Calmodulin regulation of prolactin gene expression

[49]

Ca2÷/CaM

R E G U L A T I O N OF P R O L A C T I N G E N E EXPRESSION

655

[49] C a 2 + / C a l m o d u l i n R e g u l a t i o n o f P r o l a c t i n Gene Expression By BRUCE A . W H I T E a n d CARTER BANCROFT

Introduction The rat prolactin gene is one of the very few examples of a gene for which there is evidence of its expression being regulated by Ca 2÷ and calmodulin.1 Since this chapter is concerned with the methodology of the study of prolactin gene expression in normal and transformed rat pituitary cells, we will not present extensive experimental evidence for the Ca 2÷calmodulin regulation of prolactin gene expression. Instead, we refer the reader at the appropriate section to our previous publications which contain both the experimental procedures and results of our studies on the Ca2÷-calmodulin regulation of the prolactin gene. Culture of GHa Cells in a Defined Medium Central to our studies is the use of a serum-free medium for the culture of rat pituitary tumor GH3 cells. This medium, described in White et al., z is based on a serum substitute described by Bauer et al. 3 An advantage of using this medium is that it can be used to study prolactin gene expression in the absence of the many serum components which are known to regulate prolactin production. Since this medium contains no added Ca 2÷, its use also allows the study of the effects of cellular Ca 2÷ depletion and repletion on prolactin gene expression. Components o f Serum-Free Medium

a. 1640 ml of Joklik's modified minimal essential medium for suspension culture (e.g., Gibco 410-1300) b. 184 ml of serum substitute c. 18.4 ml of HEPES solution d. 0.9 ml of lipids solution

i B. A. White and C. 7. Academic Press, 2 B. A. White, L. R. 3 R. F. Bauer, L. O.

Bancroft, in "Calcium and Cell Function" (W. Y. Cheung, ed.), Vol. New York, 1987. Bauerle, and C. Bancroft, J. Biol. Chem. 256, 5942 (1981). Arthur, and D. L. Fine, In Vitro 12, 558 (1976).

METHODS IN ENZYMOLOGY, VOL. 139

Copyright © 1987 by Academic Press, Inc. All rights of reproduction in any form reserved.

656

REGULATION OF C a AND Ca-BINDING PROTEINS

[49]

Preparation of serum substitute 50 ml of nonessential amino acids mixture (e.g., Gibco 320-1140) 50 ml of minimal essential medium sodium pyruvate solution (e.g., Gibco 320-1360) I00 ml of 0.4 mM EDTA, pH 7.0 250 ml of sterile double-distilled deionized water 50 ml of vitamin-salt-gluconolactone solution

Preparation of vitamin-salt-gluconolactone solution Vitamin solution: 2 mg biotin 2 mg vitamin B12 Add 5 ml of sterile water, dissolve by shaking, sterile filter Gluconolactone solution: Dissolve 17.8 g gluconolactone in 300 ml water. Adjust pH to 7.0 with 10 N NaOH. Adjust volume to 500 ml Salt solution: 180 mg choline chloride 40 mg protamine sulfate 6 mg ZnSO4"7H20 96 mg FeNH4(SO4)E" 12HEO Add I00 ml of water, stir well to dissolve Add 100 ml of gluconolactone solution to 100 ml of salt solution and sterile filter. Add 100/zl of vitamin solution.

Preparation of HEPES solution 2.45 g NaCI 100 g HEPES Add 200 ml water, adjust pH to 7.4. Adjust volume to 280 ml. Sterile filter and store at - 2 0 °

Preparation of lipids solution 50.5/xl 99% pure oleic acid 100 mg lecithin 195 mg cholesterol Dissolve in 50 ml of 95% ethanol at room temperature. Store at 4 °

Subculture of GH3 Cells in Serum-Free Medium. Rat pituitary tumor (GH3) cells are grown in suspension culture in Joklik's modified minimal essential medium for suspension culture containing 15% horse serum, 2.5% fetal calf serum, and 0.5% HEPES solution (see above). At the time

[49]

Ca2+/CaM REGULATION OF PROLACTIN GENE EXPRESSION

657

of an experiment, cells are centrifuged out of the growth medium at 400 g for 5 rain in sterile 50-ml conical tubes and the supernatant decanted. The cells are then gently but thoroughly resuspended in prewarrned serumfree medium. Up to six cell pellets are pooled into one 50-ml tube, and the volume adjusted to 50 ml with serum-free medium. The cells are then centrifuged at 400 g for 5 min and the supernatant decanted. During this second centrifugation, 9 ml of prewarmed serum-free medium is added to the required number of tissue culture dishes (100 x 20 mm). The washed cell pellet is then resuspended in serum-free medium, and the volume adjusted to a few milliliters greater than the number of tissue culture dishes to be used for that particular experiment (e.g., resuspend cells in 23-25 ml for a 20-dish experiment). Cap tube and mix cells by gently inverting. Aliquot 1 ml of resuspended cells to each of five dishes using a 5-ml pipet (keep the pipet almost horizontal). Gently shake each dish to disperse the cells. Repeat aliquoting cells 5 ml at a time until finished. Culture cells overnight in a 37° incubator at 2% CO2. Inspection of cells on the following morning should reveal randomly dispersed, mostly attached cells that are about 90% viable. Treatment of GH3 Cells with Test Substances. After cells have been subcultured overnight in serum-free medium, they are treated with various test substances for up to 1 week. The following treatments have been used to obtain evidence for the regulation of prolactin gene expression by Ca 2+ and/or calmodulin. Induction of prolactin gene expression by CaCl2: White et al. 2 demonstrated that addition of 0.04-0.40 mM CaCI2 to cells cultured in serumfree medium specifically increased relative levels of prolactin mRNA and prolactin synthesis, without changing total protein synthesis or relative growth hormone synthesis. An increase in prolactin mRNA is observed after a lag period of several hours, and this increase continues for several days.4 A change in the appearance of the cells accompanies the induction of prolactin gene expression, in that cells become aggregated in organized end-to-end cords within a few hours after the addition of 0.4 mM CaC12. Induction of prolactin gene expression by hormones and growth factors: White et al. 2 showed that the synthetic glucocorticoid, dexamethasone, specifically antagonized the ability of Ca 2÷ to increase prolactin mRNA levels. In the same cultures, dexamethasone increased growth hormone mRNA levels independently of Ca 2+. This is an illustration of the specificity of the Ca 2÷ effect in GH3 cells and how the level of growth hormone gene expression can be used as a control in some experiments 4 B. A. White and C. Bancroft, J. Biol. Chem. 258, 4618 (1983).

658

REGULATION OF C a AND Ca-BINDING PROTEINS

[49]

with GH3 cells. White and Bancroft 4 also used culture of GH3 cells in serum-free medium to show that both thyrotropin-releasing hormone and epidermal growth factor require Ca 2÷ for their ability to maximally induce prolactin gene expression. Wark and Tashjian 5 have used similar culture conditions to demonstrate that the steroid hormone, 1,25-dihydroxyvitamin D3, also requires Ca 2+ for its ability to increase prolactin mRNA levels. Inhibition o f prolactin gene induction by calmodulin inhibitors : White 6 utilized subculture of GH3 cells in serum-free medium for the purpose of examining the effects of calmodulin inhibitors on the ability of epidermal growth factor plus Ca 2÷ to increase prolactin mRNA levels. The phenothiazine, calmidazolium, was used in most of these studies. Calmidazolium was purchased from Boehringer Mannheim Biochemicals, dissolved in 95% ethanol at a concentration of 5 mM, and stored at - 2 0 ° for up to 6 months. Calmidazolium inhibits prolactin gene expression in a concentration range of 0.05 to 1.00/zM. The specificity of the drug was assessed by measuring its effects on total protein and RNA synthesis (see below) and cellular aggregation. As mentioned above, GH3 cells in serumfree medium are randomly spread over the tissue culture dish. Addition of Ca 2÷ induces both prolactin gene expression and an end-to-end aggregation of cells. One indication of the specificity of calmidazolium was its ability to completely inhibit an increase in prolactin mRNA levels without disrupting the Ca2+-induced cell-cell aggregation. The Use of Primary Cultures of Rat Pituitary Cells for Ca 2÷ Regulation Studies The serum-free medium described above was employed by Gick and Bancroft 7 to study regulation by Ca 2+ of prolactin and growth hormone gene expression in normal rat pituitary cells. The techniques employed in these studies have been described in detail previously. 7 Briefly, when monolayer cultures of pituitary cells were incubated in serum-free medium in the presence of exogenously added Ca 2÷ at concentrations up to 1.0 mM, no effect on prolactin or growth hormone mRNA levels was observed. The use of the metallochromatic Ca 2÷ indicator arsenazo IIIs showed that the serum-free medium as prepared contained Ca 2÷ concentrations in the range of 10 to 40/zM. Hence, it seemed possible that these 5 j. D.-Wark and A. H. Tashjian, Jr., J. Biol. Chem. 258, 12118 (1983). 6 B. A. White, J. Biol. Chem. 260, 1213 (1985). 7 G. G. Gick and C. Bancroft, J. Biol. Chem. 260, 7614 (1985). s A. Scarpa, F. J. Brinkley, T. Tiffert, and G. R. Dubyak, Ann. N.Y. Acad. Sci. 307, 86 (1978).

[49]

Ca2÷/CaM REGULATION OF PROLACTIN GENE EXPRESSION

659

low Ca 2÷ levels were maximally stimulating production of these two mRNAs. To examine this question, the Ca 2÷ concentration of a given preparation of serum-free medium was lowered to any desired value by adding the appropriate amount of EGTA, calculated by employing a value of 0.15/zM for the Ko of Ca2÷-EGTA. It was then found that pituitary cells can maintain basal levels of prolactin and growth hormone mRNA down to quite low (micromolar) levels of extracellular Ca 2÷. However, as extracellular Ca 2÷ concentrations were lowered below 1.0/xM, cellular levels of both prolactin and growth mRNA exhibited a precipitous, dosedependent decrease. The observation that lowering extracellular Ca 2÷ to these levels did not decrease either overall RNA synthesis or the cellular levels of the mRNA for another metal-regulated protein, metallothionein, showed that Ca 2÷ exhibits specificity in its regulation of prolactin and growth gene expression in these cells. Finally, nuclear run-on transcription assays have shown that these effects of Ca 2÷ depletion are exerted at least partially at the transcriptional level. 9 Measurement

of Relative Levels of Prolactin mRNA

We have demonstrated previously that Ca 2÷ increases prolactin gene expression when measured at the level of transcription, 9 steady state nuclear levels of gene transcripts, 4 steady state cytoplasmic mRNA levels, 2 or the relative level of prolactin synthesis. 2 However, for routine assays, we employ the cytoplasmic dot hybridization technique described below to mesaure prolactin mRNA.

Cytoplasmic Dot Hybridization This procedure was originally developed by White and BancroW ° as a rapid assay of relative prolactin and growth hormone mRNA levels in multiple samples of GH3 cells and normal pituitary tissue. It should be noted that since the original publication of this procedure, a large number of reports have appeared in which cytoplasmic dot hybridization was used to measure the relative levels of a wide range of RNA species in many different cell and tissue types. Thus, we would expect this method to be useful for investigating whether other genes are regulated by Ca 2÷ or Ca2+-binding proteins. We employ the following procedure to measure prolactin mRNA levels. 9 C. Bancroft, G. G. Gick, M. E. Johnson, and B. A. White, in "Biochemical Actions of Hormones" (G. Litwack, ed.), Vol. 12, p. 173. Academic Press, New York, 1985. i0 B. A. White and C. Bancroft, J. Biol. Chem. 257, 8569 (1982).

660

REGULATION OF C a AND Ca-BINDING PROTEINS

[49]

Reagents for Preparation of Cytoplasmic Lysate i. Phosphate-buffered saline (PBS), sterilized by filtering (e.g., Nalgene type SCN, 0.2 ~m) ii. 10 mM Tris-HCl, pH 7.4, 1.0 mM EDTA (TE buffer). Sterilize by filtering or autoclaving. TE buffer is prepared in baked glassware with diethylpyrocarbonate (DEP)-treated H20. DEP-treated H20 is prepared by adding 300/~1 DEP (Sigma) to 500 ml of double-distilled H20, shaking hard to mix, loosening cap to vent, incubating at 37° overnight, and finally autoclaving iii. 200 mM vanadyl-ribonucleoside complex (VRC) as a ribonuclease inhibitor. VRC can be purchased (e.g., BRL) or made as described. H We prepare 20-ml batches, and store in 500-t~l aliquots under nitrogen gas at - 8 0 °. VRC is not absolutely essential for all tissues, but is a worthwhile precaution iv. 5% Nonidet P-40 (NP-40): Purchase 100% NP40 from Sigma, and dilute in DEP-treated H20. Store in sterile plastic test tube at 4° v. 37% formaldehyde solution. We purchase this from Fisher (analytical grade), store at room temperature, and use for 6 months vi. 20× SSC (3 M NaCI, 0.3 M sodium citrate, pH 7.0). Pass through 0.2-/~m filter, and then treat with DEP as described above for H20. Store at room temperature in baked bottles

Procedure for Preparation of Cytoplasmic Lysate of 5 × 105 GH3 Cells Grown in a Culture Dish. Scrape cells and transfer into a sterile plastic test tube or a baked glass test tube. Pellet cells in a refrigerated centrifuge, decant supernatant, and resuspend cells in 1 ml of cold PBS. Transfer cells to a 1.5-ml conical microcentrifuge tube, spin 10 sec at 7000-12,000 g and aspirate supernatant (use a Pasteur pipet that has a narrowed tip by pulling over a flame). It is important to remove the PBS as completely as possible without disturbing the cell pellet. Fully resuspend the pellet in 90/zl of a TE-VRC solution (make fresh by adding 50 tzl of VRC to 950/zl TE). Add 10/.d 5% NP40, mix, and incubate for 5 min on ice. Mix once during this incubation. Centrifuge the sample for 2.5 min at 7000-12,000 g to pellet the nuclei, and transfer 100 ~1 of cytoplasmic supernatant to 100/zl of an SSC-formaldehyde solution (make fresh by adding 400/xl formaldehyde to 600 ttl of 20× SSC). Incubate in 60 ° water bath for 15 min, and store at - 2 0 °. In preparing the cytoplasmic lysate, it is not important that the volume be kept at 100 ~1, but it is important to lyse the cells or tissue. Thus, some investigators homogenize samples in 200-500/~1 of TE. Another variable is the amount of VRC used. Samples with higher endogenous ribonuclease activity may require more VRC. H S. L. Berger and C. S. Birkenmeier,

Biochemistry 18, 5143 (1979).

[49]

Ca2+/CaM REGULATION OF PROLACTIN GENE EXPRESSION

661

Dilution series of samples are easily made by vortexing samples and then pipetting varying amounts (I-20/zl) of sample into microtiter plate (e.g., Falcon No. 3911 Microtest III) wells containing 15x SSC to make the final volume 200/zl. If the sample was originally homogenized in a large volume, then more can be applied to the nitrocellulose. The amount of sample applied to the nitrocellulose is limited by the concentration of protein (i.e., more than about 5-8 x 105 GH3/0.4 mm well will block flow through the filter). The samples are applied to nitrocellulose (Schleicher and Schuell or Millipore) using a vacuum manifold (e.g., Schleicher and Schuell Minifold). The nitrocellulose is then baked for 1.5-2.0 hr at 80° in a vacuum oven, prehybridized, hybridized, and washed essentially as described, lz

Minigel Blot Analysis The use of RNA gel blot hybridization is required if the sizes or number of the particular RNA species being studied needs to be determined. By combining some of the features of cytoplasmic dot hybridization 1° and the microanalytical guanidine hydrochloride precipitation method of Cheley and Anderson, 13we have devloped a rapid and convenient method for RNA gel blot (Northern) hybridization using formaldehyde gels.

Reagents Reagents that are needed in addition to those listed above are the following: i. Agarose (ultrapure DNA/RNA grade) ii. 10× MOPS buffer: 200 mM MOPS (Sigma or Research Organics), pH 7, 50 mM sodium acetate, 7 mM EDTA. Treat with DEP as described above. There is no need for concern about the yellow color that develops after autoclaving iii. Deionized formamide. We purchase formamide from International Biotechnologies, Inc. (New Haven), and deionize it by mixing it with 20% volume of a mixed bed resin (Bio-Rad AG 501-X8) at room temperature for 15-20 min. Following removal of the resin by filtration through Whatman No. 1 paper, aliquots are stored at - 7 0 ° iv. Ethidium bromide solution. 2 mg/ml in DEP-treated water v. Sample loading buffer. 50% glycerol, 0.4% bromphenol blue in 1 × MOPS vi. 7.6 M guanidine-HC1 in 0.1 M potassium acetate buffer, pH 5 vii. 95% Ethanol, stored at - 2 0 ° in baked glassware 12 p. Thomas, Proc. Natl. Acad. Sci. U.S.A. 77, 5201 (1980). 13 S. Cheley and R. Anderson, Anal. Biochem. 137, 15 0984).

662

REGULATION OF Ca AND Ca-BINDING PROTEINS

[49]

Procedure for Minigel Blot Analysis. We prepare cytoplasmic RNA by extracting the 100/zl cytoplasmic sample (see procedure for cytoplasmic dot hybridization above) with 800/.d of 7.6 M guanidine-HCl solution and precipitating with 550/zl of 95% ethanol at - 2 0 ° overnight. The sample is then centrifuged in a microcentrifuge at 5000 g for 30 min and the supernatant decanted. The RNA pellet is dried by leaving the centrifuge tube inverted on a Kim-Wipe at room temperature for 15-20 min. The RNA pellet is then dissolved in 20-40 /.d of TE buffer plus 1 /zl of RNasin (Promega Biotech). Repipetting and brief heating (60° for 15-30 sec) help to completely dissolve the RNA pellet. About 2-3 ~1 of sample is removed into 1 ml of water for measuring the RNA concentration by optical density at 260 nm. Samples for gel electrophoresis are prepared by adding up to 4.7 ~1 of RNA sample to 10/zl deionized formamide, 2/~1 10 x MOPS buffer, and 3.3/xl of 37% formaldehyde. Samples are incubated at 60 ° for I0 min, and then 1 ~I of ethidium bromide solution and 3/zl of sample loading buffer are added. The samples are run at 50 V in a 1% agarose-formaldehyde gel (see below) submersed in 1 × MOPS buffer for about 2-3 hr. After photography of the ethidium bromide-stained ribosomal RNA bands, the RNA bands are then transferred to nitrocellulose overnight as described) 2 In addition to nitrocellulose, we have also obtained good results with nylon (Amersham HyBond). The 1% agarose-formaldehyde gel is formed by melting 200 mg of agarose in 14.6 ml of DEP-treated water plus 2 ml of 10× MOPS. After the agarose dissolved, it is placed in a 60° water bath for 10-15 min. A 3.2-ml volume of 37% formaldehyde is added, and 10-15 ml of the gel solution pipetted onto a 5 × 7.5 mm microscope slide or into a minigel apparatus. This last step should be performed in a fume hood. Quantitation of Results We routinely scan our films with a scanning densitometer that is interfaced with an integrator. Arbitrary scanning units can then be expressed per dish, per cell number, per amount of DNA, per amount of protein, or, in the case of RNA blots or gel blots, per OD260 units of total RNA. It is important to obtain a series of film exposures so that dots with different intensities will produce scanning signals within the linear response range of the film and/or the scanner. The use of multiple dilutions of each sample also ensures obtaining accurate relative values. Some investigators have cut out the dots and counted the radioactivity, and have obtained the same results as scanning) As routinely employed, cytoplasmic dot hybridization quantitates the relative levels of an mRNA in various samples. We have recently de-

[49]

Ca2+/CaM REGULATION OF PROLACTIN GENE EXPRESSION

663

scribed how an internal RNA, standard synthesized in vitro can be employed to measure absolute levels of an mRNA by this technique? 4

Measurement of Overall Protein or RNA Synthesis in Samples Prepared for Cytoplasmic Dot Hybridization In experiments in which cellular calcium homeostasis is disrupted, it is important to monitor parameters of general cellular health. We routinely measure total RNA or total protein synthesis in the same samples that are used to measure prolactin mRNA levels by cytoplasmic dot hybridization, by the following method: 1. Add 50-100/zl of [3H]uridine or [3H]leucine (1 mCi/ml stock) to the cell cultures 15-30 min before scraping up the cells. 2. Process cells for cytoplasmic dot hybridization as described above up to the point at which 100/zl of cytoplasm is to be transferred to the SSC-formaldehyde mix. Transfer only 50/zl to an equal volume of SSC-formaldehyde for denaturation, and use the remaining 50 /zl for TCA precipitation of labeled macromolecules. We also use this second 50-/zl aliquot for determination of total protein concentration. Detection of Calmodulin-Binding Proteins and Ca2+-Calmodulin Dependent Protein Kinase Activity in the Nuclear Matrix Fraction of GH3 Cell Nuclei The results from the calmodulin inhibitor experiments 5 indicate a probable involvement of calmodulin in the regulation of prolactin gene expression. Calmodulin and calmodulin-regulated proteins have been detected in nuclei of several cell types (reviewed in Ref. 1). Additionally, recent evidence has been obtained for regulated nuclear Ca 2÷ levels. ~5 These findings have led us to examine the possibility that calmodulin may regulate prolactin gene expression through a direct interaction with nuclear proteins. The following methods were recently utilized by White and Preston 16 for detection of [~25I]calmodulin-binding proteins and Ca 2÷calmodulin-dependent protein kinase activity in the nuclear matrix fraction of GH3 cells.

t4 B. A. White, T. Lufkin, G. M. Preston, and C. Bancroft, this series, Vol. 24, p. 269. 15 D. A. Williams, K. E. Fogarty, R. Y. Tsien, and F. S. Fay, Nature (London) 318, 558 (1985). 1~ B. A. White and G. M. Preston, submitted for publication.

664

REGULATIONOF Ca AND Ca-BINDINGPROTEINS

[49]

Detection of Nuclear Matrix Calmodulin-Binding Proteins by Protein Blotting Preparation of Nuclear Fractions. GH3 cells (5 × 108 cells) are pelleted, washed once in phosphate-buffered saline (PBS), and resuspended in 9 ml (final volume) of PBS. The suspended cells are divided into two 4.5-ml aliquots and lysed by addition of 500/~1 of 5% Nonidet P-40 and incubation on ice for 5 min. At 4 min, 50/zl of 5% deoxycholate is added for the final minute of cell lysis. The cell lysates are centrifuged for 5 min at 1000 g, and the centrifugates (crude nuclear pellets) washed twice in 5 ml of 10 mM HEPES, pH 7.5, 0.5 mM spermidine, 1 mM EDTA, 0.25 mM EGTA, 5 mM NaCI, 2 mM dithiothreitol, and 10% glycerol (NW buffer). Washed nuclei are resuspended in 5 ml of NW buffer containing 30% sucrose, and centrifuged 20 min at 10,000 g in a Sorvall HB-4 rotor. The nuclear pellets are resuspended in 1 ml of 10 mM Tris-HCl, pH 7.5, 0.2 mM EDTA, 0.075 M NaC1, and 0.1% sarkosyl by homogenization in a 2ml glass-glass homogenizer, and centrifuged 3 min at 7000 g in a Beckman Microfuge 12. The chromatin pellet is extracted by homogenization in 0.4 M Nacl, 5 mM sodium phosphate, pH 7.0, followed by centrifugation as above. This low salt-extracted chromatin pellet is resuspended in 2 ml of 0.1 M NaCI, 10 mM sodium phosphate, ph 7.2 (CaM buffer), and an equal volume of 4 M NaCI added. The sample is homogenized on ice in a Dounce type tissue grinder with 20-30 strokes using the " A " (tight fitting) pestle, followed by centrifugation at 16,000 g in a Sorvall HB-4 rotor at 4 ° for 30 min. This homogenization and centrifugation is performed three times. The final nuclear pellet matrix is resuspended in CaM buffer at a protein concentration of 1-2 mg/ml. All reagents employed in nuclei isolations and extractions contain 0.2 mM paramethylsulfonyl fluoride (PMSF) as a protease inhibitor. This nuclear matrix fraction contains no detectable level of the mitochondrial enzyme, cytochrome c oxidase, or of the microsomal enzyme, NADH cytochrome c reductase. 16 Protein Blotting and [lesI]Calmodulin Binding Nuclear matrix proteins are fractionated on SDS gels containing a 7 15% linear polyacrylamide gradient. Proteins are electophoretically transferred from the gel to nitrocellulose at 80 V for 2 hr, using the buffer system of Towbin et a l . 17 The blot "sandwich," going from cathode to anode, consisted of a Scotch-Brite pad, one sheet of Whatman 3MM paper presoaked in double-distilled water, one sheet of 3MM paper presoaked in 0.5% SDS, the gel, one sheet of nitrocellulose prewetted in 17 H. Towbin, T. Staehelin, and J. Gordon, Proc. Natl. Acad. Sci. U.S.A. 76, 4350 (1979).

[49]

CaE+/CaM REGULATION OF PROLACTIN GENE EXPRESSION

665

water, two sheets of 3MM paper presoaked in water, and a Scotch-Brite pad. TM The transfer is performed in a Bio-Rad Transblot apparatus immersed in an ice bath. Calmodulin is iodinated according to the procedure of Bolton and Hunter.19 A 500-/zCi sample of 125I-labeled N-succinimidyl3-(4-hydroxyphenol) proprionate (Bolton-Hunter reagent) is evaporated to dryness under N2 in a conical tip vial, and 30/~g of eel calmodulin in 100 /zl of 0.2 M borate, pH 8.0, is added. After a 60-min incubation at 4°, the [125I]calmodulin is separated from the 125I-labeled Bolton-Hunter reagent by Sephadex G-25 column chromatography and stored in 50-/.d aliquots at - 2 0 °. The specific activity of [~25I]calmodulin ranges from 50 to 120 Ci/ mmol. Protein blots are preincubated overnight at ambient temperature in 10-20 ml of CaM buffer containing 2% bovine serum albumin (BSA) in a seal-a-meal freezer bag. The preincubation buffer is then removed, and an equal volume of CaM buffer containing 5-10 × 106 cpm of [125I]calmodulin added. The blots are incubated overnight at ambient temperature with shaking, and then washed in three changes of CaM buffer containing 0.5% Nonidet P-40 for about 1 hr total washing time. The strips are blotted on Kim-Wipes, wrapped in plastic wrap, and exposed to XAR-5 (Kodak) film at - 8 0 ° with a Cronex intensifying screen. After blotting proteins in whole nuclei, 0.4 M NaCl-extracted nuclear fraction, 2 M NaCl-extracted nuclear fraction, and the 2 M NaCl-insoluble nuclear matrix fraction, and incubation of the blot with ~zSI-calmodulin, White and Preston ~6 observed that nonhistone calmodulin-binding proteins resided almost exclusively in the nuclear matrix fraction. Several nuclear matrix proteins bound [125I]calmodulin in a manner that was inhibited by EGTA or by nanomolar levels of unlabeled calmodulin. Residual histones in the nuclear matrix significantly bound calmodulin (CH) but in a calcium-independent manner. A major calmodulin-binding protein of Mr 56,000, termed NMP 56, was consistently detected. Calmodulin binding to NMP 56 is totally reversed by EGTA and significantly inhibited by nanomolar levels of unlabeled calmodulin. Thus, the use of protein blotting has afforded us the ability to rapidly screen different nuclear fractions and to detect several specific calmodulin-binding proteins in a specific fraction.

Detection of Ca2+-Calmodulin-Dependent Protein Kinase Activity in the Nuclear Matrix Fraction of GH3 Cells The nuclear matrix fraction is prepared as described above, except that 50 mM EGTA is present during the three high salt extractions. Four ~a D. Watkins and B. A. White, J. Biol. Chem. 260, 5161 (1985). 19 A. E. Bolton and W. M. Hunter, Biochem. J. 133, 529 (1973).

666

REGULATION OF C a AND Ca-BINDING PROTEINS

[49]

of these nuclear matrix preparations were assayed for cytochrome c oxidase and N A D H cytochrome c reductase and neither of these enzymes was detected. 16 Based on the limit of detection of these assays, contamination of the nuclear matrix fraction by these enzymes was less than 0.5% of total cellular activity. Nuclear matrix proteins (30/zg) are incubated for 5 rain in a total volume of 0.25 ml containing 40 mM HEPES, pH 7.0, 0. I mM dithiothreitol, 0.1 /xM ATP, 20 mM MgCI2, and 25 /.~Ci [32p]ATP (specific activity = 5 mCi//.~M) plus various test substances. Samples are then boiled in sample buffer, resolved on 7-15% linear acrylamide-SDS gels, stained in 50% methanol, 10% acetic acid, 0.125% Coomassie blue, destained in 7% acetic acid, 5% methanol, dried under vacuum onto 3MM paper, and exposed to XAR-5 film.16 Several nuclear matrix proteins were phosphorylated in a manner that was not altered by addition of either 0.1 mM CaC12 or 10/xM calmidazolium. Addition of both 0.1 mM CaCI2 and 2.4/zM calmodulin increased the extent of phosphorylation of a 56-kDa protein, which may be NMP 56. This Ca2+-calmodulin-dependent phosphorylation of the 56-kDa nuclear matrix protein is blocked by 10/zM calmidazolium or replacement of 0.1 mM CaC12 with 10 mM EGTA. Summary Despite the extensive literature on the biological actions of Ca 2+ and calmodulin, very little is known about their involvement in nuclear functions, e.g., regulation of specific gene expression. To date, the only genes other than prolactin and growth hormone shown to be regulated by perturbations in cell Ca 2+ are those coding for two glucose-regulated proreins. 2° However, there is a growing body of indirect evidence for nuclear functions of Ca 2+ and calmodulin (reviewed in Ref. 1), and we suspect that other examples of Ca2+-regulated genes will emerge. We have described in this chapter several different experimental approaches which we have employed to examine first whether prolactin gene expression is regulated by changes in cell Ca 2+ content, and then to begin searching for the components of the mechanism by which Ca 2+ exerts its effects on the prolactin gene. The tentative identification of 55kDa nuclear matrix protein as both a calmodulin-binding protein and a substrate of a CaE+-calmodulin-dependent protein kinase suggests that NMP 55 may be a subunit of a multifunctional CaE+-calmodulin-protein kinase. This enzyme was recently detected in the nuclear matrix fraction 20 E. Resendez, Jr., J. W. Attenello, A. Grafsky, C. S. Chang, and A. S. Lee, Mol. Cell. Biol. 5, 1212 (1985).

[50]

PROTEIN METHYLATION

667

of neuronal nuclei, and was shown to phosphorylate a chromatin protein similar to high mobility group protein 17 (HMG 17). 21 Since HMG 17 is associated with actively transcribed chromatin, 22 its phosphorylation in GH3 cells might play a role in the Ca2+-calmodulin-dependent regulation of prolactin gene expression by hormones and growth factors. Acknowledgments This research was supported by NIH Grant AM 32836 to B.A.W. and NSF Grant DCB 8316415 to C.B. 21 N. Sahyoun, H. Levine, III, and P. Cuatrecasas, Proc. Natl. Acad. Sci. U.S.A. 81, 4311 (1984). 22 S. Weisbrod, Nature (London) 297, 289 (1982).

[50] C a l m o d u l i n as a n A c t i v a t o r a n d a S u b s t r a t e o f Methyltransferase Enzymes l By FRANK L. SIEGEL, LYNDA S. WRIGHT, a n d PAUL M. ROWE

Posttranslational methylation of a wide variety of eukaryotic and prokaryotic proteins has been described; these modifications are either Nmethylation or carboxylmethylation reactions. N-Methylation usually occurs at lysine, arginine, or histidine residues, while carboxylmethylation esterifies glutamate residues in prokaryotic organisms and aspartate residues in eukaryotes. 2 In most cases the significance of methylation is not known, nor is the means by which methylation is regulated. A notable exception is the methylation of receptors involved in bacterial chemotaxis. 3 Calmodulin, myosin, histones, and cytochrome c are all methylated at basic amino acid residues. Calmodulin is N-methylated at lysine-115 in most species by a highly specific N-methyltransferase,4 Sadenosyl-L-methionine : calmodulin(lysine)N-methyltransferase,CLNMT (E.C. 2. I. 1.60), and while the significance of this modification is not fully understood, methylation selectively modulates the ability of calmodulin ZAbbreviations used: CLNMT, S-adenosyl-L-methionine: calmodulin(lysine) N-methyltransferase; DTT, diothiothreitol; HEPES, 4-(2-hydroxyethyl)-l-piperazineethane sulfonic acid; SDS, sodium dodecyl sulfate. 2 W. K. Paik and S. Kim, "Protein Methylation." Wiley, New York, 1980. 3 S. Clarke, Anna. Rev. Biochem. pp. 479-506 (1985). 4 A. Sitaramayya, L. S. Wright, and F. L. Siegel, J. Biol. Chem. 255, 8894 (1980).

METHODS IN ENZYMOLOGY, VOL. 139

Copyright © 1987by Academic Press, inc. All rights of reproduction in any form reserved.