0013-7227/92/1304-1942$03.00/O Endocrinology Copyright 0 1992 by The Endocrine
Vol. 130, No. 4 Printed in U.S.A.
Society
Differential Actions of Growth Hormone Like Growth Factor-I on Tissue Protein Dwarf Mice J. M. PELL*
and InsulinMetabolism in
AND P. C. BATES*
Department of Endocrinology and Animal Physiology,Agricultural and Food ResearchCouncil Institute for Grasslandand Animal Production (J.M.P.), Hurley, Maidenhead, Be&s, SL.65NB United Kingdom; and l\Tutrition ResearchUnit (P.C.B.), London, NW1 2PE United Kingdom
ABSTRACT. The actions and interactions of exogenous insulin-like growth factor-1 (IGF-I) and bovine GH (bGH) on protein metabolism were investigated in uiuo using Snell dwarf mice. Mice were administered a daily dose of 1.5 or 20 rg bGH in the presence or absence of 20 gg IGF-I. IGF-I and GH stimulated significant increases in whole body weight gain. Serum IGF-I Concentrations increased dramatically in-mice administered IGF-I. but more modestlv in GH-treated mice. However, greater im&ases in tissue IGFII content were observed for GH- than for IGF-treated mice, implying that GH exerted ita anabolic actions by local IGF-I synthesis. Skeletal muscle (combined gastrocnemius plus plantaris) weight was significantly increased in GH-treated mice and tended to increase in IGF-treated mice. Muscle protein synthesis
0
NE OF the most important aspects of GH physiology in normal growth and development is its stimulation of protein anabolism at the expense of fat accretion. This has relevance for both medical and agricultural applications; GH inhibits the degradation of muscle that accompanies many clinical conditions and also promotes the efficient conversion of nutrient into lean tissue. Administration of GH stimulates increased rates of tissue protein synthesis in both normal (1) and hypopituitary (2) animals, and this is associated with increased circulating insulin-like growth factor-I (IGF-I) concentrations. However, the extent to which the protein anabolic action of GIj is mediated via IGF-I is equivocal at present, even though IGF-I-stimulated growth has been demonstrated (3), and muscle protein synthesis may be increased, albeit modestly, in vitro by IGF-I (4). The potential actions and interactions of GH and IGFReceived September 13,199l. Address all correspondence and requests for reprints to: Dr. J. M. Pell, Department of Molecular and Cellular Physiology, Institute of Aniial Physiology and Genetics Research, Cambridge, CB2 4AT United Kingdom. * Current address: Agricultural and Food Research Council Institute of Animal Physiology and Genetics Research, Babraham, Cambridge, CB2 4AT United Kingdom.
was stimulated by about 50% in mice treated with IGF-I alone and the lower dose of GH and by over 100% in the group treated with 20 ag/day GH compared with that in saline-treated mice; further additive increases in synthesis rates were observed for mice administered both IGF-I and GH. In all cases, this stimulation was due to both increased RNA content and efficiency of protein synthesis, expressed as grams of protein synthesized per g RNA/day. Liver weight and protein synthetic rata were increased by as much as 25% and 34%, respectively, in GH-treated mice, but IGF-I inhibited hepatic protein metabolism, tending to decrease synthesis rates and inducing a decrease in the efficiency of protein synthesis. Thus, IGF-I and GH have specific and differential effecta on tissue protein metabolism in this model. (Endocrinologv 130: 1942-1950,1992)
I on other anabolic processes, such as bone growth, are complex, and this intricacy is also likely for a process as diverse as protein synthesis. At one extreme, GH may act directly via its own tissue receptors to stimulate protein metabolism; alternatively, its action may be dependent exclusively on IGF-I. If IGF-I does mediate the protein anabolic aspect of GH activity, then the source and function of the IGF-I must be considered. This may be endocrine, being released from the liver into the circulation for action at peripheral tissues, or it may be paracrine/autocrine and be synthesized locally at its site of action. A further possibility is that cooperation between both GH and IGF-I is required for an anabolic response. Evidence can be found to support all of these suggestions, and no unequivocal situation exists. GH receptors are present in both liver and muscle (5), and therefore, GH action could be direct. The data of Guler et al. (6), showing that exogenous IGF-I can mimic the actions of GH, are in agreement with the original somatomedin hypothesis (7) that circulating IGF-I is the mediator of GH action. However, the lack of sensitivity to exogenous IGF-I, in quantitative terms, found in most studies comparing GH and IGF-I activities (6, 8) does cast some doubt on this interpretation of GH action,
1942
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 12:29 For personal use only. No other uses without permission. . All rights reserved.
GH, IGF-I,
AND
TISSUE
implying that even though circulating IGF-I can stimulate tissue growth, blood-borne IGF-I might not be the most appropriate or physiological route in uiuo (9). Recent reports suggesting that GH and IGF-I each stimulate the growth of different specific tissues (10) indicate independent roles for GH and IGF-I in the control of anabolism. It is also possible that the imbalance between free IGF-I and its binding proteins that must occur when exogenous IGF-I is administered acutely means that the IGF-I may be rapidly degraded or is not targeted to relevant tissues; this must be more of a problem in hypopituitary animals, since they already have decreased GH-dependent binding protein concentrations (11). Studies in vitro (12-14) have demonstrated that direct local action of both GH and IGF-I is required for the effective growth and development of many tissues, such as bone. In this case, tissue-synthesized IGF-I may have more relevance than circulating IGF-I. Clearly, it is important to separate the actions of GH and IGF-I; this is difficult to achieve, since the administration of GH will generally induce IGF-I synthesis. However, genetically hypopituitary animals, such as Snell dwarf mice, which are unable to synthesize GH, provide an excellent model for the study of IGF-I action alone as well as in combination with exogenous GH. The aims of this investigation were, therefore, to determine 1) whether IGF-I could exert anabolic actions on protein metabolism in the absence of GH, and 2) whether IGF-I exhibited additive or interactive effects with GH. Doses of IGF-I and GH were selected so that IGF-I was administered with minimal or moderate amounts of GH (15). Three different tissues were investigated: an expendable, nutritionally sensitive muscle type (skeletal), a nonexpendable muscle (cardiac), and an essential central protein pool (liver). Materials
and Methods
Animals Snell dwarf mice were bred from a colony at the Institute for Grassland and Animal Production. They were maintained at a temperature of 25 C on a 12-h light, 12-h dark cycle and were offered a powdered diet ad libitum (25% crude protein, consisting of milk powder, wheatgerm, Spratts diet-l, and egg powder in the proportions
100:25:70:3 by weight);
water was always
available. At 8-10 weeks of age, 36 mice were randomly allocated to 1 of 6 treatment groups (see below) and housed G/cage, 1 from each group. Experimental
design
Mice were weighed daily for 3 days and then were treated daily for 7 days with saline (SAL), 20 fig recombinant human IGF-I (IGF), 1.5 pg pituitary bovine (b) GH (LG), 20 pg bGH (HG), 20 rg IGF-I plus 1.5 rg bGH (IGF + LG), or 20 pg IGFI plus 20 pg bGH (IGF + HG). Each daily dose was divided
PROTEIN
METABOLISM
1943
into two aliquots and administered at 0800 and 1600 h as a loo-p1 SCinjection. GH was dissolved in carbonate-buffered saline [25 mM sodium carbonate, pH 9.4, in 0.9% (wt/vol) sodiumchloride] and IGF-I in PBS (100mM sodiumphosphate, pH 7.2). On day 7, mice were killed by decapitation exactly 2 h after their morning treatment. Blood was collected from the neck, kept on ice until clotted, and then centrifuged at 2000x g for 15 min; serumwasremoved and stored at -20 C. Skeletal muscleof mixed fiber type (combinedgastrocnemiusand plantaris), liver, and heart (ventricle) were dissectedout rapidly, cooledon ice, weighed,and stored in liquid nitrogen. Food was removed from the gastrointestinal tract, and the remainderof the carcasswas frozen in liquid nitrogen. Measurement
of protein
synthesis rate
Exactly 15 min before death, mice were injected ip with a large dose of [2,6-3H]phenylalanine (New England Nuclear, Stevenage, Herts, United Kingdom) and 30 pmol unlabeled phenylalanine in a volume of 0.2 ml 0.9% (wt/vol) sodium chloride for the determination of protein synthesis rates (ks). This method for the determination of protein synthesis rate “floods” all precurser pools, and the resultant rate of [3H] phenylalanine incorporation into protein, therefore, only depends on the synthetic rate (16). Samplesof muscle,liver, and heart were homogenizedin ice-cold 10% (wt/vol) trichloroacetic acid using a Polytron homogenizer(Northern Media Supplies, Hessle,North Humberside, United Kingdom) and centrifuged at 2800 x g for 20 min. Carcasseswere homogenizedin 4 vol water using the Polytron homogenizer. A 2-ml portion was removed, and an equal volume of 20% (wt/vol) trichloroacetic acidwasadded;further procedureswere carried out asdescribed for the tissue samples.Trichloroacetic acid-solublephenylalanine-specific radioactivity (Sr) and protein-bound phenylalanine-specific radioactivity (SB) in the trichloroacetic acid-pre-
cipitated pellets were measured by the enzymic-fluorimetric method of Garlick et al. (16). Tissuetotal protein (17) and total RNA (18) concentrations were also determined, and RNA contents were calculated in relation to protein contents as RNA/protein ratios. RibosomalRNA accounts for 80-85% of the total RNA, and therefore, changes in total RNA concentrations largely represent changesin ribosomal RNA (19). Whole body and tissue fractional protein synthesis rates were calculated as a percentage of the total protein pool per day, based on the equation: ks = 100 X SB/(Sr X t), where t is the time of exposure to [3H]phenylalanine in days (20). The efficiency of the protein synthetic rate in terms of the amount of protein synthesizedper unit ribosomalRNA/day (gramsof protein per g RNA/day, known as RNA activity) was alsocalculated from: RNA activity = fractional protein synthesis rate X (tissue protein concentration/tissue RNA concentration). Measurement
of serum IGF-I
concentration
Serum IGF-I concentrations were determined in acidethanol-extracted samplesusing the method of Daughadayet al. (21), with an additional overnight precipitation of binding proteins at -20 C after initial incubation and centrifugation of protein in acid-ethanol. This adaptation is analogousto that describedrecently by Breier et al. (22) and results in a more complete reduction in interference by binding proteins. Ex-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 12:29 For personal use only. No other uses without permission. . All rights reserved.
1944
GH, IGF-I,
AND TISSUE
tracted IGF-I concentrations were then determined by RIA using polyclonal anti-human IGF-I antiserum suppliedby the Institute of Child Health (University of London, United Kingdom). Assessmentof tissueIGF-I abundance Additional mice (n = lo/group) were treated with saline, 20 pg IGF-I, or 1.5 rg GH in exactly the same manner as that describedabove, except that protein synthesis rates were not determined, and dissectedtissueswere cooled using liquid nitrogen-cold isopentanebefore storage in liquid nitrogen. The tissue IGF-I concentration was assessed using an immunocytochemical technique. Cryostat sections of liver and muscle were fixed in formalin (4% in 0.1 M sodium phosphatebuffer, pH 7.0), washedusing distilled water followed by Tris buffer [O.Ol M in 0.9% sodium chloride (wt/vol), pH 7.61, and then incubated with undiluted nonimmune rabbit serum. The sections were then incubated for 5-6 h at room temperature with a specific sheepantihuman IGF-I antiserum diluted lo-fold (liver) or 4-fold (muscle) in Tris buffer. Bound antibody was visualizedby incubation overnight at 4 C with donkey antisheep gold conjugate (Immunogold, Biocell Research Laboratories, Cardiff, United Kingdom) diluted loo-fold in Tris buffer, followed by silver lactate enhancement. Three slides were prepared per tissue sample:complete negative (no Immunogold), nonspecificbinding (replacement of sheepanti-IGF-I by nonimmunesheepserum),and test (asdescribedabove).The degree of staining was quantified using an image analyzer (Quantimet model 520, Leica Cambridge Ltd., Cambridge, United Kingdom). IGF-I abundancewasassumedto be equivalent to specific staining, which was calculated by subtracting staining for the nonspecificbinding slide from the correspondingtest slide. Measurementof serumglucoseconcentrations Blood glucoseconcentrations were obtained at death by the use of a Hypocount glucose meter (Hypoguard Ltd., Woodbridge, Suffolk, United Kingdom). Statistical analysis Statistical differences between treatment groups were assessedusing a 3 x 2 analysis of variance. Main effects of GH (both doses),IGF-I, and 1.5 pg (LG) us. 20 pg (HG) GH were tested as well as any interaction between GH and IGF-I. Data are presented with the pooled SE of the difference (SED) obtained from the analysis of variance calculations as well as individual group SES. When appropriate, the day 0 weights of the mice were usedas a covariate; thus, there were 29 residual degreesof freedom.
PROTEIN
METABOLISM
fore, initial weight was used as a covariate for their statistical analysis. Final body weight increased in GHtreated mice in a dose-dependent manner (P c O.OOl), but only tended to increase in IGF-I-treated mice. However, when whole body weight gain was calculated (Fig. l), IGF-I induced a significant 4-fold increase (P < 0.005), and GH as much as a lo-fold increase. The increases in daily gain were additive for the IGF + LG group, but no further increase in weight gain was observed for the IGF + HG us. the HG mice. Both GH (P c 0.001) and IGF-I (P < 0.004) induced increases in whole body protein synthetic rate, and as with whole body daily gain, the effects of exogenous IGF-I were not apparent in mice additionally treated with the high dose of GH (SAL, 19.2; IGF, 23.0; LG, 20.8; IGF + LG, 26.0; HG, 27.4; IGF + HG, 27.8%/day; SED = 1.46; n = 6). Serum glucose concentrations
Even though serum glucose concentrations tended to decrease in mice treated with IGF-I, statistical significance was not achieved, although it is possible that glucose concentrations had been further decreased during the time between administration of IGF-I, and death and blood sampling (2 h). In contrast, GH induced a
significant increase (P < 0.025) in serum glucose concentrations (SAL, 6.28; IGF, 5.60; LG, 8.40; IGF + LG, 7.14; HG, 8.45; IGF + HG, 7.42 mM; SED = 1.32; n = 6). Muscle (gastrocnemius plus pluntaris) Muscle weight increased significantly
Whole body weight gain and protein synthesis rates
The mean initial body weight (day 0 of treatment) of the mice did not differ significantly between allocated treatment groups (Table l), but when final body weight and tissue weights were regressed against initial weight, significant correlations were obtained (P < 0.001); there-
in a dose-de-
pendent manner in GH-treated mice (Table 1). Even though IGF-I induced an apparent increase in muscle weight, this was not statistically significant. The muscle protein concentration was unaffected by IGF-I, but decreased in GH-treated mice (P < 0.004; SAL, 160.5; IGF, 159.8; LG, 157.4; IGF + LG, 159.4; HG, 151.7; IGF + HG, 151.0 mg protein/g wet wt; SED = 2.8; n = 6),
probably due to increased water content. Figure 2a shows muscle protein synthesis rates, which increased significantly by about 50% in mice treated with
IGF-I alone and low GH and by over 100% in the high GH group compared with values in saline-treated mice. Further increases in synthesis rates were observed for mice administered GH and IGF-I together; no statistical interaction was found for these combined treatment
groups, implying Results
Enda - 1992 Vol13O*No4
that the combined
actions of GH and
IGF-I were additive, not synergistic in skeletal muscle. Two basic mechanisms may mediate an increase in protein synthesis. First, the amount of protein synthetic material (ribosomal RNA) can increase and, second, the efficiency of that RNA in terms of the amount of protein synthesised per unit RNA in any given time (RNA
activity)
may increase. As shown in Fig. 2b, the total
RNA/protein
ratio increased significantly in response to
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 12:29 For personal use only. No other uses without permission. . All rights reserved.
GH, IGF-I,
AND TISSUE
PROTEIN
METABOLISM
1945
TABLE 1. Whole body and tissue weights of Snell dwarf mice treated for 7 days with IGF-I and GH (SAL, saline-treated controls; IGF, 20 pg/day IGF-I; LG, 1.5 pg/day GH; HG, 20 pg/day GH)
Initial BW (g) Final BW (g) Muscle wt (mg) Liver wt (mg) Heart wt (mg)
SAL
IGF
LG
LG +IGF
HG
HG + IGF
9.10 9.51 81.1 451 32.0
9.07 9.89 84.0 424 32.9
9.48 10.19 85.3 503 35.1
9.53 10.57 90.5 478 36.5
9.88 11.00 97.9 565 43.3
9.30 11.16
Significance
SED(n
= 6)
IGF
GH
LG us. HG
0.28 0.23
NS 0.087 NS NS NS
NS
Vol. 130, No. 4 Printed in U.S.A.
Society
Differential Actions of Growth Hormone Like Growth Factor-I on Tissue Protein Dwarf Mice J. M. PELL*
and InsulinMetabolism in
AND P. C. BATES*
Department of Endocrinology and Animal Physiology,Agricultural and Food ResearchCouncil Institute for Grasslandand Animal Production (J.M.P.), Hurley, Maidenhead, Be&s, SL.65NB United Kingdom; and l\Tutrition ResearchUnit (P.C.B.), London, NW1 2PE United Kingdom
ABSTRACT. The actions and interactions of exogenous insulin-like growth factor-1 (IGF-I) and bovine GH (bGH) on protein metabolism were investigated in uiuo using Snell dwarf mice. Mice were administered a daily dose of 1.5 or 20 rg bGH in the presence or absence of 20 gg IGF-I. IGF-I and GH stimulated significant increases in whole body weight gain. Serum IGF-I Concentrations increased dramatically in-mice administered IGF-I. but more modestlv in GH-treated mice. However, greater im&ases in tissue IGFII content were observed for GH- than for IGF-treated mice, implying that GH exerted ita anabolic actions by local IGF-I synthesis. Skeletal muscle (combined gastrocnemius plus plantaris) weight was significantly increased in GH-treated mice and tended to increase in IGF-treated mice. Muscle protein synthesis
0
NE OF the most important aspects of GH physiology in normal growth and development is its stimulation of protein anabolism at the expense of fat accretion. This has relevance for both medical and agricultural applications; GH inhibits the degradation of muscle that accompanies many clinical conditions and also promotes the efficient conversion of nutrient into lean tissue. Administration of GH stimulates increased rates of tissue protein synthesis in both normal (1) and hypopituitary (2) animals, and this is associated with increased circulating insulin-like growth factor-I (IGF-I) concentrations. However, the extent to which the protein anabolic action of GIj is mediated via IGF-I is equivocal at present, even though IGF-I-stimulated growth has been demonstrated (3), and muscle protein synthesis may be increased, albeit modestly, in vitro by IGF-I (4). The potential actions and interactions of GH and IGFReceived September 13,199l. Address all correspondence and requests for reprints to: Dr. J. M. Pell, Department of Molecular and Cellular Physiology, Institute of Aniial Physiology and Genetics Research, Cambridge, CB2 4AT United Kingdom. * Current address: Agricultural and Food Research Council Institute of Animal Physiology and Genetics Research, Babraham, Cambridge, CB2 4AT United Kingdom.
was stimulated by about 50% in mice treated with IGF-I alone and the lower dose of GH and by over 100% in the group treated with 20 ag/day GH compared with that in saline-treated mice; further additive increases in synthesis rates were observed for mice administered both IGF-I and GH. In all cases, this stimulation was due to both increased RNA content and efficiency of protein synthesis, expressed as grams of protein synthesized per g RNA/day. Liver weight and protein synthetic rata were increased by as much as 25% and 34%, respectively, in GH-treated mice, but IGF-I inhibited hepatic protein metabolism, tending to decrease synthesis rates and inducing a decrease in the efficiency of protein synthesis. Thus, IGF-I and GH have specific and differential effecta on tissue protein metabolism in this model. (Endocrinologv 130: 1942-1950,1992)
I on other anabolic processes, such as bone growth, are complex, and this intricacy is also likely for a process as diverse as protein synthesis. At one extreme, GH may act directly via its own tissue receptors to stimulate protein metabolism; alternatively, its action may be dependent exclusively on IGF-I. If IGF-I does mediate the protein anabolic aspect of GH activity, then the source and function of the IGF-I must be considered. This may be endocrine, being released from the liver into the circulation for action at peripheral tissues, or it may be paracrine/autocrine and be synthesized locally at its site of action. A further possibility is that cooperation between both GH and IGF-I is required for an anabolic response. Evidence can be found to support all of these suggestions, and no unequivocal situation exists. GH receptors are present in both liver and muscle (5), and therefore, GH action could be direct. The data of Guler et al. (6), showing that exogenous IGF-I can mimic the actions of GH, are in agreement with the original somatomedin hypothesis (7) that circulating IGF-I is the mediator of GH action. However, the lack of sensitivity to exogenous IGF-I, in quantitative terms, found in most studies comparing GH and IGF-I activities (6, 8) does cast some doubt on this interpretation of GH action,
1942
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 12:29 For personal use only. No other uses without permission. . All rights reserved.
GH, IGF-I,
AND
TISSUE
implying that even though circulating IGF-I can stimulate tissue growth, blood-borne IGF-I might not be the most appropriate or physiological route in uiuo (9). Recent reports suggesting that GH and IGF-I each stimulate the growth of different specific tissues (10) indicate independent roles for GH and IGF-I in the control of anabolism. It is also possible that the imbalance between free IGF-I and its binding proteins that must occur when exogenous IGF-I is administered acutely means that the IGF-I may be rapidly degraded or is not targeted to relevant tissues; this must be more of a problem in hypopituitary animals, since they already have decreased GH-dependent binding protein concentrations (11). Studies in vitro (12-14) have demonstrated that direct local action of both GH and IGF-I is required for the effective growth and development of many tissues, such as bone. In this case, tissue-synthesized IGF-I may have more relevance than circulating IGF-I. Clearly, it is important to separate the actions of GH and IGF-I; this is difficult to achieve, since the administration of GH will generally induce IGF-I synthesis. However, genetically hypopituitary animals, such as Snell dwarf mice, which are unable to synthesize GH, provide an excellent model for the study of IGF-I action alone as well as in combination with exogenous GH. The aims of this investigation were, therefore, to determine 1) whether IGF-I could exert anabolic actions on protein metabolism in the absence of GH, and 2) whether IGF-I exhibited additive or interactive effects with GH. Doses of IGF-I and GH were selected so that IGF-I was administered with minimal or moderate amounts of GH (15). Three different tissues were investigated: an expendable, nutritionally sensitive muscle type (skeletal), a nonexpendable muscle (cardiac), and an essential central protein pool (liver). Materials
and Methods
Animals Snell dwarf mice were bred from a colony at the Institute for Grassland and Animal Production. They were maintained at a temperature of 25 C on a 12-h light, 12-h dark cycle and were offered a powdered diet ad libitum (25% crude protein, consisting of milk powder, wheatgerm, Spratts diet-l, and egg powder in the proportions
100:25:70:3 by weight);
water was always
available. At 8-10 weeks of age, 36 mice were randomly allocated to 1 of 6 treatment groups (see below) and housed G/cage, 1 from each group. Experimental
design
Mice were weighed daily for 3 days and then were treated daily for 7 days with saline (SAL), 20 fig recombinant human IGF-I (IGF), 1.5 pg pituitary bovine (b) GH (LG), 20 pg bGH (HG), 20 rg IGF-I plus 1.5 rg bGH (IGF + LG), or 20 pg IGFI plus 20 pg bGH (IGF + HG). Each daily dose was divided
PROTEIN
METABOLISM
1943
into two aliquots and administered at 0800 and 1600 h as a loo-p1 SCinjection. GH was dissolved in carbonate-buffered saline [25 mM sodium carbonate, pH 9.4, in 0.9% (wt/vol) sodiumchloride] and IGF-I in PBS (100mM sodiumphosphate, pH 7.2). On day 7, mice were killed by decapitation exactly 2 h after their morning treatment. Blood was collected from the neck, kept on ice until clotted, and then centrifuged at 2000x g for 15 min; serumwasremoved and stored at -20 C. Skeletal muscleof mixed fiber type (combinedgastrocnemiusand plantaris), liver, and heart (ventricle) were dissectedout rapidly, cooledon ice, weighed,and stored in liquid nitrogen. Food was removed from the gastrointestinal tract, and the remainderof the carcasswas frozen in liquid nitrogen. Measurement
of protein
synthesis rate
Exactly 15 min before death, mice were injected ip with a large dose of [2,6-3H]phenylalanine (New England Nuclear, Stevenage, Herts, United Kingdom) and 30 pmol unlabeled phenylalanine in a volume of 0.2 ml 0.9% (wt/vol) sodium chloride for the determination of protein synthesis rates (ks). This method for the determination of protein synthesis rate “floods” all precurser pools, and the resultant rate of [3H] phenylalanine incorporation into protein, therefore, only depends on the synthetic rate (16). Samplesof muscle,liver, and heart were homogenizedin ice-cold 10% (wt/vol) trichloroacetic acid using a Polytron homogenizer(Northern Media Supplies, Hessle,North Humberside, United Kingdom) and centrifuged at 2800 x g for 20 min. Carcasseswere homogenizedin 4 vol water using the Polytron homogenizer. A 2-ml portion was removed, and an equal volume of 20% (wt/vol) trichloroacetic acidwasadded;further procedureswere carried out asdescribed for the tissue samples.Trichloroacetic acid-solublephenylalanine-specific radioactivity (Sr) and protein-bound phenylalanine-specific radioactivity (SB) in the trichloroacetic acid-pre-
cipitated pellets were measured by the enzymic-fluorimetric method of Garlick et al. (16). Tissuetotal protein (17) and total RNA (18) concentrations were also determined, and RNA contents were calculated in relation to protein contents as RNA/protein ratios. RibosomalRNA accounts for 80-85% of the total RNA, and therefore, changes in total RNA concentrations largely represent changesin ribosomal RNA (19). Whole body and tissue fractional protein synthesis rates were calculated as a percentage of the total protein pool per day, based on the equation: ks = 100 X SB/(Sr X t), where t is the time of exposure to [3H]phenylalanine in days (20). The efficiency of the protein synthetic rate in terms of the amount of protein synthesizedper unit ribosomalRNA/day (gramsof protein per g RNA/day, known as RNA activity) was alsocalculated from: RNA activity = fractional protein synthesis rate X (tissue protein concentration/tissue RNA concentration). Measurement
of serum IGF-I
concentration
Serum IGF-I concentrations were determined in acidethanol-extracted samplesusing the method of Daughadayet al. (21), with an additional overnight precipitation of binding proteins at -20 C after initial incubation and centrifugation of protein in acid-ethanol. This adaptation is analogousto that describedrecently by Breier et al. (22) and results in a more complete reduction in interference by binding proteins. Ex-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 12:29 For personal use only. No other uses without permission. . All rights reserved.
1944
GH, IGF-I,
AND TISSUE
tracted IGF-I concentrations were then determined by RIA using polyclonal anti-human IGF-I antiserum suppliedby the Institute of Child Health (University of London, United Kingdom). Assessmentof tissueIGF-I abundance Additional mice (n = lo/group) were treated with saline, 20 pg IGF-I, or 1.5 rg GH in exactly the same manner as that describedabove, except that protein synthesis rates were not determined, and dissectedtissueswere cooled using liquid nitrogen-cold isopentanebefore storage in liquid nitrogen. The tissue IGF-I concentration was assessed using an immunocytochemical technique. Cryostat sections of liver and muscle were fixed in formalin (4% in 0.1 M sodium phosphatebuffer, pH 7.0), washedusing distilled water followed by Tris buffer [O.Ol M in 0.9% sodium chloride (wt/vol), pH 7.61, and then incubated with undiluted nonimmune rabbit serum. The sections were then incubated for 5-6 h at room temperature with a specific sheepantihuman IGF-I antiserum diluted lo-fold (liver) or 4-fold (muscle) in Tris buffer. Bound antibody was visualizedby incubation overnight at 4 C with donkey antisheep gold conjugate (Immunogold, Biocell Research Laboratories, Cardiff, United Kingdom) diluted loo-fold in Tris buffer, followed by silver lactate enhancement. Three slides were prepared per tissue sample:complete negative (no Immunogold), nonspecificbinding (replacement of sheepanti-IGF-I by nonimmunesheepserum),and test (asdescribedabove).The degree of staining was quantified using an image analyzer (Quantimet model 520, Leica Cambridge Ltd., Cambridge, United Kingdom). IGF-I abundancewasassumedto be equivalent to specific staining, which was calculated by subtracting staining for the nonspecificbinding slide from the correspondingtest slide. Measurementof serumglucoseconcentrations Blood glucoseconcentrations were obtained at death by the use of a Hypocount glucose meter (Hypoguard Ltd., Woodbridge, Suffolk, United Kingdom). Statistical analysis Statistical differences between treatment groups were assessedusing a 3 x 2 analysis of variance. Main effects of GH (both doses),IGF-I, and 1.5 pg (LG) us. 20 pg (HG) GH were tested as well as any interaction between GH and IGF-I. Data are presented with the pooled SE of the difference (SED) obtained from the analysis of variance calculations as well as individual group SES. When appropriate, the day 0 weights of the mice were usedas a covariate; thus, there were 29 residual degreesof freedom.
PROTEIN
METABOLISM
fore, initial weight was used as a covariate for their statistical analysis. Final body weight increased in GHtreated mice in a dose-dependent manner (P c O.OOl), but only tended to increase in IGF-I-treated mice. However, when whole body weight gain was calculated (Fig. l), IGF-I induced a significant 4-fold increase (P < 0.005), and GH as much as a lo-fold increase. The increases in daily gain were additive for the IGF + LG group, but no further increase in weight gain was observed for the IGF + HG us. the HG mice. Both GH (P c 0.001) and IGF-I (P < 0.004) induced increases in whole body protein synthetic rate, and as with whole body daily gain, the effects of exogenous IGF-I were not apparent in mice additionally treated with the high dose of GH (SAL, 19.2; IGF, 23.0; LG, 20.8; IGF + LG, 26.0; HG, 27.4; IGF + HG, 27.8%/day; SED = 1.46; n = 6). Serum glucose concentrations
Even though serum glucose concentrations tended to decrease in mice treated with IGF-I, statistical significance was not achieved, although it is possible that glucose concentrations had been further decreased during the time between administration of IGF-I, and death and blood sampling (2 h). In contrast, GH induced a
significant increase (P < 0.025) in serum glucose concentrations (SAL, 6.28; IGF, 5.60; LG, 8.40; IGF + LG, 7.14; HG, 8.45; IGF + HG, 7.42 mM; SED = 1.32; n = 6). Muscle (gastrocnemius plus pluntaris) Muscle weight increased significantly
Whole body weight gain and protein synthesis rates
The mean initial body weight (day 0 of treatment) of the mice did not differ significantly between allocated treatment groups (Table l), but when final body weight and tissue weights were regressed against initial weight, significant correlations were obtained (P < 0.001); there-
in a dose-de-
pendent manner in GH-treated mice (Table 1). Even though IGF-I induced an apparent increase in muscle weight, this was not statistically significant. The muscle protein concentration was unaffected by IGF-I, but decreased in GH-treated mice (P < 0.004; SAL, 160.5; IGF, 159.8; LG, 157.4; IGF + LG, 159.4; HG, 151.7; IGF + HG, 151.0 mg protein/g wet wt; SED = 2.8; n = 6),
probably due to increased water content. Figure 2a shows muscle protein synthesis rates, which increased significantly by about 50% in mice treated with
IGF-I alone and low GH and by over 100% in the high GH group compared with values in saline-treated mice. Further increases in synthesis rates were observed for mice administered GH and IGF-I together; no statistical interaction was found for these combined treatment
groups, implying Results
Enda - 1992 Vol13O*No4
that the combined
actions of GH and
IGF-I were additive, not synergistic in skeletal muscle. Two basic mechanisms may mediate an increase in protein synthesis. First, the amount of protein synthetic material (ribosomal RNA) can increase and, second, the efficiency of that RNA in terms of the amount of protein synthesised per unit RNA in any given time (RNA
activity)
may increase. As shown in Fig. 2b, the total
RNA/protein
ratio increased significantly in response to
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 12:29 For personal use only. No other uses without permission. . All rights reserved.
GH, IGF-I,
AND TISSUE
PROTEIN
METABOLISM
1945
TABLE 1. Whole body and tissue weights of Snell dwarf mice treated for 7 days with IGF-I and GH (SAL, saline-treated controls; IGF, 20 pg/day IGF-I; LG, 1.5 pg/day GH; HG, 20 pg/day GH)
Initial BW (g) Final BW (g) Muscle wt (mg) Liver wt (mg) Heart wt (mg)
SAL
IGF
LG
LG +IGF
HG
HG + IGF
9.10 9.51 81.1 451 32.0
9.07 9.89 84.0 424 32.9
9.48 10.19 85.3 503 35.1
9.53 10.57 90.5 478 36.5
9.88 11.00 97.9 565 43.3
9.30 11.16
Significance
SED(n
= 6)
IGF
GH
LG us. HG
0.28 0.23
NS 0.087 NS NS NS
NS
