Regulation of protein synthesis in ventricular myocytes by vasopressin. The role of sarcoplasmic/endoplasmic reticulum Ca2+ stores.

Protein synthesis in H9c2 ventricular myocytes was subject to rapid inhibition by agents that release Ca2+ from the sarcoplasmic/endoplasmic reticulum, including thapsigargin, ionomycin, caffeine, and arginine vasopressin. Inhibitions were attributable to the suppression of translational initiation and were coupled to the mobilization of cell-associated Ca2+ and the phosphorylation of eIF2alpha. Ionomycin and thapsigargin produced relatively stringent degrees of Ca2+ mobilization that produced an endoplasmic reticulum (ER) stress response. Translational recovery was associated with the induction of ER chaperones and resistance to translational inhibition by Ca2+-mobilizing agents. Vasopressin at physiologic concentrations mobilized 60% of cell-associated Ca2+ and decreased protein synthesis by 50% within 20-30 min. The inhibition of protein synthesis was exerted through an interaction at the V1 vascular receptor, was imposed at physiologic extracellular Ca2+ concentrations, and became refractory to hormonal washout within 10 min of treatment. Inhibition was found to attenuate after 30 min, with full recovery occurring in 2 h. Translational recovery did not involve an ER stress response but rather was derived from the partial repletion of intracellular Ca2+ stores. Longer exposures to vasopressin were invariably accompanied by increased rates of protein synthesis. Translational inhibition by vasopressin, but not by Ca2+-mobilizing drugs, was both preventable and reversible by treatment with phorbol ester, which reduced the extent of Ca2+ mobilization occurring in response to the hormone. Larger and more prolonged translational inhibitions occurred after down-regulation of protein kinase C. This report provides the first compelling evidence that hormonally induced mobilization of sarcoplasmic/endoplasmic reticulum Ca2+ stores is regulatory upon mRNA translation.

Ca 2ϩ sequestered by the endoplasmic reticulum (ER) 1 is essential for optimal rates of protein synthesis occurring in nucleated mammalian cells (reviewed in Ref. 1). This control involves coupling of the rates of mRNA translation to those of protein translocation into the organelle for subsequent processing or folding. Depletion of ER sequestered Ca 2ϩ to slow glycoprotein processing and transport competence, or introduction of a reducing environment to suppress ER processing of proteins with disulfide cross-links, results in the activation of the double-stranded RNA-activated protein kinase, the phosphorylation of eIF2␣, the inhibition of eIF2B, and the slowing of translational initiation (2)(3)(4)(5)(6). An adaptive response, characterized by translational accommodation to continued depletion of ER Ca 2ϩ stores by drugs such as ionomycin or thapsigargin or to the continued presence of a thiol-reducing agent, occurs over several hours. This adaptive response is dependent on increased expression of the ER resident chaperone, GRP78/BiP (1,4,7,8). Inductions of this chaperone and of recovery from translational inhibition depend on activation of grp78 transcription and, in some cell types, a growth-promoting factor. The acute suppression of translational initiation by agents that inhibit ER protein processing, the induction of the ER chaperones GRP78 and GRP94, and the recovery of translational activity are characteristic of the "ER stress response" (1,9,10). Translational suppression, however, is not required for expression of the later events. Both GRP78 and translational tolerance can be induced by Ca 2ϩ -mobilizing or thiol-reducing agents at concentrations that do not suppress protein synthesis (7).
The physiologic significance of translational suppression by conditions provoking the ER stress response is unclear. However, Ca 2ϩ -mobilizing hormones, including epinephrine, angiotensin, and vasopressin, have been found to slow protein synthesis over several minutes in isolated hepatocytes (11)(12)(13). Inhibitions were reduced during incubations in Ca 2ϩ -depleted media and overturned at supraphysiologic Ca 2ϩ concentrations (11). Corresponding reductions were observed in the polysomal contents of excised portions of perfused rat liver in response to hormones and manipulations of the Ca 2ϩ content of the perfusing medium (14). It was proposed that hormonally induced changes in intracellular Ca 2ϩ homeostasis provide a mechanism for regulating the rate of protein synthesis in normal hepatocytes. Given the transient viability and low synthetic rates of dispersed hepatocytes and the technical limitations associated with perfused rat liver, it has not been possible to correlate translational rates with changes in cell-associated Ca 2ϩ or eIF2␣ phosphorylation, to examine the reversibility of hormonally imposed inhibitions, or to ascertain whether an adaptive response occurs during continuous hormonal treatments. Although cultured cells are better suited to such studies, cell lines exhibiting translational suppression in response to hormones that mobilize Ca 2ϩ from the ER to the cytosol have not been identified.
We now report that H9c2 ventricular myocytes respond to drugs that deplete the sarcoplasmic/endoplasmic reticulum (S(E)R) of Ca 2ϩ and to physiologic concentrations of arginine vasopressin with a rapid inhibition of mRNA translation at initiation. Inhibitions are coupled to the mobilization of cellassociated Ca 2ϩ and to the phosphorylation of eIF2␣. Ionomycin and thapsigargin produced relatively stringent degrees of Ca 2ϩ mobilization in which translational recovery was associated with the induction of ER chaperones. Translational recovery from inhibition by vasopressin did not involve an ER stress response but derived instead from the partial repletion of intracellular Ca 2ϩ stores. Inhibition of protein synthesis by vasopressin became refractory to hormonal washout within 10 min of treatment. These findings support the validity of hormonally mediated regulation of mRNA translation involving mobilization of Ca 2ϩ sequestered within the S(E)R in ventricular myocytes.

EXPERIMENTAL PROCEDURES
Materials-H9c2(2-1) rat embryonic ventricular myocytes were obtained from the American Type Culture Collection at passage number 12. Arginine vasopressin, lysine vasopressin, desmopressin, endothelin-1, angiotensin II, phenylephrine, caffeine, and phorbol 12-myristate 13-acetate (PMA) were purchased from Sigma. Ionomycin and thapsigargin were from Calbiochem. Monoclonal antibody to eIF2␣ was the gift of Dr. Lynn O'Brien, University of Rochester. Ampholines were purchased from Galliard-Schlesinger (pH 4 -8) and from Oxford Glycosystems (pH 3.5-10). L- [4,  General Methodology-Stock H9c2 cells were propagated in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and subcultured before confluence. For experiments, monolayers were allowed to reach confluence after which cultures were maintained for 1 day in Dulbecco's modified Eagle's medium supplemented with 5% horse serum to promote differentiation of myoblasts into myotubes (15). Before treatments, cells were equilibrated for 5-15 min with serum-free Ham's F-10 medium modified to contain 25 mM leucine and 0.2 mM Ca 2ϩ . Leucine pulse incorporation was measured as described previously (16) for monolayers in multiwell trays (2.4 cm 2 /well) and for 15-min incubations. Incubations were conducted in triplicate and results are presented as the mean Ϯ S.E. of values obtained. Findings were reproduced on at least two separate occasions. [ 35 S]Methionine labeling, one-dimensional 7.5% polyacrylamide gel electrophoresis (SDS-PAGE) of detergent-solubilized extracts of methionine-labeled cells, and autoradiography were conducted as described previously (17). Ribosomal and polyribosomal size distributions were measured by density gradient centrifugation as described previously (18).
Measurement of Cell-associated 45 Ca 2ϩ -Cells in 12-well (4.8 cm 2 / well) plates were pre-equilibrated for 2 h in 1 ml of modified F-10 medium containing 0.2 mM Ca 2ϩ and 45 CaCl 2 (0.02 Ci/mmol) before treatments. After treatments, monolayers were washed twice with 2 ml of ice-cold buffered saline containing 2.5 mM LaCl 3 , dissolved in 500 l of 1% sodium dodecyl sulfate, and analyzed for radioactivity. 45 Ca 2ϩ in combination with 5 mM EGTA was added to non-equilibrated samples and the preparations washed and solubilized immediately to assess the contribution of extracellular 45 Ca 2ϩ . These values (3% of experimental values) were subtracted from experimental values. Results, presented in nanomoles of cell-associated Ca 2ϩ /mg of protein, are expressed as the mean Ϯ S.E. of values obtained for triplicate incubation samples. Findings were reproduced on at least two separate occasions.
Determination of the Phosphorylation State of eIF2␣-Duplicate monolayers (2.4 cm 2 ) of multiwell trays were lysed at 40°C in 200 l of sample buffer containing 3% ampholines (4 parts pH 4 -8 and 1 part pH 3-10), 0.4% Tween 20, 2% ␤-mercaptoethanol, and 9.5 M urea. Lysates were then subjected to slab gel isoelectric focusing in the presence of 9.5 M urea as described (19) to separate the phosphorylated and nonphosphorylated forms of the ␣-subunit of eIF2. Gels were treated with 1 M Tris, pH 8.8, and blotted onto polyvinylidene difluoride membranes under basic conditions. Phosphorylated and non-phosphorylated eIF2␣ were immunodetected with monoclonal antibody to eIF2 and chemiluminescence as described previously (5). Films were scanned by UMAX Magic Scan version 1.3.3. Analysis of the relative amounts of phosphorylated and non-phosphorylated subunit was performed on a Macintosh Quadra 700 computer using the public domain NIH Image program. 2

Inhibition of Protein Synthesis in H9c2 Myocytes by Vasopressin: the Roles of S(E)R Ca 2ϩ Stores and Protein
Kinase C-Agents that deplete ER Ca 2ϩ stores inhibit protein synthesis in non-muscle cells in approximately 6 -12 min (1)(2)(3)(4)(5)(6)(7)17). Similar inhibitions were observed in H9c2 myocytes exposed for 30 min to agents that release Ca 2ϩ from the sarcoplasmic reticulum or ER (Table I). The divalent cation ionophore, ionomycin, inhibited leucine incorporation and mobilized cell-associated Ca 2ϩ , with larger effects observed at 25 as compared with 5 nM drug. Thapsigargin (1 M), an irreversible inhibitor of Ca 2ϩ uptake by the S(E)R, mobilized 75% of cell-associated Ca 2ϩ while inhibiting protein synthesis by 93%. Caffeine (5 mM), which releases Ca 2ϩ from the sarcoplasmic reticulum of cardiac muscle (20), mobilized 45% of H9c2 cell-associated Ca 2ϩ and inhibited leucine incorporation by 40%. Hormones that have been reported to increase cytosolic free Ca 2ϩ ([Ca 2ϩ ] i ) in normal neonatal or H9c2 ventricular myocytes (21)(22)(23)(24)(25) were also examined for the ability to suppress protein synthesis and mobilize Ca 2ϩ stores. Phenylephrine (10 M) and epinephrine (data not shown) elicited marginal responses, whereas angiotensin II and endothelin-1 were ineffective. Arginine vasopressin (1 M), however, mobilized approximately two thirds of the total cell-associated Ca 2ϩ and suppressed protein synthesis by 45%. Inhibitions of protein synthesis in the range of 35-40% occurred at quite low arginine vasopressin concentrations (10 nM), maximized to slightly higher values (Ӎ50%) at 0.1-1 M hormone, and were attributable to an interaction at the V1 vascular receptor (Fig. 1). Lysine vasopressin, a slightly less potent agonist selective for the V1 vascular receptor, elicited similar degrees of inhibition at 0.01-10 M, whereas desmopressin, a long acting synthetic agonist selective for the V2 renal receptor, at identical concentrations did not affect leucine incorporation.
The inhibition of leucine incorporation in H9c2 cells that developed in response to the mobilization of sequestered Ca 2ϩ was readily reversed by restoration of the cation to the extracellular medium (Fig. 2). To generate Ca 2ϩ -depleted preparations, H9c2 cells were pretreated for 15 min in medium lacking Ca 2ϩ and containing 0.1 M ionomycin and 1 mM EGTA. Preparations were then washed with albumin-containing medium to remove ionophore (26) and equilibrated in medium containing 1 mM EGTA and graded increases in Ca 2ϩ . At Ca 2ϩ concentrations in the physiological range (1 mM in excess of EGTA), leucine incorporation was stimulated approximately 2 Software was developed at the National Institutes of Health and is available from the Internet by anonymous ftp from zippy.nimh.nih.gov or on floppy disk from NTIS, 5285 Port Royal Rd., Springfield, VA 22161 (PB95-500195GEI).  (27). The Ca 2ϩ dependence of the vasopressin inhibition of protein synthesis was examined as a function of increasing extracellular concentrations of the cation (Fig. 2). Vasopressin had no effect on leucine incorporation at low extracellular Ca 2ϩ concentrations but reduced incorporation at 1 mM free cation by half. Supraphysiologic (2-3 mM) extracellular Ca 2ϩ concentrations did not reverse the inhibition attributable to vasopressin. The time dependence of the inhibition of protein synthesis by vasopressin was determined for cells incubated for 2 h with 0.2 mM Ca 2ϩ , which is sufficient to maintain rates of leucine incorporation in untreated cells (Fig. 3). Within 20 -30 min of hormone addition, maximal inhibition of leucine incorporation was observed. Thereafter, pulse incorporation rates were found to rise, with full recovery of activity observed at 2 h. The addition of EGTA in excess of Ca 2ϩ produced declining pulse incorporation rates that plateaued at 50% of control values from 1-2 h. Vasopressin added in combination with the chelator provoked larger inhibitions of leucine incorporation (67% at 30 min), which were sustained through 2 h of incubation.
Activation of phospholipase C in response to occupation of V1 vascular receptors is associated with the intracellular generation of both inositol trisphosphate and diacylglycerol (27). The potential involvement of protein kinase C in translational regulation by vasopressin was therefore explored utilizing PMA as an activator of the enzyme. In control experiments, exposure of H9c2 cells to PMA for up to 30 min did not affect either leucine incorporation or cell-associated Ca 2ϩ (data not shown). By 45 min, however, leucine incorporation tended to rise modestly. The effects of vasopressin on leucine incorporation were examined in cells that had been either pretreated for 15 min with 1 M PMA or carried as untreated controls (Table II). PMA pretreatment was found to abolish the inhibition of leucine incorporation that occurred in untreated cells in response to the subsequent addition of vasopressin. PMA also fully reversed (within 5 min) pre-existing inhibitions of leucine incorporation attributable to vasopressin (data not shown). In contrast, suppression of leucine incorporation occurring in H9c2 cells treated with ionomycin, thapsigargin, sodium arsenite, or dithiothreitol or with heat shock (43°C for 20 min), was neither prevented nor reversed by PMA (data not shown). Ca 2ϩ mobilization in response to vasopressin was reduced, but not abolished, by PMA (Table II). Vasopressin mobilized approximately 37% of cell-associated Ca 2ϩ in PMA-pretreated cells as compared with 50% release in non-treated cells; this degree of mobilization was completed within 10 min. Chronic exposure to phorbol ester was also employed to investigate the effects of protein kinase C down-regulation on vasopressin signaling. Conditions were chosen that are reported to produce down-regulation of the kinase in neonatal cardiomyocytes (28). H9c2 cells were pretreated for 16 h in complete growth medium with or without the addition of PMA. The time dependence of vasopressin inhibition of protein synthesis was then examined (Fig. 4). Non-pretreated preparations responded to hormone with a rapid suppression of protein synthesis that maximized at 20 min. Thereafter, pulse incorporation rates increased such that the 2-h values exceeded those of untreated controls. In preparations exposed chronically to PMA, vasopressin evoked larger (80%) inhibitions of leucine incorporation that were sustained for 40 min. Protein synthesis in these preparations recovered from inhibition by the hormone slowly, with only partial restoration of incorporation rates observed at 2 h. The protein kinase C inhibitors, calphostin C and chelerythrine, were tested in H9c2 cells at concentrations reported to inhibit the enzyme in vivo (29,30). Each of these agents produced inhibition of leucine incorporation and the release of cell-associated Ca 2ϩ within 30 min by indeterminant mechanisms (data not shown).
Inhibition of Translational Initiation and Phosphorylation of eIF2␣ by Vasopressin-The polyribosomal contents of H9c2 cells were examined after treatments with thapsigargin, vasopressin, PMA, and PMA in combination with vasopressin ( Fig.  5). As anticipated, polysomes almost completely disappeared in response to thapsigargin, which slows initiation relative to peptide chain elongation in non-muscle cells (17). Vasopressin also reduced polysomal content, but not as dramatically as thapsigargin. Polyribosomes were preserved in incubations with PMA alone, and polysome disaggregation in response to vasopressin was reduced after brief pretreatment with the phorbol ester. To correct for potential differences in loading of gradients, the amounts of 80 S monosomes, small polysomes, and large polysomes were quantitated by absorbance at 254 nm and related to each other. For each treatment condition, the ratio of 80 S monosome:small polysome:large polysome was determined to be: control preparation, 30:10:10; thapsigargin, 67:3:0; vasopressin, 45:11:9; PMA, 24:10:10; PMA ϩ vasopressin, 35:12:12. By this analysis, vasopressin caused 80 S monosomes to increase and polysomes to decrease, typical of a slowing of initiation. Large polysomes were reduced preferentially. By contrast, PMA, which stimulates amino acid incorporation, caused polysomes to accumulate at the expense of 80 S monosomes, consistent with increased rates of initiation. Effects of vasopressin were largely overturned by PMA, such that ribosomal profiles under this condition were similar to those of controls. Thapsigargin produced almost complete initiation blockade.
Suppression of translational initiation in non-muscle cells in response to drugs that deplete ER Ca 2ϩ stores is attributable to the phosphorylation of the ␣-subunit of eIF2 by doublestranded RNA-activated protein kinase (1)(2)(3)(4). The phosphorylated form of eIF2␣ is known to inhibit the GTP/GDP exchange factor, eIF2B, thereby preventing eIF2 recycling (19). Because eIF2B is usually present at much lower concentrations than eIF2, phosphorylations of 20 -30% of the eIF-2␣ pool cause  Translational Recovery from Inhibition by Vasopressin-Neither the acute inhibition of protein synthesis by hormone nor recovery from this inhibition was altered by actinomycin D at concentrations inhibitory to gene transcription, by rapamycin at concentrations that inhibit the ribosomal p70 S6 kinase (33), or by cAMP-elevating agents and analogs at concentrations that inhibit signal transduction dependent on Ras (34, 35) (data not shown). Earlier findings (see Table II and Figs. 3 and 4) were consistent with roles for protein kinase C activation and Ca 2ϩ in overturning the inhibition of leucine incorporation by vasopressin. The degree to which H9c2 cells replenish their Ca 2ϩ stores during recovery from translational inhibition by vasopressin was therefore examined. Leucine incorporation and cell-associated Ca 2ϩ were measured after varying times of treatment with vasopressin (Fig. 7). Inhibition of incorporation paralleled the decline in cell-associated Ca 2ϩ during the first 30 min of hormonal treatment. At 20 -30 min, 60% of Ca 2ϩ stores were mobilized and protein synthesis was inhibited 50%. Protein synthesis steadily recovered from inhibition thereafter, with full recovery being observed by 100 min. At 2 h of treatment, pulse incorporation rates were 20% higher than those in untreated controls. During the recovery period (30 -100 min) cell-associated Ca 2ϩ also increased but at slower rates than observed for leucine incorporation. Full recovery of pulse incorporation was associated with the restoration of approximately one-fourth of the cation initially mobilized.
A variety of cell types respond to protracted ER Ca 2ϩ mobilization with an ER stress response that includes the induction of GRP78 and the development of translational tolerance to Ca 2ϩ -mobilizing agents. H9c2 cells that had recovered from translational suppression by vasopressin over 2 h and corresponding untreated controls were examined for the development of such tolerance upon challenge with various concentrations of ionomycin, with thapsigargin at 1 M, or with vasopressin at 10 M (Table III). Comparable rates of leucine incorporation were observed for the hormonally pretreated and non-treated cells before challenge. Leucine incorporation in both preparations was reduced in a dose-dependent manner in response to increasing concentrations of ionomycin and was inhibited extensively in response to thapsigargin. A brief challenge with vasopressin, however, suppressed leucine incorporation in the non-treated controls by 68% but had no effect on incorporation in the vasopressin-pretreated preparations. Cellassociated Ca 2ϩ was also determined for non-treated and vasopressin-pretreated preparations before and after challenge with Ca 2ϩ -depleting drugs (Table III). The Ca 2ϩ contents of vasopressin-pretreated cells were 45% of those of non-treated controls. Ionomycin mobilized Ca 2ϩ in a dose-dependent fashion from both the non-treated and the vasopressin-pretreated preparations. The Ca 2ϩ contents of both preparations were reduced to comparably low values by 150 nM ionomycin and by 1 M thapsigargin. After brief challenge with vasopressin, 65% of cell-associated Ca 2ϩ was mobilized from the non-treated controls but no Ca 2ϩ was released from the hormone-pretreated preparations.
To verify that H9c2 cells are capable of expressing the ER stress response, myocytes were incubated for 8 h at varying concentrations of ionomycin or with 0.5 M thapsigargin. Preparations were washed to remove ionophore, and proteins were pulse-labeled with [ 35 S]methionine and sampled for SDS-PAGE (7.5%) and autoradiography (Fig. 8). Preferential labeling of GRP78 and GRP94 was observed in response to treatments with 10 -100 nM ionophore, with optimal labeling occurring at 30 nM (lanes c-f as compared with lane a). A modest increase in labeling of other proteins was also apparent in the treated preparations. Incubation with 5 nM ionomycin did not promote increased expression of the GRPs or stimulation of overall protein synthesis (lane b). The GRPs were also labeled preferentially in thapsigargin-pretreated preparations (lane g). Thapsigargin was not removed by the washing procedure, however, as evidenced by the decrease in overall protein labeling. Additional samples were pretreated for 2 h in the absence and presence of vasopressin before incubations in the absence or presence of ionomycin (30 nM). Preparations were washed and proteins were pulse-labeled and analyzed by SDS-PAGE and autoradiography as above (Fig. 8). Although labeling of most cellular proteins was modestly increased in the vasopressin-treated sample, expression of the GRPs was not selectively increased by the hormone (lane k as compared with lane i). No selective labeling of the GRPs was observed during shorter incubations with vasopressin (data not shown). However, pretreatment with vasopressin followed by incubation with ionomycin resulted in greater pulse-labeling of GRP78 and GRP94 than was observed in samples incubated with ionophore alone (lane l as compared with lane j).
To ascertain whether translational tolerance to inhibition by vasopressin accompanies induction of the GRPs, H9c2 cells were pretreated for 8 h in the absence or presence of ionomycin or thapsigargin (irreversible). Preparations were washed and then challenged with ionomycin at increasing concentrations, with thapsigargin, or with vasopressin. After 30 min of treatments, samples were analyzed by measurement of leucine pulse incorporation (Table IV). Incorporation in the control preparations was inhibited in the typical fashion by ionophore, thapsigargin, and vasopressin. By contrast, protein synthesis in ionomycin-pretreated preparations was markedly resistant to suppression by these agents. No inhibition by ionophore was observed except at the highest (150 nM) concentration tested, and 10 M vasopressin was ineffective. Although thapsigargin suppressed incorporation in ionomycin-pretreated preparations, the extent of this suppression was not as great as in controls. Samples pretreated with thapsigargin had somewhat lower synthetic rates as compared with untreated controls, but protein synthesis was completely unaffected after challenge with ionomycin at all concentration tested, with thapsigargin, or with vasopressin.
Reversibility of Translational Inhibition by Vasopressin-Physiologic responses to vasopressin are known to persist despite receptor internalization, which, depending on cell type, at 37°C occurs within 3-20 min of receptor occupation (36 -38). It was nonetheless unclear whether the translational inhibition imposed by vasopressin in H9c2 cells was reversible or persisted upon removal of the hormone. Cells were therefore treated with vasopressin for 0, 2, 5, 10, 15, or 20 min, and the preparations were washed three times with medium lacking hormone. Washed preparations were then incubated for 30 min in fresh medium lacking or containing vasopressin, followed by measurements of leucine pulse incorporation (Table V). Inhibitions of protein synthesis attributable to vasopressin ranged from 55% at 2 min to 69% at 20 min of treatment. These inhibitions were reversible by removal of hormone at 2 and 5 min of treatment, but thereafter were irreversible.

DISCUSSION
Although Ca 2ϩ sequestration by the ER has been previously established to maintain optimal rates of translational initiation in mammalian cells, the drugs utilized to mobilize the cation invariably led to the development of an ER stress response. This report provides the first compelling evidence that hormonally induced mobilization of S(E)R Ca 2ϩ stores is regulatory upon translation in that it is receptor-mediated, observ-able at physiological extracellular Ca 2ϩ concentrations, and reversible without the generation of a stress response. Translational inhibitions coinciding with mobilization of S(E)R Ca 2ϩ stores of H9c2 ventricular myocytes were imposed by arginine vasopressin at concentrations that raise [Ca 2ϩ ] i in normal cardiomyocytes (39,40). Inhibitions were exerted as a consequence of a hormonal interaction at the V1 vascular receptor (39,40), the receptor expressed in normal cardiac tissue (27). Vasopressin exerted a persistent, albeit decreasing, inhibition of translation for periods of 1-2 h. This result was not surprising in that vasopressin receptors are known to undergo internalization within 3-20 min in various tissues after hormonal binding. The physiologic effects of the hormone persist after internalization (27). The ability of the hormone to signal translational inhibition in H9c2 cells was reversible by washing for only several minutes. Circulating vasopressin has a relatively long half-life (approximates 20 min) and undergoes degradation in the liver and kidney but not in the pulmonary circulation (41).
Inhibition of the translational process in H9c2 cells by treatment with vasopressin involved imposition of the same mechanisms as those exerted by ionomycin, which destroys intracellular Ca 2ϩ gradients, and thapsigargin, which blocks the Ca 2ϩ -ATPase of the S(E)R. Vasopressin, like ionomycin and thapsigargin (1)(2)(3)(4)(5)(6)(7)17), slowed translation at the initiation step and caused eIF2␣ to become phosphorylated. Phosphorylation of this initiation factor correlated closely with hormonal inhibitions of protein synthesis under all experimental conditions tested. As has been observed invariably with agents that inhibit translational initiation by mobilizing ER Ca 2ϩ stores (1,4,5,7,8,17), protein synthesis in cells expressing increased concentrations of ER chaperones was tolerant to inhibition by vasopressin. The mechanism through which vasopressin caused Ca 2ϩ stores of H9c2 cells to be mobilized was not explored. Unlike depletion of Ca 2ϩ stores by thapsigargin or ionophores, Ca 2ϩ depletion in response to vasopressin began to attenuate after 30 min. Attenuation presumably involved either decreased production of, enhanced degradation of, or declining responsiveness to intracellular messenger(s) that mobilize Ca 2ϩ from the S(E)R to the cytosol. Vasopressin acting at V1 vascular receptors of various tissues is established to signal the activation of phospholipase C and the generation of inositol trisphosphate during the first few min of treatment (27). Inositol trisphosphate is widely documented as an intracellular mediator of hormone-dependent Ca 2ϩ mobilization, although cyclic ADP-ribose has also been advanced as a putative physiologic regulator of ryanodine-sensitive Ca 2ϩ -dependent release  processes in intact mammalian systems, including heart (20,42,43). Mediation of the Ca 2ϩ -mobilizing effects of vasopressin by cyclic ADP-ribose in H9c2 cells would be consistent with the actions of caffeine, a pharmacologic activator of the cardiac ryanodine receptor (20), in promoting Ca 2ϩ release, translational suppression, and eIF2␣ phosphorylation. Caffeine affects these parameters in a quantitatively similar manner to vasopressin. Eicosanoids have also been proposed to mediate the Ca 2ϩ -mobilizing actions of vasopressin in smooth muscle cells (27). As in various non-muscle cell types (1,4,5,7,8,17), recovery of H9c2 cells from translational inhibition by Ca 2ϩ -mobilizing drugs was contingent upon induction of expression of ER chaperones. In contrast, recovery from translational inhibition by vasopressin depended on partial re-accumulation of Ca 2ϩ and, most probably, the activation of protein kinase C. Indirect evidence, including abolition of vasopressin inhibition in the presence of a phorbol ester and prolonged inhibition of translation under conditions wherein the enzyme is known to be down-regulated, supports a role for protein kinase C in this event. Both calphostin C and chelerythrine at concentrations used widely to investigate the involvement of protein kinase C in various processes, produced the mobilization of Ca 2ϩ and the phosphorylation of eIF2␣. It was not possible to determine whether the effects of these inhibitors derived from inhibition of protein kinase C activity or from actions at other sites. Any action through protein kinase C would imply that the enzyme is partially coupled to translational rates in H9c2 cells in the absence of hormonal influences. No other requirements for recovery of translational activity could be identified. Other protein kinases that are signaled in response to vasopressin, such as ribosomal S6 kinases (44) and mitogen-activated protein kinase (45,46), could not be implicated. Although activation of V1 vascular receptors is known to signal increased transcription of the immediate early genes (27), transcriptional events were not required for translational recovery.
Recovery from translational inhibition by vasopressin was associated with partial, rather than full, restoration of cellassociated Ca 2ϩ , presumably at the S(E)R. Information is lacking regarding whether H9c2 or other cell types possess critical ER or S(E)R "pools" or subcompartments of Ca 2ϩ supporting translation. Translational inhibition in GH 3 and HepG2 cells occurs at somewhat greater degrees of ER Ca 2ϩ depletion than are required to impede protein processing within the organelle (1,7,26). Inhibition of protein processing invariably appears to trigger the phosphorylation of eIF2␣ in response to Ca 2ϩ depletion. Ca 2ϩ re-accumulation in H9c2 cells after 2 h of vasopressin treatment, therefore, may restore ER function sufficiently that protein synthesis can resume. PMA, presumably by activating protein kinase C, reduced the degree to which vasopressin lowered cell-associated Ca 2ϩ .
Although ionomycin (30 nM) and vasopressin (10 M) generated comparable degrees of cation mobilization and eIF2␣ phosphorylation over 30 min, only ionomycin produced a subsequent ER stress response. The ionophore at 30 nM clearly signaled comparable degrees of GRP78 induction and translational tolerance in H9c2 cells to those observed in other cell types (1,4,5,(7)(8)(9)(10)17). Continued exposures to either ionomycin or thapsigargin elicited strong inductions of GRP78 and GRP94 and development of translational tolerance to ER stressors in H9c2 cells without additional requirements for auxiliary promoters (serum, PMA, growth factors, cAMP analogs) observed in GH 3 , NIH-3T3, and myeloma cells (1,4,5,7,17). Vasopressin treatment of H9c2 cells differs from other Ca 2ϩmobilizing drugs by permitting partial recovery of cell-associated Ca 2ϩ after 30 min. This recovery, which is accompanied by increasing rates of protein synthesis, may sufficiently restore S(E)R function as to ablate the development of an ER stress response.
Vasopressin exerts inotropic actions on cardiovascular performance that was until recently, attributed to hormonal effects at the vasculature and kidney. The hormone is now appreciated to stimulate lipid metabolism in perfused hearts (47) and H9c2 cells (24), to increase [Ca 2ϩ ] i in cardiomyocytes (39), to potentiate ventricular L-type currents via V1 vascular receptor stimulation (40), and to cause atrial natriuretic factor to be secreted from cardiomyocytes (48). The translational suppression observed in H9c2 cells in response to vasopressin provides additional evidence that the heart is directly targeted by this hormone. The functional significance of the short-term suppression of ventricular protein synthesis accompanying mobilization of cell-associated Ca 2ϩ by the hormone remains to be clarified. Both muscle contraction, which is supported by [Ca 2ϩ ] i , and mRNA translation, which is supported by S(E)R sequestered Ca 2ϩ , are energy-intensive processes. Reduced translation accompanying release of S(E)R Ca 2ϩ to the cytosol may therefore function to divert ATP production toward supporting increased contractile activity. In addition to its antidiuretic actions, vasopressin serves broadly as a growth factor and, depending on cell type, can promote either hypertrophy or hyperplasia (27). Induction of specific gene expression and enhanced rates of protein synthesis are both required for growth responses to vasopressin. In support of a growth-promoting effect of vasopressin on H9c2 cells, prolonged exposure to the hormone was invariably accompanied by increased rates of amino acid incorporation into protein. It is anticipated that H9c2 cells should provide a highly useful model system for investigating both the acute and the chronic effects of vasopressin on the biochemistry and physiology of ventricular myocytes.