Complementation of a temperature sensitive Escherichia coli rpoD mutation using Lactobacillus sigma factors

Abstract

Housekeeping sigma factors in the σ⁷⁰ family, as components of the RNA polymerase holoenzyme, are responsible for regulating transcription of genes related to vegetative growth. While these factors are well understood in model organisms such as Escherichia coli and Bacillus subtilis, little experimental work has focused on the sigma factors in members of the Lactobacillus genus such as Lactobacillus brevis and Lactobacillus plantarum. This study evaluates the ability of putative σ⁷⁰ proteins from L. brevis () and L. plantarum () to complement a temperature sensitive mutation in the E. coli 285c σ⁷⁰ protein. This report is the first to show that these heterologous sigma factors were capable of restoring the viability of E. coli 285c for growth at 40-43.5 °C, indicating the and are capable of initiating transcription in a complex with the E. coli 285c RNA polymerase. These heterologous sigma factors may therefore be useful for improving biochemical knowledge of the sigma factor family or for use in the expression of hetereologous genomic libraries.

Introduction

Two members of the Lactobacillus genus, Lactobacillus brevis and Lactobacillus plantarum, have been shown to possess many desirable characteristics including superior acid, butanol, and ethanol tolerance that are of interest for industrial usage [1–5]. These species have only recently been analyzed [6–9] and as such their biochemical and genetic corpus is lacking compared to the model organism Escherichia coli. In particular, the transcriptional programs of the Lactobacilli and regulation therein under many conditions are poorly understood, though characterization efforts are underway [10–12]. Housekeeping sigma factors (σ⁷⁰ or σ^A) [13] such as those putatively found in L. brevis and L. plantarum genomes, given their role in controlling gene transcription through promoter recognition during exponential growth, play an important role in the physiology of these organisms. Experimental study of the Lactobacilli sigma factors will improve our understanding of how these factors control the Lactobacillus transcriptional program.

The importance of sigma factors arises from their transient contact with the eubacterial RNA polymerase core enzyme (α₂ββ′ω) that allows RNAP to specifically recognize promoter sequences upstream of genes [14–18]. Without these proteins, gene transcription is impossible. Sigma factors are divided into protein families based on structural homology, with factors responsible for regulating gene expression during exponential growth belonging to group 1 [13, 15, 19, 20]; by acting in concert, the various sigma factors within an organism effectively divide the transcriptional program of an organism into multiple sub-routines capable of responding to environmental conditions rapidly [21]. While the understanding of the biological roles of sigma factors found in well characterized organisms is extensive [22–26], the lack of experimental data on L. brevis and L. plantarum physiology means that the putative sigma factors in these organisms are understood only by analogy to the related Bacillus subtilis [27] and through sequence analysis of known promoters. In light of their role in regulating genes critical for growth, the housekeeping sigma factors of L. brevis () and L. plantarum () are the best candidates for in vivo examination to better understand the transcriptional programs of these organisms.

Analysis of the Lactobacilli sigma factors poses significant challenges given the relative paucity of genetic tools available for these organisms. We chose to examine the behavior of and within E. coli as a result, in spite of the significant differences between the organisms. This approach has been successful in the past with non-E. coli housekeeping sigma factors [28–32] and other factors regulating flagellar production (σ^F) and response to extracellular stress (σ^E) [33–35] in complementing E. coli null phenotypes. Sigma factors that are too structurally divergent from those found in E. coli tend to be non-functional in vivo [36]. Given the diversity of heterologous sigma factors that are functional in E. coli, there must exist a significant amount of functional flexibility of σ-RNA polymerase interactions. An improved understanding of the minimal sigma factor sequence required for functioning would improve our understanding of both a basic aspect of bacterial physiology and enhance efforts for sigma factor engineering.

To that end, this work analyzes the functionality of and within an E. coli with a temperature sensitive defect in the native protein [37–40]. We first compare the structure of Lactobacilli sigma factors to other functional, heterologous sigma factors that have been expressed in E. coli in the past to detect any common features amongst these proteins. Two expression plasmids containing the Lactobacilli sigma factors promoter are then constructed and are shown to complement the defective under non-permissive conditions. Potential applications for expression of foreign promoters in metabolic engineering are subsequently discussed.

Experimental Procedures

Bacterial strains and growth conditions

E. coli (BW25113 and 285c), Lactobacillus brevis (ATCC 367), and Lactobacillus plantarum (WCFS1) were used in this study. E. coli cultures were grown in LB broth at 37°C; 50 µg/mL of streptomycin was added as necessary. The E. coli mutant strain 285c with a temperature sensitive defect in its σ⁷⁰ protein and the parent strain P9OA5c [37, 41, 42] were generously provided by Robert Calendar (UC Berkeley) and cultured at 30°C in LB media and 25 µg/ml streptomycin (if needed) to prevent selection of revertant mutations in σ^H [40]. For the complementation assay, the E. coli 285c strain were transformed with empty pCLE vector backbone or plasmids containing L. brevis or L. plantarum σ-factors. Cultures were grown overnight at 30°C in LB media with 25 µg/ml streptomycin. The growth of the P9OA5c, 285c/pCLE, 285c/pCLE-Lp, and 285c/pCLE-Lb strains was monitored spectrophotomerically at 42°C using a TECAN plate reader to assess the complementation effect of the heterologous sigma factors. Batch studies of growth at 43.5-46°C were conducted in a shaking water bath at 200 rpm.

σ-factor selection

Putative σ⁷⁰ growth factors were identified on the basis of sequence annotation (L. plantarum: LP-1926, L. brevis: LVIS-0756) [43] for study. Heterologous sigma factors were placed under the control of P_lac with the native ribosome binding site to ensure equivalent translation efficiency. Primers for these sequences were designed from the E. coli-MG1655, L. brevis ATCC 367, and L. plantarum WCFS1 strains using Primer-BLAST [43] (Table S1). Sequence alignments of E. coli, Lactobacilli, and other sigma factor proteins [28–31, 44] using Clustal Omega [45].

Genetic manipulation

The low copy number pCL1920 plasmid [46] was used to express the σ⁷⁰-factors from L. plan-tarum and L. brevis in E. coli. First, the 300 base pair region upstream of the native E. coli σ⁷⁰ gene containing the native RBS was introduced into pCL1920 to create pCL1920-EcS70 (pCLE). Note that the pCLE plasmid does not contain the gene. Subsequently, each of the σ factor gene was amplified directly from the corresponding genomic DNA and subsequently purified using the Clean and Concentrator kit (Zymo Research). The fragments were digested with the appropriate restriction enzymes and ligated into the digested pCLE plasmid using T4 DNA ligase (New England Biolabs) to create (pCLE-Lb) and (pCLE-Lp). These expression plasmids were then introduced into E. coli (BW25113 and 285c) by electroporation (BioRad GenePulser XL) and plated onto selective media. All cloning PCR steps were performed using the Phusion polymerase (Finnzymes) and the constructs verified by sequencing (MCLAB Inc).

Results

Sequence Conservation in Heterologous Sigma Factors

In order to better understand why the Lactobacilli sigma factors are functional, we compared the sequences of all housekeeping sigma factors that have been expressed in E. coli to this date to identify any conserved structural features (Figure 3). All aligned sigma factors shared at least 44% amino acid similarity, and the L. plantarum and L. brevis factors are 66%-70% similar. Based on sequence alignments, domain 1 and the linker connecting it to domain 2 [13] appear to be entirely absent from the gram positive sigma factors, though there is a high degree of conservation in domains 2 and 4 in all of the heterologously expressed proteins. These sigma factor domains control interactions with RNA polymerase and DNA, and are therefore critical for promoter recognition [13]. Housekeeping sigma factors from Caulobacter crescentus, Rickettsia prowazekii, Microcystis aeruginosa, and Bordetella pertussi were also very similar to E. coli σ⁷⁰ in domains 2 and 4, and all but M. aeruginosa contained a poorly conserved domain 1. Based on the data today, it appears that the only requirement for functional housekeeping sigma factor in E. coli may be sufficient similarity between domains 2 and 4 in the native and heterologous proteins [19].

Lactobacilli sigma factors complement native σ⁷⁰ defects

The first step in evaluating the properties of the heterologous sigma factors is to assess whether the sigma factors from L. brevis and L. plantarum are capable of complementing a temperature sensitive defect in the protein found in the E. coli 285c strain [37], as has been done repeatedly in the past for other heterologous sigma factors [29–31, 44, 47, 48]. First, low-copy plasmids containing the Lactobacillus sigma factors were introduced into E. coli 285c (see Methods and Materials). The growth kinetics of the 285c/pCLE-Lb, 285c/pCLE-Lp, 285c/pCLE, and the parent strains (P9OA5c) as shown in Figure 1 are quite similar at the permissive temperature (30°C), indicating that there is no defect in activity at low temperatures. However, subjecting the strains to the non-permissive temperature (42°C) reveals significant differences between the strains (Figure 2). Little to no growth is seen in the 285c/pCLE strain, while the 285c/pCLE-Lb and 285c/pCLE-Lp strains containing the heterologous sigma factors are not only viable, but indeed grow fairly robustly. The doubling time of the 285c/pCLE-Lb and 285c/pCLE-Lp strains are significantly greater than that of the parent strain (∼255 min versus 108 min for P9OA5c), suggesting that while and can restore viability at elevated temperatures, the growth rate of 285c is still impaired overall due to imperfect complementation or differences in promoter specificity compared to the native σ⁷⁰. These results strongly support the conclusion that the L. brevis and L. plantarum sigma factors can successfully bind to the E. coli RNA polymerase and initiate transcription in vivo effectively, as otherwise the elevated temperature would have permanently arrested growth in all strains.

Figure 1:

Growth of all 285c strains and the parent strain at 30°C. Each strain was inoculated from overnights into LB with 25 µg/ml streptomycin (antibiotic omitted for P9OA5c) and cultured at 30°C to determine the growth kinetics at the permissive temperature. As can be seen, the growth rates and the yields of the strains are similar.

Figure 2:

Complementation challenge at 42°C. Each strain was inoculated from overnights into LB with 25 µg/ml streptomycin (antibiotic omitted for P9OA5c) and cultured at 42°C to determine if the heterologous sigma factors could complement a temperature sensitive defect in the native σ⁷⁰ protein. Robust growth is observed for 285c/pCLE-Lb and 285c/pCLE-Lp while the 285c/pCLE strain is not viable at the elevated temperature.

Figure 3:

Alignments of E. coli σ⁷⁰ with heterologously expressed rpoD homologs from other organisms [27–31, 44]. Conserved residues are shaded in gray or black depending while Alignments were performed with Clustal Omega [45] and visualized using Boxshade.

There are several possible physiological roles for the heterologous sigma factors within E. coli that can be inferred from this complementation effect. Firstly, these proteins may in fact act as direct replacements for factor, enabling the transcription of genes required for exponential growth at high temperatures. This hypothesis is lent support by the fact that the σ⁷⁰ promoter consensus sequences for lactic acid bacteria including those of L. brevis and L. plantarum are close to that of E. coli [49–54]. Efficient transcription of several E. coli promoters by the Lactobacillus acidophilus RNA polymerase holoenzyme has also been previously identified [55], providing additional weight to this interpretation. Suppressor mutations, such as those in rpoH or proteases, increase the stability of in vivo and restore the viability of 285c at elevated temperatures [56]. To test for compensatory mutations that restore growth at elevated temperatures, the ability of the 285c/pCLE-Lb and 285c/pCLE-Lp to grow at 43.5°C was assayed in batch cultures. Both strains grow robustly at 43.5°C (data not shown), indicating that they most likely do not have defects in proteolysis related genes that would stabilize . It is still possible that rpoH suppressor mutations exist in the strains but this phenotypic reversion is extremely rare (frequency ≈ 10^-8) and unlikely to have occurred in independent replicates of 285c/pCLE-Lb and 285c/pCLE-Lp simultaneously. It was also recently demonstrated that expression of confers enhanced transcription of L. planatarum gene fragments [32], providing further evidence that these sigma factors are functional when heterologously expressed. Based on these growth kinetics and small likelihood of undetected compensatory mutations, we conclude that the Lactobacillus σ⁷⁰ proteins are responsible for the observed complementation effect.

Conclusions

We have demonstrated that two primary sigma factors from L. brevis and L. plantarum are able to complement a temperature sensitive growth defect in E. coli 285c that results from loss of the native σ⁷⁰ activity at elevated temperatures. This finding appears to be one of the first demonstrating complementation with a Gram-positive primary sigma factor, and therefore provides additional insight into the plasticity of both the RNA polymerase core enzyme and the sigma factors themselves. Techniques that modify the native E. coli transcriptional program such as global transcription machinery engineering (gTME) may also see a benefit from the use of the Lactobacilli sigma factors given their ability to mediate transcription in complex with RNA polymerase. Engineering E. coli to transcribe L. brevis or L. plantarum genes effectively may also depend on the expression of the sigma factors in situations where compatible promoters cannot be introduced, such as during the protoplast fusion of these organisms.

Acknowledgements

We gratefully acknowledge funding from the National Science Foundation Graduate Student Fellowship program, US NSF grant CBET-1032487, and the Texas Experimental Engineering Station, as well as experimental assistance from Maria Priscilla Almario Falls.