Effect of diabetes and hypothyroidism on the predominance of cardiac myosin heavy chains synthesized in vivo or in a cell-free system.

Rat cardiac ventricular myosins and RNA were prepared from normal, diabetic, insulin-treated diabetic, hypothyroid, and 3,3',5-triiodo-L-thyronine-treated hypothyroid rats. Myosin heavy chains isolated from purified myosin or synthesized in vitro from cardiac RNAs were subjected to partial protease digestion during sodium dodecyl sulfate-polyacrylamide gel electrophoresis. It was found that peptide maps obtained from cardiac myosin heavy chains of hypothyroid and diabetic rats were identical but differed from the maps of myosin heavy chain from control and hormone-treated animals. The same results were obtained, whether the heavy chains were isolated from purified myosin synthesized in the intact heart or from translation products coded for by cardiac RNAs added to the modified reticulocyte lysate. These results indicate that the myosin heavy chain RNA species present in the hypothyroid heart is also expressed during insulin deficiency but differs from the species expressed in normal animals. The expression of the two myosin heavy chain RNA species found in the rat cardiac ventricle appears to be independently regulated by these two hormones.

the highest Ca2+-activated ATPase activity is the predominant form in young, healthy adults; V3 with the lowest Caz+activated ATPase activity is the major embryonic form. An isozyme indistinguishable from embryonic V3 myosin by electrophoretic migration is also present in small amounts in the young adult and becomes the predominant form in animals rendered severely hypothyroid or diabetic (1, 3, 11).
A minimum of 5 MHC genes are present in the rat and at least 2 are expressed in the heart (12, 13). Indeed, Schwartz et al. (7) were able to demonstrate by specific antibodies and chymotrypsin mapping that MHC expressed in adult hypothyroid heart was identical to the MHC expressed in embryonic heart but differed from the MHC expressed in normal young hearts. In addition, since two partial cDNA clones exhibiting slightly different nucleotide sequences have been obtained from young adult rat cardiac mRNA (12), at least two different MHC genes are expressed in the heart of young rats. These two clones hybridized only weakly with embryonic cardiac mRNA, suggesting that either two allelic forms of MHCa are expressed in young adults as proposed by the authors, or MHCp expressed in young adults is not identical to that expressed in embryos. In the rabbit heart, myosin can also be expressed as isoforms and it has been established that MHCs present in V, and V, are coded for by two different mRNAs (14). Different MHCs giving rise to different peptide maps, however, have been reported to be generated by posttranslational modification of a single species of MHC in the chicken skeletal muscle (15). Therefore, although electrophoretically identical isoforms of native myosin Vs can be isolated from hypothyroid or diabetic heart in the rat, it is not evident that these two pathological conditions, induced by the lack of hormones that have very different mechanisms of action, result in the presence of a common MHC in myosin V3.
In this report, we present evidence that MHCs translated from RNA isolated from hypothyroid and diabetic animals are identical. They differ, however, from MHC coded for by RNA isolated from the heart of either control animals or hypothyroid animals treated with thyroid hormone and diabetic animals treated with insulin. Thus, although the action of T3 is mediated via nuclear receptors while that of insulin is via membrane receptors, these two hormones either directly or indirectly appear to regulate the expression of the two cardiac mRNAs which code for MHCa and MHCP.  (17) and TB was determined by radioimmunoassay (18). Four weeks after the onset of hypothyroidism, half of the animals received daily injections of 0.4 pg of T3/100 g of body weight while the remaining animals were kept completely hypothyroid. Diabetes was induced by administration of 65 mg of streptozotocin/kg body weight. Blood glucose levels in control animals were 125 & 18 mg/deciliter and in diabetic animals 435 -C 30 mg/deciliter. Blood glucose levels did not change significantly during the course of the experiments. Three days after diabetes induction, half of the animals were given daily intramuscular injections of 2 units of protamine zinc insulin/ 100 g body weight for three weeks. The other half were kept for the 2036 Cardiac Myosin Heavy Chains in Diabetes and Hypothyroidism same length of time without insulin administration. Control animals matched for age were included in the stay-Animals were killed by exsanguination. Fresh hearts pooled from 2 to 3 animals were used for each RNA preparation. Individual hearts frozen in liquid nitrogen and stored at -80 "C for 2 to 8 days were used for the preparation of myosin.

Animal Preparations-Male
Peptide Map of Myosin Heavy Chain-Myosin was purified and Ca2+-activated myosin ATPase was measured as previously described (19). Peptide mapping of myosin heavy chain from the 5 groups of animals was performed as described by Cleveland et 01. (20). Myosin heavy chain was either digested in solution without prior purification by incubating purified myosin with Staphylococcus aureus Vs protease (Miles Laboratories) as indicated in Fig. 1 or was prepurified from either myosin or total translation products on 5% polyacrylamide gels using the system described by Laemmli (21).
Minor modifications were introduced in the technique described by Cleveland et al. (20). Polyacrylamide slab gels cross-linked with AcrylAide and covalently bound to Gelbond PAG films (FMC C') were used to analyze the polypeptides after myosin digestion in solution. Polyacrylamide slab gels, also cross-linked with AcrylAide but not bound to a support, were used to separate MHC from other subunits. The gels, stained for 15 min, were destained for only 15 min in methanol/acetic acid, then for 30 min in distilled water. MHC bands were cut and incubated in Cleveland buffer (20) for 2 h and applied to a second gel containing 12.5% polyacrylamide cross-linked with bisacrylamide. Because of the large molecular mass of MHC compared to that of the proteolytic enzymes, segregation of intact substrate and enzymes already occurred in the stacking gels. Therefore, the volume of the enzyme solutions layered onto the preparative gel pieces containing MHC was increased to 30 pl and the length of the 3% polyacrylamide stacking gel was reduced to 0.5 cm to prevent further digestion of peptides generated from MHC digestion during migration through the preparative gel piece. The enzyme solutions were overlaid with 10 pl of 10% sucrose containing 0.01% bromphenol blue and the gels were run at a constant current of 4 mA/gel until the tracking dye reached the stacking gel and at 15 mA/gel thereafter. The concentrations of S. aureus Va protease and chymotrypsin were as indicated in the text. After electrophoresis, the peptides were stained with Coomassie brilliant blue R 250 or silver (23).
RNA Preparation and Translation-RNA was prepared from fresh hearts as previously described (16). Unfractionated RNA was translated in the nuclease-treated rabbit reticulocyte lysate prepared in our laboratory according to Pelham and Jackson (22).
Ten to 15 pCi of [%]methionine (900-1400 Ci/mmol) and 1-2 pg of RNA were added per 10-pl reaction. After 1-h incubation at 30 'C, the reactions were stopped by the addition of 1 pg of RNase A and 20 pg of methionine. After 15 min at room temperature, 1 M dithiothreitol was added to a final concentration of 50 mM. Incorporation of ["Slmethionine into acid-insoluble material was measured by the method of Mans and Novelli (24). ["SIMethionine incorporation into MHC was quantitated as follows. To aliquots containing 100,OOO acid-precipitable cpm, were added 10 pg of purified myosin. MHC was then separated from other translation products by electrophoresis on a 10% polyacrylamide-SDS gel (21). Proteins were visualized by staining, and the MHC band was cut out, thoroughly destained in 45% methanol, 10% acetic acid, dried, and counted in 5 ml of Betamax (WestChem). Translation aliquots containing the same amount of ["SIMHC were used for peptide mapping. After the addition of 10 pg of purified myosin to each sample, mapping was performed as described for purified myosin. At the end of electrophoresis, the gels were prepared for fluorography as previously described (16).

RESULTS
The hormonal status of hypothyroid and diabetic animals was documented as previously described (3, 16). Administration of 0.4 pg of T3/100 g body weightlday to hypothyroid rats and 2 units of protamine zinc insulin/100 g body weight/ day to diabetic animals normalized Ca2+-activated myosin ATPase activity and the myosin isoenzyme distribution pattern.
The predominance of MHC among total cardiac proteins was similar for animals of different hormonal status; the mobility of MHC in SDS-polyacrylamide slab gels, however, appeared to be very slightly increased when myosin was isolated from hormone-deficient hearts (Fig. 1). There was no S. aureus VB protease digestion of rat cardiac myosin heavy chain. Fifteen pg of purified myosin were applied to a 5% polyacrylamide-AcrylAide separating gel as described under "Materials and Methods." Slices containing MHCs were placed into wells of the analytical gel and overlaid with 30 pl of protease as indicated. A 1.5mm thick, 12.5% polyacrylamide gel was used to separate the peptides which were visualized by silver staining. Other conditions as described in Fig. 1. difference between the 5 myosin preparations used, although a slight increase in minor contaminants may be present in the myosin isolated from insulin-treated diabetic animals ( Fig. 1, I). Limited S. aureus Vs protease digestion of myosin isolated from the hearts of normal and hormone-injected rats led to peptide maps different from those obtained from myosin Cardiac Myosin Heavy Chains of hypothyroid and diabetic animals (Fig. 1). Six peptides characteristic of myosin isolated from the heart of normal (C), Tdreated (T), or insulin-treated (I) rats were seen at the lower protease concentration (0.08 pg). None of these peptides were present in the digests of either diabetic (D) or hypothyroid (H) myosin; they were replaced by seven new peptides not seen or only faintly visible in the normal and hormone-treated myosin. Unique peptides characteristic of either diabetic and hypothyroid or control and hormonetreated myosin were observed at the two concentrations of V8 protease. The two lowest peptides seen at the higher protease concentration may not originate from MHC since myosin light chain 2 was also digested at this protease concentration. When MHC was first separated from other myosin subunits and digestion was allowed to proceed during electrophoresis, a smaller set of peptides was obtained even a t a protease concentration which resulted in only partial digestion of MHC (Fig. 2). However, as observed for myosin digested in solution, MHC, while at least 4 peptides were unique for diabetic and hypothyroid MHC. Similar results were obtained when Ve protease was replaced by chymotrypsin; one set of peptides common to control (C) and hormone-treated (T, I ) MHC (Fig. 3, arrows on the left) were not observed in the hormonedeficient myosins (D, H), and 5 peptides seen in the latter were not present or barely visible in the control group (Fig. 3, arrows on the right). It should be mentioned that the control MHC map differs from the hormone-treated MHC maps by 2 peptides (broken arrows).

MHC
To determine whether the differences in the peptide maps of MHC resulted from post-translational modifications of a single MHC species or from two MHC polypeptides coded for by different mRNA species, RNAs isolated from the heart of normal, hormone-deficient, and hormone-treated animals were translated in a cell-free system, and [35S]methioninelabeled MHC was subjected to limited chymotrypsin or S. aureus V8 protease digestion. The distribution of total translational products coded for by RNA isolated from the five groups of animals was analyzed by SDS-polyacrylamide gel electrophoresis. MHC migrated with the same mobility in each group and represents one of the major translational products (data not shown).
[%]Methionine present in MHC was measured as described under "Materials and Methods" and aliquots containing the same amounts of radioactive MHC were used for limited protease digestion. Results presented in Fig. 4 (S. aureus V8 protease digests) and Fig. 5 (chymotrypsin digests) show that peptide maps obtained from diabetic and hypothyroid MHC are identical and differ from peptide maps obtained from the control and hormone-treated MHC. It should be noted that the T3-treated sample appears to contain peptides characteristic of both types of MHC, indicating that the dose of T3 used to treat animals in this particular experiment was not sufficient to completely normalize MHC mRNA.

DISCUSSION
This report demonstrates for the first time that mRNA isolated from the heart of hypothyroid and diabetic rats synthesizes an identical MHC which differs from MHC coded

Cardiac Myosin Heavy Chains in Diabetes and Hypothyroidism
for by cardiac RNA isolated from normal or hormone-treated animals. The same MHC RNA species must therefore be present in hypothyroid and diabetic ventricles, and must differ from mRNA present in the heart of normal and hormonetreated animals. The results indicate that at least two different mRNA species coding for two myosin heavy chains must be expressed in the rat heart, as it has been shown for the rabbit heart (14).
Although others have reported that rat cardiac MHC expressed in hypothyroid rats differs from that occurring in normal rats (1,7), the possibility that posttranslation modification of a single MHC species resulted in the observed differences in peptide maps could not be excluded in these reports. In the chicken fast twitch muscle, for example, MHCs having different peptide maps are coded for by a single mRNA species (15). In the present study, different maps were obtained from hormone-deficient and normal MHCs whether MHC was obtained from intact myosin purified from cardiac ventricles or from translation products synthesized in vitro from cardiac RNAs. The slight difference seen between control and hormone-treated myosin digests only when chymotrypsin was used to digest MHC synthesized in vivo may indicate different levels of post-translational modification in the 2 groups.
MHC is a large polypeptide which represents a significant fraction of the total cardiac proteins. In comparing peptide maps of MHCs obtained from hormone-deficient and control tissues, we have assumed that the prominent band, having a molecular mass of 200,000 daltons, is homogeneous and represents MHC. Although it is possible that this band contains other polypeptides in addition to MHC, no other peptide having a similar M , could be identified by two-dimensional gel electrophoresis (16). In addition, Chizzonite et at. (25) obtained similar peptide maps from rabbit cardiac myosins before and after immunoadsorption, indicating that MHC is the major polypeptide present in the 200,000-dalton band. Although our results do not permit determination of whether the different MHC RNA species result from the transcription of different MHC genes or are due to differences in the processing of a single hnRNA precursor species transcribed from one gene (26), it is most likely that these 2 mRNA species are transcribed from two different genes. Data obtained by Shina et al. (14) using specific cDNA clones derived from MHC mRNA from normal and thyrotoxic rabbit hearts strongly suggest that the a and /3 isoforms of myosin heavy chain are the products of two different genes. In addition, heteroduplex analysis of genomic clones (12,27) indicate the presence of an MHC multigene family in rats and rabbits.
Recent immunological studies demonstrate that the expression of each MHC is not restricted to different cardiac myocyte populations but that both isoforms are present together in a large fraction of the cell population (28), suggesting multifactorial regulation of gene expression in the heart. Whether diabetes-and hypothyroidism-induced increase in MHCP mRNA predominance and decrease in MHCa mRNA predominance result from mechanisms common to both conditions or indicate multihormonal control of the MHC genes is currently unclear. T4 and T3 hormone levels are decreased in diabetic rats; however, they are not lowered to the hypothyroid range. In addition, previous studies have shown that physiological Tx replacement doses which normalize myosin isoenzyme distributions in hypothyroid rats do not do so in diabetic animals (29). The lower T4 and T3 levels of diabetic rats therefore cannot explain the myosin V3 predominance which occurs in diabetes.