Spatial regulation of protein A in Staphylococcus aureus

Summary

Surface proteins of Staphylococcus aureus play vital roles in bacterial physiology and pathogenesis. Recent work suggests that surface proteins are spatially regulated by a YSIRK/GXXS signal peptide that promotes cross-wall targeting at the mid-cell, though the mechanisms remain unclear. We previously showed that protein A (SpA), a YSIRK/GXXS protein and key staphylococcal virulence factor, mis-localizes in a ltaS mutant deficient in lipoteichoic acid (LTA) production. Here, we identified that SpA contains another cross-wall targeting signal, the LysM domain, which, in addition to the YSIRK/GXXS signal peptide, significantly enhances SpA cross-wall targeting. We show that LTA synthesis, but not LtaS, is required for SpA septal anchoring and cross-wall deposition. Interestingly, LTA is predominantly found at the peripheral cell membrane and is diminished at the septum of dividing staphylococcal cells, suggesting a restriction mechanism for SpA septal localization. Finally, we show that D-alanylation of LTA abolishes SpA cross-wall deposition by disrupting SpA distribution in the peptidoglycan layer without altering SpA septal anchoring. Our study reveals that multiple factors contribute to the spatial regulation and cross-wall targeting of SpA via different mechanisms, which coordinately ensures efficient incorporation of surface proteins into the growing peptidoglycan during the cell cycle.

Introduction

Staphylococcus aureus is a Gram-positive bacterium that frequently colonizes human nares and skin, and it is a leading cause of both hospital-acquired and community-acquired infections (von Eiff et al., 2001, Tong et al., 2015). The cell envelope of S. aureus plays essential roles in bacterial colonization and disease development. Like other Gram-positive bacteria, the cell envelope of S. aureus consists of a thick cell wall peptidoglycan layer, teichoic acids including wall teichoic acid (WTA) and lipoteichoic acid (LTA), polysaccharides and a number of surface proteins (Rajagopal & Walker, 2017).

The cell wall anchored surface proteins are key components of the staphylococcal cell envelope. Many of them fulfill virulence functions such as adhesion, biofilm formation, nutrient acquisition, antibiotic resistance, and immune evasion (Foster et al., 2014, Schneewind & Missiakas, 2019). Surface protein precursors contain an N-terminal signal peptide and a C-terminal cell wall sorting signal. The N-terminal signal peptide mediates secretion of precursors across the cytoplasmic membrane (Yu et al., 2018). Once translocated across the membrane, the signal peptide is cleaved off by the signal peptidase SpsB (Bae & Schneewind, 2003, Yu et al., 2018). The resulting surface protein precursors are covalently attached to the peptidoglycan precursors, the lipid II molecules, via sortase A (SrtA)-mediated cleavage of the C-terminal sorting signals (Mazmanian et al., 1999). The lipid II-linked surface protein precursors are then incorporated into the mature peptidoglycan layer during cell wall biosynthesis (Perry et al., 2002).

Strikingly, many surface protein precursors contain a specific N-terminal signal peptide with a highly conserved YSIRK/GXXS motif (Rosenstein & Götz, 2000, Tettelin et al., 2001). The YSIRK/GXXS signal peptide promotes protein trafficking to the division septum and subsequent anchoring to the cross-wall peptidoglycan (Carlsson et al., 2006, DeDent et al., 2008, Raz et al., 2012, Yu & Götz, 2012, Yu et al., 2018). Upon cell separation, cross-wall anchored surface proteins are displayed on the surface of the newly synthesized hemisphere of the daughter cells (Cole & Hahn, 1962, DeDent et al., 2008, Raz et al., 2012, Yu et al., 2018). Surface proteins eventually cover the entire cell surface after several rounds of cell division. The mechanisms of how YSIRK/GXXS signal peptides promote cross-wall targeting are largely unknown. Staphylococcal protein A (SpA) is well-known virulence factor that binds to host immunoglobulin and disrupts the host immune response (Forsgren & Sjöquist, 1966). We previously showed that SpA precursor containing the YSIRK/GXXS signal peptide engages the SecA-mediated secretion pathway (Yu et al., 2018). SecA does not determine SpA septal localization as SecA is found at both the septal and peripheral cell membranes. In contrast, the cross-wall localization of SpA is abolished in a ltaS depletion mutant deficient in synthesizing LTA (Yu et al., 2018). Thus, LtaS-mediated LTA synthesis provides a heretofore unknown mechanism that enriches the YSIRK/GXXS protein precursors at the septum.

LTA is another key component of staphylococcal cell envelope. It is a glycerol-phosphate polymer tethered to the cell membrane via a diglucosyl-diacylglycerol (Glc₂-DAG) glycolipid anchor (Percy & Grundling, 2014, Schneewind & Missiakas, 2014). The glycolipid anchor is synthesized in the cytosol by YpfP and translocated across the membrane by LtaA (Gründling & Schneewind, 2007a). At the outer leaflet of the membrane, LtaS synthesizes LTA by repeatedly transferring glycerol phosphate moieties from phosphatidylglycerol (PG) to Glc₂-DAG (Gründling & Schneewind, 2007b). The LtaS protein consists of N-terminal transmembrane domains and an extracellular enzymatic domain. Interestingly, the extracellular enzymatic domain of LtaS is cleaved by the signal peptidase SpsB, resulting in inactivation of the enzyme (Wormann et al., 2011). LtaS-mediated LTA synthesis is essential for bacterial viability. In S. aureus, ltaS depletion leads to enlarged cell size and severe growth and cell division defects (Gründling & Schneewind, 2007b). Deletion of ltaS is not possible unless the mutant is maintained in osmotically stabilizing medium or acquires additional suppressor mutations (Corrigan et al., 2011, Baek et al., 2016, Karinou et al., 2019). Earlier work revealed that LtaS localizes at the septum whereas YpfP and LtaA are found throughout the entire cell (Reichmann et al., 2014). A model is proposed, in which LTA is synthesized at the septum and distributed to the cell envelope during cell division and separation (Reichmann et al., 2014).

After backbone synthesis, LTA is heavily decorated with D-alanylation and glycosylation. D-alanylation is carried out by the DltABCD system (Neuhaus et al., 1996, Neuhaus & Baddiley, 2003) and glycosylation is catalyzed by the CsbB-GtcA-YfhO three-component system (Kho & Meredith, 2018). Both D-alanylation and glycosylation occur at the hydroxyl groups at position C2 of LTA glycerol phosphate backbone. Under standard laboratory culture conditions, the majority (80-90%) of the LTA repeat units are decorated with D-alanines while only up to 2% of the repeat units are decorated with glycosyls in S. aureus (Kho & Meredith, 2018). LTA and LTA modifications have been shown to impact the localization and production of surface proteins. For instance, LTA is known to modulate autolysin localization (Zoll et al., 2012), bind to surface protein Sbi in S. aureus (Smith et al., 2012), interact with the GW modules of Internalin B in Listeria monocytogenes (Jonquieres et al., 1999) and affect surface protein production in Lactobacillus acidophilus and Streptococcus gordonii (Selle et al., 2017, Lima et al., 2019). The addition of D-alanyl esters adds positive charges to the otherwise negatively charged LTA polymer, a modification that has been implicated in modulating protein secretion and folding. For example, the dlt mutation suppresses the secretion deficiency of the prsA mutation that encodes for a secretion chaperone in Bacillus subtilis (Hyyrylainen et al., 2000). In other cases, the dlt mutant displays a thickened cell wall that traps secreted proteins (Nouaille et al., 2004). Nevertheless, most of the literature studies non-covalently bound surface proteins; little is known about the interactions between LTA, LTA modifications and covalently anchored surface proteins.

When analyzing SpA sequence, we noticed that SpA contains a cell wall binding motif, the LysM (Lysin Motif) domain (Buist et al., 2008) (Fig.1A). The LysM domain has been shown to bind to peptidoglycan at specific sites, which is regulated by teichoic acids (Buist et al., 2008). For instance, the LysM domain of AcmA, the major autolysin of Lactococcus lactis, binds to the mid-cell where LTA is absent (Steen et al., 2003). In S. aureus, excreted cytoplasmic proteins fused with LysM domain can be targeted to the cross-wall (Ebner et al., 2015); LysM domain promotes the cross-wall targeting of murein hydrolases LytN and Sle1, and WTA prevents its binding to the peripheral wall (Frankel & Schneewind, 2012). Hence, teichoic acids appear to regulate LysM-mediated cross-wall targeting.

Figure 1.

LysM domain enhances SpA cross-wall targeting. (A) S. aureus Newman SpA domain organization (UniProt: A0A0H3K686). SP, signal peptide; IgBDs, immunoglobulin-binding domains; Xr, region X; CWS, cell wall sorting signal; numbers indicate amino acid residues. (B) Surface deposition of the newly synthesized SpA and SpA_ΔLysM. Wild-type spa and spa_ΔlysM are expressed by ATc-inducible P_tet promoter from plasmid pCLitet in S. aureus SEJ1 (RN4220Δspa). Staphylococcal cells were trypsin-digested to remove the pre-deposited SpA, and refreshed in medium containing trypsin inhibitor for 20 min to allow SpA regeneration. Cells were then fixed, and immunostained with SpA-specific antiserum and Alexa Fluor 488-conjugated secondary antibody (green). Bacterial cell membrane was counterstained with Nile red (red). Vector: pCLitet empty vector. Representative images are from three independent experiments. (C) Quantification of SpA cross-wall localization from the images in panel B. The percentage is calculated by the numbers of cells with cross-wall SpA localization divided by the total cell numbers. At least three random images per sample per experiment were taken and quantification is based on three independent experiments. Unpaired, two-tailed t-test was used for statistical analysis; n.s. no statistical difference. (D) Representative histogram showing SpA fluorescence intensity along a line perpendicular to the cross-wall of a cell in panel B. The peak represents cross-wall signal intensity (C) and the two shoulders represent peripheral wall signals (P). (E) Violin plot of relative SpA cross-wall fluorescence intensity. The value is the ratio of cross-wall signal intensity versus the average of peripheral wall signal intensity of the same cell. Cells analyzed are from three independent experiments: SpA, n=155; SpA_ΔLysM, n=148. Unpaired, two-tailed t-test was used to analyze statistical difference; ****, p<0.0001. (F) Plasmids pCLitet-spa, pCLitet-spa_ΔlysM, and pCLitet were transformed to S. aureus RN4220ΔspaΔsbi; spa and spa_ΔlysM were expressed with ATc induction. S. aureus cultures were fractionated into cytoplasm (C), membrane (M), cell wall (W) and culture supernatant (S) and analyzed by SpA immunoblotting. Sortase A (SrtA, MW 23.5 kD) immunoblotting was used as fractionation and loading control. Numbers on the left indicate protein ladders in kDa.

In this study, we investigated the roles of the LysM domain, LTA and LTA modifications in SpA cross-wall targeting. We found that LysM, LTA, and LTA D-alanylation all contribute to SpA spatial regulation, but functions at different stages of SpA biogenesis. These factors coordinately promote efficient incorporation of SpA into the growing peptidoglycan during the cell cycle.

Results

The LysM domain enhances SpA cross-wall targeting, but is dispensable for SpA mis-localization in the ltaS mutant

The full-length SpA contains an N-terminal YSIRK/GXXS signal peptide, five IgG binding domains (IgBDs), repeats region X (Xr), LysM domain and C-terminal cell wall sorting signal (Fig. 1A). To examine the role of the LysM domain in SpA cross-wall targeting, we generated a SpA LysM domain deletion variant by deleting amino acid residues 413-457 of SpA. SpA and SpA_ΔLysM were expressed under ATc inducible promoter in plasmid pCLitet. The recombinant plasmids were chromosomally integrated in a S. aureus Δspa mutant strain SEJ1 (RN4220Δspa). The cross-wall localization of SpA was examined by de novo SpA surface localization immunofluorescence microscopy. Staphylococcal cells were treated with trypsin to remove the existing SpA on the bacterial surface; trypsin-treated cells were grown in fresh media for 20 min to allow SpA regeneration and deposition on the cell surface; the 20 min samples were fixed and subjected to immunofluorescence microscopy using SpA-specific antiserum. Imaging results revealed that both SpA and SpA_ΔLysM localized at the cross-wall of the newly divided cells (Fig. 1B). The pCLitet empty vector control and samples without ATc induction showed little immunofluorescence signals, indicating that the fluorescence signals are SpA-specific and the expression of SpA requires ATc induction. Quantification of the percentage of SpA cross-wall localization exhibited no statistical difference between SpA and SpA_ΔLysM (Fig. 1C), implying that the LysM domain is not required for SpA cross-wall localization. However, the SpA fluorescence signals at the cross-wall were strongly reduced in SpA_ΔLysM (Fig. 1B). To quantify the level of reduction, a line was drawn perpendicular to the cross-wall at the middle of a cell and a fluorescence intensity histogram was generated along the line using Image J. Fig. 1D displays representative histograms of SpA and SpA_ΔLysM, where the cross-wall peak (“C”) was clearly higher in SpA than SpA_ΔLysM. The peripheral wall peaks (“P”) were similar between SpA and SpA_ΔLysM. The ratio of cross-wall peak (“C”) versus the peripheral wall peak (“P”) of SpA was about 1.5-fold higher than that of SpA_ΔLysM, demonstrating that LysM significantly enhanced SpA cross-wall targeting (Fig. 1E). To examine whether the LysM domain contributes to SpA secretion and anchoring, we performed SpA immunoblotting. However, the presence of Sbi, a non-specific IgG-binding protein in S. aureus interfered with SpA_ΔLysM detection (data not shown). To overcome this issue, plasmids expressing SpA and SpA_ΔLysM were transformed to WY110 (RN4440ΔspaΔsbi). Staphylococcal cultures were fractionated to cytosolic (C), membrane (M), cell wall (W) and supernatant (S) and analyzed by SpA immunoblotting (Fig. 1F). Both SpA and SpA_ΔLysM were mostly found in the cell wall fractions and some proteins were released into the supernatant. There was no visible accumulation of SpA precursors in the cytosolic or membrane fractions, indicating that the secretion and anchoring of SpA were not affected by LysM domain. We concluded that the LysM domain enhances SpA cross-wall targeting independent of secretion and anchoring.

To examine whether the LysM domain is involved in SpA mis-localization in the ltaS mutant, pCLitet-spa and pCLitet-spa_ΔLysM were transformed to SEJ1iltaS (RN4220Δspa P_spac-ltaS). Under ltaS inducing conditions (+IPTG), SpA and SpA_ΔLysM showed clear cross-wall localization. Upon ltaS depletion (−IPTG), both SpA and SpA_ΔLysM mis-localized to the peripheral sites (Fig. S1A). The level of reduction in cross-wall localization was similar between SpA and SpA_ΔLysM (Fig. S1B). Similar results were obtained in a ΔltaS suppressor mutant (Fig. 2). Taken together, these results demonstrate that the LysM domain enhances SpA cross-wall targeting, but is not absolutely required for SpA cross-wall localization. Moreover, the LysM domain is dispensable for SpA mis-localization upon ltaS depletion.

Figure 2.

Septal anchoring and cross-wall deposition of SpA and SpA_ΔLysM are abolished in ANG1786 that lacks ltaS but has normal cell morphology. (A) Surface deposition of the newly synthesized SpA and SpA_ΔLysM in ANG1786 and its parental strain SEJ1; spa and spa_ΔlysM are expressed with ATc induction. Vector, pCLitet empty vector. Representative images are from three independent experiments. (B) Quantification of SpA cross-wall localization from the images in panel A. Unpaired, two-tailed t-test was used for statistical analysis; ****, p<0.0001. (C) SpA localization in protoplasts. Staphylococcal cells were trypsin-digested to remove pre-deposited SpA and fixed. Fixed cells were digested with or without 20 μg/ml lysostaphin for 2 min on slide. The resulting protoplasts were fixed and immunostained with SpA-specific antiserum and Alexa Fluor 647-conjugated secondary antibody (red). BF, bright-field images; BODIPY® fluorescent vancomycin stains the cell wall (Van-FL, green); Hoechst 33342 stains DNA (blue). Representative images are from three independent experiments. Wild-type spa and spa_ΔlysM were expressed with ATc induction in SEJ1, ANG1786, and SEJ1ΔsrtA. (D) SpA immunoblotting of culture supernatant (S) and lysostaphin-digested bacterial pellet (P) samples. Sortase A (αSrtA) blot serves as a loading control. Asterisk denotes non-specific Sbi bands. Numbers on the left indicate protein ladders in kDa.

SpA mis-localization in ΔltaS is not due to its cell division defect

LtaS-mediated LTA synthesis has pleotropic functions such as regulating cell division and separation. The ltaS depletion strain exhibits strong cell morphological alterations and cell division defects (Gründling & Schneewind, 2007b) (Fig. S1A). As SpA cross-wall localization is closely related to cell division, we asked whether SpA mis-localization in the ltaS mutant is due to its cell division defect. A ΔltaS suppressor strain ANG1786 (RN4220ΔspaΔltaS 4S5) has been characterized earlier (Corrigan et al., 2011). In this strain, a point mutation in the gdpP gene rescues the lethality and restores cell morphology of ΔltaS. Lacking ltaS while having relatively normal cell morphology, ANG1786 provides a way to distinguish whether SpA mis-localization is due to the lack of ltaS or due to its cell division defect. ANG1786 and its parental strain SEJ1 were transformed with pCLitet-spa and pCLitet-spa_ΔLysM, and grown with ATc to induce the expression of spa and spa_ΔLysM. Fluorescence microscopy of de novo SpA surface localization revealed that both SpA and SpA_ΔLysM mis-localized to the peripheral sites in ANG1786 (Fig. 2AB), which is consistent with ltaS depletion results, demonstrating that SpA mis-localization in the ltaS mutant is not caused by its cell division defect but is indeed due to the loss of LtaS.

LTA synthesis is required for SpA septal localization and cross-wall deposition

To further reveal the localization of SpA underneath the thick peptidoglycan layer, we used a protoplast-based immunofluorescence microscopy method that we developed earlier (Yu et al., 2018). In this experiment, after trypsinization, staphylococcal cells were digested with the murein hydrolase lysostaphin; the resulting protoplasts were fixed, permeabilized and examined with SpA immunofluorescence microscopy. In the control samples with trypsin but without lysostaphin digestion, no SpA signals could be detected and the cell wall could be clearly visualized by Van-FL staining (Fig. 2C, +ATC, +Trypsin, −Lysostaphin). In contrast, lysostaphin digestion removed most of the peptidoglycan and separated the two dividing daughter cells, which allows the detection of SpA at the division septum underneath the peptidoglycan layer (Fig. 2C, +ATC, +Trypsin, +Lysostaphin). The imaging results revealed that SpA and SpA_ΔLysM localized at the septum in SEJ1, but mis-localized all over the cytoplasmic membrane in ANG1786 (Fig. 2C). Little SpA signals could be detected in the control of ΔsrtA, suggesting that the SpA detected here are anchored SpA (Fig. 2C). Thus, ΔltaS not only alters SpA cross-wall deposition but also disrupts SpA septal anchoring. Immunoblotting analysis of staphylococcal cultures separated to culture supernatant (S) and cell pellet (P) demonstrated that the abundance of cell wall anchored SpA and SpA_ΔLysM was reduced in ANG1786 (Fig. 2D). This is consistent with our previous ltaS depletion results, supporting the hypothesis that cross-wall targeting promotes efficient incorporation of surface proteins to the cell envelope (Carlsson et al., 2006, DeDent et al., 2008, Raz et al., 2012, Yu et al., 2018).

The enzymatic activity of LtaS is required for SpA septal localization

Next, we asked whether SpA cross-wall localization is dependent on LtaS protein or its product LTA. To address this, ANG1786 containing pCLitet-spa (stain WY77) was complemented with either the wild-type ltaS or a catalytically inactive variant ltaS_T300A. The LtaS_T300A variant has been shown to abolish LtaS enzymatic activity while maintaining stable protein production (Lu et al., 2009). Both alleles were expressed from IPTG-inducible promoter in plasmid pJK4. Protoplast-based microscopic analysis indicated that wild-type ltaS, but not ltaS_T300A, restored SpA septal localization (Fig. 3A). Consistent with earlier studies, the LtaS_T300A mutation abolished LTA production but produced stable LtaS protein (Fig. 3BC) (Lu et al., 2009). These data suggest that the enzymatic activity of LtaS is required for SpA septal targeting.

Figure 3.

The enzymatic activity of LtaS is required for SpA septal localization. (A) Strain WY477 (ANG1786 pCLitet-spa) was complemented with wild-type ltaS or ltaS_T300A expressed from IPTG-inducible promoter in pJK4 plasmid; spa is expressed from ATc-inducible promoter in plasmid pCLitet. Bacterial cultures were grown with 1 mM IPTG to induce plasmid-borne ltaS alleles and with 200 ng/ml ATc to induce spa expression. Fluorescence images showing SpA localization in protoplasts. Staphylococcal cells were trypsin-digested to remove pre-deposited SpA and fixed. Fixed cells were digested with or without 20 μg/ml lysostaphin for 2 min on slide. The resulting protoplasts were fixed and immunostained with SpA-specific antiserum and Alexa Fluor 647-conjugated secondary antibody (red). Van-FL stains the cell wall (Van-FL, green); Hoechst 33342 stains DNA (blue). Vector, pJK4 empty vector. Representative images are from three independent experiments. The same bacterial culture used for panel A was examined with LTA and LtaS immunoblotting (panel B and C). (B) LTA immunoblotting. Staphylococcal membranes were collected and LTA was detected with LTA-specific monoclonal antibody mAb 55 (αLTA). The protein ladders (in kDa) are indicated on the left. (C) LtaS immunoblotting. The cell lysates (lysostaphin-digested cell pellet) were immunoblotted with LtaS-specific antiserum (αLtaS, the MW of full-length LtaS is 74.4 kD) or SrtA antiserum (αSrtA).

LTA predominantly localizes at the peripheral membrane in dividing staphylococcal cells

The above results suggest that the production of LTA, but not LtaS, is required for SpA localization. Previous microscopic studies revealed that LTA is not detectable from the cell surface, but rather occupies the periplasmic space between the cell wall and cytoplasmic membrane (Aasjord & Grov, 1980, Matias & Beveridge, 2008, Reichmann & Grundling, 2011, Reichmann et al., 2014, Shiraishi et al., 2018). However, detailed subcellular localization of LTA has not been elucidated. To examine the localization of LTA, we used the same protoplast-based immunofluorescence microscopy method described above. Staphylococcal strain SEJ1 and LTA-negative strain (ANG1786) were grown to mid-log phase, fixed with paraformaldehyde, digested with lysostaphin, fixed again and stained with LTA-specific monoclonal antibody. The cell membrane was stained with Nile red (Fig. 4A). In the majority of lysostaphin-treated protoplasts (75% of the cell population), LTA was predominantly found at the peripheral cell membrane and absent at the septum (Fig. 4AB, cell type 1). 10% of the protoplasts exhibited weak or incomplete septal signals (Fig. 4AB, cell type 2), and 15% of the protoplasts localized LTA over the entire cell membrane (Fig. 4AB, cell type 3). The cells with peripheral LTA localization appeared to be actively dividing cells, in which the two daughter cells were still connected and maintained flat septal membrane before splitting (Fig. 4A, yellow arrows). In contrast, the cells with LTA covering the entire membrane appeared to be separated individual cells (Fig. 4A, yellow arrowhead). Consistent with previous study (Reichmann et al., 2014), no LTA signals could be detected on the bacterial cell surface without lysostaphin treatment, indicating that LTA localizes close to the cell membrane underneath the peptidoglycan layer (Fig. 4A). The immunofluorescence signals were LTA-specific as no signals could be detected in ANG1786. Together, the data suggest a cell cycle-dependent LTA maturation process: the old peripheral sites are occupied by mature LTA; the synthesis of LTA occurs at the septum; the concentration or the length of LTA gradually increases at the septum during cell division; the growing LTA eventually covers septal membrane when the daughter cells split. The opposite localization pattern between LTA and SpA suggests that in the active dividing cells, peripheral localized LTA restricts SpA to be localized at the septum.

Figure 4.

Cell-cycle dependent LTA localization in S. aureus. (A) LTA subcellular localization in protoplasts. Cells of SEJ1(RN4220Δspa) and its isogenic ltaS deletion strain ANG1786 were fixed and digested with or without 20 μg/ml lysostaphin for 2 min on slide. The protoplasts were fixed and immunostained with LTA-specific monoclonal antibody followed by Alexa Fluor 488-conjugated secondary antibody (green). Nile red stains the cell membrane (red). Numbers denote different LTA localization patterns: 1. LTA localizes at the peripheral membrane of dividing cells; 2. LTA localizes at both peripheral and septal membrane of dividing cells; 3. LTA localizes all over the membrane in cells that have been separated. Representative images are from three independent experiments. (B) Quantification of different LTA localization patterns from panel A. Images from three independent experiments were quantified. One-way ANOVA with Tukey’s multiple comparisons test was used for statistical analysis; ****, p<0.0001.

Lack of LTA D-alanylation abolishes SpA cross-wall deposition without affecting SpA septal anchoring

We further asked whether LTA D-alanylation or glycosylation affects SpA cross-wall localization. When analyzed by SpA immunofluorescence microscopy, the transposon mutant of yfhO deficient in glycosylation did not exhibit difference compared to the wild-type (data not shown). A dltC mutant was generated by transducing dltC::kan to SEJ1 containing pCLitet-spa (strain WY478). Immunofluorescence microscopy of de novo SpA surface localization revealed that the cross-wall localization of SpA was significantly diminished in the dltC mutant (Fig. 5AB). The phenotype can be complemented by plasmid-borne expression of dltABCD operon (Fig. 5AB). SpA immunoblotting of the cell pellet fraction demonstrated that SpA cell wall anchoring was not affected in ΔdltC (Fig. 5C). However, the abundance of SpA was increased in the culture supernatant of ΔdltC (Fig. 5C). Interestingly, the abundance of another IgG-binding protein, Sbi, was reduced in the cell pellet of ΔdltC (Fig. 5C). To our surprise, the sortase A (αSrtA) blot, which we routinely use as the loading control, showed increased protein level in the supernatant (Fig. 5C). The same was observed for another loading control, the house-keeping sigma factor A blot. These observations invite the question of whether ΔdltC releases more proteins to the supernatant in general. This was indeed the case as revealed by Coomassie blue-stained SDS-PAGE (Fig. 5D). Together, the data indicate that the dltC mutation abolished SpA cross-wall deposition and generated cell envelope defects such as increased cell lysis and release of cellular proteins to the culture supernatant.

Figure 5.

dltC mutation abolishes SpA cross-wall deposition. (A) Surface deposition of the newly synthesized SpA in SEJ1, SEJ1ΔdltC and the complementation strain. Cultures were grown with ATc to induce spa expression and with IPTG to induce the expression of dltABCD in the complementation plasmid. SEJ1 vector (pCLitet empty vector) is a negative control of SpA specificity. Representative images are from three independent experiments. (B) Quantification of SpA cross-wall localization from panel A. Ordinary one-way ANOVA with Tukey’s multiple comparisons test was used for statistical analysis; ****, p<0.0001; n.s., no statistical difference. (C) SpA, SrtA and SigA (sigma factor A) immunoblotting of lysostaphin-digested bacterial cell pellet (P) and culture supernatant (S). Asterisk denotes non-specific Sbi bands. Numbers on the left indicate protein ladders in kDa. (D) Coomassie blue stained SDS-PAGE of lysostaphin-digested bacterial cell pellet (P) and culture supernatant (S).

When working with the dltC mutant, we observed heterogenous colonies on the agar plates even after repeated transduction. To further examine if the SpA mis-localization phenotype is specific to LTA D-alanylation, but not to DltC protein per se, and to rule out any potential suppressor mutations, a CRISPR-interference knockdown approach was used. A knockdown plasmid pKK2-dltA containing dCas9-mediated interference with dltA sgRNA was constructed. The expression of dcas9 was ATc-inducible. The knockdown plasmid and the empty vector control were transformed to strain RN4220, WY743 (SEJ1 pCL55-spa) and WY744 (SEJ1 pCL55-spa_ΔlysM). Northern blot analysis using dltD RNA probe demonstrated that 20 ng/ml ATc induction diminished the abundance of dlt operon transcripts in all three strains tested, indicating that the knockdown strategy was effective and the expression of dlt operon could be transcriptionally repressed (Fig. 6A). When analyzed by immunofluorescence microscopy, dlt knockdown diminished SpA cross-wall deposition of both SpA and SpA_ΔLysM (Fig. 6BC). Consistent with the results from ΔdltC, the dlt knockdown strains exhibited decreased Sbi in the cell pellet, increased SpA, SrtA, SigA and overall protein abundance in the supernatant (Fig. 6DE). Thus, not DltC protein per se, but the lack of LTA D-alanylation abolishes SpA cross-wall deposition. Moreover, the mis-localization phenotype is independent of the LysM domain.

Figure 6.

dlt knockdown abolishes SpA cross-wall deposition. (A) Northern blot analysis of dlt transcriptional repression by CRISPR interference. The knockdown plasmid containing ATc-inducible dCAS9 and dltA sgRNA was transformed to strains RN4220 and SEJ1 harboring pCL55-spa or pCL55-spa_ΔlysM, in which spa or spa_ΔlysM are expressed from spa native promoter. Total RNA was extracted and hybridized with for an RNA probe targeting dltD. (B) Surface deposition of the newly synthesized SpA in dlt knockdown strains. Cultures were grown with 20 ng/ml ATc to repress dlt expression. Vector, pKK2 CRISPRi empty vector; dltA, pKK2-dltA; SEJ1 is a negative control of SpA specificity. Representative images are from three independent experiments. (C) Quantification of SpA cross-wall localization from panel B. Unpaired, two-tailed t-test was used for statistical analysis; ****, p<0.0001; ***, p<0.001. (D) Immunoblotting of SpA, SrtA and SigA from lysostaphin-digested bacterial pellet (P) and culture supernatant (S). Asterisk denotes non-specific Sbi bands. Numbers on the left indicate protein ladders in kDa. (E) Coomassie blue stained SDS-PAGE of lysostaphin-digested bacterial pellet (P) and culture supernatant (S).

We finally asked whether LTA D-alanylation affects the septal anchoring of SpA underneath the peptidoglycan layer. Protoplast-based SpA immunofluorescence microscopy revealed similar septal localization patterns between ΔdltC and its parental strain, indicating that LTA D-alanylation does not affect SpA septal anchoring (Fig. 7A). Rather, LTA D-alanylation seems to affect the distribution of SpA across the thick peptidoglycan layer. In line with this hypothesis, the dltC mutant displayed noticeably enlarged cell size and reduced cross-wall formation when analyzed by cell wall Van-FL staining (Fig. 7BCD). The Van-FL fluorescence signals were decreased at the cross-walls and increased at the peripheral walls, suggesting that the peptidoglycan layer is less cross-linked and weakened in the dltC mutant (Fig. 7BE). The abnormal cell wall structure in the dltC mutant likely disrupts SpA cross-wall deposition. Taken together, the above data indicate that LTA D-alanylation is required for proper SpA cross-wall deposition, but is not required for SpA septal anchoring. The disruption of SpA cross-wall deposition in the dlt mutant is likely due to its altered peptidoglycan structure and growth.

Figure 7.

dltC mutation does not affect SpA septal anchoring, but displays enlarged cell size and altered cell wall structure. (A) SpA localization in protoplasts of SEJ1 and ΔdltC. (B) Van-FL staining of SEJ1, SEJ1ΔdltC and the complementation strain. Representative images are from three independent experiments. (C) Cell size quantification with the samples from panel B. Cells were divided into two groups: with or without visible cross-walls, representing dividing or non-dividing cells. (D) Quantification of percentage of visible cross-walls in the cell population from panel B. (E) Quantification of the ratio between cross-wall Van-FL intensity and peripheral wall Van-FL intensity. All quantifications (panel C, D, E) are from three independent experiments; Ordinary one-way ANOVA with Tukey’s multiple comparisons test was used for statistical analysis in panel C, D, E; ****, p<0.0001; ***, p<0.001; **, p<0.01; *, p<0.05.

Discussion

While the biochemical mechanisms of surface protein biogenesis in Gram-positive bacteria have been studied in great detail, little is known about their dynamic spatial and temporal regulation (Dramsi & Bierne, 2017). Here, we show that LysM, LTA, and LTA D-alanylation spatially regulate SpA biogenesis at different stages: LysM enhances SpA cross-wall targeting independent of secretion and anchoring, LTA controls SpA septal anchoring and cross-wall deposition, and LTA D-alanylation regulates SpA cross-wall deposition post septal anchoring.

Previous studies have shown that the YSIRK/GXXS signal peptide promotes SpA cross-wall targeting. The current work explores the role of a second cross-wall-targeting motif of SpA, the LysM domain, which enhances SpA cross-wall targeting, albeit by a different mechanism. Initially identified from bacterial phage lysin (Garvey et al., 1986), the LysM domain typically consists of 44 to 65 amino acid residues and has been found in both prokaryotes and eukaryotes (Buist et al., 2008). In bacteria, the LysM domain is frequently found in murein hydrolases. It has been well documented in the literature that the LysM domain binds to peptidoglycan (Steen et al., 2003, Buist et al., 2008, Visweswaran et al., 2014). In S. aureus, previous work has shown that purified LysM_SA0710-mCherry fusion protein binds to purified peptidoglycan from S. aureus; LysM_SA0710 fused with two excreted cytoplasmic proteins aldolase and enolase targets these proteins to the staphylococcal cross-wall (Ebner et al., 2015). The same results were obtained in another study, whereby purified LysM_LytN-mCherry and LysM_Sle1-mCherry recombinant proteins bind to purified staphylococcal peptidoglycan as well as to the cross-wall of staphylococcal cells (Frankel & Schneewind, 2012). SA0710, LytN and Sle1 are all murein hydrolases. It is proposed that the mature murein hydrolase enzyme binds back to the peptidoglycan via the LysM domain after secretion, and teichoic acids restrict LysM binding to specific sites such as the cross-wall (Steen et al., 2003, Frankel & Schneewind, 2012). Here we show that the LysM domain present in SpA, a cell wall anchored protein, promotes SpA cross-wall targeting. In the case of SpA, we propose that following SpA translocation across the membrane, the LysM domain binds to the nascent peptidoglycan or the lipid II molecule, resulting in enhanced cross-wall targeting. The LysM domain is present in other cell wall anchored YSIRK/GXXS proteins as well. In a recent study, Bai et al analyzed over 10,000 YSIRK/GXXS proteins from different bacterial species and the LysM domain is found in 0.29% of YSIRK/GXXS proteins containing glycosyl hydrolase domains (Bai et al., 2020). Besides the LysM domain, other cell wall binding domains, such as GW, SH3 domains, can also be found in the YSIRK/GXXS family proteins (Pfam: PF04650). Why would a cell wall anchored protein harbor multiple cross-wall targeting motifs? We propose that the dual cross-wall targeting mechanisms ensure maximum incorporation of SpA to the newly synthesized peptidoglycan at the cross-wall. Our current results indicate that LysM is not required for the SpA mis-localization phenotype in the ltaS mutant. However, the results do not exclude the possibility that LTA regulates LysM domain-mediated binding to the cross-wall peptidoglycan. It would be interesting to further investigate this question.

We previously demonstrated that SpA mis-localizes in the ltaS depletion strain. Here, we further elucidated the mechanism. Our results establish that the mis-localization is not caused by the cell division defect of ΔltaS, as evidenced by the ΔltaS suppressor mutant ANG1786. Complementation experiments with a catalytically inactive variant, LtaS_T300A, reveals that SpA cross-wall localization requires the LtaS enzymatic activity. It is unlikely that the single amino acid mutation of T300A affects LtaS folding or localization, which in turn would affect SpA localization, since the point mutation is located in the extracellular enzymatic domain of LtaS, and the protein could be stably produced.

One of the key findings of this study is the cell cycle-dependent localization of LTA in staphylococci. Previous immunoelectron microscopy and immunofluorescence microscopy studies indicated that LTA is embedded underneath the peptidoglycan layer and occupies the periplasmic space between the peptidoglycan layer and the cytoplasmic membrane (Aasjord & Grov, 1980, Matias & Beveridge, 2008, Reichmann & Grundling, 2011, Reichmann et al., 2014, Shiraishi et al., 2018). In agreement with these previous results, LTA was not detected on the staphylococcal cell surface in our experiments. Protoplast-based immunofluorescence microscopy enabled us to detect LTA underneath the peptidoglycan layer during cell division. Intriguingly, LTA covers the old hemispheres (the peripheral sites) of the daughter cells, but is hardly detectable at the septum in actively dividing cells. The absence of LTA at the septum cannot be explained by incomplete lysostaphin digestion or incomplete daughter cells separation, as SpA is readily detected at the septum using the same protocol. We propose that the concentration of the newly synthesized LTA is too low or the length of LTA is too short at the septum, which limits its detection in our experiments. In comparison, the concentration of mature LTA is high at the peripheral membrane. An earlier study has proposed a model whereby LTA is synthesized at the septum and distributed to the cell envelope during cell division and separation (Reichmann et al., 2014). Our observations motivate some modification of this model. We propose that mature LTA occupies the periphery of staphylococcal cells; during cell division, new LTA is synthesized at the septum and gradually becomes mature and abundant; LTA eventually covers the septal membrane before the daughter cells are ready to split; after cell separation, LTA covers the entire cell membrane of the new daughter cells, which becomes the periphery of the cells and is ready for the next round of cell division.

The distinct localization pattern of LTA and SpA led us to propose a ‘restriction model’ in which the peripherally localized mature LTA restricts SpA to be at the septum. The localization of LTA underneath peptidoglycan layer suggests that LTA performs functions close to the cytoplasmic membrane. We envision that mature LTA occupies the periplasmic space at the peripheral sites, which blocks the secretion or restricts the diffusion of SpA in the membrane. In contrast, the concentration of LTA is low at the septum, which allows SpA to be efficiently secreted, diffused, and captured by SrtA. Early studies have revealed similar “restriction” mechanisms whereby WTA spatially restricts PBP4 and major autolysin Atl to be at the septum in S. aureus (Atilano et al., 2010, Schlag et al., 2010). However, an ‘attraction model’ cannot be completely ruled out at this point. The enzymatic activity of LtaS at the septum that produces LTA and generates membrane lipid turnover may attract protein secretion and localization. Besides regulating surface protein substrates, an alternative explanation could be that LTA synthesis regulates SrtA localization. The localization of SrtA has been examined in several Gram-positive bacteria. In S. pyogenes, SrtA localizes as foci that are primarily associated with septum, and SrtA is recruited to the septum at the early stage of cell division (Raz & Fischetti, 2008). In S. agalactiae, SrtA localizes at the division septum (Brega et al., 2013). In Enterococcus faecalis and S. mutans, SrtA localizes as single foci mostly found at the septum and the polar region (Hu et al., 2008, Kline et al., 2009). In S. pneumoniae, SrtA distributes over the entire cell (Tsui et al., 2011). The localization of SrtA in S. aureus has not been reported. As surface proteins are anchored at both the septum (YSIRK positive) and cell periphery (YSIRK negative) in S. aureus, we anticipate that SrtA may show a cell cycle-dependent localization, but will not be restricted to the septum. LtaS-mediated LTA synthesis maybe involved in supporting SrtA localization, which would be interesting for future studies. Collectively, our study provides new insights into the vital roles of teichoic acids in regulating the biogenesis of cell envelope components.

In addition, our study uncovers a new role of LTA D-alanylation in SpA spatial regulation. Interestingly, unlike the ltaS mutant, the dlt mutant diminishes SpA cross-wall deposition without affecting SpA septal anchoring. The dlt mutant displayed enlarged cell size, altered cell wall structure and increased cell lysis, which is in agreement with earlier observations (Santa Maria et al., 2014). These results led us to propose that D-alanylation likely affects peptidoglycan structure and cell wall growth, which in turn regulates SpA spatial distribution. This mechanism is reminiscent of ActA polar localization in L. monocytogenes, in which the polarization of ActA results from the cell wall growth rates and dynamics along the bacterial cell (Rafelski & Theriot, 2006). D-alanylation of LTA is known to regulate various properties of the cell envelope, such as maintaining cation homeostasis, modulating autolysin activity, and defining cell wall electromechanical properties (Peschel et al., 1999, Neuhaus & Baddiley, 2003). However, little is known about how the cell wall grows spatially in S. aureus and how D-alanylation facilitates proper cell wall growth. Future studies are needed to elucidate the mechanisms by which D-alanylation regulates dynamic cell wall growth and spatial distribution of surface proteins.

Experimental Procedures

Bacterial strains and growth conditions

Escherichia coli strains were grown in Luria-Bertani broth (LB) or on LB agar plate. S. aureus strains were grown in tryptic soy broth (TSB) or on tryptic soy agar plate (TSA). 100 μg/ml ampicillin (Amp) was used for plasmids selection in E. coli; 7.5 μg/ml chloramphenicol (Chl), 10 μg/ml erythromycin (Ery) and 50 μg/ml kanamycin (Kan) were used in S. aureus. 1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) was used to induce ltaS expression from P_spac promoter in SEJ1iltaS, WY541, WY543 and WY545; 0.1 mM IPTG was used to induce dltABCD expression in pLOW-dltABCD. 200 ng/ml anhydrotetracycline (ATc) (Takara) was used to induce gene expression from the P_tet promoter in pCLitet and 20 ng/ml ATc was used to induce dCas9 in pKK2. If not specified, overnight cultures were grown with appropriate antibiotics and 1:100 refreshed to TSB the next day. Appropriate inducers (ATc or IPTG) were added to the refreshed cultures. Refreshed cultures were grown for 2-3 hours to mid-log phase (OD₆₀₀ of 0.8), and proceeded with different experiments. The strains and plasmids used in this study are listed in Table 1.

Table 1.

Strains and plasmids used in this study

Strains or plasmids	Descriptiona	Reference or source
E. coli
DC10B	Cloning strain	(Monk et al., 2012)
S. aureus
SEJ1	RN4220Δspa, WY112	(Gründling & Schneewind, 2007a)
SEJ1iltaS	RN4220Δspa Pspac-ltaS, EryR, IPTG inducible, WY443	(Gründling & Schneewind, 2007b)
ANG1786	SEJ1ΔltaS suppressor 4S5, WY449	(Corrigan et al., 2011)
WY109	RN4220 sbi::φNΣ, EryR	(Yu et al., 2018)
WY110	RN4220ΔspaΔsbi::φNΣ, EryR	(Yu et al., 2018)
WY478	SEJ1 pCLitet-spa, ChlR, ATc inducible	This study
WY745	SEJ1 pCLitet-spaΔlysM, ChlR, ATc inducible	This study
WY480	SEJ1 pCLitet, ChlR, ATc inducible	This study
WY1026	WY110 pCLitet-spa, ChlR, ATc inducible	This study
WY946	WY110 pCLitet-spaΔlysM, ChlR, ATc inducible	This study
WY1024	WY110 pCLitet, ChlR, ATc inducible	This study
WY747	SEJ1iltaS pCLitet-spa, ChlR, EryR, ATc and IPTG inducible	This study
WY748	SEJ1iltaS pCLitet-spaΔlysM, ChlR, EryR, ATc and IPTG inducible	This study
WY794	SEJ1iltaS pCLitet, ChlR, EryR, ATc and IPTG inducible	This study
WY477	ANG1786 pCLitet-spa, ChlR, ATc inducible	This study
WY777	ANG1786 pCLitet-spaΔlysM, ChlR, ATc inducible	This study
WY479	ANG1786 pCLitet, ChlR, ATc inducible	This study
WY541	ANG1786 pCLitet-spa, pJK4, ChlR, KanR, ATc and IPTG inducible	This study
WY543	ANG1786 pCLitet-spa, pJK4-ltaS, ChlR, KanR, ATc and IPTG inducible	This study
WY545	ANG1786 pCLitet-spa, pJK4-ltaST300A, ChlR, KanR, ATc and IPTG inducible	This study
WY849	SEJ1ΔdltC::kan pCLitet-spa, KanR, ChlR, ATc inducible	This study
WY850	SEJ1ΔdltC::kan pCLitet-spa, pLOW-dltABCD, KanR, ChlR, EryR, ATc and IPTG inducible	This study
WY743	SEJ1 pCL55-spa, ChlR	This study
WY744	SEJ1 pCL55-spaΔlysM, ChlR	This study
WY851	SEJ1 pCL55-spa, pKK2-dltA, ChlR, EryR, ATc inducible	This study
WY852	SEJ1 pCL55-spa, pKK2, ChlR, EryR, ATc inducible	This study
WY853	SEJ1 pCL55-spaΔlysM, pKK2-dltA, ChlR, EryR, ATc inducible	This study
WY854	SEJ1 pCL55-spaΔlysM, pKK2, ChlR, EryR, ATc inducible	This study
Plasmids
pCL55	S. aureus chromosomal integration vector, AmpR (E. coli), ChlR (S. aureus)	(Lee et al., 1991)
pCL55-spa	Full length spa with its native core promoter cloned in pCL55, AmpR (E. coli), ChlR (S. aureus)	(Yu et al., 2018)
pCL55-spaΔlysM	Deletion of lysM (encoding 413-457 aa of SpA) in pCL55-spa, AmpR (E. coli), ChlR (S. aureus)	This study
pCLitet	pCL55 containing ATc-inducible Ptet promoter, AmpR (E. coli), ChlR (S. aureus), ATc inducible	(Gründling & Schneewind, 2007b)
pCLitet-spa	Full length spa cloned in pCLitet, AmpR (E. coli), ChlR (S. aureus), ATc inducible	This study
pCLitet-spaΔlysM	lysM deletion in pCLitet-spa, AmpR (E. coli), ChlR (S. aureus), ATc inducible	This study
pJK4	Complementation plasmid, KanR (E. coli& S. aureus), IPTG inducible	(Kern & Schneewind, 2010)
pJK4-ltaS	Full length ltaS cloned in pJK4. KanR (E. coli& S. aureus), IPTG inducible	This study
pJK4-ltaST300A	T300A variant of LtaS cloned in pJK4. KanR (E. coli& S. aureus), IPTG inducible	This study
pLOW-dltABCD	dltABCD expressed from Pspac promoter, AmpR (E. coli), EryR (S. aureus), IPTG inducible	(Santa Maria et al., 2014)
pKK2	CRISPR-interference knockdown vector, AmpR (E. coli), EryR (S. aureus), ATc inducible	This study
pKK2-dltA	pKK2 with dltA sgRNA, AmpR (E. coli), EryR (S. aureus), ATc inducible	This study

^a.Abbreviations: Chl, chloramphenicol; Ery, erythromycin; Kan, kanamycin; Amp, ampicillin; ATc, anhydrotetracycline; IPTG, Isopropyl β-d-1-thiogalactopyranoside.

Construction of plasmids and strains

The primers used in this study are listed in Table 2. PCR-mediated mutagenesis was used to construct pCL55-spa_ΔlysM, pCLitet-spa_ΔlysM, and pJK4-ltaS_T300A. Primers containing desired deletion or mutation were used to amplify the corresponding template plasmids (pCL55-spa, pCLitet-spa_, pJK4-ltaS). The PCR product was digested with DpnI and transformed to E. coli DC10B (Monk et al., 2012). Primers 443/444 was used to amplify spa from S. aureus Newman genomic DNA. The resulting PCR product was digested with AvrII and SacII and ligated with pCLitet (Gründling & Schneewind, 2007a), generating pCLitet-spa. Primers 587/588 were used to PCR amplify ltaS from RN4220 genomic DNA. The PCR product was restricted by XbaI and KpnI and ligated with plasmid pJK4 (Kern & Schneewind, 2010), generating pJK4-ltaS. The positive clones confirmed by DNA sequencing were electroporated to the desired S. aureus strains. Chromosomal integration of pCL55 and pCLitet derivatives was confirmed by PCR using primers pairs 184/185 and 186/187.

Table 2.

Primers used in this study

Primer No.	Sequence	Description
494F	agaagatggtaacgataagaagcaaccagcaaacca	spalysM deletion
495R	tgctggttgcttcttatcgttaccatcttctttacca	spalysM deletion
443F	cccCTAGGtcgttatattatgatgactttaca	pCLitet-spa
444R	cggGGTACCCCGCGGttatagttcgcgacgacgtcca	pCLitet-spa
587F	gggTCTAGAatgagttcacaaaaaaagaaaattagt	pJK4-ltaS
588R	gggGGTACCttattttttagagtttgctttaggt	pJK4-ltaS
373F	aggtaaagcatctgactctgaattta	pJK4-ltaST300A
374R	agtcagatgctttaccttgacctgt	pJK4-ltaST300A
184F	agcaacaccacataatggttcaca	184-F-attL
185R	atcgccatcaattcaaggatagca	185-R-attL
186F	tgacgttgagcctcggaaccct	186-F-attR
187R	tgatcgtagacctaatgacga	187-R-attR
TM371	ccatatagtttttgtatacgg	origin, forward
TM1885	gacttagaagcaaacttaagagtg	ermB, forward
KK1897	CATGGCTCTAGTCGACgagtaaacttggtctgacag	AmpR, forward
KK1898	acaaaaactatatggGTTAGAAGACGTCAGGTGG	AmpR, reverse
KK1899	GTTTGCTTCTAAGTCcatgaccaaaatcccttaacg	origin, reverse
KK1900	CTGTTGAACTCTCGAGggaactatgaaaaaggaacg	ermB, reverse
KK1915	AAACatgattttattcggtcacat	dltA sgRNA, forward
KK1916	ATGTatgtgaccgaataaaatcat	dltA sgRNA, reverse
KK1921	ccttattgttcgtctactcg	sgRNA sequencing, forward
KK2296	cgaatcaaacaagtattatcctatatacg	dltD probe, reverse
KK2297	GTTTATAATACGACTCACTATAGGGAGAagacatgtttttctgctggaac	dltD probe, forward

To construct CRISPR-interference knockdown plasmid pKK2, pJMP1337 (Peters et al., 2019) was digested with XhoI and SalI to obtain the dCas9 cassette. Separately, the Amp^R cassette was amplified from pJMP1337 with primers KK1897 and KK1898. The origins pBR322 and pSK41 were amplified from pLOW (Liew et al., 2011) with primers TM371 and KK1899. The ermB cassette was amplified from pTM239 with primers TM1885 and KK1900 (Santiago et al., 2015). The plasmid was then assembled using the In-Fusion cloning kit (Takara Bio USA) to yield pKK2. The dltA sgRNA oligonucleotides were annealed in a 30 μl reaction containing 50 μM of each oligo in 50 mM NaCl by heating at 95°C for 5 min and cooling to 4°C at a rate of 0.1°C/second. The annealed oligo was cloned into pKK2 using the BsaI-HF®v2 Golden Gate Assembly Kit (NEB) at 37°C for 60 minutes followed by 60°C for 5 min. Assembly transformants were screened by colony PCR using TM1885 and KK1916 and sequenced with KK1921 to confirm.

Strain ANG1786 is a kind gift from Angelika Gründling’s lab. Strains of WY849, WY850 were generated by transducing dltC::kan from TM852 (RN4220ΔdltC::kan) (Kho & Meredith, 2018) to the desired target strains. Strains of WY897, WY899 were generated by transducing srtA::φNΣ from Newman srtA::φNΣ.

Protein A immunofluorescence microscopy

Two protocols described previously were used with modification (Yu et al., 2018, Scaffidi et al., 2021). Protocol A, surface deposition of newly synthesized SpA. 2 ml of mid-log cultures were harvested by centrifugation, washed once with PBS and incubated in 1 ml PBS containing 0.5 mg/ml trypsin (Sigma) for 1 h to remove the pre-deposited SpA on the bacterial surface. Subsequently, cells were washed twice with PBS, resuspended in fresh TSB containing 1 mg/ml soybean trypsin inhibitor (Sigma) and incubated at 37°C for 20 min with rotation. Immediately following 20 min incubation, 250 μl of the cell suspension was taken and fixed with the fixation solution (2.4% paraformaldehyde and 0.01% glutaraldehyde in PBS). The samples were fixed for 20 min at room temperature, washed three times with PBS and applied to poly-L-lysine coated 8-well glass slides (MP Biomedicals). Excess and non-adherent cells were washed away with PBS and samples were proceeded with SpA immunofluorescence microscopy.

Protocol B, SpA localization in protoplasts. 2 ml of mid-log phase cultures were harvested, washed once with PBS and resuspended in 1 ml PBS containing 0.5 mg/ml trypsin. After incubation at 37°C for 1 h, staphylococcal cells were washed twice with PBS and fixed with 50% of fixation solution for 15 min at room temperature and 15 min on ice. The fixed cells were washed three times with PBS and resuspended in 1 ml GTE buffer (50 mM glucose, 20 mM Tris-HCl pH 7.5, 10 mM EDTA). 20 μg/ml lysostaphin (AMBI) was added to the cell suspension, and 50 μl of cell suspension was immediately applied to poly-L-lysine coated slides and incubated for 2 min. Non-adherent cells were sucked away and cells were washed once with PBS. Excessive liquid was aspirated and the slides were air-dried. Dried slides were immediately dipped in methanol at −20°C for 5 min followed by dipping in acetone at −20°C for 30 s. After the slides were completely dried, the samples on the slides were re-hydrated with PBS for 5 min and proceed with SpA immunofluorescence microscopy.

Samples prepared from the above two protocols were blocked with freshly made 3% BSA for 45 min and incubated with rabbit anti-SpA_KKAA antiserum (Kim et al., 2010) (1:4000 in 3% BSA) for 1 h. Cells were washed 8 times with PBS and further incubated in dark with anti-rabbit IgG conjugated with either Alexa Fluor 488 or Alexa Fluor 647 (1:500 in 3% BSA) (Invitrogen) for 1 h. Cells were washed 10 times with PBS and incubated with 50 μg/ml Hoechst 33342 DNA dye, 5 μg/ml Nile red (Sigma) or 1 μg/ml BODIPY® FL-conjugated vancomycin (Van-FL) (Invitrogen) for 5 min in dark followed by washing 3 times with PBS. A drop of SlowFade™ Diamond Antifade Mountant (Invitrogen) was applied to the samples before sealing with coverslips. Fluorescence images were captured on a Leica SP8 3X STED Laser Confocal Microscope with HC PL APO 63×oil objective (1.4NA, WD 0.14 mm).

To quantify the frequency of SpA cross-wall localization, the number of cells with cross-wall SpA localization and the total number of cells from the same image were counted. The frequency was calculated by dividing the numbers of cells with SpA cross-wall localization by total cell numbers. To quantify the intensity of SpA signals at the cross-wall, a method described earlier was used with modification (Yu & Götz, 2012). A line was drawn at the middle of the cell and perpendicular to cross-wall in Image J, generating a histogram of fluorescence intensity along the line. The relative cross-wall signal intensity is the ratio of cross-wall signals versus the peripheral signals of the same cell. At least three random images were taken for each sample in one experiment. Minimum 50 cells per sample per experiment were taken for quantification. Three independent experiments were performed to analyze statistical difference. All the images were analyzed in Image J software (Schneider et al., 2012).

LTA immunofluorescence microscopy

The immunofluorescence microscopy protocol for SpA protoplast localization described above was used to detect LTA localization. Briefly, 2 ml of mid-log phase staphylococcal cultures (SEJ1 and ANG1786) were sedimented by centrifugation, washed once with PBS and fixed in 1 ml 50% fixation solution. Cells were washed and digested with lysostaphin on slide. The resulting protoplasts were fixed and permeabilized with methanol and acetone. Samples were stained with LTA-specific monoclonal antibody (mAb 55, Novus Biologicals) (1:50 dilution in 3% BSA) followed by incubation with Alexa Fluor 488-conjugated anti-mouse IgG (1:500 dilution in 3% BSA). Fluorescence images were captured on a GE Applied Precision DeltaVision Elite deconvolution fluorescence microscope using a Photometrics CoolSnap HQ2 camera. At least three random images were taken for each sample in one experiment. Three independent experiments were performed to analyze statistical difference. All the images were analyzed in Image J software (Schneider et al., 2012).

Cell fractionation and SpA immunoblotting

The cell fractionation experiment was performed according to the protocol described previously (Yu et al., 2018). Briefly, bacteria cultures (WY946, WY1024 and WY1026) were grown to mid-log phase with ATc induction and normalized to OD₆₀₀ of 1. One milliliter of culture was centrifuged at 18,000 × g for 5 min in a table centrifuge. The supernatant was carefully transferred to a clean tube (supernatant fraction). The pellet was resuspended in 1 ml TSM [50 mM Tris-HCl (pH 7.5), 0.5 M sucrose, 10 mM MgCl₂] and incubated with 20 μg/ml lysostaphin for 10 min at 37°C. After centrifugation at 18,000 × g for 5 min, the supernatant (cell wall fraction) was transferred to a clean tube. The protoplast pellet was resuspended in 1 ml Tris buffer [50 mM Tris-HCl (pH 7.5), 150 mM NaCl] and subjected to five times freeze-thaw cycle in dry ice/ethanol and warm water bath. Membranes were in the cell lysate were sedimented by centrifugation at 20,000 × g for 30 min. Supernatant was transferred to another tube (cytosolic fraction) whereas the pellet (membrane fraction) was suspended in 1 ml Tris buffer. Proteins from different fractions were precipitated by 10% trichloroacetic acid (TCA), washed with acetone, and solubilized in 100 μl 1×SDS sample buffer (62.5 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 5% 2-mercaptoethanol, 0.01% bromophenol blue).

To collect proteins from cell pellet (P) and supernatant (S), 1 ml of mid-log phase staphylococcal culture was centrifuged at 18,000 × g for 5 min. The supernatant was carefully transferred to a clean tube. The cell pellet was washed once and resuspended in 1 ml of Tris buffer. Subsequently, the cell suspension was incubated with 20 μg/ml lysostaphin at 37°C for 30 min. Proteins from the culture supernatant and lysostaphin-digested cell pellet fraction were precipitated with 10% TCA, washed with acetone, and solubilized in 100 μl 1×SDS sample buffer.

SpA immunoblotting protocol has been described previously in detail (Yu et al., 2018). Protein samples were separated in 10% or 12% SDS-PAGE, blotted with primary antisera (αSpA_KKAA 1:10,000 dilution, αSrtA 1:20,000 dilution, αSigA 1:5,000 dilution) and secondary anti-rabbit IgG linked with HRP (1:10,000 dilution). αSpA_KKAA and αSrtA antisera are kind gifts from Schneewind-Missiakas lab; αSigA (sigma factor A) antiserum is a kind gift from Prahathees Eswara’s lab.

LtaS and LTA Immunoblotting analysis

For LtaS immunoblotting, protein samples from the lysostaphin-digested cell pellet faction were collected as described above and separated in 10% SDS-PAGE. 50 μl human IgG (Sigma) was added to 10 ml block-solution (5% milk) to block SpA cross-reaction. Rabbit anti-LtaS serum (Richter et al., 2013) was used at 1:5,000 dilution. The protocol described in (Gründling & Schneewind, 2007a) was adopted for LTA immunoblotting. Mid-log phase staphylococcal cultures were normalized to OD₆₀₀ of 4 per ml. 1 ml of the normalized culture was mixed with 0.5 ml of 0.1 mm glass beads. The cells were disrupted in a Mini-Beadbeater (BioSpec) by agitating violently and repetitively for six times; each time with 1 min agitating and 1 min cooling on ice. The glass beads were sedimented by centrifugation at 200 × g for 1 min. 0.5 ml of the supernatant was transferred to a new tube and centrifuged at 18,000 × g for 10 min. The pellet was suspended in 50 μl 1×SDS sample buffer. The samples were separated in 15% SDS-PAGE and immunoblotted with LTA-specific monoclonal antibody (mAb 55, Novus Biologicals) with a dilution of 1:2500.

RNA extraction and Northern blotting

S. aureus harboring pKK2 were grown overnight in TSB containing 5 μg/ml erythromycin at 37°C and diluted 1:100 into 3 ml of fresh media containing 5 μg/ml erythromycin with either no ATc, 1 ng/ml ATc, or 20 ng/ml ATc. Cultures were grown at 37°C to OD₆₀₀ =0.8 whereby 1 ml of each sample was harvested by centrifugation. Cell pellets were suspended in 900 μl of Qiazol lysis reagent (Qiagen) and kept at 4°C for disruption by bead-beating in 2 ml screw cap tubes using a Mini-Beadbeater (Biospec Products) at 3800 rpm for two 30-sec cycles. The beads were collected by brief centrifugation and 600 μl of lysate was transferred into microcentrifuge tubes. The beads were washed with an additional 300 μl of Qiazol and the wash was combined with the lysate. RNA was extracted using the Qiagen RNeasy Plus Universal Mini Kit following the manufacturer’s protocol. 1 μg of each RNA sample was used for a formaldehyde-based northern analysis as described previously (Armbruster et al., 2019). The template for the RNA probe targeting dltD was amplified with primers KK2296 and KK2297.

Supplementary Material

si1

Figure S1. LysM domain is dispensable for SpA mis-localization upon ltaS depletion. (A) Surface localization of newly deposited SpA and SpA_ΔLysM upon ltaS depletion. Wild-type spa and spa_ΔlysM were expressed from ATc-inducible P_tet promoter in SEJ1iltaS (RN4220Δspa P_spac-ltaS). The localization of the newly deposited SpA on bacterial cell surface was analyzed in cells with ltaS induction (+IPTG) and depletion (−IPTG). SpA-specific fluorescence signals are in green. Bacterial cell membrane was stained with Nile red (red). Vector, pCLitet empty vector. Representative images are from three independent experiments. (B) Quantification of SpA cross-wall localization from the samples in panel G. Unpaired, two-tailed t-test was used for statistical analysis; ****, p<0.0001.

Data availability statement

The data that support the findings of this study are available from the corresponding: author upon reasonable request.