INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CARDIAC WORK is an essential mediator of heart size and myocyte growth. However, the precise role of mechanical force, independent of calcium-related signaling, has yet to be determined. The aim of the present study was to discover to what extent mechanical activity vs. calcium regulates translation and localization of -myosin heavy chain (-MyHC) mRNA in myocytes from neonatal rats. For this purpose, 2,3-butanedione monoxime (BDM), a well-known inhibitor of skeletal and cardiac muscle contraction, was used (3, 10). Increasing BDM concentrations up to 18 mM has been shown to progressively decrease isometric tension and stiffness of rabbit psoas muscle fibers because affinity of nucleotide binding to the myosin head was increased and more cross bridges were in the detached state (29). In cardiac muscles, it was reported that at concentrations lower than 10 mM, BDM blocks actin-myosin cross-bridge cycling and force development without any significant effects on calcium transients (1). Also, exposure of spontaneously contracting neonatal rat cardiocytes to 7.5 mM BDM produced rapid and near-complete inhibition of contractile activity without a significant effect on either basal or peak intracellular calcium concentration ([Ca²⁺]_i) transient amplitude (3).

The total RNA and protein levels are higher in beating than nonbeating neonatal myocytes in culture (19, 26). Also, RNA and protein contents can increase in myocytes that are either stimulated to beat by electrical pacing (14) or as a result of increased hemodynamic load (12, 13, 20). In contrast, reduction in cardiac load leads to atrophy accompanied by decreases in transcription of most muscle mRNAs and the rate of protein synthesis (3, 17, 26). Surprisingly, contractile arrest increases the level of -MyHC mRNA, although its protein content is decreased (22). The increased -MyHC mRNA abundance could be explained by an increase in transcription or by a change in the stability of the message. Our previously published data (8) showed that translational control is involved in the regulation of -MyHC gene expression when beating was arrested by the calcium channel blocker verapamil. The stabilization of -MyHC mRNA in verapamil-treated myocytes was due to translational block at a postinitiation stage.

Localization of mRNA is dependent on functional changes in different cell types. For example, within minutes the addition of serum to starved fibroblasts causes an accumulation of both the -actin mRNA and its translated reporter protein in the lamellipodia of the leading edge of the cell (15, 16, 18). The rapidity of this event suggests that an existing zip code RNA binding protein modulates new synthesis of actin filaments (24). Also, mechanical tension and integrin signaling may play a role in mRNA and ribosome redistribution in fibroblasts and endothelial cells (4). Our group has shown that -MyHC mRNA is distributed throughout the cytoplasm in spontaneously beating cardiac myocytes, whereas in verapamil-arrested myocytes the mRNA is localized perinuclearly (7). We have recently shown that the microtubules are necessary to carry the mRNA outward from the nucleus to the periphery of the cardiac myocytes and that this distribution is dependent on the translational state of the mRNA (21).

In normal excitation-contraction coupling, the transient increase in intracellular calcium triggers cross-bridge activation. However, by arrest of beating in a more specific way, we can allocate the signal between the mechanical and calcium transduction pathways. In the present study, we used BDM to inhibit cross-bridge cycling in a concentration that leaves the calcium transient unaffected. This is a better way to allocate the signal components than in our earlier studies using verapamil (8), which blocks both calcium cycling and subsequent contractility. First, polysome distribution profiles and -MyHC mRNA levels were analyzed from fractions of a sucrose gradient. Our data showed that BDM-induced arrest leads to an increase in free -MyHC mRNA levels with no significant change in the content of the heavier polysomes that are translating -MyHC mRNA. Next, the localization of -MyHC mRNA was assessed by in situ hybridization and local concentration recorded in optical density (OD) units as a function of the distance from the nucleus. The -MyHC mRNA was dispersed throughout the cytoplasm in spontaneously beating cells. When beating was arrested with BDM, a similar cytoplasmic pattern was detected (no significant difference). Taken together, these data indicate that mechanical signals regulate initiation of -MyHC mRNA translation but do not appear to be important for its localization.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Neonatal rat primary cardiac myocyte culture. Cardiac myocytes were dissociated from 1- to 3-day-old neonatal Sprague-Dawley rat hearts. The cells were cultured as described before (6) and plated at high density (1,000-2,000 cells/mm²). Cells were grown on either collagen-coated (Vitrogen 100, Celtrix, Santa Clara, CA) round glass coverslips for in situ hybridization or on 10-cm dishes for ribosome distribution and slot-blot analysis. The culture medium used was a 1:1 mixture of DMEM and Ham's F-12 nutrient mixture, without L-glutamine and supplemented with penicillin, streptomycin, fungizone, and 5% fetal bovine serum. The cells were kept in medium with serum for 48 h before being changed to medium lacking serum. Thyroxine (0.5 nmol/l) was used to maintain a pure -MyHC expressing culture. The addition of cytosine -D-arabinofuranoside (5 mg/ml) and the omission of L-glutamine from the medium were used to control the division of the fibroblast population. All chemicals and culture materials were purchased from Sigma Chemical (St. Louis, MO) unless otherwise stated.

Pharmacological treatments. The cardiac myocyte cultures were treated by the addition of pharmacological agents to the extracellular medium starting 24 h after switching to serum-free medium to rule out any effect from serum factors. Cells were treated with BDM (7.5 mM, 6 h) to inhibit spontaneous contractile activity or treated with cycloheximide (10 mM, 2 h) to inhibit protein synthesis without visibly affecting contractility (7). Alongside the experimental cultures, control cultures were run at each time point with the addition of fresh medium.

Polysome distribution profile and slot-blotting analysis. Cardiocyte cultures in 10-cm dishes were rinsed twice with cold PBS. Cells were then scraped into homogenization buffer: 50 mM Tris, 250 mM KCl, 25 mM MgCl₂, 2% Triton X-100, 200 mM sucrose, 1 mM EGTA, pH 7.4; supplemented with 0.25 mM dithiothreitol, 0.5% Nonidet P-40, 0.5% sodium deoxycholate, 200 mg/ml of heparin sodium, 1 mg/ml of cycloheximide, and 4 U/ml of ribonuclease inhibitor. Homogenates were centrifuged first at 4,000 rpm for 5 min and then at 11,000 rpm for 15 min to pellet cell membrane, insoluble proteins, nuclei, and mitochondria. Postmitochondrial supernatants were layered out on linear sucrose gradients (0.4-1.2 M) and centrifuged for 90 min at 40,000 rpm in a SW41 rotor. The gradients were fractionated into 0.5-ml fractions, starting with the top of the gradient. The total RNA content was monitored spectrophotometrically at 254 nm. To determine the amount of total RNA in the free, nonpolysomal fraction relative to the amount in translating, polysomal fraction, the gradients were divided into two components: nonpolysomal (lighter) and polysomal (heavier). Then we integrated the area under the curves to obtain total amounts of RNA for lighter and heavier fractions. These components of RNA from BDM- and cycloheximide-treated cells were expressed as the percentage of untreated control.

1

1

Isolation of RNA. Total RNA was extracted from 1 × 10⁶ cells with the commercially available RNeasy kit (Qiagen, Santa Clarita, CA). The concentration was determined photometrically at 260 nm.

Cytochemistry. The cells were first washed in cold PBS, pH 7.4, and then fixed in 4% paraformaldehyde-PBS-MgCl₂ for 10 min. Cells were stored in fresh 70% ethanol at 20°C until use. The myofibrils were stained with rhodamine-conjugated phalloidin (3.3 mM; Molecular Probes, Eugene, OR) in PBS for 45 min. The coverslips were then mounted in PBS-glycerol (1:3) containing 0.1% phenylenediamine to prevent fading of fluorescence. Myocytes were viewed through a dual band-pass barrier filter on a Nikon Microphot FXA epifluorescent microscope, and fields were selected in a nonbiased manner for evaluation of myofibrillar architecture.

In situ hybridization for -MyHC. In vitro transcription of the -MyHC probe was achieved by first linearizing vector pH 110 (a kind gift from Dr. L. A. Leinwand) with BamH I (antisense) and Hind III (sense), and transcribing with T₇ (antisense) and T₃ (sense) polymerases. The probes were labeled with digoxigenin-UTP during the transcription reaction, and the transcription was performed according to the manufacturer's instructions (Boehringer-Mannheim Biochemicals, Indianapolis, IN). After the cells were treated with BDM for 6 h, they were fixed and stored as previously described. The fixed cells were prehybridized in 50% deionized formamide-4× SSC at 65°C for 10 min. The coverslips were then hybridized in the denatured probe mixture overnight at 42°C in a humidified hybridization oven. The probe mixture for each coverslip contained 20 ng of probe, 6 mg of salmon sperm DNA, and 3 mg tRNA. The mixture was lyophilized, resuspended in 10 ml of deionized formamide, and then boiled for 10 min to denature the probe. To this an equal volume of hybridization solution (4× SSC, 20% dextran sulfate, 0.1 M dithiothreitol) was added. After the hybridization, the coverslips were incubated for 30 min in RNase A and RNase T1 mixture at 37°C to remove nonhybridized RNA. The RNase treatment and a series of washing steps with SSC were followed by enzymatic detection method using digoxigenin-UTP label. The coverslips were incubated with digoxygenin antibody solution conjugated to alkaline phosphatase at room temperature for 30 min. Finally, the colorimetric enzymatic step with 4-nitro blue tetrazolium chloride and X-phosphate/5-bromo-4-cloro-3-indolylphosphate was carried out for ~1 h. All chemicals for in situ hybridization were purchased from Boehringer-Mannheim Biochemicals (Indianapolis, IN).

Image analysis. The cells hybridized in situ were viewed using the Nikon Microphot-FXA light microscope. The images were digitized with ImagePoint video image analysis system (Photometric Image Point cooled charged couple devise video camera, Photometrics, Tucson, AZ) for a quantitative analysis of OD as a function of the distance from the nucleus (mm). Image analysis was performed using Image Pro Plus software (Media Cybernetics, Silver Spring, MD). To ensure a nonbiased sample, a cell at the center of every other high magnification field (×60 objective) was analyzed yielding at least 30 cells per experiment. A profile of the OD was collected for each cell at 2.5-mm intervals outward from the edge of the nucleus. For each experiment the average OD (n = 30) was plotted against the distance from the edge of the nucleus, which was defined as 100%.

Statistical analysis. Data are given as means ± SE, with five separate cultures used for each experimental condition. Statistical probability was assessed by Kruskal-Wallis nonparametric test followed by Mann-Whitney U-test for total RNA levels, 18S rRNA, and -MyHC mRNA content. ANOVA followed by Student's t-test was used to test the differences in the slopes between control and the treatment groups. P < 0.05 was considered to indicate a statistically significant difference between groups.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Increased level of total RNA in nonpolysomal pool after BDM arrest. We studied the regulation of mRNA translation in response to decreased cardiac load in BDM-treated cells. The spectrophotometric profiles in Fig. 1 show that the RNA is distributed between the light, nonpolysomal and heavier, polysomal RNA pools. The lighter fractions contain free mRNA, 40S and 60S subunits, and 80S ribosomes. We first analyzed extracts from control, spontaneously beating cells (Fig. 1A). This tracing demonstrates the resolution of the sucrose gradient procedure and shows the RNA content in both the nonpolysomal and polysomal fractions from beating myocytes. The arrest of contraction by BDM shifted the total RNA from the heavy pool into the light, nonpolysomal pool (Fig. 1B). A significantly smaller amount of RNA is found in the polysomal fraction of BDM-arrested myocytes.

View larger version (39K): [in this window] [in a new window]   Fig. 1.   Spectrophotometric profiles of polysome gradients from cultured neonatal rat myocytes. Migration of free mRNA, 40S, 60S, and 80S ribosomes, and polysomal regions are indicated. Beginning of polysomal fraction (heavier) is indicated by arrow. Representative tracings from control, 2,3-butanedione monoxime (BDM)- and cycloheximide (CHX)-treated cells are shown. A: control, spontaneously contracting myocytes. B: myocytes arrested by BDM (7.5 mM) for 6 h. C: myocytes treated with cycloheximide (10 mM) for 2 h. D: analysis of total RNA in polysome gradients from BDM- and CHX-treated cells. Quantitative analysis of total RNA in nonpolysomal and polysomal fractions isolated from cells treated for 6 h with 7.5 mM BDM and 2 h with 10 mM CHX. Untreated myocytes were used as control (CON). Data are means ± SE (n = 5). * Significant difference from control (P < 0.05).

BDM arrest stabilizes -MyHC mRNA in nonpolysomal pool. Because contractile arrest of cardiac myocytes stabilizes the half-life of -MyHC mRNA and also results in a reduction of -MyHC protein synthesis (8, 22), we wanted to analyze regulation of translation in BDM-treated cells further. We used the RNA slot-blot method to detect -MyHC mRNA in control and BDM-treated cells (Fig. 2A). The same blot was then reprobed with the 18S rRNA probe (Fig. 2B). The amount of -MyHC mRNA from BDM-arrested cells in nonpolysomal and polysomal fractions was quantified for five experiments and standardized to control cells (Fig. 2C). Summary data show that there is no difference in -MyHC mRNA content in polysomal fractions between arrested and control cells (108.1 ± 8.21%, P = 0.69). However, a significant increase in the -MyHC mRNA level is detected in the nonpolysomal fraction when beating is blocked (120.5 ± 7.7%, P < 0.05 vs. beating cells). Note that -MyHC mRNA is more abundant in BDM-arrested cells.

View larger version (72K): [in this window] [in a new window]   Fig. 2.   A and B: representative slot-blot images for -myosin heavy chain (-MyHC) mRNA and 18S rRNA, respectively, from control (CON) and BDM-treated cells. On the basis of spectrophotometric profile, the beginning of polysomal fraction (heavier) is determined for every blot and indicated by arrow. C and D: -MyHC mRNA and 18S rRNA content in polysome gradients after BDM treatment (6 h with 7.5 mM BDM). Quantification of -MyHC mRNA and 18S rRNA was performed for nonpolysomal and polysomal fractions using RNA slot-blotting analysis. Data are means ± SE (n = 5). * Significant difference from control (P < 0.05).

Myofibril cytoarchitecture. Cells were labeled for actin with rhodamine to observe myofibrillar architecture. Contracting myocytes have well-defined, striated myofibrils in culture. However, when the cells are inhibited from contracting with BDM for 6 h, the myofibrils begin to disassemble (data not shown) and eventually lose all myofibrillar organization after 48 h, as previously seen by Byron et al. (3).

Determination of -MyHC mRNA localization in BDM-treated cardiac myocytes. In situ hybridization was used to determine the distribution of -MyHC mRNA in cardiac myocytes after BDM treatment for 6 h (Fig. 3). The localization of the -MyHC mRNA was detected by the distribution of the dark stain produced during the final colorimetric enzymatic step after the hybridization to antisense RNA probe that binds to the endogenous -MyHC mRNA within the cell. Hybridization of the myocytes with the control, sense probe did not show any detectable signal within the cells under any treatment as reported earlier (7).

View larger version (116K): [in this window] [in a new window]   Fig. 3.   In situ hybridization showing localization of -MyHC mRNA in cardiac myocytes. A: spontaneously beating myocytes. B: cells arrested with BDM for 6 h. Darkness of staining within cells indicates level of mRNA abundance. Arrows are 10 mm from edge of nucleus.

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Optical density (OD) profiles showing -MyHC mRNA abundance as function of distance from nucleus. -MyHC mRNA level at each location was estimated by OD after in situ hybridization and expressed as %OD at edge of nucleus. , Spontaneously beating cardiac myocytes (CON); , myocytes arrested for 6 h with BDM. Data are means ± SE (n = 5). There were no statistical differences at any given distances between control and BDM-arrested cells (P > 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Posttranscriptional regulation of cardiac contractile mRNA expression is an important determinant of cardiac myocyte plasticity. However, the mechanism by which mechanical activity of cardiac cells modulates the translation and localization of the mRNA is not yet known. In this study we used BDM to block beating of rat myocytes to determine the contribution of the mechanical signaling pathway to these processes, independent of calcium changes. We have concentrated on the -MyHC isoform because its translation is blocked during contractile arrest (8, 22, 23), whereas other mRNAs, such as -myosin, -cardiac actin, and glyceraldehyde-3-phosphate dehydrogenase are not (26).

Verapamil and BDM are widely used inhibitors of muscle contraction, and both eliminate cross-bridge cycling. Verapamil does this by blocking the entry of calcium into the cell through L-type calcium channels therefore blocking the subsequent steps in excitation-contraction coupling. In contrast, at low doses, BDM influences the contractile machinery directly (29) but leaves calcium unchanged (3). Our published data showed that verapamil arrest resulted in increased abundance of 18S rRNA and -MyHC mRNA in the heavier polysome complex (8). This finding is similar to that seen for cycloheximide, which indicates a postinitiation block of translation for both total RNA and -MyHC mRNA after the removal of calcium influx. BDM arrest, however, was not studied previously so the locus of action has not been assigned directly to a mechanical event, a calcium event, or to both. Surprisingly, we now find that verapamil and BDM have very different effects on translation. Now we report that cross-bridge inhibition with BDM caused a redistribution of total RNA and 18S rRNA from polysomal to nonpolysomal pool, which is opposite to verapamil-induced shift to the heavier polysomal fraction. Interestingly, we found that the -MyHC mRNA does not completely follow the general behavior of 18S rRNA and, presumably, the majority of mRNAs in BDM-arrested cells. Specifically, the increased amount of -MyHC message in the nonpolysomal pool is detected, but the polysomal pool did not decline. Thus we confirm that -MyHC mRNA is more abundant in BDM-arrested cells because of stabilization (22, 26). Apparently, BDM not only stabilizes -MyHC mRNA but also affects other translational steps. In contrast, verapamil predominantly affects these latter steps, whereas -MyHC mRNA block of initiation is not detectable (8).

There is general agreement that BDM under 10 mM interferes with actin-myosin cross-bridge formation and inhibits force generation with negligible effects on the [Ca²⁺]_i transient in rat myocytes in culture (3). Therefore, at 7.5 mM BDM, the concentration we have used in this study, [Ca²⁺]_i transients were unaffected, whereas the contractile activity was greatly depressed. This literature needs to be carefully evaluated because the effective dose varies with species and age of myocytes. For example, an even lower BDM concentration (3 mM) affects calcium transients in isolated adult myocytes from guinea pig (10). Another potential problem to be considered is that oximes such as BDM are chemical phosphatases (9). For example, it has been shown that 100 mM BDM decreased the phosphorylation state of the inhibitory subunit of troponin and phospholamban in guinea pig ventricular myocytes, which could in part be due to activation of type 1 or 2A phosphatase activity. This suggested that at very high concentration BDM affects the phosphorylation state of the cardiac regulatory proteins via activation of their phosphatases (30).

Phosphorylation regulates several proteins in the translational pathway (2, 27), and it is necessary to consider whether BDM may affect them directly. In adult feline cardiocytes, increased cardiac load is coupled to accelerated rates of protein synthesis (13). Phosphorylation of translational initiation factor eIF-4E appears to be a mechanism for regulation of protein synthesis (27). Inhibition of that phosphorylation was detected in the presence of 7.5 mM BDM, but BDM did not affect eIF-4E phosphorylation in response to either insulin or phorbol ester treatment. This shows that BDM inhibition of eIF-4E phosphorylation is caused by blockade of active tension development and not because of the intrinsic phosphatase activity of BDM. We confirm this general control of initiation via mechanical signals. Namely, our present data (Fig. 1) showed a shift of total RNA from the polysomal into the nonpolysomal pool after BDM arrest. This indicates that mechanical signals are necessary for the initiation step of translation regardless of the change in calcium transients.

The distribution of myosin mRNA has been studied by us in cardiac muscle cells in tissue and in culture (5, 25, 28). In vivo, the mRNA is found radiating outward from the nucleus in spoke-like arrays between the dense contractile fibrillar masses. These arrays follow the microtubular pattern, and we have examined the relationship between mRNA and microtubules in vitro (21). In vivo studies show increase in the level of myosin mRNA after thyroid treatment, and we observed the anticipated denser accumulation of myosin mRNA with in situ hybridization. There was, however, no change in the spatial distribution of mRNA. In vivo models having contractile deficiencies are problematic because one cannot decrease contractility in vivo without serious cardiac deficits to the animal. However the effects of decreased activity can readily be done in culture systems.

We reexamined the -MyHC mRNA localization to see how calcium and/or mechanical signals were controlling subcellular distribution. In the present study we confirmed (data not shown) that treating the cardiac myocytes with calcium-blocker verapamil results in perinuclear localization of -MyHC mRNA (7, 11, 21). However, cells treated with 7.5 mM BDM showed that -MyHC mRNA was distributed throughout the cytoplasm. Taking our published results with the new data on BDM, we conclude that -MyHC message localization in cardiac myocytes requires calcium signaling without significant involvement of cross-bridge activity.

The mechanism by which mechanical strain controls movement of mRNA may be related to focal adhesion complexes (4). Using magnetic forces to twist the cells, magnetometry detected a decrease in redistribution of poly(A)⁺ RNA, and ribosomes after endothelial cells were treated with 20 mM BDM. This demonstrated that tension per se is a trigger for these redistribution events. The discrepancies between these data and ours may be due to the higher BDM concentration they used. Although we know that this high BDM concentration would be sufficient to introduce calcium effects in the heart cells, we do not know how similar endothelial cells would respond to BDM treatment in terms of calcium transients.

In summary, we now demonstrate a differential regulation of the translation processes between pure mechanical (BDM treatment) and compound calcium and mechanical signaling (verapamil treatment). With BDM arrest, under normal calcium transients, we see increased levels of free, untranslatable, -MyHC message, suggesting that calcium is not necessary to block initiation. Thus pure mechanical signaling predominantly regulates the initiation stage of translation for -MyHC mRNA, whereas calcium cycling affects postinitiation steps of the translational process. Surprisingly, cross-bridge activity alone does not appear to be important for -MyHC message localization, whereas calcium cycling seems essential for the distribution of that mRNA throughout the cytoplasm. We conclude that direct inhibition of contractile machinery, but not calcium, regulates initiation of -MyHC mRNA translation. However, calcium, not pure mechanical signals, appears to be important for message localization.