Gene order conservation of the T3SS export apparatus

A. Overview of the genetic organization of the genes of the T3SS export apparatus in selected organisms. Genes are color coded according to homology. B. Assembly of the T3SS export apparatus. The export apparatus assembles in the bacterial inner membrane (IM), starting with the SctR pentamer assembly. Subcomplexes with SctT, SctS, SctU, and SctV subunits are sequentially assembled in the indicated order. The export apparatus then serves as a nucleus for downstream T3SS assembly. IM, inner membrane. Panel B is reproduced from Figure 2 from (Pais et al., 2023). C. In silico analysis of the genetic organization of the export apparatus was performed using the Cblaster program (Gilchrist et al., 2021), with export apparatus genes from S. Typhimurium flagella (strain LT2) as query sequences. Among the Cblaster cluster hits, NCBI reference genomes containing all sctRSTU and the homologous fliPQRS genes were included for synteny analysis (n = 1152). Genes separated by more than 500 bp or located on different strands were considered disconnected (indicated by “–”). The taxonomic distribution of export apparatus gene clusters was mapped and visualized as a circular tree: ring 1 indicates the taxonomic class, ring 2 the synteny category, and ring 3 the T3SS type (injectisome vs. flagellum), with all categories color-coded as in the legend. D. Synteny distributions from panel C were quantified separately for injectisome and flagellar export apparatus gene clusters. The panels on the right show part-to-whole graphs for each system.

Plasmid-based complementation of chromosomal deletion mutants of individual export apparatus genes reveals a requirement for sctT expression within the conserved genetic organization of the export apparatus

A. SipA-NL secretion assay. Secretion function was assessed in S. Typhimurium strains with sctR, sctS, sctT or sctU deletions in a SipA-NL background after complementation with low copy number pT10 plasmid derivatives expressing export apparatus subunits. Luminescence values are scaled relative to a wild-type (WT) strain that expresses export apparatus proteins from the chromosome, which was set to a reference value of 100%. Data represent the mean of biological replicates (n ≥ 3), with standard deviations shown as error bars. Each complementation was compared with pT10-sctRSTU within each strain with an unpaired t-test. B. BN- and SDS-PAGE of solubilized crude membrane samples after plasmid complementation. SctRFLAG, SctTFLAG, and SctUFLAG were visualized by immunoblotting of the triple FLAG tag, which is abbreviated as ‘F’. Bands corresponding to export apparatus assembly intermediates and the needle complex (NC) are annotated on the left of the BN-PAGE panels. The intensity of the SDS-PAGE bands was quantified using Image Studio software (Li-Cor, v5.2) and is shown as relative values compared to the samples of wild-type strain transformed with a pT10-sctRSTU plasmid containing the corresponding FLAG tags, which was set to a reference number of 1. Data of SctRFLAG, SctTFLAG, and SctUFLAG are shown in blue, orange, and green, respectively. Data represent the mean of biological replicates (n ≥ 3), with standard deviations shown as error bars. P-values were calculated with one sample t-test, where the sample values were compared to the reference value of 1. ***, P ≤ 0.001, **, P ≤ 0.01, *, P ≤ 0.05. F, 3xFLAG.

Disconnection of the sctS-sctT gene order impairs export apparatus assembly and type III secretion

A. SipA-NL secretion was measured in a S. Typhimurium ΔsctRSTU mutant strain with a SipA-NL background, which was transformed with pT10 plasmid derivatives with scrambled order (sctSTRU, sctTRSU and sctTSRU) or disconnection (sctR_STU and sctRS_TU, where 20 bp spacer is denoted as _) of export apparatus genes, each of which contained the 3xFLAG tag on sctR. P-values of each sample were calculated using an unpaired t-test, comparing them to the wild-type (sctRSTU) value. ALU, arbitrary luminescence unit. B. The same samples used in the secretion assay were analyzed by BN- and SDS-PAGE, which were performed as described in Fig. 2. The intensity of the SDS-PAGE bands was quantified using Image Studio software (Li-Cor, v5.2) and is shown as relative values to the wild-type gene order (sctRSTU) samples, which were set to a reference value of 1. The data represent the mean of biological replicates (n ≥ 2) with the standard deviations shown as error bars. P-values were calculated with one sample t-test, where the sample values were compared to the reference value of 1. ****, P ≤ 0.0001. ***, P ≤ 0.001, **, P ≤ 0.01.

sctS and sctT are translationally coupled through a mRNA structure in sctS

A. Predicted secondary structure of the sctS mRNA (5’-3’) including the first 9 nucleotides of sctT was generated using SPOT-RNA2 (Singh et al., 2021). The Shine-Dalgarno (SD) sequence and start codon of sctT are indicated in orange, and sctS stop codon is indicated in red. Codons in sctS where the ochre stop codons were introduced are indicated on the graph. B. sctS and sctT genes were replaced with mCherry and sfGFP to use fluorescence as a readout for SctS and SctT expression, respectively. The stem-loop sequence of sctS was fused to mCherry (mCherry:sctSS64-G86), to mimic the native translational coupling between sctS and mCherry. Stem-loop destabilizing silent mutations (dest.; sctS L72 CTC::CTG, A82 GCG::GCA) or a 20 bp spacer (purple bar) were introduced to the pT12-mCherry-sfGFP construct as indicated in the top box. In the bottom box, mutations for varying the distance between the sctS stop and sctT start codons are shown, which were introduced to the pT12-mCherry-sfGFP construct. The distance between sctS stop and sctT start codons was reduced by 3 or 6 nucleotides (−3, -6) or increased by 3 (+3). Complementary ‘c’ mutations were further introduced on the other side of the stem-loop for maintaining the stem-loop structure. As a control, only the complementary mutation ‘oc’ was introduced without the distance mutations. C. Fluorescence of mCherry and sfGFP was acquired simultaneously, and the sfGFP/mCherry fluorescence ratio was calculated. In each plot, black lines mark the regression coefficients, and 95% confidence intervals are indicated in red. D. sctT was replaced with the NanoLuc (NL) gene to use luminescence as a readout for SctT expression. The pT12-sctS-NL plasmid was used as the basis for introducing the stop codons indicated in Fig. 4A. E. Luminescence measurements from the constructs with each ochre stop codon are shown in the bar graph. The vertical line separates codons located before and after the stem-loop. Three independent biological replicates were performed. The standard deviations are shown as error bars. ALU, arbitrary luminescence unit.

Translational coupling through the sctS stem-loop and ordered gene organization prevents SctT overexpression and ensure sequential assembly of the export apparatus

A. Schematic diagram showing the translational coupling between mCherry and sctT, established by fusing a remnant of the sctS stem-loop sequence to mCherry (mCherry:sctSS64-G86). B. SipA-NL secretion was measured in the S. Typhimurium ΔsctRSTU mutant strain with a SipA-NL background, which was transformed with pT10 plasmid derivatives harboring either the wild-type gene order or the sctRS-mCh-sctTU construct, with the mCherry-sctT connection as described in Fig. 5A. P-values of each sample were calculated using an unpaired t-test. Three independent biological replicates were performed. The standard deviations are shown as error bars. ALU, arbitrary luminescence unit. ns, not significant.C. Partial structure of the export apparatus (PDB 6F2D) (Kuhlen et al., 2018) showing the SctR-SctS-SctT interface, with the SctR V170 residue highlighted in red. SctR, SctS, and SctT are shown in blue, purple, and orange, respectively. D. In vivo photo-crosslinking of SctR and SctT in S. Typhimurium ΔsctRSTU mutant strain, transformed with pT10 plasmid derivatives harboring either the wild-type gene order or sctRS-mCh-sctTU construct, with and without amber stop codon sites (indicated as X). E. Schematic diagram showing constructs with different sctR-sctT arrangements including the wild-type (WT), ‘directly connected’ (C), and ‘stem-loop’ (SL). In the wild-type, the sctS stem-loop (olive green) contains the sctT Shine-Dalgarno (SD) sequence (red). In the C connection, sctS was deleted and the SD was placed in sctR. In SL connection, sctS4-207 was deleted, but the stem-loop sequence was retained. R* mutation (SctR A223K and T224G) were introduced to the C connection to accommodate sctT’s SD within sctR. R* alone was included as a control for the C connection. F. SipA-NL secretion was measured in S. Typhimurium ΔsctRSTU mutant strain with a SipA-NL background, which was transformed with pT10 plasmid derivatives harboring either the wild-type gene order or C and SL constructs in sctRTSU and sctSRTU gene orders as described in Fig. 5E. The statistical analysis was conducted as described in Fig. 5B. G. BN and SDS-PAGE were performed as described in Fig. 2. The intensity of the SDS-PAGE bands was quantified using Image Studio software (Li-Cor, v5.2) and is shown as relative values compared to the wild-type gene order (sctRSTU) samples, which were set as 1. The data represent the mean of three biological replicates, with the standard deviations shown as error bars. P-values were calculated using one way ANOVA. ****, P ≤ 0.0001. ***, P ≤ 0.001, **, P ≤ 0.01.

Reducing the translation of SctT compensates for the lack of translational coupling regulation through sctS stem-loop

A. S. Typhimurium ΔsctRSTU mutant strain in SipA-NL background was transformed with pT10 plasmids sctTRSU or sctTSRU, with ATG, GTG or TTG start codons on sctT. BN- and SDS-PAGE were performed as described in Fig. 2. The intensity of the SDS-PAGE bands was quantified using Image Studio software (Li-Cor, v5.2) and are shown as relative values to the wild-type gene order (sctRSTU) samples, which were set as a reference value of 1. The data represents the mean of four biological replicates with the standard deviations shown as error bars. P-values of each sample were calculated using one way ANOVA. P-values calculated from comparisons against the wild-type (sctRSTU) are shown above each bar graph. The comparisons between the mutants are shown below the graph with a line with the P-value. B. SipA-NL secretion of the S. Typhimurium transformants. The data represents the mean of ≥3 biological replicates with the standard deviations shown as error bars. P-values calculated from comparisons against the wild-type (sctRSTU) are shown above each bar graph. The comparisons between the mutants are shown with a line with the P-value. ALU, arbitrary luminescence unit. C. The growth curve of S. Typhimurium transformants was measured at OD600 absorbance every 5 minutes for 9 hours. The data represents the mean of three biological replicates with the standard deviations shown as error bars. ****, P ≤ 0.0001. ***, P ≤ 0.001, **, P ≤ 0.01, *, P ≤ 0.05.

Uncontrolled expression results in futile SctT multimerization

A. Partial structure of the export apparatus (PDB 6F2D) (Kuhlen et al., 2018), showing the SctR-SctS-SctT interface (left). SctR, SctS, and SctT are shown in blue, purple, and orange, respectively. A hypothetical SctT-SctT interface is shown on the right, with SctT molecules in shades of orange. SctT F132 is indicated in red. B. Futile SctT-SctT interactions upon SctT overexpression as shown by in vivo photo-crosslinking. S. Typhimurium ΔsctRSTU mutant strains expressing pT10 plasmid derivatives with or without crosslinking residue (SctT F132X, where X denotes an amber stop codon encoding pBpa) were irradiated or not with UV irradiation for 30 min. Crude membrane fractions were subjected to SDS-PAGE and analyzed by Western blotting detecting 3xFLAG tag. Cross-linked interacting partners of SctT are indicated to the right of each blot. Asterisk (*) indicates a SctT-specific band, which may represent a crosslink of a SctT degradation product.

Summary of export apparatus assembly and sctS-sctT translational coupling as a fine-tuning mechanism

The export apparatus genes sctRSTU are transcribed in one transcript, which exhibits a high degree of synteny. The expression of sctT is under tight control involving translational coupling with sctS through the sctS mRNA stem-loop structure. Recognition of sctT SD by a ribosome is inhibited unless the ribosome translating sctS melts the stem-loop. In the absence of control by the stem-loop, SctT is overexpressed and self-assembles into futile multimers. SctT self-assembly under conditions of unregulated SctT expression disrupts the export apparatus assembly and causes cytotoxicity. Control of sctT translation by the sctS stem-loop ensures appropriate levels of SctT and swift complexation of SctT with the SctR pentamer. The SctR5T1 subcomplex subsequently recruits SctS and SctU subunits for finalizing the helical assembly of the export apparatus. Successfully assembled export apparatus then nucleates the further assembly of the needle complex base. SRP: signal recognition particle. SR: SRP receptor. OM, outer membrane. IM, inner membrane. The injectisome illustration was reproduced from Figure 1 from (Pais et al., 2023).

Chromosomally introduced spacer between sctS and sctT results in SctT overexpression and reduced secretion function

A. Equal amounts of crude membrane samples of wild-type and spacer mutant S. Typhimurium strains with a SipA-NL background were loaded on BN- and SDS-PAGE. The intensity of the SDS-PAGE bands was quantified using Image Studio software (Li-Cor, v5.2) and is shown as relative values to the wild-type samples, which were set to a reference value of 1. The data represent the mean of two biological replicates with the standard deviations shown as error bars. B. SipA-NL secretion was measured from the wild-type and spacer mutant S. Typhimurium strains with SipA-NL background. The P-values were calculated with an unpaired t-test, compared to the wild-type. ****, P ≤ 0.0001. ***, P ≤ 0.001, **, P ≤ 0.01.

SctT multimer formation prediction by AlphaFold2

Cryo-EM structure of SctR5T1 complex (PDB 6F2D) (Kuhlen et al., 2018) and AlphaFold2 multimer predictions (Evans et al., 2021) of SctT pentamer and hexamer. In the SctR5S4T1 structure, SctR is shown in cyan, SctS in purple, and SctT in red. The confidence level of AlphaFold multimer predictions is displayed in gradients from blue (low confidence) to red (high confidence). The corresponding pTM and ipTM scores are shown below each prediction.

Conservation of stem-loop structure of sctS across many T3SS-expressing bacteria

Predicted secondary mRNA structures in the intragenic region of sctS and sctT homologs in other pathogens. The sctS stop codon, sctT start codon, and sctT SD sequence are shown in red, green, and blue, respectively. Overlapping sctS stop and sctT start codons are marked with a triangle and the number of overlapping bases. mRNA structure predictions were performed with SPOT-RNA2 (J. Singh et al., 2021). SPI, Salmonella pathogenicity island. SD, Shine-Dalgarno.

Unmodified immunoblots shown in the study.

The western blots were quantified and included in the statistical analyses presented in the corresponding figures. The intensity of the SDS-PAGE bands was quantified using the Image Studio software (Li-Cor, v5.2). Blots highlighted with red dotted boxes are those shown in the main figures.