Pergamon 0 0 0 3 - 9 9 6 9 ( 9 5 ) 0 0 0 1 6 - X

Archs oral Biol. Vol. 40, No. 7, pp. 639-644, 1995 Copyright © 1995 Elsevier Science Ltd

Printed in Great Britain. All rights reserved 0003-9969/95 $9.50 + 0.00

EXPRESSION OF A FUNCTIONAL RAT SALIVARY CYSTATIN S POLYPEPTIDE IN ESCHERICHIA COLI

ASHU SHARMA, l BRIAN C. O 'CONNELL, 2 L A W R E N C E A. TAB AK 3 and G U R R I N D E R S. BEDI 4.

~Department of Oral Biology, State University of New York at Buffalo, Buffalo, NY, zCIPCB, National Institute of Dental Research, Bethesda, MD, 3Department of Biochemistry and Dental Research, University of Rochester, NY and 4Department of Microbiology and Immunology, The Medical College

of Pennsylvania, Philadelphia, PA 19129, U.S.A.

(Accepted 3 January 1995)

Summary~ystat in S is a cysteine proteinase inhibitor that is transiently expressed during rat sub- mandibular gland development and can be induced by isoproterenol in the adult. A cDNA for rat cystatin S which included the entire coding sequence of the secreted cystatin was cloned. A coding region of the cystatin gene was amplified by polymerase chain reaction and cloned into the pGEX-2T expression vector. The chimeric plasmid was transformed into Escherichia coli, and protein expression was induced by isopropyl-fl-D-thiogalactopyranoside. The expressed protein was purified from insoluble inclusion bodies after solubilization with urea and fast protein liquid chromotography on a MonoQ column. The purified recombinant cystatin reacted with antibodies to cystatin S purified from rat submandibular glands and showed an amino-terminal amino acid sequence identical to that of rat cystatin S. The recombinant protein exhibited papain inhibition activity comparable to natural cystatin. This was a successful expression and purification of a functionally and immunologically reactive recombinant cystatin from E. coli, an approach which will be used later towards generating recombinant variants to study the binding and functional domains of thi~ cysteine protease inhibitor.

Key words: cystatin, cystein~proteinase inhibitor, expression, submandibular, salivary.

INTRODUCTION

The cystatin superfamily includes a number of cys- teine proteinase inhibitors that are widely distributed throughout mammalian tissues (Barrett et al., 1986). There are three families of cystatins, classified accord- ing to the complexity of their structure. An inducible cysteine proteinase inhibitor belonging to cystatin family 2 has been described in rat submandibular gland. Originally named LM protein (Menaker et al., 1974; Naito, 1981), rat salivary cystatin (cystatin S) has been isolated and sequenced from the sub- mandibular glands of adult rats exposed chronically to the fl-adrenergic agonist isoproterenol (Bedi, 1989a, b). The cDNA for rat salivary cystatin has been cloned and sequenced (Shaw et al., 1988), and more recently the gene for rat cystatin S has also been isolated and sequenced (Cox and Shaw, 1992). Cystatin S was shown to be transiently expressed in rat submandibular glands during acinar development and the m R N A was localized to submandibular acinar cells by in situ hybridization (Shaw et al., 1988; 1990).

*To whom all correspo:adence should be addressed. Abbreviations: BAPNA, BZ-DL-Arg-NPhNO2; IPTG, iso-

propyl-fl-D-thiogalactopyranoside; SDS-PAGE, sodium dodecyl sulphate-polyacrylamide gel electi'ophoresis; TBS, tris-buffered saline.

Limited information is available about the structural requirements for the protease-inhibitor interaction of cystatins in general and salivary cystatin in particular. Recombinant DNA techniques are now being used to understand the molecular mechanism of inhibition of cysteine proteinases by their protein inhibitors (Machleidt et al., 1991). In order to begin understand- ing the mechanism of inhibition by cystatin S, it is important to generate recombinant variants lacking one or more active binding domains. We now present a report of the cloning of a cDNA that codes for the entire rat cystatin S, and describe the expression and purification of a biologically active recombinant cystatin in Escherichia coli.

MATERIALS AND METHODS

c D N A library screening fo r cystatin

2 gt 11 cDNA libraries were available from previous studies, constructed from the poly(A + ) R N A derived from submandibular glands of both control and isoproterenoi-treated rats (Cooper, Elia and Tabak, 1991). Clones expressing cystatin S were identified immunologically with a rabbit antiserum (Bedi, 1989a). In brief, competent Y1090 (Promega) cells were infected with a port ion from each cDNA library representing 105 plaque-forming units. IPTG-soaked nitrocellulose filters were placed on the plaques as soon as they became visible and were incubated overnight.

639

640 Ashu Sharma et al.

The filters were rinsed in TBS (50 mM tris, pH 7.5, 150 mM NaC1), blocked with 0.5% gelatin/TBS, and added to a 1:200 dilution of antiserum. The anti- serum had been presorbed against a Y1090 cell lysate. The filters were rinsed and incubated with 150 ng of [~25I]protein A (>30#Ci /#g , ICN) for 4h. Finally, the filters were rinsed, dried and placed on film.

Plaques that appeared positive on the autoradio- gram were located, picked from the plates and stored in 0.5 ml of buffered medium (20mM tris-HC1, pH 7.6, 10mM MgSO4, 100mM NaCI, 0.01% gelatin) with 5/~1 of chloroform. A portion of the plaque supernatant was used to infect cells. These clones were screened as previously described. Positive plaques after the second screening were picked, lysates prepared and DNA extracted for restriction analysis (Sambrook, Fritsch and Maniatis, 1989).

Amplification o f eystatin cDNA

Oligonucleotide primers with sequences 5'-AAG- CTTGACTCCTGGAGCCCG and 5 '-CTGCAGT- A A T G G T A G C G A C were based on the sequence flanking the EcoRI cloning site in 2 g t l l (Moran et al., 1990). Restriction sites at the 5' ends of the primers for HindIII and PstI (underlined) were incorporated to facilitate subcloning of the product. An Applied Biosystems model 391 DNA Synthesizer was used to make the primers.

Amplification was by polymerase chain reaction for 30 cycles in a DNA thermal cycler (Perkin Elmer Cetus, Norwalk, CN). The cystatin plasmid clone 13.3.2, described in Results, was used as a template and with the following variables: denaturation, 94°C, 2 min; annealing, 45°C, 2 min; and primer extension, 70°C, 3 min. The product was digested with PstI and HindIII, then ligated into pBluescript (designated as pBSCys2).

Construction o f eystatin-expression chimeric plasmid

For construction of the cystatin-expression plasmid, the coding region of the cystatin S was amplified from the pBSCys2 plasmid clone by polymerase chain reaction using a forward (sense) primer based on nucleotide residues 125-141 of the rat cystatin cDNA and a T3 promoter primer of the vector as a reverse primer. The forward primer incorporated a Barn HI restriction site for in-frame cloning into the Barn HI site of the pGEX-2T expression vector. The primer sequences (with restriction site underlined) are as follows: forward primer, GCGCGGATCCGGTCA- CTTTCTGGGTGG; reverse primer, ATTAACCC- TCACTAAAG. The 5' end of the forward primer also contained a 4-bp overhang to facilitate digestion. The amplified DNA was 'end-polished' by the Klenow fragment of DNA polymerase according to standard procedures (Sambrook et al., 1989) and was then digested with BamHI (Promega Corp.). The DNA fragment was purified from agarose gel by the 'Gene- Clean' procedure (Biol01 Inc., La Jolla, CA). The purified DNA was ligated into a pGEX-2T expression vector (Pharmacia Inc.), which was linearized by restriction enzymes BamHI and Sinai.

Selection o f recombinants and cystatin expression

The ligated chimeri c plasmid was transformed into E. coli XL Blue (Strategene Inc.) by electroporation

(Gene Pulsar; Bio-Rad, Richmond, CA) and the colonies were selected by plating on LB agar plates containing 50#g of ampicillin/ml. Plasmid DNA was isolated according to the method of Birnboim and Doly (1979). DNA sequencing was done on the double-stranded plasmid DNA with the Sequenase version 2.0 sequencing kit (U.S. Biochemicals, Cleveland, OH). A bacterial colony containing the recombinant plasmid, pGEXCys, with the correct in-frame coding sequence was used for protein expression. In brief, a bacterial colony was grown in 3 ml LB broth containing 50/~g ampicillin/ml until the optical density at 590 nm reached 0.6. IPTG was then added to a final concentration of 1 mM, and the cultures incubated for another 1 h to induce protein synthesis. The cells were harvested by centrifugation, and the cell pellet was boiled in 50 p l of Laemmli (1970) buffer for 10 min and electrophoresed on 12.5% polyacrylamide gels as described by Bedi (1989a).

Nucleic acid sequencing

DNA was sequenced by the dideoxy chain termination method using a Sequenase 2.0 kit. The T3 and T7 primers (Strategene) were used to sequence the pBSCys2 clone. Oligonucleotide primers based on the internal region of the cystatin cDNA fragments were utilized for sequencing the pGEXCys plasmid DNA. The sequencing reactions were electrophoresed on 6% acrylamide/TBE gels containing 8 M urea. Gels were fixed in 10% acetic acid/10% methanol, dried on to Whatman's 3-mm paper and autoradiographed.

Purification o f recombinant cystatin

Recombinant fusion proteins were obtained from 1-1 cultures. The overnight culture of transformed E. eoli (50 ml) was inoculated into 1 I of fresh LB/amp broth. The culture was incubated for 4 h at 37°C with vigorous shaking, IPTG was then added to a final concentration of 1 mM, and the culture was incu- bated for a further 1.5 h. The cells were harvested by centrifugation at 500g at 4°C and recombinant cystatin was purified from insoluble inclusion bodies by the procedure of Marston (1987). In brief, the cells were lysed with lysozyme followed by treatment with deoxycholic acid. Contaminating nucleic acids were removed by DNase I. The insoluble material was washed once with 2 M urea and finally solubilized in lysis buffer (50 mM tris-HCl, pH 8.0, 1 mM EDTA, 100 mM NaCI) containing 8 M urea. The solubilized components from the inclusion bodies were dialysed stepwise against decreasing concentrations of urea (starting with 8, 6, 4 and 2 M of urea) in lysis buffer to allow for refolding of the denatured protein. Finally, the sample was dialysed against 2 M urea and 2.5% fl mercaptoethanol, and centrifuged at 10,000g for 15 min. The fusion protein was finally dialysed against 0.01 M tris-HC1, pH 8.0, and digested with 100 U of thrombin for 6 h at 37°C.

Purification of recombinant cystatin was achieved by ion-exchange chromatography on MonoQ using a fast protein liquid chromatography system. The thrombin digest from the above step was applied to a MonoQ (30 x 1 cm) column equilibrated with 0.05 M tris-HCl buffer, pH 8.0. After washing the column with two volumes of starting buffer, the elution was accomplished by a linear gradient of 0-1 M NaC1 in

641

kDa

0.05 M tris-HCl buffer, pH 8.0. Fractions of 6 ml were collected at a flow rate of 3 ml/min and assayed for protease inhibition activity and for protein estim- ation at 280 nm. Portions of selected fractions were analysed by SDS-PAGE.

S D S - P A G E and immunoblot analysis

SDS-PAGE was by the method of Laemmli (1970), using a 10% polyacrylamide gel in a mini-gel system (SE 280 vertical gel unit, Hoefer). For immunoblot analysis, the proteins separated by SDS-PAGE were transferred to nitrocellulose membranes by using the TE 22 Mighty Small transphor unit (Hoefer) in 25 mM tris-glycine buffer, pH 8.3, containing 15% methanol (Towbin, Staehlin and Gordon, 1979). The unoccupied binding sites on the membrane were blocked by incubating for 1 h with 20 mM tris-HCl buffer, pH 8.0, containing 2% bovine serum albumin. The membranes were then incubated with a 1:200 dilution of antibodies overnight at 4°C. The mem- branes were washed and treated with 1:2000 diluted goat anti-rabbit IgG conjugated with horseradish per- oxidase for 1 h. After washing, the bound antibodies were visualized by addition of the 4-chloro-1-naphthol coiour-developing rezgent. The reaction was stopped by immersing the membranes in distilled water.

Automated Edman degradation

The amino acid se£uence of the purified recombin- ant cystatin was determined by automated stepwise sequencing on an Applied Biosystems model 477A gas-phase sequencer with an on-line model 120A PTH analyser (Bedi, 1989b).

Enzyme inhibition assay

Chromogenic substrate BAPNA was used for enzyme inhibition assays. The activity of papain was assayed by a slight :modification of the method of Barrett (1976). The enzyme was preincubated for 10min with an eque,l volume (50pl) of activation buffer (0.1 M phosphate buffer, pH 6.0 containing 0.2 mg/ml dithiothreitol and 0.5 mg/ml Na 2EDTA) with or without cystatin. The reaction was started by adding 25/~1 of 10mM solution of BAPNA and stopped by the addition of 25 #1 glacial acetic acid. The amount of liberated p-nitroaniline was deter- mined at 410 nm. Tlz~e amount of enzyme used was adjusted to obtain a final optical density of between 0.5 and 0.7 in positive controls containing no inhibitors. Rat submandibular gland kallikrein (Bedi, 1992) and cysteine proteinase from Porphyromonas gingivalis (Bedi and Williams, ;(994) were assayed as described previously.

. . . . <.T..T. - - - - - - - - - ---- - - ---- - - - - - - . . . . . . . . .

C 2 " ~. . . . . . . . . . . . . . . . * C1

N ORF I I I

1 100 200 300 400 500 600 700

Fig. l. Schematic representation of the two cDNA frag- ments (C 1 and C2), from which the sequence of rat cystatin S was derived. The dashed lines indicate the produce of at least two sequencing reactions. The proposed open reading frame (ORF) is shown as open box and the N-terminus of

the mature protein is marked as N.

Fig. 1, extending from the Eco RI site at 389 bp to the polyadenylated tail at + 707.

As the full-sized insert was not recovered from the clone 13.3.2 with EcoRI, a cDNA was amplified by polymerase chain reaction. From primers flanking the EcoRI cloning site of 2gt l l , a 536-bp cDNA frag- ment, C2, was generated (Fig. 1). The entire 707-bp cDNA sequence deduced from fragments C1 and C2 is the same as that reported by Shaw et al. (1988), though there are two conservative substitutions in the eDNA at + 205 and + 208 (not shown). The sequence we found contains an additional T at position + 121 and also extends 5' by 64 bp more than previously described. A thymidine residue is substituted for cytidine at + 105. The eDNA fragment C2 included

1

Expression of rat salivary cystatin

97.4 I

66.2 |

42 .7 ! --. Fusion protein

31.0

G S T

RESULTS

Initial screening of the cDNA library from saline- treated rats using antibody to cystatin did not reveal any positive plaques. The library from isoproterenol- treated animals yielded 35 potential positives. Twelve of these plaques were selected to be replated. After secondary screening, 12 plaques were picked and the clone 13.3.2 was found to contain the largest insert. EcoRI digestion of tile phage released a 316-bp frag- ment that was ligated to the plasmid pBluescript tm (Strategene). This fragment (CI) is represented in

21.5 1 14.4

A B C Fig. 2. SDS-PAGE o f proteins expressed in E. coil Lanes A and B, extracts of cells from two separate colonies containing pGEXCys clone; lane C, extract of ceils contain- ing pGEX-2T vector (control) (10 #1 of sample was loaded

in each lane).

642 Ashu Sharma et al.

( A ) k D a

( B )

1 4 2 . 9 -

9 7 . 2 -

5 0 . 0 -

3 5 . 1 -

2 9 . 7 -

2 1 . 9 -

!

1 2 3 4 1 2 3 4

Fig. 3. SDS-PAGE (A) and Western blot (B) analysis of the expressed cystatin protein. Lane 1, recombinant fusion protein extracted from inclusion bodies; lane 2, thrombin digest of fusion protein; lane 3, purified cystatin from recombinant fusion protein; lane 4, purified cystatin S from rat submandibular

glands.

the entire coding sequence of the secreted cystatin (Bedi, 1989b).

Expression of cystatin cDNA in E. coli

A DNA fragment encoding cystatin polypeptide was amplified from the cystatin cDNA fragment C2 by polymerase chain reaction, and used to construct the cystatin expression vector. A DNA fragment of approx. 475 bp, the predicted size based on the primers, was obtained. Direct DNA sequencing of the amplified insert agreed with sequence of cloned cystatin cDNA.

To keep the cystatin cDNA fragment in a correct reading frame, the amplified fragment encoding the full-length secreted cystatin polypeptide was treated with Klenow and digested with Bam HI. The resulting DNA fragment with a 5' cohesive and 3' blunt end was subsequently cloned into the BamHI and SmaI site of pGEX-2T. Extracts of E. coli containing the pGEXCys plasmid (Fig. 2) showed a prominent band at 40 kDa (lanes A and B shown for two independent colonies) when compared to E. coli transformed with pGEX-2T alone (lane C). Triton-X 100 (1%) or 2 M urea were not able to solubilize the expressed fusion protein from the bacterial pellet, indicating that the protein was present as insoluble inclusion bodies. The 40-kDa expressed fusion protein reacted with antibodies to cystatin (Fig. 3B, lane 1).

Purification and characterization of recombinant eystatin

Recombinant cystatin was purified from insoluble inclusion bodies after cleavage with thrombin and

fractionation by fast protein liquid chromatography on an anion exchanger MonoQ column. When the fractions were analysed by SDS-PAGE, fractions 45-51, eluting at 300mM NaC1 salt concentration, contained cystatin polypeptide. The protein peak containing cystatin polypeptide was dialysed against water and lyophilized. The purified protein migrated at an apparent molecular mass of 14 kDa and reacted with antibodies to cystatin (Fig. 3).

The sequence of the amino-terminal 11 residues of the immunoreactive 14-kDa protein determined by

100 -o o c .,~ s o

,_~ s o >

40

• 2 0 u o

a . 0

= Cystatin .

tatln --~

I I ~ I I I

0 5 10 15 20 25

C o n c e n t r a t i o n (IJg)

Fig. 4. Inhibition of cysteine protease papain by native and recombinant (rec) cystatin. The enzyme inhibition curves represent the percentage papain activity retained in the presence of varying amounts of native and recombinant

cystatin.

Expression of rat

solid-phase microsequencing was GSGHFLGGIE K . This sequence is identical to cystatin purified from rat submandibular gland:~ except that it contained two additional residues, gly and ser, at the N-terminal. The recombinant cystatin polypeptide was tested for its ability to inhibit papain activity; its inhibitory effect was similar to that of the purified native cystatin (Fig. 4). The inhibitor did not show any inhibition of serine p~coteinases trypsin and rat sub- mandibular gland kallikrein and cysteine proteinase from P. gingivalis.

DISCUSSION

Although cystatins may be found in a number of tissues, including skin, liver, plasma, granulocytes and human salivary glands, the only example of an inducible cysteine proteinase inhibitor is rat salivary cystatin (Bedi, 1991). "['he amino acid sequence for rat cystatin S is very similar to that of other family 2 cystatins, such as human cystatins C, SA-I and SN, chicken cystatin and bovine cystatin. The cDNA in this report encodes for a mature protein which is the same as a variant described by Bedi (1989b) using amino acid sequencing. The cDNA has a putative translational start site at +44, which conforms to the initiation consensus sequence PuXXATGG (Kozak, 1981). Conceptual translation indicated there is a 27-amino acid signal peptide containing mainly hydrophobic residue,,', in its central region. This organization is similar to that proposed for human cystatins C, SA and SIN, which have signal peptides of 26, 24 and 23 residues, respectively, as deduced from their nucleotide sequences (Saitoh et al., 1987). Hence, the gene for rat cystatin S may be related to that of the human cystatins. Signal sequences from the cystatin family are not homologous to those of proline-rich proteins or glutamine/glutamic acid-rich proteins, though these transcripts are also induced by isoproterenol (Cooper et al., 1991; Carlson, 1993). The difference between the signal sequences of cystatin S and glutamine/glutamic acid-rich proteins (both acinar products) suggests that a specific motif is not required for expression in this cell type.

Cystatins are pote.nt competitive inhibitors of cysteine proteinases, with very high affinities for some of the target enzymes. The limited information available about the binding sites of cystatins with cysteine-proteinases is ,;peculative, based on the known sequence of the conserved sites present in various members of cystatin superfamily (Barrett, 1987). The X-ray crystal structure of recombinant human stefin B in complex with papain further confirms that the conserved residues form a tripartite wedge, which slots into the papain active site (Stubbs et al., 1990). Currently, recombinant DNA techniques are being applied in studies of the molecular mechanism of inhibition of cysteine proteinases by cystatins (Genenger et al., 1991; Machleidt et al., 1991; Hall et al., 1993). Over the past decade, genes for human cystatin C (Abrahamson et al., 1988; Strauss et al., 1988a), human salivary cystatin (Bobek, Wang and Levine, 1993), stefin A (Strauss et al., 1988b), rat cystatin ct (Katunuma et al., 1988), and stefin B (Jerala et al., 1988; Th:iele et al., 1988) were produced by recombinant DNA methods. In this study we

salivary cystatin 643

have successfully constructed a recombinant plasmid encoding the rat salivary cystatin S molecule and efficiently expressed this in E. coli. The papain- inhibiting activity of E. co l i -p roduced cystatin S was very similar to that of native purified rat salivary cystatin and showed immunological reactivity with antisera prepared against purified cystatin. The authenticity of the purified protein was further confirmed by the finding that the amino-terminal amino acid sequence of the thrombin cleavage product was identical to the amino acid sequence of rat submandibular gland cystatin (Bedi, 1989b). As expected, the purified protein contained two additional amino acid residues, gly and ser, generated from the cleavage of fusion protein by thrombin, but these additional residues did not alter either the immuno- logical or functional characteristics of the purified protein. The expression system described here will be useful for producing variants of cystatin S for structure/function studies.

Acknowledgements--This work was supported in part by National Institutes of Health Grant DE-09690. We thank Christine Hamilton for typing this manuscript.

REFERENCES

Abrahamson M., Dalboge H., Olafsson I., Carlsen S. and Grubb A. (1988) Efficient production of native, bio- logically active human cystatin S by Escherichia coli. FEBS Lett. 236, 14-18.

Barrett A. J. (1976) An improved color reagent for use in Barrett's assay of cathepsinB. Analyt. Biochem. 76, 374 376.

Barrett A. J. (1987) The cystatins: a new class of peptidase inhibitors. Trends Biochem. Sci. 12, 193-196.

Barrett A. J., Rawlings N. D., Davies M. E., Machleidt W., Salvesen G. and Turk V. (1986) Cysteine proteinase inhibitors of the cystatin superfamily. In Proteinase Inhibitors. Research Monographs in Cell and Tissue Physi- ology (Eds Barrett A. J. and Salvesen G.), pp. 515 569. Elsevier, Amsterdam.

Bedi G. S. (1989a) Purification and characterization of an inducible cysteine proteinase inhibitor from submandibular glands of isoproterenol-treated rats. Arch. Biochem. Biophys. 270, 335-343.

• Bedi G. S. (1989b) Amino acid sequence of an inducible cysteine proteinase inhibitor (cystatin) from submandibular glands of isoproterenol-treated rats. Arch. Biochem. Biophys. 273, 245553.

Bedi G. S. (1991) The effect of adrenergic agonists and antagonists on cysteine proteinase inhibitor (cystatin) in rat saliva. Archs oral Biol. 36, 611 618.

Bedi G. S. (1992) Purification and characterization of kallikrein-like serine proteases from rat submandibular glands. Prep. Biochem. 22, 67-81.

Bedi G. S. and Williams T. (1994) Purification and characterization of a collagen-degrading protease from Porphyromonas gingivalis. J. biol. Chem. 269, 599- 606.

Birnboim H. C. and Doly J. (1979) A rapid alkaline lysis extraction procedure for screening recombinant plasmid DNA. Nucleic Acid Res. 7, 1513-1523.

Bobek L., Wang X. and Levine M. J. (1993) Efficient production of biologically active human salivary cystatin in Escherichia coli. Gene 123, 203-210.

Carlson D. M. (1993) Salivary proline-rich proteins: bio- chemistry, molecular biology, and regulation of expression. Crit. Rev. oral Biol. Med. 4, 495-502.

644 Ashu Sharma et al.

Cooper L. F., Elia D. M. and Tabak L. A. (1991) Secretagogue-coupled changes in the expression of glut- amine/glutamic acid-rich proteins (GRPs). Isoproterenol induced changes in GRP transcript expression and changes in isoforms secreted. J. biol. Chem. 266, 3532 3539.

Cox J. L. and Shaw P. A. (1992) Structure, organization and regulation of a rat cysteine proteinase inhibitor-encoding gene. Gene 110, 175 180.

Genenger G., Lenzen S., Mentele R., Assfalg-Machleidt I. and Auerswald E. A. (1991) Recombinant Q53E- and Q53N-chicken egg white cystatin variants inhibit papain, actinidin and cathepsin B. Biomedica Biochimica Acta 50, 621-625.

Hall A., Dalbage H., Grubb A. and Abrahamson M. (1993) Importance of the evolutionarily conserved glycine residue in the N-terminal region of human cystatin C (Gly-ll) for cysteine endopeptidase inhibition. Biochem. J. 291, 123 129.

Jerala R., Trstenjak M., Lenarcic B. and Turk V. (1988) Cloning a synthetic gene for human stefin B and its expression in E. coli. FEBS Lett. 239, 41-48.

Kozak M. (1981) Possible role of flanking nucleotides in recognition of the AUG initiator codon by eukaryotic ribosomes. Nucleic Acid Res. 9, 5233 5262.

Katunuma N., Yamato M., Kominami E. and Ike Y. (1988) Total synthesis of the cystatin alpha gene and it's expression in E. coli. FEBS Lett. 238, 116 118.

Laemmli U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London)227, 680-685.

Machleidt W., Thiele U., Assfalg-Machleidt I., Forger D. and Auerswald E. A. (1991) Molecular mechanism of inhibition of cysteine proteinases by their protein inhibitors: kinetic studies with natural and recombinant variants of cystatins and stefins. Biomedica Biochimica Acta 50, 613-620.

Marston F. A. O. (1987) The purification of eukaryotic polypeptides expressed in E. coil DNA Cloning: A Practical Approach (Ed. Clover D. M.), pp. 59 88. IRL Press, Oxford.

Menaker L., Sheetz J. H., Cobb C. M. and Navia J. M. (1974) Gel electrophoresis of whole saliva and associated histologic changes in submandibular glands of isoproterenol treated rats. Lab. Invest. 30, 341-349.

Moran L. S., Maina C. V., Poole C. B. and Slatko B. E. (1990) Nucleotide sequence of the phage lambda gt l l SacI-KpnI lac Z region. Gene 93, 163-164.

Naito Y. (1981) Studies on LM protein appearing in submandibular glands of isoproterenol-treated rats. I. Purification and characterization of its appearance. Chem. Pharm. Bull. 29, 1365-1372.

Saitoh E., Kim H-S., Smithies O. and Maeda N. (1987) Human cysteine proteinase inhibitors: Nucleotide sequence analysis of three members of the cystatin gene family. Gene 61, 329 338.

Sambrook J., Fritsch E. and Maniatis T. (1989) Molecular Cloning, a Laboratory Manual. Cold Spring Harbor Laboratory Press.

Shaw P. A., Barka T., Woodin A., Schachter B. S. and Cox J. L. (1990) Expression and induction by fl-adren- ergic agonists of the cystatin S gene in submandib- ular glands of developing rats. Biochem. J. 265, 115 120.

Shaw P. A., Cox J. L., Barka T. and Naito Y. (1988) Cloning and sequencing of cDNA encoding a rat salivary cysteine proteinase inhibitor inducible by fl-adrenergic agonists. J. biol. Chem. 263, 18,133-18,137.

Strauss M., Bartsch F. O., Stollwerk J., Trstenjak M., B6hning A., Gassen H. G., Machleidt W. and Turk V. (1988a) Chemical synthesis of a gene for human cystatin C and its expression in E. coli. Biol. Chem. Hoppe-Seyler 369, 209 218.

Strauss M., Stollwerk J., Lenarcic B., Turk V., Jany K. D. and Gassen H. G. (1988b) Chemical synthesis of a gene for human stefin A and its expression in E. co6. Biol. Chem. Hoppe-Seyler 369, 1019 1030.

Stubbs M. T., Laber B., Bode W., Huber R., Jerala R., Lenarcic B. and Turk V. (1990) The refined 2-4A x-ray crystal structure of recombinant human steffin B in complex with the cysteine proteinase papain: a novel type of proteinase inhibitor interaction. EMBO J. 9, 1939 1947.

Thiele U., Auerswald E. A., Gebhard W., Assfalg- Machleidt I., Popovic T. and Machleidt W. (1988) lnhibitorily active recombinant human stefin B. Gene synthesis, expression and isolation of an inhibitory active MS-2 pol-stefin B fusion protein and preparation of Des[Metl,2(2)] stefin B. Biol. Chem. Hoppe-Seyler 369, 1167 1178.

Towbin H., Staehlin T. and Gordon J. (1979) Electro- phoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some application. Proc. natn. Acad. Sci. U.S.A. 76, 4350 4354.