The VanRS Homologous Two-Component System VnlRS_Ab of the Glycopeptide Producer Amycolatopsis balhimycina Activates Transcription of the vanHAX_Sc Genes in Streptomyces coelicolor, but not in A. balhimycina

Abstract

In enterococci and in Streptomyces coelicolor, a glycopeptide nonproducer, the glycopeptide resistance genes vanHAX are colocalized with vanRS. The two-component system (TCS) VanRS activates vanHAX transcription upon sensing the presence of glycopeptides. Amycolatopsis balhimycina, the producer of the vancomycin-like glycopeptide balhimycin, also possesses vanHAX_Ab genes. The genes for the VanRS-like TCS VnlRS_Ab, together with the carboxypeptidase gene vanY_Ab, are part of the balhimycin biosynthetic gene cluster, which is located 2 Mb separate from the vanHAX_Ab. The deletion of vnlRS_Ab did not affect glycopeptide resistance or balhimycin production. In the A. balhimycina vnlR_Ab deletion mutant, the vanHAX_Ab genes were expressed at the same level as in the wild type, and peptidoglycan (PG) analyses proved the synthesis of resistant PG precursors. Whereas vanHAX_Ab expression in A. balhimycina does not depend on VnlR_Ab, a VnlR_Ab-depending regulation of vanY_Ab was demonstrated by reverse transcriptase polymerase chain reaction (RT-PCR) and RNA-seq analyses. Although VnlR_Ab does not regulate the vanHAX_Ab genes in A. balhimycina, its heterologous expression in the glycopeptide-sensitive S. coelicolor ΔvanRS_Sc deletion mutant restored glycopeptide resistance. VnlR_Ab activates the vanHAX_Sc genes even in the absence of VanS. In addition, expression of vnlR_Ab increases actinorhodin production and influences morphological differentiation in S. coelicolor.

Introduction

Bacteria need to respond to changes in their environment. Therefore, they require adequate means to gain and process information on the immediate surroundings. Such means are represented by two-component systems (TCSs), which are ubiquitous in all prokaryotes. A typical TCS consists of a sensor histidine kinase (HK) and a response regulator (RR).¹ The HK measures a specific external signal and autophosphorylates at a conserved histidine residue within the cytosol. This phosphoryl group is transferred to the associated RR. The activated RR initiates the cellular response.² Most RRs are transcription factors that not only change the gene expression pattern of one or more genes of the cell, but also post-transcriptional and post-translational regulation of RNAs and proteins, respectively, by RRs has been reported.³ The ability of bacteria to sense the signal enables them to react with an adaptive response.

Of special interest is the glycopeptide-sensing TCS VanRS that controls the expression of glycopeptide resistance genes in gram-positive pathogens,^4,5 some glycopeptide producers, and other actinomycetes.⁶ VanS is a membrane-standing HK. Its C-terminus extends into the cytoplasm and contains the kinase domain and the phosphorylation site.⁷ VanS senses the presence of glycopeptides and catalyses adenosine triphosphate-dependent autophosphorylation of a specific histidine residue. Subsequently, VanS transfers the phosphate group to an aspartate residue of VanR, which then activates the transcription of the resistance genes. However, under noninduction conditions, VanS acts as a phosphatase, removing the phosphate group from VanR.⁸

Glycopeptides such as vancomycin, teicoplanin, and telavancin are used for treating infections caused by gram-positive pathogens. They act by binding to the N-acyl-d-alanyl-d-alanine (d-Ala-d-Ala) termini of peptidoglycan (PG) and its precursor lipid II. This binding effectively sequesters the substrate for the transglycosylases and the d,d-transpeptidases, two key enzymes of cell wall synthesis, resulting in an inability to grow and subsequently to cell death.

Glycopeptide resistance is mediated by reprogramming cell wall biosynthesis. Ten types of resistances have been characterized so far (VanA-N).^9,10 In each case, the terminal d-alanine (d-Ala) in the pentapeptide side chain of the PG of gram-positive bacteria is substituted either by a d-lactate (d-Lac) (VanA, B, D, F, and M) or a d-serine (VanC, E, G, L and VanN). These substitutions result in a 1000-fold¹¹ or 6-fold¹² decreased binding affinity of the glycopeptide to its target, respectively. Categorization into the different phenotypes is based on the inducibility, the breadth of resistance to individual compounds, and the level of resistance.⁹

Three of those phenotypes, VanC,¹³ VanD,¹⁴ and VanN,^15,16 are constitutively expressed. All others are inducible to different degrees by different glycopeptides. It was shown that enterococcus and staphylococcus strains expressing glycopeptide resistance genes constitutively are impaired in growth in comparison with strains where the genes are inducible.^17,18 Apparently, careful control of the expression of these genes is advantageous.

Streptomyces coelicolor A3(2) is neither a pathogen nor a glycopeptide producer, but it is likely to encounter glycopeptides in its natural habitat. Therefore, it benefits from carrying vanRS_Sc, vanHAX_Sc, vanK_Sc, and vanJ_Sc (Fig. 1C). VanH_Sc is a d-stereospecific lactate dehydrogenase that converts pyruvate to d-Lac. VanA_Sc is a d-Ala-d-Ala-ligase family protein that ligates d-Ala and d-Lac to d-Ala-d-Lac-depsipeptides. VanX_Sc is a highly selective carboxypeptidase that cleaves the remaining d-Ala-d-Ala-dipeptide. VanK_Sc belongs to the Fem family of enzymes, which add the cross-bridging amino acid(s) to the stem pentapeptide of PG precursors.¹⁹ VanY is a membrane protein conferring resistance to teicoplanin. To identify the precise nature of the ligand signal that activates glycopeptide resistance in S. coelicolor A3(2), the VanB-type HK VanS_Sc, sensing vancomycin, but not teicoplanin, was investigated.^21,22 Investigation on VanS_Sc revealed opposed results. On the one hand, it was shown by cross-linking experiments that vancomycin is the direct ligand of the VanS_Sc.²¹ On the other hand, Kwun et al.²² demonstrated that VanS_Sc is activated by vancomycin in complex with the d-Ala-d-Ala termini of PG precursors.

FIG. 1.

(A, B) EMBOSS stretcher pairwise sequence alignment of VnlRS_Ab and VanRS_Sc. (A) EMBOSS stretcher pairwise sequence alignment of VnlS_Ab and VanS_Sc. Transmembrane domains are indicated in blue. The extracytosolic domain is highlighted by a red box. (B) EMBOSS stretcher pairwise sequence alignment of VnlR_Ab and VanR_Sc. The site of aspartate phosphorylation is indicated by red and that of the proposed autophosphorylation by green arrow. (C) Organization of the resistance genes in Streptomyces coelicolor compared with that of Amycolatopsis balhimycina. “-” for a mismatch or a gap; “.” for any small positive score; “:” for a similarity, which scores more than 1.0; and “I” for an identity where both sequences have the same residue.

The actinomycete Amycolatopsis balhimycina produces the glycopeptide balhimycin, a vancomycin-type glycopeptide differing from vancomycin only in the glycosylation pattern.^23,24 The balhimycin biosynthesis gene cluster contains all genes necessary for balhimycin production²⁵ as well the vanRS-like regulatory genes, vnlRS_Ab and the accessory resistance gene vanY_Ab (Fig. 1C).²⁶ VanYAb is a carboxypeptidase which cleaves the endstanding D-Ala-D-Ala-dipeptide from the PG precursors.²⁷ However, the vanHAX_Ab genes are encoded more than 2 Mb apart from the balhimycin biosynthesis cluster. Although VnlR_Ab does not regulate glycopeptide resistance in A. balhimycina,²⁷ its heterologous expression in the glycopeptide-sensitive S. coelicolor strain M600 J2301⁶ (ΔvanRS_Sc) revealed unexpected effects. VnlR_Ab activates the vanHAX_Sc genes, increases actinorhodin production, and influences morphological differentiation in S. coelicolor.

Materials and Methods

Bacterial, strains, plasmids, and primers

The strains and plasmids used for this study are listed in Table 1, the primers used for this study are listed in Table 2.

Table 1.

Bacterial Strains Used in This Study

	Relevant feature(s)	References
Strains
Streptomyces coelicolor A3(2)
M600	SCP1- SCP2-	30
M600 ΔvanRSSc (J2301)	vanRSSc deletion mutant	6
M600 ΔvanRSSc [vnlRSAb]	ΔvanRSSc complemented with pRM4vnlRSAb	This study
M600 ΔvanRSSc [vnlRAb]	ΔvanRSSc complemented with pRM4vnlRAb	This study
M600 ΔvanRSSc [vnlSAb]	ΔvanRSSc complemented with pRM4vnlSAb	This study
M600 ΔvanRSSc [vnlRAbD51A]	ΔvanRSSc complemented with pRM4vnlRAbD51A	This study
Amycolatopsis balhimycina DSM 5908
A. balhimycina WT DSM 5908	Wildtype	24
A. balhimycina ΔvnlRAb	vnlRAb deletion mutant	27
A. balhimycina [vnlRAb]	Overexpression of vnlRAb in A. balhimycina, using pRM4vnlRAb	This study
A. balhimycina ΔvnlRAb [vnlRAb]	ΔvnlRAb complemented with pRM4vnlRAb	This study
A. balhimycina ΔvnlSAb	vnlSAb deletion mutant	This study
Escherichia coli
XL1-blue	recA1; endA1; gyrA96; tji-1; hsdR17; supE44; relA1; lac [F'proAB, lacqZΔM15Tn10(tetr)]	28
ET 12567 pUZ8002	pUZ8002; kanr	34
ET 12567	F-; dam13::Tn9; dcm-6; hsdM; hsdR; recF143; zjj201::Tn10; galK2; galT22; ara14; lacY1; xyl15; leuB6; thi1; tonA31; rpsL136; hisG4; tsx78; mtli; glnV44	29
Plasmids
pRM4	pSET152 ermEp*, RBS, Φ31 attP-int-derived integration vector	31
pRM4vnlRSAb	Expression plasmid for vnlRSAb	This study
pRM4vnlRAb	Expression plasmid for vnlRAb	This study
pRM4vnlSAb	Expression plasmid for vnlSAb	This study
pRM4vnlRAbD51A	Expression plasmid for vnlRAb,	This study
	Exchange of Asp at position 51 with Ala
pSP1	Inactivation vector in A. balhimycina	32
	Erythromycin and ampicillin resistance
pSPΔvnlSAb	pSP1 carrying a 1579 bp upstream and a 1509 bp downstream fragment of vnlSAb	This study

WT, wild type.

Table 2.

Primers Used in This Study

	Primer	Sequence	Relevant feature(s)	References
1	vnlRS-compl.1	CATCGGCATATGCGCGTGCTGATCGTCGAG	Cloning of vnlRSAb	This study
2	vnlRS-compl. 2	GAATTCCTGCGCGACTCCAGCGTTTCTCAGCGGAAG
3	vnlR-over	GAATTCAGGCGTAGCTGAGG	Cloning of vnlRAb	This study
4	vnlS-cloning	ATCATATGAGCGTCCGCCTCAAAC	Cloning of vnlSAb	This study
5	vnlRD51A1	GAATATCCCGGgCGAGGACGG	Recombinant primers for aa exchange	This study
6	vnlRD51A2	CCGTCCTCGcCCGGGATATT
7	vnlSdelFrg1for	TTAGAATTCGATTGTCCGCGAGAAATG	Cloning of vnlSAb upstream region	This study
8	vnlSdelFrg1rev	TAATCTAGACCCGGCCGCTCTGTC
9	vnlSdelFrg2for	TTATCTAGACCGGCCGGGTCACCTC	Cloning of vnlSAb downstream region	This study
10	vnlSdelFrg2rev	ATTGCATGCGGGCGCAAGTGAGTTTCGGTCATCG
11	pSETerme rev	ATGCTAGTCGCGGTTGA	Integration of Plasmid and Insert	This study
12	attBli-fwd	TTCTGGAAATCCTCGAAGGC	Integration of plasmid through ΦC31
13	attPint-rev	TGTGCATGCGCCCACGAATG
14	ery for	AAGGGAGAAAGGCGGACAGG	Proof of erythromycin resistance cassette	This study
15	ery rev	GTCGCTTCTGCGCAAGTACC
16	vnlSproof for	TGCTCGAAGTCCTCGTTTCC	Integration and deletion verification	This study
17	vnlSproof rev	GCAAGTACGTGAGCGATCAG
18	vanSSc_RT_1	CTCCAACTGACCACGAACCT	RT-PCR analysis in S. coelicolor M600	This study
19	vanSSc_RT_2	GGTCGGTGTGTATGCGTTC
20	vanRSc_RT_1	TGCTGAGTGTCAACGCCTAC
21	vanRSc_RT_2	CGAACTGCTTCCTGGTCAAC
22	vanASc_RT_1	ACCGTGACAGGAGACGAGAC
23	vanASc_RT_2	CTGGTGGATCCGGAAGAAT
24	hrdBSc_RT_1	TGACCAGATTCCGGCCACTC		This study
25	hrdBSc_RT_2	CTTCGCTGCGACGCTCTTTC
26	sigB for	CGTAGGTCGAGAACTTGAAC	RT-PCR analysis in A balhimycina DSM5908	41
27	sigB rev	GTGTCTACCTCAACGGTATC
28	vanH1	GGGACAAGCCCATCAAGAAC		27
29	vanA2	GAGCGGACTTGACGGAGATG
30	vanY_RT_fwd	TCGGCACGAGGATTG
31	vanY_RT_rev	TTCACGCACAGTTCG

RT-PCR, reverse transcriptase polymerase chain reaction.

Escherichia coli XL1-blue²⁸ was used for cloning purposes, and the methylation-deficient strain E. coli ET12567²⁹ was used to obtain unmethylated DNA for A. balhimycina transformations.

A. balhimycina²⁴ is the balhimycin-producing wild type (WT) and was used to generate the vnlS deletion as well as the vnlR-overexpressing strains (this study). Furthermore, ΔvnlR deletion²⁷ was used for complementation (this study).

S. coelicolor M600³⁰ were used to generate S. coelicolor M600 ΔvanRS.⁶ This deletion strain was used to generate complementations with vnlRS_Ab, vnlR_Ab, vnlS_Ab, and vnlR_AbD51A (this study).

The overexpression plasmids pRM4vnlRS_Ab, pRM4vnlR_Ab, pRM4vnlS_Ab, and pRM4vnlR_AbD51A are derived from pRM4³¹, a pSET152-derived nonreplicative, ΦC31 integration vector with an integrated constitutive ermEp* promoter, an artificial ribosomal binding site, and an apramycin resistance cassette.

The deletion vector pSPΔvnlS_Ab is derived from pSP1³² in which flanking regions of vnlS_Ab were cloned.

Media and culture conditions

A. balhimycina grown in 100 ml TSB medium (Difco) for 48 hr and 2 ml of this preculture were used to inoculate the main cultures either in 100 ml R5³⁰ or in TSB medium. R5 medium was used to stimulate balhimycin production, while TSB medium was used when balhimycin production should be prevented. After 48 hr of cultivation, the mycelium was used to isolate PG precursors, to extract DNA, or to perform resistance assays against different glycopeptides. To isolate RNA, the cells were grown 15/39/63 hr. Balhimycin production assays were performed after 5 days of growth.

S. coelicolor M145 and M600 were grown on Cullum-agar plates for sporulation. Isolated spores were used to inoculate 10 ml R5 medium as preculture for cell wall precursor extraction or DNA extraction. For RNA isolation, 2 ml of a 48-hr-old TSB preculture was used to inoculate 100 ml of HA medium. The cells were harvested after 69 hr.

To compare the growth of the different S. coelicolor strains, 10 μl spores (∼1.5 × 10⁷) of each strain were streaked on a YM plate. The plate was incubated for 7 days.

A. balhimycina and S. coelicolor were grown at 30°C, and liquid cultures were shaken at 180 rpm.

A. balhimycina and S. coelicolor strains were cultivated in 100 ml of R5 medium in an orbital shaker (220 rpm) in 500-ml baffled Erlenmeyer flasks at 27°C.

Liquid/solid media were supplemented with 100 μg/ml apramycin to select for strains carrying integrated antibiotic resistance genes.

E. coli was grown in Luria-Bertani broth (Roth) at 37°C using 100 μg/μl apramycin or 150 μg/μl ampicillin for selection of plasmid-containing colonies. Liquid cultures were shaken at 180 rpm.

Plasmid construction

For the heterologous expression in S. coelicolor ΔvanRS_Sc, the entire coding regions of the vnlR_Ab (Table 2 primer 1 + 3), vnlRS_Ab (Table 2 primer 1 + 2), and vnlS_Ab (primer 4 + 2) were amplified using Kapa-Hifi proofreading polymerase and the corresponding primers in brackets.

The vnlR_Ab (758 bp), vnlRS_Ab (1908 bp), and vnlS_Ab (1198 bp) polymerase chain reaction (PCR) products were integrated into pRM4³¹ through the primer-attached restriction sites (Nde-EcoRI) downstream of the ermEp* promoter.

Site-directed mutagenesis by overlap extension³³ was performed for the exchange of aspartate at position 51 to an alanine with the primers 5 + 6 (Table 2). The 758 bp PCR product vnlR_AbD51A was integrated into pRM4 through the primer-attached restriction sites (Nde-EcoRI) downstream of the ermEp* promoter.

For the in-frame deletion of vnlS_Ab (1125 bp), a 1579 bp upstream fragment (Table 2 primer 7 + 8) and a 1509 bp downstream fragment (Table 2 primer 9 + 10) of vnlS_Ab were amplified from A. balhimycina genomic DNA using Kapa-Hifi proofreading polymerase and the corresponding primers in brackets. The plasmid pSPΔvnlS was constructed by integration of the fragments in pSP1³⁰ through the primer-attached restriction sites at the 5′ and 3′ ends (EcoRI/XbaI and XbaI/SphI) (pSPΔvnlS).

DNA transfer

Transformation of E. coli XL1-blue²⁸ and ET12567 (pUZ8002)³⁴ was performed as described previously.^35,36

Plasmids pRM4vnlRS_Ab, pRM4vnlR_Ab, pRM4vnlS_Ab, and pRM4vnlR_AbD51A were transferred into S. coelicolor through intergeneric conjugation.³⁰ Plasmid integration was confirmed by colony PCR using the primer pair 12 + 13 (Table 2) or primer 11 (Table 2) in combination with a reverse primer of corresponding gene.

pRM4vnlR_Ab and pRM4vnlR_AbD51A were transferred into A. balhimycina through the direct transformation method^32,37 using unmethylated plasmid DNA isolated from E. coli ET12567. Integration of plasmid was verified by PCR using primer pair 11 + 3 (Table 2).

For deletion of vnlS_Ab, A. balhimycina WT was transformed with pSPΔvnlS_Ab by direct transformation. The integration of the plasmid into the chromosome through homologous recombination was confirmed by PCR screening for the erythromycin resistance cassette, using primers ery for and ery rev (Table 2). To obtain deletion mutants, a second homologous recombination event was provoked by stressing plasmid-carrying colonies as described by Puk et al.³⁸ Colonies were examined for sensitivity to erythromycin, and the deletions were verified by PCR analysis, using primers 16 + 17 (Table 2).

Sequence alignment

The amino acid (AA) sequences of VnlRS_Ab, VanR_Sc, and VanS_Sc are available under accession number Y16952 (named VanRS), (SCO3590), and (SCO3589), respectively.

Alignment of the AA sequences was performed by EMBOSS stretcher³⁹; (www.ebi.ac.uk/Tools/psa/emboss_stretcher/).

Resistance test, reverse transcriptase polymerase chain reaction analyses, PG precursor, and cell wall analysis

Resistance test, reverse transcriptase polymerase chain reaction (RT-PCR) analyses, extraction of PG precursors, PG isolation, and the high-performance liquid chromatography–mass spectrometry (HPLC-MS) analyses were performed as described.^27,40

Balhimycin concentration

The balhimycin concentration in 1 ml culture was quantified using HPLC with a photodiode array detector (HPLC-DAD) as described.⁴⁰ The balhimycin concentration was calculated to 100 μg/ml total DNA.

Inference of biomass concentration from DNA quantification

For the quantification of total DNA in 1 ml culture, an acid extraction of DNA coupled with a colorimetric method⁴¹ was performed by measuring the absorbance at 600 nm. To analyze the amount of DNA, a standard curve with salmon sperm DNA was generated.

Results

A. balhimycina includes a VanRS homologous TCS (VnlRS_Ab) encoded in the balhimycin biosynthetic gene cluster

In most of the antibiotic-producing bacteria, the antibiotic biosynthetic gene clusters include resistance genes. One exception is the balhimycin producer A. balhimycina. In this study, the glycopeptide resistance genes vanHAX_Ab are located 2 Mb apart from the balhimycin biosynthetic gene cluster. In addition, the resistance is characterized by another unusual feature: the counterpart of the well-known TCS VanRS, which is known to regulate vanHAX expression in pathogens and in S. coelicolor is encoded by genes (vnlRS_Ab), which are part of the biosynthetic gene cluster and are therefore not colocated with the vanHAX_Ab genes.

VanRS_Sc of S. coelicolor was reported to sense glycopeptides and to activate the expression of the vanHAX_Sc genes.¹⁹ To elucidate the differences of the two actinomycete TCSs, we compared the AA sequence of VnlRS_Ab with the sequence of VanRS_Sc. Sequence alignment using EMBOSS stretcher³⁹ revealed 82% sequence similarity between VnlR_Ab and VanR_Sc (Sco3590) and 73% between VnlS_Ab and VanS_Sc (Sco3589) (Fig. 1A, B). Based on the high similarity, a corresponding function of both RRs could be proposed.

In S. coelicolor, VanS_Sc phosphorylates VanR_Sc at the aspartate at AA position 51. Replacement of this residue with an alanine completely destroyed the activity of VanR_Sc.⁶ It has been shown that in S. coelicolor, in the absence of vancomycin, acetylphosphate phosphorylates VanR_Sc, whereas VanS_Sc acts as a phosphatase to decrease the level of VanR_Sc∼P. On exposure to vancomycin, VanS activity switches from a phosphatase to a kinase and vancomycin resistance is induced.⁶ Furthermore, Novotna et al.⁴² specified a serine residue at AA position 69 important for autophosphorylation through acetyl phosphate.

Sequence comparison revealed that VnlR_Ab contains both, a conserved aspartate at AA position 51 (D51) and a serine at AA position 69, the position that probably becomes autophosphorylated through acetyl phosphate (Fig. 1B), indicating an analogous phosphorylation pattern of VnlR_Ab compared with VanR_Sc.

The RR VnlR_Ab does not control expression of the vanHAX_Ab genes in A. balhimycina

In A. balhimycina, vanHAX_Ab expression does not depend on VnlRS_Ab. The deletion of vnlR_Ab had no effect on glycopeptide resistance and did not result in any obvious phenotype.²⁷ This raises the interesting question on the function of this TCS in A. balhimycina.

Since overexpression of RRs of two-component signal transduction systems often modulates multidrug resistance,^43,44 we overexpressed vnlR_Ab in A. balhimycina to analyze the effects on resistance and antibiotic production. vnlR_Ab was cloned under the control of the constitutive promoter ermE*p into the integrative plasmid pRM4 (pRM4vnlR_Ab). pRM4vnlR_Ab was transferred into A. balhimycina WT and into the A. balhimycina ΔvnlR_Ab mutant,²⁷ resulting in the recombinant strains A. balhimycina [vnlR_Ab] and A. balhimycina ΔvnlR_Ab [vnlR_Ab], respectively. The phenotypes of the recombinant strains overexpressing VnlR_Ab and (as a control) that of the deletion mutant A. balhimycina ΔvnlR_Ab were compared with the WT phenotype (Fig. 2). All strains produced balhimycin at the same level (Fig. 2A). No differences in resistance against balhimycin were observed. Using a method optimized for actinomycetes,²⁷ muropeptides from all A. balhimycina strains cultivated under balhimycin production conditions were isolated. HPLC/MS chromatograms showed the similar muropeptide composition pattern for all strains (Fig. 2B).

FIG. 2.

Analysis of balhimycin production, muropeptide composition, and PG precursors in A. balhimycina WT (1), A. balhimycina ΔvnlR_Ab (2), A. balhimycina [vnlR_Ab] (3), and A. balhimycina ΔvnlR_Ab [vnlR_Ab] (4). (A) Production of balhimycin measured by HPLC (n = 5). (B) HPLC/MS chromatogram of the muropeptides (positive mode). The first bracket embraces the peaks representing muropeptide monomers, the second the muropeptide dimers. (C) Extracted ion chromatograms of the negative mode from the PG precursors isolated from cells grown in R5 (balhimycin production) and in TSB (no balhimycin production). d-Lac, Pentapeptide precursors ending on d-Ala-d-Lac 1194 m/z at retention time ∼18 min. d-Ala, Pentapeptide precursors ending on d-Ala-d-Ala 1193 m/z at retention time ∼12 min. HPLC, high-performance liquid chromatography; MS, mass spectrometry; PG, peptidoglycan; WT, wild type.

In addition to muropeptides, the PG precursors were analyzed. For this purpose, we cultivated the strains under balhimycin production conditions and conditions under which balhimycin production is disabled. Under production as well as under nonproduction conditions, A. balhimycina WT, A. balhimycina [vnlR_Ab], A. balhimycina ΔvnlR_Ab, and A. balhimycina ΔvnlR_Ab [vnlR_Ab] produced resistant PG precursors ending with d-Ala-d-Lac (Fig. 2C). Only A. balhimycina WT and A. balhimycina ΔvnlR_Ab [vnlR_Ab] produced traces of precursors ending with d-Ala-d-Ala under nonproduction conditions (Fig. 2C). These results suggest that VnlR_Ab does not regulate the synthesis of resistance PG in A. balhimycina.

RT-PCR analyses revealed that a vanHAX_Ab transcript was detectable in A. balhimycina ΔvnlR_Ab, confirming that the expression of vanHAX_Ab is independent of vnlR_Ab (Fig. 3).

FIG. 3.

RT-PCR analyses of vanHAX_Ab and vanY_Ab in A. balhimycina WT, A. balhimycina ΔvnlR_Ab, and in A. balhimycina WT overexpressing vnlR_Ab (WT [vnlR_Ab]). RNA was isolated at different time points (15/39/63 hr) from the three strains cultivated in balhimycin production medium R5. sigB: transcription of the housekeeping gene sigB. vanHAX_Ab and vanY_Ab: transcription of vanHAX_Ab and vanY_Ab. For PCR, genomic DNA (gDNA) was used as positive control.

In A. balhimycina, sensing of glycopeptides through VnlS_Ab is not required for expressing the resistance genes

In enterococci and in S. coelicolor, the RR VanR_Sc becomes phosphorylated by the HK VanS_Sc. To analyze whether and how VnlR_Ab interacts with VnlS_Ab, we constructed an in-frame ΔvnlS_Ab mutant of A. balhimycina using the inactivation plasmid pSPΔvnlS_Ab. This plasmid containing a 1509 bp downstream fragment and a 1579 bp upstream fragment of vnlS_Ab was introduced into A. balhimycina through direct transformation. Successive homologous recombination resulted in the deletion of vnlS_Ab. A. balhimycina ΔvnlS_Ab showed neither a defect in balhimycin production nor resistance toward glycopeptides. In addition, no changes in the PG precursor and in the nascent PG composition in comparison with A. balhimycina WT were observed (data not shown). These results suggested that sensing the presence of glycopeptides does not correlate with balhimycin production and glycopeptide resistance. Apparently, the expression of the vanHAX_Ab genes occurs independently of VnlS_Ab.

VnlR_Ab is able to activate vanHAX_Sc transcription in S. coelicolor

In silico analyses revealed similar characteristics of VnlRS_Ab compared with VanRS_Sc. However, as shown above, the VnlRS_Ab system in A. balhimycina, in contrast to VanRS_Sc in S. coelicolor, does not regulate the vanHAX_Ab. To clarify the contradictory findings, the genes encoding the TCS VnlRS_Ab as well as VnlR_Ab and VnlS_Ab individually were transferred into the S. coelicolor mutant strain, in which the vanRS_Sc genes were deleted,⁶ to elucidate the ability of VnlR_Ab to activate the vanHAX_Sc genes in the S. coelicolor mutant. vnlRS_Ab, vnlR_Ab, and vnlS_Ab were introduced into S. coelicolor ΔvanRS_Sc under the control of the constitutive promoter ermEp* using the integrative plasmid pRM4vnlR_Ab. The growth of the recombinant strains was tested on glycopeptide-containing plates.

Introduction of vnlRS_Ab and of vnlR_Ab alone into S. coelicolor M600 ΔvanRS_Sc resulted in balhimycin-resistant strains (Fig. 4). In contrast, expression of vnlS_Ab alone did not change the glycopeptide-sensitive phenotype of the S. coelicolor ΔvanRS_Sc mutant. These results indicated that VnlR_Ab from A. balhimycina is able to activate the transcription of vanHAX_Sc in S. coelicolor M600 also in the absence of VnlS_Ab. Since in S. coelicolor M600 VanR_Sc∼P can be generated in a VanS_Sc-independent manner using acetylphosphate,⁶ we suggest a similar activation of VnlR_Ab in the absence of VnlS_Ab or VanS_Sc.

FIG. 4.

Growth and resistance of the S. coelicolor M600 ΔvanRS_Sc complemented with different combinations of vnlRS_Ab. (A) Growth on YM agar containing no antibiotic. (B) Growth on YM agar containing apramycin (100 mg/ml) (Apra 100) to prove plasmid integration. (C) Growth on YM agar containing balhimycin (10 mg/ml) (Bal 10). (D) Growth on YM agar containing teicoplanin (10 mg/ml) (Teico 10). (E) Growth on YM agar containing no antibiotic. M600, S. coelicolor M600.

The activation of the vanHAX_Sc genes in the complemented S. coelicolor M600 ΔvanRS_Sc mutant with vnlRS_Ab and vnlR_Ab was further analyzed by RT-PCR. For this purpose, RNA was isolated from 25-hr-old liquid cultures grown without addition of any glycopeptide. A vanHAX_Sc transcript was detected when S. coelicolor M600 ΔvanRS_Sc was complemented with vnlRS_Ab or with vnlR_Ab alone. However, in S. coelicolor M600 and in the S. coelicolor M600 ΔvanRS_Sc mutant, transcription of the vanHAX_Sc failed (Fig. 5), confirming the functionality of VnlR_Ab as transcriptional activator in S. coelicolor.

FIG. 5.

RT-PCR analyses of S. coelicolor M600 and different S. coelicolor M600 mutants. RNA was isolated after 25 hr of cultivation in the absence of any glycopeptide. hrdB: transcription of the housekeeping gene hrdB_Sc. vanS_Sc, vanR_Sc, and vanHAX_Sc: transcription of vanS_Sc, vanR_Sc, and vanHAX_Sc. For PCR, genomic DNA (gDNA) was used as positive control.

To investigate whether transcription of the vanHAX_Sc genes indeed resulted in the formation of glycopeptide-resistant PG precursors, which caused the resistant phenotype, we used HPLC/MS to analyze the PG precursor composition of S. coelicolor M600 ΔvanRS_Sc and of S. coelicolor M600 ΔvanRS_Sc complemented either with vnlRS_Ab or vnlR_Ab. Complementing S. coelicolor M600 ΔvanRS_Sc with vnlRS_Ab or with vnlR_Ab restored the synthesis of resistant PG precursors. In the presence of balhimycin exclusively, PG precursors ending with d-Ala-d-Lac were synthesized (Fig. 6C, D). The PG precursor composition of the glycopeptide-sensitive S. coelicolor M600 ΔvanRS_Sc mutant was analyzed after growing the strain in the absence of balhimycin. In this mutant, only sensitive cell wall precursors ending on d-Ala-d-Ala (1193 m/z) eluting at a retention time of 10–11 min were detected (Fig. 6A).

FIG. 6.

Extracted ion chromatograms (negative mode) of the PG precursors isolated from cells grown in R5 medium without antibiotic or with 25 mg/ml balhimycin (bal25). M600 ΔvanRS_Sc: S. coelicolor M600 ΔvanRS_Sc; d-Lac, pentapeptide precursors ending on d-Ala-d-Lac 1194 m/z at retention time ∼18 min; d-Ala, pentapeptide precursors ending on d-Ala-d-Ala 1193 m/z at retention time ∼14 min.

The phosphorylation site D51 is essential for the function of VnlR_Ab

To define the phosphorylation site of VnlR_Ab, D51, which was identified as a likely phosphorylation site by sequence composition (Fig. 1), was replaced by an alanine by exchanging nucleotide A to C at position 161 of vnlR_Ab using the recombinant PCR method. The exchange was verified by sequence analysis. The mutated gene was cloned into the integrative vector pRM4 under the control of the ermEp* promoter and introduced into S. coelicolor M600 ΔvanRS_Sc. The resulting recombinant strain was not able to grow in the presence of the tested glycopeptides (Fig. 4). Therefore, we propose D51 as the VnlR_Ab phosphorylation site.

VnlR_Ab expands the glycopeptide resistance in S. coelicolor

The glycopeptide resistance mechanism in S. coelicolor belongs to the VanB type of resistance, meaning that glycopeptide resistance can only be induced by vancomycin or vancomycin-type glycopeptides, whereas teicoplanin (a type IV glycopeptide) fails to activate resistance,⁴² resulting in a teicoplanin-sensitive phenotype of S. coelicolor. In contrast, A. balhimycina is resistant against vancomycin- as well as teicoplanin-type glycopeptides. To analyze whether the RR is responsible for determination of the glycopeptide resistance type, the recombinant strains S. coelicolor ΔvanRS_Sc [vnlRS_Ab], S. coelicolor ΔvanRS_Sc [vnlR_Ab], S. coelicolor ΔvanRS_Sc [vnlS_Ab], and S. coelicolor ΔvanRS_Sc [vnlRS_Ab D51A] were grown on teicoplanin-containing plates. Surprisingly, the recombinant strains (S. coelicolor ΔvanRS_Sc [vnlRS_Ab], S. coelicolor ΔvanRS_Sc [vnlR_Ab]) were able to grow also on teicoplanin-containing plates, whereas growth of S. coelicolor M600 WT was inhibited (Fig. 4). These results indicated that VnlR_Ab is able to induce teicoplanin resistance in S. coelicolor M600 by probably activating further genes required for teicoplanin resistance.

VnlR_Ab influences antibiotic production in S. coelicolor

To analyze if the heterologous expression of VnlR_Ab, in addition to the activation of the vanHAX_Sc genes, causes further (morphological) changes in S. coelicolor M600, the growth and production of actinorhodin were investigated without the addition of any antibiotic. Similar titers of spores (1.5 × 10⁷) of S. coelicolor M600, S. coelicolor ΔvanRS_Sc, S. coelicolor ΔvanRS_Sc [vnlRS_Ab], S. coelicolor ΔvanRS_Sc [vnlR_Ab], S. coelicolor ΔvanRS_Sc [vnlS_Ab], and S. coelicolor ΔvanRS_Sc [vnlRS_Ab D51A] were plated on YM medium. Surprisingly, the heterologous expression of vnlRS_Ab or vnlR_Ab alone in S. coelicolor M600 ΔvanRS_Sc caused retardation in growth and increased actinorhodin production (Fig. 4E).

These results suggested that VnlR_Ab is not only able to activate the vanHAX_Sc genes in S. coelicolor M600 and to change its glycopeptide resistance type but it also has effects on other genes in S. coelicolor M600.

VnlR_Ab is responsible for the activation of vanY_Ab

Heterologous expression of VnlR_Ab in S. coelicolor confirmed that it can take over the VanR_Sc function to induce the expression of the vanHAX_Ab genes and, in addition, can apparently induce the expression of further genes. In contrast, it is not involved in regulation of the vanHAX_Ab genes in A. balhimycina. Since regulatory genes are often colocalized with its target genes, we speculated that VnlRS_Ab might control vanY_Ab, which is located directly adjacent to vnlR_Ab and which encodes a carboxypeptidase. Previous studies showed that VanY_Ab cleaves the d-Ala-d-Ala dipeptide from the PG precursors, but it is not able to cleave the d-Ala-d-Lac depsipeptide.²⁷ To investigate, whether VnlR_Ab regulates the expression of vanY_Ab, transcriptional analyses were performed. RT-PCR analyses revealed that vanY_Ab was only transcribed when vnlR_Ab was expressed under the control of the strong promoter ermE*p (Fig. 2 (A. balhimycina [vnlR_Ab]). In A. balhimycina WT, transcription was detectable on a low level only after 63 hr of cultivation and in the A. balhimycina ΔvnlR_Ab mutant, vanY_Ab transcription was not induced at all (Fig. 2). This result was confirmed by RNA-seq analyses where we compared the transcription level of vanY_Ab in the A. balhimycina WT and A. balhimycina ΔvnlR_Ab (data not shown). The transcription of vanY_Ab was 25-fold decreased in A. balhimycina ΔvnlR_Ab compared with A. balhimycina WT. We therefore concluded that the RR VnlR_Ab in A. balhimycina is involved in controlling the expression of resistance mediated by VanY_Ab.

Discussion

Glycopeptide resistance in pathogens and in S. coelicolor is mediated by the action of VanHAX. The expression of the vanHAX genes is regulated by the TCS VanRS, the genes of which are colocalized with vanHAX. In the presence of glycopeptides, VanS becomes autophosphorylated and phosphorylates VanR, which subsequently activates transcription of vanHAX. VanH, VanA, and VanH reprogram the biosynthesis of the PG precursors, resulting in lipid II with an N-terminal d-Ala-d-Lac depsipeptide instead of the normally occurring d-Ala-d-Ala termini, the target of the glycopeptides. A. balhimycina produces the vancomycin-like glycopeptide balhimycin and has to protect itself from the action of the glycopeptide.⁴⁵ The genome of A. balhimycina includes vanHAX_Ab genes and vanRS-like genes (vnlRS_Ab). However, in contrast to other glycopeptide-resistant bacteria, the vanHAX_Ab genes in A. balhimycina are located 2 Mb apart from the vnlRS_Ab genes, which are part of the balhimycin biosynthetic gene cluster.⁴⁰

RT-PCR experiments revealed that VnlR_Ab is not involved in the activation of the vanHAX_Ab genes in A. balhimycina. Subsequent PG analyses confirmed that a vnlR_Ab deletion mutant cannot synthesize resistant muropeptides. Since vnlR_Ab is colocalized with the balhimycin biosynthetic genes, an alternative role of VnlR_Ab as regulator of balhimycin synthesis was assumed, but the deletion of the vnlR_Ab did not affect balhimycin production. Hence, VnlR_Ab is not the central regulator activating the vanHAX_Ab resistance genes or the balhimycin biosynthetic genes.

To further investigate the potential target gene(s) of VnlR_Ab, we analyzed the transcription of vanY_Ab, which encodes a carboxypeptidase and which is located adjacent to the vnlRS_Ab genes in the balhimycin biosynthetic gene cluster.²⁵ RT-PCR and RNA-seq analyses revealed that vanY_Ab expression was 25-fold decreased in A. balhimycina ΔvnlR_Ab compared with A. balhimycina WT.

VanY_Ab is a d,d-carboxypeptidase, which cleaves the endstanding d-Ala from lipid II, resulting in the formation of tetrapeptides.²⁷ In contrast to other described carboxypeptidases,⁴⁶ VanY_Ab has no d,d-carboxyesterase activity. The tetrapeptides are the substrates for the l,d-transpeptidase (Ldt), which subsequently cross-links the tetrapeptide acyl donors at the third AA. This results in PG with 3–3 cross-linked tetra- and tripeptides, which are devoid of the d-Ala-d-Ala-ending peptides, and which can therefore not serve as target of glycopeptides anymore.⁴⁷ Investigations of the PG of A. balhimycina revealed the presence of 3–3 cross-linked tetra- and tripeptides.^40,45 Furthermore, we could identify at least three ldt genes in the genome of A. balhimycina.⁴⁵ We therefore speculate that by activating the expression of vanY_Ab, VnlR_Ab is involved in regulating an alternative, VanHAX_Ab-independent glycopeptide resistance mechanism in A. balhimycina. This fact is further confirmed by RT-PCR analysis, where it was shown that VanY_Ab is expressed in A. balhimycina ΔvanHAX_Ab.⁴⁰

This observation is in accordance with the findings in Nonomuraea ATCC 39727, the producer of the dalbavancin precursor A40926. Nonomuraea ATCC 39727 does not encode VanHAX homologs, but possesses a VanY homolog (VanY_n) for the synthesis of a resistant PG precursor.⁴⁸ As described for A. balhimycina, VanY_n cleaves the C-terminal d-Ala from the pentapeptide as well as from the d-Ala-d-Ala dipeptide. The tetrapeptides are subsequently cross-linked by Ldt, resulting in glycopeptide-resistant cell wall.⁴⁷ The surprising features of VnlR_Ab are that although it does not regulate the transcription of the vanHAX_Ab genes in A. balhimycina, it is able to activate vanHAX_Sc transcription in S. coelicolor, and that it activates teicoplanin resistance in S. coelicolor.

Activation of the vanHAX_Sc transcription can be explained by the binding of VnlR_Ab at the promoter region of vanHAX_Sc. Sequence comparison of the promoter regions of vanHAX_Sc and vanHAX_Ab not only revealed conserved motives but also some differences (data not shown). Although many attempts have been made to analyze putative promoter sequences in gel mobility assays, no shifts could be observed. This is probably due to the fact that after purification, the protein lost its functionality (data not shown). Therefore, determination of the exact binding motive of VnlR_Ab still requires alternative approaches.

VnlR_Ab was not only able to restore vancomycin resistance in an S. coelicolor ΔvanRS_Ab mutant after heterologous expression but it even conferred teicoplanin resistance to this mutant, although S. coelicolor WT is sensitive toward teicoplanin. Recent comparative study of the VanR-VanS systems from two Streptomyces strains, S. coelicolor and Streptomyces toyocaensis (the producer of the sugarless glycopeptide A47934), indicated that the glycopeptide antibiotic inducer specificity is determined solely by the differences between the AA sequences of the VanR-VanS TCS present in each strain rather than by any inherent differences in general cell properties, including cell wall structure and biosynthesis.⁴² On the one hand, the results obtained in this work support this finding; since vnlR_Ab is under the control of the ermEp* promoter, VnlR_Ab is constitutively expressed and activates the transcription of the vanHAXJK_Sc genes in S. coelicolor independent from the presence of any glycopeptide. The activation of vanHAXJK_Sc resulted in the synthesis of PG with pentapeptides ending on d-Ala-d-Lac depsipeptide, which are resistant against vancomycin and teicoplanin. On the other hand, the second explanation contradicts the work of Novotna et al.⁴²; it is likely that VnlR_Ab activates the transcription of additional unknown genes, which mediate teicoplanin resistance.

The diverse functionality of VnlR_Ab in the glycopeptide producer A. balhimycina and in the nonproducer S. coelicolor provides the starting point of evolutionary analyses of glycopeptide resistance. In pathogenic bacteria and in S. coelicolor of glycopeptides, resistance is strictly regulated and is induced by the presence of glycopeptides. In contrast, glycopeptide producers overcome this regulation not only by the constitutive expression of the vanHAX genes but also by the development of a vanHAX-independent resistance mechanism. However, the ability of VnlR_Ab to activate transcription of vanHAX in S. coelicolor is an indication of a common origin of the three glycopeptide resistance mechanisms, the inducible one, the constitutively expressed, and the vanHAX-independent mechanism. Whether and how the complex resistance mechanism has evolved in the glycopeptide producers and whether and how it was transferred into resistance pathogens have to be subjects of future investigation.

Acknowledgments

The Deutsche Forschungsgemeinschaft (DFG) supported this work by the SFB 766 program TP-A03. The authors thank Matthew I. Hutchings, Hee-Jeon Hong, and Mark J. Buttner for providing the S. coelicolor J3201 mutant and Tobias Busche and Joern Kalinowski for the RNA-seq analyses. The authors further thank Vera Rosenberger and Tobias Martin for competent assistance during the experiments, Sigrid Stockert for technical support, and Günther Muth for sharing his excellent expertise in the research and writing process.

Disclosure Statement

The authors disclose that there are no commercial associations that might create a conflict of interest in connection with submitted manuscripts.