB, EsaA, other conserved ESS module, as well as EssE and EssC leads to the creation of partially regulated protein secretion across the E. coli inner membrane (Fig. 5b, strain S3). SDS-PAGE analyses of subcellular fractions revealed an increased number of proteins in the periplasm of strain S3 that produced the conserved module along with EssE and EssC. Periplasmic contents of control strain S1 carrying vectors only or strain S2 carrying the conserved module and EssE, appeared normal. All fractionation experiments were otherwise performed on normalized cultures, indicating that the increased abundance of proteins detected in the periplasmic fraction of strain S3 can be attributed to the assembly of partially regulated secretion mechanism. We surmise that the non-specific protein secretion mechanism in strain S3 leaks proteins in the periplasm in a non-specific manner albeit that this leakage is not lethal. It is possible that gene products otherwise responsible for selective gating of proteins remain missing in the reconstituted E. coli system. The results also point to the conclusion that the exact factor(s) otherwise required for the specific selection of WXG substrates must still be missing in E. coli.

Whereas very little information is available regarding the mechanism by which WXG substrates of the ESS system translocate across bacterial membranes into host cells, the concept of gating various substrates through bacterial membranes in the distant T7aSS remains unclear, especially from the perspective of the pathogen’s ability to mediate membrane piercings in infected host cells (Conrad et al. 2017). For example, ESX-1-dependent cell membrane lysis in Mycobacteria was in one study attributed to gross membrane disruptions rather than the formation of distinct openings in host membranes (Conrad et al. 2017). Aside from the host perspective, we propose that EssB and EsaA, together with other members of the conserved module and EssE as well as EssC are sufficient to create partially regulated protein passage that can be tolerated by E. coli BL21(DE3) cells. This hypothesis is in agreement with findings that EssE likely interacts with EssC (Cao et al. 2016; Anderson et al. 2017). Collectively, the data presented in this study shed several new insights into the nature of ESS pathway and the biological contribution of EssB and EsaA to ESS system maturation and function.

Acknowledgements

We thank Olaf Schneewind and Chloe Schneewind for careful reading of the manuscript, and members of the Aly laboratory for the useful insights. Research in the Aly laboratory is supported by capacity and core facility equipment from Sinai University, El-Arish, North Sinai, Egypt.

References

  1. Abdallah AM, Gey NC, van Pittius et al. (2007) Type VII secretion—mycobacteria show the way. Nat Rev Microbiol 5(11):883–891 [DOI] [PubMed] [Google Scholar]
  2. Aly KA, Anderson M et al. (2017) Isolation of a membrane protein complex for type VII secretion in Staphylococcus aureus. J Bacteriol 199(23):e00482–17. 10.1128/JB.00482-17 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Anderson M, Chen YH et al. (2011) EsaD, a secretion factor for the Ess pathway in Staphylococcus aureus. J Bacteriol 193(7):1583–1589 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Anderson M, Aly KA et al. (2013) Secretion of atypical protein substrates by the ESAT-6 secretion system of Staphylococcus aureus. Mol Microbiol 90(4):734–743 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Anderson M, Ohr RJ et al. (2017) EssE promotes Staphylococcus aureus ESS-dependent protein secretion to modify host immune responses during infection. J Bacteriol 199(1):e00527–16 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Baptista C, Santos MA et al. (2008) Phage SPP1 reversible adsorption to Bacillus subtilis cell wall teichoic acids accelerates virus recognition of membrane receptor YueB. J Bacteriol 190(14):4989–4996 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Burts ML, Williams WA et al. (2005) EsxA and EsxB are secreted by an ESAT-6-like system that is required for the pathogenesis of Staphylococcus aureus infections. Proc Natl Acad Sci USA 102(4):1169–1174 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Burts ML, DeDent AC et al. (2008) EsaC substrate for the ESAT-6 secretion pathway and its role in persistent infections of Staphylococcus aureus. Mol Microbiol 69(3):736–746 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cao Z, Casabona MG et al. (2016) The type VII secretion system of Staphylococcus aureus secretes a nuclease toxin that targets competitor bacteria. Nat Microbiol 2:16183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Champion PA, Champion MM et al. (2009) ESX-1 secreted virulence factors are recognized by multiple cytosolic AAA ATPases in pathogenic mycobacteria. Mol Microbiol 73(5):950–962 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chen YH, Anderson M et al. (2012) Characterization of EssB, a protein required for secretion of ESAT-6 like proteins in Staphylococcus aureus. BMC Microbiol 12:219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Conrad WH, Osman MM et al. (2017) Mycobacterial ESX-1 secretion system mediates host cell lysis through bacterium contact-dependent gross membrane disruptions. Proc Natl Acad Sci USA 114(6):1371–1376 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cukier RI (2014) Simulations of potentials of mean force for separating a leucine zipper dimer and the basic region of a basic region leucine zipper dimer. J Phys Chem B 118(35):10341–10354 [DOI] [PubMed] [Google Scholar]
  14. David MZ, Daum RS (2010) Community-associated methicillin-resistant Staphylococcus aureus: epidemiology and clinical consequences of an emerging epidemic. Clin Microbiol Rev 23(3):616–687 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Falugi F, Kim HK et al. (2013) Role of protein A in the evasion of host adaptive immune responses by Staphylococcus aureus. MBio 4(5):e00575–00513 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fink A, Sal-Man N et al. (2012) Transmembrane domains interactions within the membrane milieu: principles, advances and challenges. Biochim Biophys Acta 1818(4):974–983 [DOI] [PubMed] [Google Scholar]
  17. Garufi G, Butler E et al. (2008) ESAT-6-like protein secretion in Bacillus anthracis. J Bacteriol 190(21):7004–7011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Garza I, Christie PJ (2013) A putative transmembrane leucine zipper of agrobacterium VirB10 is essential for t-pilus biogenesis but not type IV secretion. J Bacteriol 195(13):3022–3034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Gerber D, Shai Y (2001) In vivo detection of hetero-association of glycophorin-A and its mutants within the membrane. J Biol Chem 276(33):31229–31232 [DOI] [PubMed] [Google Scholar]
  20. Gerber D, Sal-Man N et al. (2004) Two motifs within a transmembrane domain, one for homodimerization and the other for heterodimerization. J Biol Chem 279(20):21177–21182 [DOI] [PubMed] [Google Scholar]
  21. Groschel MI, Sayes F et al. (2016) ESX secretion systems: mycobacterial evolution to counter host immunity. Nat Rev Microbiol 14(11):677–691 [DOI] [PubMed] [Google Scholar]
  22. Jongerius I, von Kockritz-Blickwede M et al. (2012) Staphylococcus aureus virulence is enhanced by secreted factors that block innate immune defenses. J Innate Immun 4(3):301–311 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Karimova G, Pidoux J et al. (1998) A bacterial two-hybrid system based on a reconstituted signal transduction pathway. Proc Natl Acad Sci USA 95(10):5752–5756 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Krogh A, Larsson B et al. (2001) Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J Mol Biol 305(3):567–580 [DOI] [PubMed] [Google Scholar]
  25. Maniatis T, Fritsch EF et al. (1982) Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor [Google Scholar]
  26. Maxon ME, Wigboldus J et al. (1990) Structure-function studies on Escherichia coli MetR protein, a putative prokaryotic leucine zipper protein. Proc Natl Acad Sci USA 87(18):7076–7079 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Meselson M, Yuan R (1968) DNA restriction enzyme from E. coli. Nature 217(5134):1110–1114 [DOI] [PubMed] [Google Scholar]
  28. Neu HC, Heppel LA (1965) The release of enzymes from Escherichia coli by osmotic shock and during the formation of spheroplasts. J Biol Chem 240(9):3685–3692 [PubMed] [Google Scholar]
  29. Poulsen C, Panjikar S et al. (2014) WXG100 protein superfamily consists of three subfamilies and exhibits an alpha-helical C-terminal conserved residue pattern. PLoS One 9(2):e89313. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Qin L, Reger AS et al. (2015) Structures of cGMP-dependent protein kinase (PKG) Ialpha leucine zippers reveal an interchain disulfide bond important for dimer stability. Biochemistry 54(29):4419–4422 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Schagger H, Aquila H et al. (1988) Coomassie blue-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for direct visualization of polypeptides during electrophoresis. Anal Biochem 173(1):201–205 [DOI] [PubMed] [Google Scholar]
  32. Sundaramoorthy R, Fyfe PK et al. (2008) Structure of Staphylococcus aureus EsxA suggests a contribution to virulence by action as a transport chaperone and/or adaptor protein. J Mol Biol 383(3):603–614 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Sysoeva TA, Zepeda-Rivera MA et al. (2014) Dimer recognition and secretion by the ESX secretion system in Bacillus subtilis. Proc Natl Acad Sci USA 111(21):7653–7658 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Thammavongsa V, Kim HK et al. (2015) Staphylococcal manipulation of host immune responses. Nat Rev Microbiol 13(9):529–543 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Unnikrishnan M, Constantinidou C et al. (2017) The enigmatic Esx proteins: looking beyond mycobacteria. Trends Microbiol 25(3):192–204 [DOI] [PubMed] [Google Scholar]
  36. van Belkum A, Melles DC et al. (2009) Co-evolutionary aspects of human colonisation and infection by Staphylococcus aureus. Infect Genet Evol 9(1):32–47 [DOI] [PubMed] [Google Scholar]
  37. Ventola CL (2015) The antibiotic resistance crisis: part 2: management strategies and new agents. P T 40(5):344–352 [PMC free article] [PubMed] [Google Scholar]
  38. Warne B, Harkins CP et al. (2016) The Ess/Type VII secretion system of Staphylococcus aureus shows unexpected genetic diversity. BMC Genom 17:222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Zhang Q, Wang D et al. (2016) EsxA membrane-permeabilizing activity plays a key role in mycobacterial cytosolic translocation and virulence: effects of single-residue mutations at glutamine 5. Sci Rep 6:32618. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Zoltner M, Fyfe PK et al. (2013a) Characterization of Staphylococcus aureus EssB, an integral membrane component of the Type VII secretion system: atomic resolution crystal structure of the cytoplasmic segment. Biochem J 449(2):469–477 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Zoltner M, Norman DG et al. (2013b) The architecture of EssB, an integral membrane component of the type VII secretion system. Structure 21(4):595–603 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Zoltner M, Ng WM et al. (2016) EssC: domain structures inform on the elusive translocation channel in the Type VII secretion system. Biochem J 473(13):1941–1952 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES