Calcium channel compositions and methods

Abstract

Calcium channel .gamma.-subunit-encoding cDNAs, and related compositions and methods, are provided.

Description

TECHNICAL FIELD

BACKGROUND OF THE INVENTION

Calcium channels are membrane-spanning, multisubunit proteins that allow controlled entry of Ca.sup.+2 ions into cells from the extracellular fluid. All cells throughout the animal kingdom, and at least some bacterial, fungal and plant cells, possess one or more types of calcium channel.

The most common type of calcium channel is voltage-dependent. In a voltage-dependent channel, the "opening" which allows there to begin an influx of Ca.sup.+2 ions into the cells requires a depolarization to a certain level of the potential difference between the inside of the cell bearing the channel and the extracellular medium bathing the cell. The rate of influx of Ca.sup.+2 into the cell depends on this potential difference. All "excitable" cells in animals, such as neurons of the central nervous systems, peripheral nerve cells, and muscle cells, including those of skeletal muscles, cardiac muscles, and venous and arterial smooth muscles, have voltage-dependent calcium channels.

Calcium channels are physiologically important because the channels have a central role in regulating intracellular Ca.sup.+2 levels, which levels are important for cell viability and function. Thus, intracellular Ca.sup.+2 concentrations are implicated in a number of vital processes in animals, such as neurotransmitter release, muscle contraction, pacemaker activity, and secretion of hormones and other substances.

A number of compounds useful in treating various diseases in animals, including humans, are thought to exert their beneficial effects by modulating functions of voltage-dependent calcium channels. Many of these compounds bind to calcium channels and block, or reduce the rate of, influx of Ca.sup.+2 into cells in response to depolarization of the inside and outside of the cells.

An understanding of the pharmacology of compounds that interact with calcium channels, and the ability to rationally design compounds that will interact with calcium channels to have desired therapeutic effects, have been hampered by a lack of understanding of the structure of channel subunits and the genes that code for them. Thus, it has not been possible to obtain the large amounts of highly purified channel subunits that are required to understand, at the molecular level, the nature of the subunits and their interactions with one another, with the cell membranes across which the channels allow Ca.sup.+2 ions to pass, with Ca.sup.+2 and other ions, and with low molecular weight compounds that affect channel function. For example, with the availability of large amounts of purified calcium channel subunits, functional channels could be prepared and used to screen the effects of compounds on channel function, thereby providing a basis for the design of therapeutic agents which affect the calcium channel, or various combinations of channel subunits could be crystallized and have their structures determined to high resolution employing X-ray or neutron diffraction techniques, providing yet another basis for rational design of therapeutic agents that affect channel function.

Certain diseases, such as Lambert-Eaton Syndrome, involve autoimmune interactions with calcium channels. The ready availability of each of the calcium channel subunits would make possible immunoassays for the diagnosis of such diseases and an understanding of them at the molecular level that could lead to effective methods of treatment.

The lack of information on genes that code for calcium channel subunits has prevented the understanding of the molecular properties of the mature calcium channel subunits and their precursor proteins (i.e., the mature subunits with signal peptides appended to the amino-terminus) and the regulation of expression of calcium channel subunits. An understanding of these properties, and of how expression of calcium channel subunit genes is regulated, may provide the basis for designing therapeutic agents which have beneficial effects through affecting calcium channel function or concentration. Furthermore, the availability of sequences of genes coding for calcium channel subunits would make possible the diagnosis of defects in genes coding for such subunits, which might underlie a number of diseases.

The availability of a DNA with the sequence of a segment of at least about 14 and more preferably at least about 30 nucleotides of a cDNA encoding a subunit of a calcium channel from the cells of a tissue of an animal would make possible the isolation and cloning of cDNAs, and possibly genomic DNAs, coding for the corresponding subunit of different calcium channels from the same or different tissues and animals of the same or different species. The availability of the sequences of numerous full-length cDNAs coding for corresponding subunits of calcium channels from a variety of tissues and animal species would contribute to elucidating structure-function relationships in the subunits. This knowledge, in turn, would be useful in the design of therapeutic agents whose activities are exerted through binding to calcium channels.

In skeletal muscle, where voltage-dependent calcium channels have been best characterized, voltage-dependent calcium channels are thought to consist of two large subunits, of between about 130 and about 200 kilodaltons ("kD") in molecular weight, and a number (generally thought to be two or three) of different smaller subunits, of less than about 60 kD in molecular weight. At least one of the larger subunits and possibly some of the smaller are glycosylated. Some of the subunits are capable of being phosphorylated. There has been confusion in the art concerning the naming of the various subunits of voltage-dependent calcium channels.

The two large subunits of voltage-dependent calcium channels are designated herein the ".alpha..sub.1 -subunit" and the ".alpha..sub.2 -subunit".

The .alpha..sub.1 -subunit is not detectably changed in molecular weight when treated with dithiothreitol ("DTT") or with enzymes which catalyze removal of N-linked sugar groups from glycosylated proteins. The .alpha..sub.1 -subunit has a molecular weight of about 150 to about 170 kD when analyzed by sodium dodecylsulfate ("SDS")-polyacrylamide gel electrophoresis ("PAGE") after isolation from mammalian muscle tissue and has specific binding sites for various 1,4-dihydropyridines ("DHPs") and phenylalkylamines.

The .alpha..sub.2 -subunit is somewhat less well characterized than the .alpha..sub.1 -subunit, although recent work by Ellis et al. (see PCT Application No. WO 89/09834 and Science, 241, 1661-1664 (1988)), has provided a great deal of additional information concerning this subunit. The molecular weight of the .alpha..sub.2 -subunit is at least about 130-150 kD, as determined by SDS-PAGE analysis in the presence of DTT after isolation from mammalian muscle tissue. However, in SDS-PAGE under non-reducing conditions (in the presence of N-ethylmaleimide), the .alpha..sub.2 -subunit migrates with a band of about 160-190 kD. It is presently believed that the smaller fragment (of about 30 kD), which appears to be released upon reduction of the .alpha..sub.2 -subunit, is the carboxy terminus of the primary translation product of the .alpha..sub.2 -subunit mRNA. Regardless, however, of whether the two fragments are different subunits of the calcium channel or whether both fragments are products of the same gene (and, consequently, the .alpha..sub.2 -subunit is about 160-190 kD and is split into (at least) two fragments upon reduction), there is evidence that the .alpha..sub.2 -subunit, whatever its size, and the corresponding fragment produced under reducing conditions, whether part of the .alpha..sub.2 -subunit or not, are glycosylated with at least N-linked sugars. In addition there is evidence that the .alpha.2-subunit and the corresponding fragment produced under reducing conditions do not have specified binding sites for 1,4-dihydropyridines and phenylalkylamines, which species are known to bind to the .alpha..sub.1 -subunit.

The .beta.-subunit of the calcium channel has only recently been characterized as having an apparent molecular mass of 52-65 kD (as determined by SDS-PAGE analysis). It is comprised of consensus phosphorylation sites and has been shown by biochemical methods to be phosphorylated. This subunit is insensitive to reducing conditions.

The .gamma.-subunit of the calcium channel has not been observed in all purified preparations, depending on the source of material analyzed, the investigating laboratory, and so on. Because of its irregular appearance in the hands of some investigators, this particular subunit also remains relatively poorly characterized. The native material appears to be a glycoprotein with an apparent molecular mass of 30-33 kD, as determined by SDS-PAGE analysis. The native protein is believed to be glycosylated since its apparent molecular mass decreases after digestion with neuraminidase followed by endoglycosidase F.

Reference herein to the precursor of an .alpha..sub.1 -subunit means the protein with the amino acid sequence corresponding to the sequence of the full-length mRNA which, upon translation, results, ultimately, in .alpha..sub.1 -subunit resident as part of a calcium channel in a cell membrane. The precursor protein is converted by various, not presently well understood, processing steps to produce the .alpha..sub.1 -subunit.

Similarly, reference herein to the precursor of an .alpha..sub.2 -subunit means the protein with the amino acid sequence corresponding to the sequence of the full-length mRNA which, upon translation, results, ultimately, in .alpha..sub.2 -subunit resident as part of a calcium channel in a cell membrane. The precursor protein is converted by various processing steps into the .alpha..sub.2 -subunit. As with the .alpha..sub.1 -subunit, the details of the processing between the precursor and the mature .alpha..sub.2 -subunit are not clear, but the processing presumably involves at least removal of a leader sequence (i.e., a signal peptide), glycosylation, and cleavage to yield what is now thought to be .delta.-subunit of the calcium channel.

Similarly, reference herein to the precursor of a .beta.-subunit means the protein with the amino acid sequence corresponding to the sequence of the full-length mRNA which, upon translation, results ultimately, in .beta.-subunit resident as part of a calcium channel in a cell membrane. The precursor protein is converted by various, not presently well understood, processing steps into the .beta.-subunit.

Similarly, reference herein to the precursor of a .gamma.-subunit means the protein with the amino acid sequence corresponding to the sequence of the full-length mRNA which, upon translation, results ultimately, in .gamma.-subunit resident as part of a calcium channel in a cell membrane. The precursor protein is converted by various, not presently well understood, processing steps into the .gamma.-subunit.

The cDNA and corresponding amino acid sequence of the .alpha..sub.1 -subunit precursor of a rabbit back skeletal muscle calcium channel has been reported. See Tanabe et al., Nature 328, 313-318 (1987).

The cDNA and corresponding amino acid sequences of the .alpha..sub.2 -subunit precursor of a rabbit back skeletal muscle calcium channel and a human neuronal calcium channel have also been reported. See Ellis, et al., PCT Application No. WO 89/09834 (1989) and Science 241, 1661-1664 (1988).

The cDNA and corresponding amino acid sequence of the .beta.-subunit precursor of a rabbit back skeletal muscle calcium channel has also been reported. See Ruth et al., Science 245, 1115-1118 (1989).

Up to now, however, the cDNA and corresponding amino acid sequence of the .gamma.-subunit precursor of a calcium channel have not been reported in the literature.

Calcium channel activity, measured electrophysiologically by voltage-clamp techniques, has been induced in Xenopus laevis oocytes when total mRNA isolated from mammalian brain or cardiac muscle is injected into the oocytes. Also, it has been reported that calcium channel-containing preparations, when reconstituted into lipid bilayers, confer voltage-dependent calcium channel activity on the bilayers.

However, there is no evidence that any one of the calcium channel subunits alone provides a natively functional calcium channel in oocytes, lipid bilayers or any other situation. It has been recently reported by Hofmann, et al., Trends in Pharmacology. Sci. 8, 393-398 (1987) that mRNA prepared using the cDNA of .alpha..sub.1 -subunit obtained by Tanabe, et al. was unable to induce calcium channel activity in Xenopus laevis oocytes.

SUMMARY OF THE INVENTION

In accordance with the present invention, there is provided the cDNA and corresponding amino acid sequence of the .gamma.-subunit precursor of a calcium channel.

There are now available well-characterized cDNA clones encoding each of the four calcium channel subunits, thus enabling one to incorporate various combinations of such cDNAs into recombinant expression systems. In this way, one can assess the contribution of the various calcium channel subunits to the transport and regulation of calcium ions.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 provides the cDNA nucleotide sequence and the amino acid sequence for the .gamma.-subunit of a calcium channel.

FIG. 2 is a restriction map of a cDNA sequence which encodes the .beta.-subunit of a calcium channel, and shows the orientation of various cDNAs encoding the .beta.-subunit sequence.

FIG. 3 is a restriction map of a cDNA sequence which encodes the .gamma.-subunit of a calcium channel, and shows the orientation of various cDNAs encoding the .gamma.-subunit sequence.

FIG. 4 provides the amino acid sequence for, and a nucleotide sequence encoding the .beta.-subunit of a calcium channel.

FIG. 5 provides raw data from which the amino acid sequence for the first nine residues at the N-terminus of the .gamma.-subunit of a calcium channel was ascertained.

FIG. 6 is a restriction map of a cDNA sequence which encodes the .alpha..sub.1 -subunit of a calcium channel, and shows the orientation of various cDNAs encoding the .alpha..sub.1 -subunit sequence.

FIG. 7 is a restriction map of a cDNA sequence which encodes the .alpha..sub.2 -subunit of a calcium channel, and shows the orientation of various cDNAs encoding the .alpha..sub.2 -subunit sequence.

FIG. 8 provides the amino acid sequence for, and a nucleotide sequence encoding the .alpha..sub.1 -subunit of a calcium channel.

FIG. 9 provides the amino acid sequence for, and a nucleotide sequence encoding the .alpha..sub.2 -subunit of a calcium channel.

DETAILED DESCRIPTION OF THE INVENTION

In accordance with the present invention, we have discovered a cDNA which codes for the .gamma.-subunit of an animal calcium channel (see FIG. 1).

Thus in one of its aspects, the invention is a DNA fragment which comprises a cDNA which codes for the .gamma.-subunit of an animal calcium channel, and the RNA, coding for such a subunit, made upon transcription of such a DNA according to the invention.

In another of its aspects, the invention is a substantially pure .gamma.-subunit of an animal calcium channel.

By a "substantially pure" subunit or protein is meant a subunit or protein that is sufficiently free of other polypeptide contaminants to be considered homogeneous by SDS-PAGE or to be unambiguously sequenced.

In another of its aspects, the invention entails an eukaryotic cell with an heterologous calcium channel, said cell made by a process comprising administering to said cell a first composition, which consists essentially of a first RNA or cDNA which is translatable in said cell into the precursor of the .gamma.-subunit of a calcium channel of an animal of a first species, and at least one of a second composition selected from the group consisting essentially of a second RNA or cDNA which is translatable in said cell into the precursor of the .alpha..sub.1 -subunit of a calcium channel of an animal of a second species, a third RNA or cDNA which is translatable in said cell into the precursor of the .alpha..sub.2 -subunit of a calcium channel of an animal of a third species, and a fourth RNA or cDNA which is translatable in said cell into the precursor of the .beta.-subunit of a calcium channel of an animal of a fourth species, wherein said first, second, third and fourth species can be the same or different. Preferred cells for this purpose are Xenopus laevis oocytes.

In yet another of its aspects, the present invention entails a method for preparing an heterologous calcium channel by expressing the RNA or cDNA administered to the above-described eukaryotic cells.

In another of its aspects, the invention entails a method for assaying a compound for calcium channel agonist or antagonist activity which comprises electrophysiologically measuring the calcium channel activity of an RNA- or cDNA-containing eukaryotic cell prepared as described above when such cell is exposed to a solution of the compound being tested for such activity. For similar methods applied with Xenopus laevis oocytes and acetylcholine receptors, see e.g., Mishina et al. Nature 313, 364 (1985) and, with such oocytes and sodium channels, see Noda et al., Nature 322, 826-828 (1986).

In a further of its aspects, the invention is an eukaryotic cell containing a DNA which comprises a cDNA which can be expressed to make the .gamma.-subunit of a calcium channel. Such a cell according to the invention can also contain at least one of the following additional DNA fragments selected from the group consisting essentially of:

a second cDNA which codes for the precursor of the .alpha..sub.1 -subunit of a calcium channel of an animal of a second species, a third cDNA which codes for the precursor of the .alpha..sub.2 -subunit of a calcium channel of an animal of a third species, and a fourth cDNA which codes for the precursor of the .beta.-subunit of a calcium channel of an animal of a fourth species, wherein said first, second, third and fourth species can be the same or different. Preferably, the .alpha..sub.1 -subunit, the .alpha..sub.2 -subunit, the .beta.-subunit or the .gamma.-subunit made from such cDNA in such a cell will be foreign to the cell, i.e., will have an amino acid sequence which differs from that of any calcium channel .alpha..sub.1 -subunit, .alpha..sub.2 -subunit, .beta.-subunit or .gamma.-subunit which occurs in a cell of the same type which does not contain a DNA from which the .alpha..sub.1 -subunit, the .alpha..sub.2 -subunit, the .beta.-subunit or the .gamma.-subunit encoded by such a cDNA is expressed. Preferred among such cells are those of mammalian origin, such as COS cells, NIH3T3 cells, mouse L cells or the like, or those of yeast such as S. cerevisiae, P. pastoris, or C. tropicalis. Methods of making such cells of the invention, i.e., by transforming cells with suitable heterologous DNAs, to be maintained in the cell as episomes or integrated into chromosomal DNA of the cell, and then culturing transformants or subculturing (or passaging, in the case of mammalian cells) from such a culture or a subculture thereof, are well known to those of ordinary skill in the art.

In yet another aspect of the present invention, there is provided a method for the production of the .gamma.-subunit of a calcium channel, comprising expressing the cDNAs contained in the eukaryotic cells as described above.

Among such cDNA-containing cells of the invention, the invention entails also an eukaryotic cell with an heterologous calcium channel, said calcium channel made by a process comprising expression of a first cDNA, which codes for the precursor of the .gamma.-subunit of a calcium channel of an animal of a first species, and at least one of the following additional DNA fragments selected from the group consisting essentially of:

a second cDNA which codes for the precursor of the .alpha..sub.1 -subunit of a calcium channel of an animal of a second species, a third cDNA which codes for the precursor of the .alpha..sub.2 -subunit of a calcium channel of an animal of a third species, and a fourth cDNA which codes for the precursor of the .beta.-subunit of a calcium channel of an animal of a fourth species, wherein said first, second, third and fourth species can be the same or different. Usually at least one of said precursors of said .alpha..sub.1 -subunit, .alpha..sub.2 -subunit, .beta.-subunit and .gamma.-subunit is foreign to said cell. Again, preferred among such cells are those of mammalian origin or those of yeast such as S. cerevisiae, P. pastoris or C. tropicalis. In a preferred embodiment, such a cell will also contain another heterologous gene, which comprises a transcriptional control element (e.g., a promoter or promoter/enhancer combination), which is active in said cell and the transcriptional activity of which responds, either directly or indirectly, to an ion or molecule capable of entering said cell through a functional calcium channel (e.g., Ca.sup.++, Ba.sup.++, Ca.sup.++ ionophores), linked operatively for expression to a structural gene for an indicator protein, such as chloramphenicol acetyltransferase, luciferase or .beta.-galactosidase.

In a further aspect of the present invention, there is provided a method to identify compounds which are agonists or antagonists of mammalian calcium channels whereby an eukaryotic cell with an heterologous calcium channel, prepared as described above, is contacted with the compound to be tested, and the effect of the compound to be tested on the calcium concentration in the cell is then measured, either directly or indirectly.

These cells of the invention, which have functional, foreign calcium channels (i.e., functional calcium channels wherein at least one of the calcium channel subunits is not native to the cell) will be useful for, among other purposes, assaying a compound for calcium channel agonist or antagonist activity. First, such a cell can be employed to measure the affinity of such a compound for the functional calcium channel. Secondly, such a cell can be employed to measure electrophysiologically the calcium channel activity in the presence of the compound being tested as well as an ion or molecule, such as Ca.sup.++ or Ba.sup.++, which is known to be capable of entering the cell through the functional channel. For similar studies which have been carried out with the acetylcholine receptor, see Claudio et al. Science 238 1688-1694 (1987). These methods for assaying a compound for calcium channel agonist or antagonist activity are also part of the present invention.

In a still further embodiment of the present invention, there is provided a method for the production of functional, foreign calcium channels by expressing the cDNA-containing cells as described above.

Such cells according to the invention, in the preferred embodiment, wherein the cell also contains an heterologous gene with a transcriptional control element, which is active in the cell and responsive, either directly or indirectly, to an ion or molecule capable of entering the cell through a functional calcium channel and is linked operatively for expression to a structural gene for an indicator protein, can also be employed, in another method according to the invention for assaying a compound for calcium channel agonist or antagonist activity. This method comprises exposing a culture of such cells to a solution of a compound being tested for such activity, together with an ion or molecule which is capable of entering the cells through a functional calcium channel and affecting, either directly or indirectly, the activity of the transcriptional control element controlling transcription of the gene for the indicator protein, and comparing the level of expression, in the cells of the culture, of the gene for the indicator protein with the level of such expression in the cells of another, control culture of such cells.

A "control culture," as clearly understood by those of skill in the art, will be a culture that is, and is treated, substantially the same as the culture exposed to the compound being assayed, except that the control culture is not exposed to the compound being assayed. Alternatively, a control culture will be a culture that is exposed to the compound being assayed, but has not been transformed with calcium channel subunit encoding sequences, but is otherwise treated substantially the same as the test culture. Levels of expression of the genes for the indicator proteins are ascertained readily by the skilled by known methods, which involve measurements of the concentration of indicator protein via assays for detectable compounds produced in reactions catalyzed by the indicator protein.

As indicated above, indicator proteins are enzymes which are active in the cells of the invention and catalyze production of readily detectable compounds, e.g., chromogens, fluorescent compounds, radioactively labeled compounds, and the like.

The invention entails also a labeled (e.g., .sup.32 P or a biotinylated) RNA or single-stranded DNA of at least 14 (preferably at least 30) bases in length in a sequence which comprises a sequence of at least 14 (preferably at least 30) contiguous bases between bases 1 and 1171, inclusive, in FIG. 1, which encodes rabbit skeletal muscle .gamma.-subunit. The use of such DNAs and RNAs as probes, to identify and isolate cDNAs coding calcium channel .gamma.-subunits or to identify tissue in which .gamma.-subunit mRNA is made, is clear to those skilled in the art. In this regard, see, e.g., Example 4.

The primary strategy for cloning cDNAs encoding the .gamma. polypeptide subunit of the DHP-sensitive calcium channel from rabbit skeletal muscle was to screen rabbit back skeletal muscle cDNA expression libraries with an antibody probe specific to the protein. See generally Ausubel et al., Current Protocols in Molecular Biology, Wiley-Interscience, New York (1987); Davis et al., Basic Methods in Molecular Biology, Elsevier Science Publishing Co., New York (1986) .

Polyclonal antisera for the .gamma. polypeptide subunit was prepared in guinea pigs as described by Sharp and Campbell, J. Biol. Chem. 264, 2816-2825 (1989). The .gamma.-specific polyclonal antisera were used for screening of approximately 1.0.times.10.sup.6 plaques of recombinant phages of a random-primed lambda gt11 cDNA library.

In an initial screen with the guinea pig .gamma.-specific polyclonal antisera, two cDNA clones were isolated and identified as J6 and J10. These clones were shown to be related to each other by partial sequencing and restriction digestion analysis. These cDNAs were used to screen an oligo dT-primed lambda gt11 cDNA library to isolate overlapping cDNA clones. Three clones were isolated from this screen, and were identified as G2, G10 and G11. One of these clones (G10) was partially sequenced. The insert from clone G10 was then oriented with respect to the inserts from clones J6 and J10. The cDNA insert from clone J10 was then used to screen pooled recombinants from a rabbit skeletal muscle cDNA library [MacLennan et al, Nature 316., 696 (1985)] constructed according to Okayama and Berg [Mol. Cell Biol. 3, 280-289 (1983)]. Of the positive clones identified from this screening, clone G4-3 was found to encompass and extend the cDNA inserts in clones J6, J10 and G10.

The cDNA clones were then analyzed to establish the coding DNA sequence of the .gamma.-subunit of the calcium channel. Approximately 1171 nucleotides of .gamma.-subunit cDNA was cloned, which is consistent with an estimated 1200 nucleotide .gamma.-subunit mRNA.

FIG. 1 represents the 1,171 nucleotides of the cDNA sequence encoding the .gamma.-subunit and its precursor, including 48 nucleotides of 5' untranslated sequence, a 666 nucleotide open reading frame, and 457 nucleotides of 3' untranslated sequence.

The 1171-nucleotide cDNA sequence set forth in FIG. 1 contains a 666-nt open reading frame coding for 222 amino acids. The deduced amino acid sequence yields a calculated molecular weight of 25,058 Dalton, which is in approximate agreement to the observed molecular mass of 32 kD for the glycosylated (1-3) and 20 kD for the chemically de-glycosylated forms of the .gamma.-subunit, as determined by SDS-polyacrylamide gel electrophoresis. The deduced amino acid sequence is also in agreement with the authentic NH.sub.2 -terminus of the .gamma.-subunit as determined by protein sequence analysis on the purified skeletal muscle protein (see Example 5).

The present invention makes available well-characterized cDNA clones encoding each of the four subunits of the rabbit skeletal muscle DHP-sensitive Ca .sup.2+ channel. It is, therefore, now possible to incorporate various combinations of these cDNAs into recombinant DNA expression systems to make functional calcium channels and to assess the relative contribution of each subunit to the formation of functional Ca.sup.2+ channels.

The present invention will now be described in greater detail by reference to the following non-limiting examples.

EXAMPLE 1

Preparation of .beta. and .gamma. Antibody Probes

The DHP receptor was extracted from heavy microsomes and triads of rabbit skeletal muscle, and then purified, following the procedures described in Sharp, et al. (1987), J. Biol. Chem. 262, 12309-12315. Individual components of the receptor were separated by SDS-PAGE on a 5-16% acrylamide gradient gel in the absence (nonreducing condition) or presence (reducing condition) of 1% 2-mercaptoethanol in the sample buffer. The gels were stained for 5-10 min with Coomassie Blue stain in 10% acetic acid, 25% isopropanol, and then destained in distilled water. Individual bands were visualized. The complex contains at least four subunits: .alpha.1 (170,000 Da), .alpha.2 (175,000 Da non-reduced/150,000 Da reduced), .beta. (52,000 Da) and .gamma. (32,000 Da), which appear to be present roughly in a 1:1:1:1 stoichiometric ratio. The purified complex or individual subunits present in sliced gel bands were then used to immunize experimental animals to induce the production of polyclonal antibodies as described below.

A. .beta.-enriched Antiserum

One sheep, identified as Anti-DHPR#1, was immunized subcutaneously with 500 .mu.g of the purified DHPR protein (not gel separated) in Freund's Complete Adjuvant. Eight weeks post immunization, a subcutaneous boost of 500 .mu.g purified DHPR protein in Freund's Complete Adjuvant was given. A serum sample was collected one week later. A second boost, identical to the first, was given one month after the first bleed, and a second bleed was performed one week after the second boost.

Aliquots of the two bleeds, and aliquots of a pre-immunized serum sample, were characterized for titer and reactivity to the DHPR subunit proteins. Two to five .mu.g of DHPR subunit proteins purified as described above were applied to nitrocellulose, and reacted with 10-fold serial dilutions of the three serum samples. Primary antibody was detected using a peroxidase-labeled secondary antibody and 1-chloro-4-naphthol as a development substrate. The aliquot of serum from the second bleed demonstrated the highest titer of .beta.-specific antisera (>1:1000), and was used in the affinity purification procedure.

B. .gamma.-specific Antiserum

The .gamma. polyclonal serum was prepared in guinea pigs as previously described [Sharp and Campbell, (1989), J. Biol. Chem. 264, 2816-2825]. Briefly, guinea pig #16 was immunized at the intervals described in Sharp and Campbell ibid. with a single, homogenized 5 mm.times.15 mm gel band corresponding to the 32 KD .gamma. subunit (.gamma. content of the gel band was approximately 2 .mu.g). The collected ascites was characterized by Immunoblot analysis as described in (A), above. Samples with titers >1:1000 for the .gamma. subunit were used for further analysis.

C. Affinity Purification

Antibodies raised against the .beta. and .gamma. subunit proteins were affinity purified from serum of injected sheep and guinea pigs, respectively, following the protocol of Sharp and Campbell (1989), J. Biol. Chem. 264, 2816-2825, using Immobilon strips. In brief, 200 .mu.g of purified rabbit dihydropyridine receptor were separated by preparative SDS-PAGE and electrophoretically transferred to Immobilon-P membranes (Millipore; resolving gel size=12.5 cm.times.13 cm). Two vertical strips were cut from the edge of the membrane and stained with the reversible stain Pounceau S to visualize the immobilized protein. The membranes were then separately reacted with 10 ml of a 1:1000 dilution of one of the two polyclonal serum to identify the bands corresponding to the DHP subunits, i.e., one membrane was reacted with anti-.beta. and the other was reacted with anti-.gamma.. Peroxidase-conjugated second antibody was used to colorimetrically detect the cross-reacting bands. The sheep antiserum reacted with the .alpha..sub.1 and .beta. subunits; the guinea pig antiserum reacted with the .gamma. subunit only.

Using the stained vertical strip as a guide, horizontal strips corresponding to the .beta. and .gamma. subunit proteins were cut from the immunoblot. To block non-specific binding sites on the nitrocellulose, the strips were incubated with BLOTTO (Bovine Lacto Transfer Technique Optimizer; 50 mM NaH.sub.2 PO.sub.4, 150 mM NaCl, pH 7. 4, 5% Nonfat dry milk) for 1 hour at 22.degree. C. The filters were then incubated overnight at 4.degree. C. with 2 ml of either the .beta. or .gamma. polyclonal serum diluted with 8 ml of TBS-BSA (50 mM Tris.HCl 7.4, 150 mM NaCl, 3% BSA). The strips were then washed three times for 15 min each time in 500 mM NaCl, 50 mM Tris.HCl (7.4), followed by three sequential 15 min washes in 100 mM NaCl, 10 mM Tris.HCl (7.4). The bound antibody was eluted with 5 ml of acid (50 mM glycine HCl, pH 2.5 ) at a temperature of about 22.degree. C. The pH of the eluted antibody wash was neutralized to 7.4 by the addition of 1.0M Tris.HCl, pH 8.0.

The affinity-purified antisera were characterized similarly to the polyclonal sera, and selection for sensitivity and subunit selectivity was done for antibody titer greater than 1:40. The .beta.specific antisera was called Affi-.beta., and the .gamma.-specific antisera was called Affi-.gamma..

EXAMPLE 2

ANTIBODY SCREENING OF A RANDOM-PRIMED RABBIT SKELETAL MUSCLE cDNA LIBRARY AND CLONE CHARACTERIZATION

A. Library Screening

A rabbit skeletal muscle cDNA library was prepared in .lambda.gt11, as described by Ellis, et al. (1988), Science 241, 1661-1664, using random primers [pd(N).sub.6 hexamers (Pharmacia, Inc., Piscathaway, N.J.)] rather than oligo-d(T) to prime single-strand cDNA synthesis. Two sets of duplicate filters of the library, each containing 720,000 plaques, were made and screened for clones containing phage with inserts coding for all or part of the .beta. and .gamma.-subunits of the rabbit skeletal muscle calcium channel. Each filter was screened using either the .beta.- or .gamma.-specific antibody as follows. BLOTTO was used for blocking of the nitrocellulose filters, dilution of the antibodies, and filter washes. Detection involved horseradish peroxidase-linked rabbit anti-sheep secondary antibody and development proceeded using 4-chloro-1-naphthol as a substrate.

The .lambda.gt11 library was plated on Y1090 in LB agar and 50 .mu.g/ml ampicillin. A Y1090 culture was grown overnight in 15 ml of LB, 0.2% maltose and 50 .mu.g/ml ampicillin. The cells were pelleted and resuspended in 3 ml of 10 mM MgSO.sub.4. Each of six plates was prepared by absorbing .about.120,000 phage of the .lambda.gt11 cDNA library to 300 .mu.l of the 3 ml cell solution and then pouring onto an LB agar plate in 10 ml soft agar containing 50 .mu.g/ml ampicillin.

The plates were grown at 42.degree. C. for 3.5 hours and then overlayed with IPTG-treated filters which were soaked in 10 mM IPTG (Boehringer Mannheim Biochemicals, Indianapolis, Ind.). The filters were dried until just moist, laid in the plates and incubated for 3.5 hours at 37.degree. C. After the incubation, the filter was oriented, removed, and a second, identical IPTG-treated filter was applied to the agar plate for an additional four hours at 37.degree. C. The orientation of this second filter was identified and marked identically to the first filter.

Two sets of six plates were prepared. One set of filters was used to screen for .beta.-subunit encoding clones and the other set was used to screen for .gamma.-subunit encoding clones. After incubation, one-half microgram of purified DHP receptor was spotted on one filter of each set as a positive control. The filters were washed for 30 min at room temperature with BLOTTO. One set of duplicate filters was then incubated with affinity purified sheep anti-.beta.-subunit polyclonal antibody, and the other set was incubated with affinity purified guinea pig anti-.gamma.-subunit polyclonal antibody. The filters were incubated with their respective antibody overnight in BLOTTO at 4.degree. C. They were then washed three times, for 10 min each time, in BLOTTO.

The filters were then incubated for one hour with BLOTTO containing an appropriate second antibody: HRP-goat anti-sheep IgG (for .beta.); or HRP-goat anti-guinea pig (for .gamma.)].

The filters were washed as described above for the first antibody, and rinsed with ddH.sub.2 O to remove BLOTTO.

The positive clones were developed using about 40 ml/plate of 4-chloro-1-naphthol reagent, which is made by dissolving 60 mg of said developer in 20 ml of ice cold MeOH and mixing 4-chloro-1-naphthol (Aldrich Chemical Company, Milwaukee, Wis.) into 100 ml of TBS containing 60.mu.l of 30% H.sub.2 O.sub.2.

Five positives were identified on both duplicate filters probed with the .beta. antibody and two positives were identified on both filters probed with the .gamma. antibody. Of the five putative .beta. clones originally identified, two were characterized and used further: K5 and K6. Both of the .gamma. clones were characterized: J6 and J10. The four clones were plaque-purified by replating and rescreening with the appropriate antibody using the same conditions as before, until a single, well-isolated plaque was identified for each clone.

B. Characterization of Putative .beta.- and .gamma.-Specific Inserts

The inserts of the putative .beta. and .gamma. clones were characterized by gel sizing, screening with antibody probes, restriction enzyme mapping, and partial sequencing.

1. Gel sizing

A mini-phage prep was made of each positive clone and the DNA was digested with EcoRI following manufacturer's instructions. The digests were then separated on a 1% agarose gel using the TAE buffer system (Maniatis, 1982), and HindIII digests of control .lambda. DNA as size standards. Bands were visualized using ethidium bromide and u.v. illumination. Based on the results of gel analyses, the approximate size of the inserts were found to be as follows:

______________________________________K5 500 bp J6 800 bpK6 900 bp J10 500 bp______________________________________

2. Restriction Enzyme Mapping

To further characterize the inserts of the putative .beta. and .gamma. clones, the inserts were restriction mapped. The EcoRI inserts were subcloned into EcoRI-cut, bovine alkaline phosphatase (BAP)-treated modified pUC19 plasmid. The modified pUC19 plasmid was made by digesting pUC19 with HincII and SmaI and ligating the blunt ends. The resultant plasmid was 17 bp shorter than pUC19 and lacked several restriction sites, i.e., SmaI, XmaI, BamHI, XbaI, HincII, AccI, and SalI. Ten micrograms of the ligation reaction were used to transform DH5.alpha. cells. AmpR colonies were selected. Plasmid preps were made from the amp.sup.R colonies and aliquots of the DNA were separately digested with selected restriction enzymes. The maps of inserts of the .beta. and .gamma. clones are shown in FIGS. 2 and 3, respectively.

3. Antibody probes screen

Plasmid-containing bacteria were grown overnight on ampicillin plates (having a colony density of about 100 colonies per plate). Replicate lifts were made using dry untreated nitrocellulose filters. The replicate lifts were then incubated for 3 hours at 37.degree. C. on filter paper wetted with LB Media plus 10 mM IPTG. Incubation was carried out with filters maintained in an inverted position to induce the production of fusion proteins.

Cells were lysed by exposing the filters to chloroform saturated vapors for 5 minutes, followed by overnight incubation at 4.degree. C. in BLOTTO containing 10 mg/ml lysozyme. The filters were then immunoblotted with the appropriate antibodies using the methods described above.

a. .beta.-specific clones

Both clones tested positive with the affinity-purified antisera Affi-62 . Additionally, only clone K5 tested positive with .beta.-specific monoclonal antibody VD21 [Leung, et al, J. Biol. Chem. 263,994-1001 (1988)].

b. .gamma.-specific clones

Neither of the two .gamma.-specific clones reacted with antisera specific for the .alpha..sub.1 -, .alpha..sub.2 -, .beta.-, or .delta.-specific antisera.

4. Partial Sequencing

a. .beta.-specific inserts

A large scale plasmid prep of K5 was prepared, and 2 g of the DNA per reaction was sequenced according to the procedure provided by United States Biochemical Corporation (Cleveland, Ohio; Sequenase DNA Sequencing Kit). The M13 reverse sequencing primer and the M13 sequencing primer (New England Biolabs, Inc. Beverly, Mass.) were used to prime the sequencing reactions.

Partial sequence of K6 was obtained by isolating two clones comprised of the K6 insert in the modified pUC19 vector in both orientations. The ends of the insert were sequenced using the reverse and forward M13 primers as described for K5. Then, using the SacI site 255 bp from the 5' end of K6 and the HindIII site 444 bp from the 3' end of the insert, in combination with the same restriction sites in the polylinker, deletions in K6 were generated and internal sequence was determined. For each restriction site there are two possible orientations and, thus, two deletions that could be generated. For example, the 255 bp SacI fragment can be released and the remaining DNA religated in order to generate sequence beginning 255 bp in from the 5' end. With the fragment in the opposite orientation, the remaining 684 bp SacI fragment can be released and sequence can then be generated beginning 684 bp in from the 3' end. Each of these deletions was generated (two for the SacI site and two for the HindIII site) and the entire K6 fragment was sequenced.

Analysis of the sequence data revealed a 167 bp overlap between clones K5 and K6, as well as predicted restriction enzyme recognition sites. The K6 insert was predicted to have at least one internal restriction site for the enzymes HindIII, SacI, and PstI. A single open reading frame is encoded by these two overlapping clones. No initiating ATG triplet or translation stop codon was identified. Based on the .beta.-subunit size of 52 kDa and a predicted coding region of at least 1400 bp, the isolated inserts were at least 150 bp short of being complete.

b. .gamma.-specific insert

The 484 bp EcoRI insert fragment was isolated from a large-scale preparation of J10 and subcloned into pGEM-3 (Promega Corp., Madison, Wis. 53711). pGEM-3 contains both the T7 and SP6 polymerase binding sites positioned on opposite ends of the polylinker sequence. Oligonucleotide primers specific for these polymerase binding sites were purchased from New England Biolabs, Inc. (Beverly, Mass. 01915). The J10 insert sequence was determined by sequencing in from each end of the insert directly out of the plasmid DNA. The Sequenase Kit was used as described above. Overlapping sequence at the center of the insert was observed, thus, the sequence of the entire insert was complete.

The J6 insert contains an internal EcoRI restriction site that divides the insert into two fragments, a .about.740 bp fragment and a .about.230 bp fragment. Initially, the larger, 740 bp fragment was subcloned into the modified pUC19 vector. The ends of the insert were sequenced using the universal and reverse primers described above, and J6 was positioned relative to J10 based on the determined sequence.

Oligonucleotide primers were synthesized based on the terminal sequences of J10. Internal J6 sequence was determined using the .gamma.-specific oligonucleotide primer based on the 5' end of J10. This determined sequence combined with the J10 sequence, completed the sequencing of the 740 bp J6 fragment and provided 819nts of continuous sequence.

To characterize the remaining sequence, the .about.230 bp J6 EcoRI fragment was subcloned into pGEM-3 and sequenced.

EXAMPLE 3

SCREENING OF OLIGO dT-PRIMED RABBIT SKELETAL MUSCLE cDNA LIBRARY AND CLONE CHARACTERIZATION

A. Screening

In an attempt to isolate inserts encoding the full length .beta.- and .gamma.-specific cDNA, or at least to isolate inserts encoding .beta.- and .gamma.-specific cDNA sequences that could be spliced to inserts from clones K5 and K6, and J6 and J10 to construct full length .beta.- and .gamma.-specific cDNAs, respectively, the inserts from clones K5 and K6, and J6 and J10 were used to screen two pairs of duplicate lifts of an oligo dT-primed rabbit skeletal muscle .lambda.gt11 cDNA library (Ellis, et al., Supra) under non-stringent conditions. Hybridization: 20% formamide, 5.times. Denhardts, 6.times. SSPE, 0.2% SDS, at 42.degree. C. Wash: 0.2.times. SSPE, 0.2% SDS, 45.degree. C.

Screening with the .beta. inserts yielded four positives on duplicate filters. The .gamma. insert screening yielded three positives on duplicate filters. The seven clones were plaque-purified and remained positive in subsequent screenings.

B. Characterization

The inserts from the positive clones were removed by EcoRI digestion as described previously and sized. The clone names and insert sizes are as follows:

______________________________________.beta.B 1800 G2 2300.beta.C 1500 G11 2000.beta.D 1500 G10 650.beta.E 2100______________________________________

The inserts listed above were subcloned into pUC19, modified as before, and restriction mapped and partially sequenced. The restriction maps allowed the various .beta. and .gamma. clones to be oriented [as shown in FIGS. 2 and 3].

The DNA sequence of insert .beta.D was determined directly out of the plasmid using the M13 primers as described previously. Insert .beta.D is 1293 bp. It extends 218 nts 5' of K5 and ends 176 nts short of the 3' end of K6 (See FIG. 2).

The sequence of the .beta.E insert at the 3' end was determined by sequencing the plasmid DNA priming with the universal primer (sequencing through the polyA tail) and priming internal sequence using an oligo synthesized based on the 3' end of K6. Overlapping sequence was determined, thus completing a continuous 1769 nt sequence from the 5' end of .beta.D to the 3' end of .beta.E.

The sequence at the 3' end of .beta.B was determined as described above. The sequence primed with the internal oligo matches the sequence determined for .beta.E. Sequence adjacent to the polyA tail was obscured and not determined. Even though individual nucleotides cannot be determined in this sequencing reaction, one can see from the general pattern that the sequence matches the sequence determined for .beta.E.

FIG. 4 shows the complete DNA sequence and deduced amino acid sequence of the cloned .beta.-subunit transcript. This sequence is a compilation of the DNA sequence determined by us and that reported by Ruth et al., Science 245, 1115 (1989). Ambiguities in the sequence determined by us were resolved by comparison of the sequence to that reported by Ruth et al. The amino acid residue through these regions are shown circled.

The termini of clone G10 were sequenced as described above and G10 was positioned relative to J6 and J10. A polyA tail was identified at the 3' end. Overlapping sequence was completed by priming G10 sequence internally using an oligonucleotide primer based on the 3' end of J10. The combined, determined sequence of J10, J6, and G10 was 1139 nt.

EXAMPLE 4

SCREENING OF OKAYAMA-BERG RABBIT SKELETAL MUSCLE cDNA LIBRARY FOR .gamma.-SUBUNIT INSERTS AND CLONE CHARACTERIZATION

A rabbit skeletal muscle cDNA library (MacLennan et al., 1985, Nature 316, 696) constructed according to Okayama and Berg was screened for .gamma. clones. Approximately 2.times.10.sup.6 recombinants were transformed into E. coli by standard methods. The transformation mix was divided into ten aliquots and each was separately inoculated into 100 mls of broth media. The transformed E. coli were grown to stationary phase and the plasmid DNA in each of the ten cultures was recovered by a standard plasmid prep procedure that was scaled down to 100 mls. Digestion of the pools with BamHI or XhoI released the insert from the Okayama-Berg vector pcD-X. The digest was probed with the J10 insert fragment. A series of fragments, the largest and most abundant fragment being .about.1.5 kb for XhoI and .about.1.4 Kb for BamHI were identified. These insert sizes are consistent with an .about.1200 nt transcript (see below) after additional sequence contributed by the Okayama-Berg vector, .about.250 nt for XhoI and .about.150 nt for BamHI, is subtracted. These fragments were present in both pool 8 and pool 10.

An aliquot of pool 8 was transformed into E. coli and .about.600,000 recombinants were probed with the J10 fragment. Approximately .about.50 to 75 duplicate positives were identified. Nine were chosen for further characterization. The insert size was identified in two clones: G4-3 and G6-1, .about.1275 nts and .about.1075 nts, respectively. The internal EcoRI site was mapped in each clone. G4-3 extends .about.475 nt 5' of the EcoRI site and 6-1 extends .about.275 nts 5' of the site. Further characterization of G4-3 by subcloning into M13 and DNA sequencing showed that G4-3 encoded the complete .gamma.-subunit coding sequence in addition to 48 nt of 5' untranslated sequence and 457 nt of 3' untranslated sequence. The .gamma.-subunit cDNA sequence and the determined amino acid sequence are shown in FIG. 1.

EXAMPLE 5

SEQUENCE OF THE .gamma. PROTEIN ISOLATED FROM RABBIT SKELETAL MUSCLE TISSUE

The .gamma.-subunit protein was electrophoresed and blotted onto immobilon by the method of Matsudaira [J. Biol. Chem. 262, 10035-10038 (1987)]. The portion of the blot that contained the .gamma.-subunit was excised from the immobilon membrane and analyzed for protein sequence. The stained immobilon pieces were placed into a sequencing reaction chamber above a biobrene-treated glass fiber filter. The Applied Biosystems 470 Gas Phase Protein Sequencer, with online phenylthiohydantoin analyzer, was run as described previously [Hunkapillar, M. W. and Hood, L. E. (1983), Science 219, 650; Hewick, R. M. et al., J. Biol. Chem. 256, 7990 (1981)], and the following sequence was determined: ##STR1##

The raw data for this sequencing analysis is presented in FIG. 5.

EXAMPLE 6

DEVELOPMENT OF CALCIUM CHANNEL SUBUNIT-EXPRESSING MAMMALIAN CELLS

A. Transient Transfection Protocol

The host cells transfected with calcium channel subunit-encoding DNAs, transfection selection plasmids, and marker construct plasmid were developed using the following protocol.

Approximately 24 hr prior to transfection, host cells are plated in 10 cm plates at a density of 2.5.times.10.sup.6 cells/plate and incubated at 37.degree. C. (5% CO.sub.2). One to eight hours prior to transfection, the cells are fed 9 ml of media [500 ml Dulbecco's Modified Eagle Medium; 4500 mg/L D-glucose; L-glutamine; 55 ml calf serum; 5 ml penicillin/streptomycin (100.times.: 10,000 U/ml/10,000 .mu.g/ml ) ].

The DNAs are prepared by combining 5 .mu.g of each subunit-encoding expression plasmid to be transfected with water to a final volume of 440 .mu.l. The water-DNA solution is mixed, and 60 .mu.l of 2M CaCl.sub.2 is added and mixed. This is called Solution B. Solution A is 500 .mu.l of 2.times.HBS (2.times.=10 g/L Hepes, 16 g/L NaCl; autoclaved; pH=7.10.+-.0.05) mixed with 10 .mu.l of 100.times.PO.sub.4 (1:1 mixture of 70 mM Na.sub.2 HPO.sub.4 and 70 Mm NaH.sub.2 PO.sub.4. A DNA precipitate is made by dripping Solution B into Solution A, while bubbling sterile air through Solution A for mixing. The precipitate is allowed to form during the next 30.+-.10 min.

The plated cells were transfected by adding, dropwise, .about.1 ml of precipitate to each plate containing cells and media and mixing well by swirling. The treated cells were returned to the incubator (37.degree. C.; 5% CO.sub.2) for five hr.

A glycerol shock is then applied to the transfected cells by removing the DNA precipitate and media and adding 2 ml of media containing 10% glycerol. After three minutes, the glycerol media is diluted with 5 ml PBS and mixed. The PBS-diluted media is aspirated off and the 5 ml PBS wash is repeated two more times. After the third PBS wash is aspirated off, the cells are fed 9 ml media and incubated at 37.degree. C. (5% CO.sub.2) for 24 to 72 hr.

The transfected cells can be analyzed 24-72 hours after DNA addition. Analysis can include 1) selection of transfectants, 2) northern analysis, 3) western analysis, 4) ligand binding studies, 5) functional analysis using a marker construct, and 6) electrophysiological measurements.

B. Mammalian Host Cells

The preferred cells for use as host for transfection with DNAs of the invention are Ltk-cells (ATCC CCL1.3).

C. Eukaryotic Expression Vectors Encoding Calcium Channel Subunit-Encoding DNAs

The calcium channel subunit expression plasmids were constructed using the following parent plasmids: (1) modified pSV2dhfr [Subramani et al., (1981) Mol. Cell. Biol. 1:854-864; modified by cutting at the unique HindII site and inserting a 5' -EcoRI-EcoRV-HindIII-3' polylinker], (2) pSV2 (plasmid pSV2dhfr was digested with HindIII and BglII, releasing the dhfr sequence, which was then replaced with a 5' -HindIII-EcoRV-EcoRI-BglII-3' adapter), and (3) pcD-X [Okayama, H. and Berg, P. (1983) Mol. Cell. Biol. 3:280-289].

For each expression construct, the insert and vector were ligated at approximately a 2:1 insert-to-vector molar ratio, generally using 50 to 100 ng of vector. Restriction maps of the .alpha.1, .alpha.2 and .beta. clone inserts, referred to in 1 and 2 below, are shown in FIGS. 6, 7 and 4, respectively.

1. pSKmCaCH.alpha.1pSV2dhfr

The 1.55 kb KpnI (polylinker sites) fragment from clone .lambda.SKmCaCH.alpha.1.8 [Ellis et al., (1988); Science 241: 1661-1664)] was inserted into a plasmid vector (e.g., pIBI24, pUC18, pUC19), and the KpnI-EcoRI fragment, encoding nucleotides -78 to 1006 of the .alpha.1 subunit, was gel purified.

The 4900 bp EcoRI-BamHI (BamHI site in pcD-X vector) fragment from clone pSKmCaCH.alpha.1.7 (Ellis et al., supra) was isolated and gel purified. The BamHI digest was a partial digest to avoid cutting at the internal BamHI site located at .about.4300 bp. This fragment was subcloned into EcoRI-BamHI-digested pIBI24 (International Biotechnologies, Inc.; New Haven, Conn.), and the .about.4900 nt 3' EcoRI-XbaI (XbaI polylinker site) fragment from that plasmid was gel purified.

The .about.1085 nt fragment originating from .about.SKmCaCH.alpha.1.8 was ligated to the .about.4900 nt fragment originating from pSKmCaCH.alpha.1.7, and the ligation was cloned into KpnI-XbaI-digested pGEM3 (Promega Corp., Madison, Wis. 53711). An upstream ATG present in the ligated insert, contributed by the EcoRI-KpnI-NcoI adapter used to construct the library from which .lambda.SKmCaCH.alpha.1.8 was isolated, was removed as follows. The plasmid (pGEM3+ insert) was linearized with KpnI, and the linearized DNA was digested with T4 polymerase in the dCTP; this limits the digestion in both directions to the first C nucleotide. The ends were then made blunt by digestion with S1 nuclease. The linear DNA was ligated to itself and transformed into DH5.alpha. cells. The deletion was confirmed by sequencing. This clone was called pSKmCaCH.alpha.1.1.DELTA.NcoI.

Modified vector pSV2dhfr was digested with EcoRI and HindIII. A partial EcoRI digest and a HindIII digest of clone pSKmCaCh.alpha.1.1.DELTA.NcoI was performed, and the .about.5950 bp insert encoding the .alpha.1 subunit was isolated on a low percentage agarose gel. The vector and insert fragments were ligated together and transformed into DH5.alpha. cells. AmpR cells were selected. Correct plasmid was confirmed by restriction mapping and was called pSKmCaCH.alpha.1pSV2dhfr. It contains the coding sequence of the .alpha.1 subunit of the rabbit skeletal muscle calcium channel, preceded by approximately 78 nucleotides of 5' sequence and followed by the 3' untranslated sequence and a stretch of As.

2. pSKmCaCH.alpha.2pSV2

Equal molar amounts of the 650 bp KpnI-HindII fragment from .lambda.SKmCaCH.alpha.2.15 and the 2700 bp HindIII-XbaI fragment from .lambda.SKmCaCH.alpha.2.3 (both described in Ellis et al., supra) were ligated together with 50 ng of KpnI-XbaI-digested and dephosphorylated pIBI24. The ligation was transformed into competent NM522 cells (Stratagene, Inc., San Diego, Calif.) and AmpR colonies were selected. The correct clone was identified by restriction mapping and was called p.alpha.2.1. Clone p.alpha.2.1 was digested with XbaI and SphI (SphI site in vector) and dephosphorylated. A second three-way ligation was performed with the XbaI-SphI fragment from p.alpha.2.1 and the 750 bp XbaI-NcoI fragment from .lambda.SKmCaCH.alpha.2.3 and the 1800 bp NcoI-SphI fragment from .lambda.SKmCaCH.alpha.2.6. The ligation was transformed into NM522 cells and AmpR colonies were selected. Correct plasmid was identified by restriction digest mapping and was called p.alpha.2.15.

Vector pSV2 was digested with EcoRI and EcoRV and ligated to the .about.3600 bp gel purified, SmaI-EcoRI fragment from p.alpha.2.15. The ligation was transformed into DH5.alpha. cells. AmpR cells were selected. Correct plasmid was identified by restriction digestion, and was called pSKmCaCH.alpha.2pSV2. It contains the coding sequence of the .alpha.2 subunit of the rabbit skeletal muscle calcium channel, preceded by approximately 56 nucleotides of 5' sequence and followed by .about.300 nt of 3' untranslated sequence.

3. pSKmCaCH.alpha.pSV2

A full-length construct of the .beta.-encoding sequence was made by ligating the 975 bp EcoRI-HindII fragment (5' end of.beta.) from p.beta.D (gel-purified) and the 775 bp HindIII-EcoRI fragment (3' end of .beta.) from p.beta.B (gel purified) together. The ligation was then digested with EcoRI and the 1700 bp EcoRI fragment was gel purified and ligated into EcoRI-digested and dephosphorylated pGEM3Z. The ligation was transformed into DH5.alpha. cells, and restriction mapping was performed to identify the correct clone. The 1700 bp insert of the correct clone was released by EcoRI digestion, gel purified, and ligated into EcoRI-digested and dephosphorylated vector pSV2. The ligation was transformed into DH5.alpha. cells. AmpR cells were selected. Correct plasmid was identified by restriction mapping and was called pSKmCaCH.beta.pSV2. It contains the coding sequence of the .beta. subunit of the rabbit skeletal muscle calcium channel, preceded by approximately 92 nucleotides of 5' sequence and followed by 3' untranslated sequence and a stretch of As.

4. pSKmCaCH.gamma.pcD-X

Clone G4-3, isolated from the Okayama-Berg library (see Example 4), was renamed pSKmCaCH.gamma.pcD-X. It contains the coding sequence of the .gamma. subunit of the rabbit skeletal muscle calcium channel, preceded by approximately 48 nucleotides of 5' sequence and followed by 3' untranslated sequence and a stretch of As. The insert is contained in the pcD-X expression vector.

D. Development of Cell Line Containing Four Calcium Channel Subunit DNAs

Cells were developed to express all four subunits of a calcium channel by transfecting Ltk-cells with 5 .mu.g each of plasmids pSKmCaCH.alpha.1pSV2dhfr (.alpha.1),pSKmCaCH.alpha.2pSV2 (.alpha.2), pSKmCaCH.alpha.1pSV2 (.gamma.),pDKmCaCH.gamma.pcD-X (.beta.), and 1 .mu.g of the selection plasmid pThx24 [Zipser et al. (1981); Proc. Natl. Acad. Sci. 78:6276-6280], following the protocol of (A). Positive transfectants were selected in HAT media (media in "A", plus 15 .mu.g/ml hypoxanthine, 1 .mu.g/ml aminopterin, 5 .mu.g/ml thymidine) following standard procedures.

While the present invention has been described in detail herein, those of ordinary skill in the art will recognize numerous variations and modifications, in what is described, that are within the spirit of the invention. Such variations and modifications are within the scope of the invention as described and claimed herein.

Claims

1. An isolated DNA fragment, comprising a sequence of nucleotides that encodes the .gamma.-subunit of a rabbit skeletal muscle calcium channel that has the amino acid sequence set forth in FIG 1.

2. The DNA fragment of claim 1, wherein the .beta.-subunit is encoded by the sequence of nucleotides set forth in FIG. 1.

3. An eukaryotic cell transformed with the DNA fragment of claim 1.

4. An eukaryotic cell transformed with the DNA fragment of claim 2.

5. An isolated DNA fragment, comprising a sequence of nucleotides that encodes the .gamma.-subunit of a DHP-sensitive rabbit skeletal muscle calcium channel.

6. An eukaryotic cell transformed with heterologous DNA, wherein the heterologous DNA comprises the DNA fragment of claim 5.