Developmental Expression of Peptidylglycine a-Amidating Monooxygenase (PAM) in Primary Cultures of Neonatal Rat Cardiocytes: A Model for Studying Regulation of PAM Expression in the Rat Heart

Jean-Yves

Maltese

and Betty A. Eipper

Department of Neuroscience Johns Hopkins University School Baltimore, Maryland 21205

of Medicine

INTRODUCTION

Primary cultures of neonatal rat atrial and ventricular cardiomyocytes were used to investigate the expression of peptidylglycine a-amidating monooxygenase (PAM), a bifunctional enzyme required for the production of cY-amidated neuroendocrine peptides. The use of assays for the individual enzymes, peptidylglycine n-amidating monooxygenase (PHM) and peptidyl-a-hydroxyglycine a-amidating lyase (PAL), demonstrated that the levels of expression observed in vitro approximated those observed in viva. Both in viva and in vitro, atrial and ventricular PAL activity greatly exceeded PHM activity. Atrial and ventricular cardiomyocytes secreted PHM and PAL activity at a constant rate throughout the culture period. lmmunofluorescence studies localized PAM proteins to the perinuclear region, with intense punctate staining. Both in viva and in vitro, PAM mRNAs encoding integral membrane proteins predominated throughout the neonatal period, with PAM-l mRNA becoming more prevalent after the first week in culture. Although PAM-2 mRNA decreased in prevalence in viva at the time when PAM-l expression increased, levels of PAM-2 mRNA remained elevated throughout 2 weeks in vitro. Western blot analysis demonstrated intact PAM-l and PAM-2 proteins in atrial cultures, with the prevalence of PAM1 increasing in older cultures. Atrial cardiomyocytes secreted only bifunctional PAM proteins. Many of the features of PAM expression, processing, and storage that are unique to cardiomyocytes as opposed to endocrine cells are faithfully replicated by primary atrial and ventricular cultures. (Molecular Endocrinology 6: 1996-2006, 1992) ome-8809/92/1998-2008$03.00/0 Molecular Endocrmology CopyrIght 0 1992 by The Endocrme

Peptidylglycine a-amidating monooxygenase (PAM; EC 1.14.17.3) is involved in the posttranslational processing of many peptide hormones and is commonly associated with the secretory granules of endocrine cells and neurons (1, 2). PAM catalyzes the formation of CYamidated peptides from peptide precursor molecules with a COOH-terminal glycine. It has recently been shown that peptide cu-amidation is a two-step reaction catalyzed by the sequential action of two enzymatic domains contained within the PAM precursor (3-6). The first enzyme, peptidylglycine cu-hydroxylating monooxygenase (PHM), produces an oc-hydroxylated intermediate in the presence of copper, ascorbate, and molecular oxygen. The second enzyme, peptidyl-a-hydroxyglycine cu-amidating lyase (PAL), cleaves the peptidylcY-hydroxyglycine intermediate to form the cu-amidated peptide and glyoxylate; this reaction can occur spontaneously under nonphysiological conditions. The two catalytic domains of the bifunctional PAM protein can be separated by endoproteolysis and can act independently (7). A single complex gene encodes PAM in the rat (8). Tissue-specific alternative splicing can generate at least seven forms of PAM mRNA (9-11). Two major forms of PAM mRNA (rPAM-1 and -2) have been characterized in the adult rat atrium (9, 10). Both forms encode bifunctional PAM precursor proteins with an NH,-terminal signal sequence, followed by the PHM and PAL catalytic domains and a single putative transmembrane domain near the COOH-terminus. PAM-l contains a noncatalytic domain (exon A) that separates the PHM and PAL domains, while PAM-2 lacks this domain. The various PAM proteins are subjected to tissue-specific posttranslational modifications that amplify the degree of diversity generated from the PAM gene (1 l-1 4). For

Socmty

1998

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 July 2015. at 08:12 For personal use only. No other uses without permission. . All rights reserved.

Developmental Expression of PAM

1999

example, in pituitary cells, exon A serves as the site for endoproteolytic cleavages separating PHM from PAL (15). In the adult atrium, the majority of the PAM proteins remain bifunctional and membrane associated (7). During fetal and neonatal development, the prevalence of rPAM-1 and -2 mRNAs in the rat cardiac atrium and ventricle varies dramatically (16). While rPAM-2 mRNA is the most abundant form from postnatal days 1-3, approximately equal amounts of both forms of PAM mRNA are present on postnatal day 5; rPAM-1 mRNA becomes more abundant by postnatal day 7 and is the major form of PAM mRNA in the adult rat atrium. Less PAM mRNA is found in the ventricle than in the atrium throughout most of development, but changes in the ratio of PAM-1 to PAM-2 mRNA follow the same time course in both tissues. Although the atrium of the heart has higher levels of PAM activity and mRNA than any other tissue (17-l 9), the atrium is not known to contain high levels of Namidated peptide, and the function of PAM in this tissue is not clear. Based on subcellular fractionation of adult rat atria, most of the PAM activity is recovered in the secretory granule-enriched fractions containing atrial natriuretic factor (ANF) (12). ANF, the major peptide hormone synthesized in the heart atrium, is stored in these atrial granules as a prohormone and is cleaved at the time of secretion (20-22); neither the prohormone nor the active form of ANF is a-amidated. Primary cultures prepared from neonatal cardiac atria have served as a useful model system for studying ANF expression in cardiomyocytes (23-25). We have used a similar primary culture system to determine if the developmental regulation of PAM expression observed in viva is mimicked in culture. Our earlier studies on PAM expression in cultured cardiomyocytes preceded the elucidation of the bifunctional nature of the enzyme; in the earlier studies, secretion of PAM activity by cultured atrial cardiomyocytes was found to be responsive to glucocorticoids and (Bu& (26). In this study we compared levels of PHM and PAL activities in cell cultures of various ages with the levels of enzyme activity in tissue extracts prepared from pups of the same age. The forms of PAM mRNA in cultures of different ages were identified by reverse transcription/ polymerase chain reaction (PCR) in order to determine whether the developmentally regulated switch from a predominance of PAM-2 mRNA to a predominance of PAM-l mRNA occurred in culture. The PAM proteins stored in the cultured cardiomyocytes were localized by immunofluorescence, and the forms of PAM protein in cell extracts and spent media were characterized by Western blot analysis.

RESULTS PHM and PAL Activities in Rat Atrium in Viva and in Primary Cell Culture

and Ventricle

Our earlier studies indicated that expression activity in atrium and ventricle varied during

of PAM develop-

ment (16). Since peptide Lu-amidation is now known to involve the sequential action of two enzymes contained within the bifunctional PAM protein, we separately measured these two activities, PHM and PAL, in extracts prepared from the two cardiac compartments of neonatal and adult rats. To determine whether primary cultures of neonatal atrial and ventricular myocytes provide a good model for investigating the unique features of PAM expression in developing cardiomyocytes, expression of PHM and PAL activities was also determined in culture extracts. The results obtained for atria and atrial cultures are shown in Fig. IA. PHM specific activity in vivo was maximal around postnatal day 11, but neither PHM nor PAL levels varied more than 2-fold throughout postnatal development. Expression of PHM and PAL activities rose over the first week in culture and remained constant over the rest of the time in culture (Fig. 1A). The specific activities of PHM and PAL in extracts of primary cultures of neonatal atrial cells were as high as or slightly higher than those observed in vivo. The protein content of the atrial cultures increased with time in

A

B

Fig. 1. Expression of PHM and PAL Activities in Atrium and Ventricle in Viva and in Vitro Tissue and primary cultures prepared on postnatal day 1 (day of birth) were extracted (see Materials and Methods) on the days indicated. Data for atrium are shown in A, and data for ventricle in 6. PHM and PAL activities were separately measured using acetyl-Tyr-Val-Gly or acetyl-Tyr-Val-a-hydroxyglycine as substrate (see Materials and Methods). Results are the mean of three individual experiments, assayed in duplicate. The SEM was less than 10% of the mean. The PAM activity profiles shown were measured using o-Tyr-Val-Gly at pH 8.5 and are replotted from the data of Ouafik et a/. (16).

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 July 2015. at 08:12 For personal use only. No other uses without permission. . All rights reserved.

MOL 2000

ENDO.

1992

culture; taking this into account, the total amount of PHM and PAL activities per culture doubled by 5-7 days in vitro and stayed at this level for the remainder of the culture period (data not shown). The results obtained for ventricle and ventricular cultures are shown in Fig. 1B. In ventricular extracts, the specific activities of PHM and PAL were maximal on the third to fifth postnatal day, with lower levels in adulthood (Fig. 1 B). In extracts of mature cultures, the specific activities of PHM and PAL were similar to those observed in vivo. The protein content of the ventricular cultures was unchanged with time in culture; taking this into account, the total amount of PHM activity in the ventricular cultures remained constant throughout the entire culture period. The results of our previous study on the developmental regulation of atrial and ventricular a-amidation in vivo (16) are plotted for comparison (Fig. 1, A and B). The enzyme assay employed in the earlier study used a different peptide substrate, o-Tyr-Val-Gly, and a pH alkaline enough to support the spontaneous conversion of peptidyl-a-hydroxyglycine intermediates into Lu-amidated products; since the assay measured production of a-amidated product, results obtained with it are referred to as PAM activity. In both atrium and ventricle, levels of PHM activity were about 5fold higher than levels of PAM activity. However, changes in PHM and PAL activities with age in vivo paralleled changes in PAM activity reported previously. Secretion

of PHM and PAL Activities

Time-course experiments performed on 1 -week-old cultures of atrial and ventricular myocytes indicated that the accumulation of both PHM and PAL activities in the medium was linear for over 20 h (data not shown). Therefore, secretion of PHM and PAL activities was measured by collecting spent medium for 24-h intervals throughout the 2-week culture period (Fig. 2). Secretion of PHM and PAL activities by atrial cultures increased 2-fold between the fifth and seventh days of culture and remained at the higher level for the rest of the culture period (Fig. 2A). Since the PHM and PAL activities in the atrial cultures doubled over the same time period, the rate of secretion of enzyme activity, expressed as a percentage of the cell content of enzyme activity, remained constant throughout the 2-week period in culture. Atrial myocytes secreted 20 + 3% of their cellular content of PHM activity and 8 + 1% of their cellular content of PAL activity per h. Secretion of PHM and PAL activities by ventricular cultures also remained constant throughout the 2-week period in culture. Ventricular cultures secreted 20 + 4% of their cellular content of PHM activity and 5 f 1% of their cellular content of PAL activity per h. To achieve the desired level of sensitivity, PHM and PAL assays were routinely carried out with concentrations of substrate substantially below the K, of each enzyme for its respective peptide substrate; the specific activities reported could thus be affected by changes

Vol6No.12

in K, as well as changes in maximum velocity. Since visualization of PAM proteins on Western blots was consistently more difficult when using spent medium than when using culture extracts, we determined the K, of PHM for its peptidylglycine substrate. While K, values of 15 PM for acetyl-Tyr-Val-Gly were measured in spent medium from atrial cultures and in extracts of neonatal atrium, a K, value of about 400 ELM was measured in extracts of g-day-old atrial cultures (data not shown). The rate at which PAM proteins are secreted may, thus, be much lower than the rate calculated based on enzyme activity measured at a low substrate concentration. Localization Cardiocytes

of PHM and PAL Proteins

in Cultured

Antibodies raised against recombinant bacterial proteins corresponding to the PHM [rPAM-(37-382)] and PAL [rPAM-(463-864)] domains of PAM were used along with fluorescein-labeled goat antirabbit antibodies to visualize PHM and PAL proteins in 7-day-old cultures of atrium and ventricle (Fig. 3). Atrial myocytes exhibited a punctate pattern of perinuclear fluorescence when visualized with antisera to PAL or PHM (Fig. 3A). The pattern observed resembled the localization of ANF in atrial myocytes (27-29) and suggested that the PAM proteins were localized with ANF in the region of the Golgi apparatus and in secretory granules. In ventricular cultures, PAL immunoreactivity was again clearly located in a perinuclear position (Fig. 3B). Although antisera to PHM gave a relatively faint signal, the signal observed was also localized to the perinuclear region and exhibited a punctate pattern. PCR Analysis Cardiocytes

of PAM mRNAs in Cultured

Levels and forms of PAM mRNA were previously shown to be subject to developmental regulation in both atrial and ventricular tissue (16). To determine whether similar changes occurred in the cardiocyte culture system, total RNA was extracted from atrial and ventricular cultures of different ages, and cDNA was synthesized using reverse transcriptase. A qualitative measure of the forms of PAM mRNA present was obtained by carrying out PCR amplification using pairs of oligonucleotide primers that distinguish among the different alternatively spliced forms of PAM mRNA (Fig. 4A). The major forms of PAM mRNA found in adult atria and ventricles are rPAM-1 and -2, which differ by the absence of exon A (315 nt) in rPAM-2 (9). The primer pair spanning exon A distinguishes between mRNAs of the PAM-l type and mRNAs of the rPAM-2, -3, -3a, or -3b type (Fig. 48); mRNAs of the rPAM-4 and -5 type are not amplified by this primer pair. Throughout the culture, rPAM-2 mRNA was the predominant form of PAM mRNA in both atrial and ventricular cells. In both atrial and ventricular cultures, mRNAs of the rPAM-1 type increased in prevalence

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 July 2015. at 08:12 For personal use only. No other uses without permission. . All rights reserved.

Developmental

Expression

A.

= s 2 g

2001

of PAM

B.

ATRIUM

14

1.4

12

1.2

10

= 3 2

8

1.0 0.8

6

0.6

4

0.4

2

0.2

0

VENTRICLE

0.0 5

7

II

14

5

7

II

14

Days mculture

Day? mculture

Fig. 2.

Secretion of PHM and PAL Activities by Atrial and Ventricular Cell Cultures PHM and PAL activities were measured In medium collected from cultures of the indicated age; cultures had been fed with fresh medium 24 h earlier. The activity reported corresponds to the total secretion from a 35mm well of cells initially plated at a density of 2 x 1 O5 cardiac cells/cm*. Results are the mean of three individual experiments, performed in duplicate. Vertical bars indicate the SEM.

after

5-7

days

in

culture.

Increased

expression

of

rPAM-1 occurred with a similar time course in atrial and ventricular cultures. Thus, the appearance of mRNAs of the rPAM-1 type around postnatal day 5 in vivo was mimicked in cultured atrial and ventricular cardiocytes. In contrast, the decline in mRNAs of the rPAM-2 type in vivo was not mimicked in vitro. The primer pair spanning the region referred to as exon B [258 nucleotides (nt)] distinguishesbetween mRNAsof the rPAM-1 or -2, -3a, -3b, and -3 types (Fig. 4C). Elucidationof the exon/intron structure of the gene encodingrat PAM (8) indicatesthat exon B iscomprised of two exons (exons 25 and 26): exon 25 (204 nt) encodes the frans-membranedomain of PAM and is followed by the 54 nt exon 26. Alternative splicingcan exclude either both or neither of these exons to generate the forms of PAM mRNA observed. In culturesof all ages,the predominantsplicingpattern includedboth of these exons; smallamounts of rPAM-Sa, -3b, and 3 were observedat every age. At no stagewere mRNAs encoding rPAM3a, -3b, or -3 predominant. Adult atrium andventricle contain only smallamounts of rPAM-4 and -5 mRNA (10). In atrial and ventricular cultures of all ages, rPAM-4 and rPAM-5 mRNAs were barely detectable by reverse transcription and PCR amplification with form-specific primers (data not shown). The amounts of rPAM-4 and rPAM-5 mRNA detected did not vary over the 2-week culture period. Fig. 3. Localization of PHM and PAL Proteins by Immunofluorescence Neonatal atrial (A) and ventricular (B) cells were cultured for 7 days on chamber slides, fixed with paraformaldehyde, permeabilized with Triton X-l 00, and incubated overnight with a 1:2000 dilution of rabbit polyclonal antibody to PHM (A) or PAL (8). lmmunoreactive cells were visualized with fluorescein isothiocyanate-conjugated goat antibody to rabbit immunoglobulin. Bar = 33 pm.

Western Blot Analysis Antibodies specificto the different domainsof the PAM1 protein (PHM, exon A, PAL, and COOH-terminal domain)were used to identify PAM proteins present in extracts of atrial tissue (Fig. 5) and atrial cultures (Fig. 6). Along with age-dependentvariation in the forms of PAM mRNA expressed, differences in rates of translation as well as oosttranslationalorocessinacan contrib-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 July 2015. at 08:12 For personal use only. No other uses without permission. . All rights reserved.

MOL ENDO. 1992 2002

Vol6No.12

A. exon A

rat PAM-l

exon B /I/,-,

-2 -3 -3a -3b -4

L,

\I V

STOP -5

VENTRICLES

ATRIA

3

4

5

7

II

IS

3

4

5

7

II

BE 56

BE4

IS

r-+

cl

-rPAM-1 -

rPAM2/3/3a/3b

C. ATRIA 3

4

5

7

BE3

VENTRICLES II

IS

3

4

5

7

II

BE 16

r---+ 4 I exon B

IS

rPAM-

I

l/2

rPAM-3b rPAM-3a rPAM-3 PCR Analysis of Forms of PAM mRNA in Cardiocyte Cultures A, Schematic representation of seven forms of rat PAM mRNA (2, 10, 11). Total RNA was extracted from cultured atrial and ventricular cells of the indicated age, and 1 pg was subjected to reverse transcription and amplification using oligonucleotide primers spanning the major sites of alternative splicing. B, PCR amplification was carried out using a primer pair (BE4/BE56) that spans exon A (315 nt) and generates fragments of 453 bp for rPAM-1 and 138 bp for rPAM-2. C, PCR amplification was carried out using a primer pair (BEB/BE16) that spans exon B (258 nt) and generates fragments of 901 bp for rPAM-1 and rPAM-2. The amplified products were fractionated on 1.2% agarose gels and stained with ethidium bromide. Plasmids containing cDNA inserts encoding each of the forms of PAM were amplified and used to identify the products. The data presented are representative of three individual experiments. Fig. 4.

ute to the presence of variable amounts of several different PAM proteins. Extracts prepared from neonatal and adult atria contained varying proportions of two major PAM proteins [120 and 105 kilodaltons (kDa)] visualized by antisera to both PHM and PAL (Fig. 5).

The 120-kDa, but not the 1 OSkDa, protein was visualized by antiserum to exon A (not shown), suggesting that these proteins represent rPAM-1 and -2, respectively. The 105-kDa PAM protein was more prevalent in the early postnatal period, while the 120-kDa PAM

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 July 2015. at 08:12 For personal use only. No other uses without permission. . All rights reserved.

Developmental Expression of PAM

2003

PHM b--ii-l

PAL 11

3

5

7

11 14 Ad 1

5. Western Blot Analysis of PAM Proteins during Atrial Development in Viva Atria from rats of the indicated ages were extracted, fractionated on 10% sodium dodecyl transferred to Immobilon-P membranes. Each sample contained about 40 pg protein and 30-40 pmol/h PAL activity). PAM proteins were visualized with anti-PHM or anti-PAL antibodies. The indicates the age of the animal in days (Ad, adult). Apparent molecular masses of the PHM (kilodaltons). Fig.

PHM

sulfate-polyacrylamide gels, and pmol/h PHM activity (450-600 number at the top of each lane and PAL proteins are indicated

exon A

PAL 15

7

11

14

t-

80

6. Western Blot Analysis of PAM Proteins in Atrial Cell Cultures Extracts of atrial cultures of different ages were prepared; samples (40 Fg protein) containing 20-30 pmol/h PHM activity and 600-800 pmol/h PAL activity were analyzed as described in Fig. 5. Blots were consecutively analyzed for PHM, PAL, and exon A. Numbers at the top of each lane indicate the duration of the cell culture in days. Fig.

protein was more prevalent at later times; this shift in prevalence was more apparent with the antiserum to

PAL than with the antiserumto PHM and agreeswith the observed changes in mRNA prevalence (16). The 97-kDa protein visualized by antiserum to PHM was alsoapparent upon longer exposure of blots visualized with the PAL antiserum; this protein could represent intact PAM-3 or a processedform of PAM-l or -2. A 75kDa PAL protein appearedwith a time course similar to that of PAM-l; a signalof similarintensity was not

observedat 75 kDa with the PHM antibody, suggesting that the 75kDa PAL protein may be produced from PAM-l by endoproteolytic cleavage.Smallermonofunctional forms of PHM and PAL were not observedin day 5-l 4 neonatalatrium. Extracts of atrial cultures ranging in age from 5-14 days were analyzed in the same fashion and found to contain PAM proteins similarto those in atrial extracts prepared from pups of the correspondingage (Fig. 6). PAM proteinsof 120 and 105 kDa were againprevalent;

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 July 2015. at 08:12 For personal use only. No other uses without permission. . All rights reserved.

MOL 2004

END0.1992

both were visualized by antisera to the COOH-terminal domain (data not shown). The 120-kDa, but not the 105-kDa, PAM protein was recognized by antiserum to exon A. Upon longer exposure, the 97-kDa protein visualized by antiserum to PHM was also visualized by antiserum to PAL. An 80-kDa protein visualized by antisera to exon A, PAL, and the COOH-terminal domain was more prevalent in extracts of older cultures; proteins of this size were only faintly visualized with antiserum to PHM. Although not identical in size to the 75kDa PAL protein observed in vivo, visualization of an 80-kDa protein by antiserum to exon A suggests that the 80-kDa protein could be derived from rPAM-1. In accordance with the increased prevalence of mRNA encoding rPAM-1 in older cultures, the 120-kDa PAM protein and the 80-kDa PAL protein become prominent only after the first week in culture. Consistent with the maintained high level of expression of PAM-2 mRNA in atrial cultures (Fig. 4) high levels of the 105-kDa PAM protein persisted throughout the 2-week period in culture. Smaller monofunctional forms of PHM and PAL were not observed in atrial cultures. PAM proteins secreted by atrial cardiomyocytes must be generated by endoproteolytic cleavage between the PAL domain and the putative trans-membrane domain. The use of a cell culture system provides the opportunity to analyze the secreted proteins. Aliquots of spent medium from atrial cultures ranging in age from 5-14 days were concentrated and subjected to Western blot analysis (Fig. 7). PHM proteins of 102, 86, and 81 kDa were detected in each sample of spent medium analyzed. All of the proteins visualized with the PHM antibody were also visualized with the PAL antibody (data not shown); no monofunctional PHM or PAL proteins were detected in the spent medium. Only the 102-kDa PAM protein was visualized by antiserum to exon A. Thus, the 102-kDa PAM protein is a product of rPAM1, while the 86- and 81 -kDa PAM proteins are products of PAM mRNAs lacking exon A (primarily PAM-2). Despite the increased prevalence of 120-kDa rPAM-1 in extracts of older cultures, secretion of the 102-kDa product derived from rPAM-1 was not increased correspondingly.

DISCUSSION Both the high level of PAM expression in the cardiac atrium and its complex pattern of regulation during development strongly suggest that this protein subserves an important role in cardiomyocytes. Our previous study on PAM expression in the developing heart used an enzyme assay that measured the overall conversion of peptidylglycine substrate into an a-amidated product (16). With elucidation of the bifunctional nature of PAM and development of assays for the component enzyme activities, we could compare levels of PHM and PAL activities during development in vivo and in primary atrial and ventricular cardiomyocytes maintained in

Vo16No.12

serum-free medium for various periods of time. Levels of PHM and PAL activities observed in vitro were stable in time and similar in magnitude to the levels observed in viva. Levels of PHM activity were significantly higher than levels of PAM activity measured previously at alkaline pH using a different peptide substrate. Both in vivo and in vitro, the amount of PAL activity greatly exceeded the amount of PHM activity in both atrium and ventricle. When assayed in a similar fashion, stably transfected AtT-20 corticotrope tumor cells and hEK293 embyronic kidney cells expressing PAM-l or PAM2 contained similar levels of PHM and PAL activities (15, 30). The different ratios of PAL to PHM activity observed in vivo and in heart cultures suggests that cardiomyocytes may process or store PAM in a different way. PAM proteins in adult bovine atrium are located in a perinuclear position in a pattern resembling that obtained with antiserum to ANF (27-29). The PHM and PAL proteins in cultured atrial cardiomyocytes also exhibited a punctate perinuclear pattern of staining; a ring of immunofluorescence surrounded the entire nucleus. Although levels of expression were lower in ventricular cardiomyocytes, PAL proteins were also localized to the perinuclear region, often outlining the entire nucleus. The PAM proteins in stably transfected AtT-20 and hEK-293 cells expressing PAM-l or PAM2 exhibited a significantly different appearance. AtT-20 cells exhibited intense punctate staining of PAM proteins in cell processes and in an eccentric perinuclear position coincident with markers for the Golgi apparatus (15). The PAM proteins in stably transfected hEK-293 cells were concentrated in an eccentric perinuclear position coincident with markers for the Golgi apparatus, with punctate staining throughout the cell and a small percentage of PAM on the cell surface (30). The distinctive perinuclear localization of endogenous PAM proteins in atrial cardiomyocytes indicates that the protein is stored somewhat differently. Developmental regulation of PAM expression has been observed in the pancreas (31) and hypothalamus (32) as well as in the heart (16). In vivo, PAM-l mRNA becomes more prevalent than PAM-2 mRNA in the heart after approximately postnatal day 7 (16). Atrial and ventricular cardiomyocytes maintained in serumfree culture exhibited a rise in the levels of PAM-l mRNA at the corresponding time; in contrast to the situation observed in viva, the amount of PAM-2 mRNA in the cultures did not decline dramatically after postnatal day 7. Thus, at least some of the developmental changes in PAM expression occur in the absence of neuronal or hormonal input. At no time were PAM mRNAs encoding soluble bifunctional or monofunctional proteins prevalent. The cultures provide a system that can be used to determine what factors control these changes in PAM expression. Similar PAM proteins were identified in Western blots of extracts of neonatal atrium and atrial cultures of the corresponding age. The 120-kDa PAM protein recognized by antisera to PHM, exon A, PAL, and the COOH-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 July 2015. at 08:12 For personal use only. No other uses without permission. . All rights reserved.

Developmental Expression of PAM

2005

PHM

exon

A

Fig. 7. Western Blot Analysis of PAM Proteins Secreted by Atrial Cell Cultures Medium from atrial cell cultures was collected 24 h after feeding, concentrated, and fractionated, as described in Fig. 5; blots were analyzed consecutively with anti-PHM and antiexon A antibodies. The duration of time in cell culture is indicated at the top of each lane (days). The samples contained between 40-90 pmol/h PHM activity.

terminal domain represents intact PAM-l and becomes more prominent both in vivo and in vitro after postnatal day 7. The 105kDa PAM protein recognized by antisera to PHM, PAL, and the COOH-terminal domain, but not by antisera to exon A, represents intact PAM-2. As observed for PAM-2 mRNA, the PAM-2 protein becomes less prevalent in vivo after postnatal day 7. In vitro the PAM-2 protein, like the PAM-2 mRNA, remains prominent throughout the culture period. The minor 97kDa protein visualized by antisera to PHM and PAL may represent intact PAM-3 or a processed form of integral membrane PAM. A collection of smaller proteins (75-80 kDa) was detected in extracts of atrium and atrial cultures by antisera to PAL, exon A, and the COOH-terminal domain only when PAM-l was a prominent product; biosynthetic labeling studies should reveal whether they are products of PAM-l processing. Like proatrial natriuretic factor (22, 24), most of the PAM protein stored in atrial cardiomyocytes is stored in an unprocessed form; the PAM proteins detected in spent medium are not stored in large amounts in atrial granules. The major PAM proteins expressed in the heart are integral membrane proteins, and secretion of enzyme activity requires at least one endoproteolytic cleavage to separate the catalytic domains from the putative transmembrane domain. Both enzyme activities could

be measured in spent medium. When expressed as the percentage of cell content of enzyme activity secreted per h, the rate of secretion of PHM activity (20%/h) by atrial or ventricular cardiomyocytes exceeded the rate of secretion of PAL activity (5-8%/h). AtT-20 cells expressing PAM-1 secrete PHM activity more rapidly than PAL activity because an endoproteolytic cleavage within exon A creates a soluble monofunctional PHM protein and an integral membrane monofunctional PAL protein. However, Western blot analysis indicates that atrial cardiomyocytes secreted bifunctional PAM proteins. Preliminary studies suggest that the kinetic parameters of PHM are altered upon secretion; a decrease in the K, for acetyl-Tyr-Val-Gly upon secretion may contribute to the apparently higher secretion rate for PHM. The 102-kDa protein visualized by antisera to PHM, exon A, and PAL must be derived from PAM-l. The 86- and 81-kDa proteins visualized by antisera to PHM and PAL, but not by antiserum to exon A could be derived from PAM-2 or the small amounts of PAM3 present. Cultured atrial cardiomyocytes secrete approximately 15% of their content of ANF/h, while cultured ventricular cardiomyocytes secrete 50% of their content of ANF/h (27). Based on the prevalence of PAM mRNAs (17) and on incorporation of [35S]methionine into newly synthesized immunoprecipitable PAM proteins, PAM is a major

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 July 2015. at 08:12 For personal use only. No other uses without permission. . All rights reserved.

MOL 2006

ENDO.

1992

Vol6No.12

protein in the atrium. Although the major peptide product of atrial cardiomyocytes, ANF, is not a-amidated, endoproteolytic cleavage of proatrial natriuretic factor after different single Arg residues could generate sites for a-amidation (33). Cardiomyocytes also produce mRNAs encoding several propeptides that contain potential sites for endoproteolytic processing at paired basic amino acid sequences and Lu-amidation: proenkephalin (34, 35) brain natriuretic peptide (36), proneuropeptide-Y (37) insulin-like growth factor-l (38, 39) and transforming growth factor-p1 (40,41). The amount of PAM mRNA and protein in the atrium far exceeds that present in endocrine cells (e.g. AtT-20) that successfully a-amidate all of the product peptides produced. Alternatively, PAM may serve an entirely different function in the atrium. Protein disulfide isomerase, an abundant protein in the endoplasmic reticulum, is also a subunit of prolyl hydroxylase (42) and d-crystallin, the major soluble protein of the lens, has arginosuccinate lyase activity (43). The cardiomyocyte cultures should provide a tool with which to investigate alternate functions for this protein.

MATERIALS Preparation

AND METHODS of Cardiocyte

Cell Cultures

ylmethylsulfonylfluoride, 1 pg/ml limabean trypsin inhibitor, 1.6 @g/ml benzamidine, 2 pg/ml pepstatin, and 20 pg/ml DNase-I. Homogenization was carried out with a Brinkmann Polytrontype homogenizer (Westbury, NY); homogenates were frozen and thawed twice and then centrifuged at 400 x g for 10 min at 4 C. The supernatants were removed and assayed for protein content using the bicinchoninic acid protein assay kit with BSA as a standard (Pierce, Rockford, IL). Spent medium was removed from cells and cleared of cell debris by centrifugation at 400 x g for 5 min. For Western blot analysis, cells were incubated in serum-free medium lacking BSA, and spent media were concentrated approximately 5-fold with Centricon 10 filter units (Amicon, Danvers, MA). To determine whether PAM proteins were stable after secretion, biosynthetically labeled PAM proteins secreted by stably transfected AtT-20 cells expressing rPAM-2 (15) were incubated with atrial cardiomyocytes for 5 h and isolated by immunoprecipitation; no detectable degradation of the PAM proteins occurred. Enzyme assays were performed as described previously (7); samples (1 Fg tissue or cell extract protein or 2 ~1 24-h spent medium) were assayed in duplicate. PHM reactions were carried out in a final volume of 40 ~1 containing 140 mM NaMES (pH 5.0) 0.5 PM CuS04, 100 pg/ml catalase, 0.5 mM ascorbate, 0.5 WM a-N-Acetyl-Tyr-Val-Gly, and 15,000 cpm [‘251]a-N-acetyl-Tyr-Val-Gly for 1 h. Dose-response studies indicated that 0.5 GM CuSO., gave the maximum level of activity. The reaction was linear in time for up to 3 h and in amount of protein up to 2 pg. PAL reactions were carried out in a final volume of 40 ~1 containing 140 mM Na 2-(A!-morpholino)ethanesulfonic acid (MES) (pH 5.0) 0.05% Lubrol-PX, 0.5 PM a-N-acetyl-Tyr-Vala-hydroxyglycine and 15,000 cpm [‘Z51]~-N-acetyI-Tyr-Val-~hydroxyglycine for 30 min. The PAL reaction was linear in time for up to 30 min and in amount of protein up to 1 pg.

Atrial and ventricular tissues were taken from 40 neonatal rats (Sprague-Dawley) on the day of birth (postnatal day 1) and

Western

processed separately using a trypsin-coljagenase dissocjation protocol (44, 45). After rinsinq in Joklik-Minimal Essential Medium, the tissue was minced-briefly, and then incubated in 5 ml Joklik-Minimal Essential Medium containing 0.1% trypsin (ICN, Costa Mesa, CA) and 0.01% collagenase (Worthington Biochemical Corp., Freehold, NJ) for 10 min at 37 C with stirring. The fragments of tissue were allowed to settle, and the supernatant, containing primarily noncardiocyte cells, was discarded. Tissue was then subjected to three cycles of digestion for 7 min at 37 C. The three supernatants were pooled and passed through a 150-pm nylon filter, and cells were pelleted by centrifugation at 400 x g for 10 min. The cells were resuspended in Dulbecco’s Modified Eagle’s MediumHam’s F-l 2 (DMEM/F12) supplemented with 10% fetal bovine serum (Sigma, St. Louis, MO) and preplated in a culture flask for 30 min; the medium was removed, and nonadherent cells were collected by centrifugation at 400 x g for 5 min. Cells were resuspended in DMEM/F12 containing 10% fetal bovine serum and plated at a density of 2 x lo5 cells/cm2 (46) on fibronectin (Sigma)-coated dishes (4 pg fibronectin/cm’). Cell yields were typically 0.6 x lo6 atrial cells and 1.5 x lo6 ventricular cells/rat pup. To limit further proliferation of fibroblasts, the cultures were treated with 10 PM cytosine-b-oarabinofuranoside (Sigma) for 24 h. On the second day of culture, the cells were fed with a serum-free medium [DMEM/ F12 containing 10 pq/ml insulin, 5 &ml transferrin, 1 Om9 M selenium, 1 mg/ml BSA, 1O-9 M TS, 26 nM corticosterone, 15 mM HEPES (DH 7.4). and 2 UM cvtosine-B-o-arabinofuranoside]. The ceiis were subsequently’maintained in this serumfree medium in the absence of cytosine-P-o-arabinofuranoside.

Aliquots of extracts and media were fractionated on 1.5-mm thick sodium dodecyl sulfate-polyacrylamide gels containing 10% acrylamide and 0.27% N,N’-methylenebisacrylamide (7). Proteins were electrophoretically transferred to Immobilon-P membranes (Millipore, Bedford, MA) in 25 mM Tris, 200 mrv glycine (pH 8.5) and 20% methanol, for 2 h at 500 mamp. Membranes were blocked by incubation in 20 mM Tris-HCI (pH 7.4), 150 mM NaCI. and 0.05% Tween-20 containina 10% dried milk, and PAM proteins were visualized with antgodies to recombinant proteins derived from several domains of the PAM protein. Recombinant PHM [rPAM-(37-382)] and PAL [rPAM-(463864)] were purified from the periplasmic space of transformed E. co/i (Yun, H.-Y., personal communication); recombinant exon A [rPAM-(409-497)] and COOH-terminal domain [rPAM(898-976)] were purified from the cytosolic fraction. Rabbit polyclonal antisera were produced at Hazleton Laboratories (Denver, PA). Antibody 475 recognizes the PHM domain, antibody 471 recognizes the PAL domain, antibody 629 recognizes exon A, and antibody 571 recognizes the COOH-terminal domain. These antibodies were used at a 1:2000 dilution in the same buffer used to dissolve the blocking agent, and cross-reactive proteins were visualized using the ECL reagent (Amersham, Arlington Heights, IL). Antipeptide antibodies specific for a peptide-in the PHM domain previously revealed 40to 55-kDa PHM proteins in extracts of atrial cultures and in spent medium (47); since PHM proteins of this size were not identified by the more potent antisera to the entire PHM domain, we must conclude that these proteins were stained nonspecifically.

Preparation of Samples Enzyme Activities

RNA Extraction/Reverse

for Measurement

of Protein

and

Cells and tissues were extracted in 1% Triton X-100, 10 mM mannitol, 20 mM Na N-tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid (pH 7.0) supplemented with 30 pg/ml phen-

Blot Analysis

Transcription-PCR

Total RNA was extracted from cell cultures following the method of Chomzinski and Sacchi (48). Reverse transcription was performed using 1 pg total RNA and avian myeloblastosis virus reverse transcriptase, as described by the manufacturer

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 July 2015. at 08:12 For personal use only. No other uses without permission. . All rights reserved.

Developmental

Expression

2007

of PAM

(Promega, Madison, WI). Amplification of cDNA derived from 100 nq total RNA was performed in a 5O-wl reaction volume with a-buffer consisting’ of 10 mM Tris-HCI (pH 8.3) 50 mM KCI. 1.5 mM MaCI,. and 0.1% aelatin in the oresence of 200 PM of each deGxy:NTP, 0.4 $I of each primer, and 2.5 U Amplitaq DNA polymerase (Perkin Elmer Cetus, Norwalk, CT). Thirty cycles of annealing at 52 C for 1 min, elongation at 72 C for 3 min, and denaturation at 95 C for 1 min were performed on a Programmable Thermal Controller (MJ Research, Inc., Watertown, MA). Samples were factionated on 1% agarose gels in 45 mM Tris-borate (pH 8.0)-l mM EDTA and visualized by staining with ethidium bromide.

5.

6.

7.

8. Cells were plated on chamber slides (LabTek, Naperville, IL) coated with fibronectin (4 pg/cm’). After 8 days of culture, cells were processed for immunofluorescence, as described by Shields et a/. (23); cells were fixed with 3% paraformaldehyde in 2.6 mM KCI, 1.5 mM KH>PO,,, 136 mM NaCI, and 8 mM Na2HP04, pH 7.4 (PBS), for 20 min, washed with PBS, incubated in 50 mM NH&I for 10 min, and permeabilized by incubation with PBS containing 0.2% Triton X-l 00 for 15 min. Slides were washed three times with PBS and blocked by incubation in 0.1% gelatin-PBS for 1 h. PAM antibodies were applied at a dilution of 1:2000 in 0.1% gelatin-PBS, and slides were incubated overnight at 4 C. Cells were then washed with PBS (three times in 1 h) and fluorescein isothiocyanate-goat antirabbit immunoglobulin G (CalTag) was applied for 1 h at a dilution of 1:400. Slides were washed for 1 h with PBS, mounted with Permafluor aqueous mounting medium (lmmunon) containing 350 mM 1,4-diazabicyclo-[2.2.2]octane and 30 mM glycine, pH 10, and examined under epifluorescence optics with a Zeiss Axioskop microscope (BP 485/20, barrier filter 520-560, Zeiss, New York, NY). Binding specificity was established using the respective preimmune sera.

9.

10.

11.

12.

13.

14.

Acknowledgments We would like to thank Hye-Young Yun and Richard Johnson for preparing the antigens used to generate antisera to PAM, and Dick Mains, Ana Maria Oyarce, Jean Husten, and Sharon Milgram for help in preparing this manuscript.

15.

16. Received July 14, 1992. Revision received August 28,1992. Accepted September 8, 1992. Address requests for reprints to: Dr. Betty Eipper, Department of Neuroscience, Johns Hopkins University School of Medicine, 725 North Wolfe Street, Baltimore, Maryland 21205. This work was supported by grants from the American Heart Association (890831) the NIH (DK-32949) and the NIDA (DA-00098).

17.

18.

19.

REFERENCES 20. Bradbury AF, Smyth DG 1991 Peptide amidation. Trends Biochem Sci 16:112-l 15 Eipper BA, Stoffers DA, Mains RE 1992 The biosynthesis of neuropeptides: peptide o-amidation. Annu Rev Neurosci 15:57-85 Katopodis AG. Ping D, May SW 1990 A novel enzyme from bovine neurointermediate pituitary catalyzes dealkylation of a-hydroxyglycine derivatives, thereby functioning sequentially with peptidylglycine a-amidating monooxygenase in peptide amidation. Biochemistry 29:61156120 Takahaski K, Okamoto H, Seino H, Noguchi M 1990 Peptidylglycine a-amidating reaction: evidence for a two-

21.

22.

23.

step mechanism involving a stable intermediate at neutral pH. Biochem Biophys Res Commun 169:524-530 Perkins SN, Husten EJ, Eipper BA 1990 The 108 kDA peptidylglycine a-amidating .monooxygenase precursor contains two seoarable enzvmatic activities involved in peptide amidation. Biochem Biophys Res Commun 171:926-932 Kato I, Yonekura HTajima M, Yanagi M, Yamamoto H, Okamoto H 1990 Two enzymes concerned in peptide hormone a-amidation are synthesized from a single mRNA. Biochem Biophys Res Commun 172:197-203 Husten EJ, Eipper BA 1991 The membrane-bound bifunctional peptidylglycine a-amidating monooxygenase protein J Biol Chem 266:17004-l 7010 Ouafik L’H, Stoffers DA, Campbell TA, Johnson RC, Bloomquist BT, Mains RE, Eipper BA 1992 The multifunctional DeDtidVlolVCine a-amidatina monooxygenase gene: exon/intron loiganization of c;alytic, processing and routing domains. Mol Endocrinol 6:1571-1584 Stoffers DA, Green CBR, Eipper BA 1989 Alternative mRNA splicing generates multiple forms of peptidyl-glytine cu-amidating monooxygenase in rat atrium. Proc Natl Acad Sci USA 86:735-739 Stoffers DA, Ouafik L’H, Eipper BA 1991 Characterization of novel mRNAs encoding enzymes involved in peptide cu-amidation. J Biol Chem 266:1701-l 707 Eipper BA, Green CBR, Campbell TA, Stoffers DA, Keutmann HT, Mains RE, Ouafik L’H 1992 Alternative splicing and endoproteolytic processing generate tissue-specific forms of pituitary peptidylglycine cu-amidating monooxygenase. J Biol Chem 267:4008-4015 Eipper BA. May V, Braas KM 1988 Membrane-associated peptidylglycine a-amidating monooxygenase in the heart. J Biol Chem 263:8371-8379 Bradbury AF, Smyth DG 1988 Peptide amidation: evidence for mutiple molecular forms of the amidating enzyme. Biochem Biophys Res Commun 154:1293-l 300 Beaudry GA, Bertelsen AH 1989 Secreted alpha-amidating enzymes are generated by specific posttranslational processing of precursors containing transmembrane domains. Biochem Biophys Res Commun 163:959-966 Milgram SL, Johnson RC, Mains RE 1992 Expression of individual forms of peptidylglycine oc-amidating monooxygenase in AtT 20 cells: endoproteolytic processing and routing to secretory granules. J Cell Biol 117:717-728 Ouafik L’H, May V, Keutmann HT, Eipper BA 1989 Developmental regulation of peptidylglycine n-amidating monooxygenase (PAM) in rat heart atrium and ventricle. J Biol Chem 264:5839-5845 Braas KM, Stoffers DA, Eipper BA, May V 1989 Tissue specific expression of rat peptidylglycine n-amidating monooxygenase activity and mRNA. Mol Endocrinol 3:1387-1398 Sakata J, Mizuno K, Matsuo H 1986 Tissue distribution and characterization of peptidylglycine a-amidating activity in rat. Biochem Biophys Res Commun 140:230-236 Braas KM, Harakall SA, Ouafik L’H, Eipper BA, May V 1992 Expression of peptidylglycine cY-amidating monooxygenase: an in situ hybridization and immunocytochemical study. Endocrinology 130:2778-2788 de Bold AJ 1985 Atrial natriuretic factor P: a hormone produced by the heart. Science 230:767-770 Glembotski CC, Wildey GM, Gibson TR 1985 Molecular forms of immunoactive atrial natriuretic peptide in the rat hypothalamus and atrium. Biochem Biophys Res Commun 129:671-678 Miyata A, Kangawa K, Toshimori T, Hatoh T, Matsuo H 1985 Molecular forms of atrial natriuretic polypeptides in mammalian tissues and plasma. Biochem Biophys Res Commun 129:248-255 Shields PP, Dixon JE, Glembotski CC 1988 The secretion of atrial natriuretic factor-(99-l 26) by cultured cardiac myocytes is regulated by. glucocrticoids. J Biol Chem 263:12619-l 2628

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 July 2015. at 08:12 For personal use only. No other uses without permission. . All rights reserved.

MOL 2008

24.

25.

26.

27.

28.

29.

30.

31

32

33

34.

ENDO.

1992

Glembotski CC, Dixon JE, Gibson TR 1988 Secretion of atrial natriuretic factor-(l-98) by primary cardiac myocytes. J Biol Chem 263:16073-l 6081 Shields PP, Glembotski CC 1989 Regulation of atrial natriuretic factor-(99-l 26) secretion from neonatal rat primary atrial cultures by activiators of protein kinases A and C. J Biol Chem 264:9322-9328 Thiele EA, Marek KL, Eipper BA 1989 Tissue-specific regulation of peptidylglycine a-amidating monooxygenase expression. Endocrinology 1252279-2288 Bloch KD, Seidman JG, Naftilan JD, Fallon JT, Seidman CE 1986 Neonatal atria and ventricles secrete atrial natriuretic factor via tissue-specific secretory pathways. Cell 47:695-702 Hassall CJS, Wharton J, Gulbenkian S, Anderson JV, Frater J, Bailey DJ, Merighi A, Bloom SR, Ploak JM, Burnstock G 1988 Ventricular and atrial myocytes of newborn rats synthesise and secrete atrial natriuretic peptide in culture: light and electron microscopical localisation and chromatographic examination of stored and secreted molecular forms. Cell Tissue Res 251:16161619 Watanabe T, Watanabe M, lshii Y, Matsuba H, Kimura S, Fujita T, Kominami E, Katunuma N, Uchiyama Y 1989 An immunocytochemical study on colocalization of cathepsin B and atrial natriuretic peptides in secretory granules of atrial myoendocrine cells of rat heart. J Histochem Cytothem 37:347-351 Tausk FA, Milgram SL, Mains RE, Eipper BA, Identification of a routing domain in a peptide processing enzyme. Mol Endocrinol, in press Ouafik L’H, Giraud P, Salers P, Dutour A, Castanas E, Boudouresque F, Oliver C 1987 Evidence for high peptide a-amidating activity in the pancreas from neonatal rats. Proc Natl Acad Sci USA 83:261-264 Faivre-Bauman A, Loudes C, Barret A, Patte C, TixierVidal A 1988 Ontogenesis of peptidylglycyl cu-amidation activity in the mouse hypothalamus in viva and in serumfree medium cultures. Dev Brain Res 40:261-267 Seidman CE, Duby AD, Choi E, Graham RM, Haber E, Homey C, Smith JA, Seidman JG 1984 The structure of rat preproatrial natriuretic factor as defined by a complementary DNA clone. Science 225:324-326 Springhorm JP, Claycomb WC 1989 Preproenkephalin

Vol6No.12

35.

36.

37. 38.

39.

40.

41. 42.

43. 44. 45.

46.

47.

48.

mRNA expression in developing rat heart and in cultured ventricular cardiac muscle cells. Biochem J 258:73-78 Low KD, Allen RG, Melner MH 1990 Association of proenkephalin transcipts with polyribosomes in the heart. Mol Endocrinol 4:1408-l 415 Minamino N, Kangawa K, Matsuo H 1988 Isolation and identification of a high molecular weight brain natriuretic peptide in porcine cardiac atrium. Biochem Biophys Res Commun 157:402-409 Marek KL, Mains RE 1991 Regulation of neuropeptide Y in neural and cardiac cultures. Ann NY Acad Sci 611:463464 Rotwein P 1986 Two insulin-like growth factor I messenger RNAs are expressed in human liver. Proc Natl Acad Sci USA 83:77-81 Adamo M, Lowe WL, Le Roith D, Roberts CT 1989 Insulinlike growth factor I messenger ribonucleic acids with alternative 5’- untranslated regions are differentially expressed during development of the rat. Endocrinology 124:2737-2744 Thompson NL, Flanders KC, Smith JM, Elingsworth LR, Roberts AB, Sporn MB 1989 Expression of transforming growth factor+1 in acute myocardial infarction in rats. Growth Factors 1:91-99 Qian SW, Kondaiah P, Casscells W. Roberts AB, Sporn MB 1991 A second messnger RNA species of transfbrming factor @ in infarcted rat heart. Cell Reg 2:241-249 Noiva R, Lennarz WJ 1992 Protein disulfide isomerase A multifunctional protein resident in the lumen of the endoplasmic reticulum. J Biol Chem 267:3553-3556 Ptiagorsky J, Wistow GJ 1989 Enzyme/crystallins: gene sharing as an evolutionary strategy. Cell 57:197-l 99 Clavcomb WC 1980 Culture of cardiac muscle cells in serum-free media. Exp Cell Res 131:231-236 Libbv P 1984 Long-term culture of contractile mammalian heah cells in a defined serum-free medium that limits non muscle cell oroliferation. J Mol Cell Cardiol 16:803-811 Specierh DW, Peace JN, McCarl RL 1981 Effects of plating density and age in culture on growth and cell division of neonatal rat heart primary cultures. In Vitro 17:863-870 Maltese J-Y, Eipper BA 1991 Peptidylglycine a-amidation exoression in orimarv heart cell culture. Phvsioloaist , 34.229 (Abstract 11.24) Chomzinski P. Sacchi N 1987 Sinale steo method of RNA isolation by guanidinium thiocyaiate phenol chloroform extraction. Anal Biochem 162:156-l 59

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 July 2015. at 08:12 For personal use only. No other uses without permission. . All rights reserved.