Main

Myocyte enlargement is the primary mechanism of cardiac growth beyond the neonatal period(1) and is controlled by complex interactions among hemodynamic workload(2, 3), multiple peptide growth factors produced by myocytes and non-myocytes(4), angiotensin II(57)1-adrenoceptor stimulation(8), and sympathetic innervation(911). Signal transduction mechanisms are correspondingly diverse but convergent, resulting in alterations of the expression of protooncogenes, contractile proteins, cell surface receptors, and embryonic markers(12). The investigation of the interactions among these stimuli is especially difficult, and they remain incompletely defined. In vitro models have proved particularly successful at defining these interactions. Sympathetic innervation of neonatal myocytes(9, 10) and fetal hearts(11) promotes cell enlargement but also modulates the growth response to glucocorticoids(13) and expression of angiotensin II type I receptors(14). The mechanism by which sympathetic innervation mediates these responses is unclear.

Using a model system of in vitro innervation of neonatal myocytes, we have shown(9, 10) that the myocyte growth effects of sympathetic innervation are not mediated by adrenergic neurotransmitters and that anatomic contact between the neurons and the myocytes is not required. This indicates that a diffusible substance released from the cells added with the neuronal explants may regulate this effect. Although the neurons are a likely candidate source for this growth-promoting activity, the cocultures of neurons and myocytes contain a heterogeneous population of cells including myocardial fibroblasts and Schwann cells. This cellular heterogeneity prevents identification of the cellular source of the growth-promoting activity. To characterize this activity, we evaluated the growth-promoting activity of a homogenous population of neuronal cells, PC12 cells(15, 16), a cell line derived from rat adrenal medulla pheochromocytoma cells. In the chromaffin state (cPC12 cells), the cells appear round, multiply logarithmically, and synthesize and secrete catecholamines as well as other neuroendocrine markers such as choline acetyltransferase, vasoactive intestinal peptide, and glutamic acid decarboxylase(17). The cells are electrically nonexcitable. Over a 2-wk period after addition of NGF to the cultures, PC12 cells convert to a nonreplicating sympathetic neuronal-like phenotype (nPC12)(15). The cells acquire neuronal characteristics of membrane excitability and long branching neurites. They cease synthesizing tyrosine hydroxylase, thereby losing the capacity to produce catecholamines and become cholinergic(15, 18).

We present results that indicate nPC12 cells share with sympathetic neurons the capacity to release a factor that promotes neonatal myocyte growth, distinct from a neurotransmitter. Myocyte growth is demonstrated by surface area measurements and [35S]methionine incorporation as an indication of protein synthesis. Both sympathetic innervation and nPC12-conditioned medium increase expression of ANP, a marker for cardiac hypertrophy. We demonstrate a dose-response relationship between myocyte size and concentration of nPC12-conditioned medium and that only medium from the nPC12 cells but not cPC12 cells contains this growth-promoting activity. This response is not inhibited by adrenergic or muscarinic blockade and cannot be mimicked by adrenergic or muscarinic agonists or several neuropeptides.

METHODS

Cell culture preparation. Primary myocyte cultures were prepared from neonatal rat 1-2 d old obtained from the WKY normotensive colony at the University of Iowa Cardiovascular Center. The dams and pups were maintained in accordance with the Public Health Service Policy on Humane Care and Use of Laboratory Animals. The experimental protocol was approved by the University of Iowa Animal Care and Use Committee.

Cultures were prepared using previously published protocols with few modifications(9, 19, 20). The ventricular apices were aseptically removed and placed in culture medium, composed of 85% MEM-Earle's salts, 15% horse serum, 4 mM L-glutamine, 20 μg/ml gentamicin, 20 mM HEPES buffer (pH 7.3), and 0.5 mM NaHCO3. The tissue was rinsed for 5 min in potassium glutamate solution containing 140 mM potassium glutamate, 16 mM NaHCO3, 0.5 mM NaH2PO4, 25 mM HEPES (pH 7.3), 16.5 mM dextrose, and 0.014 mM phenol red. Tissue was minced into 1-mm fragments then incubated with 3 mg/ml collagenase II for 20 min at 37 °C. The collagenase was discarded and replaced with 1 mg/ml trypsin in potassium glutamate solution. Cell dispersion continued with 4-6 serial, 15-min incubations with stirring at 37 °C in the trypsin solution. The supernatants were collected in culture medium at 4 °C and centrifuged, and the resulting cell pellet was resuspended in culture medium. Fibroblasts were allowed to sediment for 45 min, and 3-4 ml of supernatant medium (105 cells/ml), now enriched with ventricular myocytes, were plated onto 35-mm tissue culture dishes and incubated in the culture medium described above at 37 °C, 5% CO2, and 95% humidity. Fluorodeoxyuridine 10 μM, and uridine, 10 μM, were added to the culture medium to prevent fibroblast growth. At the time of analysis, non-myocytes comprise <10% of the total cellular population.

The sympathetic neurons were prepared with separate instruments from those used for heart dissection. The paravertebral thoracolumbar sympathetic chains of each pup were removed after the removal of the ventricular apex. The neuronal tissue was maintained in culture medium until dissection was complete. Neuronal chains were finely minced (approximately 0.1 mm in diameter) and spread onto the tissue culture dishes in 0.5 ml of culture medium. The neuronal explants attached to the dish during the time (3 h) required to complete myocardial cell preparation. Half of the ventricular myocytes were plated onto dishes with neuronal explants and the other half onto empty dishes. After 48 h in culture, culture medium in all dishes was changed. Neuronal explants comprise 10-20% of the cell population.

PC12 cell culture and treatment of myocytes with conditioned medium. PC12 cell cultures were maintained as previously described(21) at 37 °C and 6.5% CO2 on collagenized dishes in 85% RPMI 1640 with 5% fetal bovine serum and 10% donor horse serum. Cell density was 107 cells/85-mm dish. The cells were converted to the neuronal phenotype by replating the PC12 cells in RPMI 1640 with 1% donor horse serum and 50 ng/ml NGF. The cells were maintained in this medium for at least 14 d to ensure complete differentiation. More than 95% of the cells respond to NGF, observed as the extension of neurites >2-cell diameters in length. PC12 conditioned medium was prepared from the fully neuronal PC12 cells, except as otherwise noted.

The medium in dishes containing fully differentiated nPC12 cells was replaced with MEM and 1% donor horse serum and 50 ng/ml NGF. The cells were maintained in this for 2 d to allow them to condition the medium. After collection, the conditioned medium was filtered through a 0.4-μm filter to remove cells and debris. L-Glutamine and donor horse serum were added to the nPC12-conditioned medium to make it equivalent to the culture myocyte medium in all respects other than exposure to nPC12 cells. To ensure equivalence, 50 ng/ml NGF was added to the control myocyte medium. The nPC12-conditioned medium was mixed with an equal volume of fresh myocyte medium.

Conditioned medium from control myocyte cultures or innervated myocyte cultures was collected after 4 days in culture, then fultered through a 0.4-μm filter to remove cells and debris. The conditioned medium was mixed with an equal volume of fresh myocyte medium.

The myocyte medium was changed 48 h after the myocytes were plated. At this time, the medium was replaced with control culture medium or conditioned medium, and pharmacologic agents were added to appropriate cultures. Adrenergic neurotransmitter concentrations were added so that the final concentration in the medium was 2 μM(8, 22, 23). Adrenergic antagonists were added in equimolar concentrations. We have previously documented that 2 μM propranolol and prazosin are sufficient to inhibit a change in contraction frequency when adrenergic agonists are applied acutely(19). Unless otherwise stated, cell size and[35S]methionine incorporation assays were performed 48 h later, or 96 h after myocyte plating.

Surface area measurements. Cell height is not altered by innervation(9), so surface area measurements are a valid measure of cell volume(8, 22). Surface area of the myocytes was measured at 96 h in culture using on-line edge detection software with the MEA module of Bioquant IV Image Analysis Program (R & M Biometrics, Inc., Nashville, TN). Myocytes were selected if they were spontaneously contracting, free from contact with other cells, and contained only one nucleus. Innervated myocytes were selected when the attached neuron moved synchronously with myocyte contraction. The video image of a single myocyte was projected onto a monitor, and the perimeter traced in a continuous line on a digitizing tablet. Surface area was automatically calculated as the region within the boundary of the perimeter.

[35S]Methionine incorporation assay. The myocytes were cultured in 12-well dishes and plated at 1 × 105 cells/well. Twenty-four hours before cell harvest, 0.5 μCi of[35S]methionine was added to each well. Twenty-four hours later, the myocytes were washed with PBS, then lysed with 1% Triton X-100 and 0.1% SDS. The lysate was removed from the well and placed in a test tube with 3 ml of 10% trichloroacetic acid. The cell solution was allowed to precipitate on ice for 10 min, and the sample was filtered through a Millipore filter manifold onto a 2.4-cm Whatman GF/A filter. The filter was placed in a scintillation vial and radioactivity counted.

Reverse transcriptase-polymerase chain reaction. Total RNA was isolated using the standard acid guanidinium technique described by Chomczynski and Sacchi(24), with the one-step modification with TRIzol™(25) (Life Technologies, Inc, Gaithersberg, MD). The integrity of the RNA preparations was confirmed by electrophoretic patterns with ethidium bromide staining and quantified spectrophotometrically by absorbance at 260 nm. RNA was stored as an ethanol precipitate at -70 °C until analysis. The cDNA template was prepared by reverse transcription of the RNA isolated from the myocyte cultures with 2μg of total RNA with murine myeloblastoma virus reverse transcriptase. The ANP primer was designed from published sequences with the use of a primer design package, Primer 0.5 (Whitehead Institute, MIT). The sequence of the sense primer was 5′-GCA GCC TAG TCC GAT ATG and the sequence of the antisense primer was 5′-AGG GCT TCT TCC TCT TCC TCC TC. The primers were synthesized by a model 391 DNA synthesizer (Applied Biosystems, Foster City, CA). The amplifications were performed in 100 μL of reaction buffer with 1.5 mM MgCl2, 1 mM dNTP, 2.5 units of Taq polymerase, and 10μl of cDNA. The conditions for the PCR were an initial denaturation step at 94 °C for 2 min, then 27 cycles of 95 °C for 1 min, 57 °C for 1 min, and 72 °C for 1 min, followed by a final extension period of 72°C for 3 min. The samples were separated on an agarose gel with a 123-kb DNA ladder, and the gel was photographed. Densitometry was used to quantify the amplified signals. The number of amplification cycles for ANP was chosen by developing a standard curve for the primer. Fifteen to 35 PCR cycles were performed, and the cycle number was chosen from the linear portion of the curve. The housekeeping gene GAPDH was used as the internal standard. The amount of the amplified product from each gene was normalized to the amount of product amplified with the GAPDH primers. Confirmation of appropriate amplification of the product was done by DNA sequencing in the DNA Core Facility at the University of Iowa on a model 373A automated DNA sequencer(Applied Biosystems).

Statistical analysis. All data were analyzed by analysis of variance with blocking using PC SAS as the statistical software (Version 6, Cary, NC) The blocking variable was culture set. For each analysis,n was the number of cultures in that experiment, although the number of cells examined in each group is also provided on the graphs. Prior comparisons among the experimental groups were performed with Scheffe's test for significant differences. This usually entailed comparing the control values versus the other experimental values(26). A p < 0.05 was considered significant. All data are presented as the mean ±SD.

Reagents. L-Glutamine and horse serum were obtained from JRH Bioscience, Inc. (Lenexa, KS). Trypsin was from Life Technologies, Inc., and collagenase was from Worthington Biochemical Corp. (Freehold, NJ). MEM-Earle's salts and potassium glutamate solution were prepared by the Tissue Culture Hybridoma Facility, University of Iowa Cancer Center. The other reagents were from Sigma Chemical Co. (St. Louis, MO).

RESULTS

Response of myocytes to nPC12 conditioned medium. Several experiments were performed to demonstrate the effect and specificity of nPC12-conditioned medium on myocyte size. Figure 1 illustrates that a 48-h exposure to nPC12-conditioned medium results in both increased myocyte surface area and increased [35S]methionine incorporation into trichloroacetic acid-precipitable material. To demonstrate the similarity of this myocyte growth to that induced by sympathetic innervation, we compared myocyte surface area of cells exposed to nPC12-conditioned medium to that of innervated myocytes and myocytes within the cocultures, but which were distant from the explants and did not have anatomic contact with the neurons. Treatment of the myocytes with nPC12-conditioned medium mimics the growth effects of innervation.

Figure 1
figure 1

Effect of nPC12-conditioned medium on myocyte growth.(A) Surface area of cultured neonatal myocytes exposed to nPC12-conditioned medium for 48 h. Surface area of innervated myocytes and noninnervated myocytes in the culture dish are shown for comparison. n = 6 cultures, number of cells in each group is indicated on the respective column. (B) [35S]Methionine incorporation by myocytes after 48-h exposure to PC12-conditioned medium. nPC12-conditioned medium increased surface area and [35S]methionine incorporation of neonatal myocytes. Abbreviations: CON, ventricular myocytes in control culture medium; INN, innervated myocytes in control culture medium; nPC12, ventricular myocytes exposed to nPC12-conditioned medium for 48 h; N-INN, noninnervated myocytes in control culture medium, distant from the explants in the myocyte cocultures.*p < 0.05 vs CON.

It is possible that these effects were nonspecific changes in response to changes in medium, and not a distinct response to a factor in the medium of cultures with neurons. To test this, we treated myocytes with conditioned medium from control myocytes and innervated myocytes, then compared the myocyte surface area and [35S]methionine incorporation to that of innervated myocytes and those exposed to nPC12-conditioned medium (Fig. 2). The conditioned medium from myocytes alone had no effect on myocyte surface area, indicating that changing the medium was not responsible for the increased myocyte size. Surface area of myocytes treated with conditioned medium from innervated cultures was similar to that of innervated and nPC12-exposed cells. Thus, the presence of myocytes or fibroblasts in these cultures was not sufficient to promote myocyte growth. The addition of the neurons in the conditioned medium, whether they are sympathetic neuronal cells or nPC12 cells, was necessary to induce these changes. The uptake of [35S]methionine by myocytes exposed to conditioned medium from control cells was not different from control myocytes. Uptake by myocytes exposed to conditioned medium from innervated myocyte cultures was greater than that of control myocytes but did not reach statistical significance (p < 0.06). This may be a dose-related effect as the number of neurons in the innervated cultures was far less than the number in the nPC12-conditioned medium cultures.

Figure 2
figure 2

Effect of myocyte conditioned medium and innervated myocyte-conditioned medium on myocyte growth. (A) Surface area of innervated and nPC12-conditioned medium myocytes shown for comparison.(B) [35S]Methionine incorporation by myocytes after 48-h exposure to conditioned medium. Abbreviations: CON-CM, ventricular myocytes exposed to conditioned medium from control myocytes for 48 h;INN-CM, control myocytes exposed to conditioned medium from innervated myocytes for 48 h. n = 4 cultures, number of cells in each group is indicated on each column. *p < 0.05vs CON, +p < 0.06.

To demonstrate a dose-dependent relationship, we varied the density of PC12 cells used to condition the medium and also varied the proportion of nPC12-conditioned medium added to the myocytes (Fig. 3). PC12 cell density ranged from 1 × 106 to 1 × 107 cells/85-mm dish, whereas the proportion of nPC12-conditioned medium varied from 15 to 65%. Growth was measured by [35S]methionine incorporation. Increased [35S]methionine incorporation was detected at 1 × 106 cells/85-mm dish and increased with increasing nPC12 cell density. The myocytes responded in a similar manner to increasing concentration of nPC12-conditioned medium: growth was detected at 15% nPC12-conditioned medium and maximal at 50%. The growth response is dependent on the duration of exposure to nPC12-conditioned medium. The increased incorporation of[35S]methionine is not detected after 24 h in the nPC12-conditioned medium, but seen only after 48 h of exposure to the conditioned medium (Fig. 4). This would indicate that the neuronal effect may be indirect, and/or that gene expression and protein synthesis is required. To investigate this possibility, we attempted to perform cell surface area measurements of myocytes treated with nPC12-conditioned medium and translational or transcriptional inhibitors, but the prolonged drug exposure was too toxic to the myocytes.

Figure 3
figure 3

Dose response to nPC12-conditioned medium measured by[35S]methionine incorporation. (A) PC12 cells were cultured at varying cellular densities. Medium conditioned for 2 d by nPC12 cells was mixed with an equal volume of control medium. [35S]Methionine incorporation was measured after 48-h exposure to the conditioned medium. Labels refer to number of nPC12 cells/85-mm culture dish. Graph is representative of three experiments repeated in quadruplicate. (B) Myocytes were exposed to varying percentages of nPC12-conditioned medium added to control culture medium for 48 h (107 cells/85-mm dish). Percentages refer to the percentage of nPC12-conditioned medium added to the culture medium. The growth response increases with increasing amounts of nPC12-conditioned medium. Graph is representative of four experiments repeated in quadruplicate. *p < 0.05 vs control(CON).

Figure 4
figure 4

Onset of effect of nPC12-conditioned medium.[35S]Methionine incorporation measured after 24- and 48-h exposure to nPC12-conditioned medium. Graph is representative of three experiments repeated in quadruplicate. *p < 0.05 vs control at same time exposure.

To determine whether the factor was truly of neuronal origin, rather than one nonspecifically produced by any cell type, we took advantage of the change in phenotype of the PC12 cells induced by NGF. In the absence of NGF, the PC12 cells have a chromaffin phenotype, express tyrosine hydroxylase, and secrete norepinephrine and epinephrine. During a 2-wk exposure to NGF, the PC12 cells gradually acquire a neuronal phenotype, cease to express tyrosine hydroxylase, and begin to synthesize and release acetylcholine. When myocytes were treated with medium conditioned by the cPC12 cells, minimal myocyte growth was observed (Fig. 5). However, concomitant with neuronal differentiation, there was an increase in the ability of the nPC12 cells to stimulate [35S]methionine incorporation by the myocytes. This indicates that the induction of myocyte growth is associated with only the neuronal phenotype of the PC12 cells, is not a response to the change in the culture medium, and that the factor is not present in nonneuronal cells.

Figure 5
figure 5

Effect of differentiation of PC12 cells on myocyte growth. Myocytes were exposed to PC12-conditioned medium for 48 h. Days refer to number of days the PC12 cells were exposed to NGF. Graph is representative of six cultures repeated in quadruplicate. *p < 0.05vs control.

It is possible that the nPC12 cells depleted the culture medium of unlabeled methionine during the 2-wk differentiation time period. If that were the case, the specific activity of the [35S]methionine added when the conditioned medium is applied to the myocyte cultures would be higher than that of nonconditioned medium. This would result in greater incorporation of radioactive label into the conditioned medium-treated myocytes, than the control myocytes. Data shown in Figure 5 rule out this possibility because myocytes incorporate [35S]methionine equally in fresh medium and medium exposed to nonneuronal PC12 cells.

Up-regulation of ANP by sympathetic innervation and nPC12-conditioned medium. To further document the similar growth response of myocytes to sympathetic innervation and nPC12-conditioned medium we assayed the expression of ANP, a noncontractile gene known to be up-regulated during hypertrophic growth of cardiac myocytes(2729). ANP has low expression in neonatal ventricular myocytes and little to no expression in adult myocytes but can be induced by mechanical loading and diverse hormonal stimuli. When myocytes were innervated or treated with nPC12-conditioned medium, steady state levels of ANP mRNA increased (Fig. 6). Table 1 provides the relative density values.

Figure 6
figure 6

Representative gel shows RT-PCR product of ANP expression in myocytes exposed to nPC12-conditioned medium. Levels of ANP in control and innervated myocytes are shown for comparison. ANP RT-PCR signals were normalized to those of GAPDH to correct for differences in initial amounts of mRNA. Lane 1, 100-bp ladder; lane 2, GAPDH product from control myocytes; lane 3, GAPDH product from nPC12 myocytes; lane 4, GAPDH product from innervated myocytes; lane 5, ANP product from control myocytes; lane 6, ANP product from nPC12 myocytes; lane 7, ANP product from innervated myocytes. The PCR products were separated on 2% agarose gel and were then stained with ethidium bromide, photographed under UV light.

Table 1 Relative density of ANP in myocyte cultures

Lack of neurotransmitter response. Both α- andβ-adrenoceptor stimulation of myocytes have been shown to stimulate protein synthesis in other myocyte culture models(8, 30, 31). We have previously demonstrated that α- and β-adrenergic antagonists do not block the effect of innervation, and α- and β-adrenergic agonists do not reproduce the effects of innervation(9, 10). Therefore, we compared nPC12-conditioned medium to adrenoceptor stimulation. Myocytes were exposed to α-adrenoceptor stimulation and blockade with and without nPC12-conditioned medium (Fig. 7,A and B). To ensure pureα-adrenoceptor stimulation, propranolol was added simultaneously with phenylephrine to some of the cultures. Surface area measurements demonstrated no effect of phenylephrine stimulation on control cells (Fig. 7A) Cells exposed to nPC12-conditioned media and phenylephrine and propranolol, singly or in combination, show a blunted response to nPC12-conditioned media but still greater than control myocytes, indicating possibly a nonspecific effect to the drugs. The response of the cells as measured by [35S]methionine incorporation demonstrates more convincingly the lack of neurotransmitter response. [35S]Methionine incorporation in control culture medium with phenylephrine or phenylephrine and propranolol was no different from control myocytes (Fig. 7B). nPC12-conditioned medium provoked a marked increase in[35S]methionine incorporation by the myocytes, but phenylephrine did not increase this response, nor did α-adrenergic blockade prevent it (Fig. 7D).

Figure 7
figure 7

Effect of adrenergic stimulation and antagonism on myocyte growth in control culture medium and nPC12-conditioned medium as measured by surface area measurements and [35S]methionine incorporation. Adrenergic agonists, antagonists, and nPC12-conditioned medium were added at 48 h in culture, and measurements were performed 48 h later.(A) Surface area of myocytes exposed to α agonists.(B) [35S]Methionine incorporation by myocytes exposed toα agonists. (C) Surface area of myocytes exposed to β agonists. (D) [35S]Methionine incorporation by myocytes exposed to β agonists. Abbreviations: PH, 2 μM phenylephrine; PRO, 2 μM propranolol; ISO, 2 μM isoproterenol; PRA, 2 μM prazosin. For [35S]methionine experiments, graphs are representative of three experiments, repeated in quadruplicate. For surface area experiments, n = 5 cultures and number of cells in each group is indicated on respective columns.*p < 0.05 vs control (CON)†p = 0.05.

Similar experiments were performed with isoproterenol and prazosin (to ensure complete agonist stimulation) (Fig. 7,C and D).β-Adrenoceptor stimulation had minimal effect on myocyte surface area(p = 0.05) (Fig. 7C) and no effect on[35S]methionine incorporation in control culture medium (Fig. 7D) The surface area response to nPC12 was again blunted by isoproterenol and prazosin, but the incorporation of[35S]methionine by cells in nPC12-conditioned media was not enhanced by agonist stimulation nor inhibited by the antagonists.

Taken together, these data indicate that neither α- norβ-antagonism inhibited the increased [35S]methionine incorporation or increased surface area induced by nPC12-conditioned medium. Under these culture conditions, α- and β-adrenoceptor stimulation did not induce myocyte growth, and adrenergic blockade did not prevent the effects of nPC12-conditioned medium. The nPC12 cells appear to secrete a soluble, non-catecholamine factor that promotes myocyte growth.

Because muscarinic activation has been described to induce cardiac hypertrophy(32) and nPC12 cells do synthesize and release acetylcholine, myocytes were incubated with nPC12-conditioned medium and atropine (Fig. 8). As with adrenergic blockade, muscarinic blockade did not inhibit the effect of PC12-conditioned medium.

Figure 8
figure 8

Effect of muscarinic agonism and antagonism on myocyte growth exposed to nPC12-conditioned medium. Oxotrimorine and atropine were added at 48 h in culture and experiments performed 48 h later. Abbreviations:OTM, oxotrimorine; ATR, atropine. Graph is representative of three experiments, repeated in quadruplicate. *p < 0.05vs control (CON).

Neuropeptides. Neuropeptide Y and somatostain are neuropeptides known to be present within sympathetic neurons and/or nPC12 cells and appear to mediate some of the effects of α1-adrenoceptor stimulation(3336). We assessed the effects of these peptides to determine whether one or both might be the neuronal factor producing the myocyte growth. Neither agent had any effect on myocyte protein synthesis (Fig. 9).

Figure 9
figure 9

Effect of neuropeptide Y and somatostatin on[35S]methionine incorporation. Neuropeptides, 1 or 10 μM(HI) were added after 48 h in culture, and experiments were done 48 h later. Abbreviations: NPY, neuropeptide Y; SOM, somatostatin. *p <0.05 vs control(CON).

DISCUSSION

We applied two measures of cell size, surface area and new protein synthesis, to compare the effect of sympathetic innervation and nPC12-conditioned medium on neonatal myocyte growth. nPC12-conditioned medium mimicked myocyte growth induction by sympathetic innervation as judged by both methods. The experiments reported here support the following conclusions:1) nPC12-conditioned medium contains a factor that increases myocyte surface area and protein synthesis; 2) the growth-promoting activity up-regulates ANP expression, similar to other models of cardiac hypertrophy;3) the capacity to synthesize and/or release this factor is acquired with neuronal differentiation of the PC12 cells; and 4) this growth-promoting activity is not an adrenergic or cholinergic neurotransmitter, nor neuropeptide Y or somatostatin.

Sympathetic influence. The role of adrenergic stimulation in stimulating cell growth has been of considerable interest over the past decade. Simpson et al.(22) in 1982 first reported that serum-free cultured neonatal myocytes have increased surface area and protein content when incubated with norepinephrine over a 10-14-d period. This response proved to be an α1-mediated response,(8, 23) and initiates increased transcription of specific contractile proteins: c-myc, skeletal α-actin, andβ-myosin heavy chain(27, 37, 38). This pattern of gene expression is similar to other in vivo models of hypertrophy(2,39), such as hemodynamic load, but not that induced by thyroid hormone stimulation(39, 40). This model has been used extensively to provide a model of neural mediation of cardiac growth.

Neurotransmitter response alone, however, does not necessarily define sympathetic influence, because sympathetic neurons secrete many neurotransmitters that could affect myocyte growth and function. We demonstrated previously that the increase in surface area of myocytes innervated in vitro cannot be inhibited with adrenergic blockade(10). We have now extended those studies to demonstrate that this increase cannot be reproduced in this culture system by adrenergic stimulation and that protein synthesis observed with nPC12-conditioned medium can also not be suppressed by adrenergic blockade or reproduced with adrenergic stimulation. An in oculo model of beating embryonic heart development(11, 41) has demonstrated that atrial or ventricular tissue transplanted into an innervated anterior eye chamber has greater mass than tissue transplanted into a denervated chamber. In this model, α-adrenoceptor antagonism also does not inhibit the effect of innervation (D. Tucker, personal communication). Thus, in both the in vitro and in oculo models of sympathetic innervation-induced myocyte growth, an effector other than adrenergic neurotransmitters is implicated. Further support for a neuronal influence, distinct from neurotransmitter effect, is supplied by in oculo experiments wherein parasympathetic cholinergic innervation is removed. Hearts receiving parasympathetic innervation are larger than denervated hearts, and the effect is not inhibited by atropine(32).

Thus based on the data of agonist exposure and in vitro innervation, it appears that sympathetic innervation influences myocyte growth by two means. First, there is the nonadrenergic stimulus described here. This causes a robust increase in growth that occurs rapidly, detectable within 2 d after exposure of the myocytes to the factor. Second, there is theα-adrenergic-mediated growth stimulus that appears to exhibit a longer latency(22). We did not detect a growth stimulus byα-adrenergic agonists in our system, but there are several reasons why we might not have detected this. First, we assayed cell size and[35S]methionine incorporation after 24, 48, and 72 h (not shown) of innervation, nPC12-conditioned medium, or agonist treatment. The size increase described by Simpson et al.(22) as a consequence of α-adrenergic stimulation required greater than 96 h of treatment with an α-adrenergic agonist. Second, the myocytes in our culture conditions are beating and are thus under mechanical tension. This alone provides a growth stimulus(4245) that might mask anα-adrenergic response. Third, the inclusion of serum, which by itself promotes growth, may also mask the effect of α-adrenergic stimulation(22). Our studies demonstrate a nonadrenergic neuronal growth factor that induces a robust growth response in the presence of serum, indicating that serum and the neuronal factor use distinct signaling pathways to induce growth, thereby permitting additive responses.

Culture conditions. Hemodynamic load is a critical regulatory factor in myocyte size, but its precise role has been difficult to discern(46). Model systems in which myocytes retain the capacity to contract spontaneously or are mechanically stretched(7, 47) are larger and have greater rates of protein synthesis(30, 44, 45). Cells that are prevented from beating have higher rates of actin and myosin heavy chain degradation(48, 49), altered gene expression ofβ-myosin heavy chain(50), and disrupted assembly of contractile protein apparatus(51). Additionally, the response of beating cells to known stimuli of myocyte growth, i.e. norepinephrine or phorbol 12-myristate 13-acetate, is smaller than the response of quiescent cells(45). As a result, the analysis of myocyte size is critically dependent upon the culture conditions. Many model systems utilize serum-free medium that disrupts the contractile protein assembly(27). Cells do not spontaneously beat and do not experience mechanical tension. Although the presence of serum in the culture medium adds many peptides and hormones, which could potentially obscure the results, we have elected to use serum in the cultures. First, this provides a better reflection of cells in vivo where cells are continuously exposed to the peptides and hormones in serum. Second, because myocytes contract in vivo, a model system that permits beating affords a better model to assess regulators of growth and function. Third, despite the presence of serum, we have still been able to document significant increases in cell size and protein synthesis, demonstrating that serum alone does not induce a maximal growth response in the myocytes.

Direct versus indirect action of the factor. The neuronal factor may exert a direct effect on the myocytes to promote growth, or may exert an indirect effect by inducing the myocytes or non-myocytes to produce the direct stimulus. Two examples of such growth stimulants include myotrophin(52, 53), released by myocytes, and an as yet unidentified factor produced by non-myocytes(54). Alternatively the primary action of the factor produced by sympathetic neurons may be to increase contractile function of the myocytes, and this in turn leads to growth stimulation. The increase in contractile function could cause the release of an autocrine factor as has been demonstrated by Sadoshima et al.(7) with respect to angiotensin II.

Sen et al.(52) isolated and sequenced myotrophin, a 12-kD protein from hearts of spontaneously hypertensive rats that stimulates protein synthesis and the organization of myofibrils in vitro. This protein has also been isolated and cloned from normal and cardiomyopathic human hearts(53). It is more abundant in myopathic hearts than normal hearts. However, no connection to neuronal stimulation has been developed. We did not observe any increase in myofibril organization by sympathetic innervation(9), a major response of the myocytes to this protein.

Long et al.(54) reported that medium-conditioned cultures enriched with cardiac non-myocytes (30%) promote protein synthesis and increased surface area of myocytes in a serum-free system. The factor is a heat-labile peptide of molecular mass 45-50 kD, and based on heparin-Sepharose binding profiles, does not appear to be any of the known fibroblast-derived growth factors. These effects required concentration of the non-myocyte cell cultures to much higher percentages than present in our cultures. However, it is possible that the factor from sympathetic neurons enhances the non-myocyte production of this factor, which in turn promotes growth of the myocytes.

Sadoshima et al.(7) have demonstrated that mechanically stretched myocytes release angiotensin II into the medium. This conditioned medium increases protein synthesis in non-stretched myocytes. All the components, synthetic enzymes, and receptors, of the renin-angiotensin system, exist in the heart(55, 56), and the hormone is a direct cardiac growth factor(56). Angiotensin II is a potent stimulus to myocyte protein synthesis, but not DNA synthesis(57). This response is mediated via the AT1 receptor and causes rapid induction of several immediate-early genes, as well as several of the “hypertrophic” contractile proteins, i.e. skeletal α-actin and atrial natriuretic peptide. Because sympathetic innervation(57) and nPC12-conditioned media (S. H. Green P. A. Krumm, J. R. Winters, and D. L. Atkins, manuscript in preparation) increase myocyte contraction amplitude and velocity of contraction, we question whether the increased contractile function is the primary stimulus and growth is a consequence of increased contractile function.

Similarity of sympathetic innervation and nPC12-conditioned medium. Until it is fully characterized and identified, we cannot be sure that the factor produced by sympathetic neurons is identical to that produced by nPC12 cells. However, there are marked similarities in the myocyte responses induced by these two interventions. The most striking feature of the growth response is the lack of effect of neurotransmitters, and the inability to block the response with adrenergic or cholinergic antagonists in both systems. Both sympathetic innervation and exposure to nPC12-conditioned medium increase expression of ANP. The magnitude of the growth response is similar between the two stimuli. Additionally, sympathetic innervation and nPC12 cells alter contractile function of the myocytes in a similar manner. Sympathetic innervation decreases beating frequency while increasing contraction amplitude and velocity of cultured myocytes, and we have shown that nPC12-conditioned medium promotes comparable changes in the myocytes (S. H. Green et al., manuscript in preparation)(57, 58). The final indication that the factors may be similar is the observation that only the nPC12 cell-conditioned medium, but not the cPC12-conditioned medium, contains the growth-promoting activity. This indicates that the synthesis and/or release of the factor is dependent upon the sympathetic neuronal phenotype of the PC12 cells. The ability of the nPC12 cell-conditioned medium to mimic the innervation effect shows that the factor is diffusible and will provide a source for its purification.

Summary. We demonstrate the existence of a soluble neuronal factor that stimulates cardiac myocyte growth in vitro. Our data indicate that this factor is not one of several molecules known to be released by sympathetic neurons, including catecholamines, acetylcholine, neuropeptide Y, or somatostatin. The observation that a neuronal cell line is a good source of a growth-promoting factor will facilitate its identification. Further experiments to identify this factor will provide increased understanding of the mechanisms by which neurons regulate normal myocyte growth as well as mechanisms that produce pathologic hypertrophy.