Introduction

The RNase III family of enzymes is one of the key ribonuclease families that determine mRNA stability in both prokaryotes and eukaryotes. RNase III is encoded by the rnc gene in Escherichia coli and was initially characterized as playing key role in rRNA processing [3]. A recent study that utilized genome-wide analysis of E. coli transcripts showed that the abundance of a few hundred mRNA species could be regulated by RNase III [11], indicating its active role in mRNA degradation and processing in E. coli.

Among the potential RNase III targets identified by genome-wide analysis, steady-state levels of mltD mRNA were most greatly affected by changes in the cellular concentration of RNase III [11]. The mltD gene encodes for a membrane-bound lytic murein transglycosylase D (mltD), which was initially discovered by comparative analysis on the basis of the structure of the 70 kDa soluble lytic transglycosylase (Slt70) protein [6, 13]. Lytic transglycosylases are a large family of the peptidoglycan degrading enzymes that include glucosaminidases, muramidases, and amidases [2, 6, 14]. While other enzymes that belong to the lytic transglycosylase family have been relatively well studied, the physiological role and regulation of the mltD gene have not been characterized. In this study, we investigated the functional role of RNase III activity on mltD expression in E. coli.

Materials and Methods

Strains and Plasmids

The E. coli strain MG1655 rnc-14::ΔTn10 was constructed by P1 transduction of rnc-14::ΔTn10 allele from the E. coli strain HT115 [12]. MG1655mltD was constructed by deleting the mltD open reading frame in the MG1655 genomic DNA using a procedure described by Datsenko and Wanner [4]. PCR primers used in these experiments were mltD-5′-UTR-P1 (5′-ATCGGTGCCTTTTTATTATCTGGTTTGTCAGTGTAGGCTGGAGCTGCTTC-3′) and mltD-3′-UTR-P2 (5′-GTCTTTTAAGCAACTATTGACACACACATGCATATGAATATCCTCCTTA-3′). To construct pCAT924-mltD, a DNA fragment containing the mltD gene was amplified using the PCR primers, mltD pCAT924 F (5′-ATGCGGCCGCGTCTTTTAAGCAACTATTGAC-3′) and mltD pCAT924 R (5′-ATGCGGCCGCTCAGGAATCTGGCATGTTGTT-3′). The products were then cloned into the NotI site in pCAT924 [1], to generate the pCAT924-mltD construct. To construct the pCAT924-mltD-C276U plasmid, DNA fragments containing nucleotide substitution at the cleavage region of mltD mRNA were amplified using the overlap extension PCR method, digested with NotI, and subcloned into the NotI site of the pCAT924-mltD construct. The PCR primers used were mltD-C276U-R (5′-TCTGCCCGTAAAGTTACATCATGGAGATAGCTCTTATTGCG-3′), mltD +277F (5′-GATGTAACTTTACGGGCAGA-3′), mltD pCAT924 F, and mltD pCAT924 R.

Semi-quantitative RT-PCR

Semi-quantitative RT-PCR was performed and analyzed as previously described [15]. Total RNA was isolated using RNeasy miniprep kit (Qiagen) from MG1655 cells harboring pCAT924-mltD or pCAT924-mltD-C276U grown in LB containing ampicillin (100 μg/ml) at 37 °C to OD600 = 0.6. Synthesis of mltD and rnc cDNA was performed using PrimeScript 1st strand cDNA synthesis kit system for RT-PCR (Takara), l μg of total RNA, and mltD and rnc specific primers as per the manufacturer’s instructions. The primers used for RT-PCR were 5′-ATGAAGGCAAAAGCGATATT-3′ and 5′-TCAGGAATCTGGCATGTTGT-3′ for mltD, and 5′-TGTGGATGAAGGCGATATGA-3′ and 5′-CGAGAATTGACTCACGACGA-3′ for rnc.

In vitro Cleavage and Primer Extension Analysis

His-tagged RNase III purification and cleavage assays were performed as previously described [1]. The procedure for synthesis of 3′/5′-end-labeled transcripts was performed as previously described [9]. In vitro cleavage assay was performed and analyzed as previously described [7]. The following primers were used to synthesize transcripts: mltD-T7 (5′-TAATACGACTCACTATAGAAACTTTCTTGTCATCGGCTC-3′) and mltD-R (5′-TCAGGAATCTGGCATGTTGT-3′) for synthesis of full length transcript, and mltD-short T7 (5′-CTTAATACGACTCACTATAGGGAATAAGAGCTATCTCCACGA-3′) and mltD-1/3 R (5′-TGAGGATCAAAAGCGCTCTC-3′) for synthesis of model hairpin transcript. Primer extension analysis was performed using 5′-32P-labeled primer, mltD-426R (5′-CCCCGTGCTCGGAATGATCT-3′).

Analysis of mltD Protein

The procedures for western blot analysis, preparation of membrane fraction of E. coli cells, and nickel extraction was performed as previously described [8, 10].

Results

Identification of RNase III Cleavage Sites in mltD mRNA

Our previous microarray results identified about 100 genes that are downregulated by RNase III [11]. Among these, mltD mRNA levels were most significantly affected by changes in cellular RNase III concentration [11]. This observation prompted us to examine the basis of the correlation between steady-state mltD mRNA levels and RNase III concentrations. First, we tested whether cis-acting elements that are responsive to RNase III are present in mltD mRNA by performing an RNase III cleavage assay using a synthetic mltD transcript. As shown in Fig. 1a, RNase III cleavage reaction with a 5′-32P-end-labeled mltD transcript generated an RNA product of which size is between 354 and 608 nt. We were able to detect only one cleavage product. There are a few explanations for this observation. RNA fragments generated by RNase III cleavage at sites located to the either 5′- or 3′-terminus may have similar sizes, which cannot be resolved in the gel. Another possibility is that RNase III cleavages at both sites occurred rapidly and, consequently, the 5′-end labeled cleavage product generated by RNase III cleavage at the site located to the 5′-terminus of the transcript accumulated. Next, to identify RNase III cleavage sites in mltD mRNA, we performed primer extension experiments using 5′-32P-end-labeled primer (mltD-426R) that were designed to bind to a region downstream of the RNase III cleavage site as deduced from the RNase III cleavage assay shown in Fig. 1a. Total RNA purified from wild-type and rnc-deleted cells that endogenously express mltD mRNA was used for the primer extension experiments. Unfortunately, we were not able to detect any cDNA bands (Fig. 1b). It is likely that cleavage products of RNase III were not detected by primer extension analysis because they were rapidly degraded by other ribonucleases following the RNase III cleavage. This phenomenon has been reported for bdm mRNA [11]. It is also possible that expression levels of mltD mRNA are not high enough to be detected by primer extension analysis. Therefore, we reasoned that we might be able to detect cDNA products from RNase III cleavage products of mltD mRNA when it is overexpressed. For this reason, we cloned the mltD gene into a multicopy number plasmid pCAT924 that expresses the mltD protein with a C-terminal hexahistidine under the control of a constitutive trpc promoter [5]. The resulting plasmid, pCAT924-mltD, was introduced into wild-type and rnc-deleted cells, and total RNA was prepared from these strains for primer extension experiments. We observed two distinct cDNA bands that were only present in the lanes loaded with cDNA products from the reaction containing total RNA prepared from wild-type E. coli cells that overexpressed mltD mRNA (Fig. 1b). These cDNA bands corresponded to sites that were positioned in the double-stranded region of the mltD mRNA coding sequence (Fig. 1c). RNase III cleavage at these sites was predicted to produce products with an overhang of two nucleotides at the 3′-end, which is characteristic of RNase III cleavage products. These sites were designated as cleavage sites A and B.

Fig. 1
figure 1

Identification of RNase III cleavage sites in mltD mRNA in vitro and in vivo. a In vitro cleavage of the full-length mltD RNA. One picomole of 5′-32P-end-labeled mltD transcript was incubated with 2.4 pmol of purified RNase III in a cleavage buffer with (III + Mg2+) or without MgCl2 (III). Samples were withdrawn at the indicated time intervals and separated on 8 % polyacrylamide gels containing 8 M urea. The size of the cleavage product was estimated using size markers generated by internally labeled transcripts. b Primer extension analysis of mltD mRNA synthesized in vivo. Total RNA was prepared from MG1655 and MG1655rnc-14::ΔTn10 that endogenously (total 100 μg) or exogenously (pCAT924-mltD) overexpressed (total 50 μg) mltD mRNA and were hybridized with a 5′-end–labeled primer (mltD-426R). Synthesized cDNA products were analyzed on a 10 % polyacrylamide gel. Sequencing ladders were produced using the same primer used in cDNA synthesis and PCR DNA encompassing the mltD gene as a template. −, no expression; +, endogenous expression; +++, overexpression. c The predicted secondary structure of mltD mRNA. The secondary structure was deduced using the M-fold program [16]. The model hairpin RNA used for in vitro cleavage assays in d, e is shown in the right panel. The position of C276U mutation used in Fig. 2 is indicated. d, e In vitro cleavage of the model mltD hairpin RNA. One picomole of 5′-(D) or 3′-(E) 32P-end-labeled mltD model hairpin were incubated with 0.9 pmol of purified RNase III in a cleavage buffer with (III + Mg2+) or without MgCl2 (III). Samples were withdrawn at the indicated time intervals and separated on 10 % polyacrylamide gels containing 8 M urea. Cleavage products (A and B) were identified using size markers generated by alkaline hydrolysis (Hydrolysis) and predicted secondary structure of the hairpin was confirmed by analyzing the cleavage patterns of the model hairpin RNA after RNase T1 digestion. RNase T1 cleavage sites are indicated in c, d, e with arrows labeled with a–m.The relative amounts of cleaved products A and B are indicated in the parentheses. Other minor cleavage products are indicated with asterisks in d might have been produced from RNase III digestion of RNA transcripts containing an incomplete 3′- or 5′-end

Full size image

mltD mRNA cleavage by RNase III at sites A and B were further demonstrated biochemically using an in vitro synthesized model hairpin RNA (Fig. 1c, right panel) and purified RNase III. The model hairpin RNA has a nucleotide sequence between +258 and +392 nt from the mltD start codon and which contains RNase III cleavage sites A and B in the mltD mRNA. RNase III cleavage of a 5′-32P-end-labeled model hairpin RNA in vitro generated one major and one minor cleavage product, the lengths of which corresponded to cleavage sites A and B, respectively (Fig. 1d). The radioactivity in the cleavage product at site A was ~2.5 times higher than that at site B. The cleavage product at site A appeared to be more abundant since the model hairpin was labeled at the 5′-end and the cleavage product at site A accumulated during the cleavage reaction. The results from RNase III cleavage assays with 3′-32P-end-labeled model hairpin RNA confirmed cleavage sites A and B, and 1.8 times more accumulation of the cleavage product at site B (Fig. 1e), supporting that RNase III does not differentially cleave sites A and B. These results also explain why the amount of cDNA product corresponding to A site is about one-third of the amount of that corresponding to B site in the primer extension analysis of mltD mRNA in vivo (Fig. 1b).

RNase III Cleavage at A and B Sites Regulates mltD Expression

To test whether RNase III cleavage at A and B sites regulates mltD degradation, we introduced the nucleotide substitution, C276U, on the strand facing the cleavage site B in the mltD over-expression plasmid (pCAT924-mltD). The C276U mutation was chosen to avoid possible effects of a nucleotide substitution on other than RNase III cleavage activity. This mutation does not alter either the mltD mRNA’s secondary structure or subsequent amino acid sequence. Wild-type and mutant mltD mRNA was overexpressed in an mltD-null E. coli strain and mltD mRNA steady-state levels and RNase III cleavage specificity were investigated. The steady-state levels of the mutant mltD mRNA were 1.7 times greater than wild-type levels (Fig. 2a), indicating that RNase III cleavage activity on sites A and B influences mltD mRNA degradation. This conclusion was further supported by experimental results showing that RNase III was not able to efficiently cleave an in vitro synthesized model hairpin RNA containing the C276U mutation (Fig. 2b).

Fig. 2
figure 2

Inhibition of RNase III cleavage of mltD mRNA by introduction of a nucleotide substitution at the cleavage site. a Effect of a nucleotide substitution at the RNase III cleavage region on mltD mRNA decay. The plasmid pCAT924-mltD-C276U expresses mltD mRNA containing a nucleotide substitution (C276U) at the RNase III cleavage region. MG1655 harboring either the pCAT924-mltD or pCAT924-mltD-C276U plasmids were grown in LB containing ampicillin (100 μg/ml) at 37 °C to OD600 = 0.6 and total RNA samples were prepared from the cultures. Steady-state levels of mltD mRNA were assessed using semi-quantitative RT-PCR. b Effects of the C276U mutation on RNase III cleavage activity on mltD mRNA in vitro. 5′-32P-end-labeled mltD model hairpin and that containing the C276U mutation were used for in vitro RNase III cleavage assays and analyzed in the same way described in the legend to Fig. 1 d, e. H, Hydrolysis; T1, RNase T1. RNase T1 cleavage sites identified in Fig. 1 c, d, e are indicated with arrows labeled with a–f and i. c Western blot analysis of mltD protein. MG1655 and MG1655rnc-14::ΔTn10 harboring pCAT924-mltD were grown in LB medium at 37 °C. The cultures were taken in the mid-log (OD600 = 0.7) and the stationary (OD600 = 5.7) phase for preparation of cell extract and membrane fraction. mltD protein was isolated by affinity chromatography using Ni2+ column and analyzed by immunoblottin using anti-His antibody. AcrA and S1 proteins were used to provide internal standards to evaluate the amount of membrane fraction and cell extract in each lane, respectively. AcrA protein was also used to validate fractionation of membrane. L and S stand for log and stationary phase, respectively

Full size image

Next, we tested whether RNase III-mediated mltD mRNA degradation consequently affects levels of mltD protein. Wild-type and rnc-deleted strains were transformed with pCAT924-mltD and total protein was obtained from cultures of the resulting transformants in mid-log and early stationary phase. Total protein preparations were separated by SDS-PAGE gels and subjected to western blot analysis using an anti-His-tag antibody. However, we were not able to detect mltD protein in cellular extract despite it being overexpressed (Fig. 2c). Considering that the coding region of mltD mRNA folds into a stem-loop structure that contains the RNase III A and B cleavage sites, we conclude that mltD is not efficiently expressed because it has a weak ribosome binding site (Fig. 1c). For this reason, we affinity-purified and analyzed mltD from membrane fractions of cell extracts using a Ni2+ column. Our results indicate that mltD expression levels are increased by an approximately 1.5-fold in the rnc-deleted strain as compared to the wild-type strain (Fig. 2c). This suggests that mltD mRNA degradation by RNase III significantly contributes to the levels of mltD protein.

Discussion

We investigated the functional role of RNase III in the posttranscriptional regulation of mltD expression in E. coli cells and demonstrated that RNase III cleavage in the coding region of mltD mRNA contributes to the degradation of mltD mRNA, which, consequently, affects mltD protein levels (Figs. 1, 2). RNase III–mediated cleavage of mltD mRNA was inhibited by a nucleotide substitution at the cleavage site in the mltD mRNA, further demonstrating that RNase III cleavage largely contributes to mltD degradation (Fig. 2a, b). Sequence analysis of the mltD gene indicated that it is likely expressed from the sigma factor 70 promoter and that mltD mRNA appears to be inefficiently translated because the 5′-UTR does not contain a strong ribosome binding site (Fig. 1c). This notion is also supported by our western blot analysis of mltD in cells harboring the pCAT924-mltD plasmid (Fig. 2c). Such inefficient translation of mltD mRNA by ribosomes would generate regions of the mRNA that are freely folded into secondary structures, some of which allow for RNase III interaction.

A detailed characterization of the physiological roles of mltD and mechanisms that modulate RNase III activity on mltD mRNA will provide clues as to why mltD expression requires posttranscriptional regulation by RNase III.