ANALYTICAL
BIOCHEMISTRY
191,
235-241
(1990)
Molecular Cloning of Proteinase-Encoding Genes from Cancer Cells by Zymogen Assay and Direct Sequencing Nabil
G. Hagag,’
Michele
Division of Oncology, Health and Veterans Administration
Received
June
Kirchner,
and John
F. DiStefano
Sciences Center, State University of New York at Stony Brook, New York 11794, Medical Center, Northport, New York 11768
4, 1990
A method that permits the in vitro cloning and identification of proteolytic enzyme genes from cDNA expression libraries is described. The method can detect positive proteinase genes within 30 min following the transfer of plaques to nitrocellulose membrane filters. The method is based on the functional expression of fusion lot Z-proteinase protein in X gt 11 infected Y 1090 bacteria and does not require prior knowledge of either the sequence of the cDNA insert or a monoclonal antibody to its encoded antigen. This strategy when coupled with polymerase chain reaction of the cDNA insert using lac Z primer sequences that are flanking the EcoRl cloning site in gtl 1 phage permits direct sequencing of the amplified DNA. With this method we have isolated 10 genes expressing protease activity in the human small-cell carcinoma of the lung. The same procedure could be applied to isolate unknown proteinases from cDNA libraries of virtually any eukaryotic cell. o 1990 Academic
Press,
Inc.
Proteolytic enzymes have a variety of biological functions ranging from extracellular and intracellular catabolism of peptides and proteins (1) to precise processing of precursor proteins, such as prohormones and proenzymes (2,3). Proteases have been also implicated in invasion of normal human tissue by malignant cells (4-7). Recently, we have also shown that the in vitro cytotoxicity of the human epidermoid A431 carcinoma and glioblastoma Al72 cell lines is mediated by proteolytic enzymes (8,9). Proteinases are classified into four large classes including the serine, aspartic, and cysteine proteinases and metalloproteinases. In the past few years a large number of serine and cysteine proteinases have been
1 To whom 0003-2697/90
Copyright All rights
all correspondence
should
$3.00
0 1990 by Academic Press, Inc. of reproduction in any form reserved.
be addressed.
characterized. A fewer number is reported on metalloproteinases or aspartyl (acid) proteinases. Since the pro-forms of proteinases are transient, they have not been well characterized in many cells and tissue types. To date, only a small subset of known proteinases have been cloned from cDNA libraries (10-16). Thus, molecular cloning and sequencing of yet unknown enzymes will make possible more detailed analysis of the transient intermediates and the various regulatory pathways of these enzymes. The cloning vector bacteriophage X gtll that generates a fusion protein consisting of P-galactosidase and the amino acid sequence specified by the cloned insert cDNA was used (17). The screening of X expression libraries has been traditionally accomplished by immunodetection using monoclonal antibodies directed against the antigen encoded by the mRNA of the fusion protein or a closely related gene (17). Alternatively, the library can be screened by hybridization of an oligonucleotide or DNA fragment containing the desired sequence. More recently, a functional assay was utilized to clone acetyl transferase enzyme from a cDNA library (18). Positive clones once identified are normally subcloned into a smaller vector, such as Ml3 or plasmid, where the cloned DNA can easily be characterized by restriction enzyme mapping, sequencing, etc. Due to the paucity of specific probes for proteinase genes, it would be expedient to characterize unknown sequences directly within the Xvector without prior knowledge of the amino acid or nucleotide sequences. To facilitate the determination of whether the clone contains a proteinase sequence prior to subcloning, which can sometimes be a time-consuming and tedious venture; herein, we describe a method for performing plaque isolation using a nonradioactive nonimmunogenic detection method that takes advantage of the ability of these protease enzymes to digest protein substrates in a nonspecific fash235
236
HAGAG,
KIRCHNER,
ion. The cDNA library used in this study was made from human SCCL poly(A) RNA. The cloning of full-length double-stranded cDNA in an expression vector was necessary for this type of assay. This method should prove useful in the isolation of other novel proteinases from virtually any cancer cell. Furthermore, knowledge of new enzyme sequences will greatly facilitate the production of specific antibodies and DNA probes to their encoded proteins thus allowing the measurement of expression of proteinases at the protein and mRNA levels.
MATERIALS
AND
AND
DISTEFANO
10 InM IPTG’ and dried briefly under heat lamps before applying them to the plates containing the plaques.
Transfer of Plaques to Membrane
After phage growth for 6 h at 37”C, gelatin-coated nitrocellulose filters were layered on the plates and the orientation of the filters was marked using a sterile syringe needle. The membrane filters were allowed to wet completely (-2 min) and then incubated (plaque side down) for 3-6 h. Filters were removed and incubated phage side up for an additional 3 h at 37°C before being fixed.
METHODS
Detection Cells and Amplification
of the cDNA Library
Small cell carcinoma of the lung HTB 120 cell line was obtained from the ATCC. This continuous cell line was established from tumor material (22,23). The Xgtll cDNA expression library was obtained from Clontech using mRNA from NCI-H128 cell line derived from malignant pleural fluid of the small cell carcinoma of the lung (Gazdar et al., 1980, Cancer Res. 40, 3502-3507). The library’s initial titer was 10’ pfu/ml and after one round of amplification was 10” pfu/ml. Approximately 2.1 X lo6 plaque forming units of the recombinant phage (containing inserts in the range of 0.3 to 3.7 kbp cloned into the unique EcoRl site) were plated on top agar plates. An equivalent number of plaques of the corresponding X gtl0 cDNA library of small cell of the lung and another independent Burkitt’s lymphoma cell line BL41 were plated for comparison. The ampicillin-resistant Y1090 bacteria were used as the recipient cells. Escherichia eoti Y1090 strain were grown overnight to log phase in L. broth. The following morning bacteria were concentrated to one-sixth its original volume in SM buffer (100 mM NaCl, 10 mM MgSO,, 20 mM Tris- Cl, pH 7.5, and 0.1% gelatin). Aliquots of 200 ~1 were transferred to tubes, mixed with 100 ~1 of the appropriate titer of the phage and incubated at 37°C for 15 min. Infected bacteria were then mixed with 2.5 ml of top agar (or agarose), prewarmed to 45°C and plated onto loo-mm agar plates. Plates were subsequently incubated until the appearance of plaques in a 37°C incubator for an additional 4-6 h.
Preparation Filters
Supports
of GelatinfIPTG-Coated
Nitrocellulose
Nitrocellulose filters were submerged into 1% filter sterilized gelatin in 150 mM phosphate buffer, pH 7.4, for 5 min, followed by drying in a vacuum oven for at least 30 min at 80°C. Membranes were then soaked into
of Positive
Proteinase
Expressors
For detection of positive proteinase encoding cDNA, filters were submerged into staining solution containing 1:l 0.1% Congo red and Evans blue in fixing solution (40% methanol, 10% acetic acid, 50% water) for 3-5 min with gentle shaking until uniform staining was achieved. Filters were then transferred to the destaining solution (same as above without the dyes). After several washes, positive plaques were identified as those with negative (clear) staining. This procedure was repeated through primary, secondary, and tertiary screening and purification of the positive clones from the library.
Lysis of Bacteria Expressing
Cloned Proteinases
Preparation of total bacterial protein extracts was made according to the method of Metzul et al. (19). Briefly, E. coli Y1090 cells were lysogenized either with nonrecombinant gtll or with gtll clones containing cDNA inserts as described by Huynh et al. (23). Bacteria were screened for lysogeny by their inability to grow at 42°C. For the preparation of crude bacterial extracts, parallel cultures of each clone were grown at 37°C. To induce phage multiplication, the logarithmically growing cultures were rapidly heated to 42°C and maintained at this temperature for 20-30 min. The lac-Z expression was then induced by addition of 10 mM IPTG and further incubation at 37°C for 30-60 min. Care was taken that no cell lysis occurred during this time. A l-ml sample of each culture was centrifuged for 30 s at room temperature in an Eppendorf centrifuge, and the bacterial pellets were immediately frozen in liquid nitrogen and then resuspended in 0.5 ml of 20 mM potassium sodium phosphate, pH 7.5. Lysis of bacteria occurs upon thawing, yielding a crude total extract. For assay of proteolytic activity, the resuspended bacteria were thawed on ice, and lysate was used without further treatment.
* Abbreviations as.8 chain reaction.
used: IPTG,
isopropyltbiogalactose;
PCR, polymer-
PROTEINASE-ENCODING
Polyacrylamide Gelatin Produced Proteinases
GENE
Gel Electrophoresis
CLONING
of Plaque-
Proteinase activity was confirmed by electrophoretic analysis in 11% polyacrylamide sodium dodecyl sulfate gels containing copolymerized 1% gelatin as previously described (8,9,21). In this assay gelatin serves as an in situ substrate for the localization of proteinase bands by negative staining of the digested gelatin where a proteinase band is localized. This method can detect as little as 1 mU of enzyme activity (1 mU = 0.01 U equivalent of 10 ng purified trypsin protein that cleaves the synthetic substrate, N-benzoyl-L-arginine ethyl ester). Isolation and Characterization Phage DNA
of Recombinant
Preparation of phage DNA was performed as described (24). Briefly, 300 ~1 of 3 X lO’/ml fresh Y1090 were seeded with a single fresh plaque, incubated at 37°C for 15 min, mixed with 2.5 ml top agar, and poured onto a loo-mm plate. Plates were then incubated at 37°C for 4 to 6 h, or until plaques were clearly visible and when 90 to 100% of the lawn was lysed. A 3-ml sample of SM buffer was pipetted onto the plate. Top agar was scrapped off into a centrifuge tube to which 3 drops of chloroform was added. After vortexing for 10 s, the bacterial lysate was centrifuged for 10 min at 10,000 rpm in a JA21 rotor at 4°C. The supernatant was transferred to a clean tube; for each 50 ml liquid lysate, 10 ~1 of 5 mg/ml DNase and 25 ~1 of 10 mg/ml RNase were added. Incubation was continued for 1 h at 37°C. Phage particles were pelleted at 30,000 rpm for 1.5 h. The phage pellet was resuspended into 200 ~1 of 50 mM Tris . Cl, pH 8.0, transferred to an Eppendorf microfuge tube, and extracted twice with 200 ~1 of buffered phenol and once with 200 ~1 of chloroform. Phage DNA was precipitated by the addition of 20 ~1 of 3 M sodium acetate, pH 4.8, and 6 vol of 100% ethanol at room temperature followed by centrifugation at 14,000 rpm for 10 min. The pellet was washed with 70% ethanol, dried and resuspended into 100 ~1 of TE buffer. Polymerase
Chain Reaction
(PCR)
of cDNA
Inserts
Phage DNA was obtained by the rapid small-scale method (24). A 1-/lg sample of phage DNA was placed in 100 ~1 of 1X PCR buffer (10 mM Tris * Cl, pH 8.3,50 mM KCl, 10 mM MgCl,, 0.1% gelatin, dNTPs at 1.5 mM each with 1 pM forward and reverse primers). Ultrapurified primers (flanking the EcoRl cloning site in lac Z gene) were obtained from Genosys (Houston, TX). The following primer sequences were used: forward, 5’AGCAAGTTCAGCCTGGTTAAG3’, and reverse, 5’TTATGAGTATTTCTTCCAGGG3’. The PCR reaction was performed essentially as described (25,26). Briefly,
IN
CANCER
CELLS
BY
ZYMOGEN
ASSAY
237
2.5 units of Taq DNA polymerase (Cetus, Emeryville, CA) was added and the reaction was overlaid with 100 ~1 of mineral oil. Capped tubes were subjected to serial cycles of denaturation (30 s at 95°C; first cycle, 2.5 min), annealing (3 min at 55’C), synthesis, and extension (3 min at 72’C). After the final cycle, PCR products were extracted with chloroform, precipitated, and size fractionated on alkaline 4% NuSieve GTG agarose (FMC Bioproducts, Rockland, ME) gels. RESULTS
Screening
of Phage Library
for Proteinases
We have developed a new screening method that permits the direct isolation and cloning of unknown proteinase genes from cDNA libraries of cancer and eukaryotic cells. The feasibility of this screening strategy was tested by using a X gtll cDNA expression library of the small cell carcinoma of the lung. By applying the procedure outlined above, we were able to detect positive proteinase-encoding genes within 30 min following the transfer of plaques onto nitrocellulose membrane filters. The procedure is a simple and rapid calorimetric assay. Initial screening of about 600,000 plaques revealed a total of 56 positives. Figure la shows a representative filter from primary screening. The positive primary plaques were picked with an Eppendorf pipet tip and phage was eluted in SM buffer and replated at lower density and subjected to two identical cycles of screening procedure. Figures lb and lc show the results of the second and third round positives obtained from primary plaques. Comparison of the filters with agar plates revealed that only a subset of the total plaques were detected. We used two h gtl0 cDNA library plaques as controls. Screening of either BL41 (Burkitt’s lymphoma) or SCCL gtl0 plaques yielded no positive protease activity, indicating that the clearance observed on gelatin filters is a result of expressed proteinase genes encoded in the cDNA inserts of X gtll phages. Furthermore, no artifacts were observed in filters that have been overlaid on plates containing lawns of uninfected bacteria or bacteria infected with wild-type X gtll phage. Also, the number of proteinase positive plaques significantly increased with the secondary and tertiary cycles of amplifications (Fig. lc). The positive signals on the filters closely resembled the appearance of plaques on the corresponding agar plate. They were roughly circular in shape and of similar size. On some filters, the positive signals had characteristic “cometlike” tails, an artifact known in plaque hybridization procedures due to smearing during lifting of the filters from the plates. Gelatin Gel Electrophoresis
Assay
Direct evidence that the P-galactosidase fusion proteins encoded by X gtll were responsible for the proteo-
238
HAGAG,
KIRCHNER,
AND
DISTEFANO
PROTEINASE-ENCODING
GENE
CLONING
IN
CANCER
CELLS
BY
ZYMOGEN
ASSAY
FIG. 1. Identification of proteinase encoding genes by gelatin filter screening assay of cDNA expression library of the small cell carcinoma the lung. (a) Primary screening, (b) secondary screening, and (c) tertiary screening. Nitrocellulose membrane filters were precoated with gelatin and 10 mM IPTG and then layered onto lawns of bacteria infected with X gtll phage as described under Materials and Methods.
lytic activities observed on membrane filters was obtained by the gelatin gel electrophoresis assay. Figures 2A and 2B show the results of an assay of extracts derived from membranes of SCCL and X gtll 1,5, and 7 lysogens, respectively. The activity was IPTG inducible, indicating that it was a product of the lac Z fusion gene. The size of proteins were of an average molecular weight of 160,000 to 200,000 probably representing the proteinase+galactosidase fusion protein. Since the pgalactosidase portion of this fusion polypeptide has a M, of approximately 120,000, the cDNA encoded portion must have a M, of 40,000-80,000. If the phage DNA represents a single species, it is likely that the smaller size peptide is a proteolytic cleavage product.
239
of 1%
indicating that the amplified DNA in X gtll is actually a result of amplification of the cDNA insert in lac Z gene. Our preliminary data of direct sequencing of cDNA of clones indicate that the isolated clones are indeed genes coding for proteolytic enzymes with remarkable homology to the bioactive site of known serine proteases including bovine thrombin, human blood coagulation factors FX and FXI, and rat elastase II 1 (to be published elsewhere). Since these enzymes are normally synthesized in the pre- and pro-forms, and since the isolated clones exhibited proteinase activity on both gelatin filters and polyacrylamide gelatin gels, it is likely that these clones are not full-length cDNA. DISCUSSION
Amplification
of cDNA
Inserts
PCR amplification of cDNA inserts from 10 tertiary purified positive plaques revealed that the size of inserts in these clones ranged between 0.3 and 2.0 kbp, indicating that some of these inserts are not full-length cDNA. No inserts were observed in PCR reactions using either wild-type X gtll or a randomly selected X gtl0 plaque,
To facilitate the isolation of unknown human genes specifying proteolytic enzymes, we have developed a simple strategy to detect clones encoding proteinases based on their general biological activity. The proteinases were cloned by screening a cDNA expression library of the human small cell carcinoma of the lung with a solid-phase enzyme activity assay. Since it is now
240
HAGAG,
KIRCHNER,
FIG.
2. Polyacrylamide-gelatin electrophoresis of proteolytic enzymes in (A) small-cell carcinoma of the lung cells, SCCL, (B) positive plaque proteinase clones 1, 5, and 7, and (C) molecular weight markers, M, on a Coomassie blue gel run under identical conditions. Markers are from top to bottom in kDa: myosin, 200; @-galactosidase, 116.25; phosphorylase B, 97.4; bovine serum albumin, 66.2; and ovalbumin, 45.
widely accepted that proteinases are involved in mediating cell cytotoxicity and metastatic behavior, we chose the small-cell carcinoma of the lung as a model system in this study because of their known aggressive behavior in patients with this type of malignancy. Because of the propensity for early widespread metastasis to the bone marrow, lymph nodes, pleura, liver, brain, and adrenal glands at diagnosis in most patients we sought to investigate whether this cell line expresses a variety of unknown proteinase genes, some of which may be specific for their wide spectrum cytotoxicity. The strategy depends on the presumed functional expression of the fused fi-galactosidase-protein in X gtll infected Y1090 bacteria. This technique combines the use of gelatin substrate coated onto a nitrocellulose or nylon membrane filters and subsequent recognition of the enzyme activity using negative staining of digested gelatin. The identity of proteinases were further confirmed by using polyacrylamide gelatin electrophoresis assay, polymerase chain reaction amplification, and direct sequencing. The strategy is analogous to immunological screening of expression libraries using monoclonal antibody to the protein product of the gene of interest or a closely related gene. Albeit, it does not require prior knowledge of the sequence of the fusion protein. The presence of approximately the same number of second round positives on filters from several different plattings (-100 phages/plate) indicates that the positives are specific to a subset of the infecting phage. The proteolytic enzyme phage clones were purified to homogeneity by a third round of screening and characterized
AND
DISTEFANO
by restriction endonuclease map analysis and sized by PCR. Furthermore, partial sequencing of the cDNA using primers flanking the EcoRl cloning site indicates that these genes belong to the proteinase family. These results show that a X gtll cDNA insert is responsible for the proteolytic activity found in the isolated clones. It is unlikely that the enzymatic activity observed is actually bacterial proteinases. In control experiments neither bacteria alone nor bacteria infected with wild-type X gtll or h gtl0 phages possess the ability to hydrolyze gelatin on filters or in the gelatin electrophoresis assay. It also seems highly improbable that the insert is stimulating the expression of an enzyme endogenous to the bacterial host. The fact that the PCR-amplified inserts were of different sizes indicate that these are actually different genes. This method should help facilitate the quick isolation of various proteinase-encoding genes based on their substrate specificity. Isolation of a particular proteinase could be achieved by using membrane filters coated with a more specific substrate, e.g., laminin, collagen, and fibronectin. Similarly, isolation of broad spectrum proteinases could be accomplished by using albumin- or casein-coated filters. Isolation of new proteinases should make the study of the regulation and expression of these enzymes and their role in various biological processes possible.
ACKNOWLEDGMENTS This Cancer tion.
study was supported in part Society (CD 104) and a grant
by a grant from the American from the Veterans Administra-
REFERENCES 1. Barrett,
A. J., and Kirschke, H. (1981) in Methods in Enzymology (Lorand, L., Ed.), Vol. 80, pp. 535-561, Academic Press, San Diego, CA. N., Berg, M. J., and Benuck, M. (1986) Arch. Biochem. 2. Marks,
Biophys. 249,489-499. 3. Taugner,
R., Buhrle,
Histochemistry 4. Sloane,
C. P., Nobiling,
R., and Kirschke,
H. (1985)
f38,103-108.
B. J., and Honn,
K. V. (1984)
Cancer
Metastasis Rev. 3,
249-263. D. E., and Rohrlich, C. T. (1983) Biochim. Biophys. Acta 5. Mullins, 695,177-214. A. J. (1980) in Proteinases and Tumor Invasion (Strauli, 6. Barrett, P., Baici, A., and Barrett, A. J., Eds.) pp. 59-67, Raven, New York. 7. Neurath, H. (1984) Science 224,350-357. J. F., Cotto, C. A., Lane, B., and Hagag, N. G. (1988) 8. DiStefano,
Cancer Invest. 6,487-498. J. F., Cotto, 9. DiStefano, Cancer Res. 49,179-184. 10. San Segundo, B., Chan,
C. A., Lane,
B., Hagag,
S. J., and Steiner,
Acad. Sci. USA S&2320-2324.
D. (1985)
N. G. (1989)
Proc. N&l.
PROTEINASE-ENCODING
GENE
11. Faust, P. L., Kornfeld, S., and Chirgwin, Acad. Sci. USA 82,4910-4914. 12. Yoshitake, Kurachi,
J. M.
(1985)
S., Schach, B. G., Foster, D. C., Davie, K. (1984) Biochemistry 24, 3736-3750.
13. MacDonald, R. J., Stary, Chem. 257,9724-9732. 14. Friezner, S. J., Rajput, 261,6972-6985.
J. S., and B., and
Swift,
Reich,
Natl.
PFOC.
E. W.,
G. H. (1982)
E. (1986)
15. Fujikawa, K., Chung, D. W., Henrickson, (1986) Biochemistry 25,2417-2424.
CLONING
and
J. Biol.
J. Biol.
L. E., and Davie,
Chem. E. W.
16. Malinowski, D. P., Sadler, J. E., and Davie, E. W. (1894) Biochemistry 23,4243-4250. 17. Young, R., and Davis, R. (1983) PFOC. Natl. Acad. Sci. USA 80, 1194-1198. 18. Eberwine, J. H., Barches, J. D., Helwet, W. A., and Evans, (1987) PFOC. Natl. Acad. Sci. USA 84, 144991453. 19. Mutzel, R., Simon, M.-N., Biochemistry 27, 481-486.
Lacombe,
M.-L.,
and Veron,
C. J.
M. (1988)
IN
CANCER
20. Pettengill, T. J., Noll, 45,966-918.
CELLS
BY
ZYMOGEN
ASSAY
0. S., Sorenson, G. D., Wurster-Hill, W. W., Cate, C. C., and Maurerr,
241 D. H., Curphey, L. H. (1980) Cancer
21. Carney, R. N., Gazdar, A. F., Bepler, G., Guccion, J. G., Marangos, P. J., Moody, T. W., Zweig, M. H., and Minna, J. D. (1985) Cancer Res. 45,2913-2923. C., and Dowdle, E. B. (1980) Anal. Biochem. 102,19622* Heussen, 202. T. V., Young, R. A., and Davis, R. W. (1984) in DNA 23. Huynh, Cloning: A Practical Approach (Glover, D., Ed.), pp. 49-78, IRL, Oxford. 24. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A., and Struhl, K. (1987) in Current Protocols in Molecular Biology, Wiley, New York. 25. Saiki, R. K., Gelfand, D. H., Stofel, S., Scharf, S. J., Hiuchi, R., Horn, G. T., Mullis, K. B., and Erlich, H. A. (1988) Science 239, 487-491. 26. Scharf, S. J., Horn, G. T., and Erlich, H. A. (1986) Science 233, 1076-1078.
BIOCHEMISTRY
191,
235-241
(1990)
Molecular Cloning of Proteinase-Encoding Genes from Cancer Cells by Zymogen Assay and Direct Sequencing Nabil
G. Hagag,’
Michele
Division of Oncology, Health and Veterans Administration
Received
June
Kirchner,
and John
F. DiStefano
Sciences Center, State University of New York at Stony Brook, New York 11794, Medical Center, Northport, New York 11768
4, 1990
A method that permits the in vitro cloning and identification of proteolytic enzyme genes from cDNA expression libraries is described. The method can detect positive proteinase genes within 30 min following the transfer of plaques to nitrocellulose membrane filters. The method is based on the functional expression of fusion lot Z-proteinase protein in X gt 11 infected Y 1090 bacteria and does not require prior knowledge of either the sequence of the cDNA insert or a monoclonal antibody to its encoded antigen. This strategy when coupled with polymerase chain reaction of the cDNA insert using lac Z primer sequences that are flanking the EcoRl cloning site in gtl 1 phage permits direct sequencing of the amplified DNA. With this method we have isolated 10 genes expressing protease activity in the human small-cell carcinoma of the lung. The same procedure could be applied to isolate unknown proteinases from cDNA libraries of virtually any eukaryotic cell. o 1990 Academic
Press,
Inc.
Proteolytic enzymes have a variety of biological functions ranging from extracellular and intracellular catabolism of peptides and proteins (1) to precise processing of precursor proteins, such as prohormones and proenzymes (2,3). Proteases have been also implicated in invasion of normal human tissue by malignant cells (4-7). Recently, we have also shown that the in vitro cytotoxicity of the human epidermoid A431 carcinoma and glioblastoma Al72 cell lines is mediated by proteolytic enzymes (8,9). Proteinases are classified into four large classes including the serine, aspartic, and cysteine proteinases and metalloproteinases. In the past few years a large number of serine and cysteine proteinases have been
1 To whom 0003-2697/90
Copyright All rights
all correspondence
should
$3.00
0 1990 by Academic Press, Inc. of reproduction in any form reserved.
be addressed.
characterized. A fewer number is reported on metalloproteinases or aspartyl (acid) proteinases. Since the pro-forms of proteinases are transient, they have not been well characterized in many cells and tissue types. To date, only a small subset of known proteinases have been cloned from cDNA libraries (10-16). Thus, molecular cloning and sequencing of yet unknown enzymes will make possible more detailed analysis of the transient intermediates and the various regulatory pathways of these enzymes. The cloning vector bacteriophage X gtll that generates a fusion protein consisting of P-galactosidase and the amino acid sequence specified by the cloned insert cDNA was used (17). The screening of X expression libraries has been traditionally accomplished by immunodetection using monoclonal antibodies directed against the antigen encoded by the mRNA of the fusion protein or a closely related gene (17). Alternatively, the library can be screened by hybridization of an oligonucleotide or DNA fragment containing the desired sequence. More recently, a functional assay was utilized to clone acetyl transferase enzyme from a cDNA library (18). Positive clones once identified are normally subcloned into a smaller vector, such as Ml3 or plasmid, where the cloned DNA can easily be characterized by restriction enzyme mapping, sequencing, etc. Due to the paucity of specific probes for proteinase genes, it would be expedient to characterize unknown sequences directly within the Xvector without prior knowledge of the amino acid or nucleotide sequences. To facilitate the determination of whether the clone contains a proteinase sequence prior to subcloning, which can sometimes be a time-consuming and tedious venture; herein, we describe a method for performing plaque isolation using a nonradioactive nonimmunogenic detection method that takes advantage of the ability of these protease enzymes to digest protein substrates in a nonspecific fash235
236
HAGAG,
KIRCHNER,
ion. The cDNA library used in this study was made from human SCCL poly(A) RNA. The cloning of full-length double-stranded cDNA in an expression vector was necessary for this type of assay. This method should prove useful in the isolation of other novel proteinases from virtually any cancer cell. Furthermore, knowledge of new enzyme sequences will greatly facilitate the production of specific antibodies and DNA probes to their encoded proteins thus allowing the measurement of expression of proteinases at the protein and mRNA levels.
MATERIALS
AND
AND
DISTEFANO
10 InM IPTG’ and dried briefly under heat lamps before applying them to the plates containing the plaques.
Transfer of Plaques to Membrane
After phage growth for 6 h at 37”C, gelatin-coated nitrocellulose filters were layered on the plates and the orientation of the filters was marked using a sterile syringe needle. The membrane filters were allowed to wet completely (-2 min) and then incubated (plaque side down) for 3-6 h. Filters were removed and incubated phage side up for an additional 3 h at 37°C before being fixed.
METHODS
Detection Cells and Amplification
of the cDNA Library
Small cell carcinoma of the lung HTB 120 cell line was obtained from the ATCC. This continuous cell line was established from tumor material (22,23). The Xgtll cDNA expression library was obtained from Clontech using mRNA from NCI-H128 cell line derived from malignant pleural fluid of the small cell carcinoma of the lung (Gazdar et al., 1980, Cancer Res. 40, 3502-3507). The library’s initial titer was 10’ pfu/ml and after one round of amplification was 10” pfu/ml. Approximately 2.1 X lo6 plaque forming units of the recombinant phage (containing inserts in the range of 0.3 to 3.7 kbp cloned into the unique EcoRl site) were plated on top agar plates. An equivalent number of plaques of the corresponding X gtl0 cDNA library of small cell of the lung and another independent Burkitt’s lymphoma cell line BL41 were plated for comparison. The ampicillin-resistant Y1090 bacteria were used as the recipient cells. Escherichia eoti Y1090 strain were grown overnight to log phase in L. broth. The following morning bacteria were concentrated to one-sixth its original volume in SM buffer (100 mM NaCl, 10 mM MgSO,, 20 mM Tris- Cl, pH 7.5, and 0.1% gelatin). Aliquots of 200 ~1 were transferred to tubes, mixed with 100 ~1 of the appropriate titer of the phage and incubated at 37°C for 15 min. Infected bacteria were then mixed with 2.5 ml of top agar (or agarose), prewarmed to 45°C and plated onto loo-mm agar plates. Plates were subsequently incubated until the appearance of plaques in a 37°C incubator for an additional 4-6 h.
Preparation Filters
Supports
of GelatinfIPTG-Coated
Nitrocellulose
Nitrocellulose filters were submerged into 1% filter sterilized gelatin in 150 mM phosphate buffer, pH 7.4, for 5 min, followed by drying in a vacuum oven for at least 30 min at 80°C. Membranes were then soaked into
of Positive
Proteinase
Expressors
For detection of positive proteinase encoding cDNA, filters were submerged into staining solution containing 1:l 0.1% Congo red and Evans blue in fixing solution (40% methanol, 10% acetic acid, 50% water) for 3-5 min with gentle shaking until uniform staining was achieved. Filters were then transferred to the destaining solution (same as above without the dyes). After several washes, positive plaques were identified as those with negative (clear) staining. This procedure was repeated through primary, secondary, and tertiary screening and purification of the positive clones from the library.
Lysis of Bacteria Expressing
Cloned Proteinases
Preparation of total bacterial protein extracts was made according to the method of Metzul et al. (19). Briefly, E. coli Y1090 cells were lysogenized either with nonrecombinant gtll or with gtll clones containing cDNA inserts as described by Huynh et al. (23). Bacteria were screened for lysogeny by their inability to grow at 42°C. For the preparation of crude bacterial extracts, parallel cultures of each clone were grown at 37°C. To induce phage multiplication, the logarithmically growing cultures were rapidly heated to 42°C and maintained at this temperature for 20-30 min. The lac-Z expression was then induced by addition of 10 mM IPTG and further incubation at 37°C for 30-60 min. Care was taken that no cell lysis occurred during this time. A l-ml sample of each culture was centrifuged for 30 s at room temperature in an Eppendorf centrifuge, and the bacterial pellets were immediately frozen in liquid nitrogen and then resuspended in 0.5 ml of 20 mM potassium sodium phosphate, pH 7.5. Lysis of bacteria occurs upon thawing, yielding a crude total extract. For assay of proteolytic activity, the resuspended bacteria were thawed on ice, and lysate was used without further treatment.
* Abbreviations as.8 chain reaction.
used: IPTG,
isopropyltbiogalactose;
PCR, polymer-
PROTEINASE-ENCODING
Polyacrylamide Gelatin Produced Proteinases
GENE
Gel Electrophoresis
CLONING
of Plaque-
Proteinase activity was confirmed by electrophoretic analysis in 11% polyacrylamide sodium dodecyl sulfate gels containing copolymerized 1% gelatin as previously described (8,9,21). In this assay gelatin serves as an in situ substrate for the localization of proteinase bands by negative staining of the digested gelatin where a proteinase band is localized. This method can detect as little as 1 mU of enzyme activity (1 mU = 0.01 U equivalent of 10 ng purified trypsin protein that cleaves the synthetic substrate, N-benzoyl-L-arginine ethyl ester). Isolation and Characterization Phage DNA
of Recombinant
Preparation of phage DNA was performed as described (24). Briefly, 300 ~1 of 3 X lO’/ml fresh Y1090 were seeded with a single fresh plaque, incubated at 37°C for 15 min, mixed with 2.5 ml top agar, and poured onto a loo-mm plate. Plates were then incubated at 37°C for 4 to 6 h, or until plaques were clearly visible and when 90 to 100% of the lawn was lysed. A 3-ml sample of SM buffer was pipetted onto the plate. Top agar was scrapped off into a centrifuge tube to which 3 drops of chloroform was added. After vortexing for 10 s, the bacterial lysate was centrifuged for 10 min at 10,000 rpm in a JA21 rotor at 4°C. The supernatant was transferred to a clean tube; for each 50 ml liquid lysate, 10 ~1 of 5 mg/ml DNase and 25 ~1 of 10 mg/ml RNase were added. Incubation was continued for 1 h at 37°C. Phage particles were pelleted at 30,000 rpm for 1.5 h. The phage pellet was resuspended into 200 ~1 of 50 mM Tris . Cl, pH 8.0, transferred to an Eppendorf microfuge tube, and extracted twice with 200 ~1 of buffered phenol and once with 200 ~1 of chloroform. Phage DNA was precipitated by the addition of 20 ~1 of 3 M sodium acetate, pH 4.8, and 6 vol of 100% ethanol at room temperature followed by centrifugation at 14,000 rpm for 10 min. The pellet was washed with 70% ethanol, dried and resuspended into 100 ~1 of TE buffer. Polymerase
Chain Reaction
(PCR)
of cDNA
Inserts
Phage DNA was obtained by the rapid small-scale method (24). A 1-/lg sample of phage DNA was placed in 100 ~1 of 1X PCR buffer (10 mM Tris * Cl, pH 8.3,50 mM KCl, 10 mM MgCl,, 0.1% gelatin, dNTPs at 1.5 mM each with 1 pM forward and reverse primers). Ultrapurified primers (flanking the EcoRl cloning site in lac Z gene) were obtained from Genosys (Houston, TX). The following primer sequences were used: forward, 5’AGCAAGTTCAGCCTGGTTAAG3’, and reverse, 5’TTATGAGTATTTCTTCCAGGG3’. The PCR reaction was performed essentially as described (25,26). Briefly,
IN
CANCER
CELLS
BY
ZYMOGEN
ASSAY
237
2.5 units of Taq DNA polymerase (Cetus, Emeryville, CA) was added and the reaction was overlaid with 100 ~1 of mineral oil. Capped tubes were subjected to serial cycles of denaturation (30 s at 95°C; first cycle, 2.5 min), annealing (3 min at 55’C), synthesis, and extension (3 min at 72’C). After the final cycle, PCR products were extracted with chloroform, precipitated, and size fractionated on alkaline 4% NuSieve GTG agarose (FMC Bioproducts, Rockland, ME) gels. RESULTS
Screening
of Phage Library
for Proteinases
We have developed a new screening method that permits the direct isolation and cloning of unknown proteinase genes from cDNA libraries of cancer and eukaryotic cells. The feasibility of this screening strategy was tested by using a X gtll cDNA expression library of the small cell carcinoma of the lung. By applying the procedure outlined above, we were able to detect positive proteinase-encoding genes within 30 min following the transfer of plaques onto nitrocellulose membrane filters. The procedure is a simple and rapid calorimetric assay. Initial screening of about 600,000 plaques revealed a total of 56 positives. Figure la shows a representative filter from primary screening. The positive primary plaques were picked with an Eppendorf pipet tip and phage was eluted in SM buffer and replated at lower density and subjected to two identical cycles of screening procedure. Figures lb and lc show the results of the second and third round positives obtained from primary plaques. Comparison of the filters with agar plates revealed that only a subset of the total plaques were detected. We used two h gtl0 cDNA library plaques as controls. Screening of either BL41 (Burkitt’s lymphoma) or SCCL gtl0 plaques yielded no positive protease activity, indicating that the clearance observed on gelatin filters is a result of expressed proteinase genes encoded in the cDNA inserts of X gtll phages. Furthermore, no artifacts were observed in filters that have been overlaid on plates containing lawns of uninfected bacteria or bacteria infected with wild-type X gtll phage. Also, the number of proteinase positive plaques significantly increased with the secondary and tertiary cycles of amplifications (Fig. lc). The positive signals on the filters closely resembled the appearance of plaques on the corresponding agar plate. They were roughly circular in shape and of similar size. On some filters, the positive signals had characteristic “cometlike” tails, an artifact known in plaque hybridization procedures due to smearing during lifting of the filters from the plates. Gelatin Gel Electrophoresis
Assay
Direct evidence that the P-galactosidase fusion proteins encoded by X gtll were responsible for the proteo-
238
HAGAG,
KIRCHNER,
AND
DISTEFANO
PROTEINASE-ENCODING
GENE
CLONING
IN
CANCER
CELLS
BY
ZYMOGEN
ASSAY
FIG. 1. Identification of proteinase encoding genes by gelatin filter screening assay of cDNA expression library of the small cell carcinoma the lung. (a) Primary screening, (b) secondary screening, and (c) tertiary screening. Nitrocellulose membrane filters were precoated with gelatin and 10 mM IPTG and then layered onto lawns of bacteria infected with X gtll phage as described under Materials and Methods.
lytic activities observed on membrane filters was obtained by the gelatin gel electrophoresis assay. Figures 2A and 2B show the results of an assay of extracts derived from membranes of SCCL and X gtll 1,5, and 7 lysogens, respectively. The activity was IPTG inducible, indicating that it was a product of the lac Z fusion gene. The size of proteins were of an average molecular weight of 160,000 to 200,000 probably representing the proteinase+galactosidase fusion protein. Since the pgalactosidase portion of this fusion polypeptide has a M, of approximately 120,000, the cDNA encoded portion must have a M, of 40,000-80,000. If the phage DNA represents a single species, it is likely that the smaller size peptide is a proteolytic cleavage product.
239
of 1%
indicating that the amplified DNA in X gtll is actually a result of amplification of the cDNA insert in lac Z gene. Our preliminary data of direct sequencing of cDNA of clones indicate that the isolated clones are indeed genes coding for proteolytic enzymes with remarkable homology to the bioactive site of known serine proteases including bovine thrombin, human blood coagulation factors FX and FXI, and rat elastase II 1 (to be published elsewhere). Since these enzymes are normally synthesized in the pre- and pro-forms, and since the isolated clones exhibited proteinase activity on both gelatin filters and polyacrylamide gelatin gels, it is likely that these clones are not full-length cDNA. DISCUSSION
Amplification
of cDNA
Inserts
PCR amplification of cDNA inserts from 10 tertiary purified positive plaques revealed that the size of inserts in these clones ranged between 0.3 and 2.0 kbp, indicating that some of these inserts are not full-length cDNA. No inserts were observed in PCR reactions using either wild-type X gtll or a randomly selected X gtl0 plaque,
To facilitate the isolation of unknown human genes specifying proteolytic enzymes, we have developed a simple strategy to detect clones encoding proteinases based on their general biological activity. The proteinases were cloned by screening a cDNA expression library of the human small cell carcinoma of the lung with a solid-phase enzyme activity assay. Since it is now
240
HAGAG,
KIRCHNER,
FIG.
2. Polyacrylamide-gelatin electrophoresis of proteolytic enzymes in (A) small-cell carcinoma of the lung cells, SCCL, (B) positive plaque proteinase clones 1, 5, and 7, and (C) molecular weight markers, M, on a Coomassie blue gel run under identical conditions. Markers are from top to bottom in kDa: myosin, 200; @-galactosidase, 116.25; phosphorylase B, 97.4; bovine serum albumin, 66.2; and ovalbumin, 45.
widely accepted that proteinases are involved in mediating cell cytotoxicity and metastatic behavior, we chose the small-cell carcinoma of the lung as a model system in this study because of their known aggressive behavior in patients with this type of malignancy. Because of the propensity for early widespread metastasis to the bone marrow, lymph nodes, pleura, liver, brain, and adrenal glands at diagnosis in most patients we sought to investigate whether this cell line expresses a variety of unknown proteinase genes, some of which may be specific for their wide spectrum cytotoxicity. The strategy depends on the presumed functional expression of the fused fi-galactosidase-protein in X gtll infected Y1090 bacteria. This technique combines the use of gelatin substrate coated onto a nitrocellulose or nylon membrane filters and subsequent recognition of the enzyme activity using negative staining of digested gelatin. The identity of proteinases were further confirmed by using polyacrylamide gelatin electrophoresis assay, polymerase chain reaction amplification, and direct sequencing. The strategy is analogous to immunological screening of expression libraries using monoclonal antibody to the protein product of the gene of interest or a closely related gene. Albeit, it does not require prior knowledge of the sequence of the fusion protein. The presence of approximately the same number of second round positives on filters from several different plattings (-100 phages/plate) indicates that the positives are specific to a subset of the infecting phage. The proteolytic enzyme phage clones were purified to homogeneity by a third round of screening and characterized
AND
DISTEFANO
by restriction endonuclease map analysis and sized by PCR. Furthermore, partial sequencing of the cDNA using primers flanking the EcoRl cloning site indicates that these genes belong to the proteinase family. These results show that a X gtll cDNA insert is responsible for the proteolytic activity found in the isolated clones. It is unlikely that the enzymatic activity observed is actually bacterial proteinases. In control experiments neither bacteria alone nor bacteria infected with wild-type X gtll or h gtl0 phages possess the ability to hydrolyze gelatin on filters or in the gelatin electrophoresis assay. It also seems highly improbable that the insert is stimulating the expression of an enzyme endogenous to the bacterial host. The fact that the PCR-amplified inserts were of different sizes indicate that these are actually different genes. This method should help facilitate the quick isolation of various proteinase-encoding genes based on their substrate specificity. Isolation of a particular proteinase could be achieved by using membrane filters coated with a more specific substrate, e.g., laminin, collagen, and fibronectin. Similarly, isolation of broad spectrum proteinases could be accomplished by using albumin- or casein-coated filters. Isolation of new proteinases should make the study of the regulation and expression of these enzymes and their role in various biological processes possible.
ACKNOWLEDGMENTS This Cancer tion.
study was supported in part Society (CD 104) and a grant
by a grant from the American from the Veterans Administra-
REFERENCES 1. Barrett,
A. J., and Kirschke, H. (1981) in Methods in Enzymology (Lorand, L., Ed.), Vol. 80, pp. 535-561, Academic Press, San Diego, CA. N., Berg, M. J., and Benuck, M. (1986) Arch. Biochem. 2. Marks,
Biophys. 249,489-499. 3. Taugner,
R., Buhrle,
Histochemistry 4. Sloane,
C. P., Nobiling,
R., and Kirschke,
H. (1985)
f38,103-108.
B. J., and Honn,
K. V. (1984)
Cancer
Metastasis Rev. 3,
249-263. D. E., and Rohrlich, C. T. (1983) Biochim. Biophys. Acta 5. Mullins, 695,177-214. A. J. (1980) in Proteinases and Tumor Invasion (Strauli, 6. Barrett, P., Baici, A., and Barrett, A. J., Eds.) pp. 59-67, Raven, New York. 7. Neurath, H. (1984) Science 224,350-357. J. F., Cotto, C. A., Lane, B., and Hagag, N. G. (1988) 8. DiStefano,
Cancer Invest. 6,487-498. J. F., Cotto, 9. DiStefano, Cancer Res. 49,179-184. 10. San Segundo, B., Chan,
C. A., Lane,
B., Hagag,
S. J., and Steiner,
Acad. Sci. USA S&2320-2324.
D. (1985)
N. G. (1989)
Proc. N&l.
PROTEINASE-ENCODING
GENE
11. Faust, P. L., Kornfeld, S., and Chirgwin, Acad. Sci. USA 82,4910-4914. 12. Yoshitake, Kurachi,
J. M.
(1985)
S., Schach, B. G., Foster, D. C., Davie, K. (1984) Biochemistry 24, 3736-3750.
13. MacDonald, R. J., Stary, Chem. 257,9724-9732. 14. Friezner, S. J., Rajput, 261,6972-6985.
J. S., and B., and
Swift,
Reich,
Natl.
PFOC.
E. W.,
G. H. (1982)
E. (1986)
15. Fujikawa, K., Chung, D. W., Henrickson, (1986) Biochemistry 25,2417-2424.
CLONING
and
J. Biol.
J. Biol.
L. E., and Davie,
Chem. E. W.
16. Malinowski, D. P., Sadler, J. E., and Davie, E. W. (1894) Biochemistry 23,4243-4250. 17. Young, R., and Davis, R. (1983) PFOC. Natl. Acad. Sci. USA 80, 1194-1198. 18. Eberwine, J. H., Barches, J. D., Helwet, W. A., and Evans, (1987) PFOC. Natl. Acad. Sci. USA 84, 144991453. 19. Mutzel, R., Simon, M.-N., Biochemistry 27, 481-486.
Lacombe,
M.-L.,
and Veron,
C. J.
M. (1988)
IN
CANCER
20. Pettengill, T. J., Noll, 45,966-918.
CELLS
BY
ZYMOGEN
ASSAY
0. S., Sorenson, G. D., Wurster-Hill, W. W., Cate, C. C., and Maurerr,
241 D. H., Curphey, L. H. (1980) Cancer
21. Carney, R. N., Gazdar, A. F., Bepler, G., Guccion, J. G., Marangos, P. J., Moody, T. W., Zweig, M. H., and Minna, J. D. (1985) Cancer Res. 45,2913-2923. C., and Dowdle, E. B. (1980) Anal. Biochem. 102,19622* Heussen, 202. T. V., Young, R. A., and Davis, R. W. (1984) in DNA 23. Huynh, Cloning: A Practical Approach (Glover, D., Ed.), pp. 49-78, IRL, Oxford. 24. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A., and Struhl, K. (1987) in Current Protocols in Molecular Biology, Wiley, New York. 25. Saiki, R. K., Gelfand, D. H., Stofel, S., Scharf, S. J., Hiuchi, R., Horn, G. T., Mullis, K. B., and Erlich, H. A. (1988) Science 239, 487-491. 26. Scharf, S. J., Horn, G. T., and Erlich, H. A. (1986) Science 233, 1076-1078.
