four repeating domains of the endogenous inhibitor for calcium

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1988 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 263, No. 5. Issue of February 15, pp. 2364-2370.1988 Printed in U. S A .

All Four Repeating Domains of the Endogenous Inhibitor for Calcium- dependent Protease Independently Retain Inhibitory Activity EXPRESSION OF THE cDNA FRAGMENTS IN ESCHERICHZA COLI*

(Received for publication, September 21, 1987)

Yasufumi Emori, Hiroshi Kawasaki, Shinobu Imajoh, Yasufumi Minami, and Koichi Suzuki From the Department of Molecular Biology, The Tokyo Metropolitan Institute of Medical Science, Honkomagome, Bunkyo-ku, Tokyo 113, Japan

We have already determined the primary structure of the endogenous inhibitor for calcium-dependent protease (CANP inhibitor, calpastatin) from the cDNA sequence and revealed that the CANP inhibitor contains four internally repeating units which could be responsible for its multiple reactive sites (Emori, Y., Kawasaki, H., Imajoh, S., Imahori, K., and Suzuki, K. (1987) Proc. Natl. Acad. Sei. U. S. A. 84,3590-3594). Restriction fragments of the cDNA corresponding to each of the four domains (encoding 104-156 amino acid residues of the total 718 residues) were subcloned into the multicloning site of pUC9 or pUC18 in a direction and frame matched to the lacZ‘ open reading frame of the vector. Under the lac operator-promoter system, we succeeded in producing truncated fragments of the CANP inhibitor in Escherichia coli. The CANP inhibitor fragments were partially purified, and the inhibitory activities toward calcium-dependent protease (CANP) were examined. All fragments containing well conserved regions of about 30 amino acid residues (domains I-IV) located in the middle of the four units exhibited the inhibitory activity. However, their inhibitory activities varied considerably. Further truncation experiments revealed that small fragments containing 30-70 amino acid residues of the CANP inhibitor still retained inhibitory activity. From these experimental results the following conclusions can be drawn: 1) each of the four repeating units of the CANP inhibitor (about 140 amino acid residues) is a real functional unit and can inhibit CANP activity independently; and 2) domains corresponding to well conserved sequences of about 30 amino acid residues containing a consensus Thr-Ile-Pro-Pro-X-Tyr-Arg sequence are essential for the inhibitory activity, and the bordering regions are important for its modulation.

Calcium-dependent protease (CANP,’ calpain, EC 3.4.22.17) is a widely distributed intracellular protease (1-4).

* This work was supported in part by grants-in-aid for scientific research on priority areas from the Ministry of Education, Science and Culture. Research grants from the same Ministry, National Center of Neurology and Psychiatry (NCNP) of the Ministry of Health and Welfare of Japan, Taisho Pharmaceutical Co., the Naito Foundation, and the Scandinavia-Japan-Sasagawa Foundation were also used. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The abbreviations used are: CANP, calcium-activated neutral protease; g-CANP, low calcium-requiring type of CANP; m-CANP, high calcium-requiring type of CANP; SDS, sodium dodecyl sulfate.

The activity of CANP is absolutely dependent on Ca2’, and two types of CANP, a low calcium-requiring type (p-CANP) and a high calcium-requiring type (m-CANP), exist (5-7). These two isozymes are composed of different large subunits (80 kDa subunit) and identical small subunits (30-kDa subunit) (7-9). Both subunits have four potential calcium-binding regions consisting of so-called E-F hand structures in their C-terminal domains (7,9). The large subunit contains a thiol protease domain homologous to other thiol proteases such as papain (10,ll). Thus, CANP is a member of both the cysteine proteinase family and the calcium-binding protein family.

In the presence of sufficient concentrations of calcium ion, the native CANP molecule is converted to an active form by autolysis which eliminates the N-terminal regions of both subunits (12). The activity of CANP is inhibited by a specific inhibitor (the CANP inhibitor, calpastatin) (1-4). This pro- teinaceous CANP inhibitor is widely distributed in the non- lysosomal intracellular fraction of the cells together with CANP.

The molecular weight of the CANP inhibitor, estimated by SDS-polyacrylamide gel electrophoresis, is about 68,000 for the erythrocyte inhibitor and about 110,000 for the inhibitor from other tissues such as heart and liver (13, 14). Recently, we demonstrated that these two inhibitors are derived from the same gene by different posttranscriptional or posttranslational processing and that the erythrocyte inhibitor lacks the N-terminal region of the liver inhibitor (15).

The precise mode of action of the CANP inhibitor is not known, but the following interesting observations have been reported. 1) The interaction between CANP and the CANP inhibitor is reversible, and both CANP and the CANP inhibitor retain full activity after dissociation of the complex (16). 2) The association of the CANP inhibitor with CANP is dependent on Ca2+ (17). 3) One mol of the CANP inhibitor from erythyrocyte and liver can inhibit 3 and 4 mol of CANP, respectively (15).

To elucidate the inhibition mechanism, we have isolated and sequenced cDNA clones for the CANP inhibitor from rabbit and have revealed the following facts (18). 1) The CANP inhibitor contains four internal repeating structures. 2) Each repeating unit of about 140 amino acid residues contains a highly homologous sequence of about 30 amino acid residues. This suggests that the repeating unit is the functional unit of the CANP inhibitor. 3) The amino acid sequence of the CANP inhibitor is not similar to those of other cysteine proteinase inhibitors (cystatins), indicating that the CANP inhibitor is a unique cysteine proteinase inhibitor and that the mechanism of inhibition may be different from other inhibitors. To analyze further the structure-

2364

Four Domains of the CANP Inhibitor Are All Reactive 2365

function relationship of the CANP inhibitor, we expressed its cDNA fragments in Escherichia coli and tested their inhibitory activities. Our preliminary experiment showed that the C-terminal region of the CANP inhibitor retains inhibitory activity (19).

In the experiments described in this report, we constructed a number of expression plasmids. Various regions of the cDNA were then expressed in E. coli under the control of the lac promoter contained in the pUC vectors. By measuring the inhibitory activities of partially purified CANP inhibitor fragments, we demonstrated that each of the four repeating units is a real functional unit and that a small peptide of about 60 amino acid residues retains inhibitory activity against both p- and m-CANP.

EXPERIMENTAL PROCEDURES

Bacterial Strains and Plasmids-E. coli strains used were HBlOl and YA21. The plasmids used were pUC9 (20) and pUC18 (21) containing lac promoter and lacZ’ near the multiple cloning sites.

Enzymes and Chemicals-Restriction endonucleases, T4 DNA li- gase, Klenow fragment of E. coli DNA polymerase I, T4 polynucleo- tide kinase, E. coli alkaline phosphatase, exonuclease 111, Ba131 nuclease, T4 DNA polymerase, and mung bean nuclease were pur- chased from Takara Shuzo, Co. An oligodeoxynucleotide linker for insertion of a termination codon, 5’-d(CTAGCTAGCTAG)-3’ (12T linker), was synthesized with a DNA synthesizer (Applied Biosystems Inc., Model 380A). Isopropyl-8-D-thiogalactopyranoside was pur- chased from Sigma.

cDNA Clones-Two cDNA clones for the rabbit CANP inhibitor (pCI-213 and 11) which contain the cDNA inserts of the clones XCI- 213 and 11 (18) were used for the construction of expression recombinant plasmids.

Bacterial Cell Growth for Expression-E. coli cells carrying recombinant plasmids were grown to the stationary phase in x-broth (22) with 50 pg/ml ampicillin at 37 “C. After centrifugation, the cell pellet was suspended in minimal medium M9 (22) supplemented with 0.2% glycerol, 0.2% casamino acid, 50 pg/ml ampicillin, and 1 mM isopropyl-@-D-thiogalactopyranoside, and the incubation was continued at 37 “C for 1.5-2 h. For pulse-labeling experiments, 10 pCi of [3H] proline (Amersham Corp.) was added to the culture (1 ml) in the M9 medium lacking casamino acid after 30 min induction with isopropyl- @-D-thiogalactopyranoside. After 2 min labeling, cells were collected and immediately suspended in a sample buffer for SDS-polyacrylamide gel electrophoresis (23).

Preparation of Cell Extracts and Partial Purification of Expressed CANP Inhibitor Fragments-Bacterial cells were harvested by centrifugation and resuspended in 10 mM Tris-HC1, pH 7.5,2 mM EDTA. Suspended cells were disrupted with a Branson Sonifier, Model 185. Cell debris were removed by centrifugation, and the crude extract was heat-treated for 10 min in a boiling water bath and cooled to 0 “C. Precipitable materials were removed by centrifugation, and HCl was added at a final concentration of 0.02 N to the supernatant. Supernatant fractions obtained by centrifugation were neutralized by NaOH and Tris-HC1, pH 7.4, and used as partially purified CANP inhibitor fragments. The fractions were free from protease activities or inhibitory activities toward CANP, and contained mainly truncated CANP inhibitor fragments (see “Results”).

Assay of CANP Inhibitor-p-CANP and m-CANP were purified from rabbit skeletal muscle (24). The protease activities were measured at 500 pM and 5 mM CaCL for p-CANP and m-CANP, respectively, using casein as a substrate (12). The assay of the CANP inhibitor was carried out essentially as described before (13).

DNA Sequencing-DNA sequencing was performed by the dideoxy chain termination method as described by Hattori and Sakaki (25).

SDS-Polyacrylamide Gel Electrophoresis-SDS-polyacrylamide gel electrophoresis was performed using the buffer system described by Laemmli (23). Gels were composed of 12% polyacrylamide, 0.6% bisacrylamide separating gels and 3% polyacrylamide, 0.15% bisacrylamide stacking gels. Staining was carried out with Coomassie Bril- liant Blue R250, and fluorography was carried out using Amplify (Amersham Corp.).

RESULTS

Construction of Expression Plasmids for CANP Inhibitor Fragments Containing the Four Repeating Domains-Our previous studies revealed that the CANP inhibitor contains four internal repeats (18) and that a truncated fragmedt containing the C-terminal region retains inhibitory activity when expressed in E. coli (19). To determine whether all four of the repeats independently exhibit inhibitory activity, we constructed the following two series of expression plasmids. Since two restriction enzymes, BstNI and Sau3A1, do not cut the well conserved regions of 30 amino acid residues (domains I- IV) in the four repeats (Figs. 1 and 2), plasmids containing the cDNA insert for the CANP inhibitor (pCI-213 and 11) were digested with BstNI (Fig. L4) or Sau3AI (Fig. 1B). Four fragments generated by BstNI digestion (Fig. LA) were filled in at their cohesive ends by the Klenow fragment and recovered from polyacrylamide gels. The fragments were inserted into a pUC9 or pUC18 plasmid vector which had been digested with HincII and treated with bacterial alkaline phosphatase. The four fragments generated by Sau3AI digestion (Fig. 1B) were inserted similarly into pUC18 at the BamHI site. Clones containing inserts in the desired direction were selected by investigating their restriction maps. A phospho- rylated termination linker (12T linker, see “Experimental Procedures”) encoding termination codons in all three frames was inserted into selected clones (pB1-4, pS1-4) at the filled- in EcoRI sites or Hind111 sites for pUC9 or pUC18 constructs, respectively. Constructions of the final plasmids, termed pBT1-4 and pST1-4, were confirmed by sequencing from both directions in the cDNA inserts.

As shown in Fig. 2, the constructed plasmids encoded fusion proteins containing 104-156 amino acid residues of the CANP inhibitor fragments; the N-terminal 10-12 residues originated from l a d ‘ and the polylinker of the pUC vectors, and the C- terminal 0-12 residues originated from the polylinker and the 12T linker. Two sets of the four constructs derived from BstNI fragments (pBT1-4) and Sau3AI fragments (pST1-4) encoded corresponding domains I-IV having long N- and C- terminal extensions, respectively.

Expression of the CANP Inhibitor Fragments Expressed in E. coli and Their Partial Purification-Eight recombinant plasmids (Fig. 1) were expressed in E. coli YA21, which showed the most stable expression among the various E. coli strains examined. Crude extracts were partially purified as described under “Experimental Procedures” by taking advan- tage of the heat and acid stabilities of the CANP inhibitor. After two purification steps, the resultant preparations showed main bands of the CANP inhibitor fragments as shown in Fig. 3. Although the preparations still contained several minor bands, presumably corresponding to contami- nating E. coli proteins, these had no protease activity.

As shown in Fig. 4, pulse-labeled preparations of the total cellular proteins showed that the same specific bands are specifically labeled, indicating that the main bands shown in Fig. 3 are encoded by the plasmids and are not modified proteolytically after synthesis.

The molecular weights of these proteins, deduced from their electrophoretic mobilities, were much larger (20-90%) than those calculated from the deduced amino acid sequences (Ta- ble I). Furthermore, an expression plasmid encoding the 77th to 682nd residues of the CANP inhibitor ( M , about 64,000) produced a specific band with an apparent molecular weight of about 105,000 on SDS-polyacrylamide gel electrophoresis (data not shown). A similar situation has been observed with the native CANP inhibitor and its fragments (18, 19,26). Our present results confirm previous results and indicate that the

2366 Four Domains of the CANP Inhibitor Are All Reactive

A PCI-213 p a - 1 1

N 4 I II 111

I I I . ~~ . ~- 1 200 400 600 a a .

I 1 I

8 1 8 2 83 84

+ + + + dJC18IHhcll puCslnlncN

+ + E. coli HBlOl

Hind Ill (pel ) or EcoRl ( ~ 8 2 - 4 ) digestion

U pB 1-4

+ Filled-In by Klenow fragment,

+ Ligation with 121 linker

U pBT 1-4

B

1 200 400 BOO 8.8.

I I I I

s 1 52 s3 5 4

+ E.coIi HBlOl

Filled-In by Klenow fragment.

9 Ligation wHh 1 2 1 linker

E.coIi YA2 1

FIG. 1. Schema used in constructing expression plasmids pBT1-4 (A) and pST1-4 (B) . Top open bar denotes the coding region of the CANP inhibitor mRNA flanked by 5’- and 3’-noncoding regions shown by horizontal bars. The regions contained in the two cDNA clones are shown above the mRNA. The thin arrow indicates the N terminus of the mature CANP inhibitor (18). I, II, III, and ZV denote the domains (shadowed) with well conserved 30 amino acid residues in the four repeating units (18). Restriction sites for BstNI ( A ) and Sau3AI ( B ) are shown below the mRNA with nucleotide (nt) sequence numbers beginning at the first residue of the initiation methionine (18); amino acid residue numbers are in parentheses. Methods for construction are described in text.

whole region of the CANP inhibitor shows anomalous electrophoretic mobilities. Thus, although posttranslational mod- ifications have been suggested as the most probable explana- tion for these anomalies (18, 19), it can be now ascribed mainly to the nature of the protein itself.

Inhibitory Activity of CANP Inhibitor Fragments Expressed in E. coli-The partially purified preparations were adjusted to about 5-20 pg/ml as estimated by the staining shown in Fig. 3, and their inhibitory activities were measured against m-CANP and p-CANP. Although a precise comparison of inhibitory activities is difficult because preparations were not completely pure and, consequently, only qualitative analyses could be performed, the results, shown in Table 11, can be summarized as follows. 1) All the expressed proteins retain inhibitory activity against both m-CANP and p-CANP. 2) Their inhibitory activities vary considerably. For example, fragments containing domain I have a higher inhibitory activity while those containing domain I11 seem to exhibit a lower activity, especially toward p-CANP. 3) Some differences in the specificity for the target protease were observed. Namely, ppST3 is a stronger inhibitor of m-CANP than ppST2 and ppST4, whereas the situation is reversed for p-CANP. 4) The inhibitory activity seems to be determined essentially by the structure of the four domains; however, the activities of ppST2

and ppST4 are considerably lower than those of ppBT2 and ppBT4. Thus, the bordering regions also make contributions to the inhibitory activity.

Further Truncation of Expression Plasmids-To confirm further the structure-function relationship of the four domains contained in common in the two series of constructions (Figs. 1 and 2), we further constructed four expression plasmids containing shorter cDNA fragments that mainly encoded the well conserved regions (Fig. 5).

The plasmid pBTl was double-digested with Sau3AI and PstI, and a fragment of 160 base pairs was recovered and subcloned into pUC18 digested with BamHI and PstI. The three plasmids, pBT2-4, were digested with Sau3A1, and fragments of 216, 201, and 126 base pairs were recovered and subcloned into pUC18 digested with BamHI. Plasmids containing the inserts in the desired direction were selected (pBS1-4), processed, and confirmed as described for pBT1-4 and pST1-4 plasmids. The amino acid sequences encoded by the final constructs (pBST1-4) are shown in Fig. 6. These four constructs encode 59-95 amino acid residues containing 41-71 residues of the CANP inhibitor sequence which retain the well conserved sequences of about 30 amino acid residues (domains I-IV).

Encoded proteins (ppBST1-4) were partially purified as


A pBT1 MITNSSSVPGSVCKACDPKKEKKSLPAAALAEPKPDEPSGKSG~DAALDDLIDTLGEPSE- 1 2 2

pBT2 MTMITPRLAAGOEKKRKVEEDAMSDOALEALSASLGTRMA- 278

pBT3 MTMITPRLAAGQAKSNEKQPKPTGKTEESKAAVPAPVAEAVPRTSMCSIOPVPPKPASLQKSTVPDDAVEALAGSLGRKE- 381

pBT4 MTWITPR~~DALSEDFSGPSSASSLKFDDAMLSAAVSEVVSOSPASITRATAPPPDTRPSNKELDDALDKLSDSLGQRO- 616

TOEDSTAYTCPEISDPMSSTVIEELGKRE LLEKKTCVACPPPDSVTPLGPDDAIDALSSDFTCSSPVASGKEAGKEAAKSAGEVLEAESAKVMRAAAPPRPAGM~A~ 271

EPELDLSSIKEVAEAKRKEEKVEKCGEDD VPAEY KPATDKDCKPLLPEPAEKPKPRSESELIDELSKDFSQ~ 1 1. 628

381

ADPEEGKPVADKIKEKSKEEEREKLGEKE IPPDY EEAKDKDGKPLLPSEPTAQLPALSEDLLL- SI$

PDPDENKPMEDKVKERAKKEHKDKLGERD IPPEY LLDQ-

B P S T l

1 sr

PST2 31!

4 5 1

599

ASCKEACKEAAKSACEVI.EAESAKVMRAAAPPPEKKRKVEEDAMSDPALEALSASLGTR~AEPE1.6PLESTCRHASSS II I

KSNEKPPKPTCKTEESKAAVPAPVAEAVPRTSMCSIPPVPPKPASLPKSTVPDDAVEAL.ACSI.GRKEADPLESTC~ 4 5 !

P S S A S S L K F D D A M L S A A V S E V V S P S P A S I T R A T A P P P D T R P S N K E L D D A I . D K I . S D S I . B P R P P D P L E ~ 5 8 8

DSCPPAAETSPATEKDKSKTTTASSSKAAKHGDKAKDSAPTTEETSKPKANEKNAS 7 1 6

FIG. 2. The amino acid sequences of the expressed proteins encoded by pBT1-4 (ppBT1-4) ( A ) and pST1-4 (ppST1-4) (B) . The sequences were confirmed by nucleotide sequencing analyses of pBT1-4 and pST1-4. Underlined residues denote regions encoded by lacZ' or polylinker of the vector or by the 12T linker. The numbers above the sequence denote the amino acid sequence numbers of the CANP inhibitor (18). Regions corresponding to domains I-IV are marked by a bar above the sequences. The most conserved hepta-residue sequences (T-I (or V)-P-P-(or A)-X-Y-R) are boxed.

1 2 3 4 ""T -

- 35.0 -

-20.1 -

-14.3-

5 6 7 8 . -

FIG. 3. SDS-polyacrylamide gel electrophoresis of ppBT1- 4 and ppST1-4. Partially purified expressed proteins (20 pl of each preparation) prepared as described under "Experimental Procedures" were electrophoresed and stained with Coomassie Brilliant Blue R250. Positions of markers are shown as molecular weights X10-3. Lcrnes 1 4 and 5-8 are E. coli extracts carrying pBT1-4 and pST1-4, respectively.

shown in Fig. 7. Similar to ppBT1-4 and ppST1-4, these preparations mainly contained expressed proteins and were devoid of protease activity.

1 2 3 4 5 6 7 8

E

D D

D-

- 35.0

- 20.1 ,

- 14.3 .

FIG. 4. Autoradiogram of pulse-labeled proteins of E. coli carrying expression plasmids pBT1-4 and pST1-4. E. coli cells carrying expression plasmids were pulse-labeled with [3H]proline for 2 min as described under "Experimental Procedures." Cell extracts were electrophoresed and processed for fluorography. The positions of the expressed labeled proteins are shown by open triangles on the left. Samples for lunes 1-8 are the same as in Fig. 3.

Their electrophoretic mobilities were again anomalous; calculated molecular weights for ppBST1-4 are about 7,900, 10,400,10,000 and 6,700, respectively, whereas those deduced from their electrophoretic mobilities are 16,000, 22,000, 22,500, and 13,500, respectively.

Inhibitory Activity of ppBSTl-4-The results summarized in Table I11 show that all four truncated fragments containing

2368 Four Domains of the CANP Inhibitor Are All Reactive TABLE I

Molecular weights of the CANP inhibitor fragments calculated from their amino acid sequences and deduced from their mobilities on

SDS-polyacrylamide gel electrophoresis Molecular weights

Calculated” (A)

18,200 13,500 16,500 14,300 15,700 17,400 17,100 15,300

Values deduced

phoresisb (B) from electro-

32,000 19,500 24,000 17,500 23,500 25,000 32,500 24,500

Ratio (B/A)

1.8 1.4 1.5 1.2 1.5 1.4 1.9 1.6

a Calculated from the amino acid sequences shown in Fig. 2.

e ppBT1-4 and ppST1-4 denote proteins encoded by pBT1-4 and Deduced from the electrophoretic mobilities shown in Fig. 3.

pST1-4, respectively.

TABLE I1 Inhibitory activities of ppBT1-4 and ppST1-4 against

m-CANP and U-CANP Remaining CANP activity”

m-CANP p-CANP Inhibitor addedb

10 u1 50 UI 10 ul 50 ul

%

20 16 65 12

28 51 15 35

>98

7 <2 43 8

<2 12 13 14

>98

% 37 8

>98 11

42 66 86 55

>98

9 6

95 10

10 28 74 35

>98 Amounts of m-CANP and p-CANP used in each assay were 0.3

and 0.5 units, respectively. Values are the result of addition of 10 and 50 p1 of inhibitor, as shown.

Partially purified inhibitor fragments as shown in Fig. 3. Concen- trations were about 5-20 pg/ml.

Control sample from E. coli YA21 carrying pUC18 treated similarly as described in the partial purification of the extract.

41-70 amino acid residues of the CANP inhibitor still retain inhibitory activity. Particularly, ppBSTl and ppBST2 retain high activity. The activity of ppBST4 is, however, much lower than that of ppBT4. This might be partly due to the lack of the C-terminal portion of domain IV (cf ppST4 in Table 11) in ppBST4. ppBST3 is also a weak inhibitor and hardly inhibits p-CANP activity, although it contains the total sequence of domain 111. This suggests that the flanking regions of domain I11 have a significant effect on inhibitory activity. A similar, although less obvious, effect of the flanking regions was also observed for the other three domains when the inhibitory activities, shown in Tables I1 and 111, are compared. On the other hand, a region between domains I and I1 (residues 210-311) did not exhibit inhibitory activity (data not shown), indicating that flanking regions of the domains do not retain inhibitory activity in of themselves. In summary, four domains of the CANP inhibitor, comprising about 30 amino acid residues, are clearly responsible for inhibition and should contain the reactive sites.

Since the four expression plasmids did not contain regions corresponding to the four repeats (Fig. 6), a direct, precise comparison of activities among ppBST1-4 is not feasible. However, the results in Table I11 strongly suggest that the

9 9 pUCl8lEamHi lPst I pUC18lEamHI

* * €.coli HE 10 1

m H i n d 111 digestion

i) pES1-4

- 9 Filled-in by Klenow fragment,

9 Ligation with 1 2 1 linker

€.coli YA21

- 12T linker 0 pEST1-4

FIG. 5. Scheme used in constructing further truncated expression plasmids of the four domains (pBST1-4). Shadowed regions denote cDNA fragments inserted into the final constructs (pBST1-4). See text for details.

inhibitory activities of the four repeating units, which are essentially determined by the primary structures of domains I-IV, are not equal when they are isolated. Fragments containing domain I11 usually exhibit lower inhibitory activity than the others.

DISCUSSION

In our previous reports (18, 19), we demonstrated that the CANP inhibitor contains four internal repeating structures which could be responsible for its multiple reactive sites and that a C-terminal fragment of the CANP inhibitor expressed in E. coli retains inhibitory activity. Here, we have extended these studies and performed further precise expression studies of cDNA fragments in E. coli. On the basis of analyses of many fragments corresponding to various regions of the CANP inhibitor, we have clearly demonstrated that all of the four repeating domains of about 30 amino acid residues independently retain inhibitory activity and that the most conserved regions in the middle of the four domains constitute the reactive sites.

All CANP inhibitor fragments showed slower mobilities on SDS-polyacrylamide gel electrophoresis than predicted from the sequence (Figs. 3, 4, and 7). Consequently, the estimated molecular weights of CANP inhibitor fragments are much larger than the actual values. This must be the reason for the discrepancy in the molecular weights noticed before (19). Similar observations have also been reported for nuclear proteins containing a number of charged residues, especially acidic residues (27, 28). Since the CANP inhibitor contains 133 Asp and Glu residues and 100 Arg and Lys residues in a total of 639 residues in the mature protein, this may be the main cause for the anomalous electrophoretic mobility. An- other possible cause may be an unusual higher order structure of the CANP inhibitor. However, circular dichroism analyses indicate that the CANP inhibitor is mostly in a random


MITNSSSVPCDPMSSTYIEELCKRE ELLEKKTCVACPPPDSVTPLCPDDAIDAI.SSDFTCSRHASSS I n ? 24?

pBSTl

pBST2 MITNSSSVPCDLSSIKEVAEAKRKEEKVEKVEKCCEDD KPATDKDGKPI.LPEPAEKPKPRSESELlDELSKDFSQTDPLESTCRHASSS R I ! 38!

pBST3 MITNSSSVPCDPEECKPVADKIKEKSKEEEREKLGEKE EEAKDKDCKPLLPSEPTAQLPALSEDLLLLTDPLESTCRHASSS 45! 51 6

588 & I 6 2 8 pBST4 MITNSSSVPCDPDENKPMEDKVKERAKKEHKDKLCERD IPPEYR1I.LDQTCRHASSS

FIG. 6. The amino acid sequences of the expressed proteins encoded by pBST1-4 (ppBST1-4). The sequences were confirmed by nucleotide sequencing analyses of pBST1-4. Underlined residues denote those encoded by l0C.Z’ or polylinker of the vector or by the 12T linker. The numbers above the sequence denote the amino acid sequence numbers of the CANP inhibitor (18). Four well conserved regions are marked by a bar above the sequence. The most conserved sequences (T-I(or VI-P-P(or A)-X-Y-R) are boxed.

1 2 3 4

- 35.0

-20.1

-14.3

FIG. 7. SDS-polyacrylamide gel electrophoresis of pp- BST1-4. Partially purified expressed proteins (20 pl of each preparation) prepared as described under “Experimental Procedures” were electrophoresed. For other conditions, see legend to Fig. 3. Samples are pBST1-4 for lanes 1-4, respectively.

TABLE I11 Inhibitory activities of ppBSTl-4 against m-CANP and p-CANP

Remaining CANP activities Inhibitor added’ m-CANP p-CANP

10 ul 50 VI 10 ul 50 ul

% %

ppBSTl 19 15 23 21 ppBST2 22 12 23 20 ppBST3 >98 93 >98 >98 ppBST4 84 68 63 57

Partially purified inhibitor fragments as shown in Fig. 7. Protein concentrations were 5-20 pg/ml. For other conditions, see Table 11.

conformation, and abnormal structures have not yet been detected.2

All four repeating domains independently exhibited inhibitory activity, although the intensities of inhibition were different among domains. Although a precise comparison of inhibitory activities must await further kinetic studies, domain I11 seems to exhibit the lowest activity, especially against p-CANP, and domain I possesses the highest activity. Com- parison of four conserved regions suggests that domain I11 is the most typical, because its sequence seems to show the greatest sequence homology among the four domains (Ref. 18

Y. Emori, H. Kawasaki, S. Imajoh, Y. Minami, and K. Suzuki, unpublished results.

and see, for example, Fig. 2). On the other hand, domain I is considerably different from the other three but exhibits the highest activity. The cause for the different intensities of inhibition will be elucidated by analyses of the higher order structure of the CANP inhibitor and by exchanging the amino acid residues by means of site-directed mutagenesis experiments.

In the present study, we have developed small proteins of 59-95 amino acid residues which still retain inhibitory activity toward CANP. These small fragments are extremely useful for kinetic and structural studies to elucidate the inhibition mechanism and for clarification of the biological function of CANP. Muscle atrophy in muscular dystrophy has been ascribed to an increase in CANP activity (3). Low molecular weight inhibitors specific for CANP may also be useful as drugs to inhibit CANP activity specifically in various CANP- related diseases, such as muscular dystrophy.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16. 17.

18.

19.

REFERENCES

Imahori, K. (1982) in Calcium and Cell Function (Cheung, W. Y., ed) Vol. 3, pp. 473-485, Academic Press, Orlando, FL

Murachi, T. (1983) in Calcium and Cell Function (Cheung, W. Y., ed) Vol. 4, pp. 377-410, Academic Press, Orlando, FL

Kay, J. (1983) in Proteases: Potential Role in Health and Disease (Horl, W. H., and Heidland, A., eds) pp. 519-532, Plenum Publishing Corp., New York

Suzuki, K., Imajoh, S., Emori, Y., Kawasaki, H., Minami, Y., and Ohno, S. (1987) FEBS Lett. 2 2 0 , 271-277

Kawashima, S., Nomoto, M., Hayashi, M., Inomata, M., Naka- mura, M., and Imahori, K. (1984) J. Biochem. (Tokyo) 95,95- 101

Yumoto, N., Kikuchi, T., Sasaki, T., and Murachi, T. (1984) J. Biochem. (Tokyo) 96,1531-1537

Emori, Y., Kawasaki, H., Sugihara, H., Imajoh, S., Kawashima, S., and Suzuki, K. (1986) J. Biol. Chem. 261,9465-9471

Kawasaki, H., Imajoh, S., Kawashima, S., Hayashi, H., and Suzuki, K. (1986) J. Biochem. (Tokyo) 9 9 , 1525-1532

Emori, Y., Kawasaki, H., Imajoh, S., Kawashima, S., and Suzuki, K. (1986) J. Biol. Chem. 261,9472-9476

Ohno, S., Emori, Y., Imajoh, S., Kawasaki, H., Kisaragi, M., and Suzuki, K. (1984) Nature 312,566-570

Aoki, K., Imajoh, S., Ohno, S., Emori, Y., Koike, M., Kosaki, G., and Suzuki, K. (1986) FEBS Lett. 205,313-317

Imajoh, S., Kawasaki, H., and Suzuki, K. (1986) J. Biochem.

Imajoh, S., Kawasaki, H., Kisaragi, M., Mukai, M., Sugita, H.,

Takano, E., Kitahara, A., Sasaki, T., Kannagi, R., and Murachi,

Imajoh, S., Kawasaki, H., Emori, Y., and Suzuki, K. (1987)

Imajoh, S., and Suzuki, K. (1985) FEBS Lett. 187,47-50 Cottin, P., Vidalenc, P. L., Merdaci, N., and Ducastaing, A. (1983)

Emori, Y., Kawasaki, H., Imajoh, S., Imahori, K., and Suzuki, K.

Imajoh, S., Kawasaki, H., Emori, Y., Ishiura, S., Minami, Y.,

(Tokyo) 100, 633-642

and Suzuki, K. (1984) Biomed. Res. 5,481-488

T. (1986) Biochem. J. 2 3 5 , 97-102

Biochem. Biophys. Res. Commun. 146,460-467

Biochim. Biophys. Acta 743,299-302

(1987) Proc. Natl. Acad. Sci. U. S. A. 84. 3590-3594

2370 Four Domains of the CANP Inhibitor Are All Reactive

Sugita, H., Imahori, K., and Suzuki, K. (1987) FEBS Lett. 215, 274-278

20. Messing, J. (1983) Methods Enzymol. 101, 20-78 21. Yanisch-Perron, C., Vieira, J., and Messing, J. (1985) Gene

22. Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

(Amst.) 33, 103-119

23. Laemmli, U. K. (1970) Nature 227, 680-685 24. Inomata, M., Nomoto, M., Hayashi, M., Nakamura, M., Imahori,

K., and Kawashima, S. (1984) J. Biochem. (Tokyo) 95, 1661- 1670

25. Hattori, M., and Sakaki, Y . (1986) Anal. Biochem. 152, 232-238 26. Maki, M., Takano, E., Mori, H., Kannagi, R., Murachi, T., and

Hatanaka, M. (1987) Biochem. Biophys. Res. Commun. 143,

27. Kleinshmidt, J. A., Dingwall, C., Maier, G., and Franke, W. W.

28. Dingwall, C., Dilworth, S. M., Black, S. J., Kearsey, S. E., Cox,

300-308

(1986) EMBO J. 5, 3547-3552

L. S., and Laskey, R. A. (1987) EMBO J. 6,69-74

four repeating domains of the endogenous inhibitor for calcium

Documents