Elongation Factor P Is Important for Sporulation Initiation

ABSTRACT

The universally conserved protein elongation factor P (EF-P) facilitates translation at amino acids that form peptide bonds with low efficiency, particularly polyproline tracts. Despite its wide conservation, it is not essential in most bacteria and its physiological role remains unclear. Here, we show that EF-P affects the process of sporulation initiation in the bacterium Bacillus subtilis. We observe that the lack of EF-P delays expression of sporulation-specific genes. Using ribosome profiling, we observe that expression of spo0A, encoding a transcription factor that functions as the master regulator of sporulation, is lower in a Δefp strain than the wild type. Ectopic expression of Spo0A rescues the sporulation initiation phenotype, indicating that reduced spo0A expression explains the sporulation defect in Δefp cells. Since Spo0A is the earliest sporulation transcription factor, these data suggest that sporulation initiation can be delayed when protein synthesis is impaired.

IMPORTANCE Elongation factor P (EF-P) is a universally conserved translation factor that prevents ribosome stalling at amino acids that form peptide bonds with low efficiency, particularly polyproline tracts. Phenotypes associated with EF-P deletion are pleiotropic, and the mechanistic basis underlying many of these phenotypes is unclear. Here, we show that the absence of EF-P affects the ability of B. subtilis to initiate sporulation by preventing normal expression of Spo0A, the key transcriptional regulator of this process. These data illustrate a mechanism that accounts for the sporulation delay and further suggest that cells are capable of sensing translation stress before committing to sporulation.

INTRODUCTION

Translation of mRNA into protein proceeds in four stages: initiation, elongation, termination, and recycling. During elongation, elongation factor Tu (EF-Tu) delivers the aminoacylated tRNA to the ribosome A site and elongation factor G (EF-G) mediates ribosome translocation (1). Peptide bond formation is catalyzed by the large subunit of the ribosome (2, 3). Elongation proceeds at a rate of approximately 20 codons per second in bacteria (4, 5). However, some sequences are translated faster than others, and the elongation rate can regulate gene expression and affect protein folding (6–13). Polyproline tracts are often potent stalling sequences because proline is a poor donor and acceptor of peptide bond formation (14–19). Elongation factor P (EF-P) alleviates ribosome pausing at consecutive proline residues (17, 20–24).

EF-P is conserved in all domains of life (25). The eukaryotic homolog eIF5A is essential and plays a more general role in translation (26–29). In contrast, EF-P in bacteria is often not essential and loss of EF-P typically has only a modest effect on growth rate (30, 31). However, EF-P can substantially impact other aspects of physiology such as B. subtilis swarming motility (30, 32–35), where EF-P is required for the translation of a component of the flagellar hook (FliY) that contains a polyproline motif (33). Similar to swarming motility, other Δefp phenotypes have been associated with ribosome stalling at a specific motif in a single transcript (e.g., Corynebacterium glutamicum [36]). In general, Δefp mutant strains exhibit reduced translation of proteins with polyproline motifs (21, 22), although stalling is context dependent (37, 38).

Another phenotype of B. subtilis Δefp mutants is a substantial defect in sporulation, ranging from strong (~1% [39]) to severe (<0.00001% [40]). The mechanistic basis underlying this defect is not known. Although several sporulation genes encode proteins containing a PPP motif (41), whether stalling during translation of these proteins is responsible for the Δefp sporulation defect is not known. Sporulation is controlled by a series of transcription factors that function in a defined spatial and temporal fashion (42–44). Here, we show that the loss of EF-P results in a delay in the activation of multiple sporulation-specific sigma factors, consistent with a defect in sporulation initiation. The response regulator Spo0A, in conjunction with the stationary-phase sigma factor σ^H, governs entry into sporulation by directing large-scale changes in gene expression (45). Spo0A activity reflects its phosphorylation state as well as its abundance (46), and we use ribosome profiling to observe that Spo0A expression is attenuated in a Δefp mutant, suggesting that reduced Spo0A abundance is responsible for the sporulation phenotype. Consistently, we find that ectopic Spo0A expression rescues the sporulation defect.

RESULTS

EF-P is important for sporulation during nutrient exhaustion.

efp was identified in a transposon mutagenesis screen as a candidate gene required for efficient sporulation by nutrient exhaustion in Difco sporulation medium (DSM) (39). This result was confirmed by growing B. subtilis wild-type and Δefp strains for 36 h in DSM to determine sporulation efficiency. Wild-type cells sporulated with an efficiency of ~85% in DSM, whereas Δefp cells sporulated with an efficiency of ~15% (Fig. 1A). Doubling times in DSM during exponential phase were similar for the two strains (22 ± 4 min for wild type and 23 ± 4 min for Δefp mutant). Expression of efp from an IPTG (isopropyl-β-d-thiogalactopyranoside)-inducible promoter restored sporulation efficiency to wild-type levels (Fig. 1A), confirming that loss of EF-P was sufficient to explain the phenotype. Δefp cells also exhibited reduced numbers of total viable cells after 36 h of growth in DSM (Fig. 1B).

FIG 1.

Cells lacking EF-P are deficient in sporulation by nutrient exhaustion in DSM. (A) Cells were sporulated for 36 h in DSM or by resuspension in low-nutrient buffered medium. Serial dilutions were plated on LB agar before and after heating to 80°C for 20 min to enumerate total live cells and heat-resistant spores. (B) Total numbers of viable cells before heating are shown. Error bars represent standard deviations of three biological replicates. A two-tailed unpaired t test was used to determine significance. ***, P < 0.001. wt, wild type; Δefp, Δefp mutant; Δefp+efp, complemented strain.

Early markers of sporulation are delayed in the absence of efp.

To facilitate characterization of the sporulation delay, we sporulated cells by resuspension, where cells proceed through sporulation more synchronously (47). As with exponential growth in DSM, doubling times were similar for wild-type and Δefp cells (46 ± 1 min and 50 ± 2 min, respectively). In contrast to sporulation by nutrient exhaustion, we did not observe a measurable difference in sporulation efficiencies after 24 h of sporulation by resuspension (Fig. 1B). However, we did observe that sporulation of the Δefp strain was delayed. We investigated this delay using fluorescent reporters fused to sporulation-specific promoters to monitor progression of resuspended cells in sporulation over the course of 5 h (Fig. 2). Two sporulation genes encode proteins with polyproline motifs—sigE and spoIIP. A Δefp strain carrying a fluorescent reporter fused to a σ^E-dependent promoter (P_spoIID-yfp) exhibited a robust delay in expression compared to the wild type (Fig. 3A). To determine if this delay was restricted to the σ^E-dependent expression, as would be expected given the presence of the PPP motif, we analyzed strains carrying a reporter under the control of the forespore-specific σ^F sigma factor (P_spoIIQ-cfp), whose activation is independent of σ^E. To our surprise, the Δefp strain also exhibited a delay in P_spoIIQ activity (Fig. 3B), indicating that the delay occurred prior to compartment-specific gene activation of sigE.

FIG 2.

Schematic of sporulation reporters. The Spo0A transcription factor is regulated transcriptionally and is activated by phosphorylation. Spo0A~P activates transcription of spoIIE, encoding a phosphatase that activates σ^F in the forespore. This leads to activation of σ^E, the first transcription factor to be activated in the mother cell. Progress in sporulation was assessed using GFP, CFP, and YFP fusions to the spoIIE, spoIIQ, and spoIID promoters, respectively.

FIG 3.

Mother cell- and forespore-specific promoter activity is delayed in a Δefp strain. (A to C) Fluorescent proteins were fused to the promoter of spoIID (A), the forespore-specific promoter, spoIIQ (B), or the mother cell-specific promoter, spoIIE (C). (Top) Representative microscopy images taken at 1 h of resuspension (spoIIE) or 3 h of resuspension (spoIIQ and spoIID). (Middle) Time course of fluorescence intensity of >10,000 cells per promoter fusion. (Bottom) Average number of cells in which the indicated promoter fusion is “on.” A cell was defined as “on” if the fluorescence intensity value exceeded a cutoff of 3 standard deviations above the mean fluorescence value of the cells at time of resuspension. Error bars represent standard deviations of the results of at least three independent experiments. A two-tailed unpaired t test was used to determine significance. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Next, we examined spoIIE expression, since SpoIIE activates σ^F and functions upstream of both σ^E and σ^F, prior to asymmetric division. A reporter expressing green fluorescent protein (GFP) under the control of the spoIIE promoter (P_spoIIE-gfp) exhibited delayed expression in the Δefp strain compared to that in the wild type (Fig. 3C), confirming that the sporulation delay arose at an early stage of sporulation. In addition, following an initial delay, promoter activity increased similarly in the Δefp mutant and wild-type strains. P_spoIIE activity increased at a similar rate in both strains after an initial delay of 1 h (Fig. 3, bottom), and P_spoIID and P_spoIIQ promoter activities also increased similarly following a 2-h delay (Fig. 3, bottom). These data are consistent with the delay occurring early in sporulation and not as a result of EF-P directly affecting expression of compartment-specific sigma factors.

Spo0A expression is reduced in the absence of efp.

A possible source of the sporulation delay is misexpression of a protein involved in the early stage of sporulation. We therefore performed ribosome profiling to characterize the translational landscape and screen for possible differences between the wild-type and Δefp strains. Since the delay was caused by a failure of gene expression early in sporulation, we collected cells at 15 and 60 min after resuspension. Cells were flash frozen and lysed with buffer containing high magnesium to tightly associate the ribosomes to their respective messages (48). Ribosomes were purified, and exposed mRNA was digested. Ribosome-protected fragments were mapped to the B. subtilis genome. Consistent with the observations of reporter gene expression, genes involved in sporulation exhibited reduced expression in the Δefp strain (Fig. 4) (see Table S1 in the supplemental material). Of particular relevance, read density for spo0A, which encodes the transcription factor that functions as the master regulator of sporulation initiation, was substantially lower in the Δefp strain than in the wild type (Fig. 4A). Furthermore, reads mapping to genes under the control of Spo0A (49) were underrepresented in the Δefp strain (e.g., spoIIE, sigF, sigE) (Fig. 4B). Taken together, these data suggest that the sporulation delay resulted from insufficient Spo0A expression.

FIG 4.

The Δefp mutant has reduced Spo0A expression during sporulation. Wild-type and Δefp cells were sporulated by resuspension. At 15 and 60 min after resuspension, the ribosome translation levels of early sporulation genes were determined by ribosome profiling. (Left) Mapping ribosome footprints revealed reduced expression of spo0A in the Δefp mutant 1 h after resuspension. (Right) Translation levels of genes that are positively regulated by Spo0A are similar at 15 min after resuspension and reduced at 60 min after resuspension. Translation levels are plotted as reads per kilobase per million mapped reads (rpkm). Note that Spo0A positively regulates its own expression.

Ectopic expression of Spo0A rescues the Δefp sporulation defect.

Since ribosome profiling revealed reduced Spo0A expression in the efp mutant, we investigated whether the sporulation delay in the Δefp strain was corrected by additional expression of Spo0A via an inducible allele of spo0A integrated at an ectopic locus (50). spo0A induction prior to resuspension partially rescued the delay in sporulation-specific promoter activity, as monitored by a P_spoIIE-gfp reporter (Fig. 5A). And, under conditions of nutrient exhaustion in DSM, spo0A induction resulted in complete rescue of the Δefp sporulation defect (Fig. 5B). Although ectopic expression of Spo0A in DSM completely rescued the sporulation defect, it did not completely rescue cell viability (Fig. 5B), indicating that Δefp cells may also be more sensitive to nutrient exhaustion independent of sporulation.

FIG 5.

Ectopic expression of Spo0A rescues the sporulation defects caused by deletion of efp. Spo0A was expressed from an IPTG-inducible promoter prior to sporulation of cells by resuspension or by nutrient exhaustion. (A) P_spoIIE activity was monitored by fusion of this promoter to GFP. Cells were sporulated by resuspension and monitored by fluorescence microscopy. (B) Sporulation efficiency was determined by plating cells before and after heating to 80°C. The percentage of total viable cells that had formed heat-resistant spores is shown (left) as well as the total viable cell number (right). **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

DISCUSSION

EF-P is conserved in all domains of life (25). Phenotypes associated with efp deletion range from modest to lethal (27, 30, 31, 34, 51). Despite exhibiting normal growth, B. subtilis cells lacking efp failed to sporulate as efficiently as wild-type cells under nutrient exhaustion (Fig. 1), and efp mutant cells sporulated efficiently but with a substantial delay under resuspension conditions (Fig. 3). Ribosome profiling revealed reduced expression of spo0A, which encodes the master regulator of sporulation. Ectopic expression of Spo0A from an IPTG-inducible promoter significantly attenuated both sporulation phenotypes. Thus, our experiments suggest that loss of EF-P leads to reduced Spo0A levels and thereby affects sporulation. Expression of Spo0A from an IPTG-inducible promoter was sufficient to rescue the sporulation delay, suggesting that decreased spo0A transcription is likely the cause of the delay. Since ribosome profiling alone does not indicate whether loss of reads mapping to a particular transcript is due to reduced translation or an absence of the transcript itself, future studies will precisely determine whether transcription or translation of spo0A is defective and will investigate the cause of reduced expression.

EF-P alleviates ribosome stalling at XPPX motifs (21). Spo0A lacks polyproline motifs, so it is likely not a direct target of EF-P. The transcription factor σ^E, which controls early development of the mother cell, is one of the two sporulation proteins containing a polyproline motif. σ^E-dependent promoter activity proceeded normally after an ~2-h delay in the efp mutant (Fig. 3A). A similar delay was observed for a forespore-specific σ^F-dependent promoter (Fig. 3C). In addition to misexpression of σ^E- and σ^F-dependent promoters, the Δefp mutant strain exhibited a delay in P_spoIIE activity. SpoIIE is upstream of both SigF and SigE in the sporulation regulatory pathway (45), so the delay must originate early in the cascade. Importantly, activity from each of these promoters proceeded normally after the initial delay, indicating that the sporulation defect is confined to initiation. These data suggest that EF-P is not required for efficient σ^E translation, even though it has a PPP motif. Consistent with this observation, the presence of a polyproline tract is not sufficient to predict ribosome stalling, and the impact of stalling is dependent on both the expression level of the gene containing the XPPX motif and the sequence context of the motif (21, 33, 37, 38, 52).

Sporulation initiation requires both σ^H-dependent spo0A transcription and phosphorylation of Spo0A by a family of histidine kinases acting indirectly through a phosphorelay (53–55). Entry into sporulation is characterized by a gradual accumulation of Spo0A~P (46), rising to a threshold that enables the activation of specific sets of genes (56) and autoactivation of its own promoter (57), creating a positive feedback loop that causes the level of Spo0A to rise rapidly (58). The rate of the initial accumulation of Spo0A is related inversely to the growth rate, a function of nutrient availability (59). Consistently, B. subtilis cells in nutrient-limited medium can defer sporulation for ~5 cell cycles by attenuating the rate of Spo0A accumulation (60). Thus, by reflecting growth rate, levels of Spo0A couple the decision to sporulate with nutrient availability.

Here, we observe that spo0A expression is attenuated in the absence of EF-P. How might EF-P affect spo0A levels? Escherichia coli Δefp cells have higher levels of uncharged tRNA^Val due to a polyproline tract in the ValS tRNA synthetase (61). The PPP motif in ValS is required for its synthetase activity, and E. coli ValS expression is dependent on EF-P both in vivo and in vitro. While our ribosome profiling data did not indicate a clear ribosome stalling signature within B. subtilis valS during early sporulation, ValS expression levels were low under these conditions, impairing assessment of differential stalling in the Δefp strain compared to the wild type. Higher levels of uncharged tRNAs result in the synthesis of (p)ppGpp, an alarmone that directly inhibits translation initiation (62). Therefore, the absence of EF-P may result in increased ppGpp that could affect Spo0A accumulation either directly by reducing translation initiation (62) or indirectly through an as yet unknown mechanism. A subject for future work is to investigate the possible role of EF-P-dependent ValS translation in the regulation of sporulation initiation through its effects on Spo0A expression.

Similar to the efp mutant, a B. subtilis strain carrying an ileS mutation that allows tRNA^Ile mischarging has a sporulation defect without any observable phenotype during exponential growth (63). Also, similar to what we observe here, this mutation led to a substantial delay in the activation of early sporulation genes. Although the mechanistic basis for this phenotype is not understood, it is intriguing to note that issues relating to proper tRNA charging may be relevant to the phenotypes of both the efp and ileS mutations.

Sporulation is energetically costly and irreversible (64). Therefore, it is presumably advantageous to the bacterium to monitor the availability of sufficient energetic resources before committing to this transformation. As Spo0A is the master regulator of sporulation, its expression and activation reflect diverse input signals, including nutrient limitation, cell density, and intracellular GTP concentration (65). Translation is the most energetically intensive process in bacterial cells (66, 67), and this work reveals that Spo0A expression may also serve as a mechanism to sense the cellular energy status before committing to sporulation.

MATERIALS AND METHODS

Strains and media.

Strains were derived from B. subtilis 168 trpC2 and are listed in Table 1. Genomic DNA from the Δefp::kan strain from the BKK collection (68) was used to transform the wild-type parent strain. Complementation of efp was performed by amplification of efp with primers HAF178 and HAF179 (Table 1) and ligation into pDR111 at the HindIII and NheI restriction sites adjacent to the IPTG-inducible P_hyperspank. The resulting plasmid, pIPTG-efp, was digested with ScaI and transformed into the Δefp strain. Transformants were selected on spectinomycin. The Spo0A-expressing strain was constructed by amplification of spo0A with primers HF227 and HF228 and assembly into pDR111 digested with HindIII and NheI by Gibson assembly (53, 69, 70). The fragment of pDR111 containing P_hyperspank and spo0A was liberated with (resuspension) EcoRI and BamHI and cloned into ECE174 (71) cut with the same restriction enzymes. The resulting plasmid, pIPTG-spo0A, was cut with ScaI and transformed into B. subtilis with selection for double crossover at sacA on LB plates supplemented with chloramphenicol. EF-P or Spo0A was induced with 1 mM IPTG. All strains were grown in LB at 37°C except during sporulation. Where necessary, B. subtilis medium was supplemented with 100 μg/mL spectinomycin, 5 μg/mL kanamycin, and 10 μg/mL chloramphenicol. Fluorescent promoter fusions have been described previously (P_spoIIE [56] and P_spoIID and P_spoIIQ [72]). Deletion of efp from these promoter fusion strains was performed by transformation with genomic DNA from the Δefp::kan strain (BKK24450 [68]) and selection on kanamycin. Similarly, complementation of efp or construction of IPTG-inducible spo0A-expressing strains was performed by transformation with linearized pIPTG-efp or pIPTG-spo0A and selection on spectinomycin or chloramphenicol, respectively.

TABLE 1.

Strains, and primers used in the study

Strain, or primer	Description or sequence	Source or reference
Strains
168 trpC2 (wild-type)	Wild-type strain	73
168 trpC2 Δefp::kan	efp deletion strain	This study
168 trpC2 Δefp::kan amyE::hyperspank-efp	efp complementation strain	This study
168 trpC2 PspoIIE-gfp	spoIIE promoter fused to GFP, backcrossed into 168 trpC2	56
168 trpC2 PspoIIQ-cfp	spoIIQ promoter fused to CFP	72
168 trpC2 PspoIID-cfp	spoIID promoter fused to CFP	72
168 trpC2 PspoIIE-gfp Δefp::kan	spoIIE promoter fused to GFP, Δefp	This study
168 trpC2 PspoIIQ-cfp Δefp::kan	spoIIQ promoter fused to CFP, Δefp	This study
168 trpC2 PspoIID-cfp Δefp::kan	spoIID promoter fused to CFP, Δefp	This study
168 trpC2 PspoIIE-gfp Δefp::kan sacA::hyperspank-spo0A	spoIIE promoter fusion with inducible Spo0A	This study
Primers
HAF178	5′-GGCATAAGCTTAGGAGGACATTAAACATGATTTCAG-3′
HAF179	5′-GGATGCTAGCCTATGCTCTTGAAACGTAAGAACC-3′
HAF227	5′-GTGAGCGGATAACAATTAAGCTTTGGGGAGGAAGAAACGTGGAG-3′
HAF228	5′-GAATTAGCTTGCATGCGGCTAGCGCTCATGTTTAAGAAGCCTTATGCTC-3′
PCR-F	5′-AATGATACGGCGACCACCGAGATCTACAC
PCR-R	5′-CAAGCAGAAGACGGCATACGAGATNNNNNNGTGACTGGAGTTCAGACGTGTGCTCTTCC-3′

Sporulation.

Sporulation was induced either by nutrient exhaustion in DSM or by resuspension (47). For nutrient exhaustion, cultures were grown in DSM for 36 h at 37°C on a roller drum. For sporulation by resuspension, serial dilutions of overnight cultures were grown at 30°C in CH (casein hydrolysate) medium. A culture in mid-exponential phase (optical density at 600 nm [OD₆₀₀], 0.3 to 0.7) was used to inoculate 20 mL of fresh CH medium to an OD₆₀₀ of 0.05 in a 300-mL baffled flask and grown at 37°C. At an OD₆₀₀ between 0.4 and 0.6, cells were pelleted for 5 min at 3,500 relative centrifugal force (rcf) and resuspended in 20 mL of prewarmed A+B (resuspension) medium in a 300-mL baffled flask. Resuspension time courses were monitored by defining the time of resuspension as 0 h. Sporulation efficiency was determined by counting the CFU on LB plates before and after heating to 80°C for 20 min.

Sporulation and cell harvest for ribosome profiling.

Cultures of wild-type or Δefp strains were grown in CH medium overnight at 30°C. An exponential-phase overnight culture was used to inoculate 225 mL of CH medium. Cells were inoculated to a starting OD of 0.025 and grown at 37°C with shaking (220 rpm) in a baffled flask for approximately 3 h to an OD₆₀₀ of 0.4 to 0.6. Cells were pelleted at 3,200 rcf for 10 min and then resuspended in prewarmed resuspension medium according to the method of Sterlini and Mandelstam (47). Culture was harvested at 15 and 60 min after resuspension by flash-freezing the culture in liquid nitrogen and stored at −80°C.

Isolation of RNA footprints.

Ribosome profiling was performed as described previously (48). Frozen culture pellet was lysed in a cryomill with 10× lysis buffer (200 mM Tris [pH 8.0], 1,500 mM MgCl₂, 1 M NH₄Cl, 50 mM CaCl₂, 1% NP40, 4% Triton X-100). Pulverized lysate was thawed and cleared at 4°C, using 4,000 rcf for 10 min. Ribosomes were purified on a 3-mL sucrose cushion (1.1 M sucrose, 20 mM Tris [pH 8], 500 mM NH₄Cl, 10 mM MgCl₂, 0.5 mM EDTA) in a Ti70 rotor spun at 70,000 rpm for 3 h at 4°C. Pellets were washed with 20 mM Tris (pH 8.0), 15 mM MgCl₂, 100 mM NH₄Cl, and 5 mM CaCl₂ and resuspended in 100 μL of the same buffer. Twenty-five absorbance units (AU) of RNA was digested with 1,500 U micrococcal nuclease (Mnase) and mixed gently for 1 h at room temperature. Monosomes were isolated on a 10% to 54% sucrose gradient after centrifugation with an SW-41 rotor at 35,000 rpm for 2.5 h at 4°C. RNA from the fraction containing monosomes was extracted twice with acid phenol and then with chloroform. The aqueous layer was transferred to a new tube. Ammonium acetate was added to 0.3 M, and then an equal volume of cold isopropanol was used to precipitate the RNA. The pellet was dried and resuspended in RNA loading dye (NEB catalog no. B0363A) and loaded on a 15% TBE (Tris-borate-EDTA)-urea gel (Bio-Rad). The gel was stained with SYBR gold (ThermoFisher), and the ~28-bp protected fragment was excised from the gel. RNA was extracted by the crush and soak method in 300 mM sodium acetate (pH 5.5) and 1 mM EDTA. RNA was then precipitated with isopropanol.

Library preparation for sequencing.

RNA was dephosphorylated with T4 polynucleotide kinase (NEB). The universal miRNA cloning linker (NEB catalog no. S1315S) was ligated onto the RNA with T4 RNA ligase (NEB). RNA was then extracted with an oligo clean kit (Zymo), and RNA was eluted with nuclease-free water. Fragments were reverse-transcribed with Superscript III (Invitrogen) and a primer annealing to the ligated cloning linker. RNA template was degraded with 180 mM NaOH. cDNA was purified with an oligo clean kit (Zymo) and loaded on a 10% Criterion TBE gel (Bio-Rad). DNA was extracted by the crush and soak method with elution in 10 mM Tris (pH 8), 300 mM NaCl, and 1 mM EDTA. DNA was precipitated with isopropanol and circularized with CircLigase (Epicentre). Circularized product was amplified for 11 cycles with Phusion polymerase (NEB) and primers PCR-F and PCR-R. PCR product was resolved on a 10% TBE gel, extracted, and precipitated with isopropanol before sequencing on an Illumina HiSeq 2500. Sequencing reads were analyzed with custom Python scripts available at https://github.com/elifesciences-publications/2018_Bacterial_Pipeline_riboseq.

Fluorescence microscopy.

Sporulating cells were immobilized on 1% agarose pads in phosphate-buffered saline (PBS). Fluorescence microscopy was performed using a Nikon Eclipse 90i microscope and a CFI Plan Apo 100×, 1.45-numerical-aperture (NA) oil objective. Images were taken with a Hamamatsu Orca ER-AG camera and yellow fluorescent protein (YFP) (excitation [Ex] 500/20, emission [Em] 535/30), GFP (Ex470/40, Em525/50), or cyan fluorescent protein (CFP) (Ex436/20, Em480/40) filter cubes. Exposure times were 800 ms for YFP and 100 ms for GFP or CFP. Microscope images were processed using ImageJ v1.53k and an original ImageJ macro for automation that identifies bacterial cells, measures the intensity of fluorescence for each cell, and subtracts nonfluorescent background. More than 10,000 cells were analyzed per replicate. Fluorescence density plots were generated in R v4.1.2.

Data availability.

Raw sequencing files can be found in the Gene Expression Omnibus database under accession number GSE222121. Strains and plasmids are available upon request.