Constitutive expression of RyhB regulates the heme biosynthesis pathway and increases the 5-aminolevulinic acid accumulation in Escherichia coli

Authors


Abstract

In the current study, the small RNA ryhB, which regulates the metabolism of iron in Escherichia coli, was constitutively expressed in engineered E. coli DALA. The resulting strain E. coli DALRA produced 16% more 5-aminolevulinic acid (ALA) than the parent strain E. coli DALA in batch fermentation. Meanwhile, we found that addition of iron in the medium increased heme formation and reduced ALA yield, whereas the presence of iron chelator in the medium decreased heme concentration and increased the ALA production efficiency (ALA yield per OD600). The qRT-PCR analysis showed that the mRNA levels of hemB and hemH were also decreased as well as the known RyhB target genes of acnAB, sdhAB, fumA, and cydAB in E. coli DALRA. These results indicated that small RNA can be used as a tool for regulating ALA accumulation in E. coli.

Introduction

Heme is an essential iron-containing component of proteins in the electron transport chain that drives aerobic and anaerobic respiration, for example CydAB (Pomka, 1999), and is an important prosthetic group in many sensory regulatory proteins and enzymes. In heme biosynthesis pathway, 5-aminolevulinic acid (ALA) is a key intermediate that determines the accumulation of heme (Fig. 1). Recently, ALA received much attention as a pharmaceutical for cancer therapy and tumour diagnosis (Bhowmick & Girotti, 2010; Mikolajewska et al., 2010; Sakamoto et al., 2010). It can also be used as a kind of selective and biodegradable herbicide, insecticide and growth-promoting factor (Rebeiz et al., 1988; Sasaki et al., 2002).

Figure 1.

Schematic presentation of TCA cycle and heme (including ALA) biosynthesis pathway in Escherichia coli. The genes noted were investigated by qRT-PCR. The dotted arrows present feedback inhibition of heme. The regulation sites (acnAB, sdhAB, fumA and cydAB) and possible regulation sites (hemB, hemH) of RyhB are indicated. Genes: acnA, aconitate hydratase 1; acnB, bifunctional aconitate hydratase 2; sdhAB, subunits of succinate dehydrogenase; gltX, glutamyl-tRNA synthetase; hemA, glutamyl-tRNA reductase; hemL, glutamate-1-semialdehyde aminotransferase; hemB, porphobilinogen synthase; hemH, ferrochelatase. GSA, glutamate-1-semialdehyde; PBG, porphobilinogen.

Two metabolic pathways have been described for ALA biosynthesis (Sasaki et al., 2002). One is the C4 pathway, which exists in mammals, birds, yeasts, some protozoa and purple non-sulphur photosynthetic bacteria such as Rhodobacter sphaeroides. In this pathway, ALA is formed through the condensation of glycine and succinyl-CoA, a C4 intermediate of the tricarboxylic acid (TCA) cycle. The other pathway is the C5 pathway, which utilizes α-ketoglutarate, a C5 intermediate of the TCA cycle, as the carbon skeleton. This pathway is mainly present in higher plants, algae, and many bacteria. In our previous study, we engineered an Escherichia coli strain DALA that can accumulate ALA using glucose as sole carbon source through the C5 pathway (Kang et al., 2011). However, further regulation of this pathway for high ALA production met with many problems because C5 pathway is already highly regulated. For example, HemA and HemL work synergistically, whereas HemA is not stable when heme is present in excess (Wang et al., 1999). Our initial study also indicated that down-regulation of porphobilinogen synthase (HemB), an enzyme that catabolizes ALA in the heme biosynthesis pathway(Jaffe et al., 1995), did not lead to high ALA production as expected, but instead resulted in a decreased accumulation of ALA. Therefore, other strategies will have to be employed to improve the production of ALA.

Recently, a large amount of small non-coding RNA (sRNA) has been identified in bacteria, especially in E. coli, by comparative genomics, bioinformatics and other screening strategies (Wassarman et al., 2001; Vogel et al., 2003; Kawano et al., 2005). The majority of sRNAs play important roles in a variety of essential physiological processes in vivo. Among these, RyhB, a small RNA of 90 nt in length, is involved in iron consumption under iron-limiting conditions by down-regulating the expression of iron-containing proteins, including the enzymes of the TCA cycle and the aerobic respiratory chain (Massé & Gottesman, 2002; Massé et al., 2005; Semsey et al., 2006). The last step of the heme biosynthesis pathway in E. coli is the insertion of ferrous iron into the protoporphyrin IX, forming heme. As RyhB plays an important role in iron homeostasis and can reduce iron-binding proteins expression under low-iron conditions, it is very likely to down-regulate the heme synthesis pathway if it is overexpressed. Based on this supposition, we overexpressed ryhB in E. coli DALA and investigated its role in ALA biosynthesis.

Material and methods

Bacterial strains and plasmids

The strains, plasmids and primers used in this research are summarized in Supporting Information, Tables S1 and S2. Molecular cloning was done using E. coli DH5α according to the standard protocol. The ryhB gene was cloned from the plasmid pKK102-ryhB (Kang et al., 2012). Only the RNA polymerase binding region of PBAD and the ryhB gene region were amplified to allow constitutive expression of ryhB. The amplified ryhB gene fragment was digested with KpnI and BamHI, and ligated into pCLRA (Kang et al., 2011), which was digested by the same restriction enzymes. The resulting plasmid was named pCLRA-ryhB. The cloning scheme of ryhB and the construction procedure of E. coli DALRA are shown in Fig. 2.

Figure 2.

Cloning scheme of ryhB and construction procedure of Escherichia coli DALRA.

Medium and culture conditions

Luria–Bertani medium (10 g L−1 tryptone, 5 g L−1 yeast extract and 10 g L−1 NaCl, pH 7.2) was used for bacterial cultivation during genetic operation. The modified minimal medium (16 g L−1 (NH4)2SO4, 3 g L−1 KH2PO4, 16 g L−1 Na2HPO4·12H2O, 1 g L−1 MgSO4·7H2O, 0.01 g L−1 MnSO4·7H2O and 2 g L−1 yeast extract, pH 7.0; Kang et al., 2011) was used for cultivation and fermentation. Glucose (20 g L−1) was added initially as carbon source. Ampicillin (100 μg mL−1) and spectinomycin (25 μg mL−1) were added to provide selective pressure for plasmid stability during cultivation. Isopropyl-β-d-thiogalactopyranoside was used at a final concentration of 0.1 mM to induce the expression of plasmid-borne genes whose expression was under the control of the lac promoter.

Bacterial pre-culture was done in modified minimal medium using 2 g L−1 glucose as carbon source for 12 h. The pre-culture was then inoculated at 1% (v/v) volume into 50 mL fresh medium containing 20 g L−1 glucose and cultivated at 37 °C, 250 r.p.m. for 32 h. When necessary, 100 μM FeSO4 or 250 μM 2, 2′-dipyridyl was added after 4 h cultivation. The pH was monitored every 4 h and 5 M NaOH was added to maintain the pH at 7.0. Fermentation samples were taken at 4-h intervals to measure OD600, ALA and heme concentration.

Quantitative real-time PCR

The primers used in quantitative real-time PCR (qRT-PCR) are listed in Table S2. Total cellular mRNA was extracted with RNAeasy Mini Kit (Tiangen) after strains were cultivated in a shake flask for 4 h. The cDNA was obtained by reverse transcription using PrimeScript RT reagent Kit (TaKaRa) and qRT-PCR was carried out using SYBR Premix Ex TaqTM II (TaKaRa) with the LightCycler 480 Real-Time PCR system (Roche). The mRNA level of each genes was measured using three biological repeats and three technical repeats. Gene gapA was used as the control because it was stable and was expressed constantly. The qRT-PCR data were analyzed using the math formula protocol reported previously (Livak & Schmittgen, 2001).

Analytical methods

Optical density was monitored at a wavelength of 600 nm with a spectrophotometer (Shimadzu). The concentration of organic acids and glucose was analyzed by high-performance liquid chromatography (HPLC). Samples were prepared by centrifugation (12 000 g for 5 min at 4 °C) and filtration (0.22 μm syringe filter). The HPLC system (Shimadzu) was equipped with a cation exchange column (HPX-87 H; BioRad Labs) and a differential refractive index detector (Shimadzu RID-10 A). H2SO4 (0.5 mM) was used as the mobile phase at the rate of 0.6 mL min−1. The exchange column was operated at 65 °C. To analyze the concentration of extracellular ALA and intracellular heme, 1 mL of each culture sample was centrifuged (12 000 g for 5 min at 4 °C). The supernatant was used to measure ALA concentration using modified Ehrlich's reagent (Mauzerall & Granick, 1956). The cell pellet was used to analyze heme according to the method described by Sassa (1976). Glutamate was analyzed by a SBA-40C biosensor (developed by Biology Institute of Shandong Academy of Sciences) equipped with an immobilized glutamate oxidase membrane.

Results

Constitutive overexpression of ryhB increased ALA production and decreased heme accumulation

To investigate the effect of sRNA RyhB on the heme biosynthesis pathway, we subcloned ryhB into the plasmid pCLRA downstream of a constitutive promoter, generating plasmid pCLRA-ryhB (Fig. 2). The recombinant E. coli strain harbouring pCLRA-ryhB, named DALRA, was then evaluated with respect to its growth and ALA accumulation by cultivation in modified minimum medium supplemented with 20 g L−1 glucose. As expected, we found that overexpression of ryhB led to increased ALA production from 1.54 to 1.78 g L−1 in batch fermentation, 16% more than that of the control (Fig. 3). However, the accumulation of heme, which is located downstream of ALA in the heme biosynthesis pathway, decreased significantly in E. coli DALRA. The heme accumulation in strain DALRA was only 0.232 nmol OD−1, 56.6% less than that of the control. In addition, the cell growth was also slightly affected. At 32 h, the OD600 of E. coli DALA and DALRA was 13.13 and 13.95, respectively. As several metabolic pathways are affected by RyhB, overexpression of this sRNA shall have overall influence on the host, especially the heme biosynthesis pathway.

Figure 3.

Effect of ryhB overexpression on cell growth, ALA production and heme accumulation. The pre-culture was inoculated at 1% (v/v) into 50 mL medium with 20 g L−1 glucose and cultivated at 37 °C, 250 r.p.m. for 32 h. Fermentation samples were taken at an interval of 4 h for measuring OD600, ALA and heme concentration. Results are the average of three individual experiments.

Overexpression of RyhB changed metabolic flux and down-regulated iron-containing proteins

As previous studies suggested that the key genes in the TCA cycle (acnAB, sdhAB, fumA), all of which encoded iron-containing proteins, were targets of RyhB (Massé & Gottesman, 2002; Massé et al., 2005), we measured the concentration of the metabolic intermediates of E. coli DALA and DALRA (Table 1) and analyzed the transcription of these target genes by qRT-PCR (Fig. 4). When glucose was depleted, E. coli DALRA accumulated 0.237 g L−1 succinate and 0.325 g L−1 glutamate, respectively, whereas DALA accumulated 0.170 g L−1 succinate and 0.148 g L−1 glutamate. There was no significant difference in lactate and acetate accumulation between E. coli DALA and DALRA and no citrate was detected in the fermentation broth of either strain. In E. coli, the transcription of ryhB is activated under iron-limitation conditions, which makes the mRNA of sdhCDAB unstable (Massé & Gottesman, 2002). Artificial overexpression of ryhB probably destabilized the mRNA of sdhCDAB and blocked the sdhCDAB site of the TCA cycle, resulting in the accumulation of succinate and glutamate in E. coli DALRA. To confirm it, we performed a qRT-PCR analysis. Among the genes analyzed in the TCA cycle, sdhB was the most affected (Fig. 4). The relative transcription level of sdhB was down-regulated to only 0.3-fold of the control, which explained why the succinate was accumulated. The overexpression of ryhB also down-regulated the transcription of hemB and hemH genes in the heme biosynthesis pathway, but the key gene gltX, which is involved in the ALA biosynthesis, was up-regulated, which partly explains the improvement in ALA production. The proteins in the aerobic respiration chain which contain heme as essential prosthetic group, such as CydAB, were also down-regulated in E. coli DALRA (Fig. 4). We did a bioinformatics analysis of RyhB with hemB and hemH to determine possible modes of RyhB interaction with these two genes using clustalw2. Sequence alignment showed that part of RyhB could pair with the sequence within hemB and hemH mRNA. This part of the RyhB sequence is very similar to the sequence that participates in the interaction with sdhC (Massé & Gottesman, 2002) (see Fig. S1). Thus, RyhB may interact with hemB and hemH by binding to the specific site of DNA, inhibiting the transcription of these genes.

Table 1. Analysis of metabolic intermediates of Escherichia coli DALA and E. coli DALRA
Metabolic intermediatesConcentration (g L−1) DALAConcentration (g L−1) DALRA
  1. Data shown were measured from the samples taken at the time glucose was depleted. Results are the average and standard deviation of three duplicates.

Glutamate0.148 ± 0.0020.325 ± 0.025
Succinate0.170 ± 0.0060.237 ± 0.001
Lactate0.070 ± 0.0030.071 ± 0.021
Acetate4.968 ± 0.2334.776 ± 0.117
Figure 4.

Effect of ryhB overexpression on relative gene transcription in Escherichia coli DALRA compared with the control E. coli DALA. Gene gapA was selected as standard and the error bars indicate the standard deviation of the mean of three replicates.

Influence of iron and 2, 2′-dipyridyl on the production of ALA and heme accumulation

As reported previously, the transcription of ryhB was repressed when iron was present and induced when iron chelator was present (Massé & Gottesman, 2002). Thus, the addition of iron chelator appears to facilitate the ALA production, whereas addition of extra iron represses ryhB more. To measure the effect of iron and iron chelator on ALA production and heme accumulation, we cultured E. coli DALA and DALRA in the presence of 100 μM FeSO4 or 250 μM 2, 2′-dipyridyl. When FeSO4 was added, only about 0.9 g L−1 ALA was produced, but the accumulation of heme increased to about 0.55 nmol OD−1 in both E. coli DALA and E. coli DALRA (Fig. 5a). The addition of FeSO4 in the culture may switch the in vivo metabolic flux from ALA to heme formation by chelating ferrous ion with protoporphyrin IX, resulting in decreased ALA accumulation and increased heme accumulation. This implies that the effect of ryhB overexpression on heme biosynthesis and ALA accumulation is blocked by extra iron. Treatment of E. coli DALRA culture with 2, 2′-dipyridyl only led to 1.5 g L−1 ALA production (Fig. 5a). However, we found that the cell growth was obviously affected in the 2, 2′-dipyridyl treatment group due to the harmfulness of 2, 2′-dipyridyl (Fig. 5b). The OD value of E. coli DALA and DALRA of this group was 10.35 and 9.68, respectively. Calculation of the ALA production per cell density in E. coli DALRA was 0.172 g L−1 OD−1, 25% percent higher than in the control (Fig. 5c). In the presence of 2, 2′-dipyridyl, protoporphyrin IX may compete with the iron, reducing the actual concentration of iron for heme formation.

Figure 5.

Effect of iron and chelator on heme accumulation, ALA production and cell growth with ryhB overexpression. (a) ALA production and heme accumulation in FeSO4 or 2, 2′-dipyridyl-treated Escherichia coli DALA and DALRA. (b) OD value of FeSO4 or 2, 2′-dipyridyl-treated E. coli DALA and DALRA. (c) ALA concentration per OD in 2, 2′-dipyridyl-treated E. coli DALA and DALRA and control. Cultivation was carried out in a 300-mL Erlenmeyer flask, containing 50 mL modified minimal medium, at 37 °C, 250 r.p.m. for 32 h. Fermentation samples were taken at 4-h intervals to measure OD600, ALA and heme concentration. Results are the average of three individual experiments.

Discussion

The application of microRNA (micRNA) and short interfering RNA (siRNA) has increased greatly in eukaryotic cells in the last two decades due to the development of gene silence technology (Ashrafi et al., 2003; Lum et al., 2003). Nonetheless, the technology remains to be developed/optimized for bacteria in which gene knockout technology is already well established. Gene silence technology does have a potential application in necessary genes that cannot be knocked out, and the technology should be fine-tuned. Recently, inspired by the modular architecture observation of the natural sRNA genes in bacteria (Bouvier et al., 2008; Papenfort et al., 2010), researchers designed artificial synthetic sRNA genes to regulate bacteria gene expression (Man et al., 2011; Sharma et al., 2011; Na et al., 2013). Together with natural sRNA gene, artificial synthetic sRNA gene has become an effective and convenient tool for fine-tuning the expression of specific genes.

In this study, we found that the constitutive overexpression of natural small RNA RyhB influenced the transcription of iron-containing enzymes and drove the metabolic flux to ALA production. Interestingly, overexpression of ryhB down-regulated the transcription of hemB and hemH. The presence of possible targeting sites in the mRNA of hemB and hemH indicated that they could be targets of RyhB. The down-regulation of hemB and hemH by ryhB reduced the metabolic flux from ALA to heme, resulting in increased ALA production and decreased heme accumulation. Although the mechanism of the regulation of hemB and hemH by RyhB was not confirmed, there are at least three reasons for the increased ALA accumulation in E. coli DALRA: (1) RyhB up-regulated gltX expression involved in ALA biosynthesis; (2) RyhB down-regulated the gene transcription involved in ALA metabolism; (3) the smaller amount of heme accumulation caused de-repression of the feedback inhibition of ALA synthesis genes (Levicán et al., 2007; Jones & Elliott, 2010). The down-regulation of the heme-containing protein CydAB in E. coli DALRA was probably due to less heme provision in vivo. A previous study showed that cydAB was directly regulated by RyhB (Massé et al., 2005). It is possible that RyhB repressed the heme biosynthesis pathway during iron-limiting conditions and released free iron for other essential biological processes. Nevertheless, the heme accumulation in both E. coli DALA and DALRA was increased compared with the wild-type E. coli strain. As the anabolic heme biosynthesis is coupled to catabolic electron chain-driven ATP synthesis, the regulation of the heme biosynthesis pathway shall facilitate the overall metabolism of the host (Möbius et al., 2010).

In E. coli, the biosynthesis of ALA and heme is highly regulated. To promote ALA production, a strategy that can fine-tune the gene expression of this pathway is necessary. Application of native sRNA gene in this pathway is an attractive strategy and can be further exploited.

Acknowledgements

This work was financially supported by a research grant from National High-Tech Research and Development Plan of China (2012AA022104) and a grant from the National Natural Science Foundation of China (31370085) and Independent Innovation Foundation of Shandong University (IIFSDU) (2012ZD029). The authors do not have any conflicts of interest.

Ancillary