Am J Physiol Endocrinol Metab 280: E383E390, 2001.

Mechanical load increases muscle IGF-I and androgen receptor mRNA concentrations in humans
MARCAS M. BAMMAN,1,2 JAMES R. SHIPP,1 JIE JIANG,4 BARBARA A. GOWER,2.3 GARY R. HUNTER,1,3 ASHLEY GOODMAN,1 CHARLES L. MCLAFFERTY, JR.,1 AND RANDALL J. URBAN4 Departments of 1Human Studies, 2Physiology and Biophysics, and 3Nutrition Sciences, University of Alabama at Birmingham, Birmingham, Alabama 35294; and 4Department of Internal Medicine, University of Texas Medical Branch, Galveston, Texas 77550
Received 20 January 2000; accepted in nal form 25 October 2000

Bamman, Marcas M., James R. Shipp, Jie Jiang, Barbara A. Gower, Gary R. Hunter, Ashley Goodman, Charles L. McLafferty, Jr., and Randall J. Urban. Mechanical load increases muscle IGF-I and androgen receptor mRNA concentrations in humans. Am J Physiol Endocrinol Metab 280: E383E390, 2001.The mechanism(s) of loadinduced muscle hypertrophy is as yet unclear, but increasing evidence suggests a role for locally expressed insulin-like growth factor I (IGF-I). We investigated the effects of concentric (CON) vs. eccentric (ECC) loading on muscle IGF-I mRNA concentration. We hypothesized a greater IGF-I response after ECC compared with CON. Ten healthy subjects (24.4 0.7 yr, 174.5 2.6 cm, 70.9 4.3 kg) completed eight sets of eight CON or ECC squats separated by 610 days. IGF-I, IGF binding protein-4 (IGFBP-4), and androgen receptor (AR) mRNA concentrations were determined in vastus lateralis muscle by RT-PCR before and 48 h after ECC and CON. Serum total testosterone (TT) and IGF-I were measured serially across 48 h, and serum creatine kinase activity (CK), isometric maximum voluntary contraction (MVC), and soreness were determined at 48 h. IGF-I mRNA concentration increased 62% and IGFBP-4 mRNA concentration decreased 57% after ECC (P 0.05). Changes after CON were similar but not signicant (P 0.060.12). AR mRNA concentration increased (P 0.05) after ECC (63%) and CON (102%). Serum TT and IGF-I showed little change. MVC fell 10% and CK rose 183% after ECC (P 0.05). Perceived soreness was higher (P 0.01) after ECC compared with CON. Results indicate that a single bout of mechanical loading in humans alters activity of the muscle IGF-I system, and the enhanced response to ECC suggests that IGF-I may somehow modulate tissue regeneration after mechanical damage. insulin-like growth factor I; muscle hypertrophy; resistance exercise; creatine kinase; concentric; eccentric

CONVENTIONAL RESISTANCE TRAINING

involves alternating concentric (CON) and eccentric (ECC) muscle actions against a constant external load, the magnitude of which is limited by the individuals CON strength. The hypertrophic response after this biphasic training is

well documented (17, 23); however, the mechanism(s) by which resistance training induces myober hypertrophy is as yet unclear. High-resistance ECC muscle actions induce more sarcolemma damage than CON muscle actions, as evidenced by greater elevations in serum creatine kinase, myoglobin, and troponin I (32). Furthermore, ECC loading leads to more severe and more prolonged myobrillar disruption, soreness, and force decit (15). Whether these events are causally related to the growth response remains unknown; however, myober hypertrophy is blunted if the ECC phase is omitted from conventional resistance training (17). Hather et al. (17) previously found that CON/ECC training induces hypertrophy of both type I and type II myobers, whereas hypertrophy is absent in type I bers and blunted in type II bers after CON training (matched with CON/ECC for total work). ECC loading has also been shown to elevate muscle protein synthesis for a prolonged period in animals (35). It appears that ECC loading, which is submaximal only during conventional resistance exercise [ECC strength is 20 50% greater than CON (6)], is important for maximizing hypertrophy. Although no physiological consequences have been demonstrated, transient elevations in serum anabolic factors such as growth hormone (GH) and testosterone have been documented after a bout of conventional resistance exercise (24), and increased serum growth factors have also been noted across several weeks of resistance training (23). This has led to the speculation that one or more of these factors may somehow stimulate resistance exerciseinduced myobrillar protein synthesis. However, because the hypertrophic response is specic to the loaded muscle(s), activation by a systemic hormone would require load-mediated modulation of the hormones efcacy in the exercised muscle. Load-mediated modulation of receptor expression or binding afnity in the muscle might explain localization of the growth response with elevated serum anabolic
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. E383

Address for reprint requests and other correspondence: Marcas M. Bamman, Muscle Research Laboratory, EB207, 901 South 13th St., The Univ. of Alabama at Birmingham, Birmingham, AL 35294-1250 (mbamman@uab.edu). http://www.ajpendo.org

0193-1849/01 $5.00 Copyright 2001 the American Physiological Society

E384

MECHANICAL LOAD AND MUSCLE IGF-I IN HUMANS

hormones. Whether ECC loading enhances local mechanism(s) in humans is largely unknown. Increasing evidence indicates that modulation of muscle protein turnover is tightly regulated by a number of locally expressed tissue growth factors (1, 12, 13). Insulin-like growth factor I (IGF-I) is known to stimulate myoblast proliferation and differentiation in vitro (13) as well as muscle protein synthesis (22) and, as such, has received increasing attention in studies of muscle hypertrophy. Local expression of IGF-I in skeletal muscle appears to be load sensitive and acts independently of any change in serum GH or IGF-I (1). Furthermore, Yang et al. (37) recently identied two isoforms of IGF-I in skeletal muscle, one of which appears to be regulated exclusively by mechanical load. Some have proposed the name mechanogrowth factor (MGF) for this mechanically sensitive isoform (16). This local IGF-I (or MGF) is thought to induce myober hypertrophy by autocrine (i.e., direct stimulation of myobrillar protein synthesis) and/or paracrine (i.e., satellite cell proliferation, differentiation, and fusion) action (1). The efcacy of muscle IGF-I is dependent not only on its expression but also on its availability, which is regulated by a family of six IGF binding proteins (BPs) and by the abundance of the type 1 IGF receptor. For example, in muscle, IGFBP-4 has a high afnity for IGF-I and thus inhibits its myogenic effects, whereas IGFBP-5 may facilitate (13) or inhibit (21) IGF-I-stimulated differentiation under certain conditions. Additionally, IGFBP-1 has been shown to inhibit IGF-Istimulated protein synthesis (14). The increase of serum IGF-I with exogenous administration of GH or IGF-I does not appear to stimulate myober hypertrophy in the absence of mechanical load (4), nor does systemic GH treatment enhance the hypertrophic effect of resistance training (33, 38). However, induction of exogenous IGF-I directly into skeletal muscle does increase muscle mass (2), suggesting that any stimulus causing an increase in muscle IGF-I availability may lead to muscle growth. Urban et al. (34) have recently shown that exogenous testosterone administration in hypogonadal older men increases muscle strength and protein synthesis and is associated with increased muscle IGF-I mRNA concentration with a concomitant reduction in IGFBP-4 mRNA. Muscle IGF-I mRNA content is also increased in muscle after heavy exercise in humans and after stretch in animals (37). Taken together, these ndings lead to the attractive speculation that mechanical load associated with resistance exercise may increase muscle IGF-I availability, which in turn may cause increased myobrillar protein synthesis, satellite cell activation, and consequent myober hypertrophy. Because ECC loading appears to enhance the hypertrophic response to resistance training (17) and because stretch tension increases muscle IGF-I mRNA (37), it also seems plausible that ECC loading may enhance the local IGF-I response. In this study we investigated the separate effects of CON and ECC resistance exercise on muscle mRNA

concentrations of IGF-I and IGFBP-4. We hypothesized that the local IGF-I response after a single bout of resistance exercise would be greater when heavy ECC muscle actions were performed as opposed to heavy CON muscle actions. Levels of IGF-I and IGFBP-4 mRNA in vastus lateralis muscle samples were determined before and 48 h after ECC and CON. We also measured muscle androgen receptor (AR) mRNA concentration and serum levels of testosterone and IGF-I to investigate possible interactions between serum anabolic factors and local tissue responses.
METHODS

Subjects. Ten healthy subjects (7 men, 3 women) 19 yr of age participated in this study. Each subject was screened for health history by means of a health status questionnaire and for level of activity by use of the Baecke Questionnaire of Habitual Physical Activity. Exclusion criteria included current lower-body resistance training or previous history of diagnosed condition or illness that would endanger the subject during strenuous resistance exercise. All subjects were given an oral and written brieng of the study before signing informed consent forms. The study was approved by the Institutional Review Board and was conducted in the General Clinical Research Center (GCRC) at the University of Alabama at Birmingham. Familiarization sessions. Each subject completed three familiarization sessions on a Smith squat machine (York Barbell, Wright Exercise Equipment, Birmingham, AL) to learn proper execution of the exercise and to become familiar with heavy loads. The third familiarization was completed 5 days before the rst exercise session in an effort to prevent residual effects of the familiarization routine. In familiarization 1, subjects warmed up with a light weight and performed one set of 10 biphasic (CON/ECC) repetitions with a comfortable submaximal load. In familiarization 2, which was 3 days later, subjects warmed up, completed a one-repetition maximum (1RM) strength test, and concluded the session with three sets of 10 repetitions at 60% of 1RM. Approximately 34 days later, subjects returned for the nal familiarization, which consisted of a warm-up, a 1RM test, three sets of 10 repetitions at 70% of that days 1RM, and two sets of 8 repetitions at 80% of 1RM. Strength testing. As mentioned in the preceding section, each subject performed a 1RM squat strength test during the second and third familiarizations. After a sufcient warm-up period, sets of one repetition were executed with increasing load until two failed attempts occurred at a given weight. 1RM was recorded as the highest weight successfully lifted. Each attempt was separated by 2 min. Isometric maximum voluntary contraction (MVC) strength was assessed unilaterally during knee extension. MVC was determined at a knee angle of 1.91 rad (110) using a calibrated force transducer (Omega) interfaced with a desktop computer. Force output was recorded at 100 Hz, and the system provided visual numeric feedback for both subject and investigator. During each test, three MVCs (6 s duration) were performed separated by 1- to 2-min rest periods. The mean of the two highest peak forces (kg) obtained across the three trials was used for analysis. Subjects were instructed to contract as hard as possible and were verbally encouraged throughout each trial. MVC was assessed before the familiarization sessions and was repeated 48 h post-CON and post-ECC.

MECHANICAL LOAD AND MUSCLE IGF-I IN HUMANS

E385

GCRC procedures. Each subject had two exercise days; one day, ECC-only resistance exercise was performed, and the other day, CON-only resistance exercise was performed. CON and ECC bouts were separated by 610 days, and order was randomly assigned. For CON, subjects performed eight sets of eight repetitions with resistance for the rst set established at 85% of the 1RM achieved in the third familiarization. For ECC, subjects also performed eight sets of eight repetitions, but the resistance for the rst set was established at 110% of the 1RM. The resistance was set at these levels to maintain relative workloads (because ECC strength is 2050% greater than CON) and thus maximize motor unit recruitment in both CON and ECC modes. Resistance was decreased if a subject began to fatigue to the point that proper form, repetition number, and/or pace could not be maintained. If a subject reached volitional fatigue after six repetitions or fewer, the load was reduced 510 kg for the next set. Volitional fatigue during CON was dened as an inability to ascend without assistance. During ECC, fatigue was dened as a controlled descent faster than the 2-s minimum. Within subjects, a minimum squat depth was dened during familiarization sessions as a point in the range of motion at which the subjects femur was horizontal (parallel with the oor). This minimum descent was used as a criterion for successful performance during both 1RM testing and the CON and ECC exercise bouts. On each of the two exercise days, subjects reported to the GCRC in a fasted state between 0700 and 0800, and a catheter was placed in an antecubital vein shortly after arrival. The subjects lay quietly for 3040 min before a baseline blood draw. This was immediately followed by the prescribed exercise bout. Subjects remained in the GCRC for 8 h after exercise for serial blood sampling and dietary control. Postexercise blood samples (5 ml) were withdrawn at 0.5, 1, 2, 4, and 8 h. A mixed diet was prescribed by the GCRC dietetics staff. The diet was designed to be isocaloric based on the subjects body weight, and the same three meals were given on each of the two exercise days. The three meals were given at the following times: 1) just after the 1-h postexercise blood draw, 2) 3 h postexercise, and 3) 6 h postexercise. Subjects returned to the GCRC in a fasted state at 24 h postexercise for a blood draw and at 48 h postexercise for a blood draw and muscle biopsy. Also at 48 h, subjects rated perceived soreness on an 11-point scale and completed the MVC strength test to assess functional damage within the muscle. Because these tests immediately followed the muscle biopsy (right thigh), MVC and soreness were determined in the contralateral left limb. Nine of ten subjects were right-

side dominant. Each subject rated muscle soreness on a 010 scale that included corresponding descriptions of the levels of discomfort he/she was experiencing, with 0 no soreness and 10 unbearably sore and difcult to move. Muscle biopsy procedure. The vastus lateralis muscle of the right leg was sampled by percutaneous needle biopsy as we described previously (5). Each subject was biopsied on three occasions: 48 h before beginning the familiarizations, 48 h post-CON, and 48 h post-ECC. Tissue samples were snap frozen in liquid nitrogen and stored at 80C until analysis. Total RNA isolation and qualitative RT-PCR. RT-PCR was performed as described previously (30). Total RNA was isolated from muscle biopsy specimens (50100 mg) using RNAzol B (Tel-Test, Friendswood, TX). Total RNA (0.51.0 g) was then converted to cDNA by means of a reverse transcription kit (Promega, Madison, WI). Five microliters of cDNA solution were then subjected to PCR in the presence of the appropriate primers (Table 1). PCR began with 1 cycle (2 min) at 94C followed by amplication cycles consisting of 1 min at 94C, 1 min at 55C, and 1 min at 72C. The optimum number of amplication cycles was predetermined in control experiments and consisted of 26 cycles for AR and 25 cycles for both IGF-I and IGFBP-4. The products of the PCR were run on Southern gel, and amplied DNA products were sized by DNA ladder. Within subjects, samples from the three time points were run in adjacent lanes. Southern blots were then made and hybridized to oligonucleotides of the DNA fragment (Table 1). Band densities on the Southern blots were quantied by densitometry. For standardization, optical density of each band was corrected for background, and the band density for glyceraldehyde phosphate dehydrogenase was used as the adjusting factor. Serum IGF-I and total testosterone. Total IGF-I and total testosterone were determined in serum samples withdrawn before exercise and at 0.5, 1, 2, 4, 8, 24, and 48 h after CON and ECC bouts. Total IGF-I was determined by immunoradiometric assay (Diagnostic Systems Laboratories,) using 125 I. Total testosterone was determined by solid-phase radioimmunoassay (Diagnostic Products, Los Angeles, CA) using 125 I. All samples within subjects for a given hormone were assayed in random order during a single run. Mean intraassay coefcients of variation (CVs) ranged from 2 to 7%, and mean interassay CVs ranged from 3 to 10%. Serum creatine kinase. Total creatine kinase (CK) activity was determined in serum samples immediately before exercise and at 24 and 48 h after CON and ECC. CK assays of all samples were performed in random order by means of a

Table 1. PCR primers and hybridization oligonucleotides used

IGF-I AR IGFBP-4 GAP, 206 bp GAP, 473 bp sense (1 mol/l) antisense (1 mol/l) oligonucleotide sense (2 mol/l) antisense (2 mol/l) oligonucleotide sense (1 mol/l) antisense (1 mol/l) oligonucleotide sense (0.04 mol/l) antisense (0.04 mol/l) oligonucleotide sense (0.02 mol/l) antisense (0.02 mol/l) oligonucleotide 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 -AAATCAGCAGTCTTGGAACC-3 -CTTCTGGGTCTTGGGCATGT-3 -CAGGGGCTTTTATTTCAACAAG-3 -GATGCTCTACTTCGCCCCTGA-3 -CCCAGCAAATAGAATTCCATGAC-3 -CTGGGTGTGGAAATAGATG-3 -CCATCCAGGAAAGCCTGCA-3 -TGGAAGTTGCCGTTGCGGT-3 -CAGGTGCCTGCAGAAGCA-3 -GGAGTCAACGGATTTGGT-3 -GTGATGGGATTTCCATTGAT-3 -TCAGCCTTGACGGTGCCATG-3 -GGTATCGTGGAAGGACTCAT-3 -TCCACCACCCTGTTGCTGTA-3 -GTGGGTGTCGCTGTTGAAGT-3

IGF-I, insulin-like growth factor I; AR, androgen receptor; GAP, glyceraldehyde phosphate dehydrogenase.

E386

MECHANICAL LOAD AND MUSCLE IGF-I IN HUMANS

Synchron LX System (Beckman Coulter, Fullerton, CA) following manufacturers instructions. By use of this system, the within- and between-run CVs for serum in our laboratory are 2.1 and 1.2%, respectively. Data analysis. Results are reported as means SE. Muscle biopsy and isometric strength performance data were analyzed by one-way repeated-measures ANOVA with three time points, dened as preexercise baseline, 48 h post-CON, and 48 h post-ECC. A priori planned comparisons were analyzed between the baseline value and each of the two 48-h postexercise values. Serum hormone and CK data were analyzed using a new baseline blood draw for each exercise day. The new baseline was established to prevent any residual effect of previous exercise on the postexercise values after the second exercise session. Hormone levels were evaluated by repeated-measures ANOVA with two repeat factors: 1) mode (CON and ECC) and 2) eight time points (baseline and 0.5, 1, 2, 4, 8, 24, and 48 h postexercise). Total testosterone data for three female subjects were excluded from the testosterone analysis, because several values were below the detectable range of the assay. CK levels were tested by repeated-measures ANOVA in similar fashion but with three time points (baseline, 24 h postexercise, and 48 h postexercise). The least squares difference (LSD) test was employed for post hoc testing. Scores on the subjective soreness rating 48 h after CON and ECC were compared using the Wilcoxon matched pairs test. Signicance for all tests was accepted at P 0.05.
RESULTS

Table 3. Indexes of muscle soreness and/or damage

Baseline 48 h PostCON 48 h Post-ECC

MVC,* kg Soreness, ordinal scale, 010

Creatine kinase, U/l

72.4

4.6

77.6

4.2

65.3

3.9

0 (03)
Baseline 24 h postexercise

6 (28)
48 h postexercise

CON ECC

345.2 224.6

67.0 47.0

326.0 449.4

71.5 89.0

264.6 636.3

65.1 244.6

Ordinal soreness data are shown as median (range). All other values are means SE. MVC, maximum voluntary isometric knee extension contraction. Baseline MVC was tested only once (before any exercise). * Main time effect, P 0.05; different from baseline, P 0.05; different from 48 h post-CON, P 0.01.

Subject characteristics are shown in Table 2. The mean repetitions and mean loads for CON and ECC were computed from each subjects average across eight sets. As described in METHODS, the target repetition number was eight per set. Every effort was made to achieve an equivalent number of repetitions on CON and ECC days; thus individual loads were adjusted slightly for each set based on performance in the previous set (see METHODS for details). As shown in Table 2, the method of load adjustment resulted in nearly identical numbers of repetitions during CON and ECC. Moreover, the average loads for CON (82%) and ECC (102%) approximated the target loads of 85 and 110% 1RM for CON and ECC, respectively. ECC loading elicits a blunted metabolic response and activates fewer motor units than CON at the same absolute workload (3, 28). Loads were 2030% higher during ECC loading in an effort to match relative intensity Table 2. Descriptive characteristics of all subjects
Age, yr Height, cm Weight, kg 1RM squat, kg Average repetitions per set CON ECC Average load per set, kg CON ECC 24.4 174.5 70.9 136.1 7.9 7.7 111.6 139.2 0.7 2.6 4.3 10.3 0.1 0.1 7.6 11.8

and thus maximize motor unit activity in both modes of exercise. Measurements used to assess relative muscle damage after CON and ECC are presented in Table 3. Isometric MVC was unchanged 48 h after CON but was depressed 10% 48 h after ECC (P 0.05). Total CK was not different from baseline 24 and 48 h after CON. CK activity was elevated 183% 48 h after ECC (P 0.05). The average rating on the 11-point soreness scale was signicantly higher (P 0.01) 48 h after ECC compared with CON. RT-PCR results for IGF-I, IGFBP-4, and androgen receptor mRNAs are shown in Fig. 1. Repeated-measures ANOVA revealed signicant time effects (P 0.05) for IGFBP-4 and AR mRNA concentrations but not for IGF-I mRNA concentration (P 0.21). Results for IGF-I mRNA concentration were highly variable between subjects and across time. A priori planned comparisons between the baseline value and 48 h postCON and between baseline and 48 h post-ECC for IGF-I mRNA concentration showed a 62% increase

Values are means SE; n 10 subjects. 1RM, one-repetition maximum; CON, concentric exercise; ECC, eccentric exercise.

Fig. 1. Concentrations of mRNA for insulin-like growth factor (IGF) I, IGF binding protein-4 (IGFBP-4), and androgen receptor (AR) at baseline and 48 h after concentric (CON) and eccentric (ECC) loading. Values are means SE. *Different from baseline, P 0.05.

MECHANICAL LOAD AND MUSCLE IGF-I IN HUMANS

E387

after CON (P 0.005). IGF-I was elevated 48 h after CON (P 0.006), but no changes were found after ECC loading.
DISCUSSION

Fig. 2. Sample Southern blot results for IGF-I, IGFBP-4, and AR mRNAs before loading (PRE), 48 h after concentric loading (CON), and 48 h after eccentric loading (ECC). Samples from each subject were run in adjacent lanes as depicted here. A: ECC performed 610 days before CON; B and C: CON performed 610 days before ECC. GAP, glyceraldehyde phosphate dehydrogenase.

after ECC (P 0.05) with a similar trend after CON (P 0.12). Planned comparisons identied a 57% reduction in IGFBP-4 mRNA after ECC (P 0.01) and a strong trend toward decreased IGFBP-4 after CON (P 0.06). AR mRNA concentration was elevated (P 0.05) 102 and 63% after CON and ECC, respectively. Southern blot results for IGF-I, IGFBP-4, and AR mRNAs are shown in Fig. 2. Serum levels of total testosterone and IGF-I are shown in Table 4. Because testosterone was not within the detectable range for several samples among the three women, testosterone data are presented only for men. Main time effects were found for testosterone after both CON (P 0.001) and ECC (P 0.026) loading. After both exercise bouts, total testosterone levels tended to fall. Testosterone was depressed after CON at 1, 2, 4, and 8 h after exercise (P 0.05). After ECC, testosterone was lower than baseline for all time points beyond 0.5 h postexercise (P 0.05) except at 24 h. A time effect was noted for serum IGF-I (n 10)

Based on previous work linking muscle hypertrophy with local availability of IGF-I (2) or with ECC loading (17), the primary objective of this study was to determine whether ECC loading was associated with increased availability of local IGF-I in the working muscles. We noted a high degree of variability in IGF-I mRNA concentrations between subjects and across time. As a result, no main time effect was found in the ANOVA model. However, mean IGF-I mRNA levels trended upward 48 h after both CON and ECC loading, with the increase after ECC being signicant as determined by planned comparison. To our knowledge, this is the rst evidence in humans of muscle IGF-I transcriptional modulation after a single bout of heavy resistance exercise. Using immunohistochemistry, two recent investigations report increased IGF-I staining in human muscle after several bouts of high-intensity exercise (18, 31). Hellsten et al. (18) found an increased number of IGF-I immunoreactive capillaries and satellite cells in vastus lateralis muscle after 7 days of intense military training including 150 km of terrain marching with a 30-kg overload (i.e., gear). The increased IGF-I staining was accompanied by a sixfold increase in serum CK activity. Singh et al. (31) found a marked 491% increase in IGF-I staining in vastus lateralis of frail elderly after 10 wk of leg resistance training. Greater IGF-I staining at baseline was predictive of the magnitude of muscle hypertrophy (r 0.70). It should be noted that both of these exercise protocols included both CON and ECC loading. IGF-I immunoreactivity within myobers has also been shown to rise 4 days after 5 days of successive electrically stimulated ECC loading in the rat tibialis anterior (36). Others have reported increases in muscle IGF-I peptide with no concomitant increase in IGF-I gene expression after 5 days of endurance training in rats (11). The optimum time point for an acute postexercise elevation in muscle IGF-I content is unknown. Increased myober IGF-I mRNA has been detected by in situ hybridization immediately after 6 days of continuous stretch in rabbits (37). We elected to biopsy the vastus lateralis muscle 48 h after the loading bouts to

Table 4. Serum hormone levels before and after CON and ECC loading
Baseline 0.5 h 1h 2h 4h 8h 24 h 48 h

Testosterone* (ng/dl) CON ECC IGF-I (ng/ml) CON ECC Values are means

621 589 349 412

63 60 37 70

555 580 317 361

62 77 22 43

504 486 345 371

50 65 36 45

417 447 358 361

45 63 32 47

416 444 352 350

44 46 44 34

451 451 372 373

48 36 49 35

586 496 385 379

67 73 35 41

542 472 416 386

40 46 49 34

SE. * Testosterone data for men only (n

7). Main time effect, P

0.05; different from baseline, P

0.05.

E388

MECHANICAL LOAD AND MUSCLE IGF-I IN HUMANS

make the sample time coincide with previous reports of 1) increased mixed muscle protein synthesis after an acute bout of heavy resistance exercise (29) and 2) drops in isometric force and increases in myobrillar disruption after ECC loading (19). Although a direct link is not apparent, our data indicate that important changes in the local IGF-I system occur during the acute phase of tissue repair and regeneration after mechanical loading. Singh et al. (31) and Yan et al. (36) studied muscle 46 days after exercise and found large increases in immunoreactivity for IGF-I peptide, suggesting that the posttranscriptional effect continues for at least a few days. The biological activity of IGF-I is regulated by a family of six IGFBPs in serum and in extravascular tissues. In skeletal muscle, the myogenic effects of IGF-I are inhibited by IGFBP-4, which appears to exert a strong afnity for IGF-I, thereby reducing the level of free IGF-I (9, 34). In cultured myoblasts, both proliferation and differentiation are inhibited by overexpression of IGFBP-4 (9). We report a 56% drop in IGFBP-4 mRNA after ECC loading and a 31% (nonsignicant) decrease after CON. Additional components of the IGF-I system not measured in this study (including other IGFBPs and the type 1 IGF receptor) can certainly modulate the availability of IGF-I in muscle. Because, however, IGFBP-4 has been consistently shown to inhibit IGF-I action in muscle cells, we suggest that the decrease in IGFBP-4 mRNA concentration coupled with the increase in IGF-I mRNA found in this study would promote an increase in IGF-I availability within the muscle after mechanical loading. These changes occurred in the absence of any increase in serum IGF-I. This is not surprising. Several others have reported that muscle IGF-I activity is independent of changes in serum IGF-I (1). Further, this is not the rst report of unchanging serum IGF-I after a single bout of heavy resistance exercise (25). Although evidence linking local IGF-I availability to growth potentiation in muscle is mounting (1), the mechanism by which mechanical load modulates IGF-I expression is not clear. In his comprehensive review, Adams (1) proposes that this load-sensitive IGF-I in muscle works by paracrine or autocrine action to induce satellite cell proliferation and differentiation, followed by fusion of differentiated myoblasts to hypertrophying myobers. The process of myoblast fusion is thought to maintain myonuclear domain size and thus the capacity for muscle protein synthesis. Indeed, myonuclear domain is maintained during mechanical loadinduced myober hypertrophy (27). The model proposed by Adams (1) is well supported; however, the direct link between high-intensity mechanical load and muscle IGF-I modulation requires further investigation. Goldspink (16) has proposed a mechanotransduction mechanism involving cytoskeletal proteins during stretch. By this proposed mechanism, stretch tension on the basement membrane (e.g., laminin) physically activates intracellular signaling via a membranebound signaling molecule on the plasmalemma.

Based on the CK and soreness data after ECC loading, we suggest that myobrillar disruption and/or sarcolemma damage may play a role in activating the muscle IGF-I system. We have previously shown an increase in serum levels of acidic broblast growth factor (aFGF) after resistance exercise with an ECC component (7). The FGF response paralleled an increase in serum CK activity. FGFs are powerful proliferative agents in myoblast cultures (12) and have been shown to induce myotube hypertrophy (8). Additionally, FGFs are known to be released directly from mechanically wounded muscle cells (8). Because we report increased IGF-I mRNA after 48 h and others have shown increased IGF-I peptide after 46 days (31, 36), it is possible that the IGF-I response is activated by an earlier local release of FGF or some other factor(s). Based on these data, we suggest that IGF-I activity is somehow linked to mechanisms involved in tissue regeneration after mechanical damage. One problem with this hypothesis, however, is the fact that ECC exercise-induced muscle damage does not necessarily result in hypertrophy. Marked muscle damage follows unaccustomed endurance exercise with a large ECC component (e.g., downhill running) (32); however, endurance exercise is not a potent hypertrophic stimulus (23). The ECC damage associated with high-tension contractions during resistance exercise is somehow uniquely different. This is an attractive area for further study. Another possible mechanism for increasing muscle IGF-I availability is via androgen activation. Increased muscle protein synthesis after testosterone therapy in older men has been shown to be associated with increased IGF-I mRNA concentration and a concomitant reduction in IGFBP-4 mRNA concentration (34). Furthermore, inducing androgen deciency in young men decreases muscle IGF-I mRNA concentration and causes muscle atrophy (26). These data suggest that IGF-I action in muscle is secondary to androgen activity. In the seven men in the present study, serum total testosterone was not increased but fell after both CON and ECC loading. The testosterone response to resistance exercise is highly variable (24). The fall in serum testosterone across time in the present study most likely reects diurnal variation, as testosterone levels are highest in the morning and decrease into the evening hours. In support of this, testosterone values were not different from the initial morning levels 24 h after both exercise bouts. However, we found a substantial increase in AR mRNA concentration after both CON and ECC loading. To our knowledge, this is the rst report of such an increase in humans after an acute exercise bout. AR content has been shown to increase in type II muscle after resistance training in rats (10). Also in rats, gastrocnemius hypertrophy by electrical stimulation has been associated with an increased number of muscle ARs (20). Our data indicate both AR and IGF-I mRNA concentrations are upregulated by heavy mechanical load. Although the mechanism is not clear, muscle androgen and IGF-I activities may be related.

MECHANICAL LOAD AND MUSCLE IGF-I IN HUMANS

E389

We report for the rst time changes in muscle mRNAs associated with tissue growth and repair after a single bout of resistance exercise in humans. The novel approach of studying high-intensity CON and ECC loading separately enabled us to test the inuence of ECC action on the IGF-I and AR responses to mechanical load. The results indicate that high-intensity lengthening and shortening contractions both induce muscle IGF-I and AR gene transcription. Moreover, the enhanced IGF-I activation after ECC loading supports the concept that IGF-I is somehow involved in tissue regeneration after mechanical load-induced damage.
We thank the participants for their tolerance and tireless effort. We thank Leo Wright of Wright Exercise Equipment in Birmingham, Alabama, for providing the exercise equipment used in this study. This work was funded by a University of Alabama at Birmingham Provosts Faculty Development Award (M. M. Bamman), National Institutes of Health GCRC Grant M01-RR-00032, and National Institutes of Health 1 R01-AG/AR-11000. REFERENCES 1. Adams G. Role of insulin-like growth factor-I in the regulation of skeletal muscle adaptation to increased loading. Exerc Sports Sci Rev 26: 3160, 1998. 2. Adams G and McCue S. Localized infusion of IGF-I results in skeletal muscle hypertrophy in rats. J Appl Physiol 84: 1716 1722, 1998. 3. Adams GR, Duvoisin MR, and Dudley GA. Magnetic resonance imaging and electromyography as indexes of muscle function. J Appl Physiol 73: 15781583, 1992. 4. Allen DL, Linderman JK, Roy RR, Grindeland RE, Mukku V, and Edgerton VR. Growth hormone/IGF-I and/or resistive exercise maintains myonuclear number in hindlimb unweighted muscles. J Appl Physiol 83: 18571861, 1997. 5. Bamman MM, Clarke MS, Feeback DL, Talmadge RJ, Stevens BR, Lieberman SA, and Greenisen MC. Impact of resistance exercise during bed rest on skeletal muscle sarcopenia and myosin isoform distribution. J Appl Physiol 84: 157163, 1998. 6. Bamman MM, Hunter GR, Stevens BR, Guilliams ME, and Greenisen MC. Resistance exercise prevents plantar exor deconditioning during bed rest. Med Sci Sports Exerc 29: 1462 1468, 1997. 7. Clarke MS, Bamman MM, and Feeback DL. Bed rest decreases mechanically induced myober wounding and consequent wound-mediated FGF release. J Appl Physiol 85: 593600, 1998. 8. Clarke MS and Feeback DL. Mechanical load induces sarcoplasmic wounding and FGF release in differentiated human skeletal muscle cultures. FASEB J 10: 502509, 1996. 9. Damon SE, Haugk KL, Birnbaum RS, and Quinn LS. Retrovirally mediated overexpression of insulin-like growth factor binding protein 4: evidence that insulin-like growth factor is required for skeletal muscle differentiation. J Cell Physiol 175: 109120, 1998. 10. Deschenes M, Maresh C, Armstrong L, Covault J, Kraemer W, and Crivello J. Endurance and resistance exercise induce muscle ber type specic responses in androgen binding capacity. J Ster Biochem Mol Biol 50: 175179, 1994. 11. Eliakim A, Moromisato M, Moromisato D, Brasel JA, Roberts C Jr, and Cooper DM. Increase in muscle IGF-I protein but not IGF-I mRNA after 5 days of endurance training in young rats. Am J Physiol Regulatory Integrative Comp Physiol 273: R1557R1561, 1997. 12. Florini JR. Hormonal control of muscle growth. Muscle Nerve 10: 577598, 1987. 13. Florini JR, Ewton DZ, and Coolican SA. Growth hormone and the insulin-like growth factor system in myogenesis. Endocrine Rev 17: 481517, 1996.

14. Frost RA and Lang CH. Differential effects of insulin-like growth factor I (IGF-I) and IGF-binding protein-1 on protein metabolism in human skeletal muscle cells. Endocrinology 140: 39623970, 1999. 15. Gibala MJ, MacDougall JD, Tarnopolsky MA, Stauber WT, and Elorriaga A. Changes in human skeletal muscle ultrastructure and force production after acute resistance exercise. J Appl Physiol 78: 702708, 1995. 16. Goldspink G. Cellular and molecular aspects of muscle growth, adaptation and ageing. Gerodontology 15: 3543, 1998. 17. Hather BM, Tesch PA, Buchanan P, and Dudley GA. Inuence of eccentric actions on skeletal muscle adaptations to resistance training. Acta Physiol Scand 143: 17785, 1991. 18. Hellsten Y, Hansson H, Johnson L, Frandsen U, and Sjodin B. Increased expression of xanthine oxidase and insulin-like growth factor I immunoreactivity in skeletal muscle after strenuous exercise in humans. Acta Physiol Scand 157: 191197, 1996. 19. Hortobagyi T, Houmard J, Fraser D, Dudek R, Lambert J, and Tracy J. Normal forces and myobrillar disruption after repeated eccentric exercise. J Appl Physiol 84: 492498, 1998. 20. Inoue K, Yamasaki S, Fushiki T, Kano T, Moritani T, Itoh K, and Sugimoto E. Rapid increase in the number of androgen receptors following electrical stimulation of the rat muscle. Eur J Appl Physiol Occup Physiol 66: 134140, 1993. 21. James PL, Stewart CE, and Rotwein P. Insulin-like growth factor binding protein-5 modulates muscle differentiation through an insulin-like growth factor-dependent mechanism. J Cell Biol 133: 68393, 1996. 22. Jurasinski CV and Vary TC. Insulin-like growth factor I accelerates protein synthesis in skeletal muscle during sepsis. Am J Physiol Endocrinol Metab 269: E977E981, 1995. 23. Kraemer W, Patton J, Gordon S, Harman E, Deschenes M, Reynolds K, Newton R, Triplett N, and Dziados J. Compatibility of high-intensity strength and endurance training on hormonal and skeletal muscle adaptations. J Appl Physiol 78: 976989, 1995. 24. Kraemer WJ. Endocrine responses to resistance exercise. Med Sci Sports Exerc 20: S152S157, 1988. 25. Kraemer WJ, Aguilera BA, Terada M, Newton RU, Lynch JM, Rosendaal G, McBride JM, Gordon SE, and Hakkinen K. Responses of IGF-I to endogenous increases in growth hormone after heavy-resistance exercise. J Appl Physiol 79: 1310 1315, 1995. 26. Mauras N, Hayes V, Welch S, Rini A, Helgeson K, Dokler M, Veldhuis JD, and Urban RJ. Testosterone deciency in young men: marked alterations in whole body protein kinetics, strength, and adiposity. J Clin Endocrinol Metab 83: 18861892, 1998. 27. McCall G, Allen D, Linderman J, Grindeland R, Roy R, Mukku V, and Edgerton V. Maintenance of myonuclear domain size in rat soleus after overload and growth hormone/IGF-I treatment. J Appl Physiol 84: 14071412, 1998. 28. Menard MR, Penn AM, Lee JW, Dusik LA, and Hall LD. Relative metabolic efciency of concentric and eccentric exercise determined by 31P magnetic resonance spectroscopy. Arch Phys Med Rehab 72: 97683, 1991. 29. Phillips SM, Tipton KD, Aarsland A, Wolf SE, and Wolfe RR. Mixed muscle protein synthesis and breakdown after resistance exercise in humans. Am J Physiol Endocrinol Metab 273: E99E107, 1997. 30. Shefeld-Moore M, Urban R, Wolf S, Jiang J, Catlin D, Herndon D, Wolfe R, and Ferrando A. Short-term oxandrolone administration stimulates net muscle protein synthesis in young men. J Clin Endocrinol Metab 84: 27052711, 1999. 31. Singh MA, Ding W, Manfredi TJ, Solares GS, ONeill EF, Clements KM, Ryan ND, Kehayias JJ, Fielding RA, and Evans WJ. Insulin-like growth factor I in skeletal muscle after weight-lifting exercise in frail elders. Am J Physiol Endocrinol Metab 277: E135E143, 1999. 32. Sorichter S, Mair J, Koller A, Gebert W, Rama D, Calzolari C, Artner-Dworzak E, and Puschendorf B. Skeletal troponin I as a marker of exercise-induced muscle damage. J Appl Physiol 83: 107682, 1997.

E390

MECHANICAL LOAD AND MUSCLE IGF-I IN HUMANS 36. Yan Z, Biggs RB, and Booth FW. Insulin-like growth factor immunoreactivity increases in muscle after acute eccentric contractions. J Appl Physiol 74: 4104, 1993. 37. Yang S, Alnaqeeb M, Simpson H, and Goldspink G. Cloning and characterization of an IGF-1 isoform expressed in skeletal muscle subjected to stretch. J Muscle Res Cell Motil 17: 487495, 1996. 38. Yarasheski KE, Zachwieja JJ, Campbell JA, and Bier DM. Effect of growth hormone and resistance exercise on muscle growth and strength in older men. Am J Physiol Endocrinol Metab 268: E268E276, 1995.

33. Taaffe DR, Jin IH, Vu TH, Hoffman AR, and Marcus R. Lack of effect of recombinant human growth hormone (GH) on muscle morphology and GH-insulin-like growth factor expression in resistance-trained elderly men. J Clin Endocrinol Metab 81: 421425, 1996. 34. Urban R, Bodenburg Y, Gilkinson C, Foxworth J, Coggan A, Wolfe R, and Ferrando A. Testosterone administration to elderly men increases skeletal muscle strength and protein synthesis. Am J Physiol Endocrinol Metab 269: E820E826, 1995. 35. Wong T and Booth F. Protein metabolism in rat tibialis anterior muscle after stimulated chronic eccentric exercise. J Appl Physiol 69: 17181724, 1990.