A Microdomain for Protein Secretion in Gram-Positive Bacteria

See allHide authors and affiliations

Science  04 Jun 2004:
Vol. 304, Issue 5676, pp. 1513-1515
DOI: 10.1126/science.1097404


Gram-positive bacteria face unique challenges in generating biologically active conformations for their exported proteins because they lack a dedicated compartment for folding secreted polypeptides. We have discovered that protein secretion by way of the general secretory (Sec) pathway in the important human pathogen Streptococcus pyogenes proceeds through a single microdomain. Unlike other mechanisms for asymmetry involving the Sec pathway, proteins destined for secretion are targeted to a single locus distal to either cell pole that has specialized to contain the Sec translocons. This subcellular organization may represent a paradigm for secretion common to other Gram-positive pathogens with profound implications for pathogenesis.

How to secrete a polypeptide across a cellular membrane and then assist that polypeptide in folding into its final three-dimensional conformation is a fundamental problem that all organisms must solve. This is particularly true for Gram-positive bacterial pathogens such as Streptococcus pyogenes (group A streptococcus), the causative agent of numerous human diseases ranging from superficial and self-limiting (e.g., pharyngitis or “strep throat”) to invasive and life threatening (e.g., necrotizing fasciitis). To cause these diseases, S. pyogenes may secrete more than 40 different polypeptides that are ultimately released into the extracellular milieu (1). However, the pathways that promote secretion and folding of these polypeptides are not well understood.

Gram-positive bacteria lack a specialized compartment exterior to the membrane, like the periplasmic space of Gram-negative bacteria or the endoplasmic reticulum of eukaryotic cells, which organizes accessory proteins that promote the folding of nascent polypeptides after their secretion. Hence, proteins exported from a Gram-positive bacterium fold in an environment exposed to the unregulated conditions of the external milieu. The thick peptidoglycan cell wall coats the membrane, and this coating may inhibit the diffusion of proteins greater than 25 to 50 kD across the cell lattice into the external environment (2, 3). So far it is not known how the secretion machinery is organized in Gram-positive bacteria to allow proteins to fold after secretion or how transit across the cell wall is coordinated.

Despite what is not understood, recent studies have provided evidence to suggest that protein secretion is a complex and organized process in Gram-positive bacteria. For example, one of the most abundant proteins secreted from S. pyogenes is the SpeB cysteine protease. Proper targeting and maturation of nascent SpeB requires the ribosome-associated chaperone Trigger factor (4, 5) and the cell membrane–associated chaperone HtrA (6, 7). Other accessory factors contribute to SpeB maturation (8, 9), but how these factors are coordinated with the general secretory (Sec) pathway is unknown.

Examination of the sequences of multiple genomes has revealed that the streptococci contain only the Sec pathway for protein export. Thus, they lack specialized pathways like the type III secretion systems found in Gram-negative bacteria that function to inject bacterial effector proteins into the cytosol of host cells (10). A functionally analogous, although structurally different, pathway does exist in S. pyogenes that uses the poreforming cytolysin protein streptolysin O. Similar to type III secretion, considerable data suggest that cytolysin-mediated translocation (CMT) is a polarized process for injection of an effector protein into the host cell cytosol and not into the extracellular environment (10). Polarity may involve the formation of a protected channel between the bacterium and the host cell (11). However, unlike type III secretion, the Sec pathway exports both the cytolysin and the injected effector protein. It is not clear how the Sec pathway can distinguish between polarized and circumferential secretion processes.

A key to understanding the CMT pathway and the biogenesis of other secreted streptococcal virulence factors like SpeB will come from an understanding of how the Sec pathway is organized and integrated with accessory folding factors. To begin to address this question, we sought to develop “snapshots” of the secretion process that would reveal the cellular location of the Sec secretion machinery. As a model secreted polypeptide, we chose SpeB, owing to its abundance, the availability of reagents, and our ability to manipulate SpeB expression. Cultures were grown to the late exponential phase of growth to the precise time that secretion of SpeB initiates at a maximal rate, as determined by Western blotting. Cells were harvested, fixed, and then subjected to progressively lower temperature embedding. Thin sections were prepared and examined by immunogold electron microscopy (EM) after their reaction with a SpeB-specific polyclonal antiserum. Examination of stained sections revealed a single intense focus of gold particles at a discrete location adjacent to the membrane (Fig. 1, A to C) that also extends into the adjacent cell wall (Fig. 1, D and E). A single cluster in an asymmetric hemispherical location was always observed, which indicates that the antiserum is labeling SpeB that is targeted to a discrete focus rather than to an equatorial ring. In addition, the focus is located distally to both the new and old poles of the cell to a site that may be adjacent to the newly forming septum of dividing cells. Minimal gold particles were observed in cells harvested in the early exponential or late stationary phases of growth (Fig. 1, F and G), consistent with minimal expression of SpeB at these times. Similarly, no gold particles were observed for a mutant deficient in a regulatory gene that is absolutely required for expression of the SpeB gene (Fig. 1H).

Fig. 1.

SpeB is secreted from a single microdomain. Electron micrographs of wild-type (WT) streptococcal cells display a single intensely stained focus of gold particles (A, B, and C; bar, 0.5 μM) that was located adjacent to the membrane and extending into the adjacent cell wall (D and E; bar, 0.1 μM). Minimal staining was observed in logarithmic and stationary phases (F and G). No staining was observed in a mutant lacking an essential activator of transcription of the SpeB gene (ΔRopB) (H).

The pattern of SpeB secretion was also examined by an independent assay. An early exponential culture was subjected to brief sonication to disrupt streptococcal chains to single cocci and then embedded in acidic agarose to induce expression of SpeB. A pH-insensitive protease substrate (casein BODIPY TR-X) was included in the agarose to allow visualization of mature proteolytic activity in live cells (12). After cleavage, the substrate produces a fluorescent product as a result of the disruption of intramolecular quenching. When examined by fluorescent microscopy, each streptococcal cell displayed a single intense focus of fluorescence at a discrete location adjacent to the cellular membrane that was distal to the poles of the cell (Fig. 2A). The fluorescent pattern was the product of SpeB protease activity, as shown by the >90% reduction in the number of punctate cells both in the presence of a protease inhibitor and upon the introduction of a mutation that eliminates transcription of the gene encoding SpeB (Fig. 2B). Overall, this pattern of protease trafficking was identical to that observed by EM.

Fig. 2.

Live streptococcal cells secrete enzymatically active SpeB at a single locus. Representative micrographs show that each cell displays a single punctate fluorescent locus (A). Quantification revealed that fluorescence was abrogated by supplementation with protease inhibitors to the same levels observed in ΔRopB, indicating that the fluorescent pattern was dependent on catalytically active SpeB (B).

These data suggest that protein secretion is occurring at a single microdomain in the cellular membrane. However, compartmentalization of transcription and translation has been described for the Gram-positive bacterium Bacillus subtilis (13). Thus, it is possible that the staining pattern is the result of the trafficking of nascent SpeB to a membrane site adjacent to the site of transcription and translation on the chromosome. To distinguish between these possibilities, the trafficking of a protein encoded on a high copy number plasmid was examined, because it is unlikely that all copies of the plasmid would be restricted to a specific compartment. Concurrently, we examined expression of PhoZ, a secreted protein not native to S. pyogenes, to determine whether the targeting pattern was a general phenomenon or one restricted to SpeB. Analysis by immunogold EM using PhoZ antiserum revealed a pattern of gold particles clustered at a discrete focus identical to that observed for SpeB (Fig. 3, A and B). This contrasts with the circumferential distribution of gold particles observed using an antiserum directed against the M protein (Fig. 3, C and D) and the lack of gold particles observed in sections stained in the absence of the primary antibodies (Fig. 3, E and F). After its secretion from the cell, M protein becomes cross-linked to the cell wall (14) and is subsequently uniformly distributed over the cell surface in late exponential phase (15). Taken together, these data indicate that protein secretion in S. pyogenes occurs at a distinct microdomain of the cytoplasmic membrane dedicated to protein export. We have named this novel microdomain the ExPortal.

Fig. 3.

Protein secretion is targeted to a distinct microdomain. Streptococcal cells expressing the secreted alkaline phosphatase PhoZ (A and B). Cell-wall-anchored M protein displays a circumferential distribution (C and D), and omission of the primary antibody results in no detectable label (E and F). Magnification as in Fig. 1.

Asymmetric secretion and localization of proteins are essential to many processes in bacteria (16). In some cases, asymmetric secretion is the result of polar localization of a specialized secretion complex (12). For the general secretory pathway, asymmetry does not appear to be due to a restricted distribution of the Sec pathway translocons but rather to various specialized accessory factors (16, 17). Because SpeB and PhoZ are substrates for the Sec pathway, we examined whether secretion was directed by accessory targeting factors or by a restricted distribution of the Sec translocons. For this analysis, streptococci were analyzed by immunogold EM with an antiserum raised against SecA of Bacillus subtilis. This component of the Sec pathway is a highly conserved ATPase that directly interacts with the SecYEG translocase in the membrane and powers translocation of polypeptides through the Sec channel (18). This antiserum specifically recognized SecA of S. pyogenes, as determined by a Western blot analysis of cell lysates, and localized SecA exclusively to a single locus as visualized in immunogold electron micrographs (Fig. 4, A and B). Colocalization of large and small gold particles following staining with both rabbit anti-SecA and mouse monoclonal anti–SpeB antibodies confirmed that SecA is exclusively located in the ExPortal (Fig. 4, C and D). These data are consistent with a model in which targeting of secreted proteins results from the specialization of this microdomain to accumulate a high concentration of the Sec translocons.

Fig. 4.

The Sec machinery is exclusively localized. SecA is localized to a single focus (A and B). Double-labeling experiments demonstrated that only a single cluster was observed (C) that contained antibodies recognizing both SecA (18 nm beads) and SpeB (12 nm beads) (D). Note that the label is often associated with a nascent division septum. Magnification as in Fig. 1.

These data represent a mechanism for asymmetric secretion of proteins from bacteria using the Sec pathway distinct from those previously described. This localization may serve to organize accessory folding factors to coordinate their interaction with nascent unfolded polypeptides. Most examples of asymmetry in bacteria involve localization of proteins and structures to the poles of the cell, possibly by exploiting differences in lipid and cell wall content between the poles and the lateral surfaces (16). However, because the streptococcal translocons do not appear to localize to the poles, a different mechanism may be used to establish their restricted positioning. The contribution of the ExPortal to host-pathogen interactions raises intriguing possibilities, including possible cooperation with the CMT pathway for introduction of streptococcal effector proteins into the host-cell cytoplasm. Continued analysis will enrich our understanding of how streptococci cause disease and may provide greater insight into the protein secretion pathways of other Gram-positive bacteria.

Supporting Online Material

Materials and Methods


References and Notes

View Abstract

Navigate This Article