The myosin light chain IF/3F locus contains two independent promoters, MLC1F and MLC3F, which are differentially activated during skeletal muscle development. Transcription at this locus is regulated by a 3' skeletal muscle enhancer element, which directs correct temporal and tissue-specific expression from the MLC1F promoter in transgenic mice. To investigate the role of this enhancer in regulation of the MLC3F promoter in vivo, we have analyzed reporter gene expression in transgenic mice containing lacZ under transcriptional control of the mouse MLC3F promoter and 3' enhancer element. Our results show that these regulatory elements direct strong expression of lacZ in skeletal muscle; the transgene, however, is activated 4-5 d before the endogenous MLC3F promoter, at the time of initiation of MLC1F transcription. In adult mice, transgene activity is downregulated in muscles that have reduced contributions of type IIB fibers (soleus and diaphragm). The rostrocaudal positional gradient of transgene expression documented for MLC1F transgenic mice (Donoghue, M., J. P. Merlie, N. Rosenthal, and J. R. Sanes. 1991. Proc. Natl. Acad. Sci. USA. 88:5847-5851) is not seen in MLC3F transgenic mice. Although MLC3F was previously thought to be restricted to skeletal striated muscle, the MLC3F-lacZ transgene is expressed in cardiac muscle from 7.5 d of development in a spatially restricted manner in the atria and left ventricular compartments, suggesting that transcriptional differences exist between cardiomyocytes in left and right compartments of the heart. We show here that transgene-directed expression of the MLC3F promoter reflects low level expression of endogenous MLC3F transcripts in the mouse heart.

We have previously reported the identification of a distinct myosin heavy chain (MyHC) isoform in a major subpopulation of rat skeletal muscle fibers, referred to as 2X fibers (Schiaffino, S., L. Gorza, S. Sartore, L. Saggin, M. Vianello, K. Gundersen, and T. Lømo. 1989. J. Muscle Res. Cell Motil. 10:197-205). However, it was not known whether 2X-MyHC is the product of posttranslational modification of other MyHCs or is coded by a distinct mRNA. We report here the isolation and characterization of cDNAs coding a MyHC isoform that is expressed in type 2X skeletal muscle fibers. 2X-MyHC transcripts differ from other MyHC transcripts in their restriction map and 3' end sequence and are thus derived from a distinct gene. In situ hybridization analyses show that 2X-MyHC transcripts are expressed at high levels in the diaphragm and fast hindlimb muscles and can be coexpressed either with 2B- or 2A-MyHC transcripts in a number of fibers. At the single fiber level the distribution of each MyHC mRNA closely matches that of the corresponding protein, determined by specific antibodies on serial sections. In hindlimb muscles 2X-, 2A-, and 2B-MyHC transcripts are first detected by postnatal day 2-5 and display from the earliest stages a distinct pattern of distribution in different muscles and different fibers. The emergence of type 2 MyHC isoforms thus defines a distinct neonatal phase of fiber type differentiation during muscle development. The functional significance of MyHC isoforms is discussed with particular reference to the velocity of shortening of skeletal muscle fibers.

The accumulation of two myogenic regulatory proteins, MyoD and myogenin, was investigated by double-immunocytochemistry and correlated with myosin heavy chain expression in different classes of myoblasts in culture and during early myogenesis in vivo. During in vitro differentiation of fetal myoblasts, MyoD-positive cells were detected first, followed by the appearance of cells positive for both MyoD and myogenin and finally by the appearance of differentiated myocytes and myotubes expressing myosin heavy chain (MHC). A similar pattern of expression was observed in cultures of embryonic and satellite cells. In contrast, most myogenic cells isolated from newly formed somites, expressed MHC in the absence of detectable levels of myogenin or MyoD. In vivo, the appearance of both myogenin and MyoD proteins was only detected at 10.5 d postcoitum (d.p.c.), when terminally differentiated muscle cells could already be identified in the myotome. Parasagittal sections of the caudal myotomes of 10.5-d-old embryos showed that expression of contractile proteins preceded the expression of myogenin or MyoD and, when coexpressed, MHC and myogenin did not co-localize within all the cells of the myotome. In the limb bud, however, many myogenin (or MyoD) positive/MHC negative cells could be observed in the proximal region at day 11. During further embryonic development the expression of these proteins remained constant in all the muscle anlagens examined, decreasing to a low level during the late fetal period. Western and Northern analysis confirmed that the myogenin protein could only be detected after 10.5 d.p.c. while the corresponding message was clearly present at 9.5 d.p.c., strongly suggesting a posttranscriptional regulation of myogenin during this stage of embryonic development. These data show that the first myogenic cells which appear in the mouse myotome, and can be cultured from it, accumulate muscle structural proteins in their cytoplasm without expressing detectable levels of myogenin protein (although the message is clearly accumulated). Neither MyoD message or protein are detectable in these cells, which may represent a distinct myogenic population whose role in development remains to be established.

The spatial and temporal expression pattern of the muscle regulatory gene Myf-6 (MRF4/herculin) has been investigated by in situ hybridization during embryonic and fetal mouse development. Here, we report that the Myf-6 gene shows a biphasic pattern of expression. Myf-6 transcripts are first detected in the most rostral somites of the mouse embryo at 9 d of gestation and accumulate progressively in myotomal cells along the rostro-caudal axis. This expression is transient and Myf-6 mRNA can no longer be detected in myotomal cells after day 12 post coitum (p.c.). In contrast to other muscle determination genes (MyoD1, myogenin, Myf-5), Myf-6 mRNA is not detected in limb buds or visceral arches and skeletal muscle of the mouse embryo (day 8-15 p.c.). In fetal mice, Myf-6 transcripts appear at day 16 p.c. in all skeletal muscles, and the gene continues to be expressed at a high level after birth. These results suggest that early Myf-6 expression may be restricted to a population of myogenic cells that does not contribute to the embryonic muscle masses in limb buds and visceral arches. The reappearance of Myf-6 mRNA in fetal skeletal muscle coincides approximately with secondary muscle fiber formation and the onset of innervation.

Expression of the two isoforms of cardiac myosin heavy chain (MHC), MHC alpha and MHC beta, in mammals is regulated postnatally by a variety of stimuli, including serum hormone levels. Less is known about the factors that regulate myosin gene expression in rapidly growing cardiac muscle in embryos. Using isoform-specific 35S-labeled cRNA probes corresponding to the two MHC genes and the two myosin alkali light chain (MLC) genes expressed in cardiac muscle, we have investigated the temporal and spatial pattern of expression of these different genes in the developing mouse heart by in situ hybridization. Between 7.5 and 8 d post coitum (p.c.), the newly formed cardiac tube begins to express MHC alpha, MHC beta, MLC1 atrial (MLC1A), and MLC1 ventricular (MLC1V) gene transcripts at high levels throughout the myocardium. As a distinct ventricular chamber forms between 8 and 9 d p.c., MHC beta mRNAs begin to be restricted to ventricular myocytes. This process is complete by 10.5 d p.c. During this time, MHC alpha mRNA levels decrease in ventricular muscle cells but continue to be expressed at high levels in atrial muscle cells. MHC alpha transcripts continue to decrease in ventricular myocytes until 16 d p.c., when they are detectable at low levels, but then increase, and finally replace MHC beta mRNAs in ventricular muscle by 7 d after birth. Like MHC beta, MLC1V transcripts become restricted to ventricular myocytes, but at a slower rate. MLC1V mRNAs continue to be detected at low levels in atrial cells until 15.5 d p.c. MLC1A mRNA levels gradually decrease but are still detectable in ventricular cells until a few days after birth. This dynamic pattern of changes in the myosin phenotype in the prenatal mouse heart suggests that there are different regulatory mechanisms for cell-specific expression of myosin isoforms during cardiac development.

Using in situ hybridization, we have investigated the temporal sequence of myosin gene expression in the developing skeletal muscle masses of mouse embryos. The probes used were isoform-specific, 35S-labeled antisense cRNAs to the known sarcomeric myosin heavy chain and myosin alkali light chain gene transcripts. Results showed that both cardiac and skeletal myosin heavy chain and myosin light chain mRNAs were first detected between 9 and 10 d post coitum (p.c.) in the myotomes of the most rostral somites. Myosin transcripts appeared in more caudal somites at later stages in a developmental gradient. The earliest myosin heavy chain transcripts detected code for the embryonic skeletal (MHCemb) and beta-cardiac (MHC beta) isoforms. Perinatal myosin heavy chain (MHCpn) transcripts begin to accumulate at 10.5 d p.c., which is much earlier than previously reported. At this stage, MHCemb is the major MHC transcript. By 12.5 d p.c., MHCpn and MHCemb mRNAs are present to an equal extent, and by 15.5 d p.c. the MHCpn transcript is the major MHC mRNA detected. Cardiac MHC beta transcripts are always present as a minor component. In contrast, the cardiac MLC1A mRNA is initially more abundant than that encoding the skeletal MLC1F isoform. By 12.5 d p.c. the two MLC mRNAs are present at similar levels, and by 15.5 d p.c., MLC1F is the predominant MLC transcript detected. Transcripts for the ventricular/slow (MLC1V) and another fast skeletal myosin light chain (MLC3F) are not detected in skeletal muscle before 15 d p.c., which marks the beginning of the fetal stage of muscle development. This is the first stage at which we can detect differences in expression of myosin genes between developing muscle fibers. We conclude that, during the development of the myotome and body wall muscles, different myosin genes follow independent patterns of activation and accumulation. The data presented are the first detailed study of myosin gene expression at these early stages of skeletal muscle development.