Gone, 93 (1990)79-84 Elsevier
79
GENE03~i
P h o s p h a t e control of p a b S gene transcription during candicidin biosynthesis (Antibiotic biosynthesis; recombinant D N A; phosph ate regulation; transcriptional level; gene expression; PABA synthase; phosphate-deregulated mutants)
Juan A. Asturias, Paloma Liras and Juan F. Martin
Departamento de Ecologla, Gendlicay Microbiologia,Area de Microbiologla,Facultadde Biologla. Universidadde Le6n, 24071 Le6n (Spain) Received by F. Bolivar: 25 January 1990 Revised: 20 March 1990 Accepted: 23 March 1990
SUMMARY
The pabS gone of Streptomyces griseus IMRU3570 encodes the enzyme p-aminobenzoic acid synthase, which synthesizes p-aminobenzoic acid (PABA), a precursor of the antibiotic candicidin (Cd). The pabS transcript reached a peak at 12 h of incubation in batch cultures, preceding the formation of PABA synthase and the antibiotic itself. A decay of the pabS transcript was observed with an apparent half-life of 35 min. Inorganic phosphate (Pt; 7.5 mM) reduced the synthesis of the pab$ transcript, by 90-95 ~o, and consequently the formation of PABA synthase and Cd. Thirty rain after addition of 7.5 mM Pt, the cells synthesized only about 15~ as much pabS transcript compared to control cultures. However, Ps stimulated two- to threefold total RNA synthesis. The 1.7-kb pabS transcript shown by Northern hybridization was greatly reduced in amount in cells grown in 7.5 mM phosphate. Pi-deregulated mutants, described previously, were impaired in the transcriptional control exerted by Pi. It is concluded that Pi control of PABA synthase and Cd biosynthesis is exerted by repression of formation of the pab5 mRNA.
INTRODUCTION
Inorganic phosphate (P~) exerts a well known negative effect on the biosynthesis of secondary metabolites (Martin and Domain, 1980; Martin, 1989). Several antibiotic Correspondence to: Dr. J.F. Martin, Area de Microbiolosia, Facultad de
Biolosia, Universidad de Lebn, 24071 Le6n (Spain) Tel. (87)24.04.51, ext. 269; Fax (87)23.91.01. Abbreviations: bp, base pair(s); CA, ¢andicidin; kb, kilobase(s) or 1000 bp; nt, nucleotide(s); PABA, p-aminobenzoic acid; pab$, 8ene encoding PABS; PABS, PABA synthase; PD, phosphate-deregulated; P,, inorganic phosphate; Rif, rifampi©in; SDS, sodium dodecyl sulfate; SPG (soya peptone 25 mg per ml/glucose65 mg per ml 0.5 mM/ZnSO4); SSC, 0.15 M NaCI/0.015 M Na3" citrate pH 7.6; TE, 10 mM Tris. HCI pH 8.0/ I mM EDTA; TLC, thin-layer chromatography; TNS, 25 mM Tris. HCI pH 7.4/0.1 M NaCI/1% SDS/10 mM EDTA/6% Na. 4-aminosaiicilate/ 6% neutral phenol-chloroform; UV, ultraviolet; YED (yeast extract 10 mg per ml/glucose 10 mg per ml). 0378-1119/90/S03.50 © 1990Elsevier Science Publishers B.V. (Bior~dJcal Division)
biosynthetic enzymes have been found to be inhibited or repressed by Pi (reviewed by Martin, 1989). Biosynthesis of the polyene macrolide antibiotic Cd is an interesting model for studying this mechanism of control. Formation of Cd is suppressed by concentrations of Pi in the rmlge 1-10 mM (Liras et al., 1977). PABA, the precursor of the p-aminoacetophenone moiety of Cd, is formed from chorismic acid by the enzyme PABS. The synthesis of this enzyme is strongly repressed by Pi (Gil et al., 1985). The pabS gone of Streptomyces griseus IMRU3570, coding for PABS, was cloned in a 4.5-kb BamHI DNA fragment (Gil and Hopwood, 1983), Recently, we have subcloned an 114-bp DNA fragment from the upstream region of the pabS gene (Martin et al., 1988; Rebollo etal., 1989) containing a Prregulated promoter which appears to be involved in Pi control of Cd biosynthesis. Early studies on the effect of low concentrations of Pi on Cd biosynthesis in vivo suggested that the Pi regulation of
80
Cd synthesis was due to repression of de novo m R N A formation (Martin and Demain, 1976). Pi does not inhibit PABS activity at concentrations of 10raM (Gil et al., 1985). Hybridization of RNA transcripts with cloned genes could provide information on the level of transcription (Jones, 1985; Fisher and Wray, 1989; Horinouchi et al., 1989). It was, therefore, of interest to elucidate whether Pi control of Cd biosynthesis comes about through repression of the formation of PABS transcripts. The aim of the present study was to understand the regulation of transcription of the pabS gene by hybridization of total S. gr/seus RNA with an internal fragment of this gene, and to determine the role of Pi in control of PAB S at the transcriptional level.
A
100 75
CC
RESULTS AND DISCUSSION (a) Sequential synthesis of pabS transcript, PABS and Cd Dot-blot hybridizations of total R N A with the Sail pabS probe were carried out to analyze pabS m R N A synthesis by S. gr/seus during the growth phase and Cd production stage. Total R N A was isolated from mycelium at different times of growth in SPG medium with and without adch'tion of Pi. pabS m R N A was quantified as described in Fig. 1. A rapid increase in the amount of pabS transcript formed between 0 and 12 h fermentation was found. There was a sequential formation of the pabS transcript, PABS, and Cd (Fig. 1). The increase in pabS transcript preceded the rise in PABS activity that continued for another 12 h, reaching its maximum 24 h after inoculation. Between 12 and 24 h, the concentration of pabS m R N A declined, even though PABS specific activity continued increasing during this period. The onset of Cd production took place at 18 h, and synthesis ofthe antibiotic continued at a lower rate up to 96 h (not shown in Fig. 1). These
5o
E
B
18
C
400
30O . z
IO0 m,
12
24
36
48
TIME (h)
Flu. I. Time.course of the synthesis of the pubs transcript (panel A), PABS (panel B) and Cd (panel C) in cultures of$. gr/.veusIMRU3570 in SPO medium with (dk) and without (O) supplementation with 7.5 mM Nu. phosphate buffer. $. gr/aem IMRU3570 was pown and Cd was quantified as described previously(Martin and McDaniel, 1975yPABS
activity was assayed in vitro by diazotization of PABAformed(Gil et al., 1985).RNAwas isolated from50 ml of$. grOeuscultures in SPG medium (about 250 m8 of cell dry weight). Mycelinmwas collected by centrifugation, washed immediatelywith chilled distilled water and transferred quickly to -200C. Frozen cells were thawed in 2ml of P buffer (Hopwood et al., t985), treated with l0 mffmi oflysozymefor $ mittand then with 2 ml of 2 × TNS solution, and mixed with 1.5 ~ml of 81ass beads (0.45-0.$$tam, Sigma). The lysls was completed by vortexing seven times (30 s each) in ice and the iysate deproteinized by repeated phenol extraction. The RNA was then precipitated with ethanol, redissolved and reprecipitated selectivelyin 3 M Na. acetate (pH 6.0) twice. The final pellet was dissolved in 100 mM Na. acetate, $ mM M8SO4 (pH 5.0) and incubated for I h at room temperature with RNase-free DNase (20 units/ml). After extracting once with phenol and twice with chloroform, the RNA was precipitated with ethanol and redissolved in TE buffer. The RNA concentration of the different solutions was established from A,~o determinations (Sambrook et al., 1989). Aliquots of RNA were stored as ethanol-precipitated material at -20°C. All the solutions were treated with 0.1~ diethyipyrocarbonate. AHquots of ethanol.precipitated total RNA (about 75 tAB)were collected by centrifugation and vacuum-dried. The samples were dissolved in glyoxalphosphate solution (Thomas, 1983),heated for I h at 50°C and then cooled in ice. Dilutions were made in 0.1~ SDS and then spotted onto a dry nitrocellulose filter that had been soaked in water, equilibrated with 20 x SSC for 5 rain and dried under a lamp. Filters were baked for 2 h at 80°C under vacuum and heated to 100°C for 7min in 20raM Tris. HC!pH 8.0 to removeresidual$1yoxal.The probe used for quantification of pab$ mRNA was a 1.6.kb Sail DNA frasment internal to the pab$ sane purified from low melting point ngarose followingelectrophoresis (Lansridge et al., 1980).The purified DNA fragmentwas nick. translated with the nick-translation kit from Boehri~er (Matmheim, FRO) using [a-nP]dCTP as described by the manufscturer. The blots were exposed to AmershamX-rayfilm at -?0°C for 2-4 days, using an int~sifyin8 screen. Autorsdingrams were scanned with a dual wavelength Shimadzu CS-930TLC scanner as densitometer.
81 results suggest that the limiting step for prolonged Cd biosynthesis is the PABS activity which is known to decay irreversibly after 24 h (Fig. 1). The decay of PABS activity is probably due to the rapid degradation of the transcript of the pabS gene. The period of time during which high levels of PABS exist follows closely the time during which high levels of pabS transcript occur in the cells. (h) Regulation by Pt of pabS transcript formation The effect of Pi (7.5 mM) on the kinetics of formation of the pabS transcript, PABS and Cd is shown in Fig. 1. Densitometric analysis of RNA dot-blot hybridizations from cells grown in the presence or absence of P~ indicates that a 90-95% repression of the pabS transcript occurred at 12 h of fermentation (Fig. IA). Cd formation was completely blocked in cells grown in the presence of 7.5 mM Pi (Fig. IC), which correlates well with the decrease of PABS activity in Pi-supplemented cultures (Fig. 1B).
(e) P, stimulates total RNA synthesis To study if the decrease of pabS transcript observed in the presence of Pi was due to a nonspecific decrease of total RNA, the synthesis of total RNA was quantified by following the incorporation of [ t4C]uracil in cultures grown in the presence and absence of Pi (Fig. 2). The synthesis of total RNA (mainly ribosomal RNA during the rapid growth phase) was enhanced two- to threefold when the culture medium was supplemented with 7.5 mM Pt (Fig. 2). This result suggests that the repression exerted by P, on the transcription of the pabS gene is specific. There was very little total RNA synthesis in antibioticproducing cells after 24 h of incubation during the idiophase (antibiotic production phase).
(d) Time kinetics of the transcriptional control In order to establish how quickly the culture adjusts the transcriptional machinery in response to P~ supplementation, short time experiments of repression by Pi of the pabS mRNA were carried out. Pi (7.5 mM) was added to a 9-h-old culture, when the pabS mRNA was being intensively synthesized. The RNA was extracted from samples taken every 15 min and hybridized with the pabS gene probe. Results (Fig. 3) show that the transcriptional control exerted by Pi is rapid, since 30 rain after addition of P~, the cells synthesized only about 15~ ofpabS mRNA as compared to control cultures. However, a small increase of pabS mRNA was observed during the initial 15 rain following Pi addition, which suggests that the cell has to build up a substantial concentration of the intracellular effector before exerting negative control at the transcriptional level. (e) Half-life of the pabS transcript The pattern of change of the pabS transcript levels in Cd-producing S. gr/seus cultures suggested that a ~'apid decay of mRNAs specific for antibiotic biosynthetic enzymes occurs both in long-term fermentations and shortterm incubations (Figs. I and 3). To test this hypothesis, 10 pg Rif/ml was added to a 9-h-old culture. This concentration of Rif produced an 87~o inhibition of total RNA synthesis in 10 min under our experimental conditions. The decay of the pabS transcript levels following inhibition of de novo RNA synthesis is shown in Fig. 4. The apparent half-life of the pabS transcript in SPG medium was about 35 min. The apparent half life of labS transcript after 30 min of Pi addition does not seem to change significantly as compared to that determined by inhibition of RNA synthesis with Rif, i.e., levels were reduced by 50% in 30 min (Fig. 3),
100
~o
0
75
e 30
| so
~ 20
E
\
0 z
g
25
i
12
2/,
36
/.,8
60
TIME (h)
Fig. 2. Stimulation by P, ofthe rate oftotal RNA synthesis. Cultures were grown in SPG medium with ( • ) and without (@) supplementation with 7.5 mM Na.phosphate buffer. The incorporation of [=4C]uracil into RNA was measured at different times using pulses of 6 min each, as described by Liras et al. (1977).
15
t
t
30
45
t /~~ 60
150
TIME (rain)
Fig. 3. Time kinetics of the effect of 7.5 mM Na" phosphate buffer on synthesis of the pabS transcript in short-term experiments. Cultures in SPG medium at 9 h of incubation (rapid growth phase) were supplemented with Pi (arrow)(•). Control unsupplemented cultures (@). pab$ mRNA was quantified as indicated in the legend to Fig. I.
82
1
2
100 --
11
|
75 !
.
2 . 3 ~,-
so
iI
1 . 8 P,-
90 ' TIME (rnin)
6'0
120 '
150
Fig. 4. Kinetics of decay ofthe pabStranscript orS. grbeus cells ~own for 9 h in SPG medium after inhibition of de novo IUqA synthesis with 10 pg Rif/ml (Martin et al., 1977). Total RNA ,,,as isolated at different times following rifampicin addition. The app~ent half-life of the pabS transcript [calculated by using the least squares-fitted plots of the logarithmic value of the dots' intensity against time (Merino et al., 1987)] was 35 min.
but the low levels of the pabSgene transcript formed in the presence of Pl did not allow us to calculate precisely the half-life of this transcript under Pt supplementation conditions. (f) Size of the pab$ transcript The size of the pab$ transcript from $.g~sevs IMRU3$70 was determined by Northern analysis using RNA extracted as indicated in Fig. 1 legend. A main hybridization region of about 1.7 kb was observed (Fig. 5). This size correlates well with the size of the pabS gene (Rebollo et al., 1989; L.M. Criado, J.A. Gil and J.F.M., unpublished), and suggests that the pab$gene is transcribed in rive as a monocistronic mRNA. The hybridization intensity of the RNAs extracted from cells grown with and without 7.5 mM Pt supports the conclusion that Pt control is exerted at the transcriptional level
(Fig.5). (g) Prderegulated mutants are impaired in the transcriptional control Mutants partially or completely insensitive to Pi control of Cd biosynthesis [Pt-deregulated (PD) mutants] have been reported before [Martin et al., 1979). Two low Cdproducing mutants (LP5 and LP34)showed 27~ and 45 Yo production of Cd respectively as compared to the parental strain, as well as lower sensitivity to Pt control of Cd biosynthesis. LP34 is a natural revertant of a nonproducing
Fig. 5. Northern analysis of the RNA of $. Av~seuscells gown in SPG medium without (lane I) or with (lane 2) supplementation with 7.5 mM sodium phosphate bufl'er. RNA samples (30 pg) used for Northern analysis were denatured with glyoxal as described in the legend to Fig. I and
fractionated on agarose (1.1%) gel. The gel was blotted onto a nitrocellulosefilterby standard methods(Sambrooket al., 1989)and baked in a vacuum oven at 80'C for 2 h. Prehybridizationswere at 420C overnightin 50% formamide,5 x SSC,50 mM sodiumphosphatebuffer pH 6.5, 250#g/ml denatured salmon sperm DNA and 10 x Denhart's solution. Filters were hybridized in the same bulbr containing 10~ dextran sulfate at 42°C for 24 h. Afterwards,filters were washed four timesin 2 x SSC, 0.1~ SDS at 5o°Cfor 15min(Williamsand Masoon0 1985). The arrows indicate single.stranded DNA size markers in kb (Bf/l-digestedpBIU22denaturedby treatmentwith81yoxalin the same way as the RNA samples).
mutant. Both of them have a normal morphology and the same pigmentation markers as the parental strain. Formarion of PABS and synthesis of Cd by these mutants was partially insensitive to P, control (Fig. 6a,b). Formation of the pab$ transcript in these PD mutants is repressed to a lesser extent (58~o in LP$ and 7~0 in LP34, respectively) than in the parental strain (90~) when SPG medium was supplemented with 7.5 mM Pt (Fig. 6c). These results indicate that Pt-deregulated mutants are impaired in the control exert~ by Pi at the transcriptional level. There was an apparent lack of correlation between the low effect exerted by Pi on pab$ mRNA formation by mutant LP34 (7% repression) and the 4 5 ~ reduction in PABS activity
83
100 •, CmcOgmm
• P~IIAN
_r~me6wtJ~tcpt
?5
i
I
50
!
Wt LI~ LP~
wt
LP5 LP~
wt
LI~ LP3¢
Fig.6. P, inhibitionofthe synthesisof C..d(panel a), PABS(panelb) and the pabStranscript (panelc) in the wildtype (wt) $. grtveusIMRU3570 strain and in LP5 and LP34Prderegulatedmutants (Martinet al., 1979; 1984)~The culturesweregrownin SPG mediumwithor withoutsupplementationwith 7.5 mM Na. phosphatebufferand resultsare givenas the percentageofinhibitionexertedby P,. Samplesfor pab$RNAweretaken at 12h, for PABS at 24 h and for Cd assays at 36 h ff incubation.
observed in the same strain. These results indicate that an additional level of regulation (probably at the translational or posttranslational level) exists in this mutant. (h) Conclusions (1) Results indicate that transcription of the labs gene during Cd biosynthesis peaks at 12 h of growth, preceding the formation of PABS and the antibiotic itself. These results confirm previous observations on the sequential formation of macromolecules during Cd production (Liras et al., 1977). (2) Formation of pab5 transcript in 5. griseus IMRU3570 is repressed by Pi, as determined by RNA dot-blot hybridization or by Northern analysis of the transcripts (Figs. 1, 3 and 5). Despite repression ofthe pabS gene, total RNA synthesis clearly increased in Prsupplemented cultures, suggesting that ribosomal RNA genes and other genes required for the increased growth (Liras etal., 1977) (e.g., primary metabolism) are expressed more efficiently in Prsupplemented cultures. Expression of the pabS gene (and very likely other antibiotic biosynthetic genes) is, therefore, controlled differentially by the Pt concentration in the culture medium. (3) Pt exerted a severe repression of labs transcript formation (85~ after 30 min) but little or no significant effect was observed during the first 15 min following the addition of P~ to the culture. Since P~ uptake is very rapid (Martin et al., 1977) the time required to begin P~ repression may be due to the need to adjust the transcriptional machinery to repress transcription of the labs gene.
(4) Apparent half-life of the PUbStranscript in Cd-producing cultures as measured by dot-blot hybridization is 35 min. Decay of the pubs transcript explains previous observations of the decrease of the rate of antibiotic biosynthesis in batch cultures (Martin and Demain, 1980). The levels of pubs mRNA need to be replenished continuously for prolonged biosynthesis of Cd. Nothing is known about the mechanisms of mRNA degradation following translation in Streptomyces. Decay rate ofthe pubs transcript in Pi-supplemented cultures does not seem significantly different from decay in control unsupplemented cultures, i.e.,-Pi control is exerted mainly through control of transcription initiation rather than by a change in the rate of mRNA degradation. (5) Prderegulated mutants are impaired in the transcriptional control exerted by Pi. This result suggests that these mutants may have a mutation in the Pi control sequence existing upstream from the Pubs gene (ReboHo et al., 1989) or in a putative protein-effector complex that could bind to this sequence (Martin, 1989). (6) The biosynthesis of antibiotics and other secondary metabolites is probably controlled at the transcriptional level by a general mechanism (Martin and Liras, 1989). Carbon catabolite regulation of phenoxazinone synthase, an enzyme involved in actinomycin biosynthesis, is also exerted at the transcriptional level (Jones, 1985). Similarly, nitrogen regulation of giutamine synthetase in Smeptomyces coelicolor (Fisher and Wray, 1989) and the control of actinorhodin formation by the product of the afsB gene (Horinouchi et al., 1989) are exerted at the transcriptional level. The cascade of events at the molecular level that produce the 'switch-on' of the pab$ promoter following Pt depletion in SPG medium (Liras et al.0 1977; Gil et al., 1985), and the switch-off of the Pubs gene when the producing medium is supplemented with Pi, is still unknown. Two mechanisms could be involved in transcriptional regulation. One is based on transcriptional selectivity of Pt-regulated promoters by RNA polymerases with different sigma factors (Westpheling et al., 1985; Buttner et al., 1988; Takahashi et al., 1988; Westpheling and Brawner, 1989; Buttner, 1989). An alternative mechanism is the hwolvement of regulatory proteins that bind to upstream regions ofcertain actinomycetes promoters (Geistlich et al., 1989; Lin and Wilson, 1988). Positive regulatory proteins are involved in the control of the Pi regulon in E. coli(Kimura et al., 1989) suggesting that this may also be the case in Streptomyces.
ACKNOWLEDGEMENTS J.A. Asturias was supported by a fellowship of the Ministry of Education and Science, Madrid, Spain. We
84 thank J.A. Gil for providing plasmid plJ819, C. Esmahan and S. Baumberg for reading the manuscript, and M.P. Pumas, R. Barrientos, B. Martin, S. Llamas and M.I. Corrales for valuable technical assistance.
REFERENCES Buttner, MJ,: RNA polymerase heterogeneity in Streptomycescoellcolor A3(2), MUl. Microbiol. 3 (1989) 1653-1659. Buttner,MJ., Smith, A.M. and Bibb, MJ.: At leastthree differentRNA polymerasc huloenzymas direct transcription of the agarase gene (dagA) of Streptomyces coellcolorA3(2). Cell 52 (1988) 599-609. Fisher, S.H. and Wray Jr.,L.U.: Regulation of glutamine synthetase in Streptomycer coellcolor.J. BacterioL 171 0989) 2378-2383. Oeistlich, M., Irniger, S, and HOtter, R.: Localization and functional analysisof the regulated promoter from the Streptomycesglaucescens reeloperon. Mol. Microbiol. 3 (1989) 1001-1069. Oil, J.A. and Hopwood, D.A.: Cloning and expression of a p-aminobenzoic acid synthetase gone of the candicidin-producing Streptomycesgrlsesa.Gene 25 (1983) 119-132, Gil, J.A.,Nahm'ro, G., Villanueva,J.R. and Martin, J.F.:Characterization and regulation of p-aminobenzoic acid synthase from Sweptomyces 8r/sem. J. Gen. Microbiol. 131 (1985) 1279-1287. Hopwood, D.A., Bibb, MJ., Chater, K.F., Kieser, T., Bruton, CJ., Kieser, H.M., Lydiate, DJ., Smith, C.P., Ward, J.M. and Schrempf, H.: Genetic Manipulation of Streptomyces: A Laboratory Manual. John innes Foundation, Norwich, 1985. Horinouchi, S., Maipartida, F., Hopwood, D.A. and Beppu, T.: q~B stimulates transcription of the actinorhodin biosynthetic pathway in $~ptomyceJ coeltcolorA3(2) and $1r,ptomyccs ltvMans, Mol. Gen. Goner. 215 (1989) 355-357. Jones, G.H.: Regulation of phenoxuinone synthase expression in $~ptomyc¢a anttbtottcla,J. Bacteriol, 163 (1985) 1215-1221. Kimura, S., MakinG, K.. Shinqawa, H,, Amemura, M. and Nakata, A.: Regulation ofthe phosphate regulon of£schev~chtacolt: Characterization of the promoter of the pats sane. MoLOen. Goner, 215 (1989) 374-380. Lanipridse, J., LanFidge, P. and Bergquist, P.L.: Extraction of nucleic acids from agarose gels, Anal, Biochem. 103 (1980) 264-271, Lin, E. and Wilson, D.8.: Identification era colE,binding protein and its potential role in itlduction ofthe colE'gonein ~evmomonosporafmca. J. Bacteriol. 170 (1988) 3843-3846, Liras, P., Villanueva, J,R. and Martin, J,F,: Sequential expression of macromolecule biosynthesis and candicidin formation in $tmptomyce~ grlaeua, J. Oen, Microbiol, 102 (1977) 269-277. Martin, J.F,: Molecular mechanisms of the control by phosphate of the biosynthesis of antibiotics and other secondary metabolites, In Shapiro, S. (Ed,), Regulation of Secondary Metabolism in Actinomycetes. CRC Press, Boca Rat&n, FL, 1989, pp. 213-237,
Martin, J,F. and Domain, A.L,: Control by phosphate of candi~,~din production. Biochem, Biophys. Res. Commun. 71 (1~7o) 1103-1109. Martin, J.F. and Demain, A.L: Control of antibiotic biosynthesis. Microbiol. Rev. 44 (1980) 230-251. Martin, J.F. and Liras, P.: Organization and expression of genes involved in the biosynthesis of antibiotics and other secondary metabofites. Annu. Rev. Microbiul. 43 (1989) 173-206. Martin, l.F. and McDaniel, L.E.: Kinetics of biosynthesis of polyene macrolide antibiotics in batch cultures: carl maturation time. Biotechnul. Biocng. 17 (1975) 925-938. Martin, J.F., liras. P. and Dcmain, A.L.: Inhibition by phosphate ofthe activity of candicidin synthases. FEMS MicrobioL Left, 2 (1977) 173-176. Martin, J.F., Naharro, G., Liras, P. and Villanueva, J.R,: Isolation of mutants deregulated in phosphate control ofcandicidin biosynthesis. J. Antibiot. 32 (1979) 600-606. Martin, J.F., Alegre, M.T., Castro, J.M. and Liras P.: Extrachromosomal genetic elements that control specific enzymes involved i~ antibiotic biosynthesis: possible involvement of an intraceilular pleiotropic effector. In Ortiz, L., BojalH,L.F. and Yakuleff, V. (Eds.), Biological, Biochemical and Biomedical Aspects of Actinomycetc~. Academic Press, Orlando, FL, 1984. Martin, J.F., Daza, A., Asturias, J.A., Gil, 5.A. and Liras, P.: Transcrip. tional control of antibiotic biosynthesis at phosphate-regulated promoters and cloning of a gone involved in the control of the expression of multiple pathways in Stmptomyces. In Okami, Y., Beppu, T. and Ogawara, H. (Eds.), Biologyof Actinomycetes.Japan Science Society Press, Japan, 1988, pp, 424-430. Merino, E., Becerril, B., Valle, F. and Bolivar, F.: Deletion era repetitive extragenic palindromic (REP) sequence downstream from the struc, tural gone of£sc~¢ri~ia coilglutamate dehydrogenase affects stability of its mRNA. Gone 58 (1987) 305-309. Rebollo, A., Oil, J.A., Liras, P., Asturias, J.A. and Martin, J.F.: Cloning and characterization of'a phosphate-regulated promoter involved in phosphate control of candicidin biosynthesis. Gone 79 (1989)47-58. Sambrook, J., Fritsch, E.F. and Maniatis, T.: Molecular Cloning. A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spri~ Harbor, NY, 1989, Takahashi, H., Tamaka, K, and Sh0na, T,: Genetic constituent ofrpoD 8one homologues in $1reptomycesstrains, in Okami, Y., Beppu, 1".arid Ogawara, H. (Eds.), Biologyof Actinomycetes.Japan Science Society Press, Japan, 1988, pp. 58-63. Thomas, P.S.: Hybridization of denatured RNA transferred or dotted to nitrocellulose paper. Methods Enzymul. 100 (1983) 255-266. Westpheling,J. and Brawner, M.: Two transcribinl| activities are involved in expression of the $~reptomyces8aiactose operon. J. Bacteriol. 171 (1989) 1355-1361. Westphelina, J., Ranes, M. and Losick, R.: RNA polymerase betero8eneity in Streptomyces coeUcolor.Nature 313 (1985) 22-27. Williams, J.F. and Masoon, PJ.: Hybridization in the analysis of RNA. In Haines, B.D. and Hiuins, SJ. (Eds.), Nucleic Acid Hybridization: A Practical Approach. IRL Press, Oxford, 1985, pp. 139-180.
79
GENE03~i
P h o s p h a t e control of p a b S gene transcription during candicidin biosynthesis (Antibiotic biosynthesis; recombinant D N A; phosph ate regulation; transcriptional level; gene expression; PABA synthase; phosphate-deregulated mutants)
Juan A. Asturias, Paloma Liras and Juan F. Martin
Departamento de Ecologla, Gendlicay Microbiologia,Area de Microbiologla,Facultadde Biologla. Universidadde Le6n, 24071 Le6n (Spain) Received by F. Bolivar: 25 January 1990 Revised: 20 March 1990 Accepted: 23 March 1990
SUMMARY
The pabS gone of Streptomyces griseus IMRU3570 encodes the enzyme p-aminobenzoic acid synthase, which synthesizes p-aminobenzoic acid (PABA), a precursor of the antibiotic candicidin (Cd). The pabS transcript reached a peak at 12 h of incubation in batch cultures, preceding the formation of PABA synthase and the antibiotic itself. A decay of the pabS transcript was observed with an apparent half-life of 35 min. Inorganic phosphate (Pt; 7.5 mM) reduced the synthesis of the pab$ transcript, by 90-95 ~o, and consequently the formation of PABA synthase and Cd. Thirty rain after addition of 7.5 mM Pt, the cells synthesized only about 15~ as much pabS transcript compared to control cultures. However, Ps stimulated two- to threefold total RNA synthesis. The 1.7-kb pabS transcript shown by Northern hybridization was greatly reduced in amount in cells grown in 7.5 mM phosphate. Pi-deregulated mutants, described previously, were impaired in the transcriptional control exerted by Pi. It is concluded that Pi control of PABA synthase and Cd biosynthesis is exerted by repression of formation of the pab5 mRNA.
INTRODUCTION
Inorganic phosphate (P~) exerts a well known negative effect on the biosynthesis of secondary metabolites (Martin and Domain, 1980; Martin, 1989). Several antibiotic Correspondence to: Dr. J.F. Martin, Area de Microbiolosia, Facultad de
Biolosia, Universidad de Lebn, 24071 Le6n (Spain) Tel. (87)24.04.51, ext. 269; Fax (87)23.91.01. Abbreviations: bp, base pair(s); CA, ¢andicidin; kb, kilobase(s) or 1000 bp; nt, nucleotide(s); PABA, p-aminobenzoic acid; pab$, 8ene encoding PABS; PABS, PABA synthase; PD, phosphate-deregulated; P,, inorganic phosphate; Rif, rifampi©in; SDS, sodium dodecyl sulfate; SPG (soya peptone 25 mg per ml/glucose65 mg per ml 0.5 mM/ZnSO4); SSC, 0.15 M NaCI/0.015 M Na3" citrate pH 7.6; TE, 10 mM Tris. HCI pH 8.0/ I mM EDTA; TLC, thin-layer chromatography; TNS, 25 mM Tris. HCI pH 7.4/0.1 M NaCI/1% SDS/10 mM EDTA/6% Na. 4-aminosaiicilate/ 6% neutral phenol-chloroform; UV, ultraviolet; YED (yeast extract 10 mg per ml/glucose 10 mg per ml). 0378-1119/90/S03.50 © 1990Elsevier Science Publishers B.V. (Bior~dJcal Division)
biosynthetic enzymes have been found to be inhibited or repressed by Pi (reviewed by Martin, 1989). Biosynthesis of the polyene macrolide antibiotic Cd is an interesting model for studying this mechanism of control. Formation of Cd is suppressed by concentrations of Pi in the rmlge 1-10 mM (Liras et al., 1977). PABA, the precursor of the p-aminoacetophenone moiety of Cd, is formed from chorismic acid by the enzyme PABS. The synthesis of this enzyme is strongly repressed by Pi (Gil et al., 1985). The pabS gone of Streptomyces griseus IMRU3570, coding for PABS, was cloned in a 4.5-kb BamHI DNA fragment (Gil and Hopwood, 1983), Recently, we have subcloned an 114-bp DNA fragment from the upstream region of the pabS gene (Martin et al., 1988; Rebollo etal., 1989) containing a Prregulated promoter which appears to be involved in Pi control of Cd biosynthesis. Early studies on the effect of low concentrations of Pi on Cd biosynthesis in vivo suggested that the Pi regulation of
80
Cd synthesis was due to repression of de novo m R N A formation (Martin and Demain, 1976). Pi does not inhibit PABS activity at concentrations of 10raM (Gil et al., 1985). Hybridization of RNA transcripts with cloned genes could provide information on the level of transcription (Jones, 1985; Fisher and Wray, 1989; Horinouchi et al., 1989). It was, therefore, of interest to elucidate whether Pi control of Cd biosynthesis comes about through repression of the formation of PABS transcripts. The aim of the present study was to understand the regulation of transcription of the pabS gene by hybridization of total S. gr/seus RNA with an internal fragment of this gene, and to determine the role of Pi in control of PAB S at the transcriptional level.
A
100 75
CC
RESULTS AND DISCUSSION (a) Sequential synthesis of pabS transcript, PABS and Cd Dot-blot hybridizations of total R N A with the Sail pabS probe were carried out to analyze pabS m R N A synthesis by S. gr/seus during the growth phase and Cd production stage. Total R N A was isolated from mycelium at different times of growth in SPG medium with and without adch'tion of Pi. pabS m R N A was quantified as described in Fig. 1. A rapid increase in the amount of pabS transcript formed between 0 and 12 h fermentation was found. There was a sequential formation of the pabS transcript, PABS, and Cd (Fig. 1). The increase in pabS transcript preceded the rise in PABS activity that continued for another 12 h, reaching its maximum 24 h after inoculation. Between 12 and 24 h, the concentration of pabS m R N A declined, even though PABS specific activity continued increasing during this period. The onset of Cd production took place at 18 h, and synthesis ofthe antibiotic continued at a lower rate up to 96 h (not shown in Fig. 1). These
5o
E
B
18
C
400
30O . z
IO0 m,
12
24
36
48
TIME (h)
Flu. I. Time.course of the synthesis of the pubs transcript (panel A), PABS (panel B) and Cd (panel C) in cultures of$. gr/.veusIMRU3570 in SPO medium with (dk) and without (O) supplementation with 7.5 mM Nu. phosphate buffer. $. gr/aem IMRU3570 was pown and Cd was quantified as described previously(Martin and McDaniel, 1975yPABS
activity was assayed in vitro by diazotization of PABAformed(Gil et al., 1985).RNAwas isolated from50 ml of$. grOeuscultures in SPG medium (about 250 m8 of cell dry weight). Mycelinmwas collected by centrifugation, washed immediatelywith chilled distilled water and transferred quickly to -200C. Frozen cells were thawed in 2ml of P buffer (Hopwood et al., t985), treated with l0 mffmi oflysozymefor $ mittand then with 2 ml of 2 × TNS solution, and mixed with 1.5 ~ml of 81ass beads (0.45-0.$$tam, Sigma). The lysls was completed by vortexing seven times (30 s each) in ice and the iysate deproteinized by repeated phenol extraction. The RNA was then precipitated with ethanol, redissolved and reprecipitated selectivelyin 3 M Na. acetate (pH 6.0) twice. The final pellet was dissolved in 100 mM Na. acetate, $ mM M8SO4 (pH 5.0) and incubated for I h at room temperature with RNase-free DNase (20 units/ml). After extracting once with phenol and twice with chloroform, the RNA was precipitated with ethanol and redissolved in TE buffer. The RNA concentration of the different solutions was established from A,~o determinations (Sambrook et al., 1989). Aliquots of RNA were stored as ethanol-precipitated material at -20°C. All the solutions were treated with 0.1~ diethyipyrocarbonate. AHquots of ethanol.precipitated total RNA (about 75 tAB)were collected by centrifugation and vacuum-dried. The samples were dissolved in glyoxalphosphate solution (Thomas, 1983),heated for I h at 50°C and then cooled in ice. Dilutions were made in 0.1~ SDS and then spotted onto a dry nitrocellulose filter that had been soaked in water, equilibrated with 20 x SSC for 5 rain and dried under a lamp. Filters were baked for 2 h at 80°C under vacuum and heated to 100°C for 7min in 20raM Tris. HC!pH 8.0 to removeresidual$1yoxal.The probe used for quantification of pab$ mRNA was a 1.6.kb Sail DNA frasment internal to the pab$ sane purified from low melting point ngarose followingelectrophoresis (Lansridge et al., 1980).The purified DNA fragmentwas nick. translated with the nick-translation kit from Boehri~er (Matmheim, FRO) using [a-nP]dCTP as described by the manufscturer. The blots were exposed to AmershamX-rayfilm at -?0°C for 2-4 days, using an int~sifyin8 screen. Autorsdingrams were scanned with a dual wavelength Shimadzu CS-930TLC scanner as densitometer.
81 results suggest that the limiting step for prolonged Cd biosynthesis is the PABS activity which is known to decay irreversibly after 24 h (Fig. 1). The decay of PABS activity is probably due to the rapid degradation of the transcript of the pabS gene. The period of time during which high levels of PABS exist follows closely the time during which high levels of pabS transcript occur in the cells. (h) Regulation by Pt of pabS transcript formation The effect of Pi (7.5 mM) on the kinetics of formation of the pabS transcript, PABS and Cd is shown in Fig. 1. Densitometric analysis of RNA dot-blot hybridizations from cells grown in the presence or absence of P~ indicates that a 90-95% repression of the pabS transcript occurred at 12 h of fermentation (Fig. IA). Cd formation was completely blocked in cells grown in the presence of 7.5 mM Pi (Fig. IC), which correlates well with the decrease of PABS activity in Pi-supplemented cultures (Fig. 1B).
(e) P, stimulates total RNA synthesis To study if the decrease of pabS transcript observed in the presence of Pi was due to a nonspecific decrease of total RNA, the synthesis of total RNA was quantified by following the incorporation of [ t4C]uracil in cultures grown in the presence and absence of Pi (Fig. 2). The synthesis of total RNA (mainly ribosomal RNA during the rapid growth phase) was enhanced two- to threefold when the culture medium was supplemented with 7.5 mM Pt (Fig. 2). This result suggests that the repression exerted by P, on the transcription of the pabS gene is specific. There was very little total RNA synthesis in antibioticproducing cells after 24 h of incubation during the idiophase (antibiotic production phase).
(d) Time kinetics of the transcriptional control In order to establish how quickly the culture adjusts the transcriptional machinery in response to P~ supplementation, short time experiments of repression by Pi of the pabS mRNA were carried out. Pi (7.5 mM) was added to a 9-h-old culture, when the pabS mRNA was being intensively synthesized. The RNA was extracted from samples taken every 15 min and hybridized with the pabS gene probe. Results (Fig. 3) show that the transcriptional control exerted by Pi is rapid, since 30 rain after addition of P~, the cells synthesized only about 15~ ofpabS mRNA as compared to control cultures. However, a small increase of pabS mRNA was observed during the initial 15 rain following Pi addition, which suggests that the cell has to build up a substantial concentration of the intracellular effector before exerting negative control at the transcriptional level. (e) Half-life of the pabS transcript The pattern of change of the pabS transcript levels in Cd-producing S. gr/seus cultures suggested that a ~'apid decay of mRNAs specific for antibiotic biosynthetic enzymes occurs both in long-term fermentations and shortterm incubations (Figs. I and 3). To test this hypothesis, 10 pg Rif/ml was added to a 9-h-old culture. This concentration of Rif produced an 87~o inhibition of total RNA synthesis in 10 min under our experimental conditions. The decay of the pabS transcript levels following inhibition of de novo RNA synthesis is shown in Fig. 4. The apparent half-life of the pabS transcript in SPG medium was about 35 min. The apparent half life of labS transcript after 30 min of Pi addition does not seem to change significantly as compared to that determined by inhibition of RNA synthesis with Rif, i.e., levels were reduced by 50% in 30 min (Fig. 3),
100
~o
0
75
e 30
| so
~ 20
E
\
0 z
g
25
i
12
2/,
36
/.,8
60
TIME (h)
Fig. 2. Stimulation by P, ofthe rate oftotal RNA synthesis. Cultures were grown in SPG medium with ( • ) and without (@) supplementation with 7.5 mM Na.phosphate buffer. The incorporation of [=4C]uracil into RNA was measured at different times using pulses of 6 min each, as described by Liras et al. (1977).
15
t
t
30
45
t /~~ 60
150
TIME (rain)
Fig. 3. Time kinetics of the effect of 7.5 mM Na" phosphate buffer on synthesis of the pabS transcript in short-term experiments. Cultures in SPG medium at 9 h of incubation (rapid growth phase) were supplemented with Pi (arrow)(•). Control unsupplemented cultures (@). pab$ mRNA was quantified as indicated in the legend to Fig. I.
82
1
2
100 --
11
|
75 !
.
2 . 3 ~,-
so
iI
1 . 8 P,-
90 ' TIME (rnin)
6'0
120 '
150
Fig. 4. Kinetics of decay ofthe pabStranscript orS. grbeus cells ~own for 9 h in SPG medium after inhibition of de novo IUqA synthesis with 10 pg Rif/ml (Martin et al., 1977). Total RNA ,,,as isolated at different times following rifampicin addition. The app~ent half-life of the pabS transcript [calculated by using the least squares-fitted plots of the logarithmic value of the dots' intensity against time (Merino et al., 1987)] was 35 min.
but the low levels of the pabSgene transcript formed in the presence of Pl did not allow us to calculate precisely the half-life of this transcript under Pt supplementation conditions. (f) Size of the pab$ transcript The size of the pab$ transcript from $.g~sevs IMRU3$70 was determined by Northern analysis using RNA extracted as indicated in Fig. 1 legend. A main hybridization region of about 1.7 kb was observed (Fig. 5). This size correlates well with the size of the pabS gene (Rebollo et al., 1989; L.M. Criado, J.A. Gil and J.F.M., unpublished), and suggests that the pab$gene is transcribed in rive as a monocistronic mRNA. The hybridization intensity of the RNAs extracted from cells grown with and without 7.5 mM Pt supports the conclusion that Pt control is exerted at the transcriptional level
(Fig.5). (g) Prderegulated mutants are impaired in the transcriptional control Mutants partially or completely insensitive to Pi control of Cd biosynthesis [Pt-deregulated (PD) mutants] have been reported before [Martin et al., 1979). Two low Cdproducing mutants (LP5 and LP34)showed 27~ and 45 Yo production of Cd respectively as compared to the parental strain, as well as lower sensitivity to Pt control of Cd biosynthesis. LP34 is a natural revertant of a nonproducing
Fig. 5. Northern analysis of the RNA of $. Av~seuscells gown in SPG medium without (lane I) or with (lane 2) supplementation with 7.5 mM sodium phosphate bufl'er. RNA samples (30 pg) used for Northern analysis were denatured with glyoxal as described in the legend to Fig. I and
fractionated on agarose (1.1%) gel. The gel was blotted onto a nitrocellulosefilterby standard methods(Sambrooket al., 1989)and baked in a vacuum oven at 80'C for 2 h. Prehybridizationswere at 420C overnightin 50% formamide,5 x SSC,50 mM sodiumphosphatebuffer pH 6.5, 250#g/ml denatured salmon sperm DNA and 10 x Denhart's solution. Filters were hybridized in the same bulbr containing 10~ dextran sulfate at 42°C for 24 h. Afterwards,filters were washed four timesin 2 x SSC, 0.1~ SDS at 5o°Cfor 15min(Williamsand Masoon0 1985). The arrows indicate single.stranded DNA size markers in kb (Bf/l-digestedpBIU22denaturedby treatmentwith81yoxalin the same way as the RNA samples).
mutant. Both of them have a normal morphology and the same pigmentation markers as the parental strain. Formarion of PABS and synthesis of Cd by these mutants was partially insensitive to P, control (Fig. 6a,b). Formation of the pab$ transcript in these PD mutants is repressed to a lesser extent (58~o in LP$ and 7~0 in LP34, respectively) than in the parental strain (90~) when SPG medium was supplemented with 7.5 mM Pt (Fig. 6c). These results indicate that Pt-deregulated mutants are impaired in the control exert~ by Pi at the transcriptional level. There was an apparent lack of correlation between the low effect exerted by Pi on pab$ mRNA formation by mutant LP34 (7% repression) and the 4 5 ~ reduction in PABS activity
83
100 •, CmcOgmm
• P~IIAN
_r~me6wtJ~tcpt
?5
i
I
50
!
Wt LI~ LP~
wt
LP5 LP~
wt
LI~ LP3¢
Fig.6. P, inhibitionofthe synthesisof C..d(panel a), PABS(panelb) and the pabStranscript (panelc) in the wildtype (wt) $. grtveusIMRU3570 strain and in LP5 and LP34Prderegulatedmutants (Martinet al., 1979; 1984)~The culturesweregrownin SPG mediumwithor withoutsupplementationwith 7.5 mM Na. phosphatebufferand resultsare givenas the percentageofinhibitionexertedby P,. Samplesfor pab$RNAweretaken at 12h, for PABS at 24 h and for Cd assays at 36 h ff incubation.
observed in the same strain. These results indicate that an additional level of regulation (probably at the translational or posttranslational level) exists in this mutant. (h) Conclusions (1) Results indicate that transcription of the labs gene during Cd biosynthesis peaks at 12 h of growth, preceding the formation of PABS and the antibiotic itself. These results confirm previous observations on the sequential formation of macromolecules during Cd production (Liras et al., 1977). (2) Formation of pab5 transcript in 5. griseus IMRU3570 is repressed by Pi, as determined by RNA dot-blot hybridization or by Northern analysis of the transcripts (Figs. 1, 3 and 5). Despite repression ofthe pabS gene, total RNA synthesis clearly increased in Prsupplemented cultures, suggesting that ribosomal RNA genes and other genes required for the increased growth (Liras etal., 1977) (e.g., primary metabolism) are expressed more efficiently in Prsupplemented cultures. Expression of the pabS gene (and very likely other antibiotic biosynthetic genes) is, therefore, controlled differentially by the Pt concentration in the culture medium. (3) Pt exerted a severe repression of labs transcript formation (85~ after 30 min) but little or no significant effect was observed during the first 15 min following the addition of P~ to the culture. Since P~ uptake is very rapid (Martin et al., 1977) the time required to begin P~ repression may be due to the need to adjust the transcriptional machinery to repress transcription of the labs gene.
(4) Apparent half-life of the PUbStranscript in Cd-producing cultures as measured by dot-blot hybridization is 35 min. Decay of the pubs transcript explains previous observations of the decrease of the rate of antibiotic biosynthesis in batch cultures (Martin and Demain, 1980). The levels of pubs mRNA need to be replenished continuously for prolonged biosynthesis of Cd. Nothing is known about the mechanisms of mRNA degradation following translation in Streptomyces. Decay rate ofthe pubs transcript in Pi-supplemented cultures does not seem significantly different from decay in control unsupplemented cultures, i.e.,-Pi control is exerted mainly through control of transcription initiation rather than by a change in the rate of mRNA degradation. (5) Prderegulated mutants are impaired in the transcriptional control exerted by Pi. This result suggests that these mutants may have a mutation in the Pi control sequence existing upstream from the Pubs gene (ReboHo et al., 1989) or in a putative protein-effector complex that could bind to this sequence (Martin, 1989). (6) The biosynthesis of antibiotics and other secondary metabolites is probably controlled at the transcriptional level by a general mechanism (Martin and Liras, 1989). Carbon catabolite regulation of phenoxazinone synthase, an enzyme involved in actinomycin biosynthesis, is also exerted at the transcriptional level (Jones, 1985). Similarly, nitrogen regulation of giutamine synthetase in Smeptomyces coelicolor (Fisher and Wray, 1989) and the control of actinorhodin formation by the product of the afsB gene (Horinouchi et al., 1989) are exerted at the transcriptional level. The cascade of events at the molecular level that produce the 'switch-on' of the pab$ promoter following Pt depletion in SPG medium (Liras et al.0 1977; Gil et al., 1985), and the switch-off of the Pubs gene when the producing medium is supplemented with Pi, is still unknown. Two mechanisms could be involved in transcriptional regulation. One is based on transcriptional selectivity of Pt-regulated promoters by RNA polymerases with different sigma factors (Westpheling et al., 1985; Buttner et al., 1988; Takahashi et al., 1988; Westpheling and Brawner, 1989; Buttner, 1989). An alternative mechanism is the hwolvement of regulatory proteins that bind to upstream regions ofcertain actinomycetes promoters (Geistlich et al., 1989; Lin and Wilson, 1988). Positive regulatory proteins are involved in the control of the Pi regulon in E. coli(Kimura et al., 1989) suggesting that this may also be the case in Streptomyces.
ACKNOWLEDGEMENTS J.A. Asturias was supported by a fellowship of the Ministry of Education and Science, Madrid, Spain. We
84 thank J.A. Gil for providing plasmid plJ819, C. Esmahan and S. Baumberg for reading the manuscript, and M.P. Pumas, R. Barrientos, B. Martin, S. Llamas and M.I. Corrales for valuable technical assistance.
REFERENCES Buttner, MJ,: RNA polymerase heterogeneity in Streptomycescoellcolor A3(2), MUl. Microbiol. 3 (1989) 1653-1659. Buttner,MJ., Smith, A.M. and Bibb, MJ.: At leastthree differentRNA polymerasc huloenzymas direct transcription of the agarase gene (dagA) of Streptomyces coellcolorA3(2). Cell 52 (1988) 599-609. Fisher, S.H. and Wray Jr.,L.U.: Regulation of glutamine synthetase in Streptomycer coellcolor.J. BacterioL 171 0989) 2378-2383. Oeistlich, M., Irniger, S, and HOtter, R.: Localization and functional analysisof the regulated promoter from the Streptomycesglaucescens reeloperon. Mol. Microbiol. 3 (1989) 1001-1069. Oil, J.A. and Hopwood, D.A.: Cloning and expression of a p-aminobenzoic acid synthetase gone of the candicidin-producing Streptomycesgrlsesa.Gene 25 (1983) 119-132, Gil, J.A.,Nahm'ro, G., Villanueva,J.R. and Martin, J.F.:Characterization and regulation of p-aminobenzoic acid synthase from Sweptomyces 8r/sem. J. Gen. Microbiol. 131 (1985) 1279-1287. Hopwood, D.A., Bibb, MJ., Chater, K.F., Kieser, T., Bruton, CJ., Kieser, H.M., Lydiate, DJ., Smith, C.P., Ward, J.M. and Schrempf, H.: Genetic Manipulation of Streptomyces: A Laboratory Manual. John innes Foundation, Norwich, 1985. Horinouchi, S., Maipartida, F., Hopwood, D.A. and Beppu, T.: q~B stimulates transcription of the actinorhodin biosynthetic pathway in $~ptomyceJ coeltcolorA3(2) and $1r,ptomyccs ltvMans, Mol. Gen. Goner. 215 (1989) 355-357. Jones, G.H.: Regulation of phenoxuinone synthase expression in $~ptomyc¢a anttbtottcla,J. Bacteriol, 163 (1985) 1215-1221. Kimura, S., MakinG, K.. Shinqawa, H,, Amemura, M. and Nakata, A.: Regulation ofthe phosphate regulon of£schev~chtacolt: Characterization of the promoter of the pats sane. MoLOen. Goner, 215 (1989) 374-380. Lanipridse, J., LanFidge, P. and Bergquist, P.L.: Extraction of nucleic acids from agarose gels, Anal, Biochem. 103 (1980) 264-271, Lin, E. and Wilson, D.8.: Identification era colE,binding protein and its potential role in itlduction ofthe colE'gonein ~evmomonosporafmca. J. Bacteriol. 170 (1988) 3843-3846, Liras, P., Villanueva, J,R. and Martin, J,F,: Sequential expression of macromolecule biosynthesis and candicidin formation in $tmptomyce~ grlaeua, J. Oen, Microbiol, 102 (1977) 269-277. Martin, J.F,: Molecular mechanisms of the control by phosphate of the biosynthesis of antibiotics and other secondary metabolites, In Shapiro, S. (Ed,), Regulation of Secondary Metabolism in Actinomycetes. CRC Press, Boca Rat&n, FL, 1989, pp. 213-237,
Martin, J,F. and Domain, A.L,: Control by phosphate of candi~,~din production. Biochem, Biophys. Res. Commun. 71 (1~7o) 1103-1109. Martin, J.F. and Demain, A.L: Control of antibiotic biosynthesis. Microbiol. Rev. 44 (1980) 230-251. Martin, J.F. and Liras, P.: Organization and expression of genes involved in the biosynthesis of antibiotics and other secondary metabofites. Annu. Rev. Microbiul. 43 (1989) 173-206. Martin, l.F. and McDaniel, L.E.: Kinetics of biosynthesis of polyene macrolide antibiotics in batch cultures: carl maturation time. Biotechnul. Biocng. 17 (1975) 925-938. Martin, J.F., liras. P. and Dcmain, A.L.: Inhibition by phosphate ofthe activity of candicidin synthases. FEMS MicrobioL Left, 2 (1977) 173-176. Martin, J.F., Naharro, G., Liras, P. and Villanueva, J.R,: Isolation of mutants deregulated in phosphate control ofcandicidin biosynthesis. J. Antibiot. 32 (1979) 600-606. Martin, J.F., Alegre, M.T., Castro, J.M. and Liras P.: Extrachromosomal genetic elements that control specific enzymes involved i~ antibiotic biosynthesis: possible involvement of an intraceilular pleiotropic effector. In Ortiz, L., BojalH,L.F. and Yakuleff, V. (Eds.), Biological, Biochemical and Biomedical Aspects of Actinomycetc~. Academic Press, Orlando, FL, 1984. Martin, J.F., Daza, A., Asturias, J.A., Gil, 5.A. and Liras, P.: Transcrip. tional control of antibiotic biosynthesis at phosphate-regulated promoters and cloning of a gone involved in the control of the expression of multiple pathways in Stmptomyces. In Okami, Y., Beppu, T. and Ogawara, H. (Eds.), Biologyof Actinomycetes.Japan Science Society Press, Japan, 1988, pp, 424-430. Merino, E., Becerril, B., Valle, F. and Bolivar, F.: Deletion era repetitive extragenic palindromic (REP) sequence downstream from the struc, tural gone of£sc~¢ri~ia coilglutamate dehydrogenase affects stability of its mRNA. Gone 58 (1987) 305-309. Rebollo, A., Oil, J.A., Liras, P., Asturias, J.A. and Martin, J.F.: Cloning and characterization of'a phosphate-regulated promoter involved in phosphate control of candicidin biosynthesis. Gone 79 (1989)47-58. Sambrook, J., Fritsch, E.F. and Maniatis, T.: Molecular Cloning. A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spri~ Harbor, NY, 1989, Takahashi, H., Tamaka, K, and Sh0na, T,: Genetic constituent ofrpoD 8one homologues in $1reptomycesstrains, in Okami, Y., Beppu, 1".arid Ogawara, H. (Eds.), Biologyof Actinomycetes.Japan Science Society Press, Japan, 1988, pp. 58-63. Thomas, P.S.: Hybridization of denatured RNA transferred or dotted to nitrocellulose paper. Methods Enzymul. 100 (1983) 255-266. Westpheling,J. and Brawner, M.: Two transcribinl| activities are involved in expression of the $~reptomyces8aiactose operon. J. Bacteriol. 171 (1989) 1355-1361. Westphelina, J., Ranes, M. and Losick, R.: RNA polymerase betero8eneity in Streptomyces coeUcolor.Nature 313 (1985) 22-27. Williams, J.F. and Masoon, PJ.: Hybridization in the analysis of RNA. In Haines, B.D. and Hiuins, SJ. (Eds.), Nucleic Acid Hybridization: A Practical Approach. IRL Press, Oxford, 1985, pp. 139-180.
