Bacterial strains and plasmids used are listed in Table 1. Plasmids pNLI60, pNLI115, and pNLI116 were synthesized by GenScript. E. coli strains DH5α and BL21De3 were hosts for plasmid isolation and protein expression, respectively. For plasmid introduction, E. Coli strains were chemically transformed according to (Mandel and Higa, 1970). E. coli strains were cultured in lysogeny broth (LB) (Bertani, 1951). LB was prepared from 5 g bacto tryptone (Difco), 5 g NaCl (Fisher) 2.5 g of yeast extract (Difco) in 1 L H₂O, sterilized for 20 min at 121°C, and supplemented with appropriate antibiotics and 1.5% agar, as needed. Ampicillin was used at 100 μg/mL, for growth of E. coli strains. Antibiotics were purchased from Sigma-Aldrich.

Pneumococcal ComWΔ6 was expressed from pNLI60 and purified according to (Inniss et al., 2019). Pneumococcal σ^X and σ^A were expressed from pNLI94 and pNLI114, respectively. E. coli transformed with pNLI94 (SigX-V5H6, pI ∼8.42) or with pNLI114 (σ^A-V5H6, pI ∼5.02) were cultured in 3 L of LB medium at 37°C, 200 rpm to an OD₆₀₀ 0.5. To prepare a sample for SDS-PAGE, 2 mL of uninduced cells were collected and pelleted for 10 min at room temperature, and then boiled at 95°C for 2 min in a 3:1 mixture of E. coli resuspension buffer (50 mM Tris-HCl, pH 8.0, 500 mM NaCl) and Laemmli’s buffer (Bio Rad), respectively, then stored at 4°C. The large culture was induced for protein expression by addition of [1 mM]_f isopropyl β-D-1-thiogalactopyranoside (IPTG) (Gold Biotechnology). Induced cultures were incubated overnight at 20°C, 200 rpm. The next day, 2 mL of induced culture was prepped for analysis by SDS PAGE, and the large cultured collected at 5,000 rpm, 4°C, for 30 min. Cells were resuspended in 50 mL of E. coli resuspension buffer supplemented with 25 mM imidazole, 50 mM MgCl₂, 100 μg/mL DNase, and a protease inhibitor tablet (Roche). The cell resuspension was lysed using an Emulsiflex-C3 (Avestin). The soluble fraction was separated by centrifugation at 30,000 × g for 30 min at 4°C.

For σ^X purification, the soluble fraction was discarded. The pellet was washed in 15 mL of cold, filtered 20 mM L-arginine, 1 mM EDTA, 5% glycerol (RGE buffer, final pH, 10.95) supplemented with 1 mM βME and 2% w/v sodium deoxycholate (NaDOC) (Sigma Aldrich) with a homogenizer, followed by sonication for 2 min at 65% amplitude (0.8/0.2, on/off cycles). The washed pellet was spun was 30,000 × g for 30 min at 4°C. Washing, sonication, and spinning were repeated. Then the pellet was resuspended and dissolved in 15 mL of RGE supplemented with 1 mM βME and 0.6% N-lauroyl sarcosine sodium salt (Sigma-Aldrich) at 4°C with slow stirring for up to 2 days. The dissolved inclusion body was collected by centrifugation at 30,000 × g for 30 min at 4°C. The supernatant was saved and dialyzed for 8 h at room temperature in 2 L of 25 mM Tris-HCl, pH 8.3, 500 mM NaCl, 30 mM L-arginine, 10% glycerol, and 5 mM βME (CB4), and then overnight at 4°C in 2 L of CB4. The soluble supernatant was passed through a 5 mL Ni-NTA resin (Thermo Fisher Scientific) on a column equilibrated with CB4 (no βME) via gravity flow. The column was washed with 500 mL of CB4 supplemented with 5 mM βME and 40 mM imidazole. SigX-V5H6 was eluted from the column with 10 mL of CB4 supplemented with 5 mM βME and 275 mM imidazole. Subsequently, the flow through was saved and concentrated using a Vivaspin 15R 2,000 MWCO column concentrator (Sartorius). The concentrated solution was applied to a pre-packed Superdex 200 column (GE) on an AKTA (Amersham Biosciences). SigX-V5H6 was eluted at 0.3 mL/min in CB4 supplemented with 1 mM βME and collected in 1 ml fractions for analysis on SDS-PAGE gels. The fractions were pooled and concentrated again using a VivaSpin column concentrator. Protein concentration was measured at 5.6 mg/mL using a Nanodrop. Purified SigX-V5H6 was diluted to 1.2 mg/mL in CB4 supplemented with 1 mM EDTA, 1 mM βME and 20% glycerol for storage at −20°C until use.

For purification of σ^A, the soluble fraction was separated by centrifugation at 30,000 × g for 30 min at 4°C. The soluble supernatant was passed through a 5 mL Ni-NTA resin (Thermo Fisher Scientific) on a column equilibrated with Column Buffer (50 mM Tris-HCl, pH 8.0, 500 mM NaCl, 10% glycerol) via gravity. The column was washed with 500 mL of Column Buffer supplemented with 5 mM βME and 40 mM imidazole. σ^A was eluted from the column with 10 mL of column buffer supplemented with 5 mM βME and 275 mM imidazole. The eluate was concentrated using a Vivaspin 15R concentrator column into 50 mM Tris-HCl pH 8.0, 20 mM L-arginine, 500 mM NaCl, 1 mM EDTA, 10% glycerol, and 1 mM βME (DBR buffer). Protein concentration was measured as 1.85 mg/mL using a Nanodrop. SigA-V5H6 was diluted to 1 mg/mL in DBR supplemented with 20% glycerol and stored at −20°C until use.

Primers used in this study are listed in Table 1. Gene sequences for amiA, comEA, and ssbB were taken from the genome of S. pneumoniae R6 (NC_003098.1). Oligonucleotides were synthesized by IDT (Coralville, Iowa). Equal amounts (1 μM) of forward and reverse primers were used to amplify amiA (NL299/NL300), comEA (NL295/NL296), and ssbB (NL228/NL229) from 50 ng of CP2137 genomic DNA with 0.2 mM dNTP (Thermo Fisher Scientific) and 1 μL of Phire Hot Start DNA Polymerase II (Thermo Fisher Scientific) in 50 μL reactions in a thermocycler. PCR products were purified with Zymo kits, eluted in diethyl pyrocarbonate (DEPC) H₂O and visualized on a 1% agarose gel prior to storage at −20°C.

In vitro transcription from linear templates was carried out using 100 nM of E. coli RNAP obtained from NEB, pneumococcal σ^X and σ^A purified from E. coli, and ComWΔ6 purified from E. coli, when needed. To determine the minimum concentration of each σ-factor required to stimulate RNAP, initiation complexes were formed with 100 nM of E. coli RNAP, 20 nM of ssbB or amiA, and 0–200 nM of σ^X or σ^A, in 50 mM Tris-acetate, 50 mM sodium-acetate, 12.5% glycerol, 0.25 mM EDTA, 15 mM magnesium-acetate, and 35 mM βME (4X TGA buffer). Transcription from plasmid templates was carried out using 0–400 nM of E. coli RNAP and 0–320 nM of σ^A or σ^X. Solutions were incubated at 37°C for 15 min in a thermocycler, followed by addition of 1 μL of a master mix containing (120 nM) ATP, CTP, and GTP (Thermo Fisher Scientific) with ^{α P32}UTP at 0.5 μCi/μL (3000 ci/mmol, Perkin Elmer). Runoff transcription was allowed to proceed for 6 min at 37°C in a thermocycler, followed by addition of 20 nM cold UTP (0.5 μL of a 200 nM stock) and a final elongation step at 37°C for 10 min in a final volume of 5.5 μL. Reactions were stopped by addition of one volume of formamide dye (80% formamide, 10 mM EDTA pH 8.0, 1 μg/mL xylene cyonal, and 1 μg/mL bromophenol blue) and incubation at 65°C for 2 min in a thermocycler. The samples were placed on ice until gel loading.

To measure transcription stimulation by ComWΔ6 on PCR templates, 100 nM of E. coli RNAP, 80 nM σ^X or σ^A, 20 nM of comEA, ssbB, or amiA, and 0–160 nM of ComWΔ6 were mixed in 50 mM Tris-acetate, 50 mM sodium-acetate, 12.5% glycerol, 0.25 mM EDTA, 15 mM magnesium-acetate, and 35 mM βME (4X TGA buffer). Solutions were incubated at 37°C for 15 min in a thermocycler, followed by addition of 1 μL of a master mix containing (120 nM) ATP, CTP, and GTP (Thermo Fisher Scientific) with ^{α P32}UTP at 0.5 μCi/μL (3000 ci/mmol, Perkin Elmer). Runoff transcription was allowed to proceed for 6 min at 37°C in a thermocycler, followed by addition of 20 nM cold UTP (0.5 μL of a 200 nM stock) and a final elongation step at 37°C for 10 min in a final volume of 5.5 μL. Reactions were stopped by addition of one volume of formamide dye (80% formamide, 10 mM EDTA pH 8.0, 1 μg/mL xylene cyonal, and 1 μg/mL bromophenol blue) and incubation at 65°C for 2 min in a thermocycler. The samples were placed on ice until gel loading.

To visualize mRNA transcripts, 2 μL of each reaction was run on a 20 × 20, 0.375 mm thick, 6% polyacrylamide urea gel at 20 watts for 45 min. Gels were dried for 40 min at 80°C and incubated on a phosphorimager and imaged on a Typhoon 3000 (GE). The expected mRNA transcript sizes were determined using the transcriptional start sites (TSS) for each gene as determined by Aprianto et al. (2018). Experimental transcripts were compared to markers generated by in vitro transcription using the MAXIscript T7 Transcription Kit (Invirtogen) and the RNA Century-Plus Marker Templates (Invitrogen).

In vitro transcription assays confirmed that σ^X, in molar excess to RNAP, promotes transcription from combox promoters (Luo et al., 2003). However σ^X-mediated late gene transcription and transformation requires ComW in vivo (Sung and Morrison, 2005). Previous work concluded that pneumococcal RNAP exists at ∼2000 molecules per cell, and that σ^X and ComW peak at ∼3000 and ∼500 molecules per cell, respectively (Luo and Morrison, 2003; Piotrowski, 2010). Although not confirmed, σ^X might have a lower affinity for RNAP than σ^A, a trait shared among multiple alternative σ-factors (Gruber and Gross, 2003), so excess σ^X could increase the amount of σ^X holo-enzyme formation. Interestingly, levels of σ^A decrease 20–40% during the pneumococcal competence response (Luo and Morrison, 2003; Luo et al., 2003), and mutations to σ^A, that presumably decrease its affinity for RNAP, can suppress ΔcomW phenotypes (Tovpeko and Morrison, 2014; Tovpeko et al., 2016). These data hint that σ^X access to RNAP and/or competence promoters is an important step in efficient competence gene expression. In addition, we have shown that ComW, a likely σ^X binding partner (Tovpeko et al., 2016), interacts with DNA and this interaction is important for transformation (Inniss et al., 2019). We hypothesize that ComW functions at or near competence promoters to boost σ^X transcriptional activity. We tested this theory with in vitro transcription (IVT) assays.

Luo and Morrison used a ∼1.7 molar excess of σ^X to pneumococcal RNAP (RNAP_Sp) in their IVT assays, making the number of molecules of σ^X higher than that of RNAP_Sp (roughly 5.12 × 10¹² vs. 3.13 × 10¹² molecules, respectively, and based on the average molar ratio of σ^X:RNAP_Sp in cells as determined by Luo and Morrison, 2003). Thus, IVT from every competence template tested was successful and did not require ComW. As RNAP is conserved in prokaryotes (Murakami and Darst, 2003), we titrated pneumococcal σ^X and σ^A to determine the minimal amount of each σ-factor required to target 100 nM of E. coli RNAP (RNAP_Ecoli) to the pneumococcal combox and housekeeping promoters, in vitro.

The promoter regions of pneumococcal ssbB and amiA were PCR amplified for use in in vitro transcription assays (Figure 1A). At 40 nM σ^X and 100 nM RNAP_Ecoli, a 121-base (b) mRNA product from 20 nM of an ssbB linear template was produced, although the amount of mRNA produced (Figure 1B). The amount of ssbB transcript produced increased as σ^X was titrated up to 160 nM. At 80 nM, there are approximately 1.93 × 10¹¹ molecules of σ^X versus 2.41 × 10¹¹ molecules of RNAP_Ecoli at 100 nM in the reaction. This molar ratio (80 nM σ^X: 100 nM RNAP_Ecoli) allowed for observable, yet sub-optimal transcription from 20 nM of linear ssbB DNA template containing a combox promoter, without ComW. In addition, 80 nM of σ^X + 100 nM RNAP_Ecoli did not produce an mRNA product when incubated with 20 nM of linear amiA template, a gene under control of the primary σ-factor, σ^A (not shown), demonstrating that σ^X transcription specifically initiates from the combox promoter.

In contrast at 20 nM, pneumococcal σ^A produced detectable levels of a 347-base mRNA product from the housekeeping promoter upstream of amiA when incubated with RNAP_Ecoli (Figure 1B). The amount of amiA transcript produced increased as σ^A was titrated up to 160 nM, with a detectable amiA transcript at 80 nM. Interestingly, we did observe production of an ssbB transcript with 80 nM of σ^A (not shown). This might result from an orphan -10 promoter region σ^A binding site that is present 29 bp downstream of the ssbB transcription start site, according to Aprianto et al. (2018). The nature of this mRNA product must be examined more thoroughly. Nonetheless, these results demonstrate that both pneumococcal σ-factors direct RNAP_Ecoli to pneumococcal promoters, in vitro. In addition, as it appeared that more σ^X was required to stimulate RNAP_Ecoli activity at combox promoters than σ^A at housekeeping promoters, these results might indicate that more σ^X is needed to stimulate RNAP_Sp. As 80 nM of σ^X and σ^A are below saturating concentrations but still produced visible ssbB and amiA transcripts, respectively, we used 80 nM of each σ-factor in subsequent reactions.

To further explore in vitro transcription with pneumococcal σ factor, we tested transcription from plasmid templates. Supercoiling can affect bacterial gene transcription (Pruss and Drlica, 1989; Zhi et al., 2017), thus we asked whether or not our chimeric holo-enzymes could transcribe the ssbB and amiA genes from these plasmid templates. We designed two plasmids that carry pneumococcal promoters upstream of pneumococcal gene fragments, pNLI115 and pNLI116 (Figure 1C). pNLI116 contained only the combox promoter followed by an ssbB gene fragment, similar to our original PCR ssbB IVT template. pNLI115 contained the combox promoter region upstream of competence gene fragment comEA, and the housekeeping promoter region upstream of amiA, to serve as an internal control for promoter specific IVT from plasmid assays. After isolation from E. coli, each plasmid preparation contained predominantly supercoiled species, as observed on DNA agarose gels.

We titrated holo-enzymes into IVT reactions with 10 nM of either PCR DNA templates or plasmid templates. As expected, the amount of the 121 b mRNA transcript produced from linear DNA increased as the amount of σ^X holo-enzyme increased. Similar results were observed from pNLI116, as increasing amounts of a 187 b mRNA transcript was produced as the amount of σ^X holo-enzyme was titrated onto the plasmid (Figure 1D, left). In addition, a 115 b mRNA transcript was produced from pNLI115 when σ^A-RNAP_{E. coli} complex was titrated onto the plasmid (Figure 1D, right). Together these data show that our chimeric holo-enzymes can be targeted to specific promoters on either linear or plasmid DNA templates, in vitro.

Both ssbB and comEA are induced by CSP and require ComW for robust transcription (Sung and Morrison, 2005). To determine the effect of ComW on transcription from combox promoters upstream of these two genes, we titrated ComWΔ6 into IVT reactions with σ^X holo-enzyme. Transcripts from both linear ssbB and comEA templates (Figure 2) were produced when ComWΔ6 was added to IVT reactions. Specifically, the σ^X holo-enzyme incubated with ComWΔ6 at 160 nM (2X molar excess of ComW to σ^X) produced 2X as much mRNA signal from the sbbB template compared to reactions with 0 nM ComWΔ6, an increase that was statistically significant (Figure 2). Similarly, the σ^X holo-enzyme incubated with ComWΔ6 at 160 nM produced 4.5X as much mRNA signal from the comEA template compared to σ^X holo-enzyme with no ComWΔ6, also an increase that was statistically significantly (Figure 2). The 4.5X increase in comEA mRNA signal production versus the 2X increase in ssbB signal production also represented a statistically significant difference. These data demonstrate that ComW is required for robust transcription when the number of σ^X molecules is below the number of core RNAP molecules, in vitro. Furthermore, the difference in transcription stimulation suggests that the requirement for ComW at combox promoters differs. Lastly, as ComWΔ6 can stimulate σ^X-mediate transcription, in vitro, it is possible that residues ⁷³RGFISC⁷⁸ are dispensable for transcription activation in pneumococci.