INTRODUCTION

Top Abstract Introduction Experimental procedures Results Discussion References

SKELETAL MUSCLES ADAPT very rapidly to disuse by undergoing atrophy, a major phenotypic consequence. This phenomenon is one of the many aspects of widely documented muscle plasticity. Nonpathological quantitative and qualitative changes in skeletal muscle tissue are associated with pre- and postnatal development, regeneration, hormonal environment, aging, exercise, training, and disuse (for review see Ref. 30). For most muscles, these different factors control the size of muscle fibers and their fast- or slow-twitch properties, which, in turn, determine their mechanical capacities for generating short, but intense, or weaker, but longer, contractions. Muscle disuse is a very common situation that occurs each time a muscle remains inactive for an extended period, e.g., during limb immobilization or bed rest; it also occurs in zero gravity situations encountered during spaceflights. In the case of responses to disuse, many quantitative modifications have now been described and important studies have also revealed that induced atrophy is a highly regulated process that does not completely impair muscular functional activity and generally allows reverse evolution back to a nonatrophied state (16).

Different animal models concerning muscle disuse have been studied, including unweighting by hindlimb suspension (HS), physical immobilization, and exposure to microgravity environments. In these situations, a spectacular loss of muscle mass can be obtained, especially in postural predominantly slow-twitch muscles, which normally counteract gravity (28). This atrophy is due mainly to a decrease in fiber size, correlated with a decrease in myofiber protein content but usually not a decrease in the fiber population (5). Protein synthesis is actually decreased and protein degradation increased (39). In addition, mechanical properties of atrophied soleus muscles are modified. As can be expected, they lose strength but gain some fast-twitch features, with an increase in maximum unloaded shortening velocity (V_max), higher myosin adenosinetriphosphatase activity, and faster contraction/relaxation times (15, 23). However, V_max stays significantly lower than in fast-twitch gastrocnemius control muscle (19). Adaptation thus results in fibers becoming smaller but also closer to a fast-twitch type. This is supported by shifts from various slow-twitch contractile protein isoforms toward fast-twitch isoforms (11, 15, 24). This was clearly demonstrated for myosin heavy chains (MHC), with a decrease in the proportion of slow-type I MHC (10) and a corresponding increase in fast-type II MHC isoforms (2). In rat slow-twitch soleus muscle, a fast isoform (IIx MHC) that is not normally significantly expressed in this muscle (3) appears (29, 36) with an adenosinetriphosphatase activity midway between that of the two other fast-type isoforms, IIa and IIb MHC (33). MHC expression changes have been detected at the protein and RNA levels and give rise to many more hybrid fibers coexpressing different MHC isoforms than in nonatrophied soleus muscles. These contractile protein modifications occur concomitantly with increases in the ratios of glycolytic to oxidative enzyme activities (18, 24). This reduces the oxidative capacity of skeletal muscle, with a marked increase in the glycolytic metabolism. Other gene expression modulations were found to be associated with muscle atrophy, induced by HS or physical immobilization. Cytochrome-c mRNA transcripts are downregulated (5), whereas there is upregulation of muscle-specific kinase MuSK mRNA transcripts and proteins (41), dihydropyridine receptor mRNA transcripts (21), and the fast isoform of the sarcoplasmic reticulum calcium pump mRNA transcripts and proteins (35).

However, the cellular and molecular mechanisms underlying these changes are far from being clearly understood. With the aim of developing a general analysis of molecular mechanisms triggered by muscle disuse and inducing muscle atrophy, we decided to investigate genes up- or downregulated in postural slow-twitch soleus muscles in the HS rat model. In this model, hindlimb muscles undergo hypodynamia (decreased mechanical loading) and hypokinesia (decreased motor activity), whereas the hindlimbs can move freely. The basic hypothesis was that disuse modifies or creates cascades of cellular signals and some of them could interfere with the transcription levels of certain genes. Identification of such genes will provide molecular tools for analyzing the muscular atrophy process. Different techniques are now available to characterize changes in gene profile expression. The approach used here involved quantitative differential screening of a muscle cDNA library arrayed on high-density filters based on comparison of hybridization signals given by labeled total cDNA pools synthesized from HS versus control soleus muscles. The intensity of signals for a given cDNA clone was determined by its expression level in the cDNA pool probe and therefore by the expression level of the corresponding mRNA in the original tissue (27). To our knowledge, this is the first time this approach has been used to study changes induced by muscle disuse. As a first step, an already available human muscle cDNA array was used to test the feasibility of this approach (1). Three genes were found to be upregulated in HS rat soleus and one gene was found to be downregulated. Two of these four genes were already known, i.e., muscle creatine kinase (M-CK) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH). They were the target of the present study at the RNA and encoded protein levels, with particular focus on changes related to HS duration in comparison to MHC isoform changes.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Experimental procedures Results Discussion References

Animals and tissues. After 4 days of acclimatization to the animal room, female Sprague-Dawley rats weighing 200- 220 g were randomly divided into two groups (n = 9) for each time point, HS and control weight-bearing groups. The HS groups were submitted to 2, 4, 8, 21, or 28 days of HS, using an apparatus similar to that described by Morey (25). All animals were maintained on a standard diet with water ad libitum. They were housed in a room with a regulated temperature (22°C) and 12:12-h light-dark cycle. At the end of the experiments, HS and age-matched control animals were euthanized with a lethal dose of pentobarbital sodium administered intraperitoneally. Right and left soleus muscles were excised rapidly and immediately frozen in liquid nitrogen. All samples were stored at 80°C until further analyses, which were all performed on pools of 18 soleus muscles from each rat group to eliminate any individual variations. In each pool, all entire soleus muscles were used for RNA preparations to extract sufficient quantities of RNA, and equal small middle portions from all soleus muscles were used for protein extractions.

RNA preparation. Isolation of total cellular RNA from 1- to 2-g pools of 18 soleus muscles was performed using the acid guanidinium thiocyanate-phenol-chloroform method, as described by Chomczynski and Sacchi (12). Total RNA concentrations and purities were assessed by measuring absorbance at 260 and 280 nm and by agarose gel electrophoresis. Poly(A)⁺ RNAs were purified on oligo(dT) bound to magnetic spheres (PolyAtract mRNA Isolation Systems; Promega) according to manufacturer's instructions. Their qualities were assayed by Northern blot analysis probed with -actin.

Differential screening procedures. Fifty to one hundred nanograms of 1339 PCR-amplified and both end sequenced cDNAs from a human skeletal muscle were spotted onto high-density nylon filters (8 × 12 cm; Hybond N⁺, Amersham) as described in Piétu et al. (31). Two duplicate filters were probed with ³³P-labeled single-strand cDNAs derived from HS rat soleus muscles, and two other duplicate filters were hybridized in identical conditions with equivalent probes derived from control soleus muscles. ³³P-labeled single-strand cDNA probes (1 × 10⁸ to 5 × 10⁸ dpm/µg) were prepared from 500 ng of poly(A)⁺ RNA extracted from soleus muscle pools (n = 18 muscles) according to Sambrook et al. (32), in the presence of 1 µg of a poly(dA) 80 mer (27), with SuperScriptII reverse transcriptase (GIBCO BRL). Unincorporated radioactive nucleotides were removed on a NucTrap probe purification column (Stratagene). Filters were hybridized overnight with 5 × 10⁶ cpm/ml of probe in the presence of 37% formamide at 42°C, washed 3 × 15 min with 1× SSC (1× SSC is 0.15 M NaCl and 0.015 M sodium citrate, pH 7.0), 0.1% SDS at room temperature and scanned on a PhosphorImager imaging plate system (Molecular Dynamics, Sunnyvale, CA) for quantitative analysis of hybridization signal intensities with the X-dots Reader software (Cose, Le Bourget, France). Statistical analyses of the results were performed as previously described (31), and differentially expressed cDNAs were selected. Automatic DNA sequencing of the selected clones was performed by Genome Express (Grenoble, France), and sequence data were compared with public nucleotide and protein databases using BLAST and FASTA programs from the National Center for Biotechnology Information WWW server.

Northern blots. About 40 µg of total RNA prepared from soleus muscle pools (n = 18) was size fractionated in denaturing 1% agarose gel in the presence of ethidium bromide (25 µg/ml), transferred to nylon membranes, as described by the supplier (ICN), and fixed by ultraviolet irradiation. The specific sequence cDNA probes used in this study were human M-CK partial cDNA corresponding to nucleotides 692-1086 (94% homology with rat M-CK), human GAPDH 1-kb partial cDNA corresponding to nucleotides 195-1180 (87% homology with rat GAPDH), rat IIx MHC cDNA (generous gift from Dr. Stephano Schiaffino) derived from the 3'-untranslated region, mouse MyoD complete cDNA (kindly provided by Dr. C. Pinset and Dr. D. Montarras), which was 84% homologous to rat MyoD, and a mammal-specific 18S rRNA oligonucleotide (5'-GCACGGCGACTACCATCGAA-3'). Generated cDNA probes were [-³²P]dCTP labeled by random priming (DNA labeling beads, Pharmacia). Oligonucleotides were labeled at the 5'-end with T4 polynucleotide kinase and [-³²P]ATP. Labeled probes were purified through G50 spin columns. Hybridizations were carried out overnight at 42°C in 25% formamide, 5× sodium chloride-sodium phosphate-EDTA (SSPE), 5× Denhardt's solution, 100 µg/ml salmon sperm DNA. Membranes were washed in 0.1× SSPE, 0.2% SDS, for 30 min at 50°C and exposed to phosphorus screens scanned with the PCBas program (Fujix). Membrane stripping was performed in 0.1% SDS at 80°C. mRNA expression levels were standardized according to 18S rRNA hybridization signals; the error margin for the HS-to-control signal quantification ratios could be estimated at 17% after measuring eight independent Northern blots prepared with different RNA preparations and hybridized with M-CK cDNA probes.

CK analysis. According to the method described by Brosnan et al. (8), whole tissue extracts were prepared by homogenizing soleus muscle pools (n = 18) for 20 s in a 1:10 dilution of tissue to buffer containing 26 mM Tris, 0.3 M sucrose, 1% NP-40, and 20 mM -mercaptoethanol, pH 8.4. After centrifugation, supernatants were diluted to 800 µl with extraction buffer and 1 µl was loaded onto 1% agarose gel (Ciba-Corning). Electrophoresis was performed at 120 V for 20 min at 4°C. CK activity was visualized according to the CK Isoenzyme System procedure (Ciba Corning) and the enzymatic reaction product (NADPH) was directly visualized with ultraviolet light. To quantitate CK activity, regions on the gel with detectable levels of NADPH were excised and incubated for 2 h in 1 ml of 100 mM Tris at 4°C. Absorbance was measured at 340 nm. CK activity was linear over the tissue extract concentration range used. The results were standardized by muscle extract protein concentrations, as determined by the method of Bradford (6).

Electrophoretic separation of MHC isoforms. Electrophoretic separation of MHC isoforms from soleus muscle pools (n = 18) was performed using the methods described by Talmadge and Roy (37). Four different MHC isoforms were identified by this procedure (fast-type IIa, fast-type IIx, fast-type IIb, and slow-type I MHC isoforms). Gels were silver stained as described by Bio-Rad (Silver Stain Plus; Bio-Rad) and quantification of the isomyosin bands performed with an Agfa optical densitometer using Phoretix software (Biocom). The data were expressed as a percentage of total MHC isoforms, and the SE of these measurements was estimated at 1-2% (n = 4 or 5).

Statistical analyses. The results are expressed as means with SE of at least triplicate measurements. The Student's t-test was used to evaluate differences in means, with significance set at P < 0.05.

	RESULTS

Top Abstract Introduction Experimental procedures Results Discussion References

Progress of atrophy during 4 wk of HS. The effect of muscle disuse was studied in the HS rat soleus muscle model. The progress of atrophy was evaluated by comparing the mass of soleus muscles (n = 18) after 2, 4, 8, 21, and 28 days of suspension with age-matched control muscles (Fig. 1). Control soleus mass was 120 ± 7 mg (n = 45). The loss of mass appeared very rapidly: soleus muscles weighed only 80% of controls after only 2 days of suspension. Atrophy continued to progress further, but at a slower pace, and reached ~50% of the control at 28 days. Protein and RNA contents were compared with control and HS soleus muscle wet weights at each time point (Table 1). No significant differences were found, indicating no effect of disuse on total protein and RNA contents relative to soleus muscle wet weight.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Time course of atrophy. Soleus muscles from female Sprague-Dawley rats were sampled after 0, 2, 4, 8, 21, and 28 days of hindlimb suspension (HS), which was performed according to Morey's procedure (25). Data are expressed as HS soleus muscle mass percentages of age-matched control values and represent means of 18 values at each time point with ±6% SD. Differences from control were significant for all values reported (P < 0.05). Before the beginning of HS, at day 0, the soleus mass was 120 ± 7 mg.

Progress of MHC isoforms during 4 wk of HS. The atrophy pattern was compared with the expression of MHC isoforms in the present study, because their relative contents are modified in various muscle atrophy models (36). MHC isoforms from pools of soleus muscles (n =18) were separated by SDS-PAGE according to Talmadge and Roy (37) and quantified by densitometer scanning (Fig. 2). Two isoforms were modified within the first 2 days of suspension; the I MHC percentage of total MHC decreased and the IIx MHC isoform appeared specifically in HS soleus muscle, although it was absent in the control. This decrease in I MHC and increase in IIx MHC continued at a slower pace thereafter until the 28th day of suspension. After the 8th day, we observed two further modifications, with a slight decrease in the amount of IIa MHC and the appearance of IIb MHC. The overall pattern noted with our experimental model confirmed the expected slow-to-fast-twitch fiber switch and showed that the full range of known MHC isoforms, even the fast-twitch glycolytic type IIb MHC, can be expressed in slow-twitch soleus muscle (4) if the mechanical unloading lasts >8 days. Each MHC isoform exhibited its own expression profile throughout the 28 days of HS, and the most quantitatively marked modification concerned the appearance of IIx MHC, which continuously progressed and finally reached 20% of total MHC.

View larger version (39K): [in this window] [in a new window]   Fig. 2.   Time course of myosin heavy chain (MHC) isoforms. A: soleus muscle MHC isoforms from pools of soleus muscles (n =18) were separated on silver-stained SDS-PAGE gels after 0, 2, 4, 8, 21, and 28 days of HS according to Talmadge and Roy (37). Day 0 represents the control pattern. Last lane shows the MHC isoform pattern from a control diaphragm used as a reference to identify the different MHC isoforms. B: for each lane, the percentages of the 4 different MHC isoforms relative to the total MHC isoform content were deducted from scans of the SDS-PAGE gels obtained with an Agfa optical densitometer and analyzed with Phoretix software (Biocom). After 4 or 5 independent experiments the SE was estimated at 1-2% (n = 4 or 5). Differences with day 0 were significant (P < 0.05) from day 4 for I MHC, day 21 for IIa MHC, day 2 for IIx MHC, and day 21 for IIb MHC.

Search for genes up- or downregulated in soleus muscle after 4 wk of HS. To gain a better understanding of the molecular mechanisms involved in disuse muscular atrophy, differential screening of a muscle cDNA library was carried out with ³³P-labeled total cDNA pools prepared from rat soleus muscle pools (n = 18) after 28 days of HS in comparison with the control counterpart. As a first step in this strategy, the library was a subset of 1,339 end-sequenced PCR inserts of a human muscle cDNA library, organized on one high-density filter (31). About one-half of the clones gave a detectable signal with the rat probes. Twenty cDNA clones showed a differential hybridization signal between HS rat soleus probes and control probes. HS and control soleus Northern blots were probed with these 20 cDNA clones, and finally four of them exhibited significant and reproducible differential expression between three different 28-day HS and control RNAs (Fig. 3). Three cDNA clones hybridized to upregulated mRNA transcripts in HS soleus muscle; the sequences of two of them were already known: M-CK, nucleotides 692-1086, and GAPDH, nucleotides 195-1180; the third one (b-b8e08) had a sequence that did not match any others in the available public data banks. One cDNA clone (b-10g06) was downregulated in HS soleus muscle, and sequence analysis showed that it also corresponded to a new gene. The present study focused on the two known genes (M-CK and GAPDH) with regard to their importance for muscle energetic metabolism.

View larger version (40K): [in this window] [in a new window]   Fig. 3.   Comparisons of mRNA expression between 28 days HS soleus and controls. Northern blots were prepared with 40 µg of total RNA, extracted from pools of rat soleus muscles (n = 18) and fractionated on 1% agarose gels. Human cDNA probes, listed on right, were [-32P]dCTP random-labeled and incubated overnight at 42°C in the presence of 25% formamide. A mammal-specific 18S ribosomal RNA oligonucleotide was [-32P]dATP labeled and used for lane hybridization signal standardization. Hybridization signals (n = 3) were visualized on phosphorus screens and quantified with the PCBas program (Fujix). Upregulation values were 305 ± 52% for b-b8e08, 255 ± 43% for glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and 248 ± 42% for muscle creatine kinase (M-CK). b-10g06 expression was 178 ± 30% downregulated. All of these values were significantly different from control (P < 0.05). C, control.

Comparative progression of M-CK, GAPDH, and IIx MHC mRNA upregulations during 4 wk of HS. Quantitative Northern blot analyses performed with total RNAs prepared from at least three different soleus muscle pools were carried out at 2, 4, 8, 21, and 28 days of HS to assess the relative expression of M-CK and GAPDH by comparison with controls. Because the IIx MHC protein was found above (Fig. 2) to be an early and specific marker of soleus muscle disuse consequences, its mRNA expression was also measured on the same Northern blots (Fig. 4). Hybridization signals were quantified on phosphorus screens and standardized according to the 18S ribosomal RNA signal. The three mRNA expressions measured showed upregulation as early as the 2nd day of HS. M-CK mRNA showed the earliest upregulation, which reached 173 ± 40% (70% of its peak time point) after 2 days of HS treatment, although, at the same time, loss of mass was only about one-half its peak time point value. M-CK upregulation peaked at ~4 days of HS and then remained stable until the 28th day. This maximum value (referred as 100% in Fig. 4) represented an increase in mRNA expression of 248 ± 42% by comparison with age-matched controls. GAPDH mRNA upregulation showed a comparable pattern, but only peaked at 8 days, which represented 255 ± 43% upregulation (100% of the peak time point). Finally, IIx MHC mRNAs were undetectable before HS, and their expression increased the slowest to reach a maximum around day 21 of HS. In fact this slower increase of IIx MHC mRNAs represented a very important quantitative increase, because the amount detected at day 28 of HS represented an increase of 790% (100% of the peak time point) of the amount detected at day 2 of HS. The progression of upregulation of M-CK mRNAs and, to a lesser extent, GAPDH mRNAs, preceded soleus atrophy measured in terms of a loss of mass. In contrast, there was a delay in the appearance of IIx MHC mRNAs.

View larger version (57K): [in this window] [in a new window]   Fig. 4.   Time course of M-CK, GAPDH, and IIx MHC mRNA upregulation compared with atrophy progression. A: Northern blots were prepared with 40 µg of total RNA extracted from 0, 2, 4, 8, 21, and 28 day HS rat and age-matched control soleus muscles (n = 18 soleus muscles for each time point). They were hybridized overnight at 42°C in the presence of 25% formamide, with the [-32P]dCTP random-labeled human M-CK, GAPDH, and IIx MHC cDNA clones. Lane hybridization signals were standardized with 18S ribosomal RNA. B: for each probe, hybridization signals (n = 3) were quantified on phosphorus screens and the HS-to-control standardized ratios obtained at each time point were represented as percentages of the peak time point. These mRNA upregulation measurements were compared with the atrophy values expressed as percentages of the maximum loss of soleus muscle weight. For each time point, values were significantly different from control (P < 0.05).

Comparative progression of MM-CK and IIx MHC protein upregulation during 4 wk of HS. To test a potential functional significance of the earliest mRNA upregulation, i.e., M-CK mRNA, we assayed the amounts of CK protein isoforms during 4 wk of HS (Fig. 5A) in comparison with the appearance of the IIx MHC protein isoform (Fig. 2). CK is assembled in three dimeric and one octameric structure (13, 34): MM (muscle type), BB (brain type), MB (muscle/brain hybrid type), and mitochondrial type, respectively. CK activity was detected in muscle extracts analyzed on agarose gels (8). As expected, only the MM-CK isoprotein was seen in muscle extracts. Figure 5B shows the comparative variations in MM-CK and IIx MHC proteins expressed as percentages of the peak time points. Three independent experiments were quantified for each time point. The upregulation noted at the mRNA level was basically found at both protein levels but with interesting differences. MM-CK protein upregulation peaked at day 21 with a progression pattern identical to that of soleus loss of mass, with a fast increase until day 4 followed by a slower pattern. This therefore represented a delayed response compared with the early upregulation of M-CK mRNA. This maximum upregulation of MM-CK protein (100% of the peak time point) was quantified as a 295 ± 22% (n = 3) increase compared with the control value. IIx MHC protein upregulation followed with only a slight delay in mRNA upregulation, thus occurring later than the loss of mass. However, the amount of this protein increased by 520% between day 2 and day 28 of HS, this latter time point representing 100% of the peak time point.

View larger version (38K): [in this window] [in a new window]   Fig. 5.   Time course of MM-CK and IIx MHC protein upregulation compared with atrophy progression. A: MM-CK protein was assayed on agarose gels according to its enzymatic activity (8) in protein extracts prepared from 18 soleus muscles after 0, 2, 4, 8, 21, and 28 days of HS and standardized for the total amount of protein estimated according to the Bradford assay (6). Each lane contained 0.4-0.8 ng of extract protein and band intensity values were standardized according to the exact protein concentrations. Standard CK isoforms used represent human CK isoforms, which do not exactly comigrate with the rat CK isoforms. B: MM-CK amounts were obtained by densitometer scanning of MM-CK gels (n = 3) and represented for each time point as percentages of the peak time point. MM-CK data are compared with IIx MHC progression expressed as percentages of its maximum value shown in Fig. 2 and to the atrophy values expressed as percentages of the maximum loss of soleus muscle weight. Values of all time points were significantly different from control (P < 0.05).

M-CK upregulation is not correlated with a modification in MyoD expression. M-CK transcription is regulated by several transcription factors (26). One of them is the myogenic factor MyoD that binds to the E box present in the M-CK gene promoter. With regard to this role of MyoD and its key role in myogenesis, we wondered whether or not its expression would be modified during the 4 wk of HS. The same Northern blots used to analyze the expression of M-CK, GAPDH, and IIx MHC were probed with a mouse MyoD probe (Fig. 6). No differential expression was found at any time tested, indicating that upregulation of M-CK was controlled by other factors than a MyoD transcriptional modification between days 2 and 28 of HS.

View larger version (34K): [in this window] [in a new window]   Fig. 6.   Kinetic comparison of MyoD mRNA expression in HS soleus muscles and controls. Northern blots presented in Fig. 4 were stripped and rehybridized with a [-32P]dCTP labeled mouse MyoD probe. No significant difference could be detected in HS soleus muscle at any time point regarding 18S rRNA hybridization standardization.

	DISCUSSION

Top Abstract Introduction Experimental procedures Results Discussion References

Upregulation of M-CK and GAPDH by HS. Altered expression of very few genes other than myosin isoforms has been found to date, i.e., dihydropiridine receptor (21) and sarcoplasmic reticulum calcium pump (35), in the muscle disuse model involving soleus muscles from HS rats. In this context, we started a more systematic search of other up- or downregulated genes by quantitative differential screening of a muscle cDNA array to investigate transcriptional gene modulations associated with skeletal muscle adaptations to disuse. The results presented here focused on two upregulated genes, i.e., M-CK and GAPDH. Variations in the upregulations of these two mRNAs were compared with the progression of soleus muscle atrophy and to IIx MHC mRNA expression, because MHC and especially the IIx MHC isoform are major biochemical markers previously described for modifications induced by mechanical unloading. This comparison was performed at different time points over a 4-wk period of hindlimb unweighting. Upregulation of the two genes encoding metabolic proteins (MM-CK and GAPDH) started clearly before upregulation of the gene encoding the contractile IIx MHC protein. M-CK upregulation was found to be the earliest marker at both the mRNA and protein levels. Moreover, MM-CK protein upregulation exactly paralleled the progression of atrophy, therefore constituting an excellent marker for skeletal muscle adaptations to disuse. These data demonstrate that muscle loading necessary for weight support can modulate the expression of M-CK and GAPDH genes.

cDNA array differential screening strategy. The present findings for four genes whose expression was modified in HS soleus muscles confirmed our working hypothesis concerning variations in gene expression profiles associated with skeletal muscle disuse and confirmed the utility of identifying such genes to investigate molecular aspects of mechanisms induced by muscle disuse leading to muscle atrophy. This work demonstrates the feasibility of the cDNA array quantitative differential screening approach for finding up- or downregulated genes during the muscular atrophy process. Note that finally only four cDNAs with differential expression could be measured by Northern blot analysis. The main factor to explain this low yield is that we screened human muscle cDNAs with rat muscle cDNAs probes. This explains why only about one-half of the arrayed cDNAs gave a hybridization signal, although sequences occurring in very low abundance were also expected to give no detectable hybridization signals. In addition, because heterologous cDNA probes were used, filter hybridizations and washes could not be carried out under the highest stringency conditions, which is a major source of false-positive signals (20). Another technical limitation was that the screened cDNA library contained only 1,339 cDNA clones without systematic equalization of their expression frequency in muscle. However, abundant clones such as myosin cDNAs had been removed from this collection to avoid redundancy, which explained why no MHC isoform cDNA was found in the screening. Therefore the low abundant expressed cDNAs were under represented in the array. Further optimizations of the differential screening technology should include the use of rat (instead of human) muscle cDNA arrays, preferably after normalization and subtraction, to gain access to a more exhaustive cDNA population that includes rare sequences.

Importance for slow-to-fast-twitch transition. Our HS model was validated by the time course progression of the different MHC isoform proportions, which closely agreed with previous results (4, 16, 38), showing an increase in the percentages of fast MHC isoforms. In addition, this was documented in the present study with the four MHC isoforms at several time points throughout 4 wk of HS. The early decrease in the slow-twitch I MHC was compensated by neoexpression of the fast-twitch intermediate IIx MHC isoform. After 8 days of HS, MHC isoforms expressed in most glycolytic muscle fibers were amplified by a decrease in fast-twitch IIa MHC, which are expressed in fast oxidative-glycolytic fibers, and neoexpression of fast-twitch IIb MHC, which are expressed in pure fast glycolytic fibers, as already reported by Fauteck and Kandarian (17). MHC isoprotein content determines the mechanical characteristics of muscle contraction. The present results indicated that the response of an antigravity muscle such as the soleus to unloading by hindlimb unweighting was a continuously changing process that was still not stabilized after 4 wk, as indicated by the nonplateau curves for the MHC isoform percentages (Fig. 2).

Biochemical modifications occurring in unweighted soleus muscles have been mainly documented for MHC protein isoforms, which determine the fast- or slow-twitch phenotype by their adenosinetriphosphatase activities. However, it has not yet been shown whether the entire muscle fiber contractile protein content, which can also be expressed with slow or fast isoforms, is actually involved in such phenotypic changes. In addition, this slow-to-fast shift is limited; soleus muscles gain some fast-twitch-type mechanical features (24), but they never become pure fast-twitch-type muscles, which is related to the heterogeneity noted in the muscle fiber population. Upregulation of M-CK and GAPDH provides new information concerning the extent of transformation of soleus muscles to a faster type. These metabolic enzymes do not have slow and fast isoforms but they are highly expressed in muscles possessing high glycolytic potentials and low resistance to fatigue (42). Upregulation of their mRNA transcripts and upregulation of MM-CK proteins show that they are components involved in fast-type transformation of soleus muscles, which is in agreement with the key role of these two enzymes in muscle energetic metabolism. Although many slow fibers are induced to coexpress fast and slow myosin isoforms (28, 38), this is not the case for all of them. No histological results were obtained in the present study, and we thus cannot correlate the upregulation of M-CK and GAPDH with different types of atrophied soleus muscle fibers. However, it is rational to speculate that hybrid fibers coexpressing both slow and fast myosins represent the main site where these upregulations preferentially take place. In this case, such upregulations in these fibers would be much higher than the average upregulations measured with RNA extracted from all soleus fibers, including the remaining slow-twitch type. In addition, regarding the time course described in our results, M-CK then GAPDH were part of the first genes upregulated in the transformation process toward a faster phenotype. The fast-twitch-type features of crucial metabolic enzymes such as M-CK and GAPDH were turned on before contractile proteins such as IIx MHC, suggesting that metabolic adaptations precede structural adaptations. This very early adaptation of metabolic components has also been observed with the dramatic upregulation of the sarcoplasmic reticulum calcium pump (35) and with upregulation of the dihydropyridine receptor (21).

Origin of M-CK and GAPDH upregulation: importance for muscle disuse. The cause of the observed M-CK and GAPDH mRNA transcript upregulation could have a pre- or posttranscriptional origin. We investigated the involvement of one mRNA transcription factor, i.e., MyoD, but the results were negative. This observation was consistent with a recent study (14) based on hindlimb immobilization, an HS close model for muscle disuse. No alterations were found for MyoD and myogenin mRNAs, which both encode components with a M-CK E box binding activity (7). However, our results do not rule out possible modifications in the binding activities of MyoD protein, as demonstrated for the myogenin-Jun-D complex on the M-CK enhancer, in a muscle wasting situation (9). The main arguments in favor of pretranscriptional regulation are 1) this is the most common mechanism of transcript up- or downregulation and 2) Tsika et al. (40) have recently delineated 5'-regulatory regions responsible for downregulation of the M-CK gene induced by mechanical overload in fast-twitch rat plantaris muscle, which closely parallels the mechanical unloading situation studied in the present work. It would be particularly useful to study the roles of these elements and their binding transacting factors in our hindlimb unweighting model, to determine the signals controlling the regulation of M-CK by unloading. The same laboratory very recently delimited a 600-bp region in the -MHC promoter containing sequences sufficient to direct decreased transcription of -MHC in HS mouse soleus muscles (22). This other indication of mechanical loading as a signal for gene transcriptional regulation opens the way for new challenging investigations that could explain how these controls are coordinated to induce phenotypical modifications in skeletal muscles. It will be of particular interest to investigate these regulatory M-CK gene cis-acting elements that may be directly responsive to mechanical load or targets for proto-oncogenes such as Jun-D, which are known to be the very first signals modified in many cellular stresses. Their existence suggests that, in addition to being a marker of events induced by muscle disuse, M-CK is one of the upstream causes of these events. In fact, the very early M-CK upregulation in our HS model and its key role in energy maintenance supply in high energy demanding skeletal muscle tissues indicate that M-CK is a crucial component for determining the muscle phenotype.

Perspectives