Expression of a Peptide Processing Enzyme in Cultured Cells: Truncation Mutants Reveal a Routing Domain

Francisco A. Tausk, Betty A. Eipper

Sharon

L. Milgram,

Richard

Departments of Neuroscience (F.A.T., S.L.M.. anh Dermatology (F.A.T.) ’ Johns Hopkins University School of Medicine Baltimore, Maryland 21205

E. Mains,

R.E.M.,

Peptidylglycine cy-amidating monooxygenase (PAM) is a bifunctional enzyme responsible for the cu-amidation of peptides in secretory granules of neuroendocrine cells. The single gene encoding PAM undergoes tissue-specific alternative splicing and endoproteolytic processing to generate bifunctional membrane proteins with a single transmembrane domain as well as soluble proteins that are monoor bifunctional. In order to examine the endoproteolytic processing and subcellular localization of the various forms of PAM in cells lacking regulated secretory granules, we established stably transfected hEK-293 cell lines expressing naturally occurring and mutant forms of PAM. As expected, newly synthesized soluble PAM proteins were rapidly secreted into the medium. Integral membrane protein forms of PAM were largely localized in the perinuclear region with punctate staining visible throughout the cell and 2-5% of the enzyme activity detectable on the cell surface. Bifunctional PAM proteins were slowly released into the medium after expression of integral membrane protein forms of PAM. Deletion of 77 amino acids from the COOHterminus of the integral membrane forms of PAM resulted in a membrane-bound protein which retained both enzymatic activities but accumulated on the cell surface. Rapid internalization of full-length PAM proteins was observed by incubating live cells with antiserum to PAM; deletion of the COOH-terminal domain eliminated the ability of cells to internalize PAM. Thus the cytoplasmic domain of integral membrane PAM contains a routing determinant recognized by cells lacking the regulated secretory pathway. (Molecular Endocrinology 6: 2185-2196, 1992)

0666.6609/92/2165-2196$03.00/O Molecular Endocrmology Copyright 0 1992 by The Endcane

B.A.E.)

standing of the mechanisms underlying the accumulation of peptides and enzymes in secretory granules is limited. When expressed in heterologous cell lines lacking regulated secretory granules, prohormones are generally secreted rapidly without undergoing any significant amount of endoproteolytic cleavage subsequent to removal of the signal sequence (1, 2). To determine whether peptide processing enzymes might contain signals recognized by nonneuroendocrine cells, we expressed cDNAs encoding peptidylglycine a-amidating monooxygenase (PAM; EC 1 .14.17.3), which occurs naturally in both soluble and integral membrane protein forms, in cells lacking regulated secretory granules. PAM is a bifunctional enzyme responsible for the copper-, ascorbate-, and molecular oxygen-dependent a-amidation of glycine-extended peptides (3); these substrate peptides are usually biologically inactive. Peptidylglycine ru-hydroxylating monooxygenase (PHM) catalyzes the first step of the reaction, while peptidyl-ochydroxyglycine cu-amidating lyase (PAL) catalyzes the second step (3). Tissue-specific alternative splicing of the single copy PAM gene produces mRNAs encoding forms of PAM predicted to be bifunctional integral membrane proteins as well as soluble mono- and bifunctional proteins (Fig. 1) (3, 4). The largest PAM protein, rPAM1, consists of a signal and propeptide followed by the PHM catalytic domain, a noncatalytic domain referred to as exon A, the PAL catalytic domain, a transmembrane domain, and a COOH-terminal domain (5). When expressed in AtT-20 cells, a neuroendocrine cell line in which peptide products are stored in regulated secretory granules, all of the PAM proteins were subjected to endoproteolytic processing (8). The soluble forms of PAM were routed to secretory granules, and expression of integral membrane forms of PAM led to an accumulation of PAM proteins in a perinuclear position in addition to secretory granules (8). We recently demonstrated that integral membrane forms of PAM reached the surface of AtT-20 cells, from which they were rapidly internalized and returned to the perinuclear region. Truncated forms of PAM retaining the transmembrane domain but lacking the COOH-terminal domain failed to undergo internalization, resulting in an

INTRODUCTION While mature secretory granules product peptides and processing

and

are the repository of enzymes, our under-

Society

2185 The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 06:10 For personal use only. No other uses without permission. . All rights reserved.

MOL 2186

ENDO.

Membrane

1992

Vol6No.12

Predicted Molecular Mass ProPAM ( kDa )

forms

PHM

PAL o C-TERM PAM-l

106

PAM-Z

94

PAM-3

85

PAM-4

56

PAL-s

49

secretion of wild type and mutant PAM proteins by stable cell lines were evaluated by measuring PHM and PAL activity, biosynthetic labeling followed by immunoprecipitation, Western blot analysis, and immunofluorescence. These studies have led to the recognition that although very little endoproteolytic processing of PAM occurs in nonneuroendocrine cells, the noncatalytic COOH-terminal domain of PAM is recognized by these cells and is essential for the efficient internalization of PAM from the cell surface.

Soluble forms RM

PHM-s ISignaVpropeptide

?N-glycosylation

site

[Transmembrane

RESULTS Expression Transfected

40 domain

Fig. 1. PAM Proteins

Expressed in Transfected hEK-293 Ceils The region designated signal/propeptide consists of the initial 25 amino acids constituting the signal peptide and the propeptide (amino acids 26-35), and is present in all constructions. The PHM domain extends from amino acid 36-392 and is truncated at Lys 383 in PHM-s. Exons A and B include amino acids 393-497 and 832-917, respectively. The PAL domain extends from amino acid 498-831, and PAL-s is truncated at Glyae4 in exon B. The transmembrane domain is indicated by a black bar within exon B. The COOH-terminal domain (CTERM), extends from amino acid 900-976 and resides in the cytoplasm when part of PAM-l or PAM-2 and within the lumen of the secretory pathway when part of PAM-3 (5). PAM-4 terminates with a sequence of 20 amino acids that is unique to this form of PAM. All amino acids are numbered as in rPAM-

of PAM Proteins in Permanently hEK-293 Ceil Lines

Stable cell lines expressing six different PAM proteins (Fig. 1) were generated using a vector encoding PAM under control of the cytomegalovirus promoter. Two of the cDNAs encoded bifunctional PAM proteins containing the transmembrane domain (PAM-l and PAM-2), one encoded a soluble bifunctional PAM protein (PAM3), and three encoded soluble monofunctional proteins (PAM-4, PHM-s, and PAL-s); PAM-l through -4 occur naturally. Northern blot analysis demonstrated that each transfected cell line contained PAM mRNA transcripts of the predicted size (data not shown). Expression of PAM was first evaluated by assaying both cell extracts and spent media for PHM and PAL activity (Fig. 2). Enzyme activity was found to accumulate in serum-free medium at a constant rate. As ex-

1 (6, 7).

accumulation of PAM proteins on the cell surface and release of a collection of PAM proteins different from those secreted from regulated secretory granules (9). We expressed cDNAs encoding soluble and membrane forms of PAM in hEK-293 cells, a human embryonic kidney cell line containing endoproteases associated with the constitutive pathway but lacking regulated secretory granules or any significant amount of mRNA encoding prohormone convertase 1 or 2 (1 O-12). By investigating the forms of PAM protein contained within and released by the transfected cells, we sought to determine whether cells lacking the endoproteases required for cleaving prohormones but having proteases used to process membrane proteins and constitutively secreted proteins could cleave peptide processing enzymes. By determining the subcellular localization of the transfected proteins we sought to determine whether integral membrane forms of PAM traveled through these cells by a default pathway or contained routing signals recognized by cells capable only of constitutive secretion. The localization, processing, and

p I

I

E 150B

1 q PAL 1

a

z looa Y E 22 50s 0 = 5 o-

8

PAM

1 PAM

I I I 2 PAM 3 PAM 4 PHM-s

I

PAL-s

Fig. 2. Distribution

of PAM Proteins Cells were incubated in complete

between Cells and Medium serum-free medium for l8 h, and cell extracts and media were assayed for PHM and PAL activity; the results are expressed as the percentage of the cell content of each enzyme activity accumulated in the medium in 1 h. The endogenous PHM and PAL activity in wild type hEK-293 cell extracts was negligible (-0.1 pmol/pg cell protein. h). Extracts prepared from transfected hEK-293 cells had levels of PHM and/or PAL activity at least 1 OO-fold higher than wild type cells. The data represent the average + SD of 3-l 2 independent determinations.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 06:10 For personal use only. No other uses without permission. . All rights reserved.

’

Routing Domarn in Transfected PAM Proteins

2187

56

c m

c m

PAM-3

B.

PAM-4

c m

c m

PHM-s

PAL-s

PAL Antibody ; 1’ / -.

CD Antibody

cm

cmcm

cmcmcm

PAM-1

PAM-2

PAM-3

PAM-l

PAM-2

PAM-3

Fig. 3. Western Blot Analysis of PAM Proteins Expressed in hEK-293 Cells A, Cell extracts (c) and media (m) prepared from cells expressing soluble PAM proteins were subjected to Western blot analysis using antisera to PHM or PAL (PAL-s cells). The aliquots analyzed contained between 200-l 000 pmol/h PAL activity (l-l 0 fig cell extract protein) or between 100-500 pmol/h PHM activity for cells expressing only the PHM domain. No signal was observed when 20 pg wild type hEK-293 cell extract protein was analyzed. 8, Extracts and media prepared from cells expressing integral membrane PAM proteins or PAM-3 were visualized using the PAL antiserum as described in A. The membrane was then stripped and reprobed with antiserum to the COOH-terminal domain of PAM. The average molecular masses (kDa) of the PAM proteins are indicated and are the means of at least four determinations.

petted, in cells expressing soluble forms of PAM, most of the enzyme activity was recovered from the medium (40-150% of the cell content of enzyme accumulated in the medium per hour). In contrast, most of the enzyme activity recovered from cells expressing integral membrane forms of PAM was recovered from the cell extracts. Release of PHM or PAL activity from cells expressing PAM-l or -2 requires endoproteolytic cleavage; approximately 5% of the cell content of enzyme activity accumulated in the medium per hour.

In order to determine whether the PAM proteins produced by transfected hEK-293 cells underwent endoproteolytic cleavage of the type observed in AtT-20

cells(8) we subjectedcell extracts and spent mediato Western blot analysis (Fig. 3). The PAM proteins secreted from cellsexpressingsolublePAM proteinswere the samesize as or slightlylargerthan the PAM proteins remainingin the cells(Fig. 3A); as discussedbelow, the increasein size may reflect the fact that PAM proteins are glycosylated. NH,-terminal sequence analysis of

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 06:10 For personal use only. No other uses without permission. . All rights reserved.

MOL 2188

Vol6No.12

END0.1992

PAM-3

PAM-l

P

2h chase

c

c

6h chase

P

m

c

PAM-2 2h chase

6h chase

P

2h chase

c

c

m

PAM-2

cells

6h chase

46 -

PAM-3 N-glycanase

m

C

medium Neuraminidase

c

PAM-l N-glycanase +

m

C

cells Neuraminidase +

m

N-glycanase +

C

m

Neuraminidase +

Fig. 4. Pulse-Chase

Analysis of Intracellular Processing and Secretion of Soluble and Integral Membrane PAM Three equivalent wells of cells expressing PAM-l, -2, or -3 were incubated in medium containing [35S]Met for 30 min and either extracted immediately or incubated for an additional 2 or 6 h in nonradioactive complete medium. Upper panels, Cell extracts (c) and chase medrum (m) corresponding to 10% of each sample were analyzed by immunoprecipitation with a PHM antiserum. Insets, The regron of the gel containing the PAM proteins released from PAM-l and -2 cells is shown after a 5fold longer exposure to film. Lower panels, N-glycanase and neuraminidase digestions were carried out on PAM proteins immunoprecipitated from extracts of cells expressing PAM-l or -2 and from medium of cells expressing PAM-3; all samples were taken from cells that had been chased for 2 h. Control samples were incubated under the same conditions without added enzyme. Samples in the upper pane/s were analyed on gels containing 10% acrylamide; gels containing 6% acrylamide were used to achieve better resolution of the products of N-glycanase and neuraminidase digestion.

PAM-3 and PHM-s purified from spent medium indicated that the proteins began with Phe2’j, the residue immediately following the signal sequence (13). Unlike AtT-20 cells, very little endoproteolytic processing of soluble PAM proteins occurred in hEK-293 cells. The PAM proteins released from ceils expressing integral membrane forms of PAM were lo-20 kilodaltons (kDa) smaller than the PAM proteins in the cells and were detected by antisera to both PHM and PAL (Fig. 38, left). Use of an antiserum specific for the COOH-terminal domain of PAM demonstrated that this domain had been removed from the bifunctional PAM proteins released from cells expressing PAM-l or -2 and retained in the bifunctional PAM protein secreted by cells expressing PAM-3 (Fig. 3B, right). The sizes of the proteins released from PAM-l and -2 cells indicated that endoproteolytic cleavage occurred near the transmembrane domain.

Biosynthetic

Labeling

of Transfected

PAM Proteins

Secretion rates based on enzyme activity would be misleading if the cellular and secreted PAM proteins were not stable. In order to measure the synthesis,

turnover, and secretion of PAM, hEK-293 cells expressing PAM-3 and PAM-l or -2 were metabolically labeled. Triplicate wells of cells were incubated in medium containing [?S]Met for 30 min; one well was harvested immediately (pulse), and the others were incubated in unlabeled medium for 2 or 6 h (chase) (Fig. 4, upper panel). Newly synthesized PAM proteins were isolated by quantitative immunoprecipitation. The majority of the newly synthesized PAM-3 was secreted within the first 2 h after synthesis; at the end of 6 h of chase, all of the newly synthesized PAM-3 was in the medium. Cells expressing integral membrane PAM proteins exhibited distinctly different kinetics (Fig. 4, upper panel). Essentially all of the newly synthesized integral membrane PAM protein remained cell associated after a 2-h chase. After 6 h of chase, a significant amount of newly synthesized, intact PAM-l and -2 remained cell associated. Consistent with enzyme activity data, release of newly synthesized PAM-l and -2 proteins occurred slowly; even after the 6-h chase, long exposures were needed to visualize labeled PAM proteins released into the medium (Fig. 4, upper panel, insets). Biosynthetically labeled PAM-l, -2, and -3 were iso-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 06:10 For personal use only. No other uses without permission. . All rights reserved.

Routrng

Domain

In Transfected

PAM

Proterns

2189

PAL

WGA

PAM-3

PAM-l

PAM-2

Fig. 5. Localization of PAM Proteins in Transfected hEK-293 Cells Ceils expressing PAM-l, -2, or -3 were grown on fibronectin-coated glass slides, fixed, and permeabilized. PAM proteins were visualized by incubation with rabbit polyclonal antibody to PAL followed by fluorescein-conjugated goat antirabbit immunoglobulin. Cells were simultaneously incubated with rhodamine-conjugated WGA as a marker for the Golgi apparatus. Arrows mark the punctate, vesicular staining that distinguishes cells expressing integral membrane forms of PAM from cells expressing soluble forms of PAM; a similar pattern could not be observed at any focal plane in PAM-3 cells. Fluorescein and rhodamine signals were distinguished using the appropriate filters, and photographs of the same cells are shown for comparison. The scale bar represents 25 Wm.

lated by immunoprecipitation and treated with N-glycanse or neuraminidase (Fig. 4, lower panel). Digestion of all three proteins with N-glycanase decreased their apparent molecular mass by 3-4 kDa, indicating Nglycosylation of the potential site in the PAL domain (Fig. 4, lower panel). The 6-kDa increase in the apparent molecular mass of intracellular PAM-1 observed during the chase was reversed by neuraminidase digestion; in contrast, neuraminidase treatment decreased the apparent molecular mass of PAM-2 and -3 by no more than 1 kDa (Fig. 4, lower panel). Both soluble and integral membrane PAM proteins were quite stable in hEK-293 cells and in the medium (Fig. 4). Comparison of the newly synthesized PAM protein recovered from cells or medium after the 6-h

chase to the newly synthesized PAM protein in the cells after the 30-min pulse indicated that soluble and integral membrane PAM proteins were not subject to rapid turnover. When radiolabeled PAM proteins secreted by AtT-20 cells were incubated with hEK-293 cells expressing PAM-l for 5 h, no degradation was observed. Localization of Integral Membrane and Soluble PAM Proteins in Transfected Cells In neuroendocrine cells, PAM and its peptide substrates traverse the regulated secretory pathway (3, 14-17). Since hEK-293 cells lack a regulated secretory pathway and secretory granules, it was of interest to determine the distribution of soluble and integral membrane PAM

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 06:10 For personal use only. No other uses without permission. . All rights reserved.

MOL 2190

ENDO.

Vol6No.12

1992

Excm 26 b Bb

A rPAM-1 rPAM-2

rPAM-1ABb

T

TM

I976

’ 8W 917

b9,6

the staining for rhodamine-conjugated WGA (Fig. 5). In addition, punctate staining extending throughout the cell was apparent in cells expressing integral membrane forms of PAM (Fig. 5, arrows). Stainingof larger vesicular structures perhaps representing endosomesor lysosomeswas observed only in cellsexpressingintegral membraneforms of PAM. The surfaces of cells expressing integral membranePAM proteins were not heavily labeled, suggesting that these proteins were not routed to the plasmamembraneof hEK-293 cells by default. Localization of PAM Proteins COOH-Terminal Domain

C PAM-1ABb

M

C PAM-l/

M 899

C PAM-2

M I899

Fig. 8. Western Blot Analysis of PAM Proteins with Altered COOH-Terminal Domains A, The COOH-terminal ends of the mutant integral membrane PAM proteins expressed in hEK-293 cells are shown; rPAM-1 and -2 are included for comparison. rPAM-1 ABb lacks exon Bb (exon 26; amino acids 900-917). The rPAM-l/899 and rPAM-2/899 proteins are truncated at amino acid 899 but contain the transmembrane domain (TMD). B, Western blot analysis of proteins produced by transfected hEK-293 cells. Aliquots of cell extracts (C)or spent medium (M) were analyzed as described in Fig. 3, except that the proteins were visualized with the PAL antibody.

in the transfected cells (Fig. 5). Permeabilized cells were incubated with rabbit antisera to PHM or PAL and visualized with a fluorescein-conjugated second antibody. Permeabilized cells were simultaneously incubated with rhodamine-conjugated wheat germ agglutinin (WGA) in order to compare the localization of PAM proteins to the localization of this Golgi apparatus marker. The PAM antibodies yielded no signal in wild type cells. The PAM proteins in cells expressing PAM3 were visualized in a perinuclear position overlapping the signal observed with WGA (Fig. 5); the staining for PAM was more diffuse than the staining observed with WGA. When cells expressing PAM-l or -2 were stained with antisera to PHM or PAL, PAM proteins were preferentially visualized in a juxtanuclear position coincident with proteins

is Dependent

on the

Integral membraneproteins that lack routing determinants are thought to travel to the plasmamembrane via a default pathway (17-22). Since the cytoplasmic domain of many type I membrane proteins plays a critical role in routing (23-26) we establishedstable cell linesexpressingtruncated membranePAM proteins (Fig. 6A). The COOH-terminaldomain of PAM is encoded by two exons (4); rPAM-1AB,,, a naturally occurring product of alternative splicing, lacks an 18-amino acid peptide (residues 900-917) that is part of the COOH-terminaldomain. PAM-l /899 and PAM-2/899 are truncation mutants lacking all but nine aminoacids following the transmembranedomain(Fig. 6A). BifunctionalPAM proteinsof the expected sizeswere identified in cell extracts (Fig. 6B). The PAM proteins releasedby cellsexpressingPAM-l , -l/899, and -1ABb were similarin size; the PAM proteinsreleasedby cells expressing PAM-2 and -2/899 were indistinguishable. The levels of PHM and PAL activity observed were consistent with the levels of RNA and protein expression, indicating that the truncated PAM proteins were not misfoldedor destroyed rapidly after synthesis(not shown). Pulse/chasebiosynthetic labelingof cells expressingPAM-l, -l/899, -2, and -21899demonstrated a more rapid appearance of newly synthesized PAM proteinsin the mediumof cellsexpressingthe truncated forms of PAM. As observed for PAM-l and -2 (Fig. 4) a significant amount of PAM-l/899 and -21899 remainedcell associatedafter a 5-h chase incubation. To determine whether mutations in the COOH-terminaldomainaltered the localizationof PAM, hEK-293 cells were subjected to immunostaining.Proteins on the cell surface were distinguishedfrom proteins in internal membranesby comparingcells that had been fixed and then exposed to antibody (Fig. 7, right) with cells that had been fixed and permeabilized before exposureto antibody (Fig. 7, left). When the stainingof permeabilizedand nonpermeabilizedcells expressing PAM-I, -2, or -1ABb(not shown)was compared,it was clearthat only a smallproportionof the wild type integral membranePAM protein was presenton the surface of the transfected cells. Thus these integral membrane PAM proteins did not accumulateon the plasmamembrane. In contrast, cell linesexpressingPAM proteinslacking the COOH-terminaldomain showed greatly increased

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 06:10 For personal use only. No other uses without permission. . All rights reserved.

Routing

Domain

in Transfected

PAM

Proteins

2191

-TX- 100

+TX- 100

PAM-l

PAM-l I899

PAM-2

PAM-2 I899

Fig. 7. Localization

of PAM Proteins on the Surface of Transfected hEK-293 Cells PAM proteins were grown on glass chamber slides, fixed, and either permeabilized with 0.1% Triton X-l 00 (left panels) or incubated in PBS without detergent (right panels). The PAM proteins were visualized with rabbit antiserum to PAL and Cells expressing

fluorescein-conjugated

goat antirabbit

antibody.

The scale bar represents

staining on the cell surface (Fig. 7, right). When these cells were stained with PAL antiserum after permeabilization, diffuse staining was noted throughout the cell with surface staining easily visible and only a modest accumulation in the perinuclear region. Cells expressing soluble PAM-3 were examined as a control; no staining was observed in the absence of permeabilization (Fig. 7). Thus residues 918-976 in the cytoplasmic tail of integral membrane PAM contain routing information recognized by hEK-293 cells. PAL Activity on the Surface of Cells Expressing PAM The immunofluorescence experiments clearly indicated that PAM antigens were located on the surface of some

25 pm.

of the transfected cell lines. To determine whether the enzyme was active on the cell surface and to quantify the amount of PAM protein on the surface, we assayed PAL activity on live cells. Surface and released PAL activity was expressed as a percentage of the total cell content of PAL activity (Fig. 8). Transfected hEK-293 cells expressing PAM-l and -2 had low levels of PAL activity on the surface; PAM-l and -2 cells released a similarly low amount of PAL activity. In comparison, cells expressing PAM-l /899 had 6-fold more PAL activity on the surface and released PAL activity 8-fold faster than PAM-l -expressing cells. Cells expressing PAM9/ 899 had approximately 17-fold more PAL activity on the surface and released PAL activity lo-fold faster than PAM9-expressing cells.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 06:10 For personal use only. No other uses without permission. . All rights reserved.

MOL END0.1992

Vol6No.12

Internalization Surface

: .i PAM-l

PAM-l/899

PAM-2

of PAM Proteins

from the Cell

In transfected AtT-20 cells, full-length integral membrane PAM proteins were rapidly internalized from the surface; internalization was diminished upon truncation of the COOH-terminal domain at Gly8” (9). To determine whether the COOH-terminal domain functioned as an internalization signal in cells lacking a regulated secretory pathway, hEK-293 cells expressing full-length and truncated integral membrane PAM proteins were incubated in the presence of PHM antiserum for 5 min at 37 C and then either fixed or further incubated in the absence of antiserum. Internalization of the PHM antibody was determined by incubating duplicate wells with secondary antibody with or without prior permeabilization (27-29) (Fig. 9). Using this approach, PAM/PAM antibodies (Ab) on the surface of transfected cells could be distinguished from internalized PAM/PAM Ab. To rule out a significant contribution from fluid-phase uptake of the primary antiserum, cells expressing soluble PAM-3 were analyzed in the same way; no specific

1

PAM-21899

Fig. 8. PAL Activity on the Surface of Cells Expressing FullLength and Truncated PAM Proteins PAL activity was assayed on the surface, in the spent medium, and in total cell extracts prepared from hEK-293 cells expressing the indicated proteins. The amount of PAL activity found on the surface or released in 1 h is expressed as a percentage of the total PAL activity in the cell extracts. The data shown are mean values -+ so from five separate analyses.

0 time

IO min chase

Fig. 9. lnternalizatron of PAM from the Surface of PAM-2 and -2/699 Cells Cells were incubated with PHM antiserum diluted 50-fold in DMEM/F12/BSA for 5 min, rinsed, and incubated at 37 C in DMEM/ F12 for 10 min. Cells were then fixed and Incubated with fluorescein-conjugated secondary antiserum with or without prior permeabrlrzatron as Indicated. All cells were photographed under identical conditions for the same length of time. Bar = 25 pm.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 06:10 For personal use only. No other uses without permission. . All rights reserved.

Routing

Domain

in Transfected

PAM Proteins

signal was detected in the presence or absence of detergent. After the 5min incubation with PHM antiserum (Fig. 9, 0 time), a faint signal was observed on the surface of hEK-293 cells expressing PAM-2; in the presence of detergent, surface staining was visible along with a significant amount of punctate staining. After the lomin chase incubation in the absence of antibody, little staining remained on the cell surface; the internalized PAM/PAM Ab was concentrated in a juxtanuclear position. When live cells expressing truncated PAM-2/899 were analyzed in the same way, a dramatic decrease in the internalization of PAM protein from the cell surface was evident (Fig. 9). After the initial incubation with antibody, the majority of the PAM/PAM Ab remained on the cell surface. Even after a lo- or 20-min chase incubation, most of the antibody remained on the surface of PAM-2/899 cells. Although some internalization of PAM/PAM Ab was observed in PAM-2/899 cells after longer incubations, the punctate pattern seen in cells expressing PAM-2 was never observed. Similar results were obtained with PAM-l and -l/899 cells. Thus the sorting machinery in hEK-293 cells recognizes an internalization signal present in the COOH-terminal domain of PAM.

DISCUSSION The fundamental rules underlying the routing of posttranslational processing enzymes to secretory granules have not yet been elucidated. Since PAM functions late in the posttranslational processing pathway and occurs naturally in both soluble and integral membrane protein forms, it is an excellent model for trying to define the rules governing the routing of processing enzymes. In a previous study we explored the expression of soluble and membrane forms of PAM in AtT-20 cells, a neuroendocrine cell line exhibiting both constitutive and regulated secretion (8). In these cells, soluble and integral membrane protein forms of PAM were subjected to limited endoproteolytic processing, and soluble PAM proteins were released from regulated secretory granules. By stably expressing soluble and integral membrane protein forms of PAM in hEK-293 cells (IO), a kidney cell line lacking a regulated secretory pathway, we sought to define the interactions of this processing enzyme with components of the constitutive pathway. As in AtT-20 cells, active enzyme was obtained by expressing bifunctional or monofunctional, soluble or membrane, forms of PAM. PAM proteins were glycosylated when expressed in hEK-293 cells. Digestion of metabolically labeled PAM-l, -2, or -3 with N-glycanase resulted in a 3- to 4-kDa decrease in size, indicating that the single N-glycosylation site in the PAL domain was used (Fig. 1) (3). PAM-l, but not PAM-2, exhibited a 6-kDa decrease in apparent molecular mass upon digestion with neuraminidase, suggesting that sialic acid residues are attached to a site in exon A.

2193

Unlike AtT-20 cells, hEK-293 cells did not cleave the soluble forms of PAM into smaller proteins. The integral membrane forms of PAM underwent endoproteolytic cleavage to generate the large, bifunctional PAM proteins found in hEK-293 cell-spent medium. In contrast, AtT-20 cells cleaved PAM-l into soluble monofunctional PHM and PAL proteins (8) of the size observed in the bovine neurointermediate pituitary (30). Both furin and PACE-4, endoproteases thought to cleave proteins traversing the constitutive pathway, have been identified in hEK-293 cells (12). The fact that hEK-293 cells do not produce PAM proteins resembling those found in neuroendocrine tissues suggests that the enzymes required are specific to neuroendocrine cells. Release of large, bifunctional PAM proteins from hEK-293 cells expressing PAM-l or PAM-2 requires endoproteolytic cleavage at sites near the transmembrane domain. Release of the ectodomain of another type I integral membrane protein, @-amyloid precursor protein (APP), from Chinese hamster ovary cells involved cleavage after a Lys residue 12 amino acids before the transmembrane domain: -His-His-Gln-Lys”‘Leu-Val-Phe- (31). The sequence around a Lys residue eight amino acids before the transmembrane domain of rPAM exhibits significant similarity to the corresponding region of APP: -Glu-Lys-Gln-Lyss5*-Leu-Ser-Thr(6). The endoproteolytic cleavage releasing PAM may occur at the cell surface, since little of the released form is detectable in cell extracts. Based on both enzyme activity and metabolic labeling, transfected hEK-293 cells expressing PAM-l or -2 released about 5% of their cell content of PAM per hour. The rates at which the ectodomains of other type I integral membrane proteins are released from cells lacking a regulated secretory pathway vary considerably. Newly synthesized APP was completely released from transfected hEK-293 cells within 4 h (32). At the other extreme, fibroblasts expressing KexZp or furin, subtilisin-like endoproteases thought to play a role in prohormone processing (33-37) released no more than 0.2% of their enzyme content per hour. Integral membrane proteins lacking a routing determinant travel to the plasma membrane (18-22). However, expression of PAM-l and -2 in hEK-293 cells did not lead to an accumulation of PAM protein on the cell surface. Only small amounts of PAM could be visualized on the surface of cells expressing PAM-l or -2, and only l-3% of the cell content of enzyme activity could be assayed on the cell surface. Thus PAM-l and -2 contain routing information recognized by hEK-293 cells. In contrast, expression of PAM proteins truncated immediately following the transmembrane domain resulted in a redistribution of the transfected proteins to the cell surface. The fact that these truncated proteins retained full enzymatic activity minimizes the possibility that improper folding caused their redistribution. The parallel increase in surface enzyme activity and rate of release of active enzyme observed upon deletion of the COOH-terminal domain suggests that enzyme is re-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 06:10 For personal use only. No other uses without permission. . All rights reserved.

MOL 2194

ENDO.

1992

leased from the cell surface. Since deletion of the 18 amino acids immediately following the transmembrane domain had no effect on the secretion or surface expression of PAM, routinq information aooears to reside in residues 918-976 of the COOH-te&nal domain of PAM. We have recently demonstrated that integral membrane PAM proteins were rapidly internalized from the surface of transfected AtT-26 cells, and that this rapid internalization required the presence of the COOHterminal domain of PAM-l and -2 (9). To determine whether internalization of PAM-l and -2 occurred in a similar manner in hEK-293 cells, live cells were incubated with antiserum to PHM. As in AtT-20 cells, rapid internalization of the PHM/PHM Ab complex was observed in hEK-293 cells expressing PAM-l or -2; truncation of the COOH-terminal domain eliminated this rapid internalization, and most of the PHM/PHM Ab complex remained on the cell surface. While the fate of the internalized PAM proteins has not yet been investigated, the fact that more than 65% of the newly synthesized [35S]Met-labeled PAM-l and -2 was recovered intact after a 6-h chase in basal medium suggests that PAM proteins were not immediately targeted from endosomes to lysosomes for degradation. It is striking that hEK-293 cells possess the capability to recognize a signal in the COOH-terminal domain of PAM. Routing signals in P-selectin, a type I integral membrane protein found in the a-granules of platelets and Weibel-Palade bodies of endothelial cells, were not recognized by cells lacking a regulated secretory pathway. Expression of P-selectin in fibroblasts led to its accumulation on the cell surface (38, 39), while expression in AtT-20 cells led to accumulation in ACTH-containing secretory granules; removal of the COOH-terminus of P-selectin resulted in its redistribution from secretory granules to the surface of AtT-20 cells (38). As observed in stably transfected hEK-293 cells, transient expression of PAM-l in Chinese hamster ovary and NIH 3T3 cells yielded punctate staining throughout the cell and concentrated in the perinuclear region (data not shown); this observation suggests the presence of different trafficking signals in PAM and P-selectin. Kex2p, a type I integral membrane protein responsible for cleaving the a-mating factor precursor in yeast, is normally localized to the Golgi apparatus; upon expression in mammalian fibroblasts, a similar localization was observed with no enzyme activity detected on the cell surface (35-37). Truncation of the COOHterminal domain of Kex2p or expression of wild type Kex2p in yeast mutants lacking the heavy chain of clathrin led to a striking increase in the amount of enzyme activity on the cell surface (34, 40). Similar interactions may be involved in the routing of integral membrane forms of PAM. One of the two Tyr residues in tt;e COOH-terminal domain of PAM (Tyrg36) occurs in a sequence resembling those identified as routing determinants in several constitutively recycling receptors (23, 25, 41). As suggested for synaptic vesicle biogenesis (29, 42, 43), the endocytic pathway may play a

Vol6No.12

role in the routing of membrane tory granules.

MATERIALS Construction

AND

proteins to neurosecre-

METHODS

of Expression

Vectors

Construction of the pBluescript plasmids pBS.KrPAM-1, -2, 3, -4, PHM-s, and PAL-s was described previously (8, 30). To construct the form of pBS.KrPAM-1 which lacks exon Bb (exon 26), pBS.KrPAM-1 and pBS.PAM-3b (which lacks exon Bb) (44) were cleaved with BstEll and Smal to remove the fragment between nucleotides for rPAM-1). The

pBS.PAM-3b

(nt) 1897 and nt 3244 (all nt numbers

are

1.3-kilobase fragment released from was ligated to the pBS.KrPAM-1 plasmid from

which the 1.3-kilobase region had been removed. The resulting plasmid, pBS.KrPAM-1 ABb, lacks the 54 nt of exon Bb (amino acids 900-917). Sequences across restriction sites were verified by sequence analysis. The plasmid pBS.KrPAM-l/899 was created using the polymerase chain reaction to insert a stop codon after the codon for Glyegg; pBS.KrPAM-1 was amplified using a sense primer (nt 2325-2341) and an antisense primer with an Xbal site (bold) and a stop codon (italic) preceding the codon for Gly*?

5’-CCTCTAGACTATCCAAAGGCCCTTGATT-3’

(nt 2994-

2978). The amplified fragment obtained was digested with Avrll (nt 2338) and Xbal and inserted into pBS.KrPAM-1 from which the Avrll/Xbal region had been removed, creating pBS.KrPAM-l/899. The plasmid pBS.KrPAM-2/899 was constructed by inserting the Aatll (nt 2617)-Xbal fragment from pBS.KrPAM-l/899 (containing the deletion of the COOH-terminal domain) into pBS.KrPAM-2 from which the corresponding Aatll-Xbal fragment had been removed. DNA regions produced using the polymerase chain reaction were verified by sequencing. All of the KrPAM cDNAs were inserted into the pCIS.2CXXNH expression vector kindly provided by Dr. Cornelia Gorman (Genentech, South San Francisco, CA) as described (8, 45).

Transfection,

Cell Culture,

and Enzyme

Assays

Wild type hEK-293 cells were grown, transfected, and selected by measuring the enzymatic activity in cell extracts and spent medium as described previously for AtT-20 cells (8). Northern blot analyses of the transfected cell lines were performed as described (8, 11). Cell lines whose activity was not stable through the following months were subcloned by limiting dilution (8). Since serum contains significant levels of PHM and PAL activity, cells were incubated in serum-free medium when secretion of enzyme was to be measured. The medium consisted of Dulbecco’s modified Eagle’s medium (DMEM)/F12 containing insulin, transferrin, bacitracin, lima bean trypsin inhibitor, and BSA (8); spent medium was collected after time periods ranging from 2-24 h and was centrifuged to remove any cells present. Cells were homogenized in ice-cold hypotonic buffer containing detergent and protease inhibitors, and insoluble material was removed by centrifugation at 4 C for 10 min at 15,000 x g (8). PHM and PAL activity in cell extracts and spent media was measured using ‘251-labeled substrates and normalized to the protein level in the extracts (46).

Sodium Dodecyl Electrophoresis

Sulfate-Polyacrylamide Gel (SDS-PAGE) and Western Blot Analysis

Aliquots of cell extract or spent serum-free medium containing between 200-1000 pmol/h PHM or PAL activity were subjetted to SDS-PAGE and Western blot analysis (46) using a

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 06:10 For personal use only. No other uses without permission. . All rights reserved.

Routrng

Domain

in Transfected

PAM

Proteins

1:1500 dilution of rabbit antiserum raised aqainst purified bacterially expressed PHM [Ab 475 to rPAM (37-382)], PAL IAb 471 to rPAM(463-864)l. or COOH-terminal domain IAb 571 to rPAM(8981976)] (5)‘and an enhanced chemiluminkscence kit (Amersham, Arlington Heights, IL).

2195

beads and analyzed by SDS-PAGE as described (9). Immunoprecipitated proteins were released from protein A-Sepharose beads by boiling in 10 mM Na phosphate, pH 6.0, 0.1% SDS, 10 mM P-mercaptoethanol, and aliquots were digested with N-glycanse (Genzyme, Cambridge, MA) or neuraminidase (Genzyme).

Acknowledgments Fixed Cells Transfected cells were cultured on chamber slides precoated with 25 pg/ml fibronectin (Sigma, St. Louis, MO); cells were rinsed, fixed in 4% paraformaldehyde, and immunostained with primary antisera at 1 :1,500 dilution and fluorescein-conjugated goat antirabbit immunoglobulin G (CalTag, South San Francisco, CA) (8). Rhodamine-conjugated WGA (Molecular Probes, Eugene, OR) diluted l:lO,OOO was used in some experiments as a marker for the Golgi apparatus (17). The distribution of fluorescence was analyzed with a Zeiss (Thornwood, MT) Axioskop epifluorescence microscope using fluorescein (BP 485/20, barrier filter 520-560) and rhodamine (BP 546/l 2, LP 590) filters. No staining was observed when preimmune serum from the same rabbits was used to analyze transfected cells, or when wild type hEK-293 cells were analyzed. Internalization of PAM and PAM Antibodies Cells were rinsed with warm DMEM/F12 containing 1 .O mg/ml BSA (DMEM/F12/BSA) and incubated in DMEM/F12/BSA containing antiserum diluted 1:50 for 5 min. Cells were rinsed once with DMEM/F12/BSA and then either fixed or incubated further in DMEM/F12/BSA before fixina. Antisera bound to the surface were ‘visualized without detergent permeabilization, and internalized antisera were visualized after detergent treatment, using fluorescein-conjugated goat antirabbit immunoglobulin as above. Measurement

of Cell Surface

Activity

Cells were cultured in 96-well plastic culture dishes for 72 h in regular growth medrum and rinsed twice in bicarbonate-free DMEM/Fl2 medium containina 0.2 ma/ml BSA. To auantifv surface PAL activity, cells from‘iriplicat&r quadruplicate wells were incubated in an air atmosphere in 50 ~1 DMEM/F12, pH 7.0, containing 20 mM Na-HEPES but no NaHC03, plus 0.5 KM u-A!-acetyl-Tyr-Val-c-hydroxyglycine and trace amounts of the same ‘251-labeled peptide for 15 min at 37 C. Generation of amidated product represented the PAL activity present on the cell surface and activity secreted from the cells during the incubation. To correct for the contribution made by enzyme secreted from the transfected cells, 50 ~1 of the same medium lacking substrate were added to duplicate wells; spent medium was collected after a 15-min incubation at 37 C and assayed for PAL activity at pH 7.0. PAL in the cell extracts was measured as described earlier, except the pH of the buffer was adjusted to 7.0. To calculate the surface activity of transfected cells, the contribution of secreted PAL was determined using the fact that secretion was linear in time and was subtracted from the measurement using live cells to yield surface activity. There was no detectable internalization of [‘251]substrate by the cells. The amounts of secreted and surface activity were expressed as a percent of the total activity present in the cells. Biosynthetic

Labeling

Cells were incubated in methionine-deficient serum-free medium supplemented with [%]Met (575 Ci/mmol; 5 PM) for the indicated times; in some experiments cells were incubated for additional times in complete serum-free medium containing the normal concentration of methionine (47). Cell extracts were prepared as described above for assays of enzyme activity, and medium and extracts were incubated overnight at 4 C with antisera to the PHM or CD domains of PAM. Antigenantibody complexes were isolated using protein A-Sepharose

We thank Drs. Jean Husten and Ana Oyarce for helpful discussions, Richard Johnson for construction of expression vectors, Carla Berard for help in tissue culture, and Cori Gorman for the pCIS2CXXNH expression vector.

Received August 18, 1992. Revision received September 30, 1992. Accepted September 30, 1992. Address requests for reprints to: Dr. Betty A. Eipper, The Johns Hopkins University School of Medicine, Department of Neuroscience, 725 N. Wolfe Street, Baltimore, Maryland 21205. This work was supported by PHS Grants DK-32949, DA00097, and DA-00098.

REFERENCES 1. Jung L, Scheller R 1991 Peptide processing and targeting in the neuronal secretory pathway. Science 251:13301335 2. Burgess TL, Kelly RB 1987 Constitutive and regulated secretion of proteins. Annu Rev Cell Biol 3:243-293 3. Eipper BA, Stoffers DA, Mains RE 1992 The biosynthesis of neuropeptides: peptide a-amidation. Annu Rev Neurosci 15:57-85 4. Ouafik L’H, Stoffers DA, Campbell TA, Johnson RC, Bloomquist BT, Mains RE, Eipper BA 1992 The multifunctional peptidylglycine a-amidating monooxygenase gene: exon/intron organization of catalytic, processing, and routing domains. Mol Endocrinol6:1571-1584 5. Yun H-Y, Johnson RC, Mains RE, Eipper BA, Topological switching of the COOH-terminal domain of peptidylglycine a-amidating monooxygenase by alternative RNA splicing. Arch Biochem Biophys, in press 6. Stoffers DA, Green CB-R, Eipper BA 1989 Alternative mRNA splicing generates multiple forms of peptidyl-glytine a-amidating monooxygenase in rat atrium. Proc Natl Acad Sci USA 86:735-739 7. Stoffers DA, Ouafik L’H, Eipper BA 1991 Characterization of novel mRNAs encoding enzymes involved in peptide oc-amidation. J Biol Chem 266:1701-l 707 8. Milgram SL, Johnson RC, Mains RE 1992 Expression of individual forms of peptidylglycine cu-amidating monooxygenase in AtT-20 cells: Endoproteolytic processing and routing to secretory granules. J Cell Biol 117:717-728 9. Milgram SL, Eipper BA, Mains RE 1992 The COOHterminal domain of integral membrane a-amidating enzyme mediates retrieval from the cell surface and return to the Golgi apparatus. Mol Biol Cell 3:549 10. Graham F, Smiley JW, Russel W, Nairn R 1977 Characteristics of a human cell-line transformed by DNA from human adenovirus Tvpe 5. J Gen Virol 36:59-72 11. Bloomquist BT, Eipper BA, Mains RE 1991 Prohormoneconverting enzymes: regulation and evaluation of function using antisense RNA. Mol Endocrinol 5:2014-2024 12 Kiefer MC, Tucker JE, Joh R, Landsberg KE, Saltman D, Barr PJ 1991 Identification of a second human subtilisinlike protease gene in the fes/fps region of chromosome 15. DNA Cell Biol 10:757-769 13 Husten EJ, Tausk FA, Keutmann HT, Eipper BA, Functional consequences of separating the monooxygenase

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 06:10 For personal use only. No other uses without permission. . All rights reserved.

MOL 2196

14.

15.

16.

17.

18.

19.

20. 21. 22. 23. 24. 25.

26.

27.

28.

29.

30.

31.

ENDO.

1992

and lyase domains of peptidylglycine a-amidating monooxygenase (PAM). Program of the 74th Annual Meeting of The Endocrine Society, San Antonio, Texas, 1992, p 454 (Abstract) Steiner DF 1991 The biosynthesis of biologically active peptides, a perspective. In: Fricker LD (ed) Peptide Biosynthesis and Processing. CRC Press, Boston, pp l-1 5 Mains RE, Eipper BA 1990 The tissue-specific processing of pro-ACTH/endorphin: recent advances and unsolved problems. Trends Endocrinol Metab 1:388-394 Mains RE, Glembotski CC, Eipper BA 1984 Peptide 01amidation activity in mouse anterior pituitary AtT-20 cell granules: properties and secretion. Endocrinology 114:1522-1530 Rivas RJ, Moore H-P 1989 Spatial segregation of the regulated and constitutive secretory pathways. J Cell Biol 109:51-60 Pfeffer SR, Rothman JE 1987 Biosynthetic protein transport and sorting by the endoplasmic reticulum and Golgi. Annu Rev Biochem 56:829-852 Lodish HF, Kong N, Snider M, Strous GJ 1983 Hepatoma secretory proteins migrate from rough endoplasmic reticulum to Golgi at characteristic rates. Nature 304:80-83 Farquhar MG 1985 Progress in unraveling pathways of Golgi transport. Annu Rev Cell Biol 1:447-488 Bomsel M, Mostov K 1991 Sorting of plasma membrane proteins in epithelial cells. Curr Op Cell Biol 3:647-653 Simons K, Wandinger-Ness A 1990 Polarized sorting in epithelia. Cell 62:207-210 Trowbridge IS 1991 Endocytosis and signals for internalization Curr Op Cell Biol 3:634-641 Brown MS, Goldstein JL 1986 A receptor-mediated pathway for cholesterol homeostasis. Science 232:34-47 Ktistakis NT, Thomas D, Roth MG 1990 Characteristics of the tyrosine recognition signal for internalization of transmembrane surface glycoproteins. J Cell Biol 111:1393-1407 Graeve L, Patzak A, Drickamer K, Rodriguez-Boulan E 1990 Polarized expression of functional rat liver asialoglycoprotein receptor in transfected MDCK cells. J Biol Chem 265:1216-1224 Jacobowitz DM, Ziegler MG, Thomas JA 1975 In vivo uptake of antibody to dopamine P-hydroxylase into sympathetic elements. Brain Res 91:165-170 Williams MA, Fukuda M 1990 Accumulation of membrane glycoproteins in lysosomes requires a tyrosine residue at a particular position in the cytoplasmic tail. J Cell Biol 111:955-966 Matteoli M, Takei K, Perin MS, Sudhof TC, DeCamilli P 1992 Exo-endocytotic recycling of synaptic vesicles in developing processes of cultured hippocampal neurons. J Cell Biol 117:849-861 Eipper BA, Perkins SN, Husten EJ, Johnson RC, Keutmann HT, Mains RE 1991 Peptidyl-a-hydroxyglycine 01amidating lyase: purification, characterization, and expression. J Biol Chem 266:7827-7833 Wang R, Meschia J. Cotter R, Sisodia S 1991 Secretion of the p/A4 amyloid precursor protein. J Biol Chem 266:16960-l 6964

Vol6No.12

32.

33.

34.

35.

36.

37.

38.

39.

40.

41. 42.

43.

44.

45.

46.

47.

Oltersdorf T, Ward PJ, Henriksson T, Beattie EC, Neve R, Lieberburg I, Fritz LC 1990 The Alzheimer amyloid precursor protein. J Biol Chem 265:4492-4497 Fuller RS, Brake A, Thorner J 1989 Yeast prohormone processing enzyme (Kex2 gene product) is a Ca*+ dependent serine protease. Proc Natl Acad Sci USA 86:1434-l 438 Fuller RS, Brake AJ, Thorner J 1989 Intracellular targeting and structural conservation of a prohormone-processing endoproteases. Science 246:482-486 Misumi Y, Oda K, Fujiwara T, Takami N, Tashiro K, lkehara Y 1991 Functional expression of furin demonstrating its intracellular localization and endoprotease activity for processing of proalbumin and complement pro-C3. J Biol Chem 266:16954-l 6959 Germain D, Zollinger L, Racine C, Gossard F, Dignard D, Thomas D, Crine P, Boileau G 1990 The yeast Kex2 processing endoprotease is active in the Golgi apparatus of transfected NIH 3T3 fibroblasts. Mol Endocrinol 4:1572-l 579 Bresnahan PA, Leduc R, Thomas L, Thorner J, Gibson HL, Brake AJ, Barr PJ, Thomas G 1990 Human fur gene encodes a yeast Kex2-like endoprotease that cleaves-proB-NGF in vivo. J Cell Biol 111:2851-2859 Disdier M, Morrisey JH, Fugate RD, Bainton DF, McEver RP 1992 The cytoplasmic domain of P-selectin (CD62) contains the signal for sorting into the regulated secretory pathway. Mol Cell Biol 3:309-321 Koedam JA, Cramer EM, Briend E, Furie 9, Furie BC 1992 P-Selectin, a granule membrane protein of platelets and endothelial cells, follows the regulated secretory pathway in AtT-20 cells. J Cell Biol 116:617-625 Payne GS, Schekman R 1989 Clathrin: a role in the intracellular retention of a Golgi membrane protein. Science 245: 1358-l 365 Vaux D 1992 The structure of an endocytosis signal. Trends Cell Biol 2:189-l 92 Linstedt AD, Kelly RB 1991 Synaptophysin is sorted from endocytic markers in neuroendocrine PC1 2 cells but not transfected fibroblasts. Neuron 7:309-317 Cameron PL, Sudhof TC, Jahn R, DeCamilli P 1991 Colocalization of synaptophysin with transferrin receptors: implications for synaptic vesicle biogenesis. J Cell Biol 115:151-164 Eipper BA, Green CB-R, Campell TA, Stoffers DA, Keutmann HT, Mains RE, Ouafik L’H 1992 Alternative splicing and endoproteolytic processing generate tissue-specific forms of oituitarv PAM. J Biol Chem 267:4008-4015 Gorman CM, Gies DR, McCray G 1990 Transient production of proteins using an adenovirus transformed cell line. DNA Prot Engin Tech 2:3-l 0 Husten EJ, Eipper BA 1991 The membrane-bound bifunctional peptidylglycine a-amidating monooxygenase protein: exploration of its domain structure through limited proteolysis. J Biol Chem 266:17004-l 7010 Dickerson IM, Dixon JE, Mains RE 1990 Biosynthesis and posttranslational processing of site-directed endoproteolytic cleavage mutants of pro-neuropeptide Y in mouse pituitary cells. J Biol Chem 265:2462-2469

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 06:10 For personal use only. No other uses without permission. . All rights reserved.