| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Institute of Sports Medicine, Bispebjerg Hospital, Copenhagen; 2Department of Sport Science, University of Aarhus, Aarhus; 3Department of Molecular Muscle Biology, Copenhagen Muscle Research Centre, Rigshospitalet, Copenhagen; and 4Department of Biomedical Sciences, University of Copenhagen, Copenhagen, Denmark
Submitted 21 December 2006 ; accepted in final form 3 August 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
eccentric; atrogin-1; protein degradation; muscle RING finger-protein-1; forkhead box O transcription factor
Net protein turnover is a sum of anabolic and catabolic activity, and muscular activity is known to increase myofibrillar protein turnover, increasing both protein synthesis and degradation (4, 42). Whereas the activation of protein synthesis by exercise has been intensely studied in several species (37, 43), the isolated effect of exercise on protein degradation is less well characterized in humans. It is known that the rate of protein breakdown increases concomitant with that of protein synthesis in response to exercise, suggesting a molecular link between synthesis and degradation (42). However, the detailed regulation of protein degradation pathways in human skeletal muscle in relation to exercise is poorly characterized.
It is known that different loading regimens of skeletal muscle results in different outcomes with regard to adaptation and that concentric and eccentric contractions seem to complement each other in terms of muscle hypertrophy (19). Even though eccentric training produces greater soreness following exercise and in some discrete settings seems to contribute more to hypertrophy and protein turnover than concentric (15, 27), it does not seem to produce overall greater changes in protein synthesis (10, 38, 42). It is, however, not known if or how protein degradation pathways are influenced differently by the mode of muscular contraction.
Three major intracellular proteolytic systems exist: the lysosomal, the calcium-dependent, and the ubiquitin-proteasome system (UPS). It is known that calcium-dependent systems and the UPS degrade the major fraction of intracellular proteins and thus contractile proteins of skeletal muscle (17, 48). Calcium- and UPS-mediated proteolysis are thought to cooperate to some degree in the degradation of some sarcomeric proteins. Calcium-dependent processing of the sarcomeres releases myofilaments, which are then degraded by the UPS, a theory commonly referred to as the "calpain hypothesis" (2, 13, 47). The UPS degrades proteins in a tightly regulated manner, not only by regulating the amount of contractile protein, but also by affecting transcription and metabolism through coordinated degradation of transcription factors and discrete signaling roles (16). It works by attaching ubiquitin, a 76-amino acid protein, serially to lysyl residues of target proteins, which marks them for destruction in the proteasome. The process of ubiquitination involves ubiquitin being bound to a ubiquitin activating enzyme (E1 protein), transferred to a ubiquitin carrier protein (E2 protein), and ultimately a ubiquitin ligase (E3 protein) catalyzes the transfer of ubiquitin to lysine residues of target proteins, the step thought to be rate limiting in ubiquitination as a whole. At present, several hundred E3 proteins are known, this high number being necessary to generate a wide enough specificity range to cope with the multitudes of protein species needing degradation. Several of these E3 proteins display tissue-specific expression. Ubiquitin residues are then added serially to the first in a regulated manner, and when four residues have been attached, the substrate protein is terminally marked for degradation in the proteasome.
Exercise has been reported to acutely increase expression of ubiquitin and proteasome subunit-
1 gene (PSMA1) itself (55), while the response to repeated bouts of exercise is somewhat less well defined, with one study reporting upregulation of ubiquitination (51) and another reporting downregulation in terms of ubiquitin protein (55). With regard to other UPS-related components, three novel ubiquitin ligases collectively referred to as atrogenes, muscle RING finger protein 1 (MURF1), atrogin-1 (also known as MAFbx) and E3, and the forkhead box O (FOXO) transcription factors are believed to be linked to development of muscle atrophy phenomenologically (5, 28, 29) and causatively (5, 24, 26). In almost all atrophy models, MURF1 and atrogin-1 are upregulated secondary to an activation and nuclear translocation of FOXO transcription factors in the nucleus, a finding compounded by several in vitro studies indicating a regulatory role of FOXOs on expression of the ubiquitin ligases mentioned above (45, 50). While only MURF1 of the MURF family has confirmed ubiquitin ligase activity and thus constitutes an E3 protein, all the MURF proteins are highly homologous and associate with part of myocellular ultrastructure (6, 32, 33, 49). This could link MURFs to a mechanosensory function, a notion that is highly relevant with respect to the acquired exercise tolerance, as this phenomenon is known to be intensity dependent. Atrogin-1 has been shown to degrade MyoD (52), a transcription factor necessary for proper muscle development, fetal as well as postnatally (36, 54), and the domain recognized by atrogin-1 is a known transcription factor signal, thus possibly placing atrogin-1 in the role of orchestration of rather than degradation of bulk protein (52). Very little is known regarding changes in regulation and expression of these ubiquitin ligases in response to mechanical loading of skeletal muscle, with different loading modalities, as well as in the response to repeated bouts of exercise.
The aim of the present study was to examine 1) if concentric and eccentric exercise contributes in a differentiated manner to the regulated expression of markers of skeletal muscle protein degradation and 2) if application of a repeated bout of eccentric exercise is associated with changes in expression of said markers.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Exercise protocol and physiological testing. Subjects in the S and SW groups performed two bouts of stepping exercise separated by 8 wk as previously described (39), thereby performing eccentric work with one leg and concentric work with the other. Subjects performed concentric and eccentric work with the same legs in both bouts. The SW group performed the work wearing a weight vest weighing 10 kg. In both legs, soreness was evaluated and strength was measured immediately before and 3 h, 24 h, and 2, 3, 5, and 7 days after each bout (strength was also measured 7 days before each bout). Soreness was quantified on a visual analog scale. The subjects scored the soreness during rising from and sitting in a chair in the part of the thigh corresponding to the quadriceps muscle (1). Isometric strength was measured at 90° knee flexion using a strain-gauge transducer. Before the first test, subjects were familiarized to the strength testing protocol. Subjects in the control group were tested as described above, but they performed no exercise and testing was not repeated after 8 wk.
Sample collection, RNA purification and gel electrophoresis.
Regular conchotome biopsies were obtained from vastus lateralis of the quadriceps muscle as described previously (12). Biopsy samples were obtained 1 wk before each bout, 3 h after, 24 h after, and 7 days postexercise. Following the first bout, biopsies were obtained from both legs, while after the second bout biopsies were only cut from the eccentrically exercised leg. Biopsies from the control group were obtained from the right leg. Following biopsy sampling, 
50 mg biopsy tissue was cut off the specimen and immediately frozen in liquid nitrogen. The remaining part was embedded in TissueTek with fascicles arranged longitudinally and frozen in liquid nitrogen-cooled isopentane and used for further analysis. Seventy-five to one-hundred 10-µm sections were cut from biopsies on a cryostat. From these sections, RNA was extracted using the guanidinium thiocyanate-phenol-chloroform extraction method previously described by Chomzynski and Sacchi (7). RNA concentrations were quantified spectrophotometrically, and samples were diluted appropriately in RNase-free water and formaldehyde gel loading buffer (Ambion) and subjected to electrophoresis on an agarose gel [1% NuSieve GTG agarose (Cambrex), 1.5 % formaldehyde, and 1x MOPS buffer (Eppendorf)] at 7 V/cm for 65 min. Two-hundred nanograms RNA was loaded per lane. Following electrophoresis, the gel was stained in SYBR green II RNA gel stain (Cambrex) diluted 10,000-fold for 30 min and scanned on a Molecular Imager FX scanner (Bio-Rad). The relative intensity between large and small ribosomal bands was used to assess the RNA quality. The RNA content of the gel was transferred to a Positive nylon membrane (Appligene) using osmocapillary blotting with 25 mM NaOH.
Probe cloning and synthesis.
Probe templates were amplified from human muscle cDNA using the Accuprime Taq PCR (Invitrogen) and primers from MWG Biotech (Table 1) and purified using the Wizard SV Gel and PCR clean-up system kit (Promega). PCR products were inserted into the pBlueScript SKII(+) phagemid SmaI site and transformed into DH5 E. coli cells. Following cloning, DNA was purified using the High Pure Plasmid purification kit (Boehringer/Roche), and insert identity was confirmed by restriction mapping. M13 primers were used to generate double-stranded probe templates with a biotinylated sense strand. While the sense strand was retained on streptavidin-covered beads, the antisense strand was stripped and a new 32P-tagged antisense strand was synthesized on the sense strand using [-32P]dATP as previously described by Jonsdottir et al. (22). Following denaturation, the antisense strand was retrieved, and its activity was measured.
|
-32P]ATP followed by heat inactivating. The tagged oligo was recovered by precipitating the probe on a glycogen carrier, using ammonium acetate, and washing with ethanol followed by air drying. The pellet was resuspended in Tris-EDTA (TE) buffer and diluted with unmarked oligo. Northern blotting, hybridization, washing, and stripping. Membranes were preincubated at 50°C for 1 h with 5 ml hybridization buffer (UltraHyb, Ambion) followed by addition of 5–10 million counts of tagged probe. Hybridization was conducted overnight at 50°C with rotation. After incubation, the membranes were washed twice at low-stringency conditions [2x saline-sodium phosphate-EDTA (SSPE) (Invitrogen), 0.1 % SDS, room temperature] for 5 min and twice at high-stringency conditions (0.1x SSPE, 0.1 % SDS, 60°C) for 15 min. The washed membranes were exposed on Phosphor screens. Following exposure, membranes were scanned at a resolution at 50 µm using a FX Phosphorimager and photodensitometrically quantified using QuantityOne (Bio-Rad) software. Signals were normalized to backgrounds using a representative segment of the membranes (global background). Where several mRNAs binding the probe were present, all were quantified individually (except for MURF3 where 3 distinct bands were present of which 2 were inseparable) and numbered according to decreasing size. Membranes were stripped according to the EZQ stripping protocol and reprobed four to six times with probes in increasing order of band intensity.
Hybridization with 28S probe was performed with 2 pmol probe, essentially using the same protocol, but preincubation and incubation were conducted at 42°C and washing at room temperature. A representative blot for all hybridizations are shown in Fig. 1.
|
All RNA data were normalized to 28S rRNA, followed by two separate statistical procedures. First, for the 28S-normalized data, values within each bout were normalized to their respective PRE values (obtained 1 wk before each exercise bout), log-transformed, and subjected to a procedure like that described earlier, first testing for an effect of the weight vest and then testing for an effect of bout, time, or both. Second, to quantify changes in baseline expression between the first and second bout, 28S-normalized data were log-transformed and normalized to PRE/ECC1, followed by a paired t-test analysis of a null-hypothesis on PRE/ECC2. All statistics were done using SigmaStat (Systat) and Microsoft Excel.
Data for controls were subjected to one-way ANOVA for time to test if time affects the signal in any sample. If an effect for time was seen, Student's post hoc analysis was performed.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
45 mm on a VAS scale, peaking 2 days after the exercise bout. This response lasted up to 4 days. The first eccentric bout (ECC1) produced more soreness than CONC or ECC2, also documenting a repeated bout effect for soreness.
mRNA targets: structural UPS components.
For ubiquitin C polygene (UBC), two transcripts were detected, the second band representing a transcript corresponding in size to the other ubiquitin polygene, Ubiquitin B (UBB), in the human genome (as verified by hybridization with a UBB probe in our laboratory). As considerable sequence homology to this mRNA could not be avoided in design of the UBC probe, we consider it likely that this second splice form represents UBB. Both displayed a similar response pattern (Fig. 3), albeit with a stronger response for splice form I, the splice form displaying the highest absolute expression. Exercise produced a tendency toward upregulation of ubiquitin mRNA expression, peaking at 50% for splice form I and 30% for splice form II (UBB) 3 and 24 h postexercise and a downregulation of 20% 7 days postexercise. The response to the first eccentric bout was stronger than that to the concentric bout and tended to be stronger than that observed for the second eccentric bout. Baseline expression (PRE/ECC2, relative to PRE/ECC1) decreased 15% in the second bout (see Fig. 6). Also, for the first splice form an upregulation could be seen in controls at 24 h postexercise (P < 0.01).
|
|
|
|
40% lasting the first days after exercise (Fig. 3).
mRNA targets: putative atrogenes.
For the FOXO1 transcript, an upregulation of 70% was seen 3 h postexercise only with concentric exercise (Fig. 4).
FOXO3 mRNA displayed a weak downregulation, not reaching significance.
For the E3 transcript, a downregulation across the 3-h time point, with no difference within individual treatments, was the only observed finding (Fig. 4). A downregulation could be seen in controls 7 days postexercise. Baseline expression also appeared to be downregulated in the second bout (see Fig. 6).
For MURF1, concentric exercise produced a very short-lived upregulation of 150% (Fig. 4). The first eccentric exercise bout produced a nonsignificant expression decrease of 30% that persisted all through 7 days postexercise. Baseline expression also decreased almost 50% between the first and second bout of eccentric exercise (see Fig. 6).
Our atrogin-1 probe detected four transcripts of sizes 1.5 to 6.4 kb. Several splice forms have previously been observed (30). Qualitatively, the observed responses were similar for all four splice forms. Eccentric exercise produced a consistent downregulation of up to 65%, lasting up to a week after the exercise. Concentric exercise produced little if any change. For two splice forms, a significant upregulation could be observed 24 h postexercise in controls.
mRNA targets: MURF1-related proteins.
Four different transcript variants have been reported in GenBank for MURF2 (Fig. 5). The probe utilized in the present study only detects the transcript variants NM_033058
[GenBank]
and NM_184085
[GenBank]
. Two transcripts were detected in our Northern blots, displaying similar responses. Eccentric exercise produced a consistent upregulation of up to 200% in both splice forms, an effect lasting at least 24 h, whereas concentric exercise produced no changes.
According to our probe design analysis, the probe designed for MURF3 shows specificity for both the reported splice forms (accession nos. NM_032546 and NM_187841), but in our Northern blots three bands appeared. Differentiated responses were displayed between splice form I and the combined splice forms II + III, which were quantified together, because of inability to separate the bands (Fig. 5). For all splice forms, eccentric exercise only produced an upregulation of up to 90%, peaking 3 h postexercise for splice form I and at 24 h postexercise for splice form II.
Normalization control.
For GAPDH, concentric exercise produced a weak increase of 20% that peaked 3 hours post-exercise and declined afterward (Fig. 6 and 7). Eccentric exercise produced a decrease of up to 30%, peaking 7 days postexercise. Furthermore, baseline expression was decreased in the second bout (see Fig. 6).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Repeated bout effect on atrogenic and UPS component gene expression. The exercise protocol utilized in the study produced significant strength loss and soreness, and those parameters were attenuated in the second bout at several time points, thus documenting a repeated bout effect (Fig. 2). Furthermore, a recent study using a similar protocol validated the adequacy of loaded stepping exercise to produce changes in protein turnover (10), thereby supporting that molecular changes associated with muscle disruption and changes in protein turnover also increase with our intervention.
The acute transcriptional response of MURF1 to eccentric exercise differed between the first and the second bout of exercise, thus representing a very long-lasting change in response potential (Fig. 4). A tendency toward this (P < 0.1) could be seen for FOXO1, MURF2, and MURF3.
This is further supported by the fact that the difference in mRNA expression of MURF2, MURF3, PSMA1, and UBC between bouts correlated with the integrated difference between bouts for strength and soreness (R2 up to 0.84, correlations not shown).
Of particular interest, we found that prelevel expression of MURF1 had decreased by 50% before the second bout, rendering possible a downreguation of MURF1 expression as long lasting as 7 wk. Does this chronic downregulation of MURF1 then represent a depression of proteolysis 7 wk later? Probably not, as other studies show that nitrogen balance is only affected for 48–72 h postexercise with protocols of concentric or eccentric exercise or combinations thereof (37, 42). The observed downregulation may, however, represent a decrease in degradative capacity that matches a decrease in protein turnover with the accustomization to exercise (25, 41).
For MURF1, eccentric exercise seems to induce a persistent downregulation while concentric exercise induces an acute upregulation. With exercise encompassing both eccentric and concentric loading, it is hard to tell which part of the stimulus dictates the response the strongest. Most studies seem to indicate a dominant downregulation by exercise per se (14, 21, 58). Considering that MURF1 is known to ubiquitinate sarcomeric proteins in myocardium (56) and thus possibly to partake also in skeletal muscle protein degradation, this indicates that MURF1 is involved in maintenance of the repeated bout effect.
Regulation of FOXO1, FOXO3, MURF1, and atrogin-1. For FOXO1 and MURF1, a similar degree of acute upregulation in response to concentric loading exclusively was observed. As previously mentioned, this may point at these genes being regulated by metabolic stress rather than myofibrillar disruption. This is supported by previous studies showing upregulation of FOXO1 in response to energy depletion caused by exercise (23) or dietary manipulation (20). While the observed FOXO1 and MURF1 expression changes may be related to the onset of muscle damage, they do not seem to be vital components of the muscle damage inflammation process per se as this process is considerably more long-lived than 24 h.
In previous studies, FOXO transcription factors have been associated with regulation of atrogin-1 and MURF1 (45, 50). The present findings support that FOXO1 is coregulated with MURF1 or possibly involved in the regulation of it. FOXO3 has been thought to be related to regulation of atrogin-1, but in this setup no common behavior could be observed.
Most studies have shown that states of accelerated nitrogen loss are associated with upregulation of MURF1 and atrogin-1 expression (5, 18). In the present data, we can observe an acute downregulation of atrogin-1, a finding somewhat supported by previous studies (8, 58), but not all (29). Also, considering that exercise has been shown to blunt the atrogin-1 upregulation and muscle loss associated with various wasting stimuli, like immobilization (21) or hindlimb suspension (14), one might speculate that atrogin-1 is a marker of atrophy rather than a marker of increased proteolysis per se. This is supported by the fact that atrogin-1 displays specificity for a motif present in myoD and possibly other myogenic transcription factors (52), a transcription factor positively associated with hypertrophy (36, 54).
Sampling and normalization.
Previous studies have shown several genes to be induced by invasive procedures like muscle biopsy sampling, thereby in itself affecting gene expression in subsequent samples (31, 53). To take this sampling effect into account, we included a nonexercising control group. For several targets, we observed what could likely be a sampling effect, i.e., regulation in control subjects of UBC, atrogin-1, and E3 (Figs. 3 and 4). For UBC, the observed upregulation in the control group coincides with an upregulation observed in exercising subjects. Thus it is likely that the observed change in UBC expression in the exercise group is influenced by muscle traumatization due to biopsy sampling. Furthermore, the upregulation observed for two splice forms of atrogin-1 may suggest that the downregulation observed in exercising subjects is in fact blunted by an upregulation caused by sampling.
Adding to complexity, it has been previously documented that most housekeeping genes used for normalization are in fact not constantly expressed (11). This potentially has severe consequences for the interpretation of changes in gene expression as changes in normalization gene can translate into reciprocal changes in normalized gene expression. To test this, we also measured another unrelated "normalization" RNA, GAPDH, for comparison to identify potential problems with our chosen normalization RNA, 28 rRNA. GAPDH mRNA was quantified and normalized to 28S rRNA. By such means of normalization, we observed that GAPDH exhibited little but yet significant fluctuation in response to our exercise intervention, especially at the 7-day postexercise time point (Fig. 7), indicating that either 28S rRNA or GAPDH mRNA or both changed with the intervention, although both are normally considered constantly expressed (46). As most changes for our UPS targets manifest at 3 and 24 h postexercise, where the changes in GAPDH/28S ratio are smallest, this does not seem to affect any of the biologically interesting findings presented here. In fact the GAPDH/28S ratio seemed to have changed before the second exercise bout, indicating an effect on baseline expression of GAPDH or 28S lasting for 7 wk. Nevertheless, with regard to the changes in baseline expression observed for E3, UBC, and MURF1, changes in 28S expression may have influenced the E3 and UBC findings, whereas the MURF1 downregulation is too strong to be influenced by such an effect.
In summary, most of the changes we observed in the present study can be attributed to a real exercise effect and not a sampling or normalization artefact. However, our results also emphasize that great care must be taken when interpreting gene expression patterns derived from study designs such as the present one, where muscle traumatization from sampling resembles muscle damage as induced by the exercise intervention.
As for the interpretations of this study, one must take into consideration that the division in concentric and eccentric exercise used is rather unlike most kinds of physical activity, and one should take care not to extrapolate uncritically to other exercise models as it can be hard to assess eccentric and concentric contributions to the total response with other forms of exercise. Also, the exercise model used is one of muscular endurance rather than strength per se, as the exercise bouts lasted for 30 min. This also means that the mechanical load is smaller than that seen with the isokinetic knee extension apparatuses often used to generate eccentric work, and the results presented would likely have been different using such an apparatus. One could also speculate that dietary control of the subjects could have affected data. Finally, the "proteolytic marker" status of several of the measured targets is questionable and calls for validation using in vivo protein metabolism measurements.
Conclusion and perspectives. Several studies have documented the association between expression of a defined group of genes and accelerated proteolysis. As the negative protein balance seen in most states of accelerated net protein loss is accounted for by increases in protein degradation rather than decreases in synthesis, it is still somewhat unclear if the observed transcriptional changes are associated with a modulation of the net protein balance or the isolated protein degradation. One study has reported a statistically significant association between the transcription of UBB polygene mRNA and protein breakdown (3), and another study has reported an association between the amount of 14-kDa actin caspase cleavage fragment and whole body protein loss (57). Thus the in vivo significance of the presented findings are not obvious, and for a clear interpretation to made, we need to 1) identify bottlenecks in ubiquitination and proteasomal degradation and 2) determine substrate specificity of the proposed ubiquitin ligases. The presented data thus help pinpoint target gene products for future research.
Protein metabolism tracer studies indicate that with the accustomization to exercise, exercise-induced increases in net protein synthesis diminish. Such accustomization seems evident from the appearance of attenuated development of muscle damage following exercise. This long-term physiological change must be paralleled and mediated by equally long-lived molecular changes. In the present study, we support the existence of such long-lived changes induced by exercise in a system involved in the regulation of protein metabolism. Thus the present results suggest that regulation at the mRNA level of systems mediating protein degradation takes part in mediating and sustaining the accustomization to exercise.
In conclusion, UPS component expression is modulated in human skeletal muscle with exercise in a manner differentiated with loading modality. Also, the attenuated muscle damage response with repeated loading is associated with a chronic modulation of distinct UPS component expression, possibly indicating a decrease in proteolytic capacity. The present data suggest that some modulation of the proteolytic apparatus occurs, but this calls for further research using in vivo protein metabolism techniques. The pattern of expression changes associated with exercise indicates that modulation of the UPS may take part in the acquired exercise tolerance producing diminishing returns with repeated applications of a given exercise stimulus.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
