Exercise in Maintenance Hemodialysis Patients Induces Transcriptional Changes in Genes Favoring Anabolic Muscle

Results

Eighty MHD patients completed baseline testing, including a muscle biopsy, and then underwent randomization to one of the four groups: Endurance training (ET), strength training (ST), endurance plus strength training (EST) or no training (NT). Seventeen patients (six ET, two ST, five EST, and four NT) dropped out because of illness, noncompliance, moving away, or possibly reluctance to undergo another muscle biopsy. Of these, two ET, one ST, and three EST patients dropped out before (five patients) or only after 1 wk of exercise training. As a result of a laboratory accident, one or more of the muscle biopsy specimens from 12 patients were destroyed. This left data from 51 patients who completed the study.

Characteristics of the 51 MHD patients and 20 normal control subjects are shown in Tables 1 and 2. MHD patients exercised for 21.5 ± 0.7 wk. No patient had insulin-dependent diabetes. The normal individuals had normal serum creatinine levels, and none had a history of kidney disease or hypertension. Mean age and serum albumin were similar in the two groups; hemoglobin and hematocrit were only slightly lower in the MHD patients, which probably reflects the effects of erythropoietin treatment.

During the course of exercise training, the ET patients increased the duration of their ET exercises from 18 ± 1 to 35 ± 3 min, an increase of 96%. Mean change in work rate was 16 ± 9 W (35% increase). The product of these data provides a measure of the total work completed. At baseline (week 1 of training), these individuals completed 155 ± 28 kJ of work. At the end of training, defined as the last week completed, these individuals completed 330 ± 73 kJ work, an improvement of 113% in leg-cycling work. The ST patients increased the total weight lifted per week from 4001 ± 660 kg at week 1 to 15,463 ± 2073 kg during the last week of training for a 286% increase in total training volume (kg lifted per repetition × repetitions × sets). By comparison, the EST patients, between the first and last week of training, increased the duration of their ET exercises from 10 ± 0 to 18 ± 2 min (80%), the work rate from 42 ± 12 to 72 ± 16 watts (71%), and the total work from 103 ± 53 to 245 ± 63 kJ (138%). These individuals also increased the total weight lifted from 2976 ± 1069 kg during week 1 to 7423 ± 1607 kg during the final week, a 149% rise. It should be noted that these data are based on sample sizes of six, seven, and five individuals for the ET, ST, and EST groups, respectively. Regretfully, training data files of other patients are not available.

Serum C-reactive protein (CRP), TNF-α, and IL-6 concentrations in the MHD patients did not change with exercise training or in the NT group (Table 2). Serum CRP and IL-6, obtained from all four pre-exercise MHD groups combined, did not differ from the initially obtained normal values. Pre-exercise serum TNF-α was significantly greater than normal (P < 0.001). Dietary energy and protein intakes, before exercise, expressed in kcal/d or g/d or kcal/kg per d or g/kg per d, were significantly lower for the four groups combined (P < 0.05 to 0.01; Table 2) except for protein intake, expressed in g/kg per d. Dietary energy or protein intakes did not change with exercise training in any group or in the three exercising groups combined.

Skeletal muscle IGF-IEa mRNA in the four groups of MHD patients, before the onset of exercise training, was significantly lower than in the normal control subjects (Table 3). With exercise training, muscle IGF-IEa mRNA tended to increase in all groups of MHD patients, and this increase was significant in the ST and EST groups and in all three exercising MHD groups combined (Table 3). Skeletal muscle IGF-IEc mRNA also tended to be lower than normal in the MHD groups, although these values were not significantly different (Table 3). With exercise training, IGF- IEc tended to increase in all three groups, and this increase was significant in the EST group and in all three exercising groups combined (Table 3). The IGF-I receptor (IGF-IR) mRNA before exercise training in the four MHD groups combined was significantly lower than in the normal group. By the end of exercising, IGF-IR mRNA increased significantly in the three exercising groups combined.

IGF-II mRNA in the MHD patients before exercise training was significantly lower than the normal values (Table 3). By the end of exercise training, IGF-II mRNA had risen significantly in the EST group as well as in the three exercising groups combined. Before starting exercise training, IGF-IIR mRNA was not different from normal in the combined four groups of MHD patients. IGF-IIR mRNA did not change significantly with exercise training in any group. At the end of training, IGF-IIR levels were not different from normal in any group.

IGF binding protein-2 (IGFBP-2) mRNA in the MHD patients, before commencing exercising, was lower than normal (Table 3). With exercise training, IGFBP-2 mRNA increased significantly only in the ET group and in all exercising groups combined. IGFBP-3 mRNA in the MHD patients, before exercising, was not significantly lower than normal values. With exercise training, IGFBP-3 mRNA rose significantly in the three groups of exercising patients combined.

Before commencement of exercise training, there were no significant differences in the myostatin mRNA levels between the four groups of MHD patients combined and the normal control subjects (Table 3). After exercise training, there was some decrease in myostatin mRNA in each of the three exercising groups (ET 22%; ST 23%; EST 39%) and some increase in the NT group (29%). This decrease was statistically significant (P < 0.05) for all three groups combined. The fall in myostatin mRNA in the ET, ST, and EST groups combined was also significantly different from the change in the NT group (P < 0.05).

In the 29 MHD patients who had mid-training muscle biopsies, mRNA was also assayed for IGF-IEa, IGF-IEc, IGF-IR, IGF-II, IGF-IIR, IGFBP-2, IGFBP-3, and myostatin. In general, the same changes in mRNA levels present in the final muscle specimens were already present as trends in the mid-training biopsies (data shown only in Figure 1). By coincidence, the mean IGF-IEc values in the patients who underwent three muscle biopsies tended to be greater than in the patients who underwent two biopsies. We are unable to identify a cause for this discrepancy (e.g., there were no differences between the two groups with regard to age, gender, presence or absence of diabetes, or proportions of patients undergoing each type of exercise intervention).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Time course of IGF-IEc mRNA levels (A) and IGF-II mRNA levels (B) in right vastus lateralis muscle of MHD patients undergoing various types of exercise training or no exercise training. Lines indicate mean values for each group of MHD patients.

Serum IGF-I concentrations before exercise training were not different than in the normal group (Table 4). Serum IGF-I did not change significantly during exercise in any of the exercise training groups. Before exercise training, serum IGF-II was significantly greater than normal values in all four groups of MHD patients combined. Serum IGF-II levels did not change in any group after exercise training. Muscle IGF-I was greater than normal in the four groups of MHD patients combined (Table 4). Muscle IGF-I rose significantly with exercise training in the ET and ST groups and in the three exercising groups combined and by a borderline level of significance (P = 0.07) in the EST group (Table 4). Muscle IGF-II did not differ from the normal values in the MHD patients, before training, and did not change in any group of MHD patients during the study.

Possible gender differences in the response of the gene transcripts to exercise were examined. When all three groups of exercising MHD patients combined were analyzed together, the men, as compared with women, displayed a significantly greater increase in muscle mRNA for IGF-IEc and IGFBP-3 (P < 0.05 for each comparison) during exercise training. The changes in the muscle gene transcripts for IGF-IEa, IGF-IR, IGF-II, and IGFBP-2, although not statistically significant in the men versus the women, tended to be greater in the men, whereas the fall in mRNA for the antigrowth factor myostatin, which again was not statistically significant in men versus women, tended to be substantially greater in women. There was a trend in the NT female MHD patients for the mRNA levels to increase more between the first and last muscle biopsies for IGF-IEa, IGF-II, and myostatin. IGF-IR mRNA increased to a statistically significantly greater degree (P < 0.05) in the NT female versus male MHD patients. IGFBP-3 mRNA tended to fall more in these NT women.

No MHD patient or normal individual manifested edema. There were no significant differences in the pre-exercise anthropometric measurements in the four MHD patients combined versus the normal control subjects when the two genders were considered together. In the male patients considered separately, the same anthropometric similarities and differences were found except for a lower proximal thigh muscle area in the four MHD groups combined at baseline (187.7 ± 7.5 cm²) as compared with the normal men (215.3 ± 11.2 cm²; P = 0.044). The number of female patients studied was considered to be too small for separate statistical comparisons. After exercise training, there was a significant decrease in triceps and subscapular skinfold thicknesses in all three exercise groups combined (Table 5). Most striking, after exercise, anthropometry-derived percentage of body fat and total body fat decreased and fat-free mass (FFM) increased significantly in all exercising patients combined.

The dual-energy x-ray absorptiometry (DEXA) measurements showed that in the total group of MHD patients at baseline, there was a decrease in FFM and total mass in the left leg, right leg, and both legs combined and in bone mineral concentration as compared with the normal control subjects (Table 6). With exercise training in the three exercising MHD groups combined, there was a significant decrease in bone mineral concentration (Table 6). Bioelectrical impedance measurements showed no difference in total body FFM or fat mass in the combined four groups of MHD patients, before exercise, as compared with the normal values (Table 7). There also was no change in total body FFM or fat mass, expressed as kilograms of weight or as percentage of total body mass, between the pre- and end of study measurements in the ET, ST, EST, or NT groups or in the three exercising groups combined (Table 7).

When the 43 MHD patients who did not have diabetes were analyzed separately, the identical muscle mRNA, before exercise training, were found to be significantly different from the control subjects, as was observed for the entire group of 51 MHD patients with and without diabetes. Eight MHD patients had non–insulin-dependent diabetes; they all were assigned to one of the three exercising groups. In general, they showed the same trends in mRNA levels as the MHD patients without diabetes, except that the patients with diabetes, in comparison with normal individuals (none of whom had diabetes), had significantly lower pre-exercise IGF-IIR and myostatin mRNA and higher postexercise IGF-IEc mRNA. In the MHD patients with diabetes as compared with the MHD patients without diabetes, muscle IGF-II mRNA before exercise training was higher (13.4 ± 1.7 versus 6.9 ± 0.8; P = 0.003) and myostatin mRNA was lower (median [25th and 75th percentiles] 8.2 [4.2, 12.2] versus 16.4 [7.5, 33.7]; P = 0.049). Muscle IGF-I peptide concentrations in the MHD patients without and with diabetes, before exercising, were 131.6 ± 8.7 and 110.7 ± 12.4 pg/mg wet muscle tissue, respectively (NS). There was no clear association between the results of the postexercise testing and the frequency or duration of hospitalization, possibly because patients were required to continue to exercise train for a minimum of 2 wk after their last hospitalization before undergoing their final exercise testing.

Discussion

This study indicates that in right vastus lateralis muscle of sedentary MHD patients, there were decreased mRNA levels for a number of factors that may stimulate protein anabolism, suppress protein degradation, and promote accrual of protein. In particular, the mRNA for IGF-IEa, IGF-IR, IGF-II, and IGFBP-2 were decreased. Exercise training was associated with an increase in muscle mRNA for many of these growth factors, namely, IGF-IEa, IGF-IEc, IGF-IR, IGF-II, IGFBP-2, and IGFBP-3. Moreover, myostatin mRNA fell significantly. As might be expected, there was no change in the mRNA for any growth factor in the NT patients. As a result of these gene transcription changes, most of the mRNA levels were closer to normal at the end of exercise training. The only exceptions were IGFBP-2 mRNA, which after exercise training tended to be greater than normal, and myostatin mRNA, which tended to be less than normal. These findings indicate that in clinically stable, sedentary MHD patients, long-term exercise training induces a pattern of changes in gene transcription that would be likely to promote protein synthesis and reduce protein degradation.

Muscle IGF-I peptide was significantly increased whereas muscle IGF-II protein was normal in the MHD patients before they exercised. The significant further increase in muscle IGF-I protein with exercise in the ET and ST groups and in the three exercising groups combined (Table 4) is consistent with the rise in the skeletal muscle mRNA levels for IGF-IEa and IGF-IEc. These results provide further evidence that exercise training engenders an anabolic response in sedentary MHD patients. The absence of a rise in muscle IGF-II despite the rise in IGF-II mRNA with exercising may reflect posttranscriptional or posttranslational factors. It is also possible that the IGF-II protein levels rose transiently during exercise training and then fell back to normal by the time of the biopsy.

Skeletal muscle IGF-I promotes skeletal muscle protein synthesis and hypertrophy, preserves architecture, and suppresses protein degradation.¹¹ It seems that circulating IGF-I has little effect on muscle hypertrophy,¹² and most skeletal muscle IGF-I is synthesized in situ.¹³ In this regard, Nindl et al.¹⁴ described a decrease in serum total IGF-I, ternary IGF-I (IGF-I, IGFBP-3, and the acid labile substance), and the IGF-I/IGFBP-3 ratio during resistance training in 10 maintenance dialysis patients. In this study, serum IGF-I was normal and serum IGF-II was elevated before exercising, and neither peptide changed with exercise in our MHD patients (Table 4). Similar to our findings, MacDonald et al.¹⁵ described normal serum IGF-I and IGFBP-3 in 17 MHD patients; however, muscle IGF-I was low in their patients, whereas we observed increased muscle IGF-I that rose further with exercise (Table 4). The low muscle IGF-I may have been due to the wasted state of their patients, as indicated by anthropometry and body composition measurements, whereas our patients did not seem wasted. This discrepancy might be explained by the fact that our MHD patients were a more selected group of healthier individuals because they were required to undergo exercise training in our study.

Three isoforms of IGF-I are recognized in humans: IGF-IEa, IGF-IEb, and IGF-IEc. These isoforms are generated by splice variants of the IGF-I gene.¹⁶ IGF-IEa and IGF-IEc are present in skeletal muscle, whereas IGF-IEb is found primarily in liver and is also detected in muscle. These reductions in mRNA for IGF-IEa, IGF-IR, and IGF-II may predispose to the myofiber atrophy and sarcopenia that are described in MHD patients^17–19 and reduce their ability to remodel muscle protein. Although IGF-IEc mRNA was not significantly reduced in the MHD patients, the same trend toward lower mRNA values was observed for this growth factor as for IGF-IEa mRNA (Table 3). IGF-IEc, also called mechanogrowth factor, is of special interest for exercise training, because it may play a particularly important role in promoting skeletal muscle physical capacity and hypertrophy.²⁰

Myostatin is inhibitory to skeletal muscle protein synthesis and muscle hypertrophy.²¹ The finding that myostatin fell with exercise training is again consistent with a protein synthetic response to exercise training. These results are consistent with the findings of Sun et al.,²² who described an increase in IGF-I mRNA and a decrease in myostatin mRNA in plantaris muscle of rats with chronic renal failure after the rats were subjected to a chronically increased workload on that muscle.

The dietitian-performed anthropometry indicated that the MHD patients who underwent exercise training experienced a decrease in triceps and subscapular skinfold thicknesses, percentage of body fat, and total body fat mass and an increase in FFM (Table 5). The DEXA and bioelectrical impedance studies showed no changes in body composition, except for reduced bone mineral density with exercise training observed by DEXA. The reason for this discrepancy is not known. Because DEXA is often considered a more reliable measure of body composition and because the DEXA and bioelectrical impedance measurements were in agreement, we conclude that there probably were no changes in FFM or fat mass in these patients. Although no patient or normal individual had evidence of edema, MHD patients cannot regulate their water balance well; therefore, it is possible that undetected increases or reductions in total body water may have obscured our ability to detect the effects of exercise on muscle mass by DEXA or bioelectrical impedance measurements.

The increase in skeletal muscle mRNA levels for growth factors in our exercising MHD patients is similar to those reported with exercising normal humans and various animals^7–10; however, in normal humans and animals, the type of training that they undergo (e.g., endurance versus strength) influences the specific muscle mRNA levels that increase with exercise,^7–10 and skeletal muscle hypertrophy tends to occur only with resistance training.^7–10

These distinctions were not observed in our MHD patients. After exercise training, there was a significant increase in mRNA levels for one growth factor each with ET and ST and with three growth factors for EST (Table 3). The mRNA that increased with ET was for IGFBP-2, which might not be considered a growth factor. Moreover, in groups for which the mRNA levels did not change significantly, the mean values for mRNA usually moved in the same direction, as was found for the significant changes in mRNA.

There are several possible explanations for the similar responses in all three groups of exercising MHD patients. First, because of the severe degree of deconditioning and comorbidity of the MHD patients, most patients were not able to exercise vigorously; this may have obscured the ability of the various exercise groups to develop different responses to exercising. Second, the patients undergoing ST or EST might have shown more muscle hypertrophy if their training had been based on a high percentage of the maximum weight exertion performed one time (1-RM) rather than the 5-RM, which reduced their resistance load. We chose this lower resistance technique as a safety measure to reduce the risk for tendon rupture and bone fractures that have been reported in MHD patients who were subjected to high-resistance loads.^6,23,24 Also, for this latter reason and because they were deconditioned and our focus was on developing exercise capacity rather than muscle hypertrophy, the ST and EST patients generally spent their first 8 wk performing less intensive exercise training and increased the intensity of their resistance training slowly. Our objective was more focused on improving strength than muscle hypertrophy per se; therefore, we did not use such techniques as multiple sets of multiple exercises with a relatively short rest interval and eight to 12 repetitions to muscle failure, which are more effective at recruiting type IIx fibers (previously identified as type IIb or fast-twitch glycolytic fibers) in human skeletal muscle (see Sale²⁵ for motor unit recruitment patterns). Similarly, the ET patients, in general, did not exercise very vigorously (see also the Concise Methods section).

Third, many MHD patients seem to have a myopathy that might impair the response to exercise training.^2,26–28 Fourth, there may be metabolic factors in advanced renal failure that inhibit or facilitate certain mRNA responses so that the effects of different types of exercise training are obscured.²⁹ In this regard, all exercising groups manifested evidence of chronic inflammation as indicated by the significantly elevated serum TNF-α and the NS trend toward raised serum CRP and IL-6.

Finally, because the sample sizes in the various exercising groups were small, there is the possibility of a type B statistical error (see the Concise Methods section). Against this possibility is the finding that the trends toward changes in mRNA levels with exercising were virtually the same in all groups. Indeed, when the responses of all three exercising groups were pooled, a significantly increased mRNA was observed for every growth factor measured, except for myostatin, a growth inhibitor that decreased significantly, and for the IGF-II receptor, which did not change.

Our results indicated a greater increase in muscle gene transcripts for IGF-IEc and IGFBP-3 in men versus women in all three exercise training groups combined. Bamman et al.³⁰ described a greater increase in strength and in myofiber hypertrophy in older men versus older women who where undergoing similar resistance (strength) training. These differences could not be attributed to different expressions of IGF-I or myogenin mRNA, to serum IGF-I or dehydroepiandrosterone sulfate concentrations, or to changes in serum testosterone levels.

A limitation of this study is the relatively small numbers of patients in each group, which prevented a rigorous comparison of possible differences in responses to exercise training in the three exercising groups. Also, the uneven number of dropouts in the various groups could reflect a treatment bias; that most of the people who dropped out did so before commencing exercise training or were in the NT group makes this a less likely possibility. Finally, all muscle mRNA and peptide measurements were from the right vastus lateralis muscle, which may not reflect the growth factor mRNA or peptide responses in other skeletal muscles.

The exercise training regimens in this study may have also reduced muscle protein degradation. Muscle mRNA increased for some growth factors that suppress protein degradation as well as stimulate protein synthesis. Such growth factors include IGF-I and IGF-II.³¹ In addition, the increased mRNA for the IGF-I receptor or IGFBP might have inhibited protein degradation. In this regard, we have observed a significant reduction in the 14-kD fragment of actomyosin in our patients undergoing ET and EST but not ST or NT.³² Du et al.³³ described this 14-kD fragment as an indicator of the degradation rate of actomyosin protein in skeletal muscle via the caspase-3 system.

It is somewhat puzzling that our body composition measurements did not show clear evidence for increased muscle mass with exercise training; however, a rise in muscle mRNA levels and growth factor proteins with exercise training may increase protein synthesis and reduce degradation of selected proteins without engendering muscle hypertrophy. In normal mammals, ET or ST increases the synthesis of many skeletal muscle proteins that are associated with enhanced physical capacity.⁹ These changes include increased mitochondria number, particularly of the subsarcolemmal type; vascular endothelial growth factor; capillary density; fibronectin; carnitine palmitoyltransferase; and GLUT4, among others.^7,8,10,34,35 There is a redistribution of myosin heavy-chain isoforms and of the activities of many other enzymes involved with energy generation.^7,8,10 Many of these changes are associated with an increase in skeletal muscle mRNA levels for the respective proteins and also for a number of growth factors, including IGF-I and IGF-II.^7,8,10,34,35 These structural and biochemical changes in skeletal muscle are thought to contribute to the increased physical exercise capacity that occurs with ET or ST. To our knowledge, this study is the first to show that similar changes in skeletal muscle mRNA in MHD patients for certain growth factors can occur in response to ET and/or ST. It is possible that these changes in skeletal muscle mRNA observed with exercise training in this study may contribute to structural and biochemical changes in muscle that might improve the exercise capacity of MHD patients.

Concise Methods

Disclosures

None.