Molecular cloning of a vitamin D-dependent calcium ... - Europe PMC

2 downloads 0 Views 1MB Size Report
scribed by Dagert and Ehrlich (27). Hybridized plasmid (6 'ug ..... Maniatis, T., Jeffrey, A. & Kleid, D. G. (1975) Proc. Natl. Acad. Sci. USA 72, 1184-1188. 30.
Proc. Natl Acad. Sci. USA Vol. 80, pp. 4228-4232, July 1983 Biochemistry

Molecular cloning of a vitamin D-dependent calcium-binding protein mRNA sequence from chick intestine (cDNA library/comparative colony hybridization/RNA dot blot/DNA dot blot/RNA gel analysis/hybrid-selected translation)

WILLI HUNZIKER*, PAUL D. SIEBERTt, MICHAEL W. KINGt, PETER STUCKI*, ACHILLES DUGAICZYKt, AND ANTHONY W. NORMANt tDepartment of Biochemistry, University of California, Riverside, California 92521; and *Department of Veterinary Pharmacology, University of Bern, CH3012 Bern, Switzerland Communicated by Robert H. Wasserman, April 1, 1983

ABSTRACT We have constructed a recombinant cDNA library to facilitate study of the genomic actions of vitamin D3 and its hormonally active metabolite 1,25-dihydroxyvitamin D3 in initiation of the de novo biosynthesis of a 28,000-dalton vitamin Ddependent calcium binding protein (CaBP) present in chick intestine. The recombinant plasmids were prepared by the homopolymeric tailing and hybridization method using as a starting template poly(A)-enriched mRNA obtained from the intestinal mucosa of vitamin D3-replete (+D) chicks. Screening of 9,516 clones in this library was effected by using a comparative in situ colony hybridization technique with two [32P]cDNA probes; these probes were prepared from total poly(A)-RNA from chick intestinal mucosa of vitamin D-deficient (-D) chicks and a poly(A)-RNA specifically enriched for chick intestinal CaBP mRNA by immunoprecipitation of polysomes derived from vitamin D-replete (+D) chicks. We identified 26 clones that consistently displayed a significantly increased hybridization signal when comparing the -D vs. CaBP-enriched probe. Further evaluation of these clones by hybrid-selected translation showed the presence of CaBP-specific sequences. By "RNA gel" analysis of poly(A)-RNA, three independent mRNA species were found to hybridize to a CaBP clone; none of these RNA species were found in -D poly(A)-RNA. With this comparative colony hybridization procedure, we were able to identify CaBP-specific clones corresponding to a mRNA that is 0.1% of the total poly(A)-mRNA. The differential colony hybridization procedure using an enriched vs. a nonenriched probe should be of value in screening for other cDNA clones complementary to rare mRNA species.

receptor complexes in intestinal cell nuclei (13-15) suggested that 1,25(OH)2D3 acts as a steroid hormone regulating the synthesis of specific proteins at the nuclear level. There are also several other proteins in addition to CaBP that have been shown to be stimulated by 1,25(OH)2D3 at the level of biosynthesis (16-20). In an attempt to identify and characterize proteins that are regulated at the nuclear level by 1,25(OH)2D3, we have constructed a recombinant cDNA library by using total poly(A)-RNA from vitamin D-replete (+D) chick intestine. EXPERIMENTAL PROCEDURES Materials. Guanidinium thiocyanate (purum) was purchased from Fluka. Guanidinium HCI and ethidium bromide were from Sigma. [32P]dATP and [32P]dCTP (2,000-3,000 Ci/mmol; 1 Ci = 37 GBq) were purchased from Amersham. Cesium chloride (technical grade) was obtained from Kawecki Berylco (Reading, PA). Agarose (SeaKem-ME) was from FMC (Marine Colloids Division, Rockland, ME). Oligo(dT)-cellulose (type II) was from Collaborative Research (Waltham, MA). Reverse transcriptase was from J. Beard (Life Sciences, St. Petersburg, FL). All other enzymes were from Bethesda Research Laboratories except for S1 nuclease, which was from Miles. Nitrocellulose BA-85 (0.45 ,im) was from Schleicher & Schuell. Animals. One-day-old White Leghorn cockerels were raised from hatch on a vitamin D-deficient diet (21) or a vitamin Ddeficient diet supplemented with vitamin D3 (D-replete diet). Preparation of Chick Intestinal RNA. The preparation and characteristics of both total and poly(A)-enriched intestinal RNA are described in more detail elsewhere (10). Briefly, total cellular RNA was prepared from pooled intestinal mucosa scrapings using the guanidinium thiocyanate procedure described by Chirgwin et aL (22). Poly(A)-containing RNA [poly(A)-RNA] was then prepared by two cycles of binding to oligo(dT)-cellulose (23). The CaBP-enriched poly(A)-RNA was prepared by immunoprecipitation of intestinal polysomes obtained from +D chicks by the general procedure of Shapiro and Young (24). Double-Stranded cDNA Synthesis and Construction of Recombinant Plasmids. Synthesis of double-stranded cDNA complementary to total poly(A)-RNA from + D chick intestine was accomplished essentially as described by Law et al. (25). The double-stranded cDNA was then treated with S1 nuclease to remove the 3' loop, chromatographed on Sephadex G-100, and ethanol precipitated. Recombinant plasmids were prepared by insertion of the double-stranded cDNA in the Pst I site of pBR322

The secosteroid hormone 1,25-dihydroxyvitamin D3 [1,25(OH)2D3] interacts with intracellular receptors in its target tissues stimulating expression of specific genes (1, 2) and hence the production of proteins that mediate at least a portion of the biological responses attributable to vitamin D3. One of the bestcharacterized proteins regulated by 1,25(OH)2D3 is the chick intestinal vitamin D-dependent calcium-binding protein (CaBP) described by Wasserman and Taylor (3). There is extensive evidence suggesting a nuclear regulation of the synthesis of this protein. First, actinomycin D and a-amanitin have been found to block induction of CaBP by 1,25(OH)2D3 (4, 5). The activity of polysomal CaBP mRNA in chick intestine (6, 7) as well as in chick kidney (7) has been shown to be induced by 1,25(OH)2D3. More recently, nuclear (8), cytoplasmic (9), and total cellular levels (10) of chick intestinal CaBP mRNA have been shown to be dependent on 1,25(OH)2D3. Also, 1,25(OH)2D3 has been shown to stimulate general intestinal RNA metabolism (11) and chromatin template activity (12). These observations and the fact that there is rapid and specific localization of 1,25(OH)2D3-

Abbreviations: 1,25(OH)2D3, 1,25-dihydroxyvitamin D3; CaBP, calcium-binding protein; -D, vitamin D deficient; +D, vitamin D3 replete or vitamin D3-replete la-hydroxyvitamin D3-treated; DBM, dia-

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

zobenzyloxymethyl.

4228

Biochemistry:

Hunziker et aL

by using the homopolymeric tailing and hybridization method (26). Transformation of Escherichia coli with Recombinant Plasmid DNA. E. coli K-12 RR1 was rendered competent for uptake of plasmid DNA by treatment at 40C with 0.1 M CaCl2 as described by Dagert and Ehrlich (27). Hybridized plasmid (6 'ug containing 1 jig of double-stranded cDNA) was used for transformation, yielding about 10,000 independent colonies that were resistant to tetracycline and sensitive to ampicillin. Preservation of recombinant clones prior to further analysis was accomplished by overnight growth of individual cultures of clones in 96-well culture dishes containing L broth followed by addition of glycerol to 50% and storage at -200C (28). Preparation of 32P-Labeled Probes. cDNA probes were prepared essentially as described (25) for first-strand cDNA synthesis except that [32P]dATP or [32P]dCTP was used and the concentrations of unlabeled nucleotide and poly(A)-RNA were reduced to 50 AM and 60 pug/ml, respectively. Typically, the specific activity of the cDNA probes was 1-2 x 106 cpm/Ag with 15-20% copy yields. 32P-Labeled nick-translated plasmid DNA probes were prepared essentially as described by Maniatis et al. (29), using [32P]dATP or [32P]dCTP. Typically, the specific activity of the nick-translated probes was 1-2 x 108 cpm/pug. Differential Colony Hybridization. The procedure we adopted for in situ colony hybridization was a modification of the method of Grunstein and Hogness (30), described fully by Thayer (31). The protocol we used was the same except that, instead of the buffer in the original protocol (90 mM sodium citrate), 50 mM Hepes was used. After hybridization, washing, and drying, the filter was exposed to a Kodak X-Omat AR-5 x-ray film at -700C for 12-96 hr with intensifying screens. Plasmid DNA Dot Blot Analysis. DNA dot blot analysis was carried out by modifying and combining the methods of Kafatos et al (32) and Thayer (31). In brief, either crude plasmid DNA prepared by the rapid boiling method (33) or purified plasmid DNA prepared by the cleared lysate method (34) was linearized by digestion with BamHI or EcoRI, denatured with 0.3 M NaOH, and neutralized, and 5-Ml samples (about 200 ng of plasmid DNA) were spotted directly to nitrocellulose filters that had been treated with 1 M ammonium acetate. After baking at 80°C for 2 hr at reduced pressure, the DNA dot blots were hybridized to [32P]cDNA probes, washed, dried, and exposed to x-ray films exactly as described above for in situ colony hybridization. Hybrid-Selected Translation. Fifty micrograms of plasmid isolated by the cleared lysate CsCl method (34) was linearized with EcoRI. After denaturation in 0.3 M NaOH and neutralization, the DNA was bound to a 2.4-cm-diameter nitrocellulose filter by slow filtration at 4°C. After baking, three 1. 1-cm-diameter circles were cut for use in hybridization. Alternatively, the DNA was bound to diazobenzyloxymethyl (DBM)-paper (35). In this case, the plasmid pellet was taken up in 20 pl of 100 mM citric acid/Na2HPO4, briefly warmed to fully dissolve it, and then added to 100 Ml of dimethyl sulfoxide. This DNA was allowed to bind to five 1.1-cm-diameter filters overnight at room temperature and then the filters were washed several times with H20, four times with 0.4 M NaOH (1 ml per circle for 10 min at 37°C each), and several additional times with H20. Fifty micrograms of + D poly(A)-RNA was hybridized to each set of filters in 100 Mu of 70% formamide/0.4% NaDodSO4/1 mM EDTA/20 mM Pipes, pH 6.8/1 M NaCl for 3 hr at 50°C, and then the filters were washed 10 times at 65°C with 0.30 M NaCl/ 0.03 M Na citrate, pH 7/0.5% NaDodSO4 and three times with 0.30 M NaCl/0.03 M Na citrate, pH 7, over a period of about 2 hr. For elution, 200 ,ul of 10 mM Tris-HCl/1 mM EDTA, pH

Proc. Natl. Acad. Sci. USA 80 (1983)

4229

7.6, containing 12.5 Mg of E. coli tRNA per ml was added to the filters. Then, the filters were put in a boiling water bath for 45 sec and quickly chilled, the elution step was repeated once more, and the pooled eluates were ethanol precipitated. For translation, 30-40 ul of lysate mixture [32 Ml of Amersham lysate plus 8 Mul of [3'S]methionine (500 mCi/mmol)] was added directly to the RNA pellet and incubated at 30°C for 60 min Immunoprecipitation and gel electrophoresis was done as described (10). RNA Dot Blot Analyses and RNA Gel Hybridization. Both RNA dot blot analyses and RNA gel hybridizations were carried out essentially as described by Thomas (36) using [32P]-labeled nick-translated DNA probes prepared from plasmid pCI-CaBP26. For RNA dot blots, various dilutions of total intestinal RNA in 3 M NaCl/0.3 M Na citrate, pH 7, were blotted directly to

nitrocellulose without denaturation. For RNA gel hybridizations, purified total intestinal poly(A)-RNA was denatured by heating at 70°C for 2 min in the presence of 2.2 M formaldehyde and then electrophoresed on 1.2% agarose gels containing 2.2 M formaldehyde (37). After electrophoresis, the RNA was transferred and bound to nitrocellulose, hybridized, washed, and exposed to x-ray film as described by Thomas (36). RESULTS Screening for Clones Containing CaBP cDNA Sequences. Initial attempts to screen for clones containing CaBP cDNA sequences by differential colony hybridization by using a vitamin D-deficient (-D) and a +D [3 P]cDNA probe failed because of the low abundance of CaBP mRNA sequences present in the poly(A)-RNA template used to prepare the + D probe (about 0.1% as estimated by in vitro translation). We therefore used a [32P]cDNA probe prepared from poly(A)-RNA enriched for CaBP mRNA by an immunochemical technique. This RNA was found to yield the same spectrum of in vitro-synthesized proteins except for CaBP, which was enriched 15- to 20-fold to a level of about 2%. By visually comparing the intensities of the hybridization signals generated by these two probes to 9,516 colonies, 512 colonies were found to give stronger hybridization signals with the CaBP-enriched probe. A subset of these 512 colonies containing the 36 colonies having the strongest differences in the hybridization signal between the two probes was then selected for further analysis. To validate this screening procedure, hybridization of [32p]cDNA probes to purified plasmid DNA instead of to in situ-disrupted bacterial colonies was carried out. For this experiment, plasmid DNA from one of the selected 36 colonies (designated pCI-CaBP-26) (plasmid chick intestine) was isolated by the cleared lysate method (34), denatured, and spotted onto triplicate nitrocellulose filters in a quadruplicate array (Fig. 1). To have both a positive and a negative hybridization control, plasmid DNA isolated from a colony that consistently indicated a strong colony hybridization signal regardless of the probe used (pCI-23) and a colony containing only the pBR322 vector plasmid DNA were similarly spotted. The filters were then hybridized with a - D, a + D, and a CaBP-enriched [32P]cDNA probe. The result of this experiment (Fig. 1) shows that plasmid DNA derived from clone pCI-CaBP-26 (dots labeled no. 1) contains a cDNA sequence that hybridizes more strongly to the +D cDNA probe compared with the -D cDNA probe. Furthermore, this plasmid DNA hybridized more strongly to the CaBP-enriched cDNA probe than either the +D or -D cDNA probes. In contrast, the positive control clone DNA (dots labeled no. 2) indicated a strong hybridization signal with all three of the cDNA probes whereas the negative control DNA

4230

Biochemistry: Hunziker et al

FIG. 1. Validation of the differential colony hybridization screening procedure by comparative plasmid DNA dot blot hybridization: Plasmid DNA was extracted and partially purified from suspect clone pCI-CaBP-26 as well as from a positive control clone (pCI-23) and a negative control plasmid (pBR322). The preparations were then linearized by usingBamHI digestion, denatured, and bound directly to a triplicate set of nitrocellulose filters in a quadruplicate array. The DNA blots were hybridized to one of three [32P]cDNA probes: (A) probe prepared from -D poly(A)-RNA, (B) probe prepared from +D la-OH-D3-treated poly(A)-RNA, (C) probe prepared from poly(A)-RNA enriched for CaBP mRNA. For each filter, there are three types of plasmid DNA species blotted in a nearly square quadruplicate pattern connected by lines: 1, plasmid DNA from pCI-CaBP-26; 2, plasmid DNA from pCI-23; 3, plasmid pBR322 DNA. To allow for a more quantitative evaluation of the data, densitometeric tracings (A575) were obtained for each ofthe quadruplicate blots. These data are presented in Table 1.

(dots labeled no. 3) indicated only a minute hybridization signal. To allow for a quantitative comparison, the hybridization spots were densitometrically scanned at 575 nm. The data were then corrected for nonspecific hybridization by subtracting the signal indicated by negative control (pBR322) plasmid DNA. In addition, to allow comparison among the three filters, the data were normalized to the hybridization signal indicated by the positive control (pCI-23) plasmid DNA. These results are shown in Table 1. The fact that plasmid DNA derived from pCI-CaBP26 hybridizes more strongly to both the + D and the CaBP-enriched cDNA probes suggests that the cDNA insert within this colony is complementary to a portion of the CaBP mRNA. After this validation of our screening procedure, the reTable 1. Semiquantitative densitometric analysis of the DNA dot blot hybridization Plasmid DNA A575 ratio* Probe A575 of spot 0.03 0.0 pBR322 D-deficient 3.0 ± 0.60 pCI-23 cDNA 0.1 ± 0.05 pCI-CaBP-26 0.19 0.0 pBR322 D-replete 2.7 ± 0.30 pCI-23 cDNA 0.5 ± 0.20 pCI-CaBP-26 1.22 0.03 pBR322 CaBP2.1 ± 0.20 enriched pCI-23 2.6 ± 0.40 pCI-CaBP-26 cDNA Each of the four types of blots on the autoradiogram shown in Fig. 1 was scanned at 575 nm. A575 = 0.0 represents the background absorbance level of each autoradiogram. Data represent mean ± SD of quadruplicate determinations. It can be calculated from theA575 ratios that there is a 40-fold difference between the corrected and normalized -D and CaBP-enriched values. * (pCI-CaBP-26 - pBR322)/(pCI-23 - pBR322).

Proc. Natl. Acad. Sci. USA 80 (1983)

maining 35 clones selected after the initial screening were reanalyzed by the comparative - D vs. CaBP-enriched colony hybridization procedure; this process yielded 26 colonies that repeatedly hybridized more strongly to the CaBP-enriched cDNA probe. These were designated pCI-CaBP-1 to -26. Hybrid Selected Translation Analysis of CaBP-Related DNA. Plasmid DNAs from pCI-CaBP-2 and 4 were bound to nitrocellulose and that from pCI-CaBP-1 was bound to DBM-paper and these were hybridized to + D total poly(A)-RNA, eluted, and translated. Preliminary experiments showed that, during elution of the RNA from nitrocellulose, part of the immobilized DNA is eluted as well, inhibiting subsequent translation of the mRNA. This was circumvented by boiling the filters prior to hybridization in 10 mM Tris HCI, pH 7.6/1 mM EDTA for 510 min or by purification of the eluted mRNA by an oligo(dT)cellulose batch procedure. The mRNA eluted from DBM-paper on the other hand translated well without further treatment but was about 50% as pure. Gel analysis (Fig. 2) of the translation products and of the immunoprecipitated translation products shows that plasmids from clones pCI-CaBP-1 and -2 contain a cDNA sequence that hybridizes and hence selects for the CaBP mRNA. Therefore, these inserts can be considered to contain part of the CaBP mRNA nucleotide sequence. To determine the sizes of the cDNA inserts, plasmid DNA derived from 13 CaBP-related clones were digested with Pst I and resolved on a 2% agarose gel. All of the cDNA inserts were found to be fully excisable. The sizes of the cDNA inserts ranged between 200 and 450 base pairs (data not shown). RNA Dot Blot and RNA Gel Analyses of Intestinal RNA. Various amounts of either - D or + D total intestinal RNA were bound to a nitrocellulose filter and hybridized to a nick-translated [32P]DNA probe prepared from clone pCI-CaBP-26 plasmid DNA (Fig. 3). The results show that the cDNA insert from clone pCI-CaBP-26 contains a sequence complementary to a vitamin D-induced RNA. By comparing the slopes of each of the curves, we estimate that there is at least a 13-fold induction of the CaBP mRNA by the vitamin D treatment. The results of a RNA gel analysis on which the poly(A)-mRNAs from - D and + D chick intestine were separated by agarose gel 2 3 4 5 E ? 8i 9 i0

67,00043,000-

CoBP12,500-

_

_

12

-

_

_

-67,000 -43,000 -COSP -12,500

FIG. 2. Polyacrylamide gel electrophoretic analysis of the in vitro translation products coded by hybrid-selected mRNA: RNA was hybridselected, in vitro translated, and analyzed on an 8.5-20% NaDodSO4 polyacrylamide gel. The translation products were divided in half for direct electrophoresis and immunoprecipitation, except for the pCI-CaBP1-selected products which were immunoprecipitated but no total products were analyzed. Lanes: 2-5, total translation products encoded by zero exogenous RNA (lane 2), +D poly(A)-mRNA (lane 3), and mRNA selected from +D poly(A)-mRNA by the plasmid from clone pCI-CaBP2 (lane 4) or pCI-CaBP-4 (lane 5); 7-10, immunoprecipitated translation products encoded by mRNA selected by clones pCI-CaBP-1 (lane 7), pCI-CaBP-2 (lane 8), andpCI-CaBP-4 (lane 9), aswell asby +D mRNA (lane 10). Lanes 1, 6, and 11 show the migration of '25I-labeled CaBP together with the unlabeled standards bovine serum albumin, ovalbumin, and cytochrome c. Lane 12 is 125I-labeled heavy and light chains of IgG.

Biochemistry: Hunziker et al.

Proc. Natl. Acad. Sci. USA 80 (1983)

CaBP-26 and are vitamin D-dependent; these include a major hybridizing sequence approximately 2,000 nucleotides long and two other species, one of approximately 2,700 and one of 3,100 nucleotides. The slight amount of low molecular weight hybridizing RNA is believed to be an artifact because it migrated with one of the major RNA bands.

8