0o1:~-7227/9~/1315-2287$03.00/0 Endocrinology Copyright 8 1992 by The Endocrine
Posttranslational Cells: Evidence JOHN
P. MATHIS
AND
Vol. 131, No. 5 Prmted ,n U.S.A.
Society
IRIS
Processing for Cleavage
of Proenkephalin in AtT-20 at a Lys-Lys Site*
LINDBERGt
Department of Biochemistry and Molecular New Orleans, Louisiana 70112
Biology, Louisiana
State University
Medical
Center,
ABSTRACT Proteolytic processing of proenkephalin was examined in sever&l subclones -of AiTcel& stably transfected with rat proenkephalin cDNA (AT/PE cells). Proenkephalin is svnthesized in both N-elvcosylated ‘and unglycosylated for&s, as demonstrated by treatmen; &ith tunicamycin. RIAs and Western blot studies showed that AT/PE clones process proenkephalin at some, but not all, Lys-Arg sequences in a limited processing profile reminiscent of bovine adrenal chromaffin cells. Pulse-chase studies using Met’-enkephalin-Arg-Gly-Leu antiserum demonstrated that 50% of the precursor is processed within 1 h, and processing is complete after 2.5 h with the production of the 5.3. kilodalton (kDa) peptide. Further cleavage to the octapeptide Met’enkephalin-Arg-Gly-Leu is minimal. Radiosequencing results verified
the efficient cleavage of a Lys-Lys site within proenkephalin that resulted in the production of the 5.3-kDa peptide. Proenkephalin cleavage products stored within cells, which included the 5.3-kDa peptide, cbuld be released upon stimulation of cells with BaCl, (2.fold above basal levels). 8-bromo-CAMP or CRF (7- and &fold above basal levels, respectivelyj, and a mixture of BaCl,‘and 8-bromo-CAMP (20. fold above basal levels). An important difference between the processing of proenkephalin and the ACTH/endornhin orecursor (POMC) in ACT-26 cells ii efficient cleavage of’a LyslLys site in proknkephalin and not in POMC. The ability of AT/PE to process proenkephalin in a natural manner makes it a suitable model system to investigate elements involved in the processing of proenkephalin at Lys-Lys sites. (Endocrinology 131: 2287-2296,1992)
T
Arg-Arg and Arg-Lys pairs of basic amino acids are also efficiently processed by the proteolytic machinery in AtT-20 cells. However, a Lys-Lys substitution resulted in inefficient or incomplete processing (14). It has been speculated that the apparent inability of AtT-20 cells to cleave Lys-Lys sites might result from the selective specificity of endogenous proteolytic processing enzymes, possibly PC1 (12, 13). The opioid peptide precursor proenkephalin contains multiple pairs of Lys-Lys sites that must be cleaved to reveal the penta- to octapeptide enkephalins; as such, it represents an unusual example of frequent naturally occurring Lys-Lys cleavages. Previous studies have shown that AtT-20 cells are able to cleave transfected proenkephalin; one early study using cell lines stably transfected with human proenkephalin cDNA provided evidence that the proenkephalin was completely processed to Met5-enkephalin (Met5-enk) and Leu5enkephalin (Leu’-enk) (15). However, a later study, involving transient transfections of AtT-20 cells with vaccinia virus harboring proenkephalin cDNA, showed a much more limited cleavage pattern at similar expression levels (16). In neither study were specific cleavage sites addressed. In the present work we have examined the biosynthesis and posttranslational processing of proenkephalin in stably transfected AtT-20 clones. Our data indicate that these cells are able to process this precursor at Lys-Lys sites, and that the pattern of limited processing observed is not unlike that present in the bovine adrenal medulla.
HE NEUROENDOCRINE cell line AtT-20, derived from a mouse anterior pituitary tumor, has often served as a useful system to study the cellular requirements for posttranslational processing of neuropeptide precursors (1). This cell line, which possesses a regulated secretory pathway, contains all the necessary components required for the processing of POMC (2), proinsulin (3), prorenin (4), prosomatostatin (5), and proneuropeptide-Y (pro-NPY) (6) to smaller peptide products. AtT-20 cells have been shown to contain carboxypeptidase-E (7), peptidyl cr-amidating enzyme (8), and the new subtilisin-like prohormone convertase (PC) PCl, also known as PC3 (9-11). While small amounts of the related enzyme PC2 have also been reported in this cell line (9), mRNA for PC1 is by far more abundant (9-11). The involvement of PC1 in the physiological processing of POMC was recently demonstrated by antisense blockade experiments, in which clones expressing low levels of PC1 exhibited correspondingly diminished proteolytic processing of POMC (11). In vaccinia virus transfection experiments, PC1 has been shown to be more selective than PC2; PC1 cleaves POMC to produce ACTH and /3-lipotropin, but PC2 produces P-endorphin (P-end), oc-melanotropin, and an ACTH-containing intermediate from POMC (12, 13). In transfection studies, AtT-20 cells correctly process proNPY at the wild-type Lys-Arg site (6, 14). Substitutions of Received May 20, 1992. Address all correspondence and requests for reprints to: Dr. Iris Lindberg, Department of Biochemistry and Molecular Biology, Louisiana State University Medical Center, 1901 Perdido Street, New Orleans, Louisiana 70112. * This work was supported by NIDDK Grant 35199. t Supported by a Research Career Development Award from the NIDDK.
Materials Expression
plasmid
and Methods
construction
Rat proenkephalin cDNA rENK (18) by Bg111 digestion.
(17) was excised from the plasmid pEV/ After digestion, the insert was purified by
2287
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:45 For personal use only. No other uses without permission. . All rights reserved.
PROCESSING
2288
OF PROENKEPHALIN
agarose gel electrophoresis, phenol extraction, and ethanol precipitation. The insert was then ligated into pBluescript SK+ (Stratagene, La Jolla, CA) previously digested with BglII and dephosphorylated. The ligation mixture was used to transform competent E. coli (strain JMlOl), and recombinants were identified by blue/white selection. Insert orientation was determined by asymmetric digests using StuI and BglII, and all recombinants were found to contain the insert in the inverse orientation. One such plasmid (termed p2J7) was selected for further manipulations. p2J7 was digested with XbaI and Hind111 to yield a rat proenkephalin cDNA insert with a 5’ Hind111 end and a 3’ XbaI end. The insert was purified by agarose gel electrophoresis, phenol extraction, and ethanol precipitation and ligated into pRcCMV(+) (Invitrogen; previously digested with XbaI and Hind111 and dephosphorylated). Competent E. co/i (strain JM101) were transformed with the ligation mixture, and recombinants were identified by blue/white selection. The orientation of the insert was determined by asymmetric digests using StuI and XbaI, and several recombinants with a correctly oriented insert were identified. One of these recombinants was selected for use as the expression plasmid and termed pCMV/ENK. The plasmid was produced in l-liter cultures of E. coli and isolated by standard procedures. Before use in transfection procedures, the plasmid was purified twice by centrifugation through CsC12 gradients.
Cell culture and DNA
transfection
AtTm20/dv16 cells (obtained from R. E. Mains and 8. A. Eipper) were grown in Dulbecco’s Modified Eagle’s Medium (DMEM; Gibco, Grand Island, NY) containing 2.5% fetal bovine serum (Irvine Scientific, Santa Ana, CA) and 10% NuSerum (Irvine Scientific) at 37 C in 5% CO* and were split at a ratio of 1:20 twice weekly. The cells were transfected with the expression plasmid pCMV/ENK by the Lipofectin method (BRL, Gaithersburg, MD) (19). After removal of transfection medium, cells were trypsinized and distributed among 5 96-well plates. Two months later, stable transfectants resistant to the neomycin analog G418 (0.6 mg/ml; -50% active, Gibco/BRL) were screened for proenkephalin production by RIA; the level of expression ranged from 6.72-39.1 pmol Met5-Argh-Phe7-enkephalin (Met5-enk-Arg-Phe) immunoreactive peptides/lo6 cells, From these clones, 3, representing 2 medium (AT/PE 18 and AT/PE 32, 13.3 and 11.5 pmol Mets-enk-Arg-Phe/lOb cells, respectively) and 1 high expressing line (AT/PE 43; 25.8 pmol Met’-enk-ArgPhe/106 cells) were selected for further study. For Western blot and immunoprecipitation experiments, cells were grown in 35.mm wells for 2-4 days before use. For RIA, cells were grown in lo-cm dishes for 2-4 days before use. Cell extracts were collected by three methods. For Western blotting, a subconfluent well was homogenized directly in 300 ~1 Laemmli sample buffer (20) and stored at -20 C until use. For RIA, cells were homogenized in 1 ml 1 N acetic acid with 50 rnM HCl and 0.1% (vol/vol) @-mercaptoethano1, and insoluble material was removed by centrifugation. For immunoprecipitation, cells were homogenized in 1 ml 5 N acetic acid with 0.3 mM phenylmethylsulfonylfluoride (I’MSF), and insoluble material was removed by centrifugation. Recoveries of added [?S]methionine-labeled proenkephalin using the two acid homogenization solutions were comparable (>90%). The supernatant was lyophilized and stored at -20 C until analysis.
Immunoprecipitation Subconfluent 35.mm wells were used for individual time points in pulse-chase experiments. Each well was washed three times with 5-ml aliquots of warmed PBS. One milliliter of methionine/cysteine-deficient medium (ICN, Costa Mesa, CA) containing 1 mCi [35S]Translabel (average SA, 1190 Ci/mmol; >70% methionine and ~15% cysteine; ICN) was then added to each well. Cells were incubated at 37 C in 5% CO* for 20 min or 6 h. After labeling, cells were chased by incubation in DMEM with 2% well dialyzed, heat-inactivated fetal bovine serum for a period of 0, 0.5, 1, 2, or 6 h. At the end of the chase period, cells were homogenized in 1 ml 5 N acetic acid with 0.3 mM PMSF, frozen and thawed, and insoluble material was removed by centrifugation. The supernatant’was lyophilized and stored at -20 C until immunoprecipitated.
Endo. Voll31.
1992 No 5
For the tunicamycin experiment, a subconfluent 35-mm well was incubated with tunicamycin (5 pg/ml; Boehringer Mannheim, Indianapolis, IN) for 10 h at 37 C in 5% CO’. After this incubation, cells were washed three times with 5-ml aliquots of warmed PBS. One milliliter of methionine/cysteine-deficient medium (ICN) containing 1 mCi [35S] Translabel (average SA, 1190 Ci/mmol; >70% methionine and
Posttranslational Cells: Evidence JOHN
P. MATHIS
AND
Vol. 131, No. 5 Prmted ,n U.S.A.
Society
IRIS
Processing for Cleavage
of Proenkephalin in AtT-20 at a Lys-Lys Site*
LINDBERGt
Department of Biochemistry and Molecular New Orleans, Louisiana 70112
Biology, Louisiana
State University
Medical
Center,
ABSTRACT Proteolytic processing of proenkephalin was examined in sever&l subclones -of AiTcel& stably transfected with rat proenkephalin cDNA (AT/PE cells). Proenkephalin is svnthesized in both N-elvcosylated ‘and unglycosylated for&s, as demonstrated by treatmen; &ith tunicamycin. RIAs and Western blot studies showed that AT/PE clones process proenkephalin at some, but not all, Lys-Arg sequences in a limited processing profile reminiscent of bovine adrenal chromaffin cells. Pulse-chase studies using Met’-enkephalin-Arg-Gly-Leu antiserum demonstrated that 50% of the precursor is processed within 1 h, and processing is complete after 2.5 h with the production of the 5.3. kilodalton (kDa) peptide. Further cleavage to the octapeptide Met’enkephalin-Arg-Gly-Leu is minimal. Radiosequencing results verified
the efficient cleavage of a Lys-Lys site within proenkephalin that resulted in the production of the 5.3-kDa peptide. Proenkephalin cleavage products stored within cells, which included the 5.3-kDa peptide, cbuld be released upon stimulation of cells with BaCl, (2.fold above basal levels). 8-bromo-CAMP or CRF (7- and &fold above basal levels, respectivelyj, and a mixture of BaCl,‘and 8-bromo-CAMP (20. fold above basal levels). An important difference between the processing of proenkephalin and the ACTH/endornhin orecursor (POMC) in ACT-26 cells ii efficient cleavage of’a LyslLys site in proknkephalin and not in POMC. The ability of AT/PE to process proenkephalin in a natural manner makes it a suitable model system to investigate elements involved in the processing of proenkephalin at Lys-Lys sites. (Endocrinology 131: 2287-2296,1992)
T
Arg-Arg and Arg-Lys pairs of basic amino acids are also efficiently processed by the proteolytic machinery in AtT-20 cells. However, a Lys-Lys substitution resulted in inefficient or incomplete processing (14). It has been speculated that the apparent inability of AtT-20 cells to cleave Lys-Lys sites might result from the selective specificity of endogenous proteolytic processing enzymes, possibly PC1 (12, 13). The opioid peptide precursor proenkephalin contains multiple pairs of Lys-Lys sites that must be cleaved to reveal the penta- to octapeptide enkephalins; as such, it represents an unusual example of frequent naturally occurring Lys-Lys cleavages. Previous studies have shown that AtT-20 cells are able to cleave transfected proenkephalin; one early study using cell lines stably transfected with human proenkephalin cDNA provided evidence that the proenkephalin was completely processed to Met5-enkephalin (Met5-enk) and Leu5enkephalin (Leu’-enk) (15). However, a later study, involving transient transfections of AtT-20 cells with vaccinia virus harboring proenkephalin cDNA, showed a much more limited cleavage pattern at similar expression levels (16). In neither study were specific cleavage sites addressed. In the present work we have examined the biosynthesis and posttranslational processing of proenkephalin in stably transfected AtT-20 clones. Our data indicate that these cells are able to process this precursor at Lys-Lys sites, and that the pattern of limited processing observed is not unlike that present in the bovine adrenal medulla.
HE NEUROENDOCRINE cell line AtT-20, derived from a mouse anterior pituitary tumor, has often served as a useful system to study the cellular requirements for posttranslational processing of neuropeptide precursors (1). This cell line, which possesses a regulated secretory pathway, contains all the necessary components required for the processing of POMC (2), proinsulin (3), prorenin (4), prosomatostatin (5), and proneuropeptide-Y (pro-NPY) (6) to smaller peptide products. AtT-20 cells have been shown to contain carboxypeptidase-E (7), peptidyl cr-amidating enzyme (8), and the new subtilisin-like prohormone convertase (PC) PCl, also known as PC3 (9-11). While small amounts of the related enzyme PC2 have also been reported in this cell line (9), mRNA for PC1 is by far more abundant (9-11). The involvement of PC1 in the physiological processing of POMC was recently demonstrated by antisense blockade experiments, in which clones expressing low levels of PC1 exhibited correspondingly diminished proteolytic processing of POMC (11). In vaccinia virus transfection experiments, PC1 has been shown to be more selective than PC2; PC1 cleaves POMC to produce ACTH and /3-lipotropin, but PC2 produces P-endorphin (P-end), oc-melanotropin, and an ACTH-containing intermediate from POMC (12, 13). In transfection studies, AtT-20 cells correctly process proNPY at the wild-type Lys-Arg site (6, 14). Substitutions of Received May 20, 1992. Address all correspondence and requests for reprints to: Dr. Iris Lindberg, Department of Biochemistry and Molecular Biology, Louisiana State University Medical Center, 1901 Perdido Street, New Orleans, Louisiana 70112. * This work was supported by NIDDK Grant 35199. t Supported by a Research Career Development Award from the NIDDK.
Materials Expression
plasmid
and Methods
construction
Rat proenkephalin cDNA rENK (18) by Bg111 digestion.
(17) was excised from the plasmid pEV/ After digestion, the insert was purified by
2287
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:45 For personal use only. No other uses without permission. . All rights reserved.
PROCESSING
2288
OF PROENKEPHALIN
agarose gel electrophoresis, phenol extraction, and ethanol precipitation. The insert was then ligated into pBluescript SK+ (Stratagene, La Jolla, CA) previously digested with BglII and dephosphorylated. The ligation mixture was used to transform competent E. coli (strain JMlOl), and recombinants were identified by blue/white selection. Insert orientation was determined by asymmetric digests using StuI and BglII, and all recombinants were found to contain the insert in the inverse orientation. One such plasmid (termed p2J7) was selected for further manipulations. p2J7 was digested with XbaI and Hind111 to yield a rat proenkephalin cDNA insert with a 5’ Hind111 end and a 3’ XbaI end. The insert was purified by agarose gel electrophoresis, phenol extraction, and ethanol precipitation and ligated into pRcCMV(+) (Invitrogen; previously digested with XbaI and Hind111 and dephosphorylated). Competent E. co/i (strain JM101) were transformed with the ligation mixture, and recombinants were identified by blue/white selection. The orientation of the insert was determined by asymmetric digests using StuI and XbaI, and several recombinants with a correctly oriented insert were identified. One of these recombinants was selected for use as the expression plasmid and termed pCMV/ENK. The plasmid was produced in l-liter cultures of E. coli and isolated by standard procedures. Before use in transfection procedures, the plasmid was purified twice by centrifugation through CsC12 gradients.
Cell culture and DNA
transfection
AtTm20/dv16 cells (obtained from R. E. Mains and 8. A. Eipper) were grown in Dulbecco’s Modified Eagle’s Medium (DMEM; Gibco, Grand Island, NY) containing 2.5% fetal bovine serum (Irvine Scientific, Santa Ana, CA) and 10% NuSerum (Irvine Scientific) at 37 C in 5% CO* and were split at a ratio of 1:20 twice weekly. The cells were transfected with the expression plasmid pCMV/ENK by the Lipofectin method (BRL, Gaithersburg, MD) (19). After removal of transfection medium, cells were trypsinized and distributed among 5 96-well plates. Two months later, stable transfectants resistant to the neomycin analog G418 (0.6 mg/ml; -50% active, Gibco/BRL) were screened for proenkephalin production by RIA; the level of expression ranged from 6.72-39.1 pmol Met5-Argh-Phe7-enkephalin (Met5-enk-Arg-Phe) immunoreactive peptides/lo6 cells, From these clones, 3, representing 2 medium (AT/PE 18 and AT/PE 32, 13.3 and 11.5 pmol Mets-enk-Arg-Phe/lOb cells, respectively) and 1 high expressing line (AT/PE 43; 25.8 pmol Met’-enk-ArgPhe/106 cells) were selected for further study. For Western blot and immunoprecipitation experiments, cells were grown in 35.mm wells for 2-4 days before use. For RIA, cells were grown in lo-cm dishes for 2-4 days before use. Cell extracts were collected by three methods. For Western blotting, a subconfluent well was homogenized directly in 300 ~1 Laemmli sample buffer (20) and stored at -20 C until use. For RIA, cells were homogenized in 1 ml 1 N acetic acid with 50 rnM HCl and 0.1% (vol/vol) @-mercaptoethano1, and insoluble material was removed by centrifugation. For immunoprecipitation, cells were homogenized in 1 ml 5 N acetic acid with 0.3 mM phenylmethylsulfonylfluoride (I’MSF), and insoluble material was removed by centrifugation. Recoveries of added [?S]methionine-labeled proenkephalin using the two acid homogenization solutions were comparable (>90%). The supernatant was lyophilized and stored at -20 C until analysis.
Immunoprecipitation Subconfluent 35.mm wells were used for individual time points in pulse-chase experiments. Each well was washed three times with 5-ml aliquots of warmed PBS. One milliliter of methionine/cysteine-deficient medium (ICN, Costa Mesa, CA) containing 1 mCi [35S]Translabel (average SA, 1190 Ci/mmol; >70% methionine and ~15% cysteine; ICN) was then added to each well. Cells were incubated at 37 C in 5% CO* for 20 min or 6 h. After labeling, cells were chased by incubation in DMEM with 2% well dialyzed, heat-inactivated fetal bovine serum for a period of 0, 0.5, 1, 2, or 6 h. At the end of the chase period, cells were homogenized in 1 ml 5 N acetic acid with 0.3 mM PMSF, frozen and thawed, and insoluble material was removed by centrifugation. The supernatant’was lyophilized and stored at -20 C until immunoprecipitated.
Endo. Voll31.
1992 No 5
For the tunicamycin experiment, a subconfluent 35-mm well was incubated with tunicamycin (5 pg/ml; Boehringer Mannheim, Indianapolis, IN) for 10 h at 37 C in 5% CO’. After this incubation, cells were washed three times with 5-ml aliquots of warmed PBS. One milliliter of methionine/cysteine-deficient medium (ICN) containing 1 mCi [35S] Translabel (average SA, 1190 Ci/mmol; >70% methionine and
