Topologically-guided continuous protein crystallization controls bacterial surface layer self-assembly.

Many bacteria and most archaea possess a crystalline protein surface layer (S-layer), which surrounds their growing and topologically complicated outer surface. Constructing a macromolecular structure of this scale generally requires localized enzymatic machinery, but a regulatory framework for S-layer assembly has not been identified. By labeling, superresolution imaging, and tracking the S-layer protein (SLP) from C. crescentus, we show that 2D protein self-assembly is sufficient to build and maintain the S-layer in living cells by efficient protein crystal nucleation and growth. We propose a model supported by single-molecule tracking whereby randomly secreted SLP monomers diffuse on the lipopolysaccharide (LPS) outer membrane until incorporated at the edges of growing 2D S-layer crystals. Surface topology creates crystal defects and boundaries, thereby guiding S-layer assembly. Unsupervised assembly poses challenges for therapeutics targeting S-layers. However, protein crystallization as an evolutionary driver rationalizes S-layer diversity and raises the potential for biologically inspired self-assembling macromolecular nanomaterials.

A Caulobacter crescentus Microbicide Protects from Vaginal Infection with HIV-1JR-CSF in Humanized Bone Marrow-Liver-Thymus Mice.

Over 2 million people are infected with HIV-1 annually. Approximately half of these new infections occur in women residing in low-income countries, where their access to and control over HIV-1 preventative measures are often limited, indicating that female-controlled prevention options for HIV-1 are urgently needed. Microbicides that can be topically applied to the vaginal tract in advance of sexual activity represent a promising female-controlled prevention option for HIV-1. We have previously described the development of an HIV-1-specific microbicide using the surface or S-layer recombinant protein display capabilities of the nonpathogenic, freshwater bacterium Caulobacter crescentus Recombinant C. crescentus bacteria were created that displayed proteins that interfere with the HIV-1 attachment and entry process and that were able to provide significant protection of TZM-bl cells from infection with HIV-1 pseudovirus. These studies have been expanded to investigate if these recombinant C. crescentus bacteria are able to maintain efficacy with replication-competent HIV-1 and both TZM-bl cells and human peripheral blood mononuclear cells (PBMCs). In addition, we utilized the humanized bone marrow-liver-thymus (BLT) mouse model to determine if vaginal application of recombinant C. crescentus at the time of HIV-1JR-CSF infection could provide protection from HIV-1 infection. Recombinant C. crescentus bacteria expressing Griffithsin, GB virus C E2 protein, elafin, α-1-antitrypsin, indolicidin, and the fusion inhibitor T-1249 were able to protect 40 to 75% of the BLT mice from vaginal infection with HIV-1JR-CSF, with C. crescentus bacteria expressing Griffithsin being the most effective. Taken together, these data suggest that a C. crescentus-based microbicide could be a safe and effective method for HIV-1 prevention.IMPORTANCE Human immunodeficiency virus (HIV) disproportionally infects young women in sub-Saharan Africa. Current HIV-1 prevention options have had limited success among women, suggesting that alternative, female-controlled prevention options need to be developed. Microbicides that can be applied to the vaginal tract are a promising prevention option. In this study, we describe the testing of 15 potential candidates for inhibition of HIV-1 infection in a humanized mouse model of HIV-1 infection. Four of these candidates were able to provide significant protection from vaginal infection with HIV-1, with the most successful candidate protecting 75% of the mice from infection. This study describes the preclinical testing of a new strategy that could be a safe and effective option for HIV-1 prevention in women.

Generation of a Dual-Target, Safe, Inexpensive Microbicide that Protects Against HIV-1 and HSV-2 Disease.

HSV-2 infection is a significant health problem and a major co-morbidity factor for HIV-1 acquisition, increasing risk of infection 2-4 fold. Condom based prevention strategies for HSV-2 and HIV-1 have not been effective at stopping the HIV-1 pandemic, indicating that alternative prevention strategies need to be investigated. We have previously developed an inexpensive HIV-1 specific microbicide that utilizes the S-layer mediated display capabilities of Caulobacter crescentus, and have shown that recombinant C. crescentus displaying HIV entry blocking proteins are able to provide significant protection from HIV-1 infection in vitro. Here we demonstrate that recombinant C. crescentus are safe for topical application and describe 5 new recombinant C. crescentus that provide protection from HIV-1 infection in vitro. Further, we demonstrate protection from disease following intravaginal infection with HSV-2 in a murine model using C. crescentus expressing the anti-viral lectins Cyanovirin-N and Griffithsin, as well as α-1-antitrypsin and indolicidin. Interestingly, C. crescentus alone significantly reduced HSV-2 replication in vaginal lavage fluid. Protection from HSV-2 disease was strongly associated with early cytokine production in the vaginal tract. Our data support the potential for a dual-target microbicide that can protect against both HIV-1 and HSV-2, which could have an enormous impact on public health.

Environmental Calcium Controls Alternate Physical States of the Caulobacter Surface Layer.

Surface layers (S-layers) are paracrystalline, proteinaceous structures found in most archaea and many bacteria. Often the outermost cell envelope component, S-layers serve diverse functions including aiding pathogenicity and protecting against predators. We report that the S-layer of Caulobacter crescentus exhibits calcium-mediated structural plasticity, switching irreversibly between an amorphous aggregate state and the crystalline state. This finding invalidates the common assumption that S-layers serve only as static wall-like structures. In vitro, the Caulobacter S-layer protein, RsaA, enters the aggregate state at physiological temperatures and low divalent calcium ion concentrations. At higher concentrations, calcium ions stabilize monomeric RsaA, which can then transition to the two-dimensional crystalline state. Caulobacter requires micromolar concentrations of calcium for normal growth and development. Without an S-layer, Caulobacter is even more sensitive to changes in environmental calcium concentration. Therefore, this structurally dynamic S-layer responds to environmental conditions as an ion sensor and protects Caulobacter from calcium deficiency stress, a unique mechanism of bacterial adaptation. These findings provide a biochemical and physiological basis for RsaA's calcium-binding behavior, which extends far beyond calcium's commonly accepted role in aiding S-layer biogenesis or oligomerization and demonstrates a connection to cellular fitness.

Surface-layer protein from Caulobacter crescentus: expression, purification and X-ray crystallographic analysis.

Protein surface layers are self-assembling, paracrystalline lattices on the surface of many prokaryotes. Surface-layer proteins have not benefited from widespread structural analysis owing to their resistance to crystallization. Here, the successful expression of a truncated version of RsaA, the surface-layer protein from Caulobacter crescentus, from a Caulobacter protein-expression system is reported. The purification, crystallization and initial X-ray diffraction analysis of the truncated RsaA, the largest surface-layer protein studied to date and the first from a Gram-negative bacterium, are also reported.

The core and O-polysaccharide structure of the Caulobacter crescentus lipopolysaccharide.

Here we describe the analysis of the structure of the lipopolysaccharide (LPS) from Caulobacter crescentus strain JS1025, a derivative of C. crescentus CB15 NA1000 with an engineered amber mutation in rsaA, leading to the loss of the protein S-layer and gene CCNA_00471 encoding a putative GDP-L-fucose synthase. LPS was isolated using an aqueous membrane disruption method. Polysaccharide and core oligosaccharide were produced by mild acid hydrolysis and analyzed by nuclear magnetic resonance spectroscopy and chemical methods. Spectra revealed the presence of two polysaccharides, one of them, a rhamnan, could be removed using periodate oxidation. Another polymer, built from 4-amino-4-deoxy-D-rhamnose (perosamine), mannose, and 3-O-methyl-glucose, should be the O-chain of the LPS according to genetic data. The attribution of the rhamnan as a part of LPS or a separate polymer was not possible.

Development of an HIV-1 Microbicide Based on Caulobacter crescentus: Blocking Infection by High-Density Display of Virus Entry Inhibitors.

The HIV/AIDS pandemic remains an enormous global health concern. Despite effective prevention options, 2.6 million new infections occur annually, with women in developing countries accounting for more than half of these infections. New prevention strategies that can be used by women are urgently needed. Topical microbicides specific for HIV-1 represent a promising prevention strategy. Conceptually, using harmless bacteria to display peptides or proteins capable of blocking entry provides an inexpensive approach to microbicide development. To avoid the potential pitfalls of engineering commensal bacteria, our strategy is to genetically display infection inhibitors on a non-native bacterium and rely on topical application of stabilized bacteria before potential virus exposure. Due to the high density cell-surface display capabilities and the inherent low toxicity of the bacterium, the S-layer mediated protein display capabilities of the non-pathogenic bacterium Caulobacter crescentus has been exploited for this approach. We have demonstrated that C. crescentus displaying MIP1α or CD4 interfered with the virus entry pathway and provided significant protection from HIV-1 pseudovirus representing clade B in a standard single cycle infection assay. Here we have expanded our C. crescentus based microbicide approach with additional and diverse classes of natural and synthetic inhibitors of the HIV-1 entry pathway. All display constructs provided variable but significant protection from HIV-1 infection; some with protection as high as 70%. Further, we describe protection from infection with additional viral clades. These findings indicate the significant potential for engineering C. crescentus to be an effective and readily adaptable HIV-1 microbicide platform.

Evaluating secretion and surface attachment of SapA, an S-layer-associated metalloprotease of Caulobacter crescentus.

Caulobacter crescentus is used to display foreign peptides at high density as insertions into the surface (S)-layer protein (RsaA). Many recombinant RsaA proteins, however, are cleaved by SapA, a 71-kDa metalloprotease, suggesting a role in maintaining S-layer integrity. When overexpressed on a multicopy plasmid SapA was detected on the surface by fluorescent antibody only if RsaA and the O-side chain of LPS that mediates S-layer attachment were removed by mutation, indicating an outer membrane location beneath the S-layer. Secretion was mediated by the RsaA type 1 transporter since secretion was eliminated in transporter deficient strains or by C-terminal deletions in SapA (the presumed location of type 1 secretion signals). Secretion was required to become an active protease; mass spectrometry suggested this might be due to N-terminal processing during secretion, a feature shared with other type 1-secreted proteases. Overexpression leads to additional processing C-terminal to the protease domain, producing a 45-kDa protein. This was demonstrated to be self-processing. Deletion analysis revealed the C-terminal 100 amino acids were sufficient for anchoring and secretion. When protein G was fused to the last 238 amino acids of SapA it was secreted, surface attached and bound immunoglobulin, indicating potential for foreign protein display.

Enhanced neutralization of HIV by antibodies displayed on the S-layer of Caulobacter crescentus.

Innovative methods of prevention are needed to stop the more than two million new HIV-1 infections annually, particularly in women. Local application of anti-HIV antibodies has been shown to be effective at preventing infection in nonhuman primates; however, the concentrations needed are cost prohibitive. Display of antibodies on a particulate platform will likely prolong effectiveness of these anti-HIV agents and lower the cost of goods. Here, we demonstrate that the bacterium Caulobacter crescentus and its highly expressed surface-layer (S-layer) protein can provide this antibody display platform. Caulobacters displaying protein G, alone or with CD4 codisplay, successfully captured HIV-1-specific antibodies and demonstrated functional neutralization. Compared to soluble antibodies, a neutralizing anti-HIV antibody displayed on Caulobacter was as effective or more effective at neutralizing diverse HIV-1 isolates. Moreover, when an antibody reactive with an epitope induced by CD4 binding (CD4i) was codisplayed with CD4, there was significant enhancement in HIV-1 neutralization. These results suggest that caulobacters displaying anti-HIV antibodies offer a distinct improvement in the use of antibodies as microbicides. Furthermore, these reagents can specifically evaluate anti-HIV antibodies in concert with other HIV-1 blocking agents to assess the most suitable tools for conversion to scFvs, allowing for direct display within the S-layer protein and further reducing cost of goods. In summary, C. crescentus, which can be easily produced and chemically stabilized at low cost, is well suited for engineering as an effective platform, offering an inexpensive way to produce and deliver HIV-1-specific microbicides.

Analysis of the intact surface layer of Caulobacter crescentus by cryo-electron tomography.

The surface layers (S layers) of those bacteria and archaea that elaborate these crystalline structures have been studied for 40 years. However, most structural analysis has been based on electron microscopy of negatively stained S-layer fragments separated from cells, which can introduce staining artifacts and allow rearrangement of structures prone to self-assemble. We present a quantitative analysis of the structure and organization of the S layer on intact growing cells of the Gram-negative bacterium Caulobacter crescentus using cryo-electron tomography (CET) and statistical image processing. Instead of the expected long-range order, we observed different regions with hexagonally organized subunits exhibiting short-range order and a broad distribution of periodicities. Also, areas of stacked double layers were found, and these increased in extent when the S-layer protein (RsaA) expression level was elevated by addition of multiple rsaA copies. Finally, we combined high-resolution amino acid residue-specific Nanogold labeling and subtomogram averaging of CET volumes to improve our understanding of the correlation between the linear protein sequence and the structure at the 2-nm level of resolution that is presently available. The results support the view that the U-shaped RsaA monomer predicted from negative-stain tomography proceeds from the N terminus at one vertex, corresponding to the axis of 3-fold symmetry, to the C terminus at the opposite vertex, which forms the prominent 6-fold symmetry axis. Such information will help future efforts to analyze subunit interactions and guide selection of internal sites for display of heterologous protein segments.

Analysis of high-level S-layer protein secretion in Caulobacter crescentus.

Caulobacter crescentus exhibits a hexagonally arranged protein layer on its outermost surface. RsaA, the sole protein of this "S-layer", is secreted by a type I (ABC) transporter. Few type I transporters show high-level secretion, and few bacterial S-layers have been carefully examined for the amount of protein synthesis capacity needed to maintain cell coverage. Here we determined RsaA levels by quantitative immunoblotting methods, learned that very stable mRNA is a key factor in high-level secretion, and found that the transporter was capable of still higher secretion. A propensity for RsaA to aggregate was a barrier to quantitation, but with the use of S-layer shedding mutants and methods to keep RsaA soluble, we learned that approximately 31% of cell protein is RsaA. When multiple copies of rsaA were introduced, the level increased to approximately 51% of cell protein, a higher level than we are aware of for any protein in any bacterium. Unexpectedly, in comparing normal and S-layer shedding strains, an assembled S-layer was not a significant barrier to elevated secretion. The rsaA mRNA half-life was determined by real-time PCR to be 36 min, ranking with the most stable known in bacteria. A modification of the 5' region resulted in a shorter half-life and a reduction in maximum protein synthesis levels. If secretion was prevented by knockout of type I transporter genes, RsaA levels dropped to 10% or less of normal, but with no significant reduction in rsaA mRNA. Overall, normal levels of RsaA were unexpectedly high, and still higher levels were not limited by transporter capability, the presence of an assembled S-layer, or the capacity of the cell's physiology to produce large amounts of one protein. The normal upper limit of RsaA production appears to be controlled only by the level of an unusually stable message. Significant down-regulation is possible and is accomplished posttranscriptionally.

Development of an HIV-1 specific microbicide using Caulobacter crescentus S-layer mediated display of CD4 and MIP1alpha.

The development of alternative strategies to prevent HIV infection is a global public health priority. Initial efforts in anti-HIV microbicide development have met with poor success as the strategies have relied on a non-specific mechanism of action. Here, we report the development of a microbicide aimed at specifically blocking HIV entry by displaying molecular components of the HIV/host cell attachment complex on the surface of Caulobacter crescentus, a harmless aquatic bacterium. This bacterium can be readily manipulated to present heterologous proteins at high density on its surface by genetic insertion into its crystalline surface layer protein. In separate constructions, we generated bacteria displaying domain 1 of CD4 and MIP1alpha. Each moiety reacted with specific antibodies by Western immunoblot and immuno-fluorescence microscopy. Microbicide functionality was assessed using an HIV pseudotype virus assay system representing Clade B subtypes. Bacteria displaying MIP1alpha reduced infectivity by 35-78% depending on the specific subtype while CD4 display reduced infection by as much as 56%. Combinations of both constructs reduced infectivity by nearly 98%. We demonstrated that HIV infection could be inhibited using a strategy aimed at HIV-specific molecular interactions with Caulobacter surface protein display, and that sufficient protein folding and conformation could be mimicked to bind and block entry. Further, this is the first demonstration that Caulobacter surface protein display may be a useful approach to preventing HIV infection or other viruses as a microbicide. We propose that this harmless bacterium, which is inexpensive to produce and formulate, might be suitable for topical applications as a viable alternative in the search for effective microbicides to counteract the world wide incidence of HIV infection.

S-layer-mediated display of the immunoglobulin G-binding domain of streptococcal protein G on the surface of Caulobacter crescentus: development of an immunoactive reagent.

The immunoglobulin G (IgG)-binding streptococcal protein G is often used for immunoprecipitation or immunoadsorption-based assays, as it exhibits binding to a broader spectrum of host species IgG and IgG subclasses than the alternative, Staphylococcus aureus protein A. Caulobacter crescentus produces a hexagonally arranged paracrystalline protein surface layer (S-layer) composed of a single secreted protein, RsaA, that is notably tolerant of heterologous peptide insertions while maintaining the surface-attached crystalline character. Here, a protein G IgG-binding domain, GB1, was expressed as an insertion into full-length RsaA on the cell surface to produce densely packed immunoreactive particles. GB1 insertions at five separate sites were expressed, and all bound rabbit and goat IgG, but expression levels were reduced compared to those of wild-type RsaA and poor binding to mouse IgG was noted. To remedy this, we used the 20-amino-acid Muc1 peptide derived from human mucins as a spacer, since insertions of multiple tandem repeats were well tolerated for RsaA secretion and assembly. This strategy worked remarkably well, and recombinant RsaA proteins, containing up to three GB1 domains, surrounded by Muc1 peptides, not only were secreted and assembled but did so at wild-type levels. The ability to bind IgG (including mouse IgG) increased as GB1 units were added, and those with three GB1 domains bound twice as much rabbit IgG per cell as S. aureus cells (Pansorbin). The ability of recombinant protein G-Caulobacter cells to function as immunoactive reagents was assessed in an immunoprecipitation assay using a FLAG-tagged protein and anti-FLAG mouse monoclonal antibody; their performance was comparable to that of protein G-Sepharose beads. This work demonstrates the potential for using cells expressing recombinant RsaA/GB1 in immunoassays, especially considering that protein G-Caulobacter cells are more cost-effective than protein G beads and exhibit a broader species and IgG isotype binding range than protein A.

S-layer anchoring and localization of an S-layer-associated protease in Caulobacter crescentus.

The S-layer of the gram-negative bacterium Caulobacter crescentus is composed of a single protein, RsaA, that is secreted and assembled into a hexagonal crystalline array that covers the organism. Despite the widespread occurrence of comparable bacterial S-layers, little is known about S-layer attachment to cell surfaces, especially for gram-negative organisms. Having preliminary indications that the N terminus of RsaA anchors the monomer to the cell surface, we developed an assay to distinguish direct surface attachment from subunit-subunit interactions where small RsaA fragments are incubated with S-layer-negative cells to assess the ability of the fragments to reattach. In doing so, we found that the RsaA anchoring region lies in the first approximately 225 amino acids and that this RsaA anchoring region requires a smooth lipopolysaccharide species found in the outer membrane. By making mutations at six semirandom sites, we learned that relatively minor perturbations within the first approximately 225 amino acids of RsaA caused loss of anchoring. In other studies, we confirmed that only this N-terminal region has a direct role in S-layer anchoring. As a by-product of the anchoring studies, we discovered that Sap, the C. crescentus S-layer-associated protease, recognized a cleavage site in the truncated RsaA fragments that is not detected by Sap in full-length RsaA. This, in turn, led to the discovery that Sap was an extracellular membrane-bound protease, rather than intracellular, as previously proposed. Moreover, Sap was secreted to the cell surface primarily by the S-layer type I secretion apparatus.

Evaluation of a new system for developing particulate enzymes based on the surface (S)-layer protein (RsaA) of Caulobacter crescentus: fusion with the beta-1,4-glycanase (Cex) from the cellulolytic bacterium Cellulomonas fimi yields a robust, catalytically active product.

Immobilized biocatalysts, including particulate enzymes, represent an attractive tool for research and industrial applications because they combine the specificity of native enzymes with the advantage that they can be readily separated from end product and reused. We demonstrated the use of the Caulobacter crescentus surface (S)-layer protein (RsaA) secretion apparatus for the generation of particulate enzymes. Specifically, a candidate protein made previously by fusion of the beta-1,4-glycanase (Cex) from the cellulolytic bacterium Cellulomonas fimi with the C-terminus of RsaA was evaluated. Cex/RsaA cleaved the glycosidic linkage in the artificial substrate p-nitrophenyl-beta-D-cellobioside with a KM similar to that of native Cex (1.1 mM for Cex/RsaA vs 0.60 mM for Cex), indicating that the particulate Cex enzyme was able to bind substrate with wild-type affinity. By contrast, the kcat value was significantly reduced (0.08 s-1 for Cex/RsaA vs 15.8 s-1 for Cex), likely owing to the fact that the RsaA C-terminus induces spontaneous unstructured aggregation of the recombinant protein. Here, we demonstrated that not only can an RsaA fusion protein be cheaply produced and purified to a high yield (76 mg/L of dry wt for Cex/RsaA), but it can also be efficiently recycled. The Caulobacter S-layer secretion system therefore offers an attractive new model system for the production of particulate biocatalysts.

Transcriptional regulation of the S-layer protein type I secretion system in Caulobacter crescentus.

The Gram-negative Caulobacter crescentus exports RsaA, the crystalline S-layer subunit protein using a dedicated type I secretion system. The protein and two transporter genes (rsaADE) are located together, comparable to the Escherichia coli type I hemolysin hlyCABD operon, where read through of a stem loop following hlyCA results in reduced transcription of the hlyBD. Using two genetic approaches and a direct assessment of transcription from regions 5' to the genes we learned that rsaD and rsaE were transcribed together as a separate transcript from rsaA. These results are contrary to previous assumptions about the rsaADE type I secretion gene control and add another theme to the area of type I secretion transcription regulation. It may be that to accommodate the high levels of RsaA secretion, the type I transporters must be transcribed independently from rsaA.

Two outer membrane proteins are required for maximal type I secretion of the Caulobacter crescentus S-layer protein.

Transport of RsaA, the crystalline S-layer subunit protein of Caulobacter crescentus, is mediated by a type I secretion mechanism. Two proteins have been identified that play the role of the outer membrane protein (OMP) component in the RsaA secretion machinery. The genes rsaF(a) and rsaF(b) were identified by similarity to the Escherichia coli hemolysin secretion OMP TolC by using the C. crescentus genome sequence. The rsaF(a) gene is located several kilobases downstream of the other transporter genes, while rsaF(b) is completely unlinked. An rsaF(a) knockout had approximately 56% secretion compared to wild-type levels, while the rsaF(b) knockout reduced secretion levels to approximately 79%. When expression of both proteins was eliminated, there was no RsaA secretion, but a residual level of approximately 9% remained inside the cell, suggesting posttranslational autoregulation. Complementation with either of the individual rsaF genes by use of a multicopy vector, which resulted in 8- to 10-fold overexpression of the proteins, did not restore RsaA secretion to wild-type levels, indicating that both rsaF genes were required for full-level secretion. However, overexpression of rsaF(a) (with normal rsaF(b) levels) in concert with overexpression of rsaA resulted in a 28% increase in RsaA secretion, indicating a potential for significantly increasing expression levels of an already highly expressing type I secretion system. This is the only known example of type I secretion requiring two OMPs to assemble a fully functional system.

Comparison of S-layer secretion genes in freshwater caulobacters.

Our freshwater caulobacter collection contains about 40 strains that are morphologically similar to Caulobacter crescentus. All elaborate a crystalline protein surface (S) layer made up of protein monomers 100-193 kDa in size. We conducted a comparative study of S-layer secretion in 6 strains representing 3 size groups of S-layer proteins: small (100-108 kDa), medium (122-151 kDa), and large (181-193 kDa). All contained genes predicted to encode ATP-binding cassette transporters and membrane fusion proteins highly similar to those of C. crescentus, indicating that the S-layer proteins were all secreted by a type I system. The S-layer proteins' C-termini showed unexpectedly low sequence similarity but contained conserved residues and predicted secondary structure features typical of type I secretion signals. Cross-expression studies showed that the 6 strains recognized secretion signals from C. crescentus and Pseudomonas aeruginosa and similarly that C. crescentus was able to secrete the S-layer protein C-terminus of 1 strain examined. Inactivation of the ATP-binding cassette transporter abolished S-layer protein secretion, indicating that the type I transporter is necessary for S-layer protein secretion. Finally, while all of the S-layer proteins of this subset of strains were secreted by type I mechanisms, there were significant differences in genome positions of the transporter genes that correlated with S-layer protein size.

Caulobacter crescentus synthesizes an S-layer-editing metalloprotease possessing a domain sharing sequence similarity with its paracrystalline S-layer protein.

Strains of Caulobacter crescentus elaborate an S-layer, a two-dimensional protein latticework which covers the cell surface. The S-layer protein (RsaA) is secreted by a type I mechanism (relying on a C-terminal signal) and is unusual among type I secreted proteins because high levels of protein are produced continuously. In efforts to adapt the S-layer for display of foreign peptides and proteins, we noted a proteolytic activity that affected S-layer monomers with foreign inserts. The cleavage was precise, resulting in fragments with an unambiguous N-terminal sequence. We developed an assay to screen for loss of this activity (i.e., presentation of foreign peptides without degradation), using transposon and traditional mutagenesis. A metalloprotease gene designated sap (S-layer-associated protease) was identified which could complement the protease-negative mutants. The N-terminal half of Sap possessed significant similarity to other type I secreted proteases (e.g., alkaline protease of Pseudomonas aeruginosa), including the characteristic RTX repeat sequences, but the C-terminal half which normally includes the type I secretion signal exhibited no such similarity. Instead, there was a region of significant similarity to the N-terminal region of RsaA. We hypothesize that Sap evolved by combining the catalytic portion of a type I secreted protease with an S-layer-like protein, perhaps to associate with nascent S-layer monomers to "scan" for modifications.