Comparison of the Delayed Actions of Growth Hormone and Somatomedin on Adipose Tissue Metabolism JESSICA SCHWARTZ1 AND H. MAURICE GOODMAN2 Department of Physiology, University of Massachusetts Medical School, Worcester, Massachusetts 01605 ABSTRACT. Although acute insulin-like effects of growth hormone (GH) on adipose tissue are readily demonstrable in vitro, it has not been possible to reproduce in vitro the delayed (3 h) inhibition of glucose utilization seen after the administration of GH in vivo. To examine the possibility that somatomedin (Sm), the postulated mediator of GH action, might, like GH, have biphasic effects on adipose tissue metabolism and mediate the delayed inhibition of glucose utilization, Sm was prepared by gel filtration from an acid-ethanol extract of normal rat plasma. The partially purified material increased 35 SO4 incorporation into costal cartilage of hypophysectomized rats and produced in adipose tissue an acute stimulation of glucose oxidation that was not
suppressed by insulin anti-serum. Unlike GH, whose insulin-like effects disappear in 3-4 hours, Smstimulated glucose oxidation remained linear for the entire 4 hour incubation period. Although acute stimulation of tissues with GH rendered them refractory to renewed insulin-like stimulation by GH, no such refractoriness to the action of Sm was seen. Furthermore, acute stimulation with Sm failed to render tissues refractory either to itself or to GH. Finally, Sm failed to reproduce the delayed lipolytic effects produced by GH in conjunction with glucocorticoids. These results make it highly unlikely that Sm alone accounts for the delayed metabolic effects of GH. {Endocrinology 98: 730, 1976)
A LTHOUGH growth hormone (GH) has ±\. biphasic effects on glucose metabolism in hypophysectomized animals (1,2), only the acute insulin-like phase is readily demonstrable in vitro. The transitory insulin-like activity of GH subsides, and even reverses, after the first hour despite the continued presence of the hormone, and by the fourth hour after GH is administered to hypophysectomized or normal animals, a sustained depression of glucose metabolism is seen (3). However, when adipose tissue is exposed to GH in vitro, even for pro-
longed periods, basal glucose utilization is not depressed (3). One explanation for the discrepancy between the in vivo and in vitro responses to GH, and for the delay in onset of the depression in glucose utilization, is that other substance(s) formed in response to GH may mediate the later, sustained actions of the hormone. A precedent for such an idea was established with the discovery of the GH-dependent plasma sulfation factor, somatomedin (Sm), which appears to mediate the stimulation of cartilage metabolism by GH (4,5). Insulin-like effects of Sm have now been described in a number of tissues, including, in adipose tissue, the stimulation of glucose oxidation and triglyceride synthesis, and the depression of epinephrine-stimulated lipolysis in vitro (6-8). These responses to Sm, which are not suppressed by insulin antibodies (5), occur upon acute exposure to Sm, but metabolic effects during prolonged incubation with Sm have not been reported. The following experiments were undertaken to determine whether responses to Sm are biphasic, like those of GH, and whether Sm mediates the delayed metabolic effects of GH.
Received July 28, 1975. Supported in part by Grant AM 15046 and by predoctoral fellowship 5-FOl-GM 46101, both from the National Institutes of Health. This work is taken from a thesis submitted by Jessica Schwartz to Harvard University in partial fulfillment of the requirements for the degree of Doctor of Philosophy. A portion of this work has been presented in preliminary form (Fed Proc 35: 272, 1975). 1 Present address: Department of Pharmacology, New York University Medical Center, New York, New York 10016. 2 Reprint requests should be sent to Department of Physiology, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, Mass. 01605.
730
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 11:23 For personal use only. No other uses without permission. . All rights reserved.
DELAYED ACTIONS OF GH AND SOMATOMEDIN
Materials and Methods Preparation of Sm. 100 ml of plasma were obtained from normal male rats under ether anaesthesia. A 10 ml portion of the plasma was frozen and reserved for assay as reference plasma; the remainder was lyophilized and used for preparation of an acid-ethanol extract (9,10). The extract was dissolved in 1% formic acid and was applied to a 2.5 x 38 cm column of Sephadex G-75 (Pharmacia) equilibrated with 1% formic acid. A typical elution pattern is shown in Fig. 1. The fraction indicated by the arrow contained approximately 28% of the sulfation activity of the unextracted plasma, but less than 0.06% of the protein (11), indicating that nearly 500fold purification had been achieved. This represents approximately 1% of the degree of purification obtained in larger scale preparations (9). The material exhibited multiple bands when subjected to polyacrylamide gel electrophoresis (12). Corresponding fractions from four separate extractions were pooled and used in all of the experiments to be described. The partially purified Sm was tested for biological activity at 0.03-0.3 sulfation units (SU) per ml, 1 SU being defined as the sulfation activity present in 1 ml of reference plasma. One SU of partially purified Sm contained 0.14 mg protein. Sulfation activity. The Sm fractions were assayed for their ability to stimulate sulfate incorporation into cartilage by a modification of the method of Salmon and Daughaday (13), using costal carti-
731
lage segments from 70 g hypophysectomized male rats (Charles River Breeding Laboratories). Tissues were incubated individually at 37 C for 24 h in plasma or in Waymouth's medium (MB 752/1 with glutamine, Grand Island Biological), containing penicillin (50 U/ml), streptomycin (50 jug/ml), and H235SO4 (0.1 /xCi/ml, New England Nuclear). Tissues were then cleaned, weighed, digested in formic acid, and counted in Bray's Solution (14). Sulfate incorporation is expressed as cpm/mg tissue x 24 h. Insulin-like activity was assessed in replicate segments of epididymal fat from hypophysectomized rats so that one tissue from each rat was included in each experimental group. The rats (Charles River Breeding Laboratories, 110125 g) were hypophysectomized 2-3 weeks earlier and then fed a high carbohydrate, fatfree diet (Nutritional Biochemicals). Adipose tissue segments (25-50 mg) were incubated at 37 C under an atmosphere of 95% O 2 -5% CO2 in 0.5-1 ml Krebs-Ringer bicarbonate buffer pH 7.4, containing 1 mg/ml glucose (KRBG), 1% bovine albumin (Armour, fraction V), and the appropriate hormones. Production of I4CO2 from glucose uniformly labelled with 14C (0.1 /u,Ci/ml, New England Nuclear) was measured as described previously (15) and is expressed as cpm/mg tissue x h. Lipolysis was estimated using replicate segments of adipose tissue from normal male rats fasted
2.402.10-
FIG. 1. Chromatography of acidethanol extract of rat plasma on Sephadex G-75 equilibrated with 1% formic acid. Marker proteins, indicated by shaded areas, were gamma-globulin (MW 170,000 peak at 60 ml), and glucagon (MW 3485, peak 150200 ml), 1 mg/ml each. Five ml fractions were collected and the OD28o of each fraction was determined.
Q O
.3050
100
150
200
250
300
Volume ( ml )
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 11:23 For personal use only. No other uses without permission. . All rights reserved.
Entlo • 1976 Vol 98 • No 3
SCHWARTZ AND GOODMAN
732
7CN Cartilage
Fat
H40
60FIG. 2. Effect of Sm on sulfate incorporation in cartilage and glucose oxidation in adipose tissue from hypophysectomized rats. CO2 production was measured during the first h of exposure in vitro. In both experiments, Sm was present at 0.15 SU/ml (solid bars). Each bar represents the mean ± SE of at least 7 observations. The response to Sm is significant (P < .05) in both tissues.
50-
-E E o
30-
CO
2010-
C
SM
C
SM
overnight. After tissues were incubated 4 h at 37 C under an atmosphere of 95% O2-5% CO2 in 1 ml KRBG containing 3% albumin and the appropriate hormones, aliquots of the medium were taken for the analysis of glycerol (16) and free fatty acids (FFA) (17). Data were evaluated statistically by Student's t test (18). Insulin and glucagon were obtained from Eli Lilly Co., and gamma globulin (Cohn fraction II) from Sigma Chemical Co. Guinea pig antiporcine insulin serum was obtained from Schwartz-Mann and diluted 1:1000. Ovine GH (NIH-S10) was provided by the Endocrine Study Section of the National Institutes of Health.
Results
The ability of the partially purified Sm fraction to stimulate sulfate incorporation in cartilage and glucose oxidation in adipose tissue from hypophysectomized rats is depicted in Fig. 2. This fraction was the most highly enriched with both sulfation and insulin-like activities, though neither activity was limited to the one fraction. This material stimulated glucose oxidation at all concentrations tested (Table 1). The stimulation of CO2 production by 0.15 SU/ml was equiva-
lent to that produced either by 60 /u,U/ml insulin (Fig. 3) or by 1 jug/ml GH (Fig. 4) in tissues from the same animals. Anti-insulin serum completely abolished the response to insulin, but had no effect on the response to Sm (Fig. 3), indicating that the insulinlike activity of Sm could not be attributed to insulin present in the extract. In order to determine whether the insulinlike action of Sm, like that of GH, subsided after the first hour, glucose oxidation was measured over the course of 4 hours (Fig. 5). Unlike GH, Sm produced a linear stimulation of CO2 production during the entire 4 h incubation period. TABLE 1. Effect of somatomedin on glucose oxidation in adipose tissue from hypophysectomized rats CO2 production (cpm/mg x h)
Sm concentration (SU/ml)
Control
Increment due to somatomedin
P
0.03 0.15 0.30
55.9 ± 5.4* 62.5 ± 10.0 72.2 ± 8.5
21.3 ± 6.4f 42.5 ± 8.2 61.6 ± 7.8
suppressed by insulin anti-serum. Unlike GH, whose insulin-like effects disappear in 3-4 hours, Smstimulated glucose oxidation remained linear for the entire 4 hour incubation period. Although acute stimulation of tissues with GH rendered them refractory to renewed insulin-like stimulation by GH, no such refractoriness to the action of Sm was seen. Furthermore, acute stimulation with Sm failed to render tissues refractory either to itself or to GH. Finally, Sm failed to reproduce the delayed lipolytic effects produced by GH in conjunction with glucocorticoids. These results make it highly unlikely that Sm alone accounts for the delayed metabolic effects of GH. {Endocrinology 98: 730, 1976)
A LTHOUGH growth hormone (GH) has ±\. biphasic effects on glucose metabolism in hypophysectomized animals (1,2), only the acute insulin-like phase is readily demonstrable in vitro. The transitory insulin-like activity of GH subsides, and even reverses, after the first hour despite the continued presence of the hormone, and by the fourth hour after GH is administered to hypophysectomized or normal animals, a sustained depression of glucose metabolism is seen (3). However, when adipose tissue is exposed to GH in vitro, even for pro-
longed periods, basal glucose utilization is not depressed (3). One explanation for the discrepancy between the in vivo and in vitro responses to GH, and for the delay in onset of the depression in glucose utilization, is that other substance(s) formed in response to GH may mediate the later, sustained actions of the hormone. A precedent for such an idea was established with the discovery of the GH-dependent plasma sulfation factor, somatomedin (Sm), which appears to mediate the stimulation of cartilage metabolism by GH (4,5). Insulin-like effects of Sm have now been described in a number of tissues, including, in adipose tissue, the stimulation of glucose oxidation and triglyceride synthesis, and the depression of epinephrine-stimulated lipolysis in vitro (6-8). These responses to Sm, which are not suppressed by insulin antibodies (5), occur upon acute exposure to Sm, but metabolic effects during prolonged incubation with Sm have not been reported. The following experiments were undertaken to determine whether responses to Sm are biphasic, like those of GH, and whether Sm mediates the delayed metabolic effects of GH.
Received July 28, 1975. Supported in part by Grant AM 15046 and by predoctoral fellowship 5-FOl-GM 46101, both from the National Institutes of Health. This work is taken from a thesis submitted by Jessica Schwartz to Harvard University in partial fulfillment of the requirements for the degree of Doctor of Philosophy. A portion of this work has been presented in preliminary form (Fed Proc 35: 272, 1975). 1 Present address: Department of Pharmacology, New York University Medical Center, New York, New York 10016. 2 Reprint requests should be sent to Department of Physiology, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, Mass. 01605.
730
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 11:23 For personal use only. No other uses without permission. . All rights reserved.
DELAYED ACTIONS OF GH AND SOMATOMEDIN
Materials and Methods Preparation of Sm. 100 ml of plasma were obtained from normal male rats under ether anaesthesia. A 10 ml portion of the plasma was frozen and reserved for assay as reference plasma; the remainder was lyophilized and used for preparation of an acid-ethanol extract (9,10). The extract was dissolved in 1% formic acid and was applied to a 2.5 x 38 cm column of Sephadex G-75 (Pharmacia) equilibrated with 1% formic acid. A typical elution pattern is shown in Fig. 1. The fraction indicated by the arrow contained approximately 28% of the sulfation activity of the unextracted plasma, but less than 0.06% of the protein (11), indicating that nearly 500fold purification had been achieved. This represents approximately 1% of the degree of purification obtained in larger scale preparations (9). The material exhibited multiple bands when subjected to polyacrylamide gel electrophoresis (12). Corresponding fractions from four separate extractions were pooled and used in all of the experiments to be described. The partially purified Sm was tested for biological activity at 0.03-0.3 sulfation units (SU) per ml, 1 SU being defined as the sulfation activity present in 1 ml of reference plasma. One SU of partially purified Sm contained 0.14 mg protein. Sulfation activity. The Sm fractions were assayed for their ability to stimulate sulfate incorporation into cartilage by a modification of the method of Salmon and Daughaday (13), using costal carti-
731
lage segments from 70 g hypophysectomized male rats (Charles River Breeding Laboratories). Tissues were incubated individually at 37 C for 24 h in plasma or in Waymouth's medium (MB 752/1 with glutamine, Grand Island Biological), containing penicillin (50 U/ml), streptomycin (50 jug/ml), and H235SO4 (0.1 /xCi/ml, New England Nuclear). Tissues were then cleaned, weighed, digested in formic acid, and counted in Bray's Solution (14). Sulfate incorporation is expressed as cpm/mg tissue x 24 h. Insulin-like activity was assessed in replicate segments of epididymal fat from hypophysectomized rats so that one tissue from each rat was included in each experimental group. The rats (Charles River Breeding Laboratories, 110125 g) were hypophysectomized 2-3 weeks earlier and then fed a high carbohydrate, fatfree diet (Nutritional Biochemicals). Adipose tissue segments (25-50 mg) were incubated at 37 C under an atmosphere of 95% O 2 -5% CO2 in 0.5-1 ml Krebs-Ringer bicarbonate buffer pH 7.4, containing 1 mg/ml glucose (KRBG), 1% bovine albumin (Armour, fraction V), and the appropriate hormones. Production of I4CO2 from glucose uniformly labelled with 14C (0.1 /u,Ci/ml, New England Nuclear) was measured as described previously (15) and is expressed as cpm/mg tissue x h. Lipolysis was estimated using replicate segments of adipose tissue from normal male rats fasted
2.402.10-
FIG. 1. Chromatography of acidethanol extract of rat plasma on Sephadex G-75 equilibrated with 1% formic acid. Marker proteins, indicated by shaded areas, were gamma-globulin (MW 170,000 peak at 60 ml), and glucagon (MW 3485, peak 150200 ml), 1 mg/ml each. Five ml fractions were collected and the OD28o of each fraction was determined.
Q O
.3050
100
150
200
250
300
Volume ( ml )
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 11:23 For personal use only. No other uses without permission. . All rights reserved.
Entlo • 1976 Vol 98 • No 3
SCHWARTZ AND GOODMAN
732
7CN Cartilage
Fat
H40
60FIG. 2. Effect of Sm on sulfate incorporation in cartilage and glucose oxidation in adipose tissue from hypophysectomized rats. CO2 production was measured during the first h of exposure in vitro. In both experiments, Sm was present at 0.15 SU/ml (solid bars). Each bar represents the mean ± SE of at least 7 observations. The response to Sm is significant (P < .05) in both tissues.
50-
-E E o
30-
CO
2010-
C
SM
C
SM
overnight. After tissues were incubated 4 h at 37 C under an atmosphere of 95% O2-5% CO2 in 1 ml KRBG containing 3% albumin and the appropriate hormones, aliquots of the medium were taken for the analysis of glycerol (16) and free fatty acids (FFA) (17). Data were evaluated statistically by Student's t test (18). Insulin and glucagon were obtained from Eli Lilly Co., and gamma globulin (Cohn fraction II) from Sigma Chemical Co. Guinea pig antiporcine insulin serum was obtained from Schwartz-Mann and diluted 1:1000. Ovine GH (NIH-S10) was provided by the Endocrine Study Section of the National Institutes of Health.
Results
The ability of the partially purified Sm fraction to stimulate sulfate incorporation in cartilage and glucose oxidation in adipose tissue from hypophysectomized rats is depicted in Fig. 2. This fraction was the most highly enriched with both sulfation and insulin-like activities, though neither activity was limited to the one fraction. This material stimulated glucose oxidation at all concentrations tested (Table 1). The stimulation of CO2 production by 0.15 SU/ml was equiva-
lent to that produced either by 60 /u,U/ml insulin (Fig. 3) or by 1 jug/ml GH (Fig. 4) in tissues from the same animals. Anti-insulin serum completely abolished the response to insulin, but had no effect on the response to Sm (Fig. 3), indicating that the insulinlike activity of Sm could not be attributed to insulin present in the extract. In order to determine whether the insulinlike action of Sm, like that of GH, subsided after the first hour, glucose oxidation was measured over the course of 4 hours (Fig. 5). Unlike GH, Sm produced a linear stimulation of CO2 production during the entire 4 h incubation period. TABLE 1. Effect of somatomedin on glucose oxidation in adipose tissue from hypophysectomized rats CO2 production (cpm/mg x h)
Sm concentration (SU/ml)
Control
Increment due to somatomedin
P
0.03 0.15 0.30
55.9 ± 5.4* 62.5 ± 10.0 72.2 ± 8.5
21.3 ± 6.4f 42.5 ± 8.2 61.6 ± 7.8
