![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Previous Article | Next Article 
Journal of Bacteriology, March 2002, p. 1244-1252, Vol. 184, No. 5
0021-9193/02/$04.00+0 DOI: 10.1128/JB.184.5.1244-1252.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
Division of Bacterial Infection, Department of Microbiology and Immunology,1 Division of Biomolecular Imaging, Department of Basic Medical Sciences, Institute of Medical Science, University of Tokyo, Minato-ku, Tokyo 108-8639 ,2 PRESTO, Japan Science and Technology Corporation, Kawaguchi, Saitama 332-0012, Japan3
Received 4 October 2001/ Accepted 20 November 2001
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Recent studies indicate that the type III secretion complexes of Shigella flexneri and Salmonella spp. share structural similarities; the complexes are composed of two distinct parts, a basal body and a needle. The basal body is embedded in the bacterial membrane, while the needle protrudes from the bacterial surface (24, 45). The supramolecular structure of extensively purified type III secretion complexes from the envelope of S. flexneri and Salmonella enterica serovar Typhimurium reveals that the needle is around 45 to 50 nm in length and protrudes from the basal body, while the basal body consists of two pairs of upper and lower rings (24, 45). The thickness of the basal body is estimated to be approximately 31 nm, which appears to match the thickness of the bacterial envelope (around 25 nm) (45). Because the flagellar basal body also has two upper (called P and L rings) and lower (MS) rings (28, 41, 46), the basic structure of the basal body of the type III secretion machinery is similar to that of the flagellar export basal body.
The major components of the type III secretion complexes of S. flexneri and Salmonella spp. have been identified. The basal part of the Shigella type III complex is composed of MxiD, MxiG, and MxiJ (45), while that of Salmonella is composed of InvG, PrgH, and PrgK (24). The needle is composed of MxiH and MxiI in S. flexneri and PrgI and PrgJ in Salmonella, in which MxiI and PrgJ are indicated to be minor essential components (4, 22, 25, 45). The needle portion is essential, not only as the molecular bridge connecting the pathogen to the host cell but also in mediating the delivery of a set of effector proteins (25, 45).
The needle length of the type III secretion machinery in S. flexneri appears to be uniform, because it is confined to a narrow range, around 45 to 50 nm (45). When the major needle component protein of S. flexneri, MxiH, was overexpressed in the mxiH mutant, the type III secretion machinery protruded longer needles than the wild type, in which the longest needles were approximately 1,000 nm, and the bacteria invaded the host cells much more efficiently than the wild type (45). Although the distribution of the type III needle lengths in the Salmonella wild type has not yet been characterized, the lengths are thought to be distributed over a narrow range, around 50 nm (22).
When the invJ gene, located in the inv-spa operon of Salmonella pathogenicity island 1 (SPI-1), was mutated, the bacteria produced longer type III needles than the wild type and the mutant lost the ability to mediate secretion of the effector proteins into the medium (6, 7, 25, 44). Furthermore, with overexpression of HilA, a positive controller for the expression of type III-associated genes, in the invJ mutant, the type III complexes had an abundance of elongated needles, suggesting that InvJ affects the length of the needles (25). The mechanism by which InvJ controls needle length is, however, not understood. In the flagellar export system of Salmonella, the length of the hook, which is a curved appendage extending from the flagellar basal body, also had a narrow range, around 55 nm (23). When the fliK gene was mutated, however, the mutant produced a flagellar export apparatus with a longer hook than the wild type, called a polyhook (18). Therefore, FliK affects hook length, though the mechanism of hook length control by FliK is unknown.
In the Shigella type III secretion system, Spa32, a 32-kDa protein encoded by the spa32 gene in the spa operon, shows moderate amino acid homology with InvJ. The overall identity of these proteins is 15.4%, although in stretches from residues 183 to 291, the identity is 26%. Spa32 would also seem to show marginal amino acid homology with Salmonella flagellar FliK (10.9% identity) and Yersinia YscP (10.3% identity), a Yersinia type III-associated protein encoded by the yscP gene, from the sequence alignment, but not the sequence homology search, and similar locations of the spa32, invJ, fliK, and yscP genes in the respective operons (2, 19). These findings suggest that their essential roles in the secretion systems are similar.
Our previous study indicated that Spa32 was an essential factor for promoting Shigella invasion of epithelial cells and mediating secretion of effector proteins such as IpaB, IpaC, and IpaD via the type III secretion system into the medium (47). In the present study, we wished to clarify the role of Spa32 in the function and formation of the Shigella type III secretion system.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Escherichia coli K-12 MC1061 was used as the host for constructing plasmids. pBluescriptII SK+ (pBS) (Stratagene, La Jolla, Calif.), pMW119Tp (12), pGEX-4T-1 (Amersham Pharmacia Biotech, Piscataway, N.J.), and pCACTUS-Tpr (45) were used for genetic engineering experiments. pTB101-mxiH, denoted previously for its construction, was used as a vector for the overexpression of MxiH in S. flexneri grown in the presence of 1 mM isopropyl-ß-D-thiogalactopyranoside (IPTG) (45). All S. flexneri-derived strains were grown routinely in brain heart infusion (BHI) broth (Difco, Detroit, Mich.) or L-broth (Difco). Strains harboring pMW119Tp-, pTB101-, or pCACTUS-Tpr-based plasmids were grown in Mueller-Hinton (MH) broth (Difco) when selection for trimethoprim resistance was necessary.
Construction of a nonpolar spa32 mutant of S. flexneri. For construction of a nonpolar spa32 mutant, the aphA-3 (kanamycin resistance gene) cassette specifically designed for the construction of a nonpolar mutant was used (31). The aphA-3 cassette carried by plasmid pUC18K2 was obtained from Philippe J. Sansonetti (Institut Pasteur, Paris, France). The nonpolar spa32 mutant was generated as follows. A DNA fragment encompassing nucleotides from position 1505 upstream of the 5" end of the spa32 gene through to nucleotide 22 downstream from the 5" end was amplified by PCR using primers 5"-ACGCGTCGACTTTCACGGCAAACGTTG, containing an SalI site, and 5"-GCCCCATGGCCAAATATAATAGATTACGG, containing a KpnI site. The SalI-KpnI fragment was cloned into pBS, resulting in plasmid p32N. Another DNA fragment encompassing nucleotides from position 838 downstream from the 5" end of the spa32 gene through to nucleotide 2186 was amplified using primers 5"-CGCGGATCCAATAGTACAGGATATAATG, containing a BamHI site, and 5"-ACGAGATCTCGTATACGTCCGTTCTCCC, containing an XbaI site. The BamHI-XbaI fragment was cloned into pBS using the BamHI and XbaI sites, yielding plasmid p32C. The SacI-KpnI fragment digested from p32N was cloned into pUC18K2, resulting in p32NK. The SalI-BamHI fragment digested from p32NK was cloned into p32C, resulting in p32NKC. The inactivated spa32 gene digested from p32NKC at the SalI and XbaI sites was subcloned into pCACTUS-Tpr, a temperature- and sucrose-sensitive suicide vector, followed by introduction of the resultant plasmid into YSH6000 by electroporation. The transformants were grown on L-agar plates supplemented with 5% sucrose without NaCl at 42°C. The resulting sucrose- and kanamycin-resistant colonies were tested for trimethoprim sensitivity, indicative of loss of the suicide vector. One of the trimethoprim-sensitive colonies thus selected was confirmed to contain an insertion of the aphA-3 gene in the spa32 gene, as determined from restriction enzyme digestion of the PCR-amplified segment.
Preparation of antibodies. The antibodies specific for IpaB, IpaC, and IpaD used for immunoblotting were described previously (47). The antibodies specific for MxiD, MxiG, and MxiJ used for immunoblotting were also described previously (45).
The polyclonal rabbit anti-Spa32 antibody was raised against recombinant Spa32 protein. For preparation of recombinant Spa32 protein, the spa32 gene was cloned into the pGEX-4T-1 vector, which produced glutathione S-transferase (GST)-tagged Spa32. Escherichia coli K-12 cells carrying pGEX-4T-1-spa32 were cultivated in L-broth supplemented with ampicillin (50 µg/ml) for 3 h at 37°C. Expression was induced by the addition of 1 mM IPTG and incubation for 2 h at 37°C. Bacteria were disrupted by sonication four times using an ultrasonic disrupter UD-200 (TOMY) for 1 min, with incubation on ice. Purification of the GST fusion proteins with glutathione-Sepharose 4B (Amersham Pharmacia Biotech) and cleavage of the GST proteins with thrombin were performed according to the manufacturer's protocol. The purified recombinant Spa32 proteins after exclusion of the GST proteins were used to immunize rabbits. The anti-Spa32 antiserum was incubated with a nitrocellulose membrane containing transferred recombinant Spa32 protein, and the immunoglobulin fraction specific for Spa32 was eluted.
Analysis of secreted proteins from S. flexneri. A small aliquot of an overnight culture of S. flexneri grown at 30°C in L-broth was inoculated into 40 ml of BHI broth, and the bacteria were grown at 37°C for 2.5 h to an optical density of 1.3 at 600 nm. The culture (1 ml) was used for the preparation of the whole bacterial lysate. The 1-ml sample was centrifuged at 17,000 x g for 2 min at 4°C, and the resultant bacterial pellet was suspended in a sample buffer for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The lysate was boiled for 5 min and stored at 4°C. The remaining 39-ml culture was harvested, and the supernatant of the BHI broth culture was passed through a 0.45-µm-pore-size filter. The bacterial pellet was washed with 30 ml of ice-cold phosphate-buffered saline (PBS) and resuspended in 3 ml of the same buffer. The bacterial suspension was separated equally into two test tubes and incubated at 37°C for 5 min. Congo red (CR) was added to one bacterial suspension at a final concentration of 0.003%, and the bacterial suspensions were further incubated at 37°C for 10 min. After incubation, the suspensions were centrifuged at 17,000 x g for 2 min at 4°C. After centrifugation, the supernatant was passed through a 0.45-µm-pore-size filter. For the filtered BHI broth or PBS with or without CR, trichloroacetic acid (TCA) was added at a final concentration of 6%, followed by incubation on ice for 15 min. After incubation, each of the samples was subjected to centrifugation at 17,000 x g for 10 min at 4°C. The pellet was suspended in 50 µl of sample buffer for SDS-PAGE, boiled for 5 min, and stored at 4°C.
Isolation of the type III needle complexes from S. flexneri. The type III needle complexes of S. flexneri were partially or extensively purified by the method described previously (45).
Electron microscopy. (i) Observation of purified type III needle complexes. The extensively purified needle complexes of S. flexneri were placed onto carbon-coated copper grids and stained with 2% phosphotungstic acid (pH 7.0). The needle complexes were observed and their pictures were taken with a transmission electron microscope (JEM-2000ES; Jeol, Tokyo, Japan).
(ii) Observation of type III needle portions extending from S. flexneri. S. flexneri (1 ml) cultured at 37°C to an optical density of 1.3 at 600 nm was harvested, washed once with 1 ml of 5 mM phosphate buffer (pH 7.0) containing 50 mM NaCl, and suspended in 50 to 500 µl of the same buffer. The bacterial suspension was placed onto carbon-coated copper grids, stained, and observed by transmission electron microscopy (TEM) as described above for needle complex observation. The electron micrographs of the bacterial images were digitized and processed by unsharp masking on StellaImage (AstroArts, Tokyo, Japan) with some modification to cover extreme density differences. This technique is quite effective at enhancing the hidden detail of digital images, and it is sometimes applied to process astronomical and radiographic images (49).
Invasion assay. The gentamicin protection assay was used to assay the invasion of S. flexneri in HeLa cells. The assay was performed as described previously (45).
Complementation test of the spa32 mutant with the spa32, invJ, fliK, or yscP gene. The spa32 gene was amplified from YSH6000. The invJ gene was amplified from S. enterica serovar Typhimurium wild-type SB300, which was obtained from Jorge E. Galán (Yale School of Medicine, New Haven, Conn.) (14). The fliK gene was also amplified from SB300. The yscP gene was amplified from Yersinia pseudotuberculosis wild-type YP100(pIB1), which was obtained from Hans Wolf-Watz (Umeå University, Umeå, Sweden) (17). Each amplified gene was cloned into pMW119Tp in-frame at the BamHI and EcoRI sites under the lacZ promoter, and each plasmid clone of spa32, invJ, fliK, or yscP was introduced into the spa32 mutant by electroporation. The observation of needle complexes and the invasion assay of each complementary strain were performed as described above.
Amino acid sequence analysis of Spa32. Sequence homology searches were performed with BLAST2 and FASTA, and a sequence alignment was carried out with CLUSTALW ver. 1.8 on the GenomeNet server (http://www.genome.ad.jp) using standard parameters.
|
RESULTS AND DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
spa32) was investigated for the structure of the type III secretion machinery. The extensively purified type III secretion complexes from the spa32 mutant, as observed by TEM, possessed needles of various lengths (Fig. 1A, panel a). The range of needle lengths (230 individual complexes) indicated that although there was a peak at around 40 to 50 nm, the length was distributed over a very broad range from 40 to 1,150 nm (Fig. 1B, panel a). Upon introduction of a low-copy-number plasmid, pMW119Tp, clone of the wild-type spa32 gene into the spa32 mutant (spa32/pMW119Tp-spa32), the production of type III secretion complexes was restored similar to that in M94 (wild type) (Fig. 1A, panel c). Indeed, the range of needle lengths in spa32/pMW119Tp-spa32 was 42 to 45 nm, similar to that in M94, suggesting that the presence of the spa32 gene in S. flexneri is required for the narrow range of type III needle lengths (Fig. 1B, panel b, spa32/pMW119Tp-spa32; panel c, M94).
|
spa32 mutant caused by introducing pTB101-mxiH, the bacteria generated type III complexes with extremely long needles. As shown in Fig. 1A (panel e), the elongated needles were estimated to be over 5 µm, which is 100 to 200 times longer than the wild-type needles. The surface of the bacterium with negative staining as observed by TEM indicated that numerous long needles extended from the cell surface of the spa32 mutant (Fig. 1A, panel b) and spa32/pTB101-mxiH (Fig. 1A, panel f), whereas needles of normal length were not observed directly on M94 (wild-type S. flexneri) (data not shown) or spa32/pMW119Tp-spa32 (Fig. 1A, panel d) because of an accumulation of staining solution around the intact bacterium.
To confirm that the elongated needles from the spa32 mutant and spa32/pTB101-mxiH were composed of MxiH, the type III secretion complexes were extracted from M94, the spa32 mutant, spa32/pMW119Tp-spa32, spa32/pTB101-mxiH, and del-17 (type III-deficient mutant), and the protein composition of partially purified type III complexes was investigated by separation in a Coomassie brilliant blue (CBB)-stained SDS-20% PAGE gel. As shown in Fig. 2, the type III complexes extracted from all of the strains except the del-17 mutant contained MxiD, MxiG, and MxiJ, the major components of the basal body (45). They also contained an 
9-kDa protein, which was previously identified as MxiH (45). Importantly, the 9-kDa band was slightly thicker in the spa32 mutant than in M94, while the band in spa32/pMW119Tp-spa32 was similar to that in M94 (Fig. 2). Furthermore, a large amount of the 9-kDa protein was contained in the partially purified type III secretion complexes from spa32/pTB101-mxiH (Fig. 2). In fact, the 9-kDa band of the type III secretion complexes partially purified from M94, the spa32 mutant, spa32/pTB101-mxiH, and spa32/pMW119Tp-spa32 was confirmed to be MxiH by direct N-terminal amino acid sequencing (data not shown). It is worth noting that the 9-kDa protein band also contained MxiI, because the sequencing also yielded an MNYIY sequence corresponding to the N-terminal amino acids of MxiI, even though it was detected as a minor fraction, as reported previously (4).
|
spa32 mutant contained slightly larger amounts of MxiH than the wild type, the amounts of major components of the basal body such as MxiD, MxiG, and MxiJ appeared to be similar. Because Spa32 was hardly detected in the partially purified type III complexes from M94 as examined by immunoblotting with anti-Spa32 antibody (see Materials and Methods) (data not shown), the findings of this series of experiments suggested that Spa32 was not the structural component, but rather was somehow required for the narrow range of type III needle lengths in S. flexneri.
Interestingly, the peak of the type III needle lengths of the spa32 mutant was still approximately 45 nm, the needle length of the wild type (Fig. 1B, panel a). Similarly, in the Salmonella flagella, the peak of the hook lengths of the fliK mutant was the same as the hook length of the wild type (23). These results suggest that the S. flexneri needle and Salmonella hook are both still controlled even in the absence of Spa32 and FliK, respectively. Recent studies have led to the hypothesis that the C-ring compartment formed in the cytoplasm beneath the MS ring is involved in determining the length of the hook on the flagellar basal body (13, 21, 27, 30). The C-ring is predicted to function as a compartment for the storage of a fixed amount of the flagellar hook component FlgE proteins, although there is no direct evidence yet that FlgE fills the C-ring, and thus the mechanism of length control of the hook is still largely speculative (30). In S. flexneri, it is worth noting that the type III secretion machinery is also indicated to possess a compartment, like the flagellar C-ring, beneath the basal body, named the bulb structure, as investigated in the osmotically shocked Shigella envelope by TEM (3). Therefore, if the C-ring is in fact involved in controlling hook length, a similar mechanism might control the type III needle length.
Spa32 is translocated through the type III secretion machinery into the medium.
Although the extent of amino acid identity is moderate, InvJ is one of the proteins most homologous to Spa32. Importantly, InvJ is an essential secreted protein from Salmonella via the type III secretion system for the secretion of effectors such as SipB and SipC and bacterial invasiveness (6, 7). We therefore investigated whether Spa32 was secreted from S. flexneri into its environment. M94 (wild type), a mxiH mutant (mxiH, needleless mutant), and a spa47 mutant (spa47, type III-associated ATPase-deficient mutant) were grown in BHI broth at 37°C for 2.5 h, and the secretion of Spa32 into the medium was assessed by immunoblotting with anti-Spa32 antibody.
As shown in Fig. 3, Spa32 was detected in the medium from M94 but not the mxiH or spa47 mutant. The secretion of effector proteins such as IpaB, IpaC, and IpaD from S. flexneri is stimulated by suspending bacteria in PBS containing 0.003% CR, a conditional medium that stimulates type III secretion (1, 35). Thus, bacteria harvested from the culture were resuspended in PBS containing CR, and secreted proteins were checked by immunoblotting with anti-Spa32, anti-IpaB, anti-IpaC, and anti-IpaD antibodies. Although IpaB, IpaC, and IpaD were detected in PBS plus CR from M94 but not from the mxiH or spa47 mutant, Spa32 was barely detected from M94. Importantly, although Spa32 was produced by the mxiH and spa47 mutants, as determined by immunoblotting of whole bacterial lysate with anti-Spa32 antibody, its secretion was not detected at all in the medium (Fig. 3). These findings suggested that Spa32 was translocatable via the type III secretion system and that, unlike IpaB, IpaC, and IpaD, its secretion was independent of stimulation of the type III secretion system. Therefore, because the secretion of Spa32 was not the same as that of Ipa proteins and occurred during bacterial growth, we believe that Spa32 is necessary for the formation of the type III secretion complexes rather than the bacterial infection process.
|
mxiH/pTB101-mxiH), mediated secretion of IpaB, IpaC, and IpaD and, more importantly, increased the invasiveness for HeLa cells (more than twofold over the wild-type level) (45). We thus investigated the spa32 mutant and spa32/pTB101-mxiH for the ability to secrete IpaB, IpaC, and IpaD as well as to invade HeLa cells. The spa32 mutant, spa32/pTB101-mxiH, M94, spa32/pMW119Tp-spa32, and S325 (type III-deficient mutant) were examined for their ability to secrete Ipa proteins into the medium by immunoblotting with anti-IpaB, anti-IpaC, and anti-IpaD antibodies. Although M94 and spa32/pMW119Tp-spa32, but not S325, were able to secrete Ipa proteins into PBS plus CR, secretion was poorly detected from spa32/pTB101-mxiH and not at all from the spa32 mutant with suspensions in PBS plus CR (Fig. 4A).
|
spa32 mutant and spa32/pTB101-mxiH was greatly reduced compared with that of M94 and spa32/pMW119Tp-spa32, and thus the type III secretion machinery created in the absence of Spa32 in S. flexneri was not functional. Spa32 is interchangeable with Salmonella InvJ. Salmonella InvJ (35 kDa) shows moderate amino acid homology with Spa32 (Fig. 5A), while Salmonella flagellar FliK and Yersinia YscP seem to show marginal amino acid homology with Spa32, as determined by the sequence alignment (see the introduction). The latter two proteins are slightly larger (FliK, 42 kDa; YscP, 50 kDa) than Spa32 (32 kDa), and indeed, there are many gaps between the homologous amino acid regions of FliK (as well as YscP) and Spa32. However, the placements of spa32, invJ, fliK, and yscP in the respective operons were all similar, and each of the gene products was secreted (6, 32, 36, 40, 42; this study). Furthermore, the production of each protein was required for a functional type III secretion system (Spa32, InvJ, and YscP) (7, 36, 42; this study) or flagellar export system (FliK) (20, 33, 48). Consequently, we wondered whether Spa32 and its putative cognates would share similar functions.
|
spa32 mutant and the bacteria were investigated for production of the type III secretion machinery. We extracted the type III secretion complexes from each of the strains under the same conditions and analyzed the supramolecular structure by TEM. The findings indicated that expression of InvJ but not FliK or YscP in the spa32 mutant produced type III secretion complexes similar to those of the wild-type Shigella strain (Fig. 5B).
To investigate whether InvJ, FliK or YscP expressed in the spa32 mutant would compensate for the absence of Spa32 in the type III secretion system, the invasiveness of each of the strains in HeLa cells was examined using the gentamicin protection assay. As shown in Fig. 4B, although the introduction of a cloned fliK or yscP gene into the spa32 mutant was unable to restore invasiveness, the introduction of a cloned invJ gene did restore invasiveness. Although the important amino acid sequences in Spa32 that share function with InvJ remain to be identified, Salmonella InvJ shared an essential function with Shigella Spa32 in the type III secretion system. On the other hand, under the same conditions, FliK could not restore the function of the type III secretion system in the spa32 mutant. Interestingly, upon overproduction of FlgE in the Salmonella fliK mutant, the flagellar export complexes extended aberrant long hooks, called superpolyhooks (34). Thus, although their functions are not interchangeable, the role of Spa32 in the type III secretion system is thought to be similar to that of FliK in the flagellar export system.
In summary, we provide the first structural and functional evidence that Spa32 is a type III-dependent translocatable protein required for activating the type III secretion system and responsible for the narrow range of type III needle lengths. Furthermore, although the homology between Spa32 and Salmonella InvJ is moderate, the role of Spa32 in controlling Shigella type III needle length can be replaced by InvJ, providing valuable information for determining the essential features for the function of these regulatory proteins.
|
ACKNOWLEDGMENTS |
|---|
This work was supported by the Research for the Future Program of the Japan Society for the Promotion of Science.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Appl. Environ. Microbiol. | Infect. Immun. | Eukaryot. Cell |
|---|---|---|
| Mol. Cell. Biol. | J. Virol. | Microbiol. Mol. Biol. Rev. |
| ALL ASM JOURNALS |
