BIOCHEMICAL

Vol. 177, No. 3, 1991 June 28, 1991

AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 920-926

HOMOLOGY CLONING OF ASPARTIC PROTEASES FROM AN ENDOCRINE CELL LINE USING THE POLYMERASE CHAIN REACTION Nigel P. Birch and Y. Peng Loh Section on Cellular Neurobiology, Laboratory of Developmental Neurobiology, National Institute of Child Health and Human Development, National Institutes of Health, Bldg 36, Rm 2A21, Bethesda, MD 20892 Received

May 13,

1991

Summary: Oligonucleotides directed towards the active site regions of aspartic proteases were used as primers for the polymerase chain reaction to identify a unique sequence (asppcrl) from the AtT-20 anterior pituitary corticotrope cell line. Asppcrl showed the greatest similarity (85% identity) to human cathepsin E [(1989) J. Biol. Chem. 264,16748167531. Northern blot analysis of AtT-20 RNA revealed a single 1.9 kB message. Nuclease protection experiments indicated that asppcrl mRNA was present in pancreas, spleen, testis and liver at low levels and undetectable in heart and brain. This contrasted with the lysosomal aspartic protease, cathepsin D whose mRNA showed a broader tissue distribution. The restricted message distribution of asppcrl supports a more specific role for this aspartic protease in aspect(s) of cellular physiology. Q 1991Academic mess, 1°C.

Most endocrine and neuroendocrine active peptides endoproteolytic

peptides are synthesized as larger precursors.

are released in a series of post-translational

The

steps which include

cleavage (1,2). There has been considerable effort over the last 5 years to

purify and characterize the enzymes involved in the processing of these precursor proteins to the active peptides. To date, no mammalian unequivocal role in prohormone

endoproteases have been shown to have an

processing, although there are several promising candidates

(2). Several investigators have proposed a role for aspartic proteases in the maturation various prohormone

precursors. We have isolated and characterized an aspartic protease

from bovine pituitary glands and demonstrated pro-opiomelanocortin to be involved endothelin-1

of

a role for this enzyme in the processing of

and pro-vasopressin (3,4). An aspartic protease has been proposed

in the processing of ‘big’ endothelin-1

(5). Processing of anglerfish pro-somatostatin

to the potent vasoconstrictor II to somatostatin-28

has also

been proposed to involve an aspartic protease (6). The aspartic protease from bovine pituitary

secretory vesicles is present at very low levels.

proteases involved

in prohormone

maturation

It is likely that other aspartic are also present in very low amounts.

Therefore, we reasoned that possible amino acid similarities 0006-291X/91 $1.50 Copyright 0 1991 bv Academic Press. Inc. All rights of reproduction in any form restwed.

920

within the aspartic protease

Vol.

177,

No.

3, 1991

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

family may provide a tool to indirectly clone unique members of this family from endocrine cells.

The amino

determined

acid sequences of a large number

of aspartic proteases have been

and shown to contain areas of sequence similarity, particularly

active site aspartic residues (7). We designed degenerate oligonucleotide

surrounding the primers and used

the polymerase chain reaction to amplify first strand cDNA prepared from mRNA extracted from the anterior pituitary corticotrope

cell line AtT-20.

Amplified

DNA fragments were

characterized by sequence analysis.

MATERIALS

AND METHODS

Nucleic Acids

AtT-20/D16-16 cells were grown to approximately 70% of confluency in Dulbecco’s modified Eagle’s medium containing 10% fetal calf serum at 37°C in a humidified atmosphere of 10% CO,/90% air. Total RNA was prepared by centrifugation through a CsTFA cushion and poly(A+) RNA selected by two cycles of oligo(dT) chromatography as described in (8). Oligonucleotides were synthesized on an Applied Biosystems 380A oligonucleotide synthesizer. DNA Am&kation

First strand cDNA was prepared using the BRL cDNA synthesis system and 5 pg of poly (A+) RNA primed with 1 pg d(N)6 random primers (Boehringer Mannheim). Aliquots of the reverse transcriptase reaction (500 ng of cDNA) were amplified with primer concentrations of 5 PM and buffer and nucleotide concentrations as described in the Geneamp kit (Cetus) using a Cetus DNA thermal cycler. The PCR conditions were as follows: 94°C 1 min; 60°C 2 min; 72”C, 20 min for one cycle; 94”C, 40 set; 60°C 1 min; 72°C 2 min for 20 cycles; addition of an additional 5 units of Taq polymerase; 94°C 40 set, 60’32, 1 min; 72°C 2 min for 20 cycles. The amplified product was extracted with phenol/chloroform (1: l), chloroform/isoamylalcohol (24: l), ethanol precipitated and air dried. The pellet was resuspended in 20 ~1 TE and treated with 15 units of T4 DNA polymerase (BRL) and 30 units of T4 polynucleotide kinase (BRL) in the presence of 1.25 mM deoxyribonucleotides and 10 mM rATP for 30 min at 37°C. The blunt-ended and 5’ phosphorylated product was phenol extracted, precipitated with ethanol, electrophoresed on either a 5% native acrylamide gel or a 1.5% low melting point agarose gel and the DNA fragments recovered by electroelution or phenol extraction. Subclonina and Screeninp qf Amolified DNA

Fragments were ligated into M13mp18 or M13mp19 which had been cleaved with and dephosphorylated with calf intestinal alkaline phosphatase. Prior to sequencing insert-containing subclones were screened to exclude amplified cathepsin D clones using 20mer oligonucleotides specific for both the sense (5’ GAAGAACGGCACGTCCTITG 3’) and anti-sense orientation (5’ CAAAGGACGTGCCGTTCITC 3’) of mouse cathepsin D. Phage from individual recombinants were grown in 96 well rnicrotitre pIates and rephca stamps made onto nitrocellulose filters (Schleicher and Schuell). The filters were hybridized overnight with the cathepsin D-specific oligonucleotides labelled at the S-terminus with y32P-ATP (NEN 3000 Ci/mmol) and washed in 6X SSC (1X SSC = 0.015 mM sodium citrateHC1/150 mM NaCl, pH 7.0) followed by 3M tetramethylammonium chloride/50 mM Tris

ha-1

921

Vol.

177,

No.

BIOCHEMICAL

3, 1991

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

HCl, pH &0/2mM EDTA/l mg/ml NaDodS04 as described in (9) at 55°C. Subclones which failed to hybridize to either probe were either further characterized by single lane sequencing or sequenced directly by the dideoxy chain termination method (10) using modified T7 DNA polymerase ([ll], Sequenase, USB, Cleveland, OH). RNA Ana&sis

Poly(A’) RNA samples from AtT-20 cells or various mouse (CD-l) tissues were prepared as described or obtained from Clonetech Laboratories (Palo Alto, CA). AtT-20 RNA (A’, 3 pg) was electrophoresed in a 1.2% agarose gel containing 2.2 M formaldehyde and electrophoretically transferred to Nytran (Schleicher and Schuell). The filters were prehybridized for 4 hr in 50 mM phosphate buffer, pH 6.5/5X SSC/SO% formamide/5X Denhardts (1X = 200 pg/ml polyvinylpyrrolidone/200 pg/ml bovine serum albumin/200 pg/rnl Ficoll400)/250 pg/ml yeast tRNA/250 pg/ml denatured Salmon sperm DNA/O.l% SDS and hybridized for 18 hr in the same solution with random primer labelled “P-labelled asppcrl DNA Asppcrl DNA for random primer labelling and preparation of riboprobes was prepared by amplification of the appropriate Ml3 subclone with Ml3 sequencing and reverse sequencing primers (NEB) and Taq polymerase (amplification conditions: 94”C, 1 min; 5o”C, 1 min; 72”C, 2 min, 25 cycles). Blots were washed twice in 2X SSC/O.l% SDS for 20 min at room temperature followed by two 20 min washes in 0.5X SSC/O.l% SDS at 55°C. For nuclease protection experiments, asppcrl DNA or the mouse cathepsin D full length clone (isolated from the same library, Birch, unpublished data) was digested with @n-l followed by BamH-1 and size fractionated by agarose gel electrophoresis. The expected -300bp fragments were isolated and ligated into pGEM3zf(-) (Promega). DH5a cells (BRL) were transformed with the ligation mix and independent clones which contained each insert selected by lack of a-complementation. Plasmids were linearized with EcoR-1 and high specific activity RNA probes prepared using SP6 polymerase and [a-32P]-UTP (NEN, 3000 Ci/mMOL) as in (12) except the 10x transcription buffer was modified to contain 100 mM NaCl. Solution hybridization was performed (12) using 5 pg of poly A+ RNA or 5 pg yeast tRNA (control) and -50,000 c.p.m. of either the cathepsin D- or asppcrl-specific probes. Samples were hybridized for 16 h at 46°C before nuclease digestion and analysis on 6% acrylamide/urea sequencing gels.

RESULTS

AND DISCUSSION

Conservation

of the amino acid sequences around the active site aspartic acid

residues of a large number of aspartic proteases allowed us to design two degenerate oligonucleotide

primers in an attempt to clone unique aspartic proteases from endocrine

cells (Fig. 1A). Using these primers and the polymerase chain reaction, we amplified

a -600

bp fragment (Fig. 1B) which was the expected size based on the number of amino acids between the two active site aspartic residues of all members of the aspartic protease family so far cloned (7).

Among other possible sequences, we expected the amplified

DNA

fragment to contain the sequence for cathepsin D, an aspartic protease found in lysosomes and shown to have a broad tissue distribution

commensurate

with its proposed role in

general protein catabolism (13). To exclude this cDNA fragment from the amplified 922

600

Vol.

177,

No.

3, 1991

BIOCHEMICAL

A

5'

CONSENSUS

PRIMERS

BIOPHYSKAL

AP-1

3’

RESEARCH

3’

TTTCACACICCTTCITCCAC --C--T-----C---C-T---

D 0 D

COMMUNICATIONS

u-2

5’

CCCACCCACCCIIAICATCTC -----ACTT-e----A-e---

5’

CODON USAGE

hRenin r!Aenin mRenln hPepsinogenC rPepslnogenC bchymosin hcathepsin rcathepsin mcathepsin

AND

3’

5’

3’

TTTCACACTCCTTCCTCCAAC --C--T--C--C--AC-TC-T --------c-----c------

CCCTCCCTCCCATTCCTACAC --T--TAAA-TCC-CT-C--------C---CTA-T------

rDTCBSil -----A-----*-

CCLALVD --&f-----QCI---Q-IL-

-

-

-

_ -

D -

AVV-* -

1

-

-

E E E

I I I

-

-

-

B

FiFre 1. Design of active site-directed aspartic protease primers (A) and their use to amplify first strand cDNA from AtT-20 cells (B). The amino acid and DNA sequences of aspartic proteases (14-16, 19-22) were used to design primers AP-1 and AP-2. Following amplification using the polymerase chain reaction, the products were separated by agarose gel electrophoresis (lane 1). The migration positions of DNA size markers are shown in lane 2.

bp ‘pool’, we hybridized

filter replicas from 96 well plates which contained

individual

subclones of the amplified

600 bp fragment ‘pool’ with sense and anti-sense oligonucleotides

specific for mouse cathepsin D. From a total of 279 subclones three subclones failed to hybridize to these oligonucleotides. and anti-sense orientations sequence to other members similarity

These were sequenced and found to represent sense

of the same clone.

Comparison

of the translated amino acid

of the aspartic protease family revealed a high degree of

which supported its classification

as an aspartic protease (Fig. 2). A computer

search of the Genbank data base (Release 62) indicated the highest similarity was with the pepsinogen family. Asppcrl shared 50% identity and 72% similarity (including conservative substitutions) predicted

with rat pepsinogen C at the amino acid level (14).

asppcrl

amino acid sequence with mouse submaxillary

cathepsin D showed a 64% and 65% similarity

923

respectively.

However,

Comparison

of the

gland renin (15) and during the course of

Vol.

177,

No.

3, 1991

BIOCHEMICAL

AND

130 ASPPCRl

(BKl”SE)

CATHEPSTN

E

BIOPHYSICAL

110

RESEARCH

150

160

COMMUNICATIONS

no

180

*

YLSSDPQ--GGS-GSELITFFOYDPSHFSGSLNWIP”TKQAYWQIALDGIQ”GDTVM--FCSEOCQALVD (HUMAN)

PEPSnKJGEN c (RAT, RSNIN (MOVSE) CALTXEPSTN D ,HO"SE,

YMSSNPE--GGA-GSELIFGGYDHSHFSGSLNWVP"TKQAYWPIALDNIQVGGTVM--FCSEGC~AIVD YLGSQ----QGSNGGQIVFWVDKNLYTGEITWVPVTQELYWPITIDDFLIGDQASGWCSSQGCQGIVD YYNRGPHL---LGGEW-LGOSDPEHYQFHYVSLSKTDSIPITMKGV PLNRDPEG-QP-GC;ELM-LOGPDSKYYHOELSYLNVTRKAYWP"HMDQLE"GNELT--LCKGGCEAIVD

Fimre 2. Comparison of the deduced amino acid sequence of mouse asppcrl with other aspartic proteases (14-16). The two active site aspartic residues are marked by asterisks. Amino acids which are identical between all the aspartic proteases are hatched. Dashes represent gaps introduced to give optimal sequence alignment.

this study, the sequence of a new aspartic protease, human cathepsin E, was reported (16) and proved to have 85% identity and 90% similarity Northern

with asppcrl at the amino acid level.

blot analysis using random primer-labelled

asppcrl indicated

a single

message in AtT-20 cells of 1.9 kI3 (Fig. 3). This was in contrast to the human cathepsin E cDNA which detected 3 messages in a human gastric adenocarcinoma

cell line of 3.4 kB,

2.6 kB and 2.1 kB but only a single message for cathepsin D, as described by other workers (13).

The presence of three messages for cathepsin E is unique amongst the aspartic

proteases currently cloned. Recently, however, N-terminal purified human gastric cathepsin E demonstrated

amino acid sequence analysis of

that there are at least two isozymes, and

hence at least two genes (17). The significance of the differences in message number and size between the gastric adenocarcinoma

and AtT-20 cell lines is not yet clear.

The

expression of a single message in AtT-20 cells may represent the tissue-specific expression of a cathepsin E isozyme. Alternatively,

mouse asppcrl, although closely related to human

cathepsin E, may be the gene product of a unique aspartic protease. Preliminary blot analyses of the tissue distribution

of asppcrl indicated that message levels were very

low (data not shown). We therefore mapped the distribution more sensitive nuclease protection

northern

methodology

of asppcrl message using the

and compared the distribution

to that of

cathepsin D. The highest levels of mRNA protected by the asppcrl cRNA probe were seen in the pancreas with lower levels in the liver and testis (Fig. 4A). Spleen also contained comparable message abundance to that seen in the pancreas (data not shown). &ppcrl mRNA was undetectable

in brain and heart (Fig.4A).

In contrast, the mouse cathepsin D

cRNA probe protected high levels of message in testis, spleen, liver and brain and lower 924

177,

No.

3,1991

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

2.37

03 Figure 3, Northern blot analysis of poly (A’) AtT-20 mRNA probed with asppcr 1 cDNA probe. The migration positions of RNA size markers (BRL) are also shown. Firmre 4. Distribution of asppcrl (A) and cathepsin D (B) mRNA in various tissues as determined by RNA nuclease protection.

levels in the heart and pancreas (Fig.4B).

The broad distribution

of the cathepsin D mFWA

is consistent with its proposed role in the general intracellular

degradation

of proteins.

However, the more tissue-specific expression of asppcrl mRNA suggests that it may have a more specific cellular function. asppcrl in prohormone relationship

processing.

The observed distribution

does not exclude a role for

A function for asppcrl based on its close structural

to cathepsin E cannot be implied

as the physiological

role(s) for cathepsin E

is not yet known. However, subcellular fractionation studies of rat neutrophils demonstrated 925

Vol.

177,

No.

3, 1991

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

that cathepsin E is not targeted to lysosomes (18). Assignment of a definitive function for asppcrl will await the cloning, expression and functional characterization

of a full length

cDNA. ACKNOWLEDGMENTS We would like to thank Dr. M.J. Brownstein for many valuable discussions and for the synthesis of the oligonucleotides.

We would also like to thank Dr. J. Battey for helpful

advice. REFERENCES 1. Loh, Y.P., Brownstein, MJ., and Gainer, H. (1984) Annu. Rev. Neurosci. 7, 189-222. 2. Mains, R.E., Dickerson, I.M., May, V., Stoffers, D.A., Perkins, S.N., Ouafik, L., Husten, E.J. and Eipper, B.A. (1990) Front. Neuroendocrinol. 11, 52-89. 3. Loh, Y.P., Parish, D.C. and Tuteja, R. (1985) J. Biol. Chem. 260, 7194-7205. 4. Parish, D.C., Tuteja, R., Alstein, M., Gainer, H., and Loh, Y.P. (1986) J. Biol. Chem. 261, 14392-14397. 5. Sawamura, T., Kimura, S., Shimni, O., S igita, Y., Yanagisawa, M., Goto, K. and Masaki, T. (1990) Biochem. Biophys. Res. Comnun. 168, 1230-1236. 6. Ma&in, R.B., Noe, B.D., and Spiess, J. (1990) The Endocrine Society 72nd Annual Meeting, Ab. No. 1259. 7. Tang, J. and Wong R.N.S. (1987) J. Cell. Biochem. 33, 55-63. 8. Okayama, H., Kawaichi, M., Brownstein, M., Lee, F., Yokota, T., and Arai, IS. (1987) Methods Enzymol. 154, 3-28. 9. Wood, W.I., Gitschier, J., La&y, L.A., and Lawn, R.M. (1985) Proc. Natl. Acad. Sci, USA 82, 1585-1588. 10. Sanger, F., Nicklen, S. and Coulson, A.R. (1977) Proc. Natl. Acad. Sci USA 74, 54635467. 11. Tabor, S. and Richardson, C.C. (1987) Proc. Natl. Acad. Sci USA 84,4767-4771. 12. Krieg, P.A. and Melton, D.A. (1987) Methods Enzymol. 155, 397-415. 13. Faust, P.L., Kornfeld, S. and Chirgwin, J.M. (1985) Proc. Natl. Acad. Sci. USA 82,49104914. 14. Ichihara, Y., Sogawa, K., Morohashi, K., Fujii-Kuriyama, Y. and Takahashi, K. (1986) Eur. J. Biochem. 161, 7-12. 15. Panthier, J.J. and Rougeon, F. (1983) EMBO J. 2,675-678. 16. Azuma, T., Pals, G., Mohandas, T.K., Couvreur, J.M. and Taggart, R.T. (1989) J. Biol. Chem. 264, 16748-16753. 17. Athauda, S.B.P., Matsuzaki, O., Kageyama, T. and Takahashi, K. (1990) Biochem. Biophys. Res. Commun. 168, 878-885. 18. Yonezawa, S., Fujii, K., Maejima, Y., Tamoto, K., Mori, Y. and Muto, N. (1988) Arch Biochem. Biophys. 267, 176-183. 19. Imai, T., Miyazaki, H., Hirose, S., Hori, H., Hayashi, T., Kageyama, R., Ohkubo, H., Nakanishi, S. and Murakami, K. (1983) Proc. Natl. Acad. Sci. USA 80, 7405-7409. 20. Burnham, C.E., Hawelu-Johnson, C.L., Frank, B.M., and Lynch, K.R. (1987) Proc. Natl. Acad. Sci. USA 84, 5605-5609. 21. Taggart, R.T., Cass, L.G., Mohandas, T.K., Derby, P., Barr, P.J., Pals, G. and Bell, G.I. (1989) J. Biol. Chem. 264, 375-379. 22. Harris, T.J.R., Lowe, P.A., Lyons, A., Thomas, P.G., Eaton, M.A.W., Millican, T-A., Patel, T.P., Bose, CC., Carey, N.H. and Doel, M.T. (1982) Nucl. Acids Res. 10, 21772187. 926