In conditions of over-expression, WblI, a WhiB-like transcriptional regulator, has a positive impact on the weak antibiotic production of Streptomyces lividans TK24

Regulators of the WhiB-like (wbl) family are playing important role in the complex regulation of metabolic and morphological differentiation in Streptomyces. In this study, we investigated the role of wblI, a member of this family, in the regulation of secondary metabolite production in Streptomyces lividans. The over-expression of wblI was correlated with an enhanced biosynthesis of undecylprodigiosin and actinorhodin and with a reduction of the biosynthesis of yCPK and of the grey spore pigment encoded by the whiE locus. Five regulatory targets of WblI were identified using in vitro formaldehyde crosslinking and confirmed by EMSA and qRT-PCR. These included the promoter regions of wblI itself, two genes of the ACT cluster (actVA3 and the intergenic region between the divergently orientated genes actII-1 and actII-2) and that of wblA, another member of the Wbl family. Quantitative RT-PCR analysis indicated that the expression of actVA3 encoding a protein of unknown function as well as that of actII-1, a TetR regulator repressing the expression of actII-2, encoding the ACT transporter, were down regulated in the WblI over-expressing strain. Consistently the expression of the transporter actII-2 was up-regulated. The expression of WblA, that is known to have a negative impact on ACT biosynthesis, was strongly down regulated in the WblI over-expressing strain. These data are consistent with the positive impact that WblI over-expression has on ACT biosynthesis. The latter might result from direct activation of ACT biosynthesis and export and from repression of the expression of WblA, a likely indirect, repressor of ACT biosynthesis.


Introduction
Streptomyces are Gram-positive, filamentous soil bacteria of considerable biotechnological importance. Indeed this genus produces two thirds of all known antibiotics as well as other bio-active molecules, including antitumor agents, immune-suppressants, apoptosis inducers and antifungals, herbicides, insecticides etc. . . used in medicine or agriculture [1]. These bacteria are characterized by a complex developmental cycle that starts, when the nutritional conditions are favorable, by the germination of spores that develop into a substrate mycelium. Subsequently, some still poorly defined signals of nutrient limitation, are thought to trigger the development of aerial hyphae from the substrate mycelium. The tip ends of the aerial hyphae differentiate into uni-genomic spores and the production of a grey pigment encoded by the whiE locus accompanied the complete differentiation process [2,3]. This complex morphological development is mainly under the control of the bld and whi genes, that are required for the formation of aerial mycelium and spores, respectively [2,3].
The complex bld signaling cascade has been extensively studied leading to the characterization of the BldA, BldD, BldH and BldN regulons [4][5][6][7][8]. A signal molecule c-di-GMP was recently shown to induce the dimerization of the regulator BldD that is necessary to activate its DNA-binding activity [9,10]. The binding of BldD results in the repression of the sporulation genes during vegetative growth leading to a delay in the differentiation process [9,10]. One of the genes under the negative control of BldD is the σ factor, whiG [4,11,12]. The latter is necessary for the expression of numerous genes of the whi cascade, including that of the regulators whiA, whiH, whiI but not whiB [12][13][14][15]. Whi regulators are named from the white color of aerial hyphae following the mutation/disruption of their cognate encoding genes. They are known to play key roles in the differentiation of aerial hyphae into mature spores [2]. WhiB, regarded as the founding member of the WhiB-like (Wbl) family proteins, is one of the Whi regulators that is necessary for the septation steps preceding sporulation [16,17]. A recent genome-wide study demonstrated that WhiB is a transcriptional factor that binds, in cooperation with WhiA, upstream of nearly 240 transcription units required for developmental cell division [18]. Wbl proteins usually contain an unconventional helix-turn-helix motif and a [4Fe-4S] iron-sulfur cluster that have the ability to detect redox changes and regulate gene expression accordingly [19]. Such redox sensing clusters have been shown to play diverse and critical roles in actinobacterial biology, including morphological differentiation, antibiotic production, antibiotic resistance and pathogenesis [16,[20][21][22]. There are 14 WhiB-like (Wbl) regulators present in S. coelicolor, 11 Wbl proteins are encoded by chromosomal genes, and 3 are encoded by the plasmid SCP1 [23]. wblA is the most extensively studied of these regulators. It was shown to play important role in the early stage of aerial hyphal development [23,24] and to have a negative impact on oxidative stress response [25,26] and antibiotic production [24]. In contrast WhiD was shown to be essential for pre-spore maturation [27][28][29][30].
In this study we characterized another member of the Wbl gene family, WblI (SCO5046). This gene was previously shown to be a target of SCO3201, a regulator of the TetR family, whose overexpression led to strong repression of both antibiotic production and sporulation in S. coelicolor [31]. We demonstrated that the over-expression of the homologue of WblI in S. lividans TK24, greatly enhanced the weak ability of this strain to synthetize undecylprodigiosin (RED) and actinorhodin (ACT), peptidyl and polyketide secondary metabolites, respectively whereas it had a negative impact on the biosynthesis of yCPK, a type I polyketide as well as of the grey spore polyketide pigment encoded by the whiE locus. Regulatory targets of WblI were identified by formaldehyde cross linking and confirmed by EMSA and qRT-PCR. WblI constitutes a new player in the complex regulatory network governing secondary metabolite production and morphological differentiation in Streptomyces. A regulatory model consistent with the impact that WblI over-expression/deletion has on the expression of its targets and on antibiotic production is proposed and discussed. This model clarifies the hierarchical relationships between WblI and WblA.

Materials and methods
Bacterial strains, media and culture conditions S. lividans TK24 (str-6, SLP2-, SLP3-) was used in this study. Escherichia coli DH5α and BL21 were used as the hosts for routine subcloning and protein expression, respectively. SFM [32] and GYM [33] media were used for spore collection and the assessment of spore grey pigment production, respectively. R2YE medium, traditionally used for protoplast regeneration, was used for solid-grown cultures of S. lividans. To assess antibiotic production, 10 7 spores of various Streptomyces strains were spread on top of cellophane discs on R2YE agar medium. When necessary, ampicillin, kanamycin or apramycin were added to the culture medium at 50 μg/ ml, whereas thiostrepton was added to R2YE liquid medium at 5 μg/ml. Unless otherwise stated, E. coli and Streptomyces strains were incubated at 37˚C and 28˚C, respectively, and a shaking speed of 220 rpm was maintained for liquid culture.

Construction of S. lividans TK24 strains overexpressing or disrupted for wblI
In order to overexpress wblI in S. lividans TK24, the wblI gene was amplified from the genomic DNA of S. lividans TK24 by PCR using primers ExpWblI-BamHI and ExpWblI-HindIII. The PCR products were digested by BamHI and HindIII, and cloned into pWMH3-ermE Ã , a derivative of high-copy-number vector pWHM3. This contains the strong, constitutive ermE Ã promoter, an efficient ribosomal binding site and confers resistance to thiostrepton [34]. The constructed pWHM3-ermE Ã -wblI and the control empty vector pWHM3-ermE Ã were transformed into S. lividans TK24 protoplasts and transfornants were selected in the presence of thiostrepton at 50 μg/ml. A method of in-frame deletion [32] was used to construct the wblI deletion mutant. Two 1.5 kb DNA fragments flanking the wblI coding region were amplified from the S. lividans TK24 genomic DNA by PCR using primer pairs WblI-Up1/WblI-Up2, WblI-Down1/WblI-Down2. The resulting two PCR fragments were individually digested with the corresponding restriction enzymes, and subsequently cloned into pDH5 [32] cut by Hin-dIII and EcoRI using a triple ligation strategy, giving pDH5-ΔwblI. The resulting plasmid was used to delete wblI using the following procedure: protoplasts of S. lividans TK24 transformed with pDH5-ΔwblI were regenerated on R2YE medium without antibiotic selection at 28˚C until fully sporulating. Spores were harvested and plated on R2YE medium in order to have well isolated colonies. Once sporulated these well-separated colonies were replica plated on R2YE and R2YE containing thiostrepton (50 μg/ml) to identify the thiostrepton-sensitive colonies. The chromosomal structure of the wild-type strain and the thiostrepton-sensitive colonies was compared in the wblI region by PCR, using primers ExpWblI-NdeI and ExpWblI-X-hoI. The near-complete deletion of wblI gene in S. lividans TK24 was verified by the size of the amplified PCR products (396 bp for wild type strain, 96 bp for mutant strain).
approximately 50 mg of mycelium (dry weight) was collected and extracted by vortexing for 30 min at 4˚C in 1 ml methanol and in 1 ml KOH (1N) respectively. For RED assay, the extract was acidified to pH 2~3 by HCl and the OD 530nm was determined using a spectrometer (SHI-MADZU). To assay intracellular ACT, the OD 640nm of the extract was measured using the same spectrometer. For extracellular ACT determination, one-quarter of an 8 cm diameter plate of the agar R2YE medium was collected, smashed and immersed in 10 ml H 2 O for 24 h at 4˚C. The mixture was centrifuged and the resulting supernatant was transferred into a fresh tube and 10 ml KOH (1M) was added. After gentle inversion, 10 ml of HCl (3M) was added and the mixture incubated on ice for 10 min then centrifuged. The supernatant was discarded and the pellet containing ACT was re-suspended in 1 ml of KOH (1M) and the OD 640nm of the resulting solution was determined. Each sample was processed in triplicate.

Overexpression and purification of Histidine-tagged WblI in E. coli
The wblI coding sequence was amplified from the genomic DNA of S. lividans TK24 by PCR using primers ExpWblI-NdeI and ExpWblI-XhoI. The resulting DNA fragments were cloned into pET22b(+) cut by NdeI and XhoI, and the resultant plasmid transformed into E. coli BL21, a suitable host for protein expression. The resulting transformants were cultivated in LB medium supplemented with 50 μg/ml ampicillin at 37˚C until OD 600 reached 0.6. IPTG was then added at a final concentration of 1 mM, and incubation was pursued for 6 extra hours. The cells were collected by centrifugation at 12,000 g for 10 min. Then the cell pellet was resuspended in lysis buffer (50 mM Na 2 HPO 4 , pH 8.0, 500 mM NaCl) and sonicated (Hielscher Ultrasonics UP400S, 0.5 cycle and 20% amplitude) on ice until reaching complete homogeneity. After centrifugation at 14,000 rpm for 20 min, the cell extract was saved and passed through an Ni-NTA column (Cat. No. 30210; Qiagen) on a Biologic LP apparatus (Bio-Rad). The six-histidine-tagged WblI (His 6 -WblI) was purified to near homogeneity according to the manufacturer's instructions.

Isolation of the putative WblI interacting targets
The putative WblI binding targets were isolated as described previously [35]. Briefly, 100~200 pmol of purified His 6 -WblI was incubated with 10 pmol of S. lividans TK24 genomic DNA for 15 min at room temperature in the binding buffer, total volume of 1 ml. A negative control reaction that contained only genomic DNA, free of WblI, was carried out in parallel. The binding reaction was fixed by addition of 1 ml of crosslinking buffer (HEPES, NaCl, EDTA, 37% formaldehyde) and incubated at 37˚C for 10 min then at 4˚C for 1 h. Genomic DNA was sheared to an average size of 2~3 kb by sonication for 8 sec on ice (Hielscher Ultrasonics UP400S, 0.5 cycle and 20% amplitude). The complex of His 6 -WblI crosslinked with fragments of genomic DNA was isolated by passing through an Ni-NTA Agarose (Cat. No. 30210; Qiagen) column. The de-crosslinking of WblI-DNA was carried out by incubation overnight with 200 mM NaCl at 65˚C for 4 h and proteinase K (at a final concentration of 20 μg/ml). The decrosslinked DNA was recovered by EtOH precipitation. The resulting DNA pool was digested by Sau3AI, and subsequently cloned into pUC18 cut by BamHI. The inserted DNA fragments were then identified by nucleotide sequencing. The promoters present on the sequenced fragments and/or its flanking regions (within 2~3 kb) were retained for further analysis.

Electrophoretic mobility shift assay (EMSA)
The promoter regions of wblA, wblI, actVA3 and the intergenic region between actII-1 and actII-2 were amplified from S. lividans TK24 genomic DNA by PCR using primer pairs BS3579F/BS3579R, BS5046F/P5046R, BS5078F/BS5078R, BS82-83F/BS82-83R (Table 1), respectively. The resulting PCR products were 5' labeled with FITC by PCR using primer Plabel (Table 1). 50 pmol of each of the promoter regions was incubated with purified His 6 -WblI in various concentrations for 15 min at room temperature in binding buffer (10 mM Tris-HCl, 50 mM KCl, 1mM DTT, pH 7.5), total volume of 20 μl. For the competition assay, excess amounts of specific competitors of unlabeled wblI promoter or non-specific competitors of an unlabeled unrelated DNA probe were introduced. After incubation, the reaction mixtures were resolved on a 5% native polyacrylamide gel pre-run at 100 V for 30 min and run at 100 V for 90 min in a running buffer containing 45 mM Tris-HCl, pH 8.3, 45 mM boric acid, 10 mM EDTA. Visualization of the DNA signal was carried out by fluorescence imaging using a UVI Alliance 4.7 imager (UK).

Isolation of total RNA and RT-PCR
Streptomyces mycelia were collected from three independent R2YE plates at various time points and total RNA isolated using an RNAprep kit (TIANGEN, Cat. No. DP430) according to the manufacturer's instructions. Two micrograms of each RNA sample were used as a hrdB, an housekeeping gene encoding the major vegetative sigma factor whose expression is constant throughout growth, was used as an internal control to normalize the relative expression of each target gene.

WblI over-expression has a positive impact on RED and ACT biosynthesis
The production of RED and ACT was assayed in the original strain, S. lividans TK24, in the strain deleted for wblI and in derivatives of S. lividans TK24 carrying the empty plasmid pWHM3-ermE Ã or the plasmid pWHM3-ermE Ã -wblI. Results are shown in Fig 1. The overexpression of WblI, led to strong enhancement of the production of RED and ACT, the red hybrid peptide-polyketide and blue polyketide antibiotics usually weakly produced by this strain (Fig 1A) but had no impact on growth ( Figure A in S1 Appendix). Quantitative analysis showed that at 72 h, S. lividans TK24 over-expressing wblI produces 10 and 23 fold more RED and ACT, respectively than the control strain ( Fig 1B). Similar results were obtained when WblI of S. coelicolor (WblI SC ) was over-expressed (data not shown). WblI and WblI SC differ by one amino acid located outside of the WhiB helix turn helix domain ( Figure B in S1 Appendix). As anticipated, the deletion of wblI had little (or a slight negative) impact on the already weak production of these antibiotics in S. lividans TK24 (Fig 1). These results demonstrated the positive regulatory effect of WblI on ACT and RED biosynthesis in S. lividans TK24. In order to determine whether this effect was direct or indirect, attempts were made to isolate putative regulatory targets of WblI using in vitro formaldehyde crosslinking. In contrast, the over-expression of WblI was shown to be correlated with the reduced synthesis of a yellow pigment (Fig 2A) thought to be the metabolic product of the cpk gene cluster regulated by the γ-butyrolactone signaling molecule SCB1 [36][37][38]. We noticed that the spores generated by S. lividans TK24/pWHM3-ermE Ã -wblI remained white even upon prolonged incubation and never developed the usual grey color of the control strain S. lividans TK24/ pWHM3 on the GYM solid medium (Fig 2B). However, spores counts with Thomas cell or spores plating in dilution yielded similar number of spores for the two strains (data not shown). This indicated that the over-expression of WblI had little if any impact on morphological differentiation.

Isolation of the putative regulatory targets of WblI using in vitro formaldehyde crosslinking
In vitro formaldehyde cross-linking of the purified His 6 -WblI with its putative targets in the S. lividans chromosome was performed, followed by the covalent capture of WblI/DNA complex on a Ni-NTA column. After cross-linking reversal, the isolated DNA fragments were cloned into pUC18. Twenty-one clones were obtained and sequenced. Four putative target promoter regions (absent in the negative control, see materials and methods) were identified ( Table 2). Two of the targets belong to the ACT biosynthetic gene cluster. These include the region upstream of actVA3, a biosynthetic gene of the ACT cluster [39] and the intergenic region between two divergently located genes, actII-1 and actII-2. actII-1 encodes a TetR regulator known to repress the expression of the divergent gene actII-2 encoding the ACT transporter [40]. The other targets of WblI were WblI itself and another WhiB-like transcriptional regulator, WblA. The latter was shown to be essential for an early stage of aerial hyphal development [23] and to have a negative impact on the oxidative stress response [25,26] and antibiotic production [24].

WblI directly interacts with its regulatory targets in vitro
In order to confirm the direct interaction of WblI with its four putative targets, regions encompassing the promoter sequences predicted by the online Neural Network Promoter Prediction tool (www.fruitfly.org/seq_tools/promoter.html) were amplified and FITC-labeled by PCR and used as probes for EMSA. Results are shown in Fig 3A. The migration of the promoter fragments of wblI, wblA, actVA3, as well as the intergenic region between actII-1 and actII-2 were found to be retarded in the presence of purified WblI in a concentration dependent manner. In all cases, addition of an excess amount of unlabeled probes led to the fading or even disappearance of the shifted bands, whereas the competition assay was largely compromised when an irrelevant DNA probe (+1 to +378 relative to the translational start codon on wblI The four target promoter regions were inspected for the presence of conserved motifs ( Fig  3B). Three repeats of closely related sequences CTTCGAS (S standing for G or C) or CTTGA CGC were found upstream of the -35 region of actVA3. Their position is consistent with the activation of the expression of this gene by WblI. An perfect inverted repeat of the sequence (CACCGTTC TACAATG GAACGGTG) was found bracketing the -10 promoter region of ActII-1 (TetR regulator) consistent with the repression of this gene by WblI. In contrast, a single repeat of the sequence (CTTGACGC) that could not constitute an operator site was found overlapping the -35 promoter region of ActII-2 (Act transporter). An imperfect inverted repeat of the sequence CTTGACGC was found downstream of the -10 promoter sequence of wblI, suggesting a negative auto-regulation of WblI. Two direct repeats of a degenerated version of the sequence CTTCGAS (gTTCGgG CC aTTCaAG) were found between the P2 and P3 promoters  of WblA in overlap with a known negative regulatory site for AdpA [41]. Even though these putative binding sites should be confirmed by foot printing experiments, their localization fits the observed regulatory features of these genes by WblI.

Assessment of in vivo regulatory effects of WblI using quantitative RT-PCR
In order to assess in vivo, the impact of WblI on the expression of the target genes validated by EMSA, mRNA was prepared from the wild type strain of S. lividans TK24 carrying the empty vector or carrying the wblI over-expression plasmid or deleted for wblI, at 24, 48 and 72 h. The expression levels of wblA, actVA3, actII-1 and actII-2 were examined by qRT-PCR. Results are shown in Fig 4. In the wild type strain, the expression of genes of the ACT cluster (actVA3, actII-1 and actII-2) was enhanced at 48 h, a time point where ACT was detectable. At the three time points tested, the over-expression of WblI was correlated with the increased expression of the genes encoding the hypothetical protein ActVA3 and the ACT transporter ActII-2, while transcription of the divergently located TetR regulator ActII-1 and of WblA was reduced. Consistently, the deletion of wblI, at all three time points, was correlated with reduced expression of actVA3 and of actII-2 and with enhanced expression of actII-1 and wblA. Taken together with the above EMSA data, these results demonstrated that WblI positively regulates the transcription of actVA3 while negatively controlling that of actII-1 and wblA via direct interaction with their promoter regions.

Overexpression of wblI enhances ACT export
Since the TetR regulator ActII-1 that regulates negatively ActII-2, the ACT export system, was proposed as a WblI regulatory target, the ratio between extracellular (exported) and intracellular ACT was determined in strains of S. lividans TK24 carrying the empty vector (pWHM3-ermE Ã ) or the wblI overexpression plasmid (pWHM3-ermE Ã -wblI). This ratio was found to be 1.3 fold higher in the WblI overproducing strain than in the control strain indicating that WblI had a positive impact on ACT export (Table A in S1 Appendix).

Discussion
The WhiB-like transcriptional regulator [35], WblI (SCO5046), studied in this issue was first identified as a target of the transcriptional regulator of the TetR-family (SCO3201). The overexpression of SCO3201 was shown to lead to strong repression of both antibiotic production and sporulation in S. coelicolor [31]. However, since the disruption of SCO3201 did not lead to any obvious phenotype, it was proposed that SCO3201 "illegitimately" governs the expression of target genes normally under the control of other TetR regulators truly involved in the regulation of the differentiation process. It was thus inferred that the analysis of the "SCO3201" targets might lead to the identification of new players in the complex regulation of the differentiation process in S. coelicolor. Indeed, the search for SCO3021 targets allowed the identification of genes already known to be involved in the regulation of the differentiation process, validating this approach, as well as of new players including WblI (SCO5046).
In the present paper, we report the consequences of the over-expression of wblI on secondary metabolites production of S. lividans and the characterization of four of the putative regulatory targets of this regulator. The latter were identified using formaldehyde crosslinking followed by covalent capture of WblI/DNA complex on a Ni-NTA column, and were subsequently confirmed as direct targets by EMSA (Table 2, Fig 3) and qRT-PCR (Fig 4). Furthermore, an inspection of the promoter regions of the targets genes revealed some putative similar regulatory sequences. Altogether these investigations indicated that WblI represses its own transcription as well as that of the TetR regulator actII-1 that represses the expression of the divergent gene actII-2 encoding the ACT transporter [40]. qRT-PCR analysis confirmed that the transcription of actII-1 was reduced whereas that of actII-2 was enhanced in the WblI over-expressing strain. A putative WblI binding sequence was found to bracket the -10 region of actII-1 suggesting repression of the expression of the latter by WblI. Consistently an enhanced ACT export was correlated with WblI over-expression. This enhanced export ought to be correlated with an enhanced ACT biosynthesis. The latter might be achieved via the positive effect that WblI exerts on the expression of actVA3, the third gene of a succession of six genes of the ActVA3 region (Fig 4). Indeed, three repeats of the putative WblI operator sequence were found upstream of the -35 promoter region of actVA3, in a position expected for an activator site. The function of ActVA3 is unknown but we propose that it might constitute a bottleneck in the ACT biosynthetic pathway. In the wblI over-expressing strain, its enhanced biosynthesis would thus contribute to observed enhanced ACT biosynthesis.
In contrast, the over-production of WblI was correlated with a reduction of the biosynthesis of the yellow mycelial pigment and of the grey spore pigment encoded by the CPK [42] and whiE loci, respectively (Fig 2). The synthesis of grey spore pigment, regarded as a sign of completion of spore maturation, is directed by the whiE locus composed of an operon of seven genes and a gene transcribed in the opposite direction [43,44]. The transcription of genes of this locus have been shown to rely on the presence on six known whi genes (including whiB), which are required for sporulation septum formation [44]. However, none of the cpk or whiE genes were identified as putative WblI target genes in our formaldehyde cross linking experiments (Table 2). This suggests that the negative impact that WblI has on these genes may be indirect. ACT, yCPK and WhiE are all polyketides so they are likely to compete for a common precursor, acetyl-CoA. Enhanced ACT biosynthesis would thus lead to decreased acetyl-CoA availability and thus reduced yCPK and grey pigment synthesis. This view is supported by the fact that the wblI-deletion mutant did not have any obvious phenotype in relation to yellow pigment production or spore pigmentation (Fig 2).
Finally and most importantly, our RT-PCR experiments revealed reduced expression of the wblA gene in the WblI over-expressing strain (Fig 4). In S. coelicolor the expression of WblA was shown to be under the negative control of AdpA [41], a regulator playing a positive role in aerial mycelium development and ACT production. The promoter region of WblA is very complex comprising 3 promoters and 6 putative AdpA binding motifs [41]. Interestingly, two Regulatory targets of WblI, a WhiB family regulator direct repeats of a degenerated version of the putative WblI operator sequence CTTCGAS (gTTCGgG CC aTTCaAG) was noted between the P2 and P3 promoters of WblA (gTTCGgG CC aTTCaAG) in overlap of the binding site of AdpA (Fig 3B). This and our in vivo results suggested a negative regulation of WblA by WblI. WblA was shown to act as a negative regulator of ACT biosynthesis [24,41]. Consequently the repression by WblI, of WblA expression is consistent with enhanced ACT biosynthesis.
In Mycobacteria tuberculosis, Wbl proteins were shown to act as redox-sensing factors via their Fe-S clusters [45] and to trigger specific adaptive responses via their transcriptional regulator activity. The expression of the seven Wbl proteins (Whib 1-7) of this species was shown to be induced by various oxidative stresses [46]. Interestingly, in the WblA mutant of S. coelicolor, the oxidative stress response as well as RED and ACT biosynthesis were up-regulated [24][25][26]. This indicated that oxidative stress might be higher in this strain than in the original strain and suggested that oxidative stress might constitute an important signal to trigger antibiotic biosynthesis.
In summary, we provide in Fig 5 a schematic representation of the demonstrated WblI and WblA regulatory interactions. The strong enhancement that WblI over-expression has on ACT biosynthesis might result from the direct activation of ACT biosynthesis and export as well as to the repression of the expression of WblA, a likely indirect, repressor of ACT biosynthesis.
Supporting information S1 Appendix. (includes Table A, Figure A and Figure B). (DOCX)