The members of genus Streptomyces are versatile producers of a vast array of bioactive secondary metabolites including antibiotics, immunosuppressants and anticancer agents such as cephamycin C, clavulanic acid, tunicamycin and holomycin [2]. Secondary metabolite gene clusters are controlled by cluster situated regulators (CSRs) at the lowest level, and by higher-level global/pleiotropic regulators, in a complex cascade system. Thus, CSRs might have different functions as mentioned by Makitrynskyy et al. [3] and could be (i) an ultimate regulator, (ii) an ultimate regulator having cross-talk function [4], (iii) a true pleiotropic regulator [5], or (iv) regulator of a distant gene cluster [6]. In turn, global/pleiotropic regulators exert their effects on the gene clusters by controlling CSRs in response to different stimuli such as environmental signals, physiological conditions and developmental stage [7].

Tunicamycin is a fatty acyl nucleoside-type antibiotic produced by several Streptomyces species including Streptomyces lysosuperificus,Streptomyces chartreusis [8] and Streptomyces clavuligerus [9]. Also, there have been other studies showing different species having tunicamycin gene cluster resembling to that of Streptomyces [10, 11]. It has great potency against early stage of bacterial cell wall synthesis by targeting MraY (translocase I) that catalyzes the formation of peptidoglycan precursor typically referred to as lipid I [12]. In a recent study, it was shown that a marine-derived strain Streptomyces sp. DUT11 produces tunicamycin I, IV and VII showing superior anticomplement activity [13]. Since tunicamycin inhibits eukaryotic protein N-glycosylation its clinical use as antibacterial agent is not feasible yet [14]. Tunicamycin is composed of a uracil, an N-acetylglucosamine (GlcNac), an unusual 11-carbon 2 aminoaldose sugar called tunicamine and an amid-linked fatty acid [15]. The α,β-1″,11′-glycosidic linkage between tunicamine and GlcNac is also unique for this metabolite. The length of N-linked acyl chains is different among tunicamycin variants [16].

Although its structure and function are well known, biosynthetic gene cluster and metabolic pathway studies are limited [16, 17]. A total of 14 genes (tun1(A)-tun14(N)-SCLAV_4287-4274) were identified by genome mining approach coupled with cloning and heterologous expression studies, in S. chartreusis and S. clavuligerus, lacking a CSR within the cluster [18, 19] (Figs. 1a, 1b). Recently, Widdick et al. [20] have reported mutational analysis and transcriptional characterization study of tunicamycin gene cluster in S. chartreusis.

Fig. 1.
figure 1

Biochemical pathway (a) and gene cluster (b) of tunicamycin in S. clavuligerus [12, 14].

Full size image

Since there have been no regulatory genes found in the tunicamycin cluster, a possible regulation at global level needs to be elucidate to get more information for design of new tunicamycins for treatment of diseases and for strain improvement studies. In S. clavuligerus, BldG is an important pleiotropic regulator positively affecting antibiotic production and morphological differentiation [1]. It is used as an anti-anti σ-factor involved in posttranslational regulation and has positive effect on CcaR CSR and other positive modulators such as AdpA acting on antibitoic biosynthesis such as cephamycin C and clavulanic acid in S. clavuligerus [21]. The present study aims at deciphering the putative regulatory role of BldG in tunicamycin biosynthesis in S. clavuligerus.

MATERIALS AND METHODS

Bacterial strains, plasmids, media and growth conditions. The microorganisms and plasmids used in this study are listed in Table 1. S. clavuligerus strains were grown in TSB medium (Thermo Fisher Scientific, USA) at 28°C and 220 rpm to get 24 h old pre-cultures. Five ml of bacterial suspension were used to inoculate TSBY [22] supplemented with 0.5% glycerol [23]. Fermentation experiments were carried out for 4 days under the same conditions, repeated twice and two biological replicates were used. Triplicate samples were taken from the cultures to determine growth and antibiotic production. Cultures of recombinant S. clavuligerus strains were supplemented with apramycin (50 µg/mL) to maintain the plasmids.

Table 1.   Microorganisms and plasmids used or constructed in the study
Full size table

Escherichia coli DH5α and E. coli ET12567/pUZ8002 were grown in LB broth or on LB agar at 37°C with appropriate antibiotics when necessary [ampicillin (100 µg/mL), apramycin (50 µg/mL), chloramphenicol (25 µg/mL), kanamycin (50 µg/mL) and nalidixic acid (25 µg/mL)]. MS agar [24] was used to grow exconjugants following conjugation. The tunicamycin supersensitive strain Bacillus subtilis 6633 was grown in TSB medium at 28°C and 200 rpm to OD600 of 0.9–1.0 prior to bioassay studies.

Construction of plasmids and strains.S. clavuligerus bldG null mutant was kindly provided from Prof. Kapil Tahlan [1]. S. clavuligerus LB1, S. clavuligerus LB2 and S. clavuligerus pSETermE* were constructed in the present study. The bldG gene was isolated by PCR from S. clavuligerus genomic DNA using the primers indicated in Table 2. The bldG gene was subcloned to the pGEM-T® easy vector (Promega, USA) as a fragment of 452 bp to give pLB14 plasmid. The construct was verified by restriction digestion, PCR and nucleotide sequencing. (i) The bldG gene was subcloned to downstream of the promoter of glpF, a gene involved in glycerol transport [25], in pSPG, a pIJ699-derived vector. Both pLB14 and pSPG were digested with NdeI-SpeI enzymes and ligated to generate pLB1, a multicopy vector expressing bldG from the glpF promoter. (ii) The EcoRI fragment carrying bldG was ligated to EcoRI linearized pSET152ermE* to generate pLB2, an integrative single copy plasmid in which bldG is expressed from ermE* constitutive promoter [26].

Table 2.   Primers used in study
Full size table

The constructions were verified by restriction digestion, PCR and sequencing. pLB1, pLB2 and pSET152ermE* were introduced into the methylation-deficient E. coli ET12567/pUZ8002 strain to obtain non-methylated DNA. Intergeneric conjugation between Streptomyces and E. coli was carried out as described by Flett et al. [27]. Exconjugants grown on MS agar were transferred to TSB agar containing apramycin and allowed to grow for up to 4 days at 28°C. The presence of pLB1 and pLB2 in S. clavuligerus exconjugants was confirmed by PCR using the bldG reverse primer and a primer internal to the apramycin resistance gene (Table 2).

Nucleotide sequencing. DNA sequencing was carried out at BGI (Europe) through Genoks (Ankara, Turkey). Deduced nucleotide sequence was compared with the National Center for Biotechnology Information (USA) database using the BLAST search (http://www.ncbi.nlm.nih.gov/BLAST).

Growth determination. The growth of the cultures was determined by DNA quantification according to Burton [28]. DNA concentrations were calculated using herring sperm DNA as standard and expressed as μg of DNA per mL of culture.

Tunicamycin extraction and bioassay for tunicamycin production. For tunicamycin extraction from culture broths a modified procedure adapted from Tsvetanova and Price [29] and Tsvetanova et al. [15] was used. TSBYG cultures were acidified using HCl at a final concentration of 1%. Acid insoluble tunicamycin complex was centrifuged at 4000×g for 10 min and washed in 0.2 M HCl, and this step was repeated twice. The pellets were mixed with methanol and vortexed in ice to dissolve tunicamycin. Following centrifugation (4000×g, 10 min. 4°C), the methanol containing supernatants were evaporated at 40°C, and the tunicamycin extracts were dissolved in appropriate amount of methanol prior to bioassay.

Tunicamycin was quantified using the agar well diffusion method with the use of B. subtilis 6633 as indicator organism [30]. Tunicamycin concentrations were calculated based on a standard curve generated using commercial tunicamycin (Sigma-Aldrich, USA).

RNA isolation and gene expression analyses. RNA was isolated from samples taken from cultures at 36 h of growth. RNA was purified using GeneJET RNA isolation kit (Thermo Fisher Scientific, USA) according to the manufacturer’s instructions with some modifications. (i) lysozyme treatment for cell wall lysis, (ii) phenol/chlorophorm/isoamylalcohol treatment with the use of phase lock gel heavy columns (5 PRIME) and (iii) DNase treatment using DNA-freeTM kit (Thermo Fisher Scientific, USA) at final step. The integrity, purity and amount of the RNA were determined in a NanoDrop® ND-2000 spectrophotometer (Thermo Fisher Scientific, USA).

RNA samples (2 µg) were converted to cDNA using high-capacity cDNA reverse transcription kits (Applied Biosystems, USA). qRT-PCR reactions were carried out on CFX96 real time PCR robotics (BioRad, USA) using GoTaq® qPCR master mix (Promega, USA) as the specific reagent. A standard reaction included the following components: 10 µL 2×GoTaq® qPCR master mix, 0.4 µL 10 µM forward primer, 0.4 µL 10 µM reverse primer, 1 µL cDNA and 8.2 µL dH2O in a final volume of 20 µL. The qRT-PCR reactions were started with an initial denaturation step (2 min at 95°C) and proceded with 40 cycles of amplification (15 s at 95°C and 1 min at 60°C). Melting curve analyses were performed to check for specificity of the amplifications. Two independent qRT-PCR runs were performed. No template control (NTC) containing dH2O instead of cDNA were run to detect background signal and unwanted primer dimer formation. The hrdB-like gene encoding the major σ‑factor in S. coelicolor A3(2) was chosen as reference gene in relative quantification of gene expression [31]. 2-ΔΔCt method was used to analyze the data [32].

Statistical analysis of qRT-PCR data. One way ANOVA with Bonferroni post hoc testing was performed on the normalized gene expression to check whether expression data were statistically different between S. clavuligerus ATCC 27064 (wild type strain) and modified strains. Graphpad Prism Software (USA) was used for statistical evaluation of qRT-PCR data. Error bars represent 95% confidence intervals. Significance was stated as p values (*<0.05; **<0.01; ***<0.001).

RESULTS

The effect of BldG regulator in tunicamycin biosynthesis. In S. clavuligerus ΔbldG mutant, the growth was higher relative to that for wild type strain throughout the fermentation (Fig. 2). The growth of S. clavuligerus pLB1 was almost similar compared to the wild type strain during the time course of fermentation, while its growth was lower at 36 h of growth in comparison to the vector control strain (Fig. 2a). S. clavuligerus pLB2 exerted a faster growth than the wild type especially at 36 and 48 h of fermentation. However, its vector control, S. clavuligerus pSETermE* showed a higher growth pattern than S. clavuligerus pLB2 and the wild type till 96 h of incubation (Fig. 2b). The growth kinetics of bldG mutant was always higher than the manipulated strains.

Fig. 2.
figure 2

The effect of bldG on tunicamycin production. Fermentation profiles of S. clavuligerus strains grown in TSBYG medium (a, b) and specific tunicamycin production (c, d). 1—S. clavuligerus ATCC27064, 2—S. clavuligerus ΔbldG, 3—S. clavuligerus pLB1, 4—S. clavuligerus pSPG, 5—S. clavuligerus pLB2, 6—S. clavuligerus pSETermE*.

Full size image

In contrast to its high level of growth, S. clavuligerus ΔbldG produced at most 24.2 µg/mg tunicamycin throughout the fermentation corresponding to 10% specific titer of the wild type strain. Introduction of non-integrated multicopy bldG gene or integration of its extra copy into the genome of S. clavuligerus resulted in elevated tunicamycin levels. S. clavuligerus pLB1 reached 3968.3 µg/mg specific tunicamycin production corresponding to 16.4- and 1.8-fold higher tunicamycin titer than those of the wild type strain (242 µg/mg) and the vector control S. clavuligerus pSPG (2272.9 µg/mg) (Fig. 2с). S. clavuligerus pLB2 having extra copy of bldG along with ermE* promoter produced as much as 3329.3 µg/mg specific tunicamycin and this production was 13.8- and 4.5-fold more than the wild type strain and S. clavuligerus pSETermE* (731.9 µg/mg), respectively (Fig. 2d).

Transcriptional analysis of the effect ofbldGin tunicamycin C gene cluster. Expression of genes in the tunicamycin gene cluster of S. clavuligerus ΔbldG grown in TSBYG medium was compared with that of S. clavuligerus ATCC27064 by qRT-PCR (Fig. 3a). All genes studied showed significant expression levels with p values (<0.001, <0.01 or p < 0.05) with the exception of SCLAV_4274 (p > 0.05). SCLAV_4275 was the only gene that exerted 1.86-fold increased expression level in the mutant strain compared to the wild type S. clavuligerus at T36. All other tunicamycin genes were found to have lower expression values than the wild type strain ranging from 3.6- to 58-fold. The expression level of SCLAV_4287, encoding the first enzyme of tunicamycin biosynthesis pathway, was 57.9-fold lower than in the wild type. SCLAV_4286 expression also 52-fold reduced in S. clavuligerus ΔbldG after 36 h of incubation. The expression of SCLAV_4285-4283 was 30-36.7-fold lower in the bldG mutant of S. clavuligerus. The remaining genes of tunicamycin gene cluster were also downregulated with 3.6-20.9-fold decreased expression levels in the absence of bldG gene in S. clavuligerus (Fig. 3a).

Fig. 3.
figure 3

The effect of bldG in the expression of tunicamycin gene cluster in S. clavuligerus. Expression changes of tunicamycin genes (a) in S. clavuligerus ΔbldG, (b) in S. clavuligerus pLB1 and S. clavuligerus pSPG, (c) in S. clavuligerus pLB2 and S. clavuligerus pSET152ermE*, compared to the S. clavuligerus ATCC27064 (the expression was taken as 1). 1—S. clavuligerus ΔbldG, 2—S. clavuligerus pLB1 (black bars), 3—S. clavuligerus pSPG (white bars), 4—S. clavuligerus pLB2 (dark grey bars), 5—S. clavuligerus pSET152ermE* (white bars). I—36, II—72 h. Significance was stated as p values (*<0.05; **<0.01; ***<0.001).

Full size image

In contrast to expression profiles of 14 tunicamycin genes in S. clavuligerus ΔbldG after 36 h of incubation, a totally different expression levels were observed at 72 h incubation. All genes showed statistically significant and higher expression levels in the range 5.4–54.2-fold at this time point in the ΔbldG mutant in comparison to the wild type. The highest expression ratio was obtained for SCLAV_4275.

The effect of extra copy/ies of bldG on the expression levels of tunicamycin genes was also determined by qRT-PCR. In S. clavuligerus pLB1, no significant increases for SCLAV_4274, SCLAV_4275, SCLAV_4286 and SCLAV_4287 gene expression were observed at both incubation times. However, SCLAV_4276 and SCLAV_4282 genes gave statistically significant expression levels with maximum 2.7- and 4.3-fold increases, respectively, over the wild type strain at 36 h (Fig. 3b).

In S. clavuligerus pLB2, the increase in the expression levels of SCLAV_4275, SCLAV_4276, SCLAV_4280, SCLAV_4282, SCLAV_4284, SCLAV_4285 and SCLAV_4286 genes was statistically significant at 36 h of growth while all the genes did not exert such a significant expression change at the later fermentation time (Fig. 3c).

DISCUSSION

S. clavuligerus ΔbldG mutant is unable to form aerial hyphae and cannot produce clavulanic acid and cephamycin C as CSR-CcaR transcription is dependent on BldG [1]. Recently it has been shown that BldG has a wider regulatory effect in the control of clavulanic acid production in S. clavuligerus at a level above of CcaR by acting on other proteins such as BldN/AdsA, AfsR-like regulator and a TetR family protein [33]. Makitrynskyy et al. [3] showed that AdpA and BldA pleiotropic regulators directly control moenomycin antibiotic production in Streptomyces ghanaensis. Like moenomycin, tunicamycin also have a CSR-free gene cluster. In our study, tunicamycin production in the S. clavuligerus ΔbldG mutant was very low compared to the wild type at all sampling times studied. Except for SCLAV_4275, the other tunicamycin genes had lower expression value after 36 h of incubation in the bldG mutant strain. The most drastic decrease was examined in SCLAV_4286 and SCLAV_4287 expressions. All tunicamycin genes were upregulated at later incubation time in the mutant strain. Interestingly, the most upregulated gene by the absence of bldG was SCLAV_4275 with 53.2-fold increased expression level after 72 h of incubation. The present data showed that SCLAV_4275 and SCLAV_4286, the genes that have rare TTA codon in the tunicamycin gene cluster exerted diverse expression changes in the S. clavuligerus ΔbldG mutant. SCLAV_4275 (putative muT-like protein encoding gene) is involved in formation of tunicaminyluracil core together with SCLAV_4286 encoding radical SAM binding protein [18]. SCLAV_4286 involves TGGCCGGCTA, highly resembling to type II consensus sequence (TGGCCGGATT) for binding of AdpA located at the ccaR promoter region in the cephamycin C gene cluster [21]. In a recent study of Widdick et al. [20] it was reported that tunicamycin gene cluster is transcribed as a single operon from two promoters (tunp1 and tunp2) located upstream of SCLAV_4287 in S. chartreusis. Although DNA binding of AdpA to upstream region of SCLAV_4286 and SCLAV_4287 was tested in our study, no binding was obtained probably due to poor experimental conditions (data not shown). Therefore, tunicamycin biosynthesis might be directly controlled by BldG or it might exert regulatory effect on tunicamycin through mediating AdpA expression that was shown to be underrepresented in the absence of BldG [33]. However, further studies are needed to elucidate the possible regulation mechanism of BldG in tunicamycin biosynthesis.

The results obtained by bldG mutation in tunicamycin production were confirmed by bldG overexpressed S. clavuligerus pLB1 and S. clavuligerus pLB2 recombinant strains. Furthermore, the introduction of integrated copy of bldG in the cell provided more tunicamycin production than that in S. clavuligerus pLB1, multicopy bldG-containing recombinant strain, with respect to the vector control, most probably due to the use of different promoters [23]. Concomitantly, tunicamycin gene expressions in S. clavuligerus pLB1 and S. clavuligerus pLB2 were differently regulated but a general upregulation was observed in only exception that at T72, the expression differences of most tunicamycin genes in S. clavuligerus pLB2 were statistically insignificant.

* * *

The study presented here is the first report regarding to examine possible regulation role acting on expression of tunicamycin genes and production of tunicamycin. It was shown that BldG pleiotropic regulator affects tunicamycin biosynthesis at transcriptional and translational level in S. clavuligerus. However, further studies are needed to elucidate the exact mechanism.