Address correspondence and reprint requests to Dr. B. J. Jasmin at Department of Cellular and Molecular Medicine, Faculty of Medicine, University of Ottawa, 451 Smyth Road, Ottawa, Ontario, Canada K1H 8M5. E-mail: firstname.lastname@example.org
Abstract: The molecular mechanisms underlying the activity-linked plasticity of acetylcholinesterase (AChE) mRNA levels in mammalian skeletal muscle have yet to be established. Here, we demonstrate that denervation of adult muscle induces a dramatic (up to 90%) and rapid (within 24 h) decrease in the abundance of AChE mRNAs. By contrast, denervation of 14-day-old rats leads to a significantly less pronounced reduction (50% of control) in the expression of AChE mRNAs. Assessment of the transcriptional activity of the AChE gene reveals that it remains essentially unchanged in adult denervated muscles, whereas it displays an approximately two- to three-fold increase (p < 0.05) in denervated muscles from 2- to 14-day-old rats. In addition, we observed a higher rate of degradation of in vitro transcribed AChE mRNAs upon incubation with protein extracts from denervated muscles. Finally, UV-crosslinking experiments reveal that denervation increases the abundance of RNA-protein interactions in the 3′ untranslated region of AChE transcripts. Taken together, these data suggest that the abundance of AChE transcripts in mature muscles is controlled primarily via posttranscriptional regulatory mechanisms, whereas in neo- and postnatal muscles, both transcriptional and posttranscriptional regulation appears critical in dictating AChE mRNA levels. Accordingly, the activity-linked transcriptional regulation of the AChE gene appears to demonstrate a high level of plasticity during muscle development when maturation of the neuromuscular junctions is still occurring.
Acetylcholinesterase (AChE; EC 22.214.171.124) plays an essential role at cholinergic synapses of both peripheral and central nervous systems because it rapidly hydrolyzes acetylcholine released from nerve terminals. In vertebrates, several AChE catalytic subunits differing at their C-termini are produced through alternative splicing of a single gene. Distinct posttranslational modifications of these catalytic subunits, in turn, generate multiple molecular forms of AChE that are expressed in a variety of tissues and subcellular locations (for reviews, see Massoulié et al., 1993; Taylor and Radic, 1994). In skeletal muscle fibers, for instance, asymmetric forms of AChE and hydrophobic-tailed tetramers accumulate within the synaptic basal lamina (McMahan et al., 1978) and the perijunctional compartment (Gisiger and Stephens, 1988), respectively, where their expression is known to be markedly influenced by the levels of neuronal activation.
Despite the wealth of information available on the plasticity of AChE molecular forms confronted with altered levels of neuromuscular activation, our knowledge of the cellular and molecular mechanisms involved in the activity-linked regulation of AChE in muscle is still rudimentary. Several studies have begun recently to explore the molecular basis underlying the activity-linked regulation of AChE in skeletal muscle. Results of these studies have shown, for example, that muscle inactivation induced by either surgical denervation or tetrodotoxin application onto the motor nerve leads to profound reductions in the levels of AChE mRNA in mammalian skeletal muscle fibers (Cresnar et al., 1994; Michel et al., 1994; Sketelj et al., 1998). Conversely, increased neuromuscular activation achieved by exercise training or compensatory overload increases significantly the abundance of AChE transcripts (Sveistrup et al., 1995). Although these studies clearly demonstrate that nerve-evoked electrical activity constitutes a key regulator of AChE expression in skeletal muscle, the molecular mechanisms underlying this activity-linked regulation of AChE mRNAs have yet to be fully elucidated. In this context, several levels of regulation, including transcriptional as well as posttranscriptional control, may be considered. Results from several laboratories have shown, for instance, that in denervated muscle the increased expression of mRNAs encoding the various acetylcholine receptor (AChR) subunits is caused by a transcriptional activation of their respective genes (for reviews, see Hall and Sanes, 1993; Duclert and Changeux, 1995; Burden, 1998; Sanes and Lichtman, 1999). In contrast, posttranscriptional regulatory mechanisms appear to play a significant role in the regulation of AChE during differentiation of myogenic (Fuentes and Taylor, 1993; Luo et al., 1994), neuronal (Coleman and Taylor, 1996), and hematopoietic (Chan et al., 1998) cells grown in culture.
In the present study, therefore, we have begun to define the molecular mechanisms underlying the activity-linked regulation of AChE in skeletal muscle fibers. Specifically, we have assessed the contribution of transcriptional versus posttranscriptional mechanisms in the regulation of AChE mRNA levels in denervated postnatal and adult skeletal muscles by using several complementary approaches. A preliminary account of this work has appeared previously (Jasmin et al., 1998).
MATERIALS AND METHODS
Animal care and surgery
Pregnant Sprague-Dawley rats, female Sprague-Dawley rats weighing ∼250 g, and male and female rat pups (1.5-2 weeks of age) were obtained from Charles River Laboratories (St. Constant, Québec, Canada). Care and treatment of the animals were in accordance with the guidelines established by the Canadian Council on Animal Care. Hindlimb muscles were denervated by cutting and removing a short segment of the sciatic nerve in the midthigh region while the animals were anesthetized with methoxyflurane. One to 7 days later, animals were anesthetized with sodium pentobarbital (35 mg/kg i.p.) and their hindlimb muscles were dissected, excised, and rapidly frozen in liquid nitrogen. All surgical procedures were performed under aseptic conditions.
RNA extraction and RT-PCR
Total RNA was isolated from rat hindlimb muscles using 1 ml of Trizol (GibcoBRL, Burlington, Ontario, Canada) per 100 mg of tissue. Muscles were homogenized with a Polytron set at maximum speed for 2 × 15 s. Following addition of chloroform, the solution was mixed vigorously and spun at 12,000 g for 15 min at 4°C. The aqueous layer was then transferred to a fresh tube and an appropriate amount of isopropanol was added. For RNA precipitation, the samples were spun and the resulting pellets washed with 70% ethanol. Pellets were briefly air-dried and then resuspended in RNase-free water. All samples were stored at -80°C until use.
For RT-PCR analysis, all RNA samples were quantified using a Pharmacia GeneQuant II RNA/DNA spectrophotometer and the final concentration was adjusted to 80 ng/μl. Two microliters of each RNA sample was reverse-transcribed at 42°C for 45 min followed by 5 min at 99°C as previously described in detail elsewhere (Jasmin et al., 1993; Michel et al., 1994; Boudreau-Larivière et al., 1996). Negative controls consisted of the same reverse transcription mixture in which sample RNA was replaced by 2 μl of RNase-free water.
cDNAs encoding AChE, AChR α-subunit, and rRNA were amplified using PCR as described in detail elsewhere (Jasmin et al., 1993; Michel et al., 1994; Sveistrup et al., 1995; Boudreau-Larivière et al., 1996). Primers for AChE (5′-CTGGGGT-GCGGATCGGTGTACCCC; 3′-TCACAGGTCTGAGCAGC-GTTCCTG), AChR α-subunit (5′-GACTATGGAGGAGTGA-AAAA; 3′-TAGAGGTGGAAGGGATCAGC), and S12 rRNA (5′-GGAAGGCATAGCTGCTGG; 3′-CCTCGATGACATC-CTTGG) (internal control for RT-PCR experiments) were synthesized on the basis of available sequences (Boulter et al., 1985; Forster et al., 1993; Legay et al., 1993a,b). Cycle parameters for AChE included denaturation for 1 min at 94°C, followed by primer annealing and extension at 70°C for 3 min. For the AChR α-subunit, primer annealing was carried out for 1 min at 60°C and extension was for 1 min at 72°C. Primer annealing and extension for rRNA were 54°C for 1 min and 72°C for 2 min, respectively. In each experiment, the last cycle was followed by a 10-min elongation step at 72°C. The PCR products (AChE, 670 bp; AChR α-subunit, 576 bp; S12 rRNA 368 bp) were visualized on 1% ethidium bromide-stained agarose gels. Quantification of the PCR products was performed by separating PCR products in agarose gels containing the fluorescent dye VistraGreen (Amersham, Arlington Heights, IL, U.S.A.), and the labeling intensity of the PCR product, which is linearly related to the amount of DNA, was quantified using a Storm PhosphorImager and analyzed with the accompanying ImageQuant software program (Molecular Dynamics, Inc., Sunnyvale, CA, U.S.A.). All values obtained for AChE and the AChR α-subunit were corrected according to their corresponding level of rRNA present in the sample.
All RT-PCR measurements aimed at determining the relative abundance of AChE and AChR α-subunit mRNAs, as well as rRNA, in control and denervated rat muscles were performed during the linear phase of amplification (see Jasmin et al., 1993; Michel et al., 1994; Hubatsch and Jasmin, 1997). The cycle numbers were typically 37 for AChE, 35 for AChR α-subunit, and 28 for rRNA. RT-PCR conditions (primer concentrations, input RNA, choice of RT primer, cycling conditions) were initially optimized and were identical for all samples. Appropriate precautions [e.g., dedicated areas for sample preparation and analysis, use of aerosol-barrier tips (Diamed; Mississauga, Ontario, Canada)] were taken to avoid contamination and RNA degradation. Control and denervated samples, as well as negative controls (see above), were prepared using common master mixes containing the same RT and PCR reagents and were run in parallel. In all experiments, PCR products were never detected for negative controls.
Northern blot and nuclear run-on analyses
Poly(A)+ RNA extracted using the Oligotex kit (Qiagen, Chatsworth, CA, U.S.A.) was used for analysis of AChE mRNA levels by northern blotting. Samples (3 μg) obtained from control and denervated tibialis anterior (TA) muscles were first size-fractionated on 1% agarose/0.6 M formaldehyde gels and subsequently transferred onto Genescreen nylon membranes (DuPont, Wilmington, DE, U.S.A.). For hybridization, AChE (879-1, 722 bp) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH; full-length) cDNA fragments were labeled with [α-32P]dCTP using the random prime labeling method.
Nuclear run-on assays were performed using a modified version of a procedure described elsewhere (Ray et al. 1995). Nuclei were obtained by homogenizing ∼1 g of frozen control or denervated muscle in 10 volumes of lysis buffer made of 0.3 M sucrose, 60 mM KCl, 15 mM NaCl, 15 mM HEPES, pH 7.5, 2 mM EDTA, 0.5 mM EGTA, 0.15 mM spermine, 0.5 mM spermidine, 10 mM dithiothreitol, and 1 mM phenylmethylsulfonyl fluoride. Following centrifugation, pellets were resuspended in 1 ml of lysis buffer containing 0.05% Nonidet P-40 for further homogenization. Nuclei were then sedimented at 500 g and resuspended in transcription buffer containing 0.6 M (NH4)2SO4, 0.4 M Tris, pH 7.9, 0.2 M MgCl2, 0.2 M MnCl2, 1 M NaCl, 100 mM EDTA, pH 8, 0.02 M phenylmethylsulfonyl fluoride, 1 mM dithiothreitol, 10 mM creatine phosphate, 1 mM each of GTP, ATP, and CTP, 5% glycerol, 50 U of RNase inhibitor (Promega, Madison, WI, U.S.A.), and 200 μCi of [α-32P]UTP to a final volume of 200 μl. RNA was transcribed at 28°C for 30 min. Following RQ1 DNase (Promega) treatment, labeled RNA was isolated using Trizol and hybridized for 48 h with 10 μg each of linearized rat AChE (2 kb), AChR α-subunit (1.8 kb), and β-actin (2 kb) cDNAs immobilized on Genescreen Plus nylon membrane (DuPont). Following hybridization, membranes were washed thoroughly (1× saline-sodium citrate, 0.1% sodium dodecyl sulfate) at 42°C and exposed for autoradiography at -80°C for 2-5 days with intensifying screens. The intensity of the signals was quantified with a Storm PhosphorImager (Molecular Dynamics). The signals corresponding to AChE and AChR α-subunit were standardized relative to the β-actin signal.
Injection of a rat AChE promoter-reporter gene construct in muscle
A 1.9-kb fragment termed NRAP (N-box containing rat AChE promoter; see Chan et al., 1999) was subcloned into a LacZ reporter vector (Gundersen et al., 1993). This rat AChE promoter fragment is located ∼600 bp from the translational start site of the AChE gene and contains 807 bp upstream of the initiator element (see Chan et al., 1999). In recent studies, we have shown that this rat promoter fragment directs the synapse-specific expression of a reporter gene along multinucleated rat muscle fibers (Chan et al., 1999). Plasmid DNA was prepared using the Qiagen mega-prep procedure, and final pellets were resuspended in sterile phosphate-buffered saline to a final concentration of 2 μg/μl. In vivo gene transfer into control and denervated TA muscles of rat was performed as described by Gramolini et al. (1997, 1998) using an experimental framework detailed elsewhere (Walke et al., 1996). In brief, control and denervated TA muscles from either postnatal or adult rats were injected with 50 μg of the rat AChE promoter-reporter gene construct. Seven days after injection, TA muscles were excised and frozen in liquid nitrogen. β-Galactosidase activity was assayed in control and denervated neonatal muscles using a luminescent β-galactosidase detection kit (Clontech, Palo Alto, CA, U.S.A.). Activity of β-galactosidase was normalized according to a coinjected chloramphenicol acetyltransferase (CAT) plasmid (Promega) under the control of the SV40 promoter. As β-galactosidase is posttranslationally regulated in adult skeletal muscles (Gundersen et al., 1993), LacZ and CAT mRNA levels were determined by RT-PCR in mature control and denervated TA muscles.
UV-crosslinking and mRNA stability assays
The 3′ untranslated region (UTR) of the 2.4-kb AChE mRNA was isolated by using the 3′ RACE (rapid amplification of cDNA complementary ends) system (GibcoBRL). The primer specific for the AChE cDNA was located in exon 6 (alternative exon T) (5′-3′ ATAGCAAGCAGGAACGCT-GCTCA). The resultant PCR product was subcloned into pCR2.1-TOPO vector (Invitrogen), and its identity was confirmed by sequencing. In vitro RNA-binding studies were performed as previously described (Bag and Wu, 1996). In brief, 32P-labeled sense RNA was generated using the T7 polymerase. Labeled RNA (∼5 × 104 cpm) was then incubated with 100 μg of total protein extracts isolated from control and denervated adult rat hindlimb muscles. Muscle proteins were extracted using a homogenization buffer comprised of 0.3 M sucrose, 60 mM NaCl, 15 mM Tris, pH 8.0, 10 mM EDTA, 0.1 mMβ-mercaptoethanol, 0.01 mM phenylmethylsulfonyl fluoride, 0.01 mM benzamidine, 1 μg of leupeptin, 10 μg of pepstatin A, and 1 μg of aprotinin (Bag and Wu, 1996). RNA protein complexes were crosslinked by irradiation for 30 min at 4°C at 1 cm from UV lamp (3,000 μW/cm2) in the presence of RNase inhibitors. Free RNA was then digested with RNase A and T1, and the UV-crosslinked products were analyzed by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and autoradiography. In these experiments, in vitro transcribed antisense cRNAs served as controls.
To determine the stability of AChE transcript, in vitro transcribed 32P-labeled sense RNA was incubated with extracts isolated from control and denervated hindlimb muscle at 37°C. Aliquots of the samples were taken at different time intervals thereafter, and the RNA was extracted using TriPure Reagents (Boehringer Mannheim). The amount of intact radioactive RNA was determined by agarose gel electrophoresis.
Denervation affects AChE mRNA levels differentially in adult versus postnatal skeletal muscles
In a previous study, we showed that AChE mRNA levels drop by ∼10-fold in adult muscle denervated for 10 days (Michel et al., 1994; see also Cresnar et al., 1994). In the present investigation, we initially determined the time course of this response to inactivation by analyzing AChE mRNA levels in muscles denervated from 1 to 7 days. Our quantitative RT-PCR data performed within the linear range of amplification (see Jasmin et al., 1993; Michel et al., 1994) indicated that AChE mRNA levels in adult muscle decreased markedly and rapidly following denervation. As shown in Fig. 1B, for example, AChE transcript levels were reduced by ∼90% (p < 0.05) in adult muscles denervated for 2 days. Northern blot analysis performed with control and denervated adult muscles revealed that the two predominant mRNA species encoding the AChE catalytic T subunit expressed in muscle (Legay et al., 1993a,b) decreased to a similar extent over the experimental time course following denervation (see Fig. 2). In agreement with previous findings, the levels of mRNA encoding the AChR α-subunit increased significantly (p < 0.05) following denervation, reaching >20-fold in adult muscles denervated for 7 days.
Next, we examined the response of the AChE gene to denervation in postnatal muscles. For these studies, hindlimb muscles from 2-week-old rats were denervated, and levels of AChE and AChR α-subunit mRNAs were determined at 1, 2, 4, and 7 days following denervation. The pattern of increased AChR α-subunit mRNA expression in response to denervation was identical (p > 0.05) between adult and postnatal rats (data not shown). However, although AChE transcript levels in postnatal muscles were reduced following denervation, they displayed nonetheless a significantly smaller decrease (p < 0.05) than that seen in adult denervated muscles (Fig. 1B). For example, the abundance of AChE mRNAs in denervated muscles from postnatal rats was near control levels after 1 day of denervation (i.e., decreased by only 14%), whereas it was reduced by >70% in adult denervated muscles. In addition, the effects of denervation on AChE transcript levels in postnatal rats were significantly (p < 0.05) less pronounced throughout the time course of the experiment because they remained at ∼50% of control values.
Molecular mechanisms underlying the changes in AChE mRNA levels in adult versus postnatal denervated muscles
To determine whether alterations in the transcriptional activity of the AChE gene could account for the observed changes in mRNA levels following denervation, we first performed run-on assays with nuclei from control and denervated muscles excised from adult and postnatal rats. In agreement with previous studies that have shown enhanced transcription of AChR subunit genes in denervated muscles from postnatal (Fontaine and Changeux, 1989; Merlie and Kornhauser, 1989; Tsay and Schmidt, 1989; Gundersen et al., 1993; Bessereau et al., 1994; Walke et al., 1996) and adult (Gundersen et al., 1993; Merlie et al., 1994) animals, we also observed an increase in the rate of transcription of the α-subunit gene in denervated muscles (for example, see Fig. 4). In these experiments, transcription of the AChE gene was also increased two- to threefold (p < 0.05) in 2-day denervated postnatal muscles (Figs. 3A and 4). This increase was transient, however, because the rate of transcription of the AChE gene returned toward control levels by day 4. It is interesting that the pattern of expression of the AChE gene in these denervated postnatal muscles resembled that seen for the AChR α-subunit gene (see Fig. 4; see also Tsay and Schmidt, 1989). Additional experiments performed with hindlimb muscles obtained from neonatal rat pups (2-3-day-old) denervated for 3 days showed a similar two- to threefold increase in the transcriptional activity of the AChE gene in response to denervation. Conversely, no alteration in the rate of transcription of the AChE gene was observed in denervated muscles from adult animals (Fig. 3B; range of 95-129% of control for 1- to 7-day denervated muscle) despite, as expected, an increase in the transcriptional activity of the MyoD gene (see Huang et al., 1993). In these experiments, we also noted that the transcriptional activity of the AChE and β-actin genes was higher in muscles from postnatal versus adult rats (see Fig. 3). These data are entirely coherent with the observations that AChE mRNA levels are three- to fourfold higher in developing muscles compared with adult (data not shown), and with the previous results showing that transcription of the β-actin gene indeed decreases during postnatal muscle development (Cox and Buckingham, 1992).
In separate experiments, we further assessed the expression of the rat AChE gene by directly injecting into control and denervated muscles a rat AChE promoter-LacZ reporter gene construct termed NRAP (Chan et al., 1999). This approach has been used recently to examine the response of various AChR subunit genes to muscle denervation (Bessereau et al., 1994, 1998; Walke et al., 1996). In agreement with our nuclear run-on data (Figs. 3 and 4), the activity of β-galactosidase was increased by approximately threefold (p < 0.05) in denervated postnatal muscles (Fig. 5). By contrast, no significant change (control, 1.39 ± 0.16; denervated, 1.22 ± 0.08; means ± SE, arbitrary units; p > 0.05) in the expression of the reporter gene was detected in adult muscle.
Posttranscriptional mechanisms underlying the changes in AChE mRNA levels in denervated muscles
To ascertain that posttranscriptional mechanisms are indeed involved in mediating the activity-dependent regulation of AChE, we performed mRNA stability assays using in vitro transcribed AChE mRNA and protein extracts isolated from adult control and 7-day denervated muscles. In comparison with control samples, incubation of AChE transcripts with extracts from denervated muscles sharply increased the rate of degradation of the transcripts. As shown in Fig. 6A, a marked reduction in transcript levels was observed with the denervated samples following 8 h of incubation, whereas the abundance of transcripts incubated with extracts from control muscles remained largely unaffected over the same time period. As expected on the basis of our recent study showing that the levels of utrophin mRNAs remain largely unaffected by denervation (Jasmin et al., 1995), we observed no difference in the degradation pattern of in vitro transcribed utrophin RNA incubated with protein extracts isolated from control versus denervated muscles (data not shown).
Based on these results, we expected that the pattern of RNA-protein interactions would be altered in response to muscle denervation. To test this possibility, we performed UV-crosslinking assays by incubating protein extracts from adult control or denervated muscles with in vitro transcribed AChE RNA corresponding to the 3′ UTR of the 2.4-kb transcript. In these assays, we focused on the shorter of the two possible 3′ UTR because it should contain most of the key regulatory elements given that the 2.4- and 3.2-kb transcripts are equally sensitive to the effects of denervation (see Fig. 2 and Sketelj et al., 1998; see also Fuentes and Taylor, 1993). As AChE mRNA levels are compromised most significantly in adult muscles following denervation, we therefore also used protein extracts from adult muscles in these experiments. We thus determined the number of possible factors that bound to the short AChE 3′ UTR, as well as their molecular mass. In protein extracts from control muscle, we readily detected one group of complexes that consisted primarily of two bands (black arrows in Fig. 6B). The presence of these complexes was eliminated completely if the protein extracts were omitted from the incubation mixtures. Incubation of the protein extracts with antisense AChE cRNAs did not produce this banding pattern, indicating that the complexes seen with the 3′ UTR were indeed specific (data not shown). In these experiments, we observed, in agreement with our hypothesis, that denervation increased the amount of RNA-protein interactions because the intensity of the upper band was consistently increased in denervated muscle (Fig. 6B). Furthermore, we observed the presence of an additional complex that appeared strongly induced in denervated samples (white arrow in Fig. 6).
Recent studies have shown that in vivo the levels of AChE mRNA in mammalian skeletal muscle can display significant changes in response to alterations in the amount and pattern of neuromuscular activation (Cresnar et al., 1994; Michel et al., 1994; Sveistrup et al., 1995; Sketelj et al., 1998). In the present study, we have therefore examined the contribution of transcriptional versus posttranscriptional regulatory mechanisms in the activity-linked regulation of AChE transcripts in adult and postnatal skeletal muscles. As our previous work showed that denervation of adult muscles induces large reductions in AChE mRNA levels (Michel et al., 1994), we chose, in the present study, to focus on this particular model to elucidate the nature of the underlying molecular mechanisms. Our results obtained using several complementary experimental approaches indicate that posttranscriptional regulation appears as the primary mechanism controlling the levels of transcripts encoding AChE in adult skeletal muscle, whereas in neo- and postnatal muscle tissues, expression of AChE mRNA appears regulated at both transcriptional and posttranscriptional levels.
Regulation of AChE gene expression in adult muscle
Several reports have demonstrated unequivocally the important role of neural electrical activation in regulating AChE enzyme activity in skeletal muscle (for review, see Massoulié et al., 1993). For instance, muscle denervation in rodents leads to a substantial reduction in AChE enzymatic activity (Drachman, 1972; Butler et al., 1978; Fernandez and Duell, 1980; Michel et al., 1994; Boudreau-Larivière et al., 1997). In a previous study, we showed that 10 days of denervation resulted in a 10-fold reduction in enzyme activity that was accompanied by an equally pronounced decrease in AChE transcript levels (Michel et al., 1994; see also Cresnar et al., 1994). Our current results confirm and extend these initial findings on AChE mRNA levels in muscle. Indeed, we observed that the decrease in AChE mRNA levels following abolition of neuromuscular activity is extremely rapid because the amount of transcripts is reduced by ∼90% within the first 48 h following denervation (see also Sketelj et al., 1998). In addition, we show by northern blot analysis that the two predominant mRNA species encoding the T catalytic subunit expressed in adult muscle (Legay et al., 1993a, b) are reduced to a similar extent, and in parallel, in denervated muscle. As previously shown by others and in the present study, this effect of denervation stands in sharp contrast to the marked increase that occurs in the levels of the AChR α-subunit transcript in denervated muscle (for example, see Merlie et al., 1984; Goldman et al., 1985, 1988; Witzemann et al., 1991), thereby confirming that the regulation of these transcripts is clearly distinct in adult mammalian muscle.
The pronounced reduction in the levels of AChE mRNA could have resulted from alterations in the rate of mRNA synthesis and/or degradation. To examine this issue, we performed nuclear run-on assays using fast TA muscles, as well as direct injections of a rat AChE promoter-reporter gene construct. Our complementary analyses revealed that the 10-fold reduction in AChE mRNA levels in adult muscle is not accompanied by parallel modification in the rate of AChE gene transcription. Similar results would also be anticipated for slow muscles given that intact slow and fast muscles display similar levels of AChE gene transcription despite exhibiting different basal levels of AChE mRNA (Boudreau-Larivière et al., 2000). Given that transcriptional control appears as the primary mechanism regulating AChR α-subunit mRNA levels in denervated muscle (see Fontaine and Changeux, 1989; Tsay and Schmidt, 1989), our data further highlight the discordance between the regulatory mechanisms controlling expression of the genes encoding AChE and the AChR α-subunit. Taken together, our results are therefore coherent with the notion that in adult skeletal muscle, the activity-linked regulation of AChE mRNA levels occurs via posttranscriptional regulatory mechanisms.
As alterations in message stability appeared as a key posttranscriptional event accounting for the modifications in the relative abundance of AChE transcripts in skeletal muscle fibers in vivo (see also Fuentes and Taylor, 1993), we performed a series of experiments in which in vitro transcribed AChE RNAs were incubated for different time intervals with protein extracts from control and denervated muscles. Using this approach, we observed that denervation increased the rate of degradation of AChE transcripts, thereby confirming that posttranscriptional regulatory mechanisms are indeed responsible for the reduction in AChE transcripts seen in denervated muscles. In this context, it is important to note that the greater rate of mRNA degradation is unlikely related to a general increase in the activity of a ribonuclease because it is well established, for example, that denervation induces a pronounced increase (up to 50-fold) in the expression of transcripts encoding various AChR subunits (Hall and Sanes, 1993; Duclert and Changeux, 1995). Although the mechanisms responsible for the longevity of mRNAs in cells are largely undefined, specific mRNA cis-elements and trans-acting factors that appear to modify the degradation rate have been identified (for reviews, see Greenberg and Belasco, 1993; Sachs, 1993; Ross, 1995). For instance, the poly(A)+ tail of mRNAs appears to protect transcripts from degradation, whereas AU-rich elements located within the 3′ UTR have been linked to mRNA destabilization. The mRNA half-life is also dependent on interactions with trans-acting regulatory factors whose abundance or activity can be altered in response to changes in the cellular environment. In this context, we demonstrate that the large decreases in AChE mRNA levels seen in denervated muscles are accompanied by increases in the abundance of RNA-protein interactions in the 3′ UTR of the AChE mRNA. Interestingly, recent studies have shown that increased contractile activity of rat skeletal muscle decreases, on the other hand, RNA-protein interactions in the 3′ UTR of cytochrome c mRNAs (Yan et al., 1996) and enhances cytochrome c mRNA stability (Freyssenet et al., 1999). Together, these data suggest that the activity-linked regulation of specific mRNAs in muscle is mediated through alterations in the abundance or activity of RNA-binding proteins, which in turn control the stability of cellular transcripts. It will therefore become important to define specifically the cis-acting elements and trans-acting factors responsible for controlling the stability of AChE transcripts in muscle.
Regulation of AChE gene expression in postnatal muscle
In contrast to the nearly identical response of AChR α-subunit transcripts to denervation in postnatal and adult muscles, the pattern of expression of AChE mRNAs following denervation differed significantly in developing versus mature muscles. Specifically, AChE mRNA levels in adult muscles were rapidly and markedly affected following denervation because they decreased by ∼90% within the first 48 h. In denervated postnatal muscles, however, the reduction was clearly not as pronounced and, accordingly, levels of AChE transcripts were ∼50% of those seen in control muscles even after 7 days of denervation. Interestingly and in agreement with our findings, previous studies have shown, in denervated postnatal muscles, an increase in AChE activity in extrasynaptic compartments of muscle fibers (Lubinska and Zelena, 1966; Gautron et al., 1983). An agedependent response of AChE synthesis has also been documented in muscle reinnervation studies (Maldonado et al., 1984). Notably, the appearance of asymmetric forms in ectopic junctions was shown to occur more quickly in reinnervated muscles of young versus adult rats (Maldonado et al., 1984). Finally, the increased expression of AChE mRNA in denervated chick muscle was shown recently to be less dramatic in neonatal animals as compared with their adult counterparts (Rimer and Randall, 1999). Together, these findings indicate that the response of AChE to muscle denervation is agedependent.
Our nuclear run-on assays have revealed that the transcriptional activity of the AChE gene is increased three-fold in denervated postnatal muscle. It is interesting that although the regulation of AChE is distinct in rat versus chick and rabbit skeletal muscles (see, for example, Massoulié et al., 1993), a recent study also reported a three- to fourfold increase in the transcriptional activity of the AChE gene in neonatal chick muscles in response to denervation (Rimer and Randall, 1999). In the present investigation, direct plasmid injection of a rat AChE promoter-reporter gene construct into muscles from postnatal rats also showed a threefold increase in β-galactosidase activity, thereby confirming the direction of the transcriptional induction as determined by nuclear run-on assays. This elevation in AChE gene transcription likely accounts for the attenuated reduction in AChE mRNA levels observed following denervation of postnatal muscle. However, although transcriptional activation of the AChE gene occurs in denervated postnatal muscle, our experiments also suggest that posttranscriptional control is implicated in postnatal rats given the observed reduction in AChE transcript levels. Taken together with the finding obtained with adult muscles, these data indicate that the activity-linked expression of AChE mRNAs in muscle is subjected to developmental influences and that distinct molecular mechanisms operate at specific stages of muscle fiber maturation.
The reason for the differential response of the gene encoding AChE seen in neonatal and postnatal versus adult muscles remains unclear. Based on our experiments using a rat AChE promoter-reporter gene construct, however, it appears that alterations in the relative abundance of trans-acting factors in developing versus adult muscles, which ultimately regulate the transcriptional activity of the AChE gene, account for the observed differences. In this context, age-associated changes in the basic mechanisms regulating transcription have been observed in a variety of experimental systems (see Hsieh et al., 1998 and references therein), including rodent skeletal muscles (Lee et al., 1999). It would therefore be important in future studies to identify the specific 5′ DNA regulatory elements within the AChE gene that mediate this age-dependent, activity-linked transcriptional regulation.
One attractive possibility to explain the different levels of transcription factors targeting genes encoding synaptic proteins in developing versus adult muscles is that their relative abundance may be related to the level of maturity of the postsynaptic apparatus. Indeed, ultrastructural studies have shown that although exploratory motor axons reach the surface of developing myotubes at approximately embryonic day 13-14, full differentiation of the presynaptic nerve terminals and of the postsynaptic membrane of the neuromuscular junction requires several weeks and occurs not only during embryonic development, but also in neonatal and postnatal muscles (Kullberg et al., 1977). For example, polyinnervation remains for 2-3 weeks postnatally before it is eliminated, thereby indicating that the establishment of the mature neuromuscular junction occurs only once this process of synapse elimination has subsided (Dennis, 1981). It is interesting that a recent study has also highlighted a differential plasticity at the neuromuscular junctions from neonatal versus adult rats in response to partial denervation (Lubischer and Thompson, 1999). On the basis of our current results and those of others (Legay et al., 1995; see also Fontaine and Changeux, 1989; Tsay and Schmidt, 1989; Piette et al., 1993), it appears that during synaptogenesis and maturation of the neuromuscular junction, the activity-linked regulation of the AChE gene involves transcriptional and posttranscriptional regulatory mechanisms and that, in mature muscle, regulation of this synaptic protein is achieved primarily via posttranscriptional events. Accordingly, it appears therefore that the activity-linked transcriptional regulation of the AChE gene demonstrates a high level of plasticity during muscle development when maturation of neuromuscular junctions is still occurring.
We thank Drs. Claire Legay and Jean Massoulié (Ecole Normale Supérieure), Daniel Goldman (University of Michigan), and Kevin Burns (University of Ottawa) for providing the AChE, AChR α-subunit, and β-actin cDNAs, respectively, and Mr. John Lunde for expert technical assistance. This work was supported by grants from the Medical Research Council of Canada (MRC) and the University Research Fund to B.J.J. B.J.J. is an MRC Scientist.