Correlation between calmodulin-like protein, phospholipids, and growth in glucose-grown Mycobacterium phlei P. HEMALATA REDDY,'SASTRY S. B U R R AAND , ~ P. SURYANARAYANA MURTHY~ Department of Biochemistry, University College of Medical Sciences, and G. T.B. Hospital, Shahdara, Delhi- 110095, India Received July 5, 1991 Revision received November 29, 1991
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by University of Western Ontario on 11/15/14 For personal use only.
Accepted December 2, 1991 REDDY,P . H., BURRA,S. S., and MURTHY, P . S. 1992. Correlation between calmodulin-like protein, phospholipids, and growth in glucose-grown Mycobacterium phlei. Can. J. Microbiol. 38: 339-342. In Mycobacterium phlei TMC 1548 supplementation of growth medium containing 2% v/v glycerol with glucose (up to 5% W/V)resulted in an increase in growth (yield of cells), in amount of total phospholipids, and in each of the individual phospholipids (cardiolipin, phosphatidylethanolamine, phosphatidylinositol and its mannosides, and phosphatidylglycerol). However, when the medium was supplemented with a higher concentration (7.5% w/v) of glucose, both growth and phospholipid levels decreased to near control values (2% v/v glycerol alone). Cyclic AMP levels, which decreased at all concentrations of glucose, had no relation to phospholipid content or growth. The presence of a protein that possesses the property of stimulating c-AMP phosphodiesterase activity was recently demonstrated in Mycobacterium smegmatis (Falah et al. 1988. FEMS Microbiol. Lett. 56: 89-93). In M. phlei the level of this calmodulin-like protein (assayed by radioimmunoassay) changed with different concentrations of glucose in the growth medium in a manner identical with that of phospholipids. We suggest that in mycobacteria (i) intracellular calmodulin-like protein levels are affected by glucose concentration in the growth medium and (ii) there is a positive correlation between the levels of calmodulin-like protein, total and individual phospholipids, and growth (yield of cells) in glucose-grown M. phlei. Key words: Mycobacterium phlei, calmodulin-like protein, phospholipids, growth. REDDY,P . H., BURRA,S. S., et MURTHY,P . S. 1992. Correlation between calmodulin-like protein, phospholipids, and growth in glucose-grown Mycobacterium phlei. Can. J. Microbiol. 38 : 339-342. Une augmentation de la croissance (rendement de cellules), de la quantite des phospholipides totaux et individuels (cardiolipine, phosphatidylethanolamine, phosphatidylinositol et ses mannosides, phosphatidylglycerol) de Mycobacterium phlei TMC 1548 a ete obtenue apres avoir ajoute du glucose Cjusqu'a 5% p/v) au milieu de croissance contenant du glycerol (2% v/v). Cependant, la croissance et les niveaux de phospholipides ont diminue pres des valeurs temoins (glycerol seul, 2% v/v) lorsque le glucose a ete ajoute a une concentration plus elevee (7,5% p/v). Les niveaux d'AMP cyclique (AMPc), qui ont diminue a toutes les concentrations de glucose, n'avaient aucune relation avec le contenu en phospholipide ou avec la croissance. Les auteurs ont recemment demontre la presence, chez Mycobacterium smegmatis, d'une proteine qui possede la propriete de stimuler l'activite AMPc phosphodiesterase (Falah et al. 1988. FEMS Microbiol. Lett. 56: 89-93). Le niveau de cette proteine similaire a la calmoduline (dosage radio-immunologique) chez M. phlei a change, selon les differentes concentrations de glucose dans le milieu de croissance, d'une facon exactement analogue a celle des phospholipides. Les auteurs proposent (i) que chez les mycobacteries, les niveaux intracellulaires de la proteine similaire a la calmoduline sont affectes par la concentration de glucose dans le milieu de croissance et (ii) qu'il existe une correlation positive entre d'une part, les niveaux de proteine similaire a la calmoduline, de phospholipides totaux et individuels et d'autre par la croissance (rendement de cellules) de M. phlei cultive en presence de glucose. Mots clPs : Mycobacterium phlei, proteine similaire a la calmoduline, phospholipides, croissance. [Traduit par la redaction]
Mathur et al. (1976) found that when glucose (up to 5% W/V)was added as a carbon source to 2070 v/v glycerol medium, there was an increase in the content of total lipids and total phospholipids, particularly cardiolipin, in cells of Mycobacterium smegmatis. Beyond a 5% w/v glucose supplement there was a decrease in the phospholipid content of the cells at higher glucose concentrations (7.5% w/v and above). Individual phospholipid levels were not estimated. This phenomenon was attributed to catabolite repression. Later Padh and Venkitasubramanian (1980) pointed out a lack of expression of cyclic AMP (CAMP)functions and the
'present address: Department of Biochemistry, Srivenkateswara College, Dhaulakuan, New Delhi, India. 2 ~ r e s e naddress: t Department of Biochemistry, University of Texas Health Sciences Centre, San Antonio, TX, U.S.A. 3 ~ u t h oto r whom all correspondence should be addressed. Printed in Canada / lmprime au Canada
absence of CAMP receptor protein in M. smegmatis. Our present studies indicate that there is no relationship between the changes in phospholipid content and CAMP. The ubiquitous protein calmodulin is present in all eukaryotic cells, including yeasts (Manalan and Klee 1984). Evidence has been presented for the presence of calmodulinlike protein in some prokaryotes (Iwasa et al. 1981; Kerson et al. 1984; Harmon et al. 1985; Fry et al. 1986, 1991; Swan et al. 1987; Shyu and Foegeding 1989, 1991; Petterson and Bergman 1989). Recently Falah et al. (1988) demonstrated the presence of calmodulin-like protein in two species of myco bacteria, M. smegmatis and Mycobacterium bovis BCG. Nothing is known about it in other mycobacteria. Even though the gene for a calcium-binding protein was cloned and sequenced in Saccharopolyspora erythraea (Swan el a[. 1989), the function of this calcium-binding protein in this organism is unknown. No calmodulin-like properties have been reported for it other than an E F hand.
CAN. J . MICROBIOL. VOL. 38, 1992
TABLE1. Growth, total and individual phospholipids, and intracellular levels of cAMP and CAMLP in M. phlei grown with different concentrations of glucose
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by University of Western Ontario on 11/15/14 For personal use only.
Glucose concn. (% , w/v)
Growth/3 days Total PLs Cardiolipin Phosphatidylethanolamine P hosp hatidylglycerol Phosphatidylinositol Phosphatidylinositol mannosides CAMP CAMLP
1.23 110 25 18 9 15 32 15.7 4.8
2.25 (1.8) 203 (1.8) 37 (1.5) 30 (1.7) 13 (1.4) 17 (1.1) 71 (2.2) 10.6 (0.67) 6.0 (1.3)
2.08 (1.7) 310 (2.8) 42 (1.7) 33 (1.8) 19 (2.1) 29 (1.9) 138 (4.3) 9.3 (0.59) 7.5 (1.6)
NOTE:All the cultures were grown for 3 days on Youmans and Karlson's medium and contained 2% v/v glycerol irrespective of the amount of glucose added. Values of growth are expressed as grams dry weight per 100 mL medium; phospholipids are expressed as micrograms PL phosphorus, cAMP as nanomoles, and CAMLP as nanograms per 100 mg dry weight of cells. Each value is the average of three separate experiments. Numbers in parentheses denote the fold increase or decrease of the yield of cells or PL content with respect to control values.
In spite of the vast literature on the role of calmodulin in animal systems, no specific function could be attributed to calmodulin in any of the bacterial systems, although its presence was demonstrated. We therefore looked into the probable role of calmodulin-like protein (CAMLP) in Mycobacterium phlei, which has not been previously studied. We report here that there is a direct correlation between the concentration of glucose in the growth medium (up to 5% W/V)and the intracellular levels of calmodulin-like protein (but not CAMP), the total as well as all the individual phospholipids, and probably growth of M. phlei. Materials and methods
Growth of the organism Mycobacterium phlei TMC 1548 was obtained from Trudeau Mycobacterial Culture Collection Centre, National Jewish Hospital, Denver, Colo., and grown as a surface culture on Youmans and Karlson's (1947) medium containing 2% v/v glycerol as carbon source. This medium was supplemented with glucose at different concentrations (2.5, 5.0, and 7.5% w/v). The cells at mid log phase (3 days) were rapidly harvested in the cold (4°C) for various estimations. Preparation of cell walls and cell membranes Mycobacterium phlei cells were grown as described above and disintegrated in a Vibronics 15-kHz ultrasonic disintegrator, and the cell wall and cell membrane fractions were prepared by differential centrifugation as described (Reddy et al. 1975). The cell wall fraction was sedimented between 4 000 and 16 000 x g for 30 min each, and the membrane fraction was sedimented by further centrifugation of the 16 000 x g supernatant fraction at 100 000 x g for 90 min. The integrity of the membranes was checked by their ability to carry out oxidative phosphorylation.
Extraction and estimation of phospholipids and incorporation of 3 2 into ~ phospholipids Lipids were extracted by the method of Folch et al. (1957). Phospholipids were separated by thin-layer chromatography, using the solvent system chloroform-methanol-ammonia (1 15:45 :7.5, by volume). Phospholipid phosphorus was estimated by the method of Marinetti (1962). For 3 2 ~incorporation studies, the endogenous phosphate pool of the washed cell suspensions was minimized by incubation at 37OC with shaking for 2 h. Sterile carrier-free 3 2 was ~ added, followed by incubation for an additional 24 h. The phospholipids were extracted and separated as above, and radioactivity was determined (Dewan et al. 1987).
Cyclic AMP assay The cells were very rapidly chilled in 5% ice-cold trichloroacetic acid. They were disintegrated in a Vibronics 15-kHz ultrasonic oscillator with 5% trichloroacetic acid for 4 min and centrifuged at 30 000 x g in a Sorvall RC 5B centrifuge for 30 min (both at 0-4°C). In the supernatant fraction, which contains membrane and cytosolic fractions, cAMP was estimated by the method of Gilman (1970). Calmodulin-like protein assay The cells (2 g) were sonicated in 0.1 M Tris-HC1 buffer, pH 7.0 (20 mL), for 4 min in an ice bath. The supernatant fraction was heated in a boiling-water bath for exactly 4 min to destroy all endogenous proteolytic activity, immediately cooled in an ice bath, and centrifuged at 17 000 x g at 4OC for 30 min. In the supernatant fraction, CAMLP was determined following the protocol in the kit for radioimmunoassay of calmodulin from New England Nuclear (NEN) Products, Boston, Mass. Radioactivity was counted in a Kontron gamma counter, optimized for ' 2 5 ~ . This method uses antibody to bovine calmodulin.
Results Effect of glucose in growth medium on total and individual phospholipids It is evident from the results in Table 1 that there was a gradual increase in the amount of total as well as the individual phospholipids (PLs) of the whole cells of M. phlei as the concentration of glucose in the growth medium increased and reached a maximum at 5070 w/v glucose. The phospholipid and other constituents in the cells grown without glucose (which, however, had 2% v/v glycerol) served as control values. The maximum increase was seen in phosphatidylinositol mannosides (4.3-fold), while in others it was about 1.7- to 3-fold. However, at a higher glucose concentration, 7.5% w/v, the total and individual P L content came down to near control values. It was also observed that the increase in the total and individual PLs occurred both in the cell walls and membranes and in the whole cells (Table 2). Growth The yield (dry weight) of the cells harvested at mid log phase (3 days) increased at 2.5% w/v and 5.0% w/v added glucose but was much lower (near normal) at 7.5070 w/v
REDDY ET AL.
TABLE2. Composition of total and individual phospholipids in the cell walls and membranes of M. phlei supplemented with different concentrations of glucose Glucose concn. (% , w/v)
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by University of Western Ontario on 11/15/14 For personal use only.
Phospholipids
0
2.5
5 .O
7.5
Total CW CM Cardiolipin CW CM Phosphatidylethanolamine CW CM Phosphatidylglycerol CW CM Phosphatidylinositol CW CM Phosphatidylinositol mannosides CW CM NOTE: Cell walls (CW) and cell membranes (CM) were prepared from whole cells as described in Materials and methods. Other details are as for Table 1.
glucose. The lower growth in 7.5% w/v glucose medium was seen up to 5 days.
Cyclic AMP Since glucose is known to influence the cellular content were estimated' There was of the levels of a continuous decrease in cAMP level as glucose concentration increased (Table 1).
Calmodulin-like protein Since a protein similar to eukaryotes calmodulin, stimulating phosphodiesterase activity, is present in mycobacteria (Falah et al. 1988), we determined CAMLP to find out whether its levels would have any relationship to the PL content. Results in Table 1 indicate that the CAMLP content also increased gradually from the control value, reached a maximum at 5070 w/v glucose, and decreased nearly to the control value at 7.5070 w/v glucose. Calcium content After disintegrating the cells as described above, calcium ion content was estimated in the 100 000 x g supernatant by the Corning calcium analyser (model 940). In the control cells, the total calcium content was 3.3 mg (average of three experiments) per 100 mg dry weight of cells. However, calcium was not detectable in cells grown with any concentration of glucose.
Effect of trifluoperazine Trifluoperazine (0.1 mM) when added to cells at mid log ~ total PLs. In phase reduced the incorporation of 3 2 into one typical separate experiment, it reduced the counts per minute to 70% that seen in the control cells (15 x 1o4 cpm/ 100 mg dry weight cells). Discussion The gradual increase in total PL content of whole cells up to 5.0070 w/v glucose in the growth medium and the
decrease (Table 1) with a further increase in glucose concentration (7.5% w/v) indicate that glucose modulates PL content of M. phlei. It is to be kept in mind that the M. phlei was grown in a semisynthetic medium as a surface culture. It is not known whether the same situation would hold in shake cultures and in rich medium. After separation, the individual phospholipids (cardiolipin, phosphatidylethanolamine, phosphatidylglycerol, which accounted for 83-90% of the total PLs) also increased or decreased like the total PLs (Table 1). Changes similar to those observed in the PLs of whole cells were also noticed in PLs of cell wall and membrane fractions (Table 2). It appears that the glucose effect is not restricted to cardiolipin only (Mathur et al. 1976). Rather, it is a general phenomenon applicable to all the PLs on the cell walls and membranes in at least two species of mycobacteria, M. phlei and M. smegmatis. No attempt was made to find whether the above changes were due to those in the synthesis or breakdown of PLs. With changes in glucose concentration in the medium, dry weight of the cells harvested at mid log phase of growth remained higher, as in the case of PLs, up to 5070 w/v glucose (highest at 2.5% w/v) and decreased at a higher concentration. Although confirmation by viable count of the bacilli is necessary, it could perhaps be inferred that growth yield (not necessarily growth rate), like PLs, was also influenced bv- -nlucose. The question that arises is whether there is any intracellular compound influenced by glucose in the medium that could regulate PL metabolism and thereby affect the levels of all PLs. It is unlikely that cAMP is affecting .the PLs in view of .the lack of correlation between .the two. The present study on the levels of calmodulin-like protein (Table 1) at different concentrations of glucose assumes importance from two considerations. Firstly, the results show that CAMLP levels were influenced by glucose, a phenomenon demonstrated for the first time. Secondly, there is a posi-
342
CAN. J . MICROBIOL. VOL. 38, 1992
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by University of Western Ontario on 11/15/14 For personal use only.
tive correlation in M. phlei between the levels of CAMLP, the amount of total and individual PLs, and probably growth (yield) of cells (Table 1). One could argue that these changes are by coincidence and may be independent of each other. However, the above results are to be interpreted keeping in mind the following considerations. Our earlier observations (Falah et al. 1988) showed that the mycobacterial CAMLP possesses the well-recognised biological property of the eukaryotic calmodulin, namely the stimulation of phosphodiesterase. Although unlikely, the art ifactual stimulation of phosphodiesterase is possible. Furthermore, in the present study CAMLP was estimated by the radioimmunoassay kit of NEN containing antibodies to bovine calmodulin. This is an indication that there is sequence homology between the mycobacterial calmodulin-like protein and eukaryotic calmodulin, most probably in the highly conserved calcium-binding EF-hand motif. This view is supported by the observation of Swan et al. (1987, 1989) that the EF-hand motif of S.,erythraea bears a remarkable similarity to the four calcium-binding sites of human calmodulin. In short, the M. phlei CAMLP possesses one biological property of calmodulin and shows immunological relatedness to calmodulin, which regulates a wide range of cellular activities (Manalan and Klee 1984; Grand 1985; Winkler et al. 1987) in eukaryotes. Therefore, calmodulinlike protein, for which no specific function has been attributed to date, could have some role in PL metabolism in mycobacteria. From the level of calcium, it is difficult to draw any conclusion. The inhibition of incorporation of 3 2 by ~ trifluoperazine could be due to its property of antagonism to calmodulin or direct action on PL metabolism. The recent reports that calmodulin-like protein has been purified from Bacillus subtilis (Fry et al. 1991) and that calmodulin antagonists inhibit the germination of spores in Bacillus cereus (Shyu and Foegeding 1991) are of interest.
Acknowledgements The authors thank Professor S.K. Sharma. Department of Biochemistry, All-India Institute of Medical Sciences, New Delhi, for assistance with the CAMP assay. The authors also express gratitude for a grant (to P.S.M.) from the Indian Council of Medical Research, New Delhi.
Dewan, R., Gangal, S.V., Chandrasekhar, R., and Suryanarayana Murthy, P. 1987. Turnover of phospholipids in whole cells and membranes of Mycobacterium tuberculosis H37Rv. Indian J . Biochem. Biophys. 24: 278-281. Falah, A.M.S., Bhatnagar, R., Bhatnagar, N., et al. 1988. On the presence of calmodulin-like protein in mycobacteria. FEMS Microbiol. Lett. 56: 89-93. Folch, J., Lees, M., and Stanley, G.H.S. 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226: 497-509. Fry, I.J., Villa, L., Kuehn, G.D., and Hageman, J.H. 1986. Calmodulin-like protein from Bacillus subtilis. Biochem. Biophys. Res. Commun. 134: 212-21 7.
Fry, I.J., Becker, M., and Hageman, J.H. 1991. Purification and properties of calmodulin-like protein from Bacillus subtilis cells. J. Bacteriol. 173: 2506-25 13. Gilman, A.G. 1970. A protein binding assay for 3 ' , 5'-cyclic adenosine monophosphate. Proc. Natl. Acad. Sci. U.S.A. 67: 305-3 12. Grand, R. J.A. 1985. A comparative study of the occurrence, structure and function of troponin C and calmodulin. In Metalloproteins. Vol. 7. Part 2. Edited by P. Harrison. Topics in molecular and structural biology. Verlag Chemie, Basel. pp. 65-122. Harmon, A.C., Prasher, D., and Cormier, M. J. 1985. High affinity calcium-binding proteins in Escherichia coli. Biochem. Biophys. Res. Commun. 127: 31-36. Iwasa, Y., Yonematsu, K., and Matsui, K. 1981. Calmodulin-like activity in the soluble fraction of Escherichia coli. Biochem. Biophys. Res. Commun. 98: 656-660. Kerson, G. W., Miernyk, J .A., and Budd, K. 1984. Evidence for the occurrence of, and possible physiological role for cyanobacterial calmodulin. Plant. Physiol. 75: 222-224. Manalan, A.S., and Klee, C.B. 1984. Calmodulin. Adv. Cyclic Nucleotide Protein Phosphorylation Res. 18: 227-278. Marinetti, G.V. 1962. Chromatographic separation, identification and analysis of phosphatides. J. Lipid Res. 3: 1-20. Mathur, A.K., Suryanarayana Murthy, P., Saharia, G.S., and Venkitasubramanian, T.A. 1976. Studies on cardiolipin biosynthesis in Mycobacterium smegmatis. Can. J. Microbiol. 22: 354-358. Padh, H., and Venkitasubramanian, T.A. 1980. Lack of adenosine-3 ' ,5 ' -monophosphate receptor protein and apparent lack of expression of adenosine-3 ' ,5 ' -monophosphate functions in Mycobacterium smegmatis CDC 46. Microbios, 27: 69-78. Petterson, A., and Bergman, B. 1989. Calmodulin in heterocystous Cyanobacteria-biochemical and immunological evidence. FEMS Microbiol. Lett. 60: 95-100. Reddy, T.L.P., Suryanarayana Murthy, P., and Venkitasubramanian, T.A. 1975. Variations in the pathways of malate oxidation and phosphorylation in different species of mycobacteria. Biochim. Bio~hvs.Acta, 376: 210-218. Shyu, Y.T., and ~ o e ~ e d i nP.M. g , 1989. Presence of calmodulinlike calcium-binding protein in Bacillus cereus T spores. FEMS Microbial. Lett. 59: 235-240. Shyu, Y.T., and Foegeding, P.M. 1991. Calmodulin antagonists inhibit germination of Bacillus cereus T spores. J. Appli. Bacteriol. 70: 233-238. Swan, D.G., Hale, R.S., Dhillon, N., and Leadlay, P.F. 1987. A bacterial calcium-binding protein homologous to calmodulin. Nature (London), 329: 84185. Swan, D.G., Cortes, J., Hale, R.S., and Leadlay, P.F. 1989. Cloning characterization, and heterologous expression of the Saccharopolyspora erythraea (Streptomyces erythraeus) gene encoding an EF-hand calcium-binding protein. J. Bacteriol. 171: 5614-5619. Winkler, M.A., Dewitt, L.M., and Cheung, W.Y. 1987. Calmodulin and calcium channel blockers. Hypertension, 9: 217-223. Youmans, G.P., and Karlson, A.G., 1947. Streptomycin sensitivity of tubercle bacilli: studies on recently isolated tubercle bacilli and the development of resistance to streptomycin in vivo. Am. Rev. Tuberc. 55: 529-535.
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by University of Western Ontario on 11/15/14 For personal use only.
Accepted December 2, 1991 REDDY,P . H., BURRA,S. S., and MURTHY, P . S. 1992. Correlation between calmodulin-like protein, phospholipids, and growth in glucose-grown Mycobacterium phlei. Can. J. Microbiol. 38: 339-342. In Mycobacterium phlei TMC 1548 supplementation of growth medium containing 2% v/v glycerol with glucose (up to 5% W/V)resulted in an increase in growth (yield of cells), in amount of total phospholipids, and in each of the individual phospholipids (cardiolipin, phosphatidylethanolamine, phosphatidylinositol and its mannosides, and phosphatidylglycerol). However, when the medium was supplemented with a higher concentration (7.5% w/v) of glucose, both growth and phospholipid levels decreased to near control values (2% v/v glycerol alone). Cyclic AMP levels, which decreased at all concentrations of glucose, had no relation to phospholipid content or growth. The presence of a protein that possesses the property of stimulating c-AMP phosphodiesterase activity was recently demonstrated in Mycobacterium smegmatis (Falah et al. 1988. FEMS Microbiol. Lett. 56: 89-93). In M. phlei the level of this calmodulin-like protein (assayed by radioimmunoassay) changed with different concentrations of glucose in the growth medium in a manner identical with that of phospholipids. We suggest that in mycobacteria (i) intracellular calmodulin-like protein levels are affected by glucose concentration in the growth medium and (ii) there is a positive correlation between the levels of calmodulin-like protein, total and individual phospholipids, and growth (yield of cells) in glucose-grown M. phlei. Key words: Mycobacterium phlei, calmodulin-like protein, phospholipids, growth. REDDY,P . H., BURRA,S. S., et MURTHY,P . S. 1992. Correlation between calmodulin-like protein, phospholipids, and growth in glucose-grown Mycobacterium phlei. Can. J. Microbiol. 38 : 339-342. Une augmentation de la croissance (rendement de cellules), de la quantite des phospholipides totaux et individuels (cardiolipine, phosphatidylethanolamine, phosphatidylinositol et ses mannosides, phosphatidylglycerol) de Mycobacterium phlei TMC 1548 a ete obtenue apres avoir ajoute du glucose Cjusqu'a 5% p/v) au milieu de croissance contenant du glycerol (2% v/v). Cependant, la croissance et les niveaux de phospholipides ont diminue pres des valeurs temoins (glycerol seul, 2% v/v) lorsque le glucose a ete ajoute a une concentration plus elevee (7,5% p/v). Les niveaux d'AMP cyclique (AMPc), qui ont diminue a toutes les concentrations de glucose, n'avaient aucune relation avec le contenu en phospholipide ou avec la croissance. Les auteurs ont recemment demontre la presence, chez Mycobacterium smegmatis, d'une proteine qui possede la propriete de stimuler l'activite AMPc phosphodiesterase (Falah et al. 1988. FEMS Microbiol. Lett. 56: 89-93). Le niveau de cette proteine similaire a la calmoduline (dosage radio-immunologique) chez M. phlei a change, selon les differentes concentrations de glucose dans le milieu de croissance, d'une facon exactement analogue a celle des phospholipides. Les auteurs proposent (i) que chez les mycobacteries, les niveaux intracellulaires de la proteine similaire a la calmoduline sont affectes par la concentration de glucose dans le milieu de croissance et (ii) qu'il existe une correlation positive entre d'une part, les niveaux de proteine similaire a la calmoduline, de phospholipides totaux et individuels et d'autre par la croissance (rendement de cellules) de M. phlei cultive en presence de glucose. Mots clPs : Mycobacterium phlei, proteine similaire a la calmoduline, phospholipides, croissance. [Traduit par la redaction]
Mathur et al. (1976) found that when glucose (up to 5% W/V)was added as a carbon source to 2070 v/v glycerol medium, there was an increase in the content of total lipids and total phospholipids, particularly cardiolipin, in cells of Mycobacterium smegmatis. Beyond a 5% w/v glucose supplement there was a decrease in the phospholipid content of the cells at higher glucose concentrations (7.5% w/v and above). Individual phospholipid levels were not estimated. This phenomenon was attributed to catabolite repression. Later Padh and Venkitasubramanian (1980) pointed out a lack of expression of cyclic AMP (CAMP)functions and the
'present address: Department of Biochemistry, Srivenkateswara College, Dhaulakuan, New Delhi, India. 2 ~ r e s e naddress: t Department of Biochemistry, University of Texas Health Sciences Centre, San Antonio, TX, U.S.A. 3 ~ u t h oto r whom all correspondence should be addressed. Printed in Canada / lmprime au Canada
absence of CAMP receptor protein in M. smegmatis. Our present studies indicate that there is no relationship between the changes in phospholipid content and CAMP. The ubiquitous protein calmodulin is present in all eukaryotic cells, including yeasts (Manalan and Klee 1984). Evidence has been presented for the presence of calmodulinlike protein in some prokaryotes (Iwasa et al. 1981; Kerson et al. 1984; Harmon et al. 1985; Fry et al. 1986, 1991; Swan et al. 1987; Shyu and Foegeding 1989, 1991; Petterson and Bergman 1989). Recently Falah et al. (1988) demonstrated the presence of calmodulin-like protein in two species of myco bacteria, M. smegmatis and Mycobacterium bovis BCG. Nothing is known about it in other mycobacteria. Even though the gene for a calcium-binding protein was cloned and sequenced in Saccharopolyspora erythraea (Swan el a[. 1989), the function of this calcium-binding protein in this organism is unknown. No calmodulin-like properties have been reported for it other than an E F hand.
CAN. J . MICROBIOL. VOL. 38, 1992
TABLE1. Growth, total and individual phospholipids, and intracellular levels of cAMP and CAMLP in M. phlei grown with different concentrations of glucose
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by University of Western Ontario on 11/15/14 For personal use only.
Glucose concn. (% , w/v)
Growth/3 days Total PLs Cardiolipin Phosphatidylethanolamine P hosp hatidylglycerol Phosphatidylinositol Phosphatidylinositol mannosides CAMP CAMLP
1.23 110 25 18 9 15 32 15.7 4.8
2.25 (1.8) 203 (1.8) 37 (1.5) 30 (1.7) 13 (1.4) 17 (1.1) 71 (2.2) 10.6 (0.67) 6.0 (1.3)
2.08 (1.7) 310 (2.8) 42 (1.7) 33 (1.8) 19 (2.1) 29 (1.9) 138 (4.3) 9.3 (0.59) 7.5 (1.6)
NOTE:All the cultures were grown for 3 days on Youmans and Karlson's medium and contained 2% v/v glycerol irrespective of the amount of glucose added. Values of growth are expressed as grams dry weight per 100 mL medium; phospholipids are expressed as micrograms PL phosphorus, cAMP as nanomoles, and CAMLP as nanograms per 100 mg dry weight of cells. Each value is the average of three separate experiments. Numbers in parentheses denote the fold increase or decrease of the yield of cells or PL content with respect to control values.
In spite of the vast literature on the role of calmodulin in animal systems, no specific function could be attributed to calmodulin in any of the bacterial systems, although its presence was demonstrated. We therefore looked into the probable role of calmodulin-like protein (CAMLP) in Mycobacterium phlei, which has not been previously studied. We report here that there is a direct correlation between the concentration of glucose in the growth medium (up to 5% W/V)and the intracellular levels of calmodulin-like protein (but not CAMP), the total as well as all the individual phospholipids, and probably growth of M. phlei. Materials and methods
Growth of the organism Mycobacterium phlei TMC 1548 was obtained from Trudeau Mycobacterial Culture Collection Centre, National Jewish Hospital, Denver, Colo., and grown as a surface culture on Youmans and Karlson's (1947) medium containing 2% v/v glycerol as carbon source. This medium was supplemented with glucose at different concentrations (2.5, 5.0, and 7.5% w/v). The cells at mid log phase (3 days) were rapidly harvested in the cold (4°C) for various estimations. Preparation of cell walls and cell membranes Mycobacterium phlei cells were grown as described above and disintegrated in a Vibronics 15-kHz ultrasonic disintegrator, and the cell wall and cell membrane fractions were prepared by differential centrifugation as described (Reddy et al. 1975). The cell wall fraction was sedimented between 4 000 and 16 000 x g for 30 min each, and the membrane fraction was sedimented by further centrifugation of the 16 000 x g supernatant fraction at 100 000 x g for 90 min. The integrity of the membranes was checked by their ability to carry out oxidative phosphorylation.
Extraction and estimation of phospholipids and incorporation of 3 2 into ~ phospholipids Lipids were extracted by the method of Folch et al. (1957). Phospholipids were separated by thin-layer chromatography, using the solvent system chloroform-methanol-ammonia (1 15:45 :7.5, by volume). Phospholipid phosphorus was estimated by the method of Marinetti (1962). For 3 2 ~incorporation studies, the endogenous phosphate pool of the washed cell suspensions was minimized by incubation at 37OC with shaking for 2 h. Sterile carrier-free 3 2 was ~ added, followed by incubation for an additional 24 h. The phospholipids were extracted and separated as above, and radioactivity was determined (Dewan et al. 1987).
Cyclic AMP assay The cells were very rapidly chilled in 5% ice-cold trichloroacetic acid. They were disintegrated in a Vibronics 15-kHz ultrasonic oscillator with 5% trichloroacetic acid for 4 min and centrifuged at 30 000 x g in a Sorvall RC 5B centrifuge for 30 min (both at 0-4°C). In the supernatant fraction, which contains membrane and cytosolic fractions, cAMP was estimated by the method of Gilman (1970). Calmodulin-like protein assay The cells (2 g) were sonicated in 0.1 M Tris-HC1 buffer, pH 7.0 (20 mL), for 4 min in an ice bath. The supernatant fraction was heated in a boiling-water bath for exactly 4 min to destroy all endogenous proteolytic activity, immediately cooled in an ice bath, and centrifuged at 17 000 x g at 4OC for 30 min. In the supernatant fraction, CAMLP was determined following the protocol in the kit for radioimmunoassay of calmodulin from New England Nuclear (NEN) Products, Boston, Mass. Radioactivity was counted in a Kontron gamma counter, optimized for ' 2 5 ~ . This method uses antibody to bovine calmodulin.
Results Effect of glucose in growth medium on total and individual phospholipids It is evident from the results in Table 1 that there was a gradual increase in the amount of total as well as the individual phospholipids (PLs) of the whole cells of M. phlei as the concentration of glucose in the growth medium increased and reached a maximum at 5070 w/v glucose. The phospholipid and other constituents in the cells grown without glucose (which, however, had 2% v/v glycerol) served as control values. The maximum increase was seen in phosphatidylinositol mannosides (4.3-fold), while in others it was about 1.7- to 3-fold. However, at a higher glucose concentration, 7.5% w/v, the total and individual P L content came down to near control values. It was also observed that the increase in the total and individual PLs occurred both in the cell walls and membranes and in the whole cells (Table 2). Growth The yield (dry weight) of the cells harvested at mid log phase (3 days) increased at 2.5% w/v and 5.0% w/v added glucose but was much lower (near normal) at 7.5070 w/v
REDDY ET AL.
TABLE2. Composition of total and individual phospholipids in the cell walls and membranes of M. phlei supplemented with different concentrations of glucose Glucose concn. (% , w/v)
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by University of Western Ontario on 11/15/14 For personal use only.
Phospholipids
0
2.5
5 .O
7.5
Total CW CM Cardiolipin CW CM Phosphatidylethanolamine CW CM Phosphatidylglycerol CW CM Phosphatidylinositol CW CM Phosphatidylinositol mannosides CW CM NOTE: Cell walls (CW) and cell membranes (CM) were prepared from whole cells as described in Materials and methods. Other details are as for Table 1.
glucose. The lower growth in 7.5% w/v glucose medium was seen up to 5 days.
Cyclic AMP Since glucose is known to influence the cellular content were estimated' There was of the levels of a continuous decrease in cAMP level as glucose concentration increased (Table 1).
Calmodulin-like protein Since a protein similar to eukaryotes calmodulin, stimulating phosphodiesterase activity, is present in mycobacteria (Falah et al. 1988), we determined CAMLP to find out whether its levels would have any relationship to the PL content. Results in Table 1 indicate that the CAMLP content also increased gradually from the control value, reached a maximum at 5070 w/v glucose, and decreased nearly to the control value at 7.5070 w/v glucose. Calcium content After disintegrating the cells as described above, calcium ion content was estimated in the 100 000 x g supernatant by the Corning calcium analyser (model 940). In the control cells, the total calcium content was 3.3 mg (average of three experiments) per 100 mg dry weight of cells. However, calcium was not detectable in cells grown with any concentration of glucose.
Effect of trifluoperazine Trifluoperazine (0.1 mM) when added to cells at mid log ~ total PLs. In phase reduced the incorporation of 3 2 into one typical separate experiment, it reduced the counts per minute to 70% that seen in the control cells (15 x 1o4 cpm/ 100 mg dry weight cells). Discussion The gradual increase in total PL content of whole cells up to 5.0070 w/v glucose in the growth medium and the
decrease (Table 1) with a further increase in glucose concentration (7.5% w/v) indicate that glucose modulates PL content of M. phlei. It is to be kept in mind that the M. phlei was grown in a semisynthetic medium as a surface culture. It is not known whether the same situation would hold in shake cultures and in rich medium. After separation, the individual phospholipids (cardiolipin, phosphatidylethanolamine, phosphatidylglycerol, which accounted for 83-90% of the total PLs) also increased or decreased like the total PLs (Table 1). Changes similar to those observed in the PLs of whole cells were also noticed in PLs of cell wall and membrane fractions (Table 2). It appears that the glucose effect is not restricted to cardiolipin only (Mathur et al. 1976). Rather, it is a general phenomenon applicable to all the PLs on the cell walls and membranes in at least two species of mycobacteria, M. phlei and M. smegmatis. No attempt was made to find whether the above changes were due to those in the synthesis or breakdown of PLs. With changes in glucose concentration in the medium, dry weight of the cells harvested at mid log phase of growth remained higher, as in the case of PLs, up to 5070 w/v glucose (highest at 2.5% w/v) and decreased at a higher concentration. Although confirmation by viable count of the bacilli is necessary, it could perhaps be inferred that growth yield (not necessarily growth rate), like PLs, was also influenced bv- -nlucose. The question that arises is whether there is any intracellular compound influenced by glucose in the medium that could regulate PL metabolism and thereby affect the levels of all PLs. It is unlikely that cAMP is affecting .the PLs in view of .the lack of correlation between .the two. The present study on the levels of calmodulin-like protein (Table 1) at different concentrations of glucose assumes importance from two considerations. Firstly, the results show that CAMLP levels were influenced by glucose, a phenomenon demonstrated for the first time. Secondly, there is a posi-
342
CAN. J . MICROBIOL. VOL. 38, 1992
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by University of Western Ontario on 11/15/14 For personal use only.
tive correlation in M. phlei between the levels of CAMLP, the amount of total and individual PLs, and probably growth (yield) of cells (Table 1). One could argue that these changes are by coincidence and may be independent of each other. However, the above results are to be interpreted keeping in mind the following considerations. Our earlier observations (Falah et al. 1988) showed that the mycobacterial CAMLP possesses the well-recognised biological property of the eukaryotic calmodulin, namely the stimulation of phosphodiesterase. Although unlikely, the art ifactual stimulation of phosphodiesterase is possible. Furthermore, in the present study CAMLP was estimated by the radioimmunoassay kit of NEN containing antibodies to bovine calmodulin. This is an indication that there is sequence homology between the mycobacterial calmodulin-like protein and eukaryotic calmodulin, most probably in the highly conserved calcium-binding EF-hand motif. This view is supported by the observation of Swan et al. (1987, 1989) that the EF-hand motif of S.,erythraea bears a remarkable similarity to the four calcium-binding sites of human calmodulin. In short, the M. phlei CAMLP possesses one biological property of calmodulin and shows immunological relatedness to calmodulin, which regulates a wide range of cellular activities (Manalan and Klee 1984; Grand 1985; Winkler et al. 1987) in eukaryotes. Therefore, calmodulinlike protein, for which no specific function has been attributed to date, could have some role in PL metabolism in mycobacteria. From the level of calcium, it is difficult to draw any conclusion. The inhibition of incorporation of 3 2 by ~ trifluoperazine could be due to its property of antagonism to calmodulin or direct action on PL metabolism. The recent reports that calmodulin-like protein has been purified from Bacillus subtilis (Fry et al. 1991) and that calmodulin antagonists inhibit the germination of spores in Bacillus cereus (Shyu and Foegeding 1991) are of interest.
Acknowledgements The authors thank Professor S.K. Sharma. Department of Biochemistry, All-India Institute of Medical Sciences, New Delhi, for assistance with the CAMP assay. The authors also express gratitude for a grant (to P.S.M.) from the Indian Council of Medical Research, New Delhi.
Dewan, R., Gangal, S.V., Chandrasekhar, R., and Suryanarayana Murthy, P. 1987. Turnover of phospholipids in whole cells and membranes of Mycobacterium tuberculosis H37Rv. Indian J . Biochem. Biophys. 24: 278-281. Falah, A.M.S., Bhatnagar, R., Bhatnagar, N., et al. 1988. On the presence of calmodulin-like protein in mycobacteria. FEMS Microbiol. Lett. 56: 89-93. Folch, J., Lees, M., and Stanley, G.H.S. 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226: 497-509. Fry, I.J., Villa, L., Kuehn, G.D., and Hageman, J.H. 1986. Calmodulin-like protein from Bacillus subtilis. Biochem. Biophys. Res. Commun. 134: 212-21 7.
Fry, I.J., Becker, M., and Hageman, J.H. 1991. Purification and properties of calmodulin-like protein from Bacillus subtilis cells. J. Bacteriol. 173: 2506-25 13. Gilman, A.G. 1970. A protein binding assay for 3 ' , 5'-cyclic adenosine monophosphate. Proc. Natl. Acad. Sci. U.S.A. 67: 305-3 12. Grand, R. J.A. 1985. A comparative study of the occurrence, structure and function of troponin C and calmodulin. In Metalloproteins. Vol. 7. Part 2. Edited by P. Harrison. Topics in molecular and structural biology. Verlag Chemie, Basel. pp. 65-122. Harmon, A.C., Prasher, D., and Cormier, M. J. 1985. High affinity calcium-binding proteins in Escherichia coli. Biochem. Biophys. Res. Commun. 127: 31-36. Iwasa, Y., Yonematsu, K., and Matsui, K. 1981. Calmodulin-like activity in the soluble fraction of Escherichia coli. Biochem. Biophys. Res. Commun. 98: 656-660. Kerson, G. W., Miernyk, J .A., and Budd, K. 1984. Evidence for the occurrence of, and possible physiological role for cyanobacterial calmodulin. Plant. Physiol. 75: 222-224. Manalan, A.S., and Klee, C.B. 1984. Calmodulin. Adv. Cyclic Nucleotide Protein Phosphorylation Res. 18: 227-278. Marinetti, G.V. 1962. Chromatographic separation, identification and analysis of phosphatides. J. Lipid Res. 3: 1-20. Mathur, A.K., Suryanarayana Murthy, P., Saharia, G.S., and Venkitasubramanian, T.A. 1976. Studies on cardiolipin biosynthesis in Mycobacterium smegmatis. Can. J. Microbiol. 22: 354-358. Padh, H., and Venkitasubramanian, T.A. 1980. Lack of adenosine-3 ' ,5 ' -monophosphate receptor protein and apparent lack of expression of adenosine-3 ' ,5 ' -monophosphate functions in Mycobacterium smegmatis CDC 46. Microbios, 27: 69-78. Petterson, A., and Bergman, B. 1989. Calmodulin in heterocystous Cyanobacteria-biochemical and immunological evidence. FEMS Microbiol. Lett. 60: 95-100. Reddy, T.L.P., Suryanarayana Murthy, P., and Venkitasubramanian, T.A. 1975. Variations in the pathways of malate oxidation and phosphorylation in different species of mycobacteria. Biochim. Bio~hvs.Acta, 376: 210-218. Shyu, Y.T., and ~ o e ~ e d i nP.M. g , 1989. Presence of calmodulinlike calcium-binding protein in Bacillus cereus T spores. FEMS Microbial. Lett. 59: 235-240. Shyu, Y.T., and Foegeding, P.M. 1991. Calmodulin antagonists inhibit germination of Bacillus cereus T spores. J. Appli. Bacteriol. 70: 233-238. Swan, D.G., Hale, R.S., Dhillon, N., and Leadlay, P.F. 1987. A bacterial calcium-binding protein homologous to calmodulin. Nature (London), 329: 84185. Swan, D.G., Cortes, J., Hale, R.S., and Leadlay, P.F. 1989. Cloning characterization, and heterologous expression of the Saccharopolyspora erythraea (Streptomyces erythraeus) gene encoding an EF-hand calcium-binding protein. J. Bacteriol. 171: 5614-5619. Winkler, M.A., Dewitt, L.M., and Cheung, W.Y. 1987. Calmodulin and calcium channel blockers. Hypertension, 9: 217-223. Youmans, G.P., and Karlson, A.G., 1947. Streptomycin sensitivity of tubercle bacilli: studies on recently isolated tubercle bacilli and the development of resistance to streptomycin in vivo. Am. Rev. Tuberc. 55: 529-535.
