Endocrinology and Metabolism

Increased protein synthesis after acute IGF-I or insulin infusion is localized to muscle in mice

Tor H. Bark, Margaret A. McNurlan, Charles H. Lang, Peter J. Garlick


The purpose of the present study was to determine the effect of acute administration of insulin-like growth factor I (IGF-I) or insulin on in vivo protein synthesis in muscle and other organs in fasted mice and to compare this response with that produced by feeding. Recombinant IGF-I (3.3 nmol prime, 3.33 nmol/h) or insulin (0.056 nmol/h) was infused intravenously for 60 min along with glucose to prevent hypoglycemia. Fractional rates of tissue protein synthesis (FSR) were determined by injection of [2H5]phenylalanine (25 mg/100 g body wt, 40% enriched). Both IGF-I and insulin caused a 25% increase in FSR of heart (P < 0.001) and soleus muscle (P < 0.05) and a 65% increase in gastrocnemius and plantaris muscle (bothP < 0.001), thus restoring rates to those seen in fed animals. A fivefold lower dose of IGF-I also stimulated protein synthesis in gastrocnemius muscle and heart (bothP < 0.05) but not in soleus muscle. No significant effects of IGF-I on FSR were detected in liver, kidney, spleen, proximal small intestine, colon, lung, or brain. The results indicate that the ability of an overnight fast to decrease protein synthesis and the acute effects of insulin and IGF-I to stimulate protein synthesis are restricted to skeletal and cardiac muscles.

  • [2H5]phenylalanine
  • feeding
  • liver
  • intestine
  • kidney

insulin-like growth factor I (IGF-I) has been demonstrated to enhance protein synthesis and slow protein degradation (3, 6, 15, 21), but its role in the overall regulation of protein metabolism has yet to be elucidated. After trauma, there is a fall in plasma IGF-I (26) that is accompanied by a loss of body protein (11). Moreover, during growth, IGF-I levels are high, in association with high rates of protein synthesis (25). However, the involvement of IGF-I in tissue protein homeostasis in normal adults, particularly in relation to insulin and the assimilation of nutrients from meals, is not at present understood.

Infusion of IGF-I into humans for 4 h has previously been shown to stimulate protein synthesis in skeletal muscle (6) and in the whole body (21), but there appears to be little information on whether IGF-I acutely affects protein synthesis in tissues other than skeletal muscle. In studies on mice, injection of IGF-I 2.5 h before mice were killed stimulated muscle protein synthesis, but the rate did not approach that seen in fed animals (22). However, this blunted response may have resulted from the relatively long duration between the bolus injection of IGF-I and the measurement of protein synthesis, during which time the circulating concentration of IGF-I had returned to basal level. There is more information on the effect of insulin on protein synthesis. In growing rats, a stimulatory effect has been demonstrated in muscle, but this response appears to be absent from adults (2). The purpose of the present study, therefore, was to investigate whether IGF-I or insulin, given as a continuous infusion for 1 h, can acutely stimulate rates of protein synthesis to the values seen in fed animals and to compare the stimulation in skeletal muscles with different fiber types with the effects in a variety of other tissues.



Male Swiss-Webster mice 9–10 wk in age and weighing ∼35 g (Taconic Farms, Germantown, NY) were housed in a controlled environment and provided with standard chow and water ad libitum for 7 days before experimental procedures. When animals were fasted, chow was withdrawn at 2300 the evening before the study.

Experimental procedures.

The study was performed in four experiments. In the first experiment, two groups of mice, each containing eight animals, were used. One group received a primed intravenous infusion of IGF-I, and the second group received an equal volume of vehicle (0.25% BSA). Both groups were fasted. In the second experiment, three groups of mice, each containing six animals, were used. One group was fasted and received an intravenous infusion of IGF-I, as in experiment 1. The second group was fasted, whereas the third group served as a fed control group. The latter two groups were infused with an equal volume of vehicle. In the third experiment, two groups of overnight-fasted mice were used. The first group (n = 7) received an intravenous infusion of insulin, whereas the second group was infused with an equal volume of saline containing 0.25% BSA (n = 6). The fourth experiment also included two groups of overnight-fasted animals (n = 8 mice/group). One group received a primed infusion of IGF-I, as inexperiment 1, but at a fivefold lower dose, and the other group received saline containing 0.25% BSA.

On the morning of the experiment, animals were lightly restrained by placing them in separate plastic tubes. A tail vein was cannulated, using a 27-gauge needle attached to polyethylene tubing (PE-50). The mice were infused intravenously with 0.9% saline containing either IGF-I (Genentech, South San Francisco, CA) plus 0.25% BSA or only BSA for 60 min. IGF-I was administered as a primed constant infusion (25 μg + 25 μg/h in experiments 1 and2; 5 μg + 5 μg/h inexperiment 4; 25 μg is equivalent to 3.3 nmol); time-matched control animals received an equal volume (100 μl + 0.5 ml/h) of 0.25% BSA. In the third experiment, the animals were infused intravenously for 60 min with 0.9% saline-BSA with or without bovine insulin (4 mU ⋅ kg−1 ⋅ min−1, equivalent to 0.056 nmol/h; Eli Lilly, Indianapolis, IN). To counteract the hypoglycemia that normally accompanies the infusion of IGF-I or insulin, the infusate also contained 25% dextrose (experiments 1 and2) or 5% dextrose (experiment 4).

Exactly 50 min after the initiation of infusion, an intravenous bolus injection (25 mg/100 g body wt) of phenylalanine (containing 40%l-[2H5]phenylalanine; Tracer Technologies, Somerville, MA) was injected, as previously described by McNurlan et al. (17). Exactly 10 min later, each animal was killed by decapitation, and trunk blood was collected into ice-cold heparinized tubes. In the first experiment, intra-abdominal organs (liver, spleen, kidney, proximal and distal half of the small intestine, colon), heart and skeletal muscles (gastrocnemius, soleus, and plantaris muscle), lung, and brain were dissected and immediately frozen in liquid nitrogen. In the second experiment, spleen, lung, and brain were excluded. In the third and fourth experiments, only skeletal muscles, heart, and liver were sampled.

All experiments described above were approved by the Animal Care and Use Committee of the State University of New York at Stony Brook and adhered to the National Institutes of Health “Guide for the Care and Use of Laboratory Animals.”

Plasma glucose and hormone concentrations.

Glucose concentrations were assayed with the use of a Yellow Springs Instruments rapid glucose analyzer (Yellow Springs, OH). Insulin concentration was determined by RIA (Linco, St. Louis, MO) on plasma samples from all mice, except those infused with insulin. In this group, plasma insulin concentrations were assessed by RIA (DPC, Los Angeles, CA) with the use of human insulin standards. Plasma for the determination of IGF-I was extracted with the use of a modified acid-ethanol procedure as previously described (5). The eluate was evaporated, and the dried sample was reconstituted with phosphate buffer for IGF-I determination by RIA. Recombinant human [Thr59] IGF-I (gift from Upstate Biotechnology, Lake Placid, NY) was used for iodination and standards (5). The effective dose for this assay is 0.03–0.08 μg/tube; inter- and intra-assay coefficients of variation are 10 and 7%, respectively.

Measurements of protein synthesis.

The rate of protein synthesis in individual tissues was measured by the incorporation of injected [2H5]phenylalanine into tissue proteins (10, 17). The determination ofl-[2H5]phenylalanine enrichment in plasma samples and in samples of hydrolyzed muscle protein has been described recently (17). In brief, frozen organ samples were powdered under liquid nitrogen, and a portion (10–100 mg) was suspended in ice-cold 3% (wt/vol) perchloric acid. After centrifugation, the supernatant was decanted, and the pellet was further washed with perchloric acid. The enrichment of [2H5]phenylalanine in free intracellular amino acids was measured in supernatants by gas chromatography-mass spectrometry (GC-MS) of the t-butyldimethylsilyl derivative under electron impact and selective ion recording (17).

Determination of the [2H5]phenylalanine enrichment in hydrolyzed tissue protein was made by first converting the phenylalanine to phenethylamine with the enzyme tyrosine decarboxylase and then measuring the enrichment by GC-MS of the heptafluorobutyryl derivative under electron impact and selective ion recording. The ions at mass-to-charge ratio of 106 (M+2) and 109 (M+5) were monitored. The conversion to phenethylamine and the monitoring of the M+2 rather than the M+0 ion enable accurate measurements of enrichment down to 0.005 mol%, as described previously (17).

The fractional rates of protein synthesis, FSR, defined as the percentage of tissue protein renewed each day, were calculated according to the formulaFSR=(Eb×100)/(Ea×t) wheret is the time interval between injection and cooling of sampled tissue, expressed in days (10), and Eb and Ea are the enrichments of [2H5]phenylalanine in hydrolized tissue protein and in tissue free amino acids, respectively.


Results are expressed as means ± SE. Comparisons between two groups were made with the use of Student’st-test. Among three groups, comparisons were made by ANOVA and the Newman-Keuls test. Differences were considered significant at P < 0.05.


In the first experiment, the infusion of IGF-I increased the IGF-I concentration more than threefold (372 ± 56 vs. 1,261 ± 133 ng/ml, P < 0.05). Plasma insulin levels fell from 18 ± 2 μU/ml in control mice to 11 ± 2 μU/ml in IGF-I-infused animals (P< 0.05), whereas the glucose concentrations were not different between the two groups (124 ± 6 vs. 123 ± 7 mg/dl). The FSR were determined in 12 selected tissues. The most pronounced effect was a rise in FSR in all types of skeletal muscle examined in response to IGF-I (Table 1). In gastrocnemius and plantaris muscle, the increase was ∼65%, whereas in soleus and heart muscle, the increase was ∼25%. There was no significant effect of IGF-I infusion on FSR in liver, kidney, proximal small intestine, colon, lung, spleen, and whole brain (Table 1). There was a small but statistically significant decrease of ∼10% in the distal intestine.

View this table:
Table 1.

Effect of IGF-I on fractional rate of protein synthesis in various organs

Results from the second experiment indicated that an overnight fast decreased protein synthesis in all muscles examined compared with values in fed control animals (Fig. 1). The observed decrease in FSR averaged 50–20% in these tissues. The short-term infusion of IGF-I in fasted rats increased rates of protein synthesis on skeletal and cardiac muscle to values not different from those obtained in fed mice infused with vehicle (Fig. 1). Protein synthetic rates in intra-abdominal organs were not significantly influenced by fasting or the infusion of IGF-I (Fig.2). The infusion of IGF-I in this study elevated circulating IGF-I levels to 1,273 ± 124 ng/ml as compared with values seen in the two control groups (fasted 308 ± 40 vs. fed 328 ± 17 ng/ml). Insulin levels were 44% higher in the fed mice compared with fasted controls (26 ± 4 vs. 18 ± 2 μU/ml), but this change did not reach significance. Insulin was significantly decreased in animals infused with IGF-I (12 ± 2 μU/ml). Euglycemia was maintained in all three groups (control fed, 91 ± 6 mg/dl; control fasted, 75 ± 10 mg/dl; and IGF infused, 69 ± 8 mg/dl).

Fig. 1.

Rates of protein synthesis (FSR; in percent/day ± SE) in skeletal muscles [gastrocnemius (A), plantaris (B), and soleus (C)] and heart (D) of mice that were deprived of food overnight (fast), food deprived and then infused with insulin-like growth factor I (IGF-I) for 1 h (IGF-I), or given access to food ad libitum (fed). * P < 0.05 and † P < 0.001, significantly different from IGF-I and fed groups.

Fig. 2.

Rates of protein synthesis (in percent/day ± SE) in abdominal tissues of mice that were deprived of food overnight (fast), food deprived and then infused with IGF-I for 1 h (IGF-I), or given access to food ad libitum (fed). A: kidney. B: liver.C: proximal intestine.D: distal intestine.E: colon.

In the third experimental series, insulin was infused into fasted mice to raise the circulating concentration of the hormone to near postprandial levels (control 21 ± 3 μU/ml vs. insulin infused 82 ± 11 μU/ml, P < 0.05). An insulin level of 69 μU/ml has previously been observed in similar mice after feeding (24). In the present series, plasma glucose concentrations were 65 ± 8 mg/dl in the control group and 99 ± 21 mg/dl in animals infused with insulin. After 1 h, the short-term infusion of insulin had increased the protein FSR in the gastrocnemius (52%), soleus (11%), plantaris (55%), and heart (21%) (Table2) compared with control values. There was no significant effect of insulin on the FSR in liver at this time point. Plasma IGF-I levels were not significantly different between the insulin-infused mice (388 ± 45 ng/ml) and time-matched control animals (331 ± 28 ng/ml) in this protocol.

View this table:
Table 2.

Effect of insulin on fractional rate of protein synthesis in muscles and liver

The fourth experiment was to test whether IGF-I given at a fivefold lower dose would have the same effect as the higher dose used inexperiments 1 and2. The plasma level of IGF-I in these animals was 847 ± 28 ng/ml, compared with 322 ± 36 ng/ml in controls. The rates of tissue protein synthesis are shown in Table3. Both gastrocnemius muscle and heart protein syntheses were elevated significantly by IGF-I, whereas soleus muscle and liver were not affected.

View this table:
Table 3.

Effect of IGF-I at a reduced dose on fractional rate of protein synthesis in muscles and liver


The results of this study indicate that the rate of muscle protein synthesis is sensitive to an acute elevation in the circulating concentration of IGF-I or insulin. Furthermore, the ability of IGF-I or insulin to stimulate protein synthesis occurred in various individual muscles with different characteristics and fiber-type composition. Even cardiac muscle revealed an increase in synthesis rate in response to IGF-I, a finding that might be of relevance in clinical states of severe catabolism when myocardial dysfunction is evident. Moreover, infusion of IGF-I into fasted mice restored protein synthesis in all muscles to the rate observed in mice receiving an insulin infusion or in fed animals.

These observations of protein synthesis rate were made by the flooding method, using [2H5]phenylalanine. The validity of the flooding technique has previously been discussed in detail (9). The method was originally developed with the use of tritiated phenylalanine (10) but was modified to use deuterated phenylalanine for measurements in humans (17). The stable isotope also has several advantages for investigations in animals. The sensitivity of the GC-MS techniques for measuring the enrichment of [2H5]phenylalanine in protein hydrolysates and free amino acids is greater than the fluorometric assay and scintillation counting used for tritium. This permits smaller tissue samples to be measured (<0.5 mg protein or 2–5 mg tissue), for example the soleus muscle from a single mouse leg (∼5 mg). We also find the reproducibility of the measurements to be better with deuterium than with tritium. In addition, the cost of the stable isotope is far less than that of the radioactive alternative, and there is no risk of radiation and no attendant costs of radiation protection and isotope disposal.

Acute effects of IGF-I on muscle protein synthesis have not been observed previously with periods as short as 1 h in vivo, although a stimulation in vitro has been reported after 0.5 h (15). Previous measurements in growing rats 40 min after injection of IGF-I revealed no changes in FSR in muscle, jejunum, or liver (18), but in this case, no glucose was given to control for hypoglycemia. Jacob et al. (13) infused [14C]leucine into rats during 90 min of IGF-I infusion at a rate slightly lower than that in the present study but did not observe any increase in the incorporation of label into muscle protein. A similar conclusion was reached from a later experiment (14) that showed that the incorporation of label was not increased by IGF-I infusion during a 4-h infusion of [3H]phenylalanine. By contrast, other studies (22, 24) showed an increase in muscle protein synthesis at 2.5–3 h after a single subcutaneous injection of IGF-I into fasted mice. This effect was, however, smaller than that reported here, and, unlike the present study, rates were not restored to values seen in fed animals. Although in these studies (22,24) an amount of IGF-I similar to that used here was given as a bolus, no infusion of hormone was given subsequently, and plasma levels of IGF-I had returned to control values at the end of the experiment. Also, muscle types were not separated in the previous study. A stimulation of protein synthesis has also been reported in human skeletal muscle after IGF-I infusion for 3–6 h into the artery of one arm, with the other arm serving as a control (6). Protein synthesis in cardiac muscle has also been shown to be enhanced by IGF-I when infused at a high dose rate for 2 h in a perfused rat heart model (7).

In the present study, glucose was infused with IGF-I to prevent the fall in plasma glucose that would otherwise have occurred. However, care was taken to avoid hyperglycemia and resulting hyperinsulinemia so that the observed stimulation in muscle protein synthesis is not the result of insulin rather than IGF-I. In fact, insulin levels were lower in IGF-I-infused mice than in control animals. Jacob et al. (14) have suggested that insulin is permissive for the effect of IGF-I on protein synthesis. In their experiments on rats, the IGF-I infusion for 4 h caused a more pronounced fall in insulin levels than we observed, and a stimulation in protein synthesis was observed when basal insulin levels were restored by insulin infusion. In the present study, IGF-I was able to stimulate muscle protein synthesis in mice despite the moderate reduction in insulin levels. Jacob et al. also concluded that a response to IGF-I was prevented by the hypoaminoacidemia resulting from inhibition of proteolysis by the IGF-I. The fall in amino acid concentration would have been smaller in the present study because of the much shorter period of treatment (1 h), which would explain why we were able to detect a stimulation of protein synthesis in muscles by IGF-I. Fryburg (6) has suggested that the effects of IGF-I and insulin are separate, with IGF-I stimulating protein synthesis but insulin inhibiting protein degradation with no effect on synthesis. However, this conclusion was derived from studies in adult human muscle. Whereas insulin has been shown to have no effect on muscle protein synthesis in adult rats (2), a pronounced stimulation of protein synthesis by insulin has been seen in both growing rats (1) and adult mice (24). The age of the mice in the present study was 9 wk, indicating an almost fully grown animal, but at present it is not clear why adults of different species should respond differently.

In the present work, the IGF-I infusion into fasted mice raised muscle protein synthesis to the rate seen in fed animals. Moreover, the sensitivities of the various muscles and organs to the two treatments were very similar. Of the skeletal muscles, those with predominantly fast-twitch glycolytic fibers (gastrocnemius and plantaris muscles) showed the greatest changes, and the muscle with predominantly slow-twitch oxidative fibers (soleus muscle) showed the smallest. At the high dose of IGF-I, the response of cardiac muscle resembled that of soleus. However, it was notable that with the reduced dose of IGF-I (experiment 4), the soleus muscle was no longer responsive, whereas the heart and gastrocnemius were stimulated. The difference in responses between heart and soleus is apparent from the ratio of heart FSR to soleus FSR, which was significantly elevated (P < 0.01) in the IGF-I-treated group compared with the controls, suggesting a greater sensitivity of protein synthesis to IGF-I in heart compared with soleus. All other organs examined showed no change in the FSR of protein. These sensitivities are not only the same with feeding and with IGF-I or insulin infusion but closely parallel those reported previously in tissues of growing rats that were either fed or infused with insulin (1, 20). In growing rats, insulin has been shown to be required for the stimulation of muscle protein synthesis by feeding (19) but to have no effect in the fed animal (8). In accord with this, the present results show that insulin infusion into fasted mice stimulated protein synthesis, but whether insulin or IGF-I is required for the response to feeding cannot be discerned from the present data.

On a molar basis, the higher dose of IGF-I employed inexperiments 1 and2 was ∼60 times that of insulin (experiment 3). However, the infusion rate of glucose needed to maintain euglycemia was similar for insulin and the higher dose of IGF-I, suggesting that effects of the two hormones on glucose uptake were similar. With a fivefold lower dose of IGF-I (experiment 4), there was also a stimulation of protein synthesis in gastrocnemius muscle and heart (Table 3). However, even with the lower dose, we cannot rule out the possibility that some of this response could have resulted from IGF-I interacting with the insulin receptor.

The nonmuscle tissues examined showed no acute responses to IGF-I, even though the IGF-I receptor is known to be widely distributed among tissues (16). Although there is a report that protein synthesis in the liver (determined by primed constant infusion of a labeled amino acid) was increased after acute IGF-I administration (7), we were unable to confirm this response in the present study. It is possible that the hepatic effect of IGF-I requires an infusion of longer duration than 1 h and follows changes in gene transcription. In the intestine, a 3-day period of IGF-I treatment has been shown to stimulate the proliferation of the small intestinal epithelium (23). Using a continuous IGF-I infusion for 5 days in rats, Huang et al. (12) demonstrated an attenuation of gut atrophy and bacterial translocation induced by severe burn injury. Changes in intestinal cellularity were determined as increases in gut DNA and protein content. These parameters were, however, measured using mucosal scrapings, in which the rate of protein synthesis may be more sensitive to IGF-I. This beneficial effect may also have resulted from the chronic administration and a higher dose of IGF-I.

In summary, we conclude that protein synthesis is very rapidly stimulated in skeletal and cardiac muscles, but not in other organs, by infusion of IGF-I or insulin into fasting mice. The similarity of these responses to the effects of food intake suggests a possible role for these hormones in the acute control of protein deposition during feeding.


We thank Kerstin Bark, George Casella, Jie Fan, Dawn Sasvary, Barbara Tyndall, and Parminder Kang for excellent technical assistance.


  • Address for reprint requests: P. J. Garlick, Dept. of Surgery, H. S. C. T19-020, SUNY at Stony Brook, Stony Brook, NY 11794-8191.

  • This work was financially supported by National Institutes of Health Grants GM-38032, GM-32654, DK-48786, and AA-11290; by Nutricia Research Foundation, Zoetermeer, The Netherlands; and by The Swedish Medical Research Council, The Swedish Medical Society, Stockholm, Sweden.

  • Present address of T. H. Bark: Department of Surgery, Karolinska Hospital, S-171 76 Stockholm, Sweden.


View Abstract