Isolation and Sequence Analyses of cDNA Clones for the Large Subunits of Two Isozymes of Rabbit Calcium-dependent Protease*

Two sets of cDNA clones were isolated from cDNA libraries prepared from poly(A+) RNA of rabbit lung and spleen by screening with the cDNA probe for the large subunit (80-kDa subunit) of chicken calcium- dependent protease (Ca2+-protease; Ohno, S., Emori, Y., Imajoh, S., Kawasaki, H., Kisaragi, M., and Suzuki, K. (1984) Nature 312, 566-570). The two sets of clones were identified as cDNA clones for two Ca2+-protease isozymes with high (p-type) and low (m-type) calcium sensitivities from a comparison of the primary structures deduced from the nucleotide sequences with partial amino acid sequences from the two isozymes. The cDNA clones for the 80-kDa subunits of the p- and m-type Ca2+-proteases contained, in total, about 1.5- and 2.2-kilobase cDNA inserts, respectively, which correspond roughly to the C-terminal halves of the coding regions and the entire 3”noncoding regions. The two isozymes are encoded by two distinct mRNA species present in all the tissues examined, although the amount of mRNA significantly differs among the various tissues. Four E-F hand structures,

Isolation and Sequence Analyses of cDNA Clones for the Large Subunits of Two Isozymes of Rabbit Calcium-dependent Protease* (Received for publication, October 11, 1985) Yasufumi Emori, Hiroshi Kawasaki, Hidemitsu Sugihara, Shinobu Imajoh, Seiichi KawashimaS, and Koichi Suzuki Two sets of cDNA clones were isolated from cDNA libraries prepared from poly(A+) RNA of rabbit lung and spleen by screening with the cDNA probe for the large subunit (80-kDa subunit) of chicken calciumdependent protease (Ca2+-protease; Ohno, S., Emori, Y., Imajoh, S., Kawasaki, H., Kisaragi, M., and Suzuki, K. (1984) Nature 312, 566-570). The two sets of clones were identified as cDNA clones for two Ca2+protease isozymes with high (p-type) and low (m-type) calcium sensitivities from a comparison of the primary structures deduced from the nucleotide sequences with partial amino acid sequences from the two isozymes. The cDNA clones for the 80-kDa subunits of the p-and m-type Ca2+-proteases contained, in total, about 1.5and 2.2-kilobase cDNA inserts, respectively, which correspond roughly to the C-terminal halves of the coding regions and the entire 3"noncoding regions.
The two isozymes are encoded by two distinct mRNA species present in all the tissues examined, although the amount of mRNA significantly differs among the various tissues.
Four E-F hand structures, typical calcium-binding structures in various calcium-binding proteins such as calmodulin, were detected in the C-terminal regions of both isozymes, as in the case of chicken Ca2+-protease.
Comparison of the amino acid sequences of the two rabbit isozymes and the corresponding region of the chicken enzyme revealed marked homology, which indicates that these three enzymes have the same evolutionary origin. Furthermore, we suggest that the ptype rabbit Ca2+-protease, rather than the m-type, is similar to chicken Ca2+-protease, which is regarded as an m-type enzyme in the C-terminal region. The evolution and molecular basis of the differences in calcium sensitivities of the Ca2+-proteases are discussed.
Calcium-dependent protease (Ca2+-protease (calpain)) is an intracellular protease that shows a wide distribution (1)(2)(3) and comprises two subunits. The large subunit (80-kDa subunit) is a catalytic subunit (4,5) with calcium-binding regions ( 6 ) , but the function of the small subunit (30-kDa subunit) is not yet clear. In mammals, two isozymes ( p -and m-types) *This work was supported in part by research grants from the Ministry of Education, Science, and Culture, a grant-in-aid for new drug development, and grants from National Center for Nervous, Mental, and Muscular Disorders from the Ministry of Health and Welfare of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
have been isolated (7)(8)(9)(10)(11)(12). They are similar in various molecular and enzymatic properties, but differ markedly in their calcium sensitivities. The p-type enzymes have high calcium sensitivity and are active at micromolar concentrations of calcium. In contrast, the calcium sensitivity of the m-type enzymes is low, and they require millimolar calcium for their activity. On the basis of peptide mapping and amino acid sequencing,' we have shown that the small subunits of the rabbit p-and m-type enzymes are identical. The difference in the calcium sensitivity of the two isozymes is thus ascribed to the large subunits.
In the chicken, however, only a single species of Ca2+protease exists with an intermediate calcium sensitivity (13). As the first step in clarifying the mechanism of action, we determined the structure of the cDNA for the unique 80-kDa subunit of chicken Ca2+-protease (6).
As the next step, we determined the cDNA structures for the 80-kDa subunits of the two Ca2+-protease isozymes to clarify the basis of the calcium sensitivity. We isolated cDNA clones for the 80-kDa subunit of rabbit Ca2+-protease. The amino acid sequences of the C-terminal halves of the p-and m-type Ca2+-proteases were determined from the nucleotide sequences and compared with that of chicken Ca2+-protease. These results are presented together with a discussion of the molecular basis of the calcium sensitivity and of the evolution of Ca2+-protease.

EXPERIMENTAL PROCEDURES
Materink-Guanidinium isothiocyanate and cesium chloride were purchased from Wako Pure Chemical Co. Oligo(dT)-cellulose (Type VII), Escherichia coli DNA ligase, and RNase H were products of P-L Biochemicals. Reverse transcriptase and S1 nuclease were obtained from Seikagaku Kogyo Co. DNA polymerase from E. coli and its Klenow fragment, terminal deoxynucleotidyltransferase, bacterial alkaline phosphatase, and T4 polynucleotide kinase were obtained from Takara Shuzo Co. The restriction enzymes used were from Takara Shuzo CO., Toyobo Co., and P-L Biochemicals. The nick translation kit and [y-32P]ATP (-3000 Ci/mmol) were from Amersham Corp.
Construction of cDNA Libraries-Total RNA was extracted from rabbit lung or spleen by the guanidinium isothiocyanate/cesium chloride method (14). Poly(A+) RNA was purified by oligo(dT)-cellulose column chromatography (14). After single-stranded cDNA synthesis with reverse transcriptase with total poly(A+) RNA as a template, double-stranded cDNA synthesis was performed according to Gubler and Hoffman's method (15) using E. coli DNA polymerase, E. coli DNA ligase, and RNase H or the self-priming method (14) using the Klenow fragment and S1 nuclease. The double-stranded cDNA was oligo(dC)-tailed at the 3'-ends with terminal transferase, annealed with PstI-cut and dG-tailed pBR322, and then introduced into E. coli H. Kawasaki et al., manuscript submitted for publication.

9465
cDNA Cloning for Two Isozymes of Rabbit Ca2+-protease "294 by Hanahan's method (16). Isohtion of Caz+-protease cDNA Clo~s-Tetracycline-resistant transformants (about lo5 colonies) were screened by in situ hybridization (14) with nick-translated chicken Ca2+-protease cDNA as a probe in a solution containing 50 mM Tris-HC1 (pH 8.0), 1 M NaC1, 10 mM EDTA, 0.2% bovine serum albumin, 0.2% polyvinylpyrrolidone, 0.2% Ficoll 400, 0.2% SDS? and 100 pg/ml heat-denatured salmon sperm DNA (9). Washing was performed with 2 X SSC (SSC is 0.15 M NaCI, 0.015 M sodium citrate) containing 0.1% SDS at 65 "C. The eight independent clones obtained were subjected to crosshybridization among themselves, and they could be divided into two types. The longest representative of each type was sequenced, and the deduced amino acid sequences were compared with those determined with peptides obtained from the two isozymes as described below.
Partial Amino Acid Sequence Determination of Peptides of the Two Ca2+-proteme Isozymes-The two isozymes of rabbit Ca2+-protease ( p -and m-types) were purified from rabbit skeletal muscle as described previously (11). The purified enzymes were further fractionated into two subunits as described in the Miniprint S~pplement.~ The purified large subunits were digested with trypsin or lysylendopeptidase, and then the digests were fractionated on a CIS reverse-phase high pressure liquid chromatography column. The amino acid sequences of the purified peptides were determined with a gasphase protein sequenator (Applied Biosystems, Inc., Model 470A).
RNA Blot Hybridization-Total RNA was extracted from various rabbit tissues as described above. Each RNA sample ( 2 pg of total RNA) was denatured and electrophoresed in a formaldehyde-containing agarose gel (14). After electrophoresis, the RNA was transferred to a nylon membrane (Biodyne BNRG) and then hybridized with nick-translated cDNA (14) at 42 "C in a solution containing 50 mM sodium phosphate (pH 7.0), 5 X SSC, 50% (v/v) formamide, 0.1% bovine serum albumin, 0.1% polyvinylpyrrolidone, 0.1% Ficoll 400, 0.2% SDS, and 200 pg/ml heat-denatured salmon sperm DNA. The filter was washed in 0.1 X SSC, 0.1% SDS at 50 "C and then exposed to Fuji RX film at -80 "C with an intensifying screen.
DNA Sequencing-DNA sequencing was performed by the 5'-end labeling and chemical modification method of Maxam and Gilbert (17). DNA fragments prepared by digestion of recombinant plasmids with restriction enzymes were dephosphorylated with alkaline phosphatase and then 5"end-labeled with [ Y -~~P J A T P with T4 polynucleotide kinase. After the second digestion with an appropriate restriction enzyme, the 32P-labeled DNA fragments were recovered from polyacrylamide gels and processed for chemical modification and cleavage reaction.

RESULTS
Identification of Two Isozymes of Calcium-dependent Protease cDNA Clones-From about lo5 colonies, eight clones were isolated as rabbit Ca2+-protease cDNA clones using chicken Ca2+-protease cDNA as a probe. The eight clones were divided into two types on the basis of the results of cross-hybridization experiments. Two clones, representative of the two types (pLU1001 and pLM28, Fig. l), were first sequenced. The amino acid sequence deduced from the nucleotide sequence of pLUl001 contained the partial amino acid sequences of two peptides obtained from the 80-kDa subunit of the p-type Ca2+-protease (shown in the Miniprint Supplement (Table I-S) and indicated by horizontal arrows in Fig. 2 A ) . Similarly, the sequence of pLM28 contained the sequences of nine peptides from the 80-kDa subunit of the m-   Fig. 2) are indicated by solid boxes. The sites of 5'-end labeling are indicated by short vertical lines at the arrow ends. The wavy double slushes at the arrow ends indicate that the site of 5'-end labeling was the PstI site at the end of the poly(dG) .poly(dC) tail. Clone pSU20 was isolated from the spleen cDNA library, and the others were isolated from the lung cDNA library. bp, base pair. type Ca2+-protease (shown in the Miniprint Supplement (Table I-S) and indicated by horizontal arrows in Fig. 2B). Thus, pLUlOOl and pLM28 were assigned as the cDNA clones for the 80-kDa subunits of the p-and m-type Ca2+-proteases, respectively.
Nucleotide Sequences of p-and m-Type Caz+-proteose cDNAs-Restriction mapping and nucleotide sequence analyses of the two types of clones gave the following results. The p-type cDNA clones contained the C-terminal sequence of 302 residues, which corresponds to positions 404-705 of chicken Ca2+-protease 80-kDa subunit (6) (Fig. 4), and the complete 3"noncoding region (Figs. 1 and 2.4). The m-type cDNA clones contained the C-terminal424-residue sequence, corresponding to residues 282-705 of the chicken enzyme (6) (Fig. 4).
The 3"noncoding regions of mRNAs for the three enzymes (the chicken enzyme and the p-and m-type rabbit enzymes) are markedly different both in length and sequence. The chicken mRNA has 1227 nucleotides in the 3"noncoding sequence (6), whereas the rabbit p-and m-type cDNAs have 580 and 887 nucleotides, respectively. No significant sequence homology among them could be detected, although their coding regions are highly homologous, with -65% homology, as described below.
Identification and Characterization of mRNA for Ca2+-protease in Various Rabbit Tissues-The presence of mRNAs for the two isozymes was examined by RNA blot hybridization. We used the nick-translated 3"noncoding regions of the cDNAs for the two rabbit isozymes as probes because their   Table I Fig. 3 show that the mRNA for both enzymes is expressed in all the rabbit tissues examined, although the amount differs considerably from tissue to tissue. The Ca2+-protease mRNA content was the lowest in liver and the highest in lung for both types of isozyme. The length of each mRNA was about 3.5 kilobases, but the m-type mRNA was slightly larger than the p-type, probably due to the different lengths of their 3'-noncoding regions, as described above. In spleen, a smearing around the 3.5-kilobase band was always observed for the p-type enzyme. The same result FIG. 3. RNA blot analysis of total RNA from various rabbit tissues. A, p-type; B, m-type. RNA (2 pg) was transferred to a nylon membrane after agarose gel electrophoresis as described under "Experimental Procedures" and then hybridized with nick-translated cDNA fragments corresponding to the 3"noncoding regions. The DNA fragments used were from residues 908-1401 (DdeIIRsaI fragment) for the p-type (see Fig. 2 A ) and from residues 1484-2123 (RsaI fragment) for the m-type (see Fig. 2B was seen when its coding region was used as a probe. Furthermore, a faint band of about 2 kilobases was usually detected in brain for the p-type enzyme (as shown in Fig. 3A). The reasons for these phenomena are not clear.

1309
Amino Acid Sequences of the 80-kDa Subunits of the p-and m-Type Rabbit Ca2+-proteases-We previously determined the total amino acid sequence of the chicken Ca2+-protease 80-kDa subunit from the cDNA nucleotide sequence (6). The cDNA clones for the 80-kDa subunits of the two rabbit Ca2+proteases obtained here contained the C-terminal halves of their coding regions, and their amino acid sequences were compared with that of the chicken Ca2+-protease 80-kDa subunit. Remarkable sequence homology among the three enzymes was found. Their sequences can be aligned as shown in Fig. 4, with only two deletions in the sequence for rabbit m-type Ca2+-protease. The per cent homology is summarized in Table I in terms of the amino acid and nucleotide sequences. The highest amino acid sequence homology was found between the rabbit p-type enzyme and the chicken enzyme. The situation was the same when the nucleotide sequences were compared. Namely, the rabbit p-type enzyme is more similar to the chicken enzyme, an enzyme of a different species, than to the m-type enzyme of the same species.

the E-F hand structures for the putative calcium-binding regions of rabbit Ca2+-protease
In the test sequence, 0 denotes oxygen-containing amino acid residues (D, E, N, Q, S, T), L denotes hydrophobic residues (L, V, I, F, M), E denotes E, and G means G. Residues matching the test sequence are marked by asterisks below the sequences and were counted to obtain the scores. For the positions of the regions shown, see Fig. 4. The C-terminal region of about 170 residues of the chicken Ca2+-protease 80-kDa subunit has been identified as the calcium-binding domain (6). This region, as in calmodulin and parvalbumin, contains four consecutive calcium-binding regions, so-called E-F hand structures (18, 19). Alignment of the two rabbit isozymes with the chicken enzyme (Fig. 4) revealed the presence of four putative calcium-binding regions in the C termini of the two rabbit enzymes, as found for the chicken enzyme. These four regions in both isozymes show high scores for homology when the presence of the E-F hand structures was examined with the test sequence (19, 20). As summarized in Table 11, the scores for the p-type enzyme are 12, 12, 11, and 13 and for the m-type enzyme are 12, 11, 11, and 13 in order from the N terminus. The scores for the chicken enzyme are 11 for all the four regions (6), which is lower than those for the two rabbit enzymes. Thus, these four regions in both the p-and m-type rabbit enzymes can be regarded as E-F hand structures and putative calcium-binding regions. The spacing of these four E-F hand structures along the peptide chain is similar to that in calmodulin (18), as in the case of the chicken enzyme (6).

Region
The predicted secondary structures based on the sequences (21) indicate that for the putative calcium-binding loops the probabilities of the formation of either a-helix or @-sheet structures are low and that they are flanked by a-helix or @sheet structures (data not shown). These are typical features of the secondary structure for E-F hand structures (18-20). DISCUSSION We have isolated cDNA clones for the 80-kDa subunits of two isozymes of rabbit Ca2+-protease and have shown that the two isozymes are evolutionarily related. Amino acid and nucleotide sequence homology reveals that the chicken enzyme (6), which was previously designated as the m-type (13), is similar to the p-type rabbit enzyme rather than to the mtype in the C-terminal region. In chicken tissues, only one type has been isolated, and its calcium sensitivity is inter- * * * * * * * * * mediate between those of the two isozymes from mammalian sources (13). Although the total structure of the two isozymes and further comparative studies on Ca2+-protease from various species are required, we can assume that there was once only one type of Ca2+-protease, as now found in chicken, and it evolved into two species with different calcium sensitivities, as now found in mammals. The chicken enzyme might be a prototype Ca2+-protease. In contrast to the marked sequence homology in the coding regions, the 3"noncoding regions of the cDNAs for the three Ca"+-proteases show no distinct homology. This suggests that there is a greater pressure to conserve the sequence of the coding regions for functional conservation.
The two species of mRNA for the p-and m-type isozymes showed a wide distribution, being detected in all the rabbit tissues examined. No significant difference in the tissue distributions of the p-and m-type enzymes was found. But, the contents of the mRNAs in various rabbit tissues differed significantly. The content of mRNAs roughly corresponds to the amount of Ca2+-protease in various chicken and rat tissues determined by measuring the enzyme activities (22, 23). Similar tissue distribution results were reported for the chicken enzyme mRNA (6,24).
The two isozymes of rabbit Ca2+-protease differ significantly in their calcium sensitivities. The p-type enzyme requires 40 pM Ca2+ for 50% activity, and the m-type enzyme requires 700 p~ Ca2+ (11). However, both enzymes have four consecutive E-F hand structures in their calcium-binding domain, like calmodulin, and the homology scores for the calcium-binding regions of the two rabbit enzymes and the chicken enzyme are similar. Therefore, differences in calcium sensitivity of the Ca2+-proteases may be ascribed to finer structural differences in the calcium-binding regions. In this respect, it is noteworthy that the two types of rabbit enzyme clearly have different sequences at the C-terminal ends of 5 or 6 residues (Fig. 4). In addition, the sequences of the loop structures in calcium-binding regions IV and 111 vary (see Table I1 and Fig. 4). To clarify the basis of the calcium sensitivity, the total amino acid sequences of the two rabbit enzymes should be determined together with those of other species. Studies along these lines are in progress, and the results will be presented elsewhere.