Molecular Biology, Vol. 38, No. 3, 2004, pp. 329–336. Translated from Molekulyarnaya Biologiya, Vol. 38, No. 3, 2004, pp. 394–403. Original Russian Text Copyright © 2004 by Elizarov, Gavrilina, Danilenko.
GENOMICS. TRANSCRIPTOMICS. PROTEOMICS UDC 577.152.3:579.873.71
Modulation of Serine/Threonine Protein Kinase Activity in Chloramphenicol-Resistant Mutants of Streptomyces avermitilis S. M. Elizarov1, A. V. Gavrilina2, and V. N. Danilenko3 1 Bach
Institute of Biochemistry, Russian Academy of Sciences, Moscow, 119071 Russia E-mail:
[email protected] 2 Antibiotic Biotechnology Research Center, Moscow, 109004 Russia 3 State Institute of Protein Biosynthesis, Moscow, 109004 Russia Received July 14, 2003
Abstract—A mutation to chloramphenicol resistance (Cmlr) stimulates production of macrolide avermectin in Streptomyces avermitilis; production starts in the early stationary phase. By labeling in vivo, the Cmlr mutation was shown to stimulate phosphorylation of Ser and Thr in several proteins in the same growth phase. Autophosphorylation of active protein kinases (PK) was analyzed in gel after one- or two-dimensional PAGE for the original S. avermitilis strain ATCC 31272, its Cmlr mutant, and a Cmls revertant. An increase in in vivo phosphorylation was associated with an increase in autophosphorylation of Ser/Thr-PK 41K, 45K, 52K, 62K, and 85K and complete suppression of autophosphorylation of PK 66K. Comparison of the PK molecular weights and pI with the parameters deduced for putative PK encoded by S. avermitilis genes identified the 41K, 45K, 52K, 62K, and 85K proteins as pkn 24, pkn 32, pkn 13, pkn12, and pkn5, respectively. Prenylamine lactate, a modulator of calmodulin-dependent processes, substantially reduced the avermectin production, impaired the Cml resistance, and selectively inhibited Ca2+-dependent PK 85K in the Cmlr mutant. It was assumed that PK 85K is involved in regulating the avermectin production. Key words: Streptomyces avermitilis, serine/threonine protein kinases, chloramphenicol resistance, Ca2+/calmodulin, avermectin
INTRODUCTION Actinomycetes of the genus Streptomyces are well known for their ability to produce numerous secondary metabolites such as antibiotics. Regulation at several levels and various signal transduction mechanisms are undoubtedly involved in their intricate life cycle and multicellular differentiation, which resembles the growth of filamentous fungi [1]. Streptomyces species and other actinomycetes are noted for multiple resistance to antibiotics that they do not produce. Genetic changes modulating the resistance to an antibiotic are often accompanied by changes in resistance to other ones, along with phenotypic modification of sporulation, morphological differentiation, and antibiotic production; i.e., their effect is pleiotropic. We have previously observed that mutations to chloramphenicol (Cml) or aminoglycoside resistance increase the production of secondary metabolites [2]. It has been proposed to raise the yield of polyketide antibiotics by inducing Cml and kanamycin resistance (Cmlr, Kanr) mutations in industrial actinomycete strains [3–5]. Mutants with higher production of avermectin, tylosin, or erythromycin occur in Cmlr and/or Kanr isolates of S. avermitilis [3], S. fradiae [4], and
Saccharopolyspora erythraea [5], arising spontaneously at a frequency of 10–7 to 10–6. We have observed that addition of Ca2+ to the culture medium increases the rate of mutations to Cmlr and Kanr by a factor of 50–100 [3–5]. The mutants selected express a Ca2+dependent phenotype. Hence antibiotic resistance mutations have been assumed to switch some regulatory mechanisms in the bacterial cell by modulating the activities of specific protein kinases (PK) [2]. Owing to recent advance in genomics, numerous genes for serine/threonine protein kinases (Ser/Thr-PK) have been found in the actinomycete genomes. Thus complete genomic sequences has revealed 13 Ser/Thr-PK genes in Mycobacterium tuberculosis [6], 5 in Corynebacterium diphtheriae (http://www.sanger.ac.uk/Projects/ C_diphtheriae), 44 or 45 in S. coelicolor [7], and no less than 32 in S. avermitilis (http://avermitilis.ls.kitasatou.ac.jp). Some of the relevant enzymes are involved in regulating bacterial metabolism, growth, and development [8–10]. Yet the functions of the vast majority of actinomycete Ser/Thr-PK are still unclear. Metabolic pathways of avermectin biosynthesis and the structural organization of the avermectin bio-
0026-8933/04/3803-0329 © 2004 MAIK “Nauka /Interperiodica”
330
ELIZAROV et al.
synthesis gene cluster are well known [11, 12], whereas data on the regulation of avermectin biosynthesis in S. avermitilis are still scarce. In selection experiments, we have observed that avermectin production is substantially higher in particular Cmlr mutants of S. avermitilis. We have studied the biochemical mechanisms determining this association. The activity and protein targets of Ser/Thr-PK have been assayed in vitro in cell extracts of S. avermitilis ATCC 31272 and its pleiotropic Cmlr mutant with a higher avermectin production [13]. Basing on the results, we have assumed that the Cmlr mutation stimulates the function of the PK-dependent signaling in early stationary growth of a culture. Since elevated production of antibiotics is of economic and industrial interest, we studied the contribution of specific Ser/Thr-PK to the improvement of avermectin-producing strains. The objective of this work was to study the quantitative and qualitative changes in the Ser/Thr-PK pool of growing S. avermitilis cultures differing in resistance to Cml and in production of avermectin. EXPERIMENTAL Streptomycete strains. Cml resistance, avermectin production, and Ser/Thr-PK activity were established for parental S. avermitilis ATCC 31272, its Cmlr mutant, and a Cmls revertant originating from the mutant. The parental strain was obtained from the Russian Collection of Industrial Microorganisms (Moscow). The strain is Cmls (minimal inhibiting concentration 2 µg/ml Cml) and produces avermectin to about 500 µg/ml. Mutants with the Cmlr phenotype and avermectin production increased 3- to 3.5-fold arise at a frequency of (1–3) · 10–7 in S. avermitilis. Our Cmlr mutant is resistant to 30 µg/ml Cml and produces up to 1750 µg/ml avermectin. Stable spontaneous Cmls revertants arise in the Cmlr mutant at a frequency of 2 · 10–3 and are similar to the parental Cmls strain in Cml resistance and avermectin production. Our Cmls revertant was isolated from the Cmlr mutant by sensitivity to 2 µg/ml Cml and produced avermectin to about 600 µg/ml. Methods of culturing and minimal agar, complete agar, and complete liquid media were as described in [3, 13, 14]. Cell culture and extracts. For in situ analysis of protein phosphorylation, cells were cultured in a basal liquid medium [13]. Cultures were initiated by adding 5% of an inoculate (late stationary culture) and incubated at 28°ë with agitation under aerobic conditions. Aliquots were taken, and cell extracts obtained. For in vivo labeling, cells were grown in the basal liquid medium at 28°C for 40 or 48 h, collected by centrifugation at 3000 g for 15 min, resuspended in a phosphor-free buffer (100 mM Tris-HCl, pH 7.3, 100 mM NaCl, 40 mM KCl, 0.1 mM CaCl2, 0.5 mM MgCl2, 0.2 mM Na2SO4, 0.5% casamino acids, 1% glucose),
supplemented with 20 MBq/ml [32P]i (Obninsk, Russia), and incubated for 8 h. Mycelium was collected by centrifugation at 3000 g and washed twice with a standard buffer (10 mM triethanolamine, 10 mM KCl, 125 mM NaCl, 5 mM MgCl2, 6 mM 2-mercaptoethanol, 1 mM EDTA, 10% glycerol, pH 7.8). Cells were resuspended in buffer A (the standard buffer supplemented with 1 mM PMSF, 1 µg/ml leupeptin, 1 µg/ml pepstatin, 150 mM β-glycerophosphate, 100 mM NaF), and disrupted by sonication. The cell lysate was incubated with pancreatic DNase and RNase (25 µg/ml each) at 4°ë for 15 min, and cell debris was removed by centrifugation at 20,000 g for 30 min. The supernatant was combined with an equal volume of 2× sample application buffer [15] and dialyzed against 1× sample application buffer. To study the in vivo and in situ labeling, proteins were resolved by SDS-PAGE. In the case of two-dimensional SDS-PAGE, samples were prepared in the same buffer supplemented with 9 M urea. PAGE. One-dimensional SDS-PAGE in 10 or 12.5% gel was performed according to the standard protocol [15]. Gels were stained with Coomassie R, 50% trichloroacetic acid; washed with hot 7% acetic acid; dried; and exposed with a Kodak X-Omat AR X-ray film. Two-dimensional SDS-PAGE was performed as in [16]. For isoelectrofocusing we used glass tubes (diameter 2.5 mm, length 15.0 cm). A solution for gel preparation contained 2.5 ml of acrylamide–bisacrylamide (4.0% and 0.25%, respectively), 75 µl of ampholyte 3–7 (Pharmalyte, Pharmacia Biotech), and 8 µl of ampholyte 6–8. Isoelectrofocusing was carried out for 20 h at 500 V. In the second dimension, proteins were resolved in vertical 15% gel (30% acrylamide, 1% bisacrylamide). Radioactivity of gel fragments was measured by scintillation counting. The labeling of each band or spot was estimated for three independent cultures of the mutant or parental strains. PK autophosphorylation. PK activity was assayed in gel after protein renaturation according to Kameshita and Fujisawa [17]. Extracted proteins were resolved by SDS-PAGE, denatured in gel with 6 M guanidine hydrochloride, and renatured. Autophosphorylation was carried out in a phosphorylation buffer (25 mM HEPES, pH 7.4, 5 mM MgCl2) containing 2 MBq/ml [γ-32P]ATP (2.8 · 108 MBq/mM, Obninsk) at 8°C for 4–5 h. An analysis buffer contained 5 mM MgCl2 and either 1 mM CaCl2 or 1 mM EGTA. To remove an excess of labeled ATP, gels were thoroughly washed with the phosphorylation buffer containing 1% sodium pyrophosphate and 10 mM nonlabeled ATP. Gels were fixed and stained as above, dried under vacuum, and autoradiographed. Phosphoamino acid analysis. To identify the phosphoamino acids resulting from protein phosphorylation in situ or in vivo, gel fragments containing labeled proteins were hydrolyzed with 6 M HCl at MOLECULAR BIOLOGY
Vol. 38
No. 3
2004
Ser/Thr PROTEIN KINASE ACTIVITY AND Cml RESISTANCE
Avermectin assays. To quantitate avermectin, 0.5 ml of a culture were supplemented with 50% acetone, intensely shaken at 28°ë for 20 min, and centrifuged at 5000 g for 5 min. Aliquots of the supernatant were analyzed by thin-layer chromatography on glassplated silica gel (Merck) in hexane–acetone–methanol (13:7:0.5). The avermectin content was estimated with a high-speed TLC CS-920 scanner (Shimadzu, Japan) at 243 nm. In each experiment, we analyzed specimens of at least three cultures in triplicates. The effect of inhibitors on cell growth. Paper disks were used to study the effect of Cml and/or Ca2+/calmodulin modulators on the growth of Cmlr S. avermitilis cells. Spore suspension in agar (108 spores/ml) was plated on 20 ml of agar medium (0.05% K2HPO4, 0.05% MgSO4, 0.05% NaCl, 0.001% FeSO4, 0.1% KNO3, 2% glucose, 2% agar, pH 7.2) in a Petri dish. Paper disks (diameter 0.5 cm) were soaked with an aqueous solution of an inhibitor, Cml, or both and placed on the agar surface. Dishes were incubated for 2−3 days. The radius of the growth suppression area was measured in two independent cultures.
10 9 8 7 6 5 4 3 2 1 0
Avermectin, mg/l 35 30 25
Cell biomass, g/l
100°C under vacuum. Inorganic 32P and PAG hydrolysis products were removed by chromatography on Dauex 50W-X8(H+) [18, 19]. Specimens were analyzed by electrophoresis on TLC cellulose plates (Merck, Germany) in acetic acid–pyridine–deionized water (10:1:150) at 1100 V for 60 min. Electrophoregrams of 32P-labeled hydrolysates were exposed with X-ray film. Labeled phosphoamino acids were identified against reference phosphoamino acids (Sigma) detectable with ninhydrin.
331
20 15 10 5 8 12 16 20 24 28 32 36 40 44 48 52 56 60 Time, h
0
Fig. 1. Cell growth (circles) and avermectin production (squares) in the original S. avermitilis strain (open symbols) and its Cmlr mutant (filled symbols).
(a)
(b)
116 97.4 72
66 45
41
Protein was quantitated according to Bradford [10], using BSA as a standard.
37.5 29
RESULTS Protein Phosphorylation In Vivo The original Cmls S. avermitilis strain and its Cmlr derivative grew exponentially for the first 40–48 h in a liquid medium; when the stationary phase was achieved, the mutant started to produce macrolide avermectin (Fig. 1). Analysis showed that many proteins were phosphorylated in vivo before (40–48 h of growth) and after (48–56 h) the start of avermectin production (Fig. 2). In either strain, we observed three major (72K, 41K, and 37.5K) and up to 12 minor (24K–85K) labeled proteins. The set of labeled proteins was the same in both strains. Protein labeling was higher after 48–56 h of growth (compare lanes 2 and 1). On the other hand, protein labeling in Cmlr cells was higher than in cells of the original strain in both periods (the difference in labeling of the major proteins was 1.5–2.5 times). Total phosphate incorporation only slightly differed between the cultures. After acidic hydrolysis, 32P-labeled amino acids were identified as phosphoserine in the 37.5K and 41K proteins and as phosphothreonine in the 72K proMOLECULAR BIOLOGY
Vol. 38
No. 3
2004
14
1
2
1
2
Fig. 2. Autoradiography of proteins phosphorylated in vivo in cells of (a) the parental and (b) Cmlr mutant S. avermitilis strains grown in the presence of ortho-[32P]phosphate for (1) 40–48 or (2) 48–56 h. Samples were applied at 20 µg total protein per lane. Molecular weights (thousand) are indicated for reference proteins (on the left) and the major labeled polypeptides (on the right).
tein. The difference in level of protein phosphorylation between the original and Cmlr strains could result from a difference in activities of individual PK. Further experiments were designed to check this assumption. Autophosphorylation of PK in Gel Active PK in cell-free extracts were analyzed in the original and mutant S. avermitilis strains as dependent
332
ELIZAROV et al. kDa 116
(a)
(b)
(c)
97.4 66 45
29 1
2
3
4
5
6
1
2
3
4
5
6
4
5
(d)
85
85 kDa
1
2
3
4
5
Fig. 3. Changes in the set of Ser/Thr-PK during growth of (a) the original S. avermitilis strain, (b, d) Cmlr mutant, and (c) Cmls revertant in the standard medium or (d) in the presence of pernylamine lactate. (a–c) Cells were grown in the liquid medium for (1) 36, (2) 40, (3) 44, (4) 48, (5) 52, or (6) 56 h (see Fig. 1). (d) Cultures of the Cmlr mutant were grown for 48 h (1–4) without or (5) with prenylamine lactate, and PK were analyzed in the presence of (1, 4, 5) both Ca2+ and Mg2+, (2) Mg2+, or (3) EGTA. Cellfree extracts were obtained, PK tested for autophosphorylation after renaturation in gel, and gels autoradiographed. Extracts were applied at 25 µg total protein per lane.
on the culturing time. The original Cmls strain, the Cmlr mutant, and the Cmls revertant were grown for 36–56 h. Cell-free extracts were obtained and assayed for active PK. Extracted proteins were resolved by one-dimensional SDS-PAGE, denatured with guanidine hydrochloride, renatured, and tested for autophosphorylation in the presence of labeled ATP in gel. Time-related changes in autophosphorylation were compared for the three strains (Figs. 3a–3c). Extracts of cells of the original strain grown for 52–56 h (early stationary phase) contained seven radiolabeled components: 85K, 68K, 62K, 52K, 45K, 41K, and 37.5K (Fig. 3a, lanes 5, 6). Almost all of these were detectable in extracts of both mutant strains grown for 48– 56 h (Figs. 3b, 3c). The time-related changes in autophosphorylation of individual polypeptides varied among the strains. In the original strain, autophosphorylation of all but one component reached its maximum after 52 h of growth (Fig. 3a, lane 5) and slightly decreased in the next 4 h (lane 6). The exception was the 41K protein (lane 3): autophosphorylation was maximal after 44 h of growth and did not substantially change to 52 h.
In the Cmlr mutant (Fig. 3b), autophosphorylation was low in the first 36 h of cell growth and then gradually increased till 52 h. Autophosphorylation of the 85K, 68K, 52K, 45K, and 37.5K polypeptides was maximal after 52 h of growth (Fig. 3b, lane 2) and did not significantly change in the next 16 h (lanes 3–6). After 56 h of cell culturing, autophosphorylation of the 85K, 68K, and 52K minor polypeptides decreased about fourfold. In the Cmls revertant, maximal autophosphorylation of all proteins was observed after 52 h of cell growth (Fig. 3c). In general, the PK autophosphorylation profiles were similar in the original strain and in the Cmlr mutant. Yet the phosphate incorporation in the 45K, 41K, and 37.5K components in the Cmlr mutant was respectively 3.2–3.7, 2.4–2.8, and 1.9–2.1 times higher than in the original strain by two independent estimates. Autophosphorylation of these proteins correlated with the Cmlr phenotype. The Cmls revertant was similar in autophosphorylation of the 37.5K and 45K proteins to the original strain (the difference was 0.7–0.9 and 1.1–1.2 times, respectively). The autoMOLECULAR BIOLOGY
Vol. 38
No. 3
2004
Ser/Thr PROTEIN KINASE ACTIVITY AND Cml RESISTANCE
phosphorylation of the 41K protein in the revertant was 1.9–2.2 times higher than in the original strain. Thus, activation of two PK, 37.5K and 45K, directly correlated with the Cmlr phenotype. Slight suppression of the 41K PK in the revertant precluded the role of this PK in the Cmlr phenotype. In addition, the labeling of the 68K protein was somewhat more intense in the original strain after 52–56 h of culturing (Fig. 3a, lanes 5, 6). No interstrain difference was observed in the set and number of labeled proteins. Phosphoamino acid analysis of individual labeled polypeptides revealed radioactive phosphate only in phosphoserine residues in the case of the 41K and 37.5K proteins and only in phosphothreonine residues in the case of the 45K and 68K proteins. Thus, these proteins possess the Ser/Thr-PK activity, which determines their autophosphorylation. Autophosphorylation of the minor 85K polypeptide in 48-h Cmlr cultures was stimulated by Ca2+ (Fig. 3d, lanes 1–3). In the absence of Ca2+, autophosphorylation of the 85K protein was about 4.75 times lower (Fig. 3d, lanes 2, 3). Autophosphorylation of other proteins did not substantially change. Added to the incubation mixture in PK analysis, Ca2+-binding chelating agent EGTA had no effect on the autophosphorylation intensity or a set of labeled proteins (Fig. 3d, lane 3). Hence the residual phosphorylation of the 85K protein in the presence of Mg2+ was Ca2+independent (Fig. 3d, lanes 2, 3). The resolution of one-dimensional PAGE is insufficient for reliable discrimination between autophosphorylation of PK and phosphorylation of comigrating substrates. Hence two-dimensional PAGE was used to quantitate the labeling for individual proteins. Proteins were extracted from cells of the original and Cmlr S. avermitilis strains after 48 h of growth, resolved by SDS-PAGE, renatured, and tested for autophosphorylation in the presence of [γ-32P]ATP. A broad set of proteins was labeled in either strain (Fig. 4). We found eight major (37.5K, 41K, 45K, 47K, 52K, 62K, 68K, 85K) and up to 12 minor common proteins. The labeling of some proteins increased in the case of the Cmlr mutant as compared with the original strain: the factors were 2.5–3.2 for the 41K protein, 6.6–7.1 for the 45K protein, 4.5–4.9 for the 52K protein, 4.7–5.5 for the 62K protein, and 7.1–8.3 for the 85K protein by two independent estimates. On the other hand, 32Pi incorporation was not observed for 66K PK of the Cmlr mutant, whereas the enzyme was distinctly labeled in the case of the original strain (Fig. 4a). In total, these findings agreed with the above results of PK analysis with one-dimensional PAGE (Figs. 3a, 3b). Possibly, 66K PK was undetectable in the case of one-dimensional PAGE because of its low content. Phosphoamino acid analysis showed that only Thr was modified in the 45K, 52K, 66K, and 68K proteins, MOLECULAR BIOLOGY
Vol. 38
No. 3
2004
333
and only Ser in the 37.5K, 41K, and 85K proteins. No label was detected in Ser or Thr of the 47K protein. The Effect of a Ca2+/Calmodulin Inhibitor on Physiological Parameters and Activity of PK Since our Cmlr mutant had a Ca2+-dependent phenotype, we analyzed the effect of prenylamine lactate, a modulator of Ca2+ signaling, on its physiological characteristics. Prenylamine lactate acts as a potent selective inhibitor of calmodulin in the bacterial cell [21]. We studied its effects on PK activity, Cml resistance, and avermectin production in the Cmlr mutant. First, autophosphorylation of individual PK was assayed in extracts of cells grown for 36 h under the standard conditions and then for 12 h in the presence of prenylamine lactate. Proteins of 48-h cultures were resolved by SDS-PAGE, renatured in gel, and tested for autophosphorylation in the presence of labeled ATP, Mg2+, and Ca2+. The results are shown in Fig. 3d. Autophosphorylation of the 85K protein decreased about eightfold when cells were grown in the presence of 5 µM prenylamine lactate (lanes 4, 5). The labeling of other PK changed insignificantly. A decrease in autophosphorylation of the 85K protein might reflect its selective inactivation, suppression of its de novo synthesis, or degradation in the presence of the inhibitor. In addition, production or activation of the 85K PK might involve posttranslational modification similar to proteolytic activation of Ca2+/calmodulindependent enzymes [22] or phosphorylation and subsequent activation of PK [23]. Second, the effect of prenylamine lactate on the Cml resistance of the mutant was studied in paper disk assays. Prenylamine lactate used alone at up to 5 nmol/disk did not inhibit mycelium growth, and Cml used alone at 0.25 µmol/disk suppressed the cell growth in a 4.5-mm area. With the two agents used together at the same doses, the growth suppression area was 9–10 mm in radius. Thus, the Cmlr phenotype of S. avermitilis proved to depend on calmodulin. The 85K PK may serve as a target for calmodulin. Third, avermectin production was estimated for the Cmlr mutant grown in the presence of 5 µM prenylamine lactate in the time interval 36–48 h. Direct measurements with 7-day cultures showed that prenylamine lactate reduced the concentration of avermectin in the culture medium by 23–27%, having no effect on cell growth or biomass accumulation. DISCUSSION As our results demonstrate, the activities of individual PK depend on cell growth phase in S. avermitilis. We have previously observed an increase in in vitro phosphorylation of Ser/Thr in extracts for early stationary S. avermitilis cultures and for culture transi-
334
ELIZAROV et al. kDa 90
(a)
(b)
>
>
80
>
>
70
>
> >
>
60
> 50
> >
>
>
>
>
>
40
30
3.2 4.0
4.6
5.2 5.9 6.8
3.2
4.0 4.6
5.2 5.9 6.8 pl
Fig. 4. PK autophosphorylation in gel after two-dimensional PAGE of proteins extracted from 48-h cultures of (a) the parental S. avermitilis strain and (b) its Cmlr mutant. Extracts were applied at 200 µg total protein per gel. Calibration was done with comigrating standards (Bio-Rad, United States) with known molecular weights (on the left) and pI (at the bottom). Arrows indicate the positions of the 41K, 45K, 52K, 62K, 66K, 68K, and 85K proteins.
tion from logarithmic to stationary growth [13]. More recently, similar observations have been made with S. toyocaensis [24]. Thus, we were the first to demonstrate experimentally that a growth phase-dependent change in protein phosphorylation is due to activation of specific PK in Streptomyces. Both in vivo and in situ phosphorylation of Ser and Thr in proteins increased in the time interval 48–56 h. This agrees with an increase in in vitro phosphorylation of proteins in cell extracts of early stationary cultures [13]. Identification of the relevant PK and their specific substrates and elucidation of the biological role of their autophosphorylation need further experiments. Yet we found that protein phosphorylation and PK activity increase differentially when cells start to produce avermectin, suggesting an association between Ser/Thr phosphorylation of proteins and antibiotic production. Activation of the 85K PK by Ca2+ and its suppression by the inhibitor of calmodulin-dependent processes indicate that autophosphorylation of this enzyme is of physiological significance and the 85K PK contributes to the Ca2+ signaling in growing cultures of S. avermitilis. Since the 85K PK is sensitive to the calmodulin inhibitor, proteins homologous to calmodulin are probably involved in regulating the response to Ca2+ in S. avermitilis cells. Calmodulin-
like proteins have been found in many bacteria, but their targets are still unidentified [25]. Activity changes on transition of Cmlr mutant cultures from exponential to stationary growth were observed for six specific PK. Of these, five (41K, 45K, 52K, 62K, and potentially Ca2+/calmodulin-dependent 85K) are activated and one is completely suppressed on transition. Since the activities of four PK (45K, 52K, 62K, and 85K) correlated with the Cmlr phenotype and avermectin production, it is possible to identify the set of PK involved in regulating the pleiotropic effect of Cmlr mutations and to study the causeand-effect relationships between PK activation and avermectin production. Regardless of the place of these PK in the hierarchy of regulatory reactions (i.e., whether the PK activate or coordinate the enzyme systems of avermectin production), studies of their roles will open new possibilities of modulating the activity of biosynthetic processes in the cell. Recently, the S. avermitilis chromosome has been completely sequenced [26]. Although data on the genome composition still need verification, it is possible now to collate the above PK with putative Ser/Thr-PK of this bacterium. The S. avermitilis chromosome has been found to harbor 32 genes for Ser/Thr-PK varying in MOLECULAR BIOLOGY
Vol. 38
No. 3
2004
Ser/Thr PROTEIN KINASE ACTIVITY AND Cml RESISTANCE
deduced molecular weight from 23 to 93 kDa and in pI from 4.5 to 11.9. With these parameters estimated for individual enzymes (table), the 41K, 45K, 52K, 62K, 66K, and 68K proteins may be identified with a high degree of confidence as pkn 24, pkn 32, pkn 13, pkn 12, pkn 5, and pkn 28 Ser/Thr-PK, respectively. Minor deviations of the observed molecular weights and pI from the deduced parameters may be explained by posttranslational modification. Our data on protein phosphorylation in vivo provide direct evidence for this assumption. No genes were found to code for PK similar in pI to the 37.5K and 85K proteins. Possible explanation is the incompleteness of data on the genome composition or profound posttranslational modification of these PK in the cell. Thus, genes for at least six putative Ser/Thr-PK are expressed in S. avermitilis. Expression of four of these genes is activated when the cell starts production of secondary metabolites. Identification of the 37.5K and 85K proteins requires their isolation and sequencing followed by a genome search for the corresponding nucleotide sequences. Although beyond the scope of this article, the molecular mechanisms of the pleiotropic mutation resulting in the Cmlr phenotype and increasing the avermectin production are of particular interest. Such mutations are well known in actinomycetes [2, 3] and are often reversible. There are grounds to think that reversions are due to moderate amplification/deamplification of RES1, a regulatory genetic element [3]. Possibly, a similar regulatory element is responsible for Ser/Thr-PK activation in the S. avermitilis Cmlr strain. Our results indicate that the pleiotropic effect of the Cmlr mutation is mediated by a network of PK-dependent reactions. At least acquisition of the Cmlr phenotype is associated with activation of several PK in S. avermitilis cells. The results of this and previous [13] works demonstrate for the first time that the function of the regulatory PK system is switched when avermectin biosynthesis starts. Our findings provide a basis for further studying the role of specific PK in induction and coordination of avermectin biosynthesis reactions, elucidating of the biochemical effects of antibiotic resistance mutations, and improving avermectin production. The protein product of the Cml resistance gene of S. avermitilis belongs to the family of Cml 3'-O-phosphotransferases and enzymatically inactivates Cml by its ATP-dependent phosphorylation. We did not study the role of Ser/Thr-PK in regulating the activity of the enzymic system protecting cells from Cml. Yet an increase in Cml sensitivity in the presence of the Ca2+/calmodulin inhibitor implicates the Ser/Thr-PK signaling system in Cml resistance. Recently, we have directly demonstrated that specific Ser/Thr-PK modulate the activity of the antibiotic resistance system in S. rimosus: phosphorylation by endogenous Ca2+-dependent PK substantially increases the kanamycin kinase activity of aminoglycoside 3'-O-phosphotransferase [27]. MOLECULAR BIOLOGY
Vol. 38
No. 3
2004
335
Comparison of the Ser/Thr-PK under study and putative Ser/Thr-PK of S. avermitilis Actual Ser/Thr-PK (this work)
Putative Ser/Thr-PK (deduced parameters)*
MW, kDa
pI
PK
MW, kDa
pI
66300 ± 3300
4.7–4.8
pkn 5
66754
4.5
61700 ± 3200
4.6–4.7
pkn 12
57961
4.8
51800 ± 2400
4.5–4.6
pkn 13
52107
4.5
41050 ± 2200
5.9–6.1
pkn 24
43413
6.1
68100 ± 3100
5.0–5.1
pkn 28
68588
5.2
44800 ± 1800
5.3–5.4
pkn 32
44996
5.6
* Data from http://avermitilis.ls.kitasato-u.ac.jp.
Streptomyces strain 85E has been proposed as a test system for screening potential PK inhibitors [28]. Inhibitors are selected by suppression of aerial hyphae formation in this system. We think that the Cmlr gene of S. avermitilis is more promising as regards a universal sensitive test system for identifying PK inhibitors. Recent complete sequencing of the S. avermitilis chromosome [26] will facilitate identification of the PK under study, isolation of the PK regulating avermectin biosynthesis, and gene engineering manipulations with
3 2
4 1
Fig. 5. Suppression of S. avermitilis cell growth with (1) prenylamine lactate, (2) Cml, or (3, 4) both agents. Disks contained (1, 3, 4) 5 nmol of prenylamine lactate and/or (2–4) 0.25 µmol of Cml.
336
ELIZAROV et al.
the PK involved in signal transduction. This will make it possible to regulate the avermectin production and to obtain test systems for new antibacterial agents. ACKNOWLEDGMENTS This work was supported by the program New Research in Biotechnology and Biological Safety of the RF Ministry of Industry, Science, and Technology (grant no. 43.106.11.0021). REFERENCES 1. Chater K.F. 1998. Taking a genetic scalpel to the Streptomyces. Microbiology. 144, 1465–1478. 2. Danilenko V.N., Akopiants K.E. 1995. Genomic instability and silent genes of actinomycetes. In Proceedings of the 9th Int. Symp. on the Biology of Actinomycetes. Eds. Debabov G., Dudnik G., Danilenko V. Moscow: Bioinform, pp. 104–112. 3. Danilenko V.N. 2000. Streptomyces avermitilis strain NICB 132 – producer of avermectins. Russian Federation Pat. No 2147320. 4. Danilenko V.N. 2000. Streptomyces fradiae strain NICB 99 – producer of tylosin. Russian Federation Pat. No 2147319. 5. Danilenko V.N. 1999. Saccharopolyspora erythraea strain 52 – producer of erythromycin. Russian Federation Pat. No 2142509. 6. Cole S.T., Brosch R., Parkhill J. et al. 1998. Deciphering the biology of Mycobacterium tuberculosis from the complete genomic sequence. Nature. 393, 537–544. 7. Bentley S.D., Chater K.F., Cerdeno-Tarraga A.-M. et al. 2002. Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2). Nature. 417, 141–147. 8. Zhang C.C. 1996. Bacterial signaling involving eukaryotic-type kinases. Mol. Microbiol. 20, 9–15. 9. Matsumoto A., Hong S.K., Ishizuka H., Horinouchi S., Beppu T. 1994. Phosphorylation of AfsR protein involved in secondary metabolism in Streptomyces species by a eukaryotic-type protein kinase. Gene. 146, 47–56. 10. Umeyama T., Lee P.C., Ueda K., Horinouchi S. 1999. An AfsK/AfsR system involved in the response of mycelium formation to glucose in Streptomyces griseus. Microbiology. 145, 2281–2292. 11. Ikeda H., Nonomiya T., Usami M., Ohita T., Omura S. 1999. Organization of the biosynthetic gene cluster for the polyketide antihelmintic macrolide avermectin in Streptomyces avermitilis. Proc. Natl Acad. Sci. USA. 96, 9509–9514. 12. Ikeda H., Nonomiya T., Omura S. 2001. Organization of biosynthetic gene cluster for avermectin in Streptomyces avermitilis: analysis of enzymatic domains in four polyketide synthases. J. Ind. Microbiol. Biotechnol. 27, 170–176. 13. Elizarov S.M., Mironov V.A., Danilenko V.N. 2000. Changes in activity of serine/threonine protein kinases during growth of a chloramphenicol-resistant mutant
14.
15.
16.
17.
18.
19.
20.
21.
22. 23.
24.
25. 26.
27.
28.
and an original strain of Streptomyces avermitilis. Mikrobiologiya. 69, 345–351. Ikeda H., Kotaki H., Omura S. 1987. Genetic studies of avermectin biosynthesis in Streptomyces avermitilis. J. Bacteriol. 169, 5615–5621. Laemmli U.K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 227, 680–685. Malygin A.G. 1993. Two-dimensional polyacrylamide gel electrophoresis of proteins: Prospects and technology. Usp. Biol. Khim. 30, 173–212. Kameshita I. Fujisawa H. 1989. A sensitive method for detection of calmodulin-dependent protein kinase II activity in sodium dodecyl sulfate – polyacrylamide gel. Anal. Biochem. 183, 139–143. Elizarov S.M., Mironov V.A., Danilenko V.N. 2000. Calcium-induced alterations in the functioning of protein serine/threonine and tyrosine kinases in Streptomyces fradiae cells. Life. 50, 139–143. Elizarov S.M., Danilenko V.N. 2001. Multiple phosphorylation of membrane associated calcium dependent protein serine/threonine kinase in Streptomyces fradiae. FEMS Microbiol. Lett. 202, 135–138. Bradford M.M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. Natsume M. 1992. Calcium signal modulators inhibit aerial mycelium formation in Streptomyces alboniger. J. Antibiot. 45, 1026–1028. Cohen P., Klee C.B. 1987. Calmodulin. Amsterdam: Elsevier. Umeyama T., Horinouchi S. 2001. Autophosphorylation of a bacterial serine/threonine kinase, AfsK, is inhibited by KbpA, an AfsK-binding protein. J. Bacteriol. 183, 5506–5512. Neu J.M., Wright G.D. 2001. Inhibition of sporulation, glycopeptide antibiotic production and resistance in Streptomyces toyocaensis NRRL 15009 by protein kinase inhibitors. FEMS Microbiol. Lett. 199, 15–20. Norris V., Grant S., Freestone P. et al. 1996. Calcium signaling in Bacteria. J. Bacteriol. 178, 3677–3682. Ikeda H., Ishikawa J., Hanamoto A. et al. 2003. Complete genome sequence and comparative analysis of the industrial microorganism Streptomyces avermitilis. Nature Biotechnol. 21, 526–531. Elizarov S.M., Sergienko O.V., Danilenko V.N. 2002. Phosphorylation of APHVIII and putative oxytetracycline biosynthesis enzymes by serine/threonine protein kinases in Streptomyces rimosus. 1-i Mezhdunarodnyi congress “Biotekhnologiya—sostoyanie i perspektivy razvitiya (1st Int. Congr. “Biotechnology: Current State and Prospects”). 40–41. Waters B., Saxena G., Wanggui Y. et al. 2002. Identifying protein kinase inhibitors using an assay based on inhibition of aerial hyphae formation in Streptomyces. J. Antibiotics. 55, 407–416.
MOLECULAR BIOLOGY
Vol. 38
No. 3
2004
