Absence of sanA restores growth to a mutant defective in FtsI, a division-specific D, D-transpeptidase

During a genetic screen performed to obtain factors that influence cell division (as described in Materials and Methods), we observed that absence of sanA, a gene of unknown function, restored growth to a strain carrying a temperature-sensitive mutation in ftsI (ftsI23) [18]. In addition to re-establishing growth at the restrictive temperature, sanA deletion suppressed the division defects of the ftsI23 mutant (Fig 2A–2E). In E. coli, sanA is the first gene in a two-gene operon (sanA-yeiS) located approximately at 48 min on its chromosome (Fig 2B). However, deletion of the downstream gene, yeiS, did not affect the growth phenotypes of the ftsI23 mutant (S1A Fig). Furthermore, a plasmid encoding sanA cloned downstream to an IPTG-inducible promoter (P_trc::sanA) completely abrogated the suppression conferred by the deletion of sanA in the ftsI23 mutant (Fig 2C–2E).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Absence of sanA rescues the growth defects of several cell division mutants.

(A) Overnight cultures of WT, ftsI23, and ftsI23 ∆sanA strains were serially diluted and 4 μL of each dilution were spotted on LB plates and incubated at indicated temperatures. Growth phenotypes were scored after 20-24 h incubation. (B) Operonic arrangement of sanA and yeiS genes located at 48.11-48.12 min on the E. coli chromosome (not drawn to scale). Arrow indicates the direction of transcription. (C) Indicated strains were grown in LB with ampicillin at 30°C and tested for viability as described above on LB plates supplemented with ampicillin and 10 µM IPTG. (D, E) Overnight cultures of WT or its derivatives carrying pTrc99a or pTrc99a-sanA were sub-cultured at 1:100 dilution in LB supplemented with ampicillin and 10 µM IPTG at 37°C. Growth was monitored at regular intervals. Cells were collected after 3 h for visualization using Differential Interference Contrast (DIC) microscopy. Scale bars represent 5 µm. (F) ftsE and ftsE ∆sanA strains were grown overnight in LB with 1% NaCl at 30°C and cell viability was tested on indicated plates at 30°C. Microscopy was done as described above. Overnight cultures of strains carrying ftsK44, or ftsZ84 alleles were grown at 30°C, serially diluted and viability was assessed on indicated plates (LB at 30°C and LBON at 37°C).

https://doi.org/10.1371/journal.pgen.1011712.g002

Deletion of sanA restores growth and division to several other mutants defective in cell division

To examine the effect of SanA on the phenotypes of other division mutants, sanA deletion was introduced into multiple strains each carrying mutations in ftsZ (ftsZ84), ftsA (ftsA12), ftsE (ftsE::Tn10dCm), ftsX (ftsX::Tn10dCm), ftsK (ftsK44), ftsQ (ftsQ1) or ftsP (ΔftsP::Kan) [19,20]. Interestingly, deletion of sanA effectively suppressed the growth defects of all the above mutant strains except that of the strains carrying mutations in the early division genes, ftsZ (ftsZ84) or ftsA (ftsA12) (Figs 2F, and S1B,S1C). In addition, absence of SanA had no significant effect on the growth of a triple amidase (ΔamiABC) mutant defective in the process of cell separation (S1D Fig). Overall, the above results show that absence of sanA is able to overcome the growth defects of several cell division mutants.

Absence of SanA rescues the growth defects of mutants lacking other PG synthases

As sanA deletion suppressed division-defective PG synthase mutation (ftsI23), we examined its effect on the phenotypes of other PG synthase mutants. Fig 3 shows that absence of SanA is able to rescue the growth of strains depleted of either PBP1A (encoded by mrcA) or PBP1B (encoded by mrcB), the bifunctional aPBPs that facilitate wall synthesis during PG expansion (using a ΔmrcA ΔmrcB double deletion strain carrying either P_ara::mrcA or P_lac::mrcB as shelter) [21]. In addition, growth of a mutant carrying pbpA45, a hypomorphic allele of PBP2 [22], a monofunctional transpeptidase involved in scaffold-synthesis, was enhanced in absence of sanA (Fig 3C). However, sanA deletion did not confer any growth advantage to strains defective in the synthesis of other envelope components such as LPS or PL (S2A and S2B Fig). Overall, this data suggested that SanA may have a function in PG-associated processes.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Deletion of sanA suppresses the growth defects of cell elongation mutants.

(A) A double mutant of ∆mrcA ∆mrcB carrying a chromosomal copy of mrcA downstream to an arabinose-inducible promoter (P_ara::mrcA) and its sanA deletion derivative were grown overnight with 0.5% arabinose, serially diluted and tested for viability on LB plates supplemented with arabinose (Ara) at 37°C. (B) Viability of a double mutant of ∆mrcA ∆mrcB carrying a chromosomal copy of mrcB downstream to an IPTG-inducible promoter (P_lac::mrcB) or its sanA derivative was tested as described above. IPTG was used at indicated concentrations. (C) Viability of strains carrying pbpA45 or pbpA45 ∆sanA mutations grown overnight in LB at 30°C was tested on indicated plates (LB at 30°C and LBON at 42°C).

https://doi.org/10.1371/journal.pgen.1011712.g003

Absence of SanA increases nascent PG synthesis

As above results suggested a role for SanA in PG metabolism, we examined the composition of PG sacculi in WT and sanA deletion mutant. However, we did not find any significant alteration in PG composition of these strains (S2C and S2D Fig). Next, we examined the rate of nascent PG strand incorporation by measuring PG synthesis using tritiated meso-diaminopimelic acid (³H-mDAP) as described in Materials and Methods [23]. Surprisingly, sanA deletion showed a significant increase (nearly 1.5 to 2-fold) in the rate of PG incorporation compared to that of WT (Fig 4A), which is complemented by a plasmid-borne copy of sanA (Fig 4B). We also examined the rate of PG synthesis in both division and elongation mutants (carrying ftsI23, ftsE, ftsZ84, ΔmrcA ΔmrcB/ P_lac::mrcB, or pbpA45 alleles) using ³H-mDAP incorporation assay. Deletion of sanA increased mDAP counts in all the strains tested above (Fig 4C and 4D) like that of WT. Interestingly, sanA deletion increased the mDAP incorporation even in the strain carrying ftsZ84, in which the growth rescue was not observed. In summary, sanA deletion increased the nascent PG synthesis in all the strains tested, implying its basis of suppression is through a generalized increase of PG. Importantly, these results show that SanA negatively regulates PG synthesis likely contributing to balanced formation of PG in E. coli.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Effect of SanA deletion on nascent peptidoglycan (PG) synthesis.

Nascent PG strand incorporation into the sacculi was measured using ³H-mDAP as described earlier (23). All strains had a deletion of lysA to prevent mDAP conversion to lysine and grown in Min-A (minimal-A medium) supplemented with 0.2% glucose and 0.5% Casamino acids. PG synthesis was measured in normalised cultures that were grown to an OD₆₀₀ value of 0.4-0.6. (A) WT and sanA mutant strains are grown at 37°C and PG synthesis measured as described above. (B) Above strains carrying either pTrc99a or pTrc99a-sanA were grown at 37°C with 10 µM IPTG and the ³H-mDAP incorporation assay was carried out as described above. (C) Cultures of ftsI23, ftsE, and ftsZ84, along with their sanA deletion derivatives, were grown at 30°C and processed for ³H-mDAP incorporation as described earlier. (D) Derivatives of ∆mrcA ∆mrcB with P_lac:: mrcB (∆mrcAB/1B+) strain with or without sanA deletion were grown at 37°C in presence of 100 µM IPTG and ³H-mDAP incorporation assay was performed as described above. pbpA45 and pbpA45 ∆sanA were grown at 30°C for ³H-mDAP incorporation assays. All the measurements were done at least three times and ratios were calculated and plotted as bar graphs.

https://doi.org/10.1371/journal.pgen.1011712.g004

As the above results suggested a role for SanA in cell wall synthesis, we evaluated the effect of various cell wall-targeting antibiotics on the growth of sanA mutant (Fig 5A). Here, we observed that treatment with fosfomycin, which inhibits the formation of cytosolic PG precursors by targeting MurA, or moenomycin, which targets the glycosyltransferase activity of class A PBPs had no discernible effect on the growth of sanA mutant. However, interestingly, cefsulodin, a β-lactam antibiotic which targets the transpeptidase activity of class A PBPs conferred a moderate growth advantage compared to that of WT supporting a role for SanA in PG synthesis.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Phenotypes of sanA deletion mutant.

(A) Effect of cell-wall targeting antibiotics was examined on the growth of WT and ∆sanA strains on LB, LB+fosfomycin (1 µg/mL), LB+moenomycin (6 µg/mL), or LB+cefsulodin (24 µg/mL) at 37°C. (B) Viability assays of WT and ∆sanA strains were done on LB, LB + 200 µg/mL vancomycin (Van), LB + 1.0% SDS + 0.5 mM EDTA, or NA plates at indicated temperatures. (C) Growth of indicated strains was scored at 43°C on LB + 1.0% SDS + 0.5 mM EDTA and NA plates supplemented with 10 µM IPTG. (D, E) Overnight cultures of WT and ∆sanA were diluted 1:100 in Nutrient Broth (NB) and LB and grown at 43°C. Growth was monitored at regular intervals. At OD₆₀₀ of 0.6-0.8, cells were collected, immobilized on 1% agarose pads, and visualized using DIC microscopy. For cell length measurements, approximately 100 cells were used and data were analysed by ImageJ software.

https://doi.org/10.1371/journal.pgen.1011712.g005

Absence of sanA confers cell envelope defects

Considering an earlier report that sanA mutants exhibit permeability defects at elevated temperatures (43°C), we revisited this phenotype. Consistent with previous findings [24], we observed that the sanA mutant is indeed sensitive to vancomycin and treatment with SDS+EDTA at higher temperatures (Fig 5B and 5C). In addition, we found that the mutant exhibits reduced growth rate in low-osmolarity media such as Nutrient Agar (NA) which is rescued by addition of osmolytes such as NaCl or sorbitol (Figs 5B and S3A). However, all these phenotypes were evident only when cells were grown at 43°C but not at 37°C, as reported earlier [24]. To check whether the permeability defects are arising because of LPS deficiency, we examined its levels, and consistent with the earlier study [24], no alterations in LPS were observed (S3B Fig). While no obvious morphological defects were observed in LB, sanA mutants grown in NB (Nutrient Broth) were significantly smaller compared to that of WT (Figs 5D, 5E and S3C). In summary, the above experiments show that absence of SanA leads to compromised envelope barrier properties which are LPS-independent.

Interactions of sanA with ECA biosynthetic pathway

To investigate the underlying basis of the sanA mutant phenotypes, we performed a genetic screen using a multicopy plasmid library (as described in Materials and Methods), selecting colonies that grew better on NA plates at 43°C. Among the suppressors identified, a clone carrying the region encompassing both wecA and wzzE genes (early genes in the operon encoding ECA biosynthetic pathway) conferred a moderate growth advantage to sanA mutant (S4A Fig). We confirmed the suppression by constructing a clone carrying wecA-wzzE genes on a medium-copy plasmid vector downstream to an IPTG-inducible promoter (P_trc::wecA-wzzE) (Fig 6A). To understand the basis of suppression, we measured the rate of PG synthesis in strains overexpressing wecA-wzzE genes. Interestingly, overexpression of these genes decreased the incorporation of ³H-mDAP to two-fold both in the WT and sanA mutant, suggesting that wecA-wzzE overexpression generally reduces PG synthesis (Fig 6B) most likely by channelling the precursors away from PG into ECA pathway.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Genetic interactions of sanA with ECA pathway.

(A) Indicated strains were grown overnight in LB containing ampicillin and growth was checked on LB, NA, and LB + 200 μg/mL vancomycin plates supplemented with ampicillin and 10 µM IPTG. (B) The above-described strains (with additional lysA mutation) were used to measure ³H-mDAP incorporation as described earlier. Cells were grown with 10 µM IPTG. (C) Growth of the indicated strains was examined on LB, LB + 1.0% SDS + 0.5 mM EDTA or NA plates incubated at 37°C. (D, E) Growth of sanA, wecA and sanA wecA mutant strains was tested in NB at 43°C by measuring OD₆₀₀ at regular intervals and after 4h of growth, cells were collected, immobilized on 1% agarose pads, and visualized using DIC microscopy. Black arrows indicate cell lysis. The scale bar represents 5 μm.

https://doi.org/10.1371/journal.pgen.1011712.g006

We next examined the sanA mutant phenotypes in absence of ECA by introducing deletions in each of the genes involved in ECA synthesis into a sanA deletion strain. Here, we observed that deletion of wecA, wecG or wecF, the early genes in ECA biosynthesis in sanA mutant resulted in synthetic sickness on NA plates at 37 °C wherein the single mutants had no discernible defects (Fig 6C). The double deletion mutants also exhibited cell lysis in late exponential phase (Fig 6D and 6E). These observations suggested the additive sickness could be due to exacerbation of sanA’s defects in absence of ECA implying SanA functions outside of ECA pathway. In support of this, strains carrying deletion of wzxE or wzzE, the genes encoding the flippase or the co-polymerase of the ECA pathway which still make functional ECA due to the presence of redundant flippase/ polymerase (of O-antigen pathway) did not confer additive sickness (Fig 6C) [11].

SanA and ElyC work independently in PG and ECA pathways

Interestingly, ElyC, a paralog of SanA is shown to regulate ECA synthesis with elyC deletion mutants exhibiting elevated ECA levels and decreased PG synthesis [16,17]. To examine whether ElyC and SanA are functionally related to each other, we constructed sanA elyC double mutants and assessed their phenotypes. However, we find elyC single mutants themselves do not grow on NA plates (NA^S) at 43°C (Fig 7A). We then measured the rate of ³H-mDAP incorporation in these strains. In agreement with a previous study, approximately 50% less mDAP incorporation was observed in elyC single mutant [16]. More importantly, absence of sanA was able to elevate the mDAP counts approximately 1.5- to 2- fold in elyC mutant background as well (Fig 7B), suggesting the role of SanA in PG synthesis is independent of ElyC. Further, we measured ECA levels in these mutants using anti-ECA antibodies in a dot blot assay and find that the elevated ECA levels observed in elyC single mutant were not affected by absence of SanA (Fig 7C). Overall, these results suggest that the two paralogs, SanA and ElyC play distinct roles that are not interdependent on each other.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Genetic interactions of SanA with ElyC.

(A) Viability of the indicated strains was tested on LB and NA plates at indicated temperatures. (B) Above strains carrying lysA deletion were used to estimate nascent PG synthesis using ³H-mDAP incorporation. (C) Overnight cultures of the indicated strains were diluted 1:100 in LB and grown at 43°C to an OD₆₀₀ of 2.0. Cells were collected and processed for dot blot analysis as described in Materials and Methods. ΔwecA mutant is used as a negative control for detection of ECA and FtsZ as a normalization control.

https://doi.org/10.1371/journal.pgen.1011712.g007

Cleavage of PG mesh alleviates SanA⁻ phenotypes

To further dissect the function of SanA, we performed another multicopy suppressor screen (using the pACYC plasmid library) utilizing the severe NA-sensitivity phenotype of sanA wecA double deletion strain at 37°C. Among the suppressor plasmids, a clone containing a region encompassing grxD-mepH genes of the E. coli chromosome rescued both the NA- and SDS+EDTA-sensitivities of sanA mutant (S4B Fig). Subsequently, a plasmid carrying mepH alone (P_trc::mepH) [13] was found to restore the barrier defects of both sanA wecA and sanA mutants (Fig 8A–8D). MepH is a periplasmic PG endopeptidase that cleaves the peptide cross-links between the glycan strands to open up the mesh during expansion of the PG sacculus [13,25]. This observation prompted us to examine whether other cross-link specific PG endopeptidases were able to rescue the SanA⁻ phenotypes. Indeed, overexpression of either MepS or MepM, the two major cross-link cleaving PG endopeptidases [13], restored the barrier defects of sanA mutant suggesting cleaving the PG mesh is beneficial to this mutant (Fig 8D). Importantly, this observation indicated that the PG sacculus in absence of sanA could be aberrant leading to barrier defects which are alleviated by the cleavage of cross-links.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 8. Overexpression of PG cross-link cleaving hydrolases alleviates SanA⁻ phenotypes.

(A) Indicated strains were grown overnight in LB+ampicillin and growth was examined on LB, NA, LB + 1.0% SDS + 0.5 mM EDTA plates at 37°C. IPTG was used at 50 μM. (B, C) Overnight cultures of WT and its derivatives were diluted 1:100 in NB containing ampicillin and grown at 43°C. Growth was monitored at regular intervals, cells were collected after 4h of growth, and viability was tested using live-dead staining as described in Materials and Methods. Green and red indicate live and dead cells, respectively. (D) Strains were grown overnight in LB and viability was tested on LB, NA, LB + 1.0% SDS + 0.5 mM EDTA plates supplemented with indicated concentrations of IPTG at 43°C.

https://doi.org/10.1371/journal.pgen.1011712.g008

Evidence for a catalytic role of SanA in the periplasm

E. coli encodes four DUF218 domain containing proteins: SanA, ElyC, YdcF, and YgjQ. Among these, SanA, ElyC, and YgjQ are predicted to be IM-anchored with their DUF218 domains exposed to the periplasm, whereas YdcF is predicted to be a cytosolic enzyme involved in anaerobic metabolism [16,26]. In an attempt to understand whether DUF218 domain has any catalytic function, we analysed the structural and sequence alignment of these paralogs, and identified a conserved triad of threonine, histidine, and glutamate, which likely forms a catalytic core in these proteins (S5A Fig). To evaluate the functional importance of these conserved residues, we generated alanine substitution variants of SanA and assessed their ability to complement the sanA mutant. None of the three variants were able to complement the SanA⁻ phenotype, despite being expressed at levels comparable to the wild-type protein (Figs 9A and S5B). These findings suggest that the conserved triad residues are critical for its function, implying an enzymatic activity for SanA.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 9. Evidence for a catalytic role for SanA in periplasm.

(A) Complementation of SanA⁻ phenotypes by plasmid-borne sanA variants (T125A; H149A; C150A; E182A) was examined by assessing growth of indicated strains on LB or NA plates at 43°C. C150A variant was able to complement the San⁻ phenotype whereas the other variants did not. (B) Complementation of SanA⁻ phenotypes by a plasmid-borne ^ssdsbA-sanA variant was tested as described above on indicated plates at 43°C. (C) Complementation of NA^S of sanA wecA double mutant by the ^ssdsbA-sanA variant was tested as described above on indicated plates at 37°C.

https://doi.org/10.1371/journal.pgen.1011712.g009

To examine whether SanA works in the periplasm, we replaced its membrane anchor with the signal sequence from DsbA (^SSdsbA-sanA) and examined its functionality. As shown (Fig 9B and 9C), this variant complemented the SanA⁻ phenotypes confirming its site of action is in the periplasm. Overall, these results suggest that SanA functions in the periplasmic stage of PG synthesis, likely at the step of PG polymerization.

Role of SanA in PG biogenesis

During our attempts to understand septal PG synthesis, we discovered that sanA disruption suppresses not only ftsI23 mutant phenotypes, but also the defects of several mutants involved in cell division or elongation (Figs 2 and 3). This general suppressive ability of cell wall-associated mutants has prompted us to examine the effect of sanA deletion on PG synthesis. Here, we observed a significant increase in the rate of nascent PG synthesis in all the strains carrying a sanA mutation, suggesting SanA functions as a negative modulator of PG synthesis (Fig 4). A previous study from Salmonella typhimurium has shown that disruption of sanA^st (sfiX) suppressing the cell filamentation defect caused by HisH and HisF overexpression (His^c pleiotropic response), based on which, it was suggested that SanA has a role in PG synthesis [27].

In E. coli, SanA was earlier identified as a factor contributing to the resistance against vancomycin and SDS+EDTA [24]. It has also been shown to confer SDS-resistance during carbon-limited stationary-phase [28], altogether suggesting its role in the maintenance of envelope barrier properties [24,28]. Here, we find that sanA mutants are indeed sensitive to treatment with SDS+EDTA and vancomycin. In support of earlier observations [24], we also find that these barrier defects are not due to alterations in LPS (S3B Fig).

In an attempt to examine the basis of barrier defects in sanA mutant, we found that overexpression of wecA-wzzE is able to moderately rescue the SanA⁻ phenotypes. The overexpression was also able to reduce the rate of PG synthesis raising a possibility of decreased PG as the basis of sanA suppression (Fig 6A and 6B). We further tested the involvement of SanA in the ECA pathway by assessing the growth phenotypes of single and double mutants of sanA lacking each of the genes involved in ECA synthesis and observed that (1) the absence of ECA by itself does not lead to loss of barrier properties, and (2) a complete shutdown of ECA production is additively sick with a deletion of sanA likely due to a combination of dysregulated PG and loss of ECA (Fig 6C). These results allowed us to infer SanA’s role is not in ECA pathway.

Altogether, the above results showed SanA controls the formation of PG. Subsequently, identification of cross-link specific endopeptidases, MepS, -M and -H as factors suppressing the SanA⁻ phenotypes gave us a strong indication that the PG sacculus of sanA mutant could be dense and aberrant due to enhanced PG strand incorporation, leading to barrier defects (Fig 8A–8D). Although the basis of the suppression by these endopeptidases is not clear, it is plausible that cleavage of the peptide cross-links between the glycan strands may loosen the dense PG mesh thereby mitigating the barrier defects of SanA mutant. Moreover, the complementation of SanA⁻ phenotypes by ^ssdsbA-sanA variant strengthens the above result indicating SanA functions at the step of nascent PG strand polymerization, most likely by controlling the flux of new subunits into the growing PG mesh.

DUF218-family proteins contribute to Gram-negative bacterial cell envelope biogenesis

E. coli encodes four DUF218-domain containing proteins: SanA, ElyC, YdcF, and YgjQ. Topological predictions indicate that SanA, ElyC and YgjQ are IM-anchored with their DUF218 domains in the periplasmic space whereas YdcF is cytoplasmic [16,26]. Of these, ElyC is shown to control ECA levels, with its deletion exhibiting elevated ECA levels, particularly high ECA_PG [17]. Additionally, ElyC depletion leads to cell lysis at lower temperatures, accompanied with severe cell wall-related defects. This phenotype is attributed to the sequestration of the lipid carrier UndP in the ECA pathway leading to decreased PG synthesis [16].

On the other hand, SanA controls PG levels, with sanA mutants exhibiting elevated PG synthesis with compromised envelope permeability at high temperature. Although, dot blots revealed a moderate ECA decrease in sanA deletion mutant, (Fig 7C), we could not confirm through western blotting due to paucity of the antibody. Combining the above results, we propose these two DUF218 proteins work towards the maintenance of balanced PG and ECA synthesis. In support of this, both sanA and elyC genes are conserved mostly in the members of the order Enterobacterales of Gram-negative bacteria.

Among the DUF218 proteins, the structure of YdcF is known, revealing a fold similar to the adenine nucleotide α-hydrolase family. YdcF has been shown to bind S-adenosylmethionine and predicted to be an enzyme involved in anaerobic metabolism [26]. Additionally, the conserved amino acids in the DUF218 domain family hint these family members might have a potential enzymatic activity. In support of this, three residues that form a putative catalytic triad have been shown to be important for SanA function. It is interesting to speculate that both SanA and ElyC might function as enzymes working on key intermediates in the PG and ECA biosynthetic pathways, respectively.

Role of divisome proteins in septal PG synthesis

The observation that absence of sanA is a broad and general suppressor of several cell division mutants is intriguing. As sanA shows increased mDAP incorporation into the sacculus, an attractive possibility is that the basis of suppression is due to enhanced PG synthesis alleviating the division defects in these mutants. Bacterial cell division is driven by divisome, a multi-protein complex. In the early stages of division, FtsZ, along with ZipA, FtsA and other associated factors, establishes the FtsZ-ring at the division site. Subsequently, a series of Fts proteins (FtsE/X, FtsK, FtsQ, FtsL, FtsB, FtsW, FtsI, FtsN, FtsP) are recruited to facilitate the process of cell division. Of these, several proteins are known to be involved in septal PG synthesis [5,15,29,30], however, the precise molecular function of few of these proteins is not completely understood. sanA deletion suppressing the division phenotypes of mutants defective in ftsE, -X, -K, -Q and -P provides additional evidence that these factors are directly/indirectly involved in septal PG synthesis. In contrast, sanA deletion did not suppress ftsZ84 mutation, but rather exacerbated the phenotype (Fig 2F) indicating optimal levels of FtsZ could be required to establish the Z-ring as the PG sacculus in this mutant could be dense/ aberrant; however, further experiments are required to validate this idea.

Media and growth conditions

Strains were grown in LB (1% tryptone, 0.5% yeast extract, and 1% NaCl) unless otherwise indicated. LBON is LB without NaCl. Nutrient broth (NB) comprises 0.3% beef extract and 0.5% peptone. 1.5% agar is added to the respective broths to make solid media. Minimal A medium (Min-A) was supplemented with 0.005% vitamin B1, 1mM MgSO4, and 0.2% glucose before use [31]. Antibiotics were used at the following concentrations unless specified: Ampicillin (Amp-50 μg/mL), Kanamycin (Kan-50 μg/mL), and Chloramphenicol (Cm-25 μg/mL). Isopropyl β, D-thiogalactopyranoside (IPTG) was used at specified concentrations. Cells were grown at 37°C unless otherwise specified. Growth was monitored by measuring optical density at 600 nm (OD₆₀₀).

Strains and plasmid constructions

All the strains, plasmids, and primers used in this study are detailed in Supplementary Information (S1 Text).

Viability assays

Viability assays were performed by serially diluting the overnight culture (10^-2, 10^-4, 10^-5, and 10^-6), placing 4 μL of each dilution on indicated plates, and growing at a specified temperature for 18–24 h. Permeability assays were done on LB plates supplemented with 1% SDS + 0.5 mM EDTA or Vancomycin (200 μg/mL).

Microscopy

Overnight cultures were sub-cultured with 1:100 dilution and grown until required OD₆₀₀. At this point, cells were collected, washed, immobilized onto a 1% agarose pad, and visualized under the Zeiss apotome microscope in differential interference contrast mode (DIC, Normasky optics). For cell viability, Live/Dead backlight bacterial viability kit (Invitrogen) was used and cells were visualized by fluorescence microscopy with GFP and Texas red filters.

Molecular and genetic techniques

Plasmid and recombinant DNA constructions are described in S1 Text. Genomic DNA from MG1655 was used as a template for PCR amplification. All plasmids and strains were confirmed through sequencing and phenotypes. Transformations and P1 phage-mediated transductions were performed as previously described [31]. All strains are derivates of MG1655 unless otherwise mentioned. Deletion mutations are sourced from the Keio collection and transferred to the required strain background through P1-mediated phage transductions. The antibiotic resistance marker was flipped using pCP20.

Estimation of nascent PG synthesis by mDAP incorporation assay

The ³H-mDAP (tritiated meso-diaminopimelic acid) incorporation assay was performed as described earlier [23]. Indicated strains lacking LysA (to prevent the formation of lysine from mDAP) were grown overnight in LB broth. The next day, cells are washed and sub-cultured at 1:100 dilution in Minimal medium supplemented with 0.2% glucose and 0.5% CAA. At OD₆₀₀ of 0.4-0.6, normalized fractions are collected and incubated with 5 μCi/mL of ³H-mDAP (Moravek Biochemicals, USA) for 10 min with gentle shaking at 37°C. Cells were immediately lysed by addition of 3 mL of 4% SDS and boiled for 1h. The mixture was cooled overnight at RT and filtered through a 0.22 μm filter. The insoluble PG sacculi collected on the filters were washed with 30 mL of Milli-Q water and dried. The filters were used for counting radioactivity in a liquid scintillation counter (Perkin-Elmer). All the measurements were done at least three times and ratios were calculated and plotted as bar graphs.

Dot blot assay

Indicated strains were grown overnight in LB at 37°C. The next day, cells were sub-cultured with 1:100 dilution and were grown at 43°C till OD₆₀₀ value of 3. Cells were normalized to 5 OD, pelleted, washed with PBS once and resuspended in 100 µL of PBS containing 10 mM EDTA. The cell suspension was sonicated for 1 min with 10 sec on and 10 sec off cycles at 30% amplitude and briefly spun to remove un-lysed cells and debris. To normalize the sample, protein estimation was done using a Pierce BCA estimation kit (Thermo-Fisher Scientific). Normalised samples were serially diluted (10^-1, 10^-2) and 4 µL of each dilution were spotted onto the nitrocellulose membrane. Blots were developed using enhanced chemiluminescence prime-detection substrate (Amersham). Primary anti-ECA antibody was used (a kind gift from Jilong Qin and Renato Morona) at a dilution of 1:1000 and anti-FtsZ antibody was used at 1:50,000. HRP-conjugated anti-rabbit secondary antibody was used at 1:10,000.

Isolation of suppressors of ftsI23 mutant using random transposon insertion mutagenesis

To isolate suppressors of ftsI23 mutant, we performed random insertion mutagenesis using λ1098 containing Tn10dTet transposon as described [31]. Briefly, overnight culture of the MG1655 strain carrying the ftsI23 allele was sub-cultured in LB with 0.4% maltose and 10 mM MgSO4 and grown till the late logarithmic phase at 30°C. Transpositions were done as described [31] and colonies were selected on LB agar supplemented with Tetracycline (Tet; 10 μg/mL) at 32°C. Colonies that grew very well were purified and their suppressive ability was checked by reintroducing the Tn10dTet mutations into the ftsI23 strain using P1 phage-mediated transduction. Once confirmed, the Tet element from the genomic DNA of the suppressor was cloned into a plasmid vector and sequenced using Tet-specific outward primers, P1: 5’-TGGTCACCAACGCTTTTCCCGAG-3’ and P2: 5’-CTGTTGACAAAGGGAATCATAG -3’. By reading the junction sequences, we were able to identify that the insertion was in sanA ORF. The sanA::Tn10dTet insertion mutation exactly behaved like that of ΔsanA::Kan deletion from Keio collection and we subsequently used the Keio deletion mutation for all the experiments reported in this study.

Identification of multicopy plasmid suppressors of SanA⁻ phenotype

We utilized a multicopy plasmid library carrying overlapping E. coli (MG1655) genomic DNA fragments of approximately 3–5 kb in size cloned at the BamH1 site in a p15A-based plasmid, pACYC184 (obtained from Miroslav Radman’s laboratory) to perform genetic screens to identify the suppressors of either ∆sanA or ∆sanA wecA mutants. We transformed the plasmid library into the indicated strains and selected colonies that grew better on NA-Cm plates at 43°C for the ∆sanA mutants or 37°C for the ∆sanA wecA double mutants. Plasmids were isolated from these colonies and their suppressive ability was confirmed by a subsequent round of transformation and phenotype testing. Finally, the plasmids were sequenced using TetA (5′-CGCCGAAACAAGCGCTCATGAGCC-3’) and TetB primers (5′-CTATGCGCACCCGTTCTCGGAGCAC-3’) to identify the region responsible for the suppression.

Statistical analysis

All experiments were conducted at least three times. Error bars in the graphs are depicted as mean ± SD. We employed an unpaired student T-test to compare the significance between the two samples. In all the graphs, *, P < 0.05; **, P < 0.005; ***, P < 0.001 and ****, P < 0.0001.