Differential expression of B-crystallin and Hsp27 in skeletal muscle during continuous contractile activity. Relationship to myogenic regulatory factors.

AlphaB-crystallin (alphaBC) is a major structural protein (22 kDa) of the ocular lens as well as a bona fide heat shock protein in non-lens tissue. The alphaBC gene is abundantly expressed in tissues with high oxidative capacity, including the heart and type I skeletal muscle fibers, and is regulated by the MyoD family of basic helix-loop-helix transcription factors during myogenesis. To test the hypothesis that alphaBC expression may be directly regulated by the demand for oxidative metabolism, we examined the expression of alphaBC and the ancestral-related Hsp27 in rabbit tibialis anterior muscle subjected to continuous low frequency motor nerve stimulation (3 V, 10 Hz). alphaBC mRNA and protein increased within the 1st day of continuous contractile activity (5- and 2.5-fold, respectively) and achieved maximum levels (>20-and 4-fold, respectively) after 21 d of stimulation. Hsp27 mRNA and protein levels also increased with stimulation, but with a less specific and dramatic induction pattern. In agreement with the Northern analysis, in situ hybridization performed on cross sections from tibialis anterior muscle revealed progressively increasing alphaBC transcript signal, localized in a ringlet pattern, from 1 through 21 days of stimulation. Serial sections subjected to myosin immunohistochemistry revealed that alphaBC expression was confined to slow-twitch type I and a subpopulation of fast twitch type II fibers after 1 day but present in nearly all fibers after 21 days of stimulation. Transcript levels of all four myogenic regulatory factors (MyoD, myogenin, myf-5, and MRF4) also increased with stimulation in a pattern temporally similar with alphaBC, suggesting that expression of alphaBC in response to stimulation may, in part, be regulated through myogenic regulatory factor(s) interaction with the canonical E-box element located within the alphaBC promotor. These data demonstrate that expression of the small heat shock protein, alphaBC, is rapidly induced independent of the ancestrally related Hsp27 in a fiber type specific pattern in skeletal muscle subjected to the oxidative stress imposed by continuous contractile activity.

␣B-crystallin (␣BC) is a major structural protein (22 kDa) of the ocular lens as well as a bona fide heat shock protein in non-lens tissue. The ␣BC gene is abundantly expressed in tissues with high oxidative capacity, including the heart and type I skeletal muscle fibers, and is regulated by the MyoD family of basic helix-loop-helix transcription factors during myogenesis. To test the hypothesis that ␣BC expression may be directly regulated by the demand for oxidative metabolism, we examined the expression of ␣BC and the ancestral-related Hsp27 in rabbit tibialis anterior muscle subjected to continuous low frequency motor nerve stimulation (3 V, 10 Hz). ␣BC mRNA and protein increased within the 1st day of continuous contractile activity (5-and 2.5-fold, respectively) and achieved maximum levels (>20-and 4-fold, respectively) after 21 d of stimulation. Hsp27 mRNA and protein levels also increased with stimulation, but with a less specific and dramatic induction pattern. In agreement with the Northern analysis, in situ hybridization performed on cross sections from tibialis anterior muscle revealed progressively increasing ␣BC transcript signal, localized in a ringlet pattern, from 1 through 21 days of stimulation. Serial sections subjected to myosin immunohistochemistry revealed that ␣BC expression was confined to slow-twitch type I and a subpopulation of fast twitch type II fibers after 1 day but present in nearly all fibers after 21 days of stimulation. Transcript levels of all four myogenic regulatory factors (MyoD, myogenin, myf-5, and MRF4) also increased with stimulation in a pattern temporally similar with ␣BC, suggesting that expression of ␣BC in response to stimulation may, in part, be regulated through myogenic regulatory factor(s) interaction with the canonical E-box element located within the ␣BC promotor. These data demonstrate that expression of the small heat shock protein, ␣BC, is rapidly induced independent of the ancestrally related Hsp27 in a fiber type specific pattern in skeletal muscle subjected to the oxidative stress imposed by continuous contractile activity.
Crystallins constitute the major water-soluble proteins (Ͼ90%) of the ocular lens where their unique biochemical and structural properties render the lens transparent (1). Surprisingly, however, these specialized proteins are also expressed in non-lens tissues and are, in fact, related or identical to specific metabolic enzymes and stress proteins (2). This dual functionality has led to the concept of gene sharing; the idea that crystallins have been recruited to the lens through evolution for their inherent structural characteristics (i.e. ability to facilitate light refraction) apart from their distinct yet unknown functional properties within other tissues (2)(3)(4).
Among the family of crystallins (␣, ␤, ␥, and ␦) found in vertebrate lenses, the ␣B subunit of ␣-crystallin is unique in that it is expressed independent of the more lens-restricted ␣A subunit in tissues that possess high levels of mitochondria such as heart, type I and IIa skeletal myofibers, and specific regions of the kidney (5)(6)(7)(8). Although a specific enzymatic or cofactor activity has not, as yet, been ascribed to ␣B-crystallin in relation to oxidative metabolism, work over the past several years has established that ␣B-crystallin (22 kDa) is a bona fide member of the small molecular weight (25-27 kDa) heat shock protein (Hsp) 1 family (9 -12). Both ␣B-crystallin and Hsp27 share considerable sequence and structural similarity (13,14), associate in vivo (15,16), and are co-induced in response to heat and oxidative stress (14,(17)(18)(19). ␣B-crystallin, like Hsp27, also has been shown in vitro to prevent aggregation of denatured proteins in response to stress and to facilitate protein refolding upon removal of stress, confirming its status as a molecular chaperone (9,11,12).
The possibility that regulation of ␣B-crystallin expression may be directly linked to the demand for oxidative metabolism has recently been explored in skeletal muscle. In addition to a rapid decline in respiratory capacity, disuse atrophy of the soleus muscle (a slow twitch muscle composed primarily of oxidative type I and IIa fibers) induced by denervation, hind limb suspension, or tenotomy is associated with a marked reduction in both ␣B-crystallin and Hsp27 expression (5,20,21). In contrast, ␣B-crystallin is abundantly expressed in the mitochondria rich "ragged red fibers" associated with various forms of skeletal muscle mitochondrial myopathies (22). Taken together, these findings provide support for the hypothesis that expression of the small heat shock proteins may be regulated by oxidative stress produced by abnormal and/or increased demand for mitochondrial metabolism.
In the present study, we have specifically examined whether a sudden and continuous increase in the demand for oxidative metabolism directly influences the expression of ␣B-crystallin and Hsp27 in skeletal muscle. Using a well characterized model of mitochondrial biogenesis (23), our findings demonstrate that 21-day continuous contractile activity elicited by chronic low frequency motor nerve stimulation dramatically increases ␣B-crystallin expression in a fiber-type specific pat-tern. In contrast, the ancestrally related Hsp27 displays very little specific change in response to stimulation. To examine the potential mechanisms mediating the activity-induced regulation of ␣B-crystallin, we also determined the expression pattern of the myogenic regulatory factors (MRF) MyoD, myogenin, myf-5, and MRF4 (24). ␣B-crystallin is unique among all known molecular chaperones in that the 5Ј-regulatory region of the gene contains a canonical E-box element, the consensus DNA binding site for MRFs (25). In vitro studies have demonstrated that both MyoD and myogenin can bind to the E-box element in the ␣B-crystallin promotor and fully transactivate ␣B-crystallin expression in a muscle specific manner (25). In the present study, transcript levels of all four myogenic regulatory factors (MyoD, myogenin, myf-5, and MRF4) increased with stimulation in a pattern temporally similar with ␣Bcrystallin. These data raise the possibility that expression of ␣B-crystallin in response to stimulation may, in part, be regulated through MRF(s) interaction with the E-box element(s) located within the 5Ј-regulatory region of the ␣B-crystallin gene.

EXPERIMENTAL PROCEDURES
Materials-Adult New Zealand White rabbits weighing ϳ3.0 kg were purchased from Myrtles Rabbitry. Hybridization membrane (Hybond N) and all radiolabeled compounds were purchased from Amersham Corp. All restriction enzymes and other chemicals were of molecular biology grade and purchased from either Promega, Life Technologies, Inc., Sigma, or Fisher.
Animal Surgery-Rabbits were anesthetized by isoflurane inhalation and, under sterile conditions, electrodes were placed adjacent to the common peroneal nerve for stimulation of the tibialis anterior (TA) muscle of the hind limb. The leads were externalized and attached securely to pulse generators for pacing continuously at 6 -10 Hz for 4 and 8 h, and for 1, 3, 7, 14, or 21 days.
Preparation of Tissue-At the completion of the study, the rabbits were anesthetized with sodium pentobarbital (50 mg/kg, intravenously). The TA muscle was quickly removed, and ϳ500-mg pieces were quick frozen in liquid nitrogen and stored at Ϫ80°C for subsequent protein and RNA analysis. For in situ hybridization studies, portions of TA muscle from both the stimulated and contralateral unstimulated hind limbs were fixed in 4% paraformadehyde overnight, dehydrated in graded ethanols, cleared with xylene, and embedded in paraffin. Transverse sections were cut at 4 m and floated onto slides treated with Vectabond (Vector Laboratories). All protocols were reviewed and approved by the Institutional Animal Care and Research Advisory Committee and were conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals.
RNA Isolation and Northern Blot Analysis-Total RNA was isolated from ϳ200 mg of powdered TA muscle by the guanidinium thiocynatephenol-chloroform extraction method (26). RNA (15 g) was denatured and sized fractionated in duplicate by gel electrophoresis in 1.25% agarose gels containing 2.0 M formaldehyde. To assess for quality and equivalent amounts of RNA loading, the 28 and 18 S ribosomal bands were visualized by staining the gel in 0.5 g/ml ethidium bromide and were photographed by ultraviolet transillumination. The RNA was then electroblotted to Hybond N, cross-linked (Stratalinker, Stratagene Corp.), and prehybridized at 42°C for 4 h in a solution of 50% deionized formamide, 4 ϫ SSC (1 ϫ SSC ϭ 150 mM sodium chloride, 15 mM sodium citrate), 5 ϫ Denhardt's solution (50 ϫ Denhardt's solution ϭ 0.1% each of bovine serum albumin, polyvinylpyrrolidone, and Ficoll), 0.1 mg/ml yeast tRNA, 50 mM sodium phosphate (pH 7.0), 0.5 mg/ml sodium pyrophosphate, and 1% SDS. Hybridizations were carried out overnight at 42°C using the appropriate radiolabeled cDNA probe at 1-3 ϫ 10 6 cpm/ml. A BamHI-HindIII fragment containing exon III of the murine ␣B-crystallin gene was provided by J. Piatigorsky of The National Eye Institute (27). The full-length genomic clone of human hsp27 was obtained from Lee Weber at the University of Nevada, Las Vegas (28). Partial cDNAs corresponding to the myogenic regulatory factors MyoD, myogenin, myf-5, and MRF were kindly provided by E. Olson at the University of Texas Southwestern Medical Center, Dallas (24). All cDNA probes were labeled with [␣-32 P]dATP (3000 Ci/mmol) by the random priming method as described previously (29). After hybridization overnight, the membranes were rinsed for 30 min in 0.1 ϫ SSC and 0.1% SDS at room temperature followed by 10 min at 50°C and then subjected to autoradiography using Kodak SAR-5 film with intensifying screens.
Riboprobe Synthesis and in Situ Hybridization-Antisense and sense riboprobes for in situ hybridizations were generated from the exon III fragment of murine ␣B-crystallin subcloned into pCRII vector (Invitrogen, San Diego, CA) containing the SP6 and T7 RNA promoters. In vitro transcription was performed using Maxiscript kit (Ambion Inc., Austin, TX) with 35 S-UTP (400 -800 Ci/mmol) to yield the corresponding antisense and sense transcripts. To assess the spatial distribution of ␣Bcrystallin mRNA in skeletal muscle, in situ hybridization was performed as described previously (30) with modifications described by Frohman et al. (31). The tissue sections were incubated overnight at 55°C with the sense and antisense riboprobes. The slides were then treated with RNase A, washed, and coated with K.5 photographic emulsion and exposed at 4°C. The sections were developed, counterstained with hematoxylin, and examined using light-and dark-field optics. Probes were hybridized to contiguous sections and processed in parallel to facilitate comparison of gene expression in situ and fibertype specificity, as described below.
Protein Analyses-For Western blot analysis, an aliquot of powdered TA muscle (ϳ200 mg) was homogenized (Polytron) in 2 ml of cold (4°C) buffer (pH 7.4) containing 25 mM HEPES, 4 mM EDTA, 2 mM phenylmethylsulfonyl fluoride, 1 M leupeptin, 1 M pepstatin, and 0.15 M aprotinin. After homogenization, Triton X-100 was added to a final concentration of 1% before the samples were mixed by vortexing and placed on ice for 30 min. The samples were spun at 8000 rpm for 30 min at 4°C, and fractions of the supernatant were aliquoted and stored at Ϫ80°C. Protein concentration was determined by the BCA protein assay (Pierce).
For the identification of Hsp27 and ␣B-crystallin heat shock proteins, 20 g of protein extract were mixed with loading buffer (12.5 mM Tris (pH 8.0), 4.6% SDS, 20% glycerol, and 2.5% dithiothreitol), heated at 70°C for 5 min, and subjected to one-dimensional SDS-polyacrylamide gel electrophoresis. Proteins were electroblotted to NitroPure nitrocellulose transfer membranes (Micron Separations Inc.) and subsequently incubated with the following primary antibodies: 1) a rabbit polyclonal antibody to the amino-terminal 14 amino acid residues of human ␣Bcrystallin (Nova Castro) or 2) a monoclonal anti-Hsp27 (StressGen Biotechnologies Corp.). Following incubation with the appropriate secondary antibody (horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse IgG, Amersham Corp.), the proteins were visualized by the enhanced chemiluminescence detection method (Amersham Corp.).
Immunohistochemistry-Serial sections of paraffin embedded tissue used for in situ hybridization studies were processed for fast myosin immuohistochemistry to facilitate comparison of ␣B-crystallin gene expression and fiber-type composition in response to stimulation. After deparaffinizing and blocking, sections were incubated with a specific myosin antibody capable of recognizing all fast isoforms of myosin in paraffin processed tissue (monoclonal antibody MY32, Sigma). Immunoreactivity was detected by incubation with biotinylated horse antimouse IgG secondary antibody (Vector Laboratories), streptavidin alkaline phosphatase conjugate (Vector Laboratories), and, finally, visualized using an alkaline phosphatase secondary detection system (Substrate kit II, Vector Laboratories).
Statistics-Statistical analysis was performed using a one-way analysis of variance. Significant differences among the means were detected using the Dunnett's post hoc test with the level of significance set a p Ͻ 0.05.

Differential Effects of Chronic Motor Nerve Stimulation on
␣B-crystallin and Hsp27 mRNA Expression in Skeletal Muscle-Total muscle RNA content progressively increased over the 21 days of stimulation. As described previously (32,33), the stimulation-induced increase in total RNA is accounted for by proportionately equivalent increases in poly(A) and ribosomal RNA. Therefore, the mRNA data in the present study were calculated on a per mg muscle weight basis and expressed as fold changes relative to control muscle.
To determine whether the physical and metabolic stress imposed by continuous contractile activity alters the expression of ␣B-crystallin and the ancestrally related Hsp27, we performed Northern blot analysis of total RNA isolated from rabbit tibialis anterior muscle subjected to varying durations of continuous motor nerve stimulation (Fig. 1). Fig. 2 displays the average change in ␣B-crystallin (top) and Hsp27 (bottom) mRNA levels from five sets of rabbits. For comparison, the average change in total RNA yield is also plotted at each duration of stimulation. ␣B-crystallin mRNA concentration increased nearly 5-fold within the 1st day of stimulation, continued to progressively increase with longer durations of stimulation, and reached a high of 22-fold relative to unstimulated controls after 21 days of stimulation. In contrast, Hsp27 mRNA showed little specific change relative to total RNA during the 1st week of stimulation (Hsp27 increased by 4-fold, total RNA increased by 3-fold). Longer durations of stimulation (14 and 21 days) were associated with changes in Hsp27 mRNA concentration that reached statistical significance.
␣B-crystallin mRNA Is Constitutively Expressed at High Levels in the Rabbit Soleus Muscle-To further examine the relationship between ␣B-crystallin and Hsp27 expression in skeletal muscle, we determined the expression pattern of ␣Bcrystallin and Hsp27 mRNA in skeletal muscles of different fiber-type compositions (Fig. 1). In agreement with previous reports (5, 6), soleus muscle (predominantly slow twitch oxidative type I fibers) was characterized by nearly 12-and 17-fold higher levels of ␣B-crystallin expression relative to the red vastus lateralis (predominantly fast twitch oxidative type IIa fibers) and white vastus lateralis (predominantly fast twitch glycolytic type IId fibers) muscles, respectively (Fig. 3) (34). The constitutive expression of Hsp27 mRNA was also lowest in the white vastus lateralis muscle (Fig. 3). However, in contrast to ␣B-crystallin, Hsp27 mRNA levels were not different between the soleus and red vastus lateralis (Fig. 3), providing further evidence that ␣B-crystallin and Hsp27 may be under divergent, fiber type-specific regulatory mechanisms.
Continuous Motor Nerve Stimulation Induces a Marked Increase in ␣B-crystallin Protein Content-To determine whether ␣B-crystallin and Hsp27 protein levels are also increased in response to stimulation, protein extracts from control and stimulated muscles were subjected to Western blot analysis (Fig. 4). In agreement with the mRNA data, chronic motor nerve stim-ulation elicited a significant increase in ␣B-crystallin protein concentration, beginning with nearly a 2.5-fold increase after 1 day to over a 4-fold increase after 21 days of stimulation (Fig.  5). In contrast to ␣B-crystallin, Hsp27 protein content increased by less than 2-fold through 21 days of stimulation.
It is interesting to note that the increase in ␣B-crystallin protein was considerably less than the increase in ␣B-crystallin mRNA. We have also noted this relationship in cell culture  Fig. 1, as well as the total RNA/mg muscle, is presented. ␣B-crystallin and Hsp27 mRNA, quantitated as mRNA/mg muscle weight, is presented as the mean Ϯ S.E. relative to control (set to 1.0) as described under "Experimental Procedures." *Significantly different (p Ͻ 0.05) from control.
FIG. 3. Quantification of ␣B-crystallin and Hsp27 mRNA content in rabbit muscles composed of different fiber types. A summary of the densitometric analysis of the Northern data represented in the right portion of Fig. 1 is presented. ␣B-crystallin and Hsp27 mRNA content in the rabbit white vastus lateralis (predominantly fast twitch glycolytic fibers), red vastus lateralis (predominantly fast twitch oxidative and slow-twitch oxidative fibers), and soleus (predominantly slow twitch oxidative fibers) muscles is presented as the mean Ϯ S.E. *Significantly different (p Ͻ 0.05) from white vastus lateralis muscle. during differentiation of C2C12 myoblasts into myotubes in which increases in ␣B-crystallin protein are preceded by severalfold greater changes in ␣B-crystallin mRNA. 2 In the present study, it is unlikely that ␣B-crystallin protein and mRNA levels achieved steady state, as the rabbit TA muscle typically requires in excess of 50 days of continuous stimulation to "complete" the transformation process (23). Together, these data suggest that much greater levels of ␣B-crystallin mRNA may be required to generate a given change in protein, possibly due to post-translational control mechanisms. Assaying whole muscle extracts for a specific protein that is normally expressed in a fiber specific pattern may also contribute to the disparity between mRNA and protein levels typically observed with the chronic stimulation model.
␣B-crystallin mRNA Is Increased in a Fiber-specific Pattern in Response to Chronic Stimulation-To determine the induction pattern of ␣B-crystallin mRNA in response to continuous contractile activity at the level of individual muscle fibers, in situ hybridization was performed on cross sections from rabbit TA muscle subjected to varying durations of chronic motor nerve stimulation (Fig. 6A). Hybridization with the antisense riboprobe for ␣B-crystallin yielded a detectable but low level abundance of ␣B-crystallin transcript under basal conditions (Fig. 6A). This low level of expression contrasted with the high ␣B-crystallin transcript signal detected in soleus muscle under basal conditions (Fig. 6B). Within 1 day after the onset of stimulation, the signal for ␣B-crystallin mRNA in the TA muscle was clearly increased relative to unstimulated controls, particularly within a small number of fibers, and characterized by a ringlet pattern of expression (Figs. 6C and 7A, inset). The hybridization signal continued to intensify after 3 days (Fig.  6D) such that, by 21 days of continuous contractile activity, ␣B-crystallin mRNA was expressed robustly in a fiber-specific ringlet pattern (Figs. 6E and 7B, inset). Hybridizations of serial sections from the 21-day stimulated muscle with sense transcripts to ␣B-crystallin were negative (Fig. 6F).

␣B-crystallin mRNA Is Initially Expressed within Type I and a Subpopulation of Type II Fibers in Response to Chronic Motor
Nerve Stimulation-To determine whether the induction of ␣B-crystallin mRNA was fiber type-specific in response to contractile activity, we performed in situ hybridization and myosin immunohistochemistry on serial cross sections from rabbit TA muscle after 1 and 21 days of chronic motor nerve stimulation (Fig. 7). Using an antibody capable of recognizing all fast myosin isoforms in fixed, paraffin-embedded tissue and photographing under dark-field illumination, we were able to distinguish slow twitch type I fibers (dark fibers, Fig. 7, A and B) from the fast twitch type II fiber population (light fibers, Fig. 7,  A and B). After 1 day of continuous contractile activity, ␣B-  crystallin mRNA (Fig. 7A, inset) was elevated in slow twitch type I fibers (1) as well as a few fast-twitch type II fibers (22). After 21 days of stimulation, the signal for the ␣Bcrystallin transcript was evident in nearly all fibers and, similar to the results after 1 day of stimulation, was particularly intense in those fibers expressing low levels of type II myosin isoforms; i.e. fibers expressing the slow myosin isoform (dark fibers, Fig. 7B).
Transcript Levels of the MRFs Are Increased in Skeletal Muscle during Chronic Motor Nerve Stimulation-The strict transcriptional control required for expression of the ␣B-crystallin gene during myogenesis is unique among genes encoding heat shock proteins (25). To determine whether induction of ␣B-crystallin may be coordinately regulated through the muscle specific E-box element present within the 5Ј-regulatory region of the ␣B-crystallin gene, we examined the mRNA expression pattern of all four E-box-specific myogenic regulatory factors (MyoD, myogenin, myf-5, and MRF4) in response to chronic stimulation. Transcript levels of MyoD, myogenin, and myf-5 increased slightly during the 1st week of stimulation. However, these changes were minimal and of questionable significance considering the exposure time of the autorads (72 h) and the intense MyoD and myogenin signal present in Sol8 myotubes during differentiation (3 days). In contrast, MRF4 transcript levels increased within 1 day after the onset of stimulation and, similar to results with ␣B-crystallin mRNA, remained elevated through 21 days of stimulation, reaching a high of ϳ20-fold relative to unstimulated controls (Fig. 8). DISCUSSION The principal findings of the present study demonstrate that chronic low frequency motor nerve stimulation rapidly and dramatically evokes the induction of ␣B-crystallin mRNA and protein in contracting rabbit tibialis anterior skeletal muscle. In contrast, although transcript levels of the ancestrally related small Hsp27 gene were also elevated after longer durations of stimulation, Hsp27 protein increased by less than 2-fold, suggesting that distinct control mechanisms govern the expression of ␣B-crystallin and Hsp27 in contracting skeletal muscle. To this end, we provide new evidence that ␣B-crystallin may be under the regulatory influence of the myogenic regulatory factors which are coordinately induced by chronic motor nerve stimulation.
Response of ␣B-crystallin to Chronic Stimulation-Chronic low frequency motor nerve stimulation is a well characterized model of increased metabolic demand and illustrates the remarkable plasticity of skeletal muscle. Stimulation delivered over several weeks to a primarily fast twitch glycolytic muscle generates a virtual complete biochemical and morphological transformation in phenotype to that of a slow twitch oxidative muscle (23). The striking feature reported here was that the magnitude of induction of ␣B-crystallin in response to stimulation exhibited a fiber-type specific pattern of expression. We anticipated, based on the Northern blot results, that the large increase in ␣B-crystallin mRNA content observed after 1 day of stimulation likely reflected uniform expression across all fibers of the contracting TA muscle. We were surprised to find, however, using in situ hybridization and myosin immunohistochemistry to examine gene expression at the level of individual fibers, that ␣B-crystallin induction was limited to a fairly small population of fibers, primarily slow twitch type I and a small Total RNA was separated, transferred to nitrocellulose, and hybridized with DNA probes to the indicated transcripts as described under "Experimental Procedures." Ethidium bromide staining of the 28 and 18 S rRNA bands is also shown, demonstrating relative integrity and even loading of the RNA. number of fast-twitch type II fibers. 3 In agreement with these findings, we have recently found that expression of the inducible member of the major heat shock protein family, Hsp70, is also up-regulated specifically within type I and IIa myofibers within 1-3 days after the onset of stimulation. Interestingly, the fast twitch glycolytic type IIb/d fibers, which are not metabolically designed for continuous contractile activity, do not express ␣B-crystallin (this study, Fig. 7A) or Hsp70 (35). Taken together, these findings provide support for the hypothesis, originally proposed by Cadefau et al. (36), that only a subpopulation of fibers (type I and IIa) are able to meet the metabolic demand and, thus, maintain contractile activity during the first several days of stimulation. This, in turn, raises the possibility that expression of specific heat shock or stress proteins such as ␣B-crystallin and Hsp70 may identify those fibers that have initiated the adaptive response, thereby implying that recruitment during the remodeling process may proceed sequentially from type I and type IIa to type IIb/d fibers (35).
Stimulation durations extending beyond 1 day triggered a further and progressive increase of ␣B-crystallin transcript, reaching in excess of 20-fold after 21 days of stimulation, and generated an intense, ringlet pattern of expression within most of the myofibers viewed by cross section (Fig. 6E). Interestingly, this ringlet distribution pattern was especially striking in comparison to the constitutively high, but relatively homogeneous, ␣B-crystallin signal found in the rabbit soleus muscle (Fig. 6B), suggesting that the active sites of translation may primarily reside under the sarcolemma during the activityinduced remodeling of skeletal muscle.
Potential Functions of ␣B-crystallin in Skeletal Muscle-Although the precise function of ␣B-crystallin in nonlenticular tissue is not known, several biological activities have been described for crystallins, in general, and for ␣B-crystallin, in particular. ␣B-crystallin is a member of a very diverse and intriguing family of proteins (␣, ␤, ␥, and ␦) found in all vertebrate lens (3,4,13,37). Crystallins comprise approximately 90% of the total soluble protein of the lens where they exist as highly ordered protein aggregrates whose stability, in large part, accounts for the refractive properties of the lens (1). The intriguing nature of these proteins came with the surprising discovery that many crystallins are expressed in nonlenticular tissues and are, in the case of certain taxon-specific crystallins (e.g. ⑀ and ), related or identical to specific metabolic enzymes (e.g. lactate dehydrogenase and hydroxyacyl-CoA dehydrogenase, respectively) (2,4). Indeed, the recruitment of crystallin proteins to the lens is considered a later evolutionary event and has given rise to the concept of "gene sharing" advanced by Piatigorsky and Wistow (3) and Wistow (4) to describe the multiple functions that can be exhibited by the same gene product (37).
The first indication that ␣-crystallins may possess other functions outside of lens tissue came from the early observations of Ingolia and Craig (10) that ␣-crystallins share striking structural similarities with members of the small heat shock proteins of Drosophila. The biological significance of these findings was revisited when it was discovered that ␣B-crystallin is expressed in mammalian heart, skeletal muscle, lung, kidney, and brain (6,7,27). More recent work has established that ␣B-crystallin is a bona fide member of the small heat shock protein family that, like Hsp27, is induced by supraphysiological stress (13,38) and can serve as a molecular chaperone in the presence of denatured proteins (9,11,12,17,19). Consistent with their roles as molecular chaperones, our findings in the present study suggest that ␣B-crystallin and, to a lesser extent, Hsp27 may be required to support the increased protein turnover associated with isoform switching and mitochondrial biogenesis during the remodeling response of fast-twitch skeletal muscle to chronic stimulation (23). The fact that higher levels of protein turnover in slow twitch soleus muscle, which contracts almost continuously to maintain posture, correlate with ␣B-crystallin's concentration relative to other hind limb muscles (Fig. 3) (5) supports the hypothesis that ␣B-crystallin functions as a molecular chaperone within skeletal muscle.
Immunolocalization studies have revealed that ␣B-crystallin is localized to Z-lines within both slow skeletal and cardiac muscle where it is thought to interact with both actin and desmin intermediate filaments to increase stability of the Zbands (5,39). Disuse atrophy of the soleus muscle induced by hind limb suspension results in a narrowing of the Z-band, loss of ␣B-crystallin from the Z-band region, and disintegration of the myofibrillar proteins (5). Conversely, broadening of the Z-band in rabbit fast twitch TA muscle is evident within 10 days after the onset of chronic low frequency stimulation (40) and, thus, corresponds with the directional change in ␣B-crystallin observed in the present study. Whether ␣B-crystallin contributes to the stabilization of myofibrillar proteins, particularly during intermediate filament turnover triggered by the activity-induced remodeling of skeletal muscle, awaits more detailed functional analysis.
Very little is known about the post-translational modifications of ␣B-crystallin in nonlenticular tissues, although recent studies have demonstrated cAMP-dependent and independent (i.e. autokinase) activities (41). Likewise, the effect of phosphorylation on the functional properties of ␣B-crystallin is not known, although it has been suggested that phosphorylation may regulate self-aggregation or binding to other proteins (41). For example, desmin is proposed to play a role in the functional and spatial relationships between the sacromeres and the plasma membrane via the nuclear intermediate filament, lamin B (42). In view of ␣B-crystallin's association with desmin intermediate filaments (5,39) and putative biological properties, it is tempting to speculate that ␣B-crystallin may play a role either as a substrate or an activator of signal transduction pathways that link changes in contractile activity to gene expression in skeletal muscle.
Regulation of ␣B-crystallin Expression in Skeletal Muscle-There is growing evidence that ␣B-crystallin expression, particularly in skeletal muscle, may be regulated by shifts in the demand for oxidative metabolism. In support of this hypothesis, ␣B-crystallin expression during postnatal development of rats increases dramatically (Ͼ15-fold) in the soleus muscle within the same time period (ϳ10 -14 days after birth) as several oxidative enzyme markers (43), presumably reflecting the shift to oxidative metabolism required for weight bearing activity. In adult animals, ␣B-crystallin expression, outside of lens, is found exclusively in tissues with high rates of oxidative metabolism including heart, oxidative type I and IIa skeletal muscle fibers, and oxidative regions of the kidney (5)(6)(7)(8). In addition, ␣B-crystallin is abundantly expressed in the "ragged red fibers" characteristic of skeletal muscle mitochondrial myopathies, presumably as a consequence of the profound oxidative stress and compensatory proliferation of mitochondria (22). Conversely, hind limb suspension or denervation/tenotomy of the soleus muscle in adult rats decreases expression of ␣B-crystallin, consistent with the down-regulation of oxidative metabolism generated with these disuse models (5, 20, 21). Collectively, these findings raise the possibility that ␣B-crys-tallin expression, particularly in skeletal muscle, may be regulated by shifts in the demand for oxidative metabolism. The results from the present study, using a model of increased metabolic demand, establishes that ␣B-crystallin induction in skeletal muscle is a primary component of the remodeling response to chronic motor nerve stimulation and is, therefore, consistent with adaptive increase in mitochondrial-based metabolism. Whether ␣B-crystallin is indirectly required to support mitochondrial activity in tissues dependent on oxidative metabolism remains to be determined.
Several lines of evidence suggest that the induction of ␣Bcrystallin and Hsp27 in response to chronic stimulation may be modulated through common as well as distinct regulatory elements within the control regions of the two genes. Both the ␣B-crystallin and Hsp27 genes contain heat shock elements and display similar kinetics of heat inducibility in vitro (12). In addition, expression of two other heat shock proteins, Hsp70 and mitochondrial Hsp60, is up-regulated in rabbit skeletal muscle by chronic stimulation (44) and in rat skeletal muscle by exhaustive exercise (35,45), further supporting a role of the heat shock response. Importantly, because muscle temperature is increased by less than 0.5°C during chronic stimulation (35), regulation through the heat shock element is likely secondary to specific pleiotropic effects exerted within contracting myofibers, including metabolic and/or oxidative stress (46,47).
In addition to heat shock elements, the ␣B-crystallin gene also contains a muscle-specific E-box motif, the consensus binding site for the MRFs (24,25). When expressed ectopically, each of the four MRFs can activate the myogenic developmental cascade. Furthermore, in vitro studies have indicated that MyoD and myogenin can specifically bind to the E-box element located in the ␣B-crystallin promoter and fully transactivate ␣B-crystallin expression in a muscle-specific manner (25). Our study provides new evidence that contractile activity significantly induces expression of the MRFs, particularly MRF4, in a temporally similar pattern to ␣B-crystallin. Recently, there have been important questions raised concerning the function(s) of MRF4, the predominant MRF expressed in adult skeletal muscle. Gene disruption at the MRF4 locus in mice demonstrated that MRF4 is not essential for muscle development or maintenance of the skeletal muscle phenotype (48). However, whether MRF4 may play an important role in adult skeletal muscle during the adaptive response to physiological stress, aging, denervation and/or regeneration remains to be determined. The possibility that MRF4 may be critical specifically for the transcriptional regulation of ␣B-crystallin in vivo is currently under investigation.
Although their cellular function is not completely understood, heat shock proteins have been widely suggested to serve pivotal roles in the adaptation to environmental stress (49). This report is significant in that it provides new evidence that stress proteins that share similar biochemical and structural properties may be subject to different regulatory mechanisms during physiological stress in vivo, and further suggests that particular functions during cellular maintenance may be distinct from those during cellular adaptation. Our work provides support for the notion that specialized functions may exist among members of the heterogeneous but evolutionarily conserved family of heat shock proteins.