The Development of Longitudinal Variation of Myosin Isoforms in the Orbital Fibers of Extraocular Muscles of Rats

Abstract

purpose. To examine the appearance of longitudinal variation of extraocular and embryonic myosin heavy chain (MyHC) isoforms during the development of orbital singly innervated fibers of rat extraocular muscles (EOMs).

methods. EOMs were dissected from rat pups of various ages and stained with isoform-specific monoclonal antibodies to the embryonic and extraocular MyHC isoforms and to neurofilaments, as well as with labeled α-bungarotoxin. The orbital layers of whole muscles were examined by confocal microscopy. RNase protection assays for the embryonic (Myh3) and extraocular (Myh13) MyHC isoform mRNAs were also performed.

results. At 10 days postpartum, the EOM MyHC RNA was first detected by RNase protection assay. At 11 days postpartum, the extraocular isoform was detected in the orbital fibers as two thin stripes just proximal and distal to the neuromuscular junction (NMJ). Over the next few weeks, the area occupied by the extraocular isoform increased to include the entire central region of the orbital fibers at and surrounding the NMJ. At the same time, the embryonic isoform became excluded from the region of the NMJ.

conclusions. The orbital layer fibers of rat EOMs contain a longitudinal variation in MyHC isoforms not seen in other skeletal muscles. Development of this longitudinal variation begins as a late event postpartum; and the first appearance of it may be closely linked to neural contact. This targeting of MyHC isoforms to distinct domains is unique to EOMs.

Mammalian extraocular muscles (EOMs) play specialized roles as mediators of reflexive and voluntary eye movements. Because the EOMs are highly adapted to these roles, they exhibit fundamental differences from other skeletal muscles. ^{1 2 3} One of these differences, compartmentalization into a thin orbital region and a more substantial global region, is a phenomenon with an unknown purpose. What is known, however, is that the orbital and global layer fibers have unique properties, including differences in fiber types, fiber sizes, EMG characteristics, vascular content, metabolic activity, and response to botulinum toxin treatment for strabismus. ^{4 5 6 7 8 9} The most striking difference between rat orbital and global layer fibers is the longitudinal variation in properties of the orbital fibers. Both histochemistry ¹⁰ and immunohistochemistry ¹¹ had suggested that this variation extends to the myosin heavy chain (MyHC) isoforms themselves. Approximately 80% of the rat orbital fibers are singly innervated (oSIFs) ⁹ : their endplate regions contain the EOM-specific MyHC (Myh13) associated with a high speed of contraction and high stiffness minimum frequency, f _min— a measure of the kinetics of cross-bridge cycling in a fiber—whereas the flanking segments contain the embryonic MyHC (Myh3) associated with a slower speed of contraction and low f _min. ^{12 13} . The remaining orbital fibers are multiply innervated (oMIFs). These also contain the embryonic MyHC proximal and distal to the endplate region, but have a slow twitch or tonic MyHC throughout the length of the fiber. ¹³ These multiply innervated fibers generate action potentials in the region of the NMJ, but only tonic, nonpropagated potentials proximal and distal to the NMJ. ¹⁴ The central region of both oMIFs and oSIFs, moreover, corresponds to the localization of the fast sarcoplasmic reticulum Ca²⁺ pump. ¹⁵ Hence, in both orbital fiber types there is a characteristic longitudinal variation of properties: one set of properties at the NMJ, another set proximal and distal to it. 

This longitudinal arrangement of properties in the orbital fibers of rat muscles extends to the same fibers of rabbits. ¹⁶ That the pattern is constant among species examined suggests an important, but not yet demonstrated, function for these orbital fibers. Unlike the global fibers that insert on the sclera, the orbital fibers end well before the sclera. One theory ties the unique properties of orbital fibers to the active-pulley hypothesis, in which the orbital fibers attach to a specialized ring of tissue—the pulley—and alter the direction of action of the global layer fibers. This would have a significant effect on the rotational axis. ^{17 18 19}  

In addition, orbital layer fibers are recruited early during eye movements, have sustained activity throughout much of the oculomotor range, and thus are thought to be instrumental in maintaining ocular alignment. ²⁰ Consistent with this interpretation, studies suggest that the orbital layer fibers probably are the prime target of successful botulinum toxin treatment of strabismus ^{8 21} even though both orbital and global fibers are paralyzed by the toxin. Kranjc et al. ⁷ demonstrated that treatment of adult rat EOMs with botulinum toxin resulted in the permanent disappearance of the EOM-specific extraocular MyHC that normally resides in the endplate region of the oSIFs. 

In this article, we have examined how this novel, discontinuous pattern of MyHC is established during the development of the singly innervated orbital fibers. The following results show that the development of longitudinal variation in these fibers is a relatively late event in rat EOM maturation, occurring during the second week postpartum, during the time considered the “critical period” for development of the visual system. Moreover, the work strengthens our hypothesis that the longitudinal variation of isoforms is related to the development of the neuromuscular junction. 

Materials and Methods

Confocal Microscopy of Whole EOMs

The protocol for staining whole EOMs was adapted from one provided by Patricia Laboski, modified from the procedure of Dent et al. ²² Briefly, EOMs were exposed in isolated orbits and glycerinated in skinning solution (140 mM KCl, 8 mM MgCl₂, 5 mM Na₂ATP, 30 mM MOPS (3-(N-morpholino)propanesulfonic acid), 4 mM EGTA, [pH 7.1]) for 48 hours. Muscles were then excised and either rinsed in PBS before immunohistochemical staining or stored in 50% glycerine-50% skinning solution until used. After they were washed with PBS, the muscles were preincubated in PBS containing 10% fetal bovine serum and 0.5% Triton X-100. Muscles were stained overnight in this solution containing the appropriate dilution of primary antibody, washed thoroughly, and exposed overnight to secondary antibodies labeled with Alexa 488 or Alexa 546 (Molecular Probes, Eugene, OR). Some muscles were stained with Alexa 488-labeled α-bungarotoxin for 1 hour before the final washes. Muscles were examined with a laser scanning confocal microscope (LSM 510; Carl Zeiss Meditech, Jena, Germany). Through day 11 postpartum, it is difficult to isolate individual, whole EOMs from these preparations. Hence, for the earlier samples, muscles were not isolated; rather, entire orbits were stained and examined under the confocal microscope. In all images presented in this article, the microscope was focused on the orbital region only. Animals were maintained in approved accommodations at the School of Medicine, University of Pennsylvania, and used in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

RNase Protection Assay

The DNA templates for the RNA synthesis were as follows: a 246-bp fragment containing the 3′ embryonic MyHC translated and untranslated regions and a 213-bp fragment containing the 3′ extraocular MyHC translated and untranslated regions. These fragments were cloned into a pGEM-3Zf vector. PCR was performed using these clones and pUC/M13 forward primer and specific 3′ untranslated region (UTR) primer to the embryonic and the extraocular MyHCs. The PCR fragments contain a T7 RNA promoter and a specific 3′UTR. These were used as templates for RNA transcription. Antisense RNA probes were transcribed with T7 RNA polymerase in the presence of [α-³²P]UTP. The probe sizes were 156 bp for the embryonic and 115 bp for the extraocular MyHCs. Protected fragments were 116 and 95 bp, respectively. 

The RNase protection assay was performed using a kit (RPA III; Ambion, Austin, TX). Total RNA (20 μg) and 8 × 10⁴ cpm probes were mixed, coprecipitated, and resuspended in 10 μL of hybridization buffer, denatured at 95°C for 3 minutes, and incubated at 42°C overnight. Diluted RNase solution (150 μL; 1:100 RNase A/RNase T Mix in RNase digestion III buffer) was added to digest unprotected RNA at 37°C for 30 minutes. After precipitation with 225 μL RNase inactivation/precipitation III solution and yeast RNA, pellets were resuspended in 8 μL of loading buffer and the samples were separated on a 6% denaturing polyacrylamide gel, followed by exposure to film for a few hours to overnight. The 18S RNA probe, which hybridizes to 18S ribosomal RNA, was used for calibration. 

Results

In the orbital singly innervated fibers (oSIFs) of adult EOMs, a striking longitudinal variation in myosin heavy chain (MyHC) isoforms can be seen. When double stained with an antibody specific to the embryonic MyHC isoform and with one to the extraocular MyHC, the two staining patterns are overlapping, but not completely coincident. The antibody to extraocular MyHC stains a broad band in the center of the orbital fibers (Fig. 1A) . The embryonic MyHC antibody, on the other hand, stains areas proximal and distal to this central area and is sharply excluded from the central area (Fig. 1B) . The boundaries of staining with the extraocular MyHC antibody are not sharp, suggesting that there is a diffusion of this isoform from the center of the muscle fibers toward the periphery. Higher-power views of the region of overlap suggest that in this region both isoforms can contribute to the same sarcomeres (Fig. 1C) . 

We had previously hypothesized that this longitudinal variation in orbital layer MyHC isoform expression patterns is dependent on the presence of motoneuron contacts in the central region of the orbital fibers. ¹³ To test this hypothesis further, we localized the NMJs by double staining with labeled α-bungarotoxin (Fig. 2A) and with the antibody to embryonic MyHC (Fig. 2B) . With α-bungarotoxin staining of NMJs, the large NMJs of oSIFs were confined to a discrete central band, whereas the much smaller NMJs of oMIFs were distributed throughout the length of the muscle. The smaller NMJs are not seen in Figure 2 . Without exception, the population of large NMJs was found within the area from which the embryonic isoform was excluded, hence, in the center of the band containing the extraocular MyHC (Fig. 2C) . 

The appearance of longitudinal variation in MyHC isoforms in the orbital layer muscle fibers is a postnatal event. At 2 days postpartum, all fibers in the orbital region contained the embryonic MyHC through their entire length (Fig. 3A) , despite the presence of a well-formed endplate region (Fig. 3B) . Even at this early stage, axons were seen extending along the longitudinal axis of the muscle, presumably innervating the orbital and global multiply innervated fibers. At 6 days postpartum, all orbital fibers still reacted with an antibody to embryonic MyHC throughout their entire length (Fig. 3C) . The extraocular MyHC, however, was not detected at this stage (Fig. 3D) . 

By 11 days after birth, immunohistochemistry showed a small gap in staining with the anti-embryonic MyHC antibody, appearing approximately at the center of orbital muscle fibers (Fig. 4A) . At the same time, antibody staining was first detected for the extraocular MyHC isoform (Fig. 4B) . The extraocular isoform did not appear as a discrete band; rather, it arose as a doublet in the center of the muscle. The region in the center of the doublet not only excludes the extraocular MyHC isoform, but also is the same segment that similarly excludes the embryonic MyHC isoform. The relationship between the two stripes of extraocular MyHC and the developing NMJ can be seen in Figures 4C and 4D . The NMJ was localized directly between the two leaflets of the extraocular MyHC doublet. 

RNase protection assays confirm this postnatal appearance of the extraocular MyHC isoform. At 6 days postpartum, only the embryonic isoform can be detected (Fig. 5 , lane a). The extraocular isoform was first detected at 10 days (Fig. 5 , lane b) and increased thereafter (Fig. 5 , lanes c–d). Similarly, qPCR (data not shown) detected no extraocular MyHC mRNA at 6 days; 1.15 extraocular MyHC mRNA molecules/per microgram total RNA (×10⁸) could be detected at 16 days, with increasing absolute amounts through 30 days postpartum (1.99 mRNA molecules/microgram total RNA (×10⁸). In the adult, there were fewer extraocular MyHC mRNA molecules (1.00 mRNA molecules/μg total RNA (×10⁸). 

By 16 days postpartum, staining with antibodies to the extraocular and embryonic MyHCs again showed a clear absence of embryonic MyHC in the central region and a more diffuse doublet of extraocular MyHC around the center of the muscle (Fig. 6A 6B) . The embryonic isoform was now excluded from a wider area around the NMJ region than had been evident at 11 days, whereas the two stripes of extraocular MyHC isoform had widened. The NMJ area, however, still excluded both isoforms. At 20 days, however, whereas the embryonic isoform was still localized to the proximal and distal regions of the muscle, the extraocular MyHC completely filled the region around the NMJ and no longer appeared as a doublet (Figs. 6C 6D) . The only change between 20 days and later stages (Fig. 1) is a further spreading of extraocular MyHC around the NMJ region, including more overlap with the embryonic MyHC. 

Brueckner and Porter ²³ have demonstrated that dark-rearing rats during the critical period of visual development decreases the mRNA for the extraocular MyHC. This decrease was significant, but not complete. It seemed possible to us that dark rearing would disrupt the development of longitudinal variation by decreasing the extraocular MyHC and inhibiting the exclusion of the embryonic isoform from the area around the central NMJ band. Accordingly, rats were dark reared through 30 days postpartum, and their EOMs were isolated and double-stained for both MyHC isoforms. In Figure 7B , the antibody to the extraocular MyHC isoform still stains a region in the center of the muscle, but stains with significantly less intensity than in a normal animal (see Fig. 1 , for example). Staining with the embryonic antibody, on the contrary, appears undiminished in quantity and shows the normal exclusion from the presumptive NMJ region (Fig. 7A) . Hence, the disruption that occurs with dark rearing appears to separate the mechanisms of increased extraocular MyHC synthesis from establishment of the pattern of longitudinal variation of the two isoforms. Whereas this suggests that a third MyHC isoform may take the place of the extraocular MyHC in the midsection of the muscle, we have not examined this question. 

Discussion

Temporal-Spatial Changes in MyHC Expression in Orbital Singly Innervated Fibers during Development

In the present study, we show that the focal point of the postnatal development of longitudinal variation in MyHC isoforms in EOM orbital layer fibers is the area surrounding the NMJ. The extraocular MyHC first appeared as two stripes, immediately proximal and distal to the endplate band. With time, these stripes widened until the extraocular isoform appeared to occupy a broad band covering and surrounding the adult NMJ. This time course corresponds well to results presented previously by one of us, using other methods. ⁶  

The “spreading” of the extraocular isoform appeared to be similar to a mechanism seen in experiments with dystrophin. In those experiments, synthesis of the dystrophin molecule was under direct control of individual muscle nuclei, and dystrophin-positive domains were clearly identified in hybrid muscle fibers formed by fusion of dystrophin-positive and -negative myoblasts. ²⁴ Domain boundaries, however, became blurred by diffusion of dystrophin molecules from their point of synthesis. The temporal sequence of the localization of the extraocular MyHC suggests a similar sequence of events for this isoform during postnatal development. This isoform may be synthesized only by the nuclei underlying the two original stripes; and the spread of the isoform may be due to diffusion of the isoform itself. It is also possible, however, that responsibility for the synthesis of this isoform spreads to additional nuclei. The embryonic MyHC isoform, in contrast, was the dominant isoform in orbital fibers during the first week postpartum and extended through the entire length of the fibers. As soon as the doublet of extraocular MyHC appeared, the embryonic isoform also began to become excluded from the region of the NMJ, and by the third week postpartum, it was sharply excluded from this region. Although neither the extraocular nor the embryonic isoform occupied the immediate area of the NMJ at 11 days, we have not examined which MyHC isoform occupies that area. Finally, examination of myosin isoform regulation in dark-reared rats suggests that different mechanisms may be responsible for patterning versus quantitative expression of extraocular and embryonic MyHCs in EOM orbital layer fibers. 

Role of the Nerve in Patterning MyHC Gene Expression

Several observations suggest that the motoneuron also plays a role in directing the synthesis of specific isoforms and, perhaps, the repression of other isoforms in the orbital fibers of the rat EOMs. Before the appearance of the extraocular isoform, the embryonic isoform occurred throughout the length of the fibers. The initial appearance of the extraocular MyHC isoform is just proximal and distal to the NMJ, and at this time, both the extraocular and the embryonic isoforms were inhibited at the point of neuromuscular contact. Moreover, the appearance of the extraocular MyHC isoform occurs during the period of maturation of the acetylcholine receptor (AChR), with the γ-subunit being replaced by the ε-subunit (in the mouse) at ∼9 days postpartum. ²⁵ This switch causes changes in the conductance and mean open time of the channels, changes that could be related to the establishment of longitudinal arrangement of MyHC isoforms. This could repress the synthesis of extraocular and embryonic isoforms directly under the synapse. 

Other evidence suggests that neural activity is important in regulating the expression of extraocular MyHC. Brueckner and Porter ²³ have shown that this isoform can be inhibited by dark-rearing newborn pups or by disturbing vestibular activity, ²⁶ suggesting that neural activity associated with the optokinetic and vestibulo-ocular reflexes stimulate extraocular MyHC expression. Not surprisingly, significant changes in isometric contractile function, including lengthening of the time to peak tension, occurred in rat pups that were dark reared compared with normally raised littermates. ²⁷ Despite the quantitative decrease in the extraocular MyHC after dark rearing, the qualitative arrangement of neither the embryonic nor the extraocular isoform was disturbed. This is reminiscent of the control of the AChRs, which are positively regulated in subsynaptic nuclei, but negatively regulated in extrasynaptic nuclei. ^{28 29} One current theory is that the action potential itself is the signal for repression of the synthesis of the AChRs distal to the synapse, whereas positive control at the synapse is initiated by a molecule released by the nerve terminal. ³⁰ It is possible that the unusual longitudinal arrangement of MyHC isoforms in the orbital fibers of EOMs is also controlled by a combination of chemical signals from the nerve terminal and electrical activity. 

Kranjc et al. ⁷ treated adult rat EOMs with botulinum toxin, a procedure that effectively denervates the muscle. The extraocular MyHC isoform disappeared, and did not reappear—by their criteria—after the resumption of functional synaptic transmission. Although not shown in their work, the normal longitudinal variation of MyHC isoforms in the orbital fibers of the rat EOM could have been disrupted by the toxin treatment. If true, this suggests that the some neural activity is essential for the maintenance of extraocular MyHC in the oSIF. The fact that extraocular MyHC isoform expression may not have resumed after resumption of neural activity suggests that the initiation of this pattern of extraocular MyHC expression could require a complex interplay of factors found only during normal ontogeny. 

The idea, however, that the area of the nerve–muscle contact can direct the localized synthesis of a particular MyHC isoform has been confirmed in several other systems. For example, Salviati et al. ³¹ demonstrated that when a rat soleus slow-twitch muscle fiber was innervated ectopically by a fast motoneuron, synthesis of a fast myosin was induced and was constrained to the ectopic endplate region. In another, naturally occurring example, Pedrosa-Domellof et al. ³² examined the MyHC isoforms of fibers of the rat muscle spindles. The nuclear bag fibers showed a longitudinal variation of isoforms along their length; denervation of the developing fibers resulted in the synthesis of the same MyHC isoforms, but in an absence of longitudinal variation. They concluded that motor innervation regulates the expression of some isoforms along the length of intrafusal fibers, and this pattern can be disrupted by denervation of the spindle fibers. 

Possible Mechanisms for Generating Temporospatial Patterning of MyHC Expression

How can we account for the stripes of extraocular MyHC just proximal and distal to the NMJ? Similar organization of molecules can be seen in a variety of early developing systems. During development of Drosophila, for example, such stripes of gene expression can be established either by differential response to a gradient of one or more transcriptional activators or a balance between gradients of transcriptional activators and repressors (see Ref. ³³ for a discussion of the establishment of such patterns). The multinucleated nature of adult muscle fibers could be considered similar to the syncytial nature of the early Drosophila embryo. Electrical and/or chemical signaling at the maturing NMJ may establish one or more gradients that fade toward the ends of the fiber. This mechanism gives each nucleus in the myofiber positional information. In the simplest form, a single gradient could define both the stripe of extraocular MyHC and the inhibition of the embryonic isoform in the central region. More likely, in line with other developmental mechanisms, a combination of activators and inhibitors play a role in defining the longitudinal variation of these isoforms. What form a hypothetical gradient could take and why the extraocular isoform is not synthesized until the second week postpartum are additional questions and the focus of future work. 

 Supported by Grant EY11779 from the National Eye Institute.

 Submitted for publication February 3, 2004; revised April 26, 2004; accepted May 6, 2004.

 Disclosure: N.A. Rubinstein, None; J.D. Porter, None; J.F.Y. Hoh, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Confocal images of the orbital region of a single adult rat EOM double-stained with antibodies to the extraocular MyHC (red) and the embryonic MyHC (green). (A, B) Lower-power views of the entire muscle; (C) high-power view of the transitional region between the embryonic positive and the extraocular positive MyHCs. Myofibrils that are yellow contain both isoforms.

Figure 2.

 

Confocal images of an adult EOM double stained with α-bungarotoxin (A) to localize the synapses and with an antibody to the embryonic MyHC (B). (C) A superimposition of (A) and (B). The synapses are localized in the region devoid of the embryonic isoform.

Figure 3.

 

Confocal images of an EOM from a 2-day-old (A, B) and a 6-day-old (C, D) rat pup stained with antibody to the embryonic MyHC isoform (A, C), with antibody to neurofilaments (B), and with antibody to the extraocular MyHC isoform (D). Note that (A) and (B) are images of the same muscle, and (C) and (D) are the same muscle.

Figure 4.

 

Confocal images of several EOMs from 11-day-old rat pups. One EOM is double-stained with antibody to the embryonic (A) and extraocular (B) MyHC isoforms. A second EOM is double-stained with antibody to the extraocular isoform (C) and with α-bungarotoxin to identify the neuromuscular junctions (D). This higher-power view focuses on the central region of the muscle, an area similar to the positively stained segment of (B). In (A) and (B), because the image was made of an EOM still within the isolated orbit, the entire muscle is not in focus.

Figure 5.

 

RNase protection assays of the extraocular and embryonic MyHC mRNAs from 6 days postpartum through adult stages. RNA samples were exposed to both probes on days 6 (lane a), 10 (lane b), 15 (lane c), and 20 (lane d). Adult samples were exposed to either the embryonic MyHC probe (lane e) or the extraocular MyHC probe (lane f). Protected fragments were 116 bp (embryonic MyHC) or 95 bp (extraocular MyHC). The extraocular isoform mRNA appeared between 6 and 10 days postpartum, correlating well with the immunohistochemical data in Figure 4 .

Figure 6.

 

Confocal images of EOMs from rat pups on days 16 (A, B) and 20 (C, D). Muscles were double-stained with antibody to the embryonic MyHC (A, C) and antibody to the extraocular MyHC (B, D).

Figure 7.

 

Confocal images of rat EOMs after dark rearing. Rats were dark reared through 30 days postpartum, and their EOMs were isolated and double-stained for the embryonic MyHC isoform (A) and the extraocular MyHC (B).