J Mol Cell Cardiol
24, 631-639 (1992)
Di%rentialRegulationofInsulin-likeGrowthFactorIbyGrowth Hormone andThyroidH omnone in the Heart Hypophys-Rats
ofJuvenile
Joel M. Kupfw and Stanley A. Rubin From the Division of Cardiology, Department of Medicine, Ceakrs-Sinai Medical Center/UCLA Medicine, Los Angeles, CA 90048, USA
School of
(Received 25 October 1991, accepted in revisedform 23 January 1992) JOEL M. KUPFER AND STANLEY A. RUBIN, Differential Regulation of Insulin-like Growth Factor I by Growth Hormone and Thyroid Hormone in the Heart of Juvenile Hypophysectomiaed Rats. Journal of Molecular and CcNular Cardiology (1992) 24, 631-639. Recent data suggest that the heart can act as both a source and target for the actions of polypeptide growth factors. Insulin-like growth factor I (IGF-I) is a polypeptide that has both mitogenic and differentiation properties that function at the autocrine/paracrinc level, and has recently been demonstrated to be expressed in the heart. This knowledge, coupled with the observation that thyroid hormone (T3) promotes relative cardiac growth compared to the proportional increases in body and heart growth evoked by growth hormone (GH), lead us to speculate whether differential induction of cardiac IGF-I may account for the specialized trophic effects of T3 on the heart. Cardiac IGF-I gene expression was studied in an in uivo model in which cardiac growth in the hypophysectomiaed juvenile rat was stimulated with either GH, T3 or GH + T3. Two week infusions of T3 that resulted in cardiac growth, but no gain in body weight, resulted in a 4.6-fold increase in cardiac IGF-I mRNA levels compared to hypophysectomiaed controls. GH infusions that resulted in similar cardiac growth, but were accompanied by proportional body growth, had no effect on cardiac IGF-I mRNA levels. These data are the first to demonstrate stimulation of cardiac IGF-I mRNA levels by T3 and further support cardiac autocrine/paracrine actions for this polypeptide growth factor. KEY WORDS:
Autocrine;
Paracrine;
Cardiac
growth;
Myosin
Introduction In recent years it has been recognized that cellular growth and maturation can be regulated through cellular oncogenes and the local autocrine or paracrine actions of polypeptide growth factors [I, 21. Along these lines there is an emerging view that signals for cardiac growth, whether initiated at the hormonal or mechanical level, result in expression of cellular oncogenes and polypeptide growth factors that act at the autocrine or paracrine level to coordinately direct the heart towards a hypertrophic response by regulating protein metabolism at the molecular level [3, 4 for reviews see 5, 61. One important polypeptide growth factor that is known to play a pivotal role in somatic ‘Abbreviations TSH, thyroid
used in this paper: IGF-I, stimulating hormone.
insulin-like
+ 09 $03.00/O
chain,
growth and development is insulin-like growth factor I (IGF-I, footnote 1) [7, 81. According to the “somatomedin hypothesis”, IGF-I exerts its growth permissive effects through classic endocrine mechanisms in which GH, from the pituitary, stimulates hepatic production of IGF-I which then enters the circulation to act on distant tissues to promote growth. However, it is now recognized that multiple extrahepatic tissues express IGF-I, even in the hypophysectomized state, and there is increasing experimental evidence to suggest that locally produced IGF-I acts to stimulate nucleic acid synthesis, and to regulate cellular proliferation and differentiation [ 9, II]. This concept of local regulation of growth through autocrine/paracrine IGF-I
growth
Please address all correspondence and reprints to: Joel M. Center, 8700 Beverly Blvd. Los Angeles, CA 90048. 0022-2828/92/060631
heavy
factor Kupfer,
I; GH,
growth
Division
hormone;
of Cardiology,
T3, l-triiodothyronine; Cedars-Sinai
@ 1992 Academic
Medical
Press Limited
J. M. Kupfer and S.A. Rubin
632
action has received further sunnort from reports which demonstrate that other hormones and growth factors that exert trophic effects increase target organ IGF-I mRNA levels to a greater extent than GH [1.2-161. This is also apparent in muscle cells, where autocrine IGF-I expression is known to play an important role in smooth muscle cell proliferation and in skeletal muscle cell differentiation [17, 181. In the heart, myocyte IGF-I expression increases substantially during the first 50 post-natal days consistent with an important developmental role [19-211. Thyroid hormone is known to exert a strong trophic influence on the heart and promotes cardiac growth in the absence of a proportional increase in body weight [22-251. In contrast GH promotes proportional body and cardiac growth. Such data suggest that thyroid hormone, directly or indirectly, induces specific growth promoting factors within the heart. Recognizing the important growth promoting actions of IGF-I and its potential to act at the autocrine/paracrine level, we examined whether T3 preferentially regulates cardiac IGF-I expression compared to GH, which may explain at least part of the specific growth promoting action of thyroid hormone on the heart. To address this question we employed an in vivo model of cardiac growth using hypophysectomized juvenile rats treated for two weeks with physiological doses of GH, T3, or GH + T3. At the end of the treatment period left ventricular IGF-I mRNA expression was studied and correlated with the different patterns of growth. Our results demonstrate that cardiac growth resulting from T3 treatment is accompanied by significant induction of cardiac IGF-I expression compared to hypophysectomized placebo controls or chronic GH treatment. Our findings offer a potential mechanism to account for the specific trophic actions of T3 on the heart.
Methods Animal
model
Male Wistar-Furth rats were hypophysectomized by the vendor (Harlan SpragueDawley Laboratories, Indianapolis, IN, USA) immediately after weaning at about three weeks of age by the transpharyngeal
approach. To thoroughly wash pituitary hormones and to observe growth characteristics, animals were held for an additional eight weeks before entry into the study. Rats were housed in a warm environment and received tap water and chow ad libitum. This study was reviewed and approved by the Institutional Animal Care and Use Committee. Criteria for hypophysectomy Rats were considered to be hypophysectomized if their body weight increased less than 5 %/week for each of the 6 weeks prior to entry into the experiment. As a further assay of complete hypophysectomy, TSH was measured in the trunk blood in 7 placebo treated rats at the time of sacrifice with specific rat TSH materials (5153B) provided by the National Hormone Distribution Program (NIADDK, Bethesda, MD, USA). TSH was undetectable in each sample (< 1.5 lU/ml). Hormone treatment Animals received placebo or hormone treatment for 2 weeks by a mini-osmotic pump (model 2002, Alza, Palo Alta, CA, USA) implanted subcutaneously and interscapularly under volatile anaesthesia. Placebo infusion consisted of 0.045% saline. T3 (Sigma, St. Louis, MO, USA) was dissolved in alkaline water, diluted in saline and infused at a dose of 2.O~gllOOg of body weight/day. Recombinant human GH (Genentech, So. San Francisco, CA, USA) was diluted in saline and infused at a dose of 70pgf 100 g of body weight/day. The combination of T3 and GH was prepared so as to deliver through the pump the same amount as when individually metered. In a sample of 11 rats treated with either placebo or growth hormone alone, the serum level of T3 was undetectable ((25 ngfdl; normal >60 ng/dl) in 10 of the rats, and was 57ng/dl in the other. In a sample of 12 rats treated with either T3 or T3 plus GH, the serum level of T3 was 284 f 42 ng/dl. Protocol Animals were weighed in duplicate measurements on a triple beam balance on the day of pump implantation and the day of sacrifice.
Local IGF-I in T3 Regulated Cardiac Growth Fourteen days after pump implantation, the animals were decapitated and trunk blood collected. The pump was removed at the time of sacrifice and separately weighed. Animal weight does not include pump weight. The cardiac ventricles were harvested at the time of sacrifice and frozen in liquid Nz.
RNA extraction Total RNA was extracted from the ventricles by homogenization in guanidium isothiocyanate and pelleted by ultracentrifugation through a cesium chloride cushion [26]. Purity and concentration of the recovered RNA were assessed by ultraviolet spectrophotometry.
IGF-I
gene expression
The template for probe production was a subclone of an IGF-I clone [19] which contains a cDNA sequence of the 5’ end of rat IGF-I mRNA. An antisense cRNA probe was uniformly labeled with 32P UTP (3000 Ci/mM, ICN, Irvine, CA, USA) by transcription with T7 RNA polymerase (Promega, Madison, WI, USA). For the RNase protection assay, the probe and 25pg aliquots of either sample RNA or internal negative control (yeast tRNA) were solution hybridized in 75% formamide and salt buffer at a temperature of 51 OC. After overnight hybridization, the reaction was treated with a combination of RNase A and Tl (640pg/ml and 2pg/ml, respectively) (Pharmacia, Piscataway, NJ, USA) in reaction buffer for h at 30%. The mixture was incubated with sodium dodecyl sulfate and proteinase K, and then extracted with The reaction products phenol-chloroform. were precipitated with carrier glycogen by ethyl alcohol. The protected fragments were solubilized in formamide loading buffer and were separated on an 8% polyacrylamide denaturing gel. The gel was autoradiographed and the bands on the exposed film were quantitated by scanning densitometry. The homology of the probe sequence to that of rat liver RNA results in three protected fragments of 322, 297, and 241 nt, which correspond to class A’, B’ , and C’ alternatively spliced transcripts, respectively, of the 5’-untranslated regions of IGF-I mRNA.
633
Statistical analysis The study consisted of four independent treatment groups: placebo (10 rats), T3 (10 rats), GH (11 rats) and GH + T3 (7 rats). Group measurements of heart and body weight were calculated as mean f SD. Body weight could be measured both before (baseline) and after (final) treatment, whereas heart weight could be obtained only at the time of sacrifice. An estimate of baseline heart weight was made by constructing a regression formula for the placebo treated rats (final body weight vs. final heart weight) and applying the parameters of the formula to the baseline body weight of the hormone treated rats. Gene expression studies were conducted in five rats from each group, and reactions were performed in duplicate. Group measurements of IGF-I were calculated as mean f S.E.M. For each variable, group comparisons were made from a oneway analysis of variance, and, if significant, individual comparisons were made by the Bonferroni correction for multiple comparisons of the “t” test [27]. The significance level for the statistical tests was set at 0.05 (PCO.05).
RcSUltS Boc& and cardiac growth As expected, chronic administration of GH, T3 or GH + T3 had substantial effects on somatic and cardiac growth compared to placebo-treated hypophysectomized control rats. Treatment with GH resulted in significant gains in body weight compared to placebo [Fig. l(a)]. The addition of T3 (i.e. GH + T3) had no additional effect on body growth, whereas T3 treatment alone resulted in a body weight change which was no different than placebo (5% vs. 3 % change in body weight, respectively, p = N.S.). T3 administration did, however, result in significant cardiac growth compared to placebo [30% vs. 5% change in cardiac weight, respectively, PcO.05, Fig. l(b)]. GH and GH + T3 treatment also resulted in significant increases in cardiac weight. When cardiac growth was normalized with respect to accompanying body growth, it was seen that only in the presence of T3 was cardiac growth increased relative to body growth [Fig. l(c)].
634
J. M. Kupfer
PloCeDO
T3
CH
+T3
l
i .!? ; :: a i .c% E :
s .s z B
60
T 1
60 40 20 0 Placebo
T3
GH
GH+T3
Ploieao
T3
GH
CHtT3
500 400
x E 0
300
4: ;
200
F 5 t Ie
and S. A. Rubiu
100 0
FIGURE 1. Effects of placebo, thyroid hormone (T3), growth hormone (GH) or the combination (GH + T3) on body and cardiac growth. (a) Percentage change in body weight; (b) Percentage change in cardiac weight; (c)Final heart weight in milligrams per 100 grams of final body weight. Data are expressed as mean f S.D. for a sample size of 8 to 10 in each group. *, P
24, 631-639 (1992)
Di%rentialRegulationofInsulin-likeGrowthFactorIbyGrowth Hormone andThyroidH omnone in the Heart Hypophys-Rats
ofJuvenile
Joel M. Kupfw and Stanley A. Rubin From the Division of Cardiology, Department of Medicine, Ceakrs-Sinai Medical Center/UCLA Medicine, Los Angeles, CA 90048, USA
School of
(Received 25 October 1991, accepted in revisedform 23 January 1992) JOEL M. KUPFER AND STANLEY A. RUBIN, Differential Regulation of Insulin-like Growth Factor I by Growth Hormone and Thyroid Hormone in the Heart of Juvenile Hypophysectomiaed Rats. Journal of Molecular and CcNular Cardiology (1992) 24, 631-639. Recent data suggest that the heart can act as both a source and target for the actions of polypeptide growth factors. Insulin-like growth factor I (IGF-I) is a polypeptide that has both mitogenic and differentiation properties that function at the autocrine/paracrinc level, and has recently been demonstrated to be expressed in the heart. This knowledge, coupled with the observation that thyroid hormone (T3) promotes relative cardiac growth compared to the proportional increases in body and heart growth evoked by growth hormone (GH), lead us to speculate whether differential induction of cardiac IGF-I may account for the specialized trophic effects of T3 on the heart. Cardiac IGF-I gene expression was studied in an in uivo model in which cardiac growth in the hypophysectomiaed juvenile rat was stimulated with either GH, T3 or GH + T3. Two week infusions of T3 that resulted in cardiac growth, but no gain in body weight, resulted in a 4.6-fold increase in cardiac IGF-I mRNA levels compared to hypophysectomiaed controls. GH infusions that resulted in similar cardiac growth, but were accompanied by proportional body growth, had no effect on cardiac IGF-I mRNA levels. These data are the first to demonstrate stimulation of cardiac IGF-I mRNA levels by T3 and further support cardiac autocrine/paracrine actions for this polypeptide growth factor. KEY WORDS:
Autocrine;
Paracrine;
Cardiac
growth;
Myosin
Introduction In recent years it has been recognized that cellular growth and maturation can be regulated through cellular oncogenes and the local autocrine or paracrine actions of polypeptide growth factors [I, 21. Along these lines there is an emerging view that signals for cardiac growth, whether initiated at the hormonal or mechanical level, result in expression of cellular oncogenes and polypeptide growth factors that act at the autocrine or paracrine level to coordinately direct the heart towards a hypertrophic response by regulating protein metabolism at the molecular level [3, 4 for reviews see 5, 61. One important polypeptide growth factor that is known to play a pivotal role in somatic ‘Abbreviations TSH, thyroid
used in this paper: IGF-I, stimulating hormone.
insulin-like
+ 09 $03.00/O
chain,
growth and development is insulin-like growth factor I (IGF-I, footnote 1) [7, 81. According to the “somatomedin hypothesis”, IGF-I exerts its growth permissive effects through classic endocrine mechanisms in which GH, from the pituitary, stimulates hepatic production of IGF-I which then enters the circulation to act on distant tissues to promote growth. However, it is now recognized that multiple extrahepatic tissues express IGF-I, even in the hypophysectomized state, and there is increasing experimental evidence to suggest that locally produced IGF-I acts to stimulate nucleic acid synthesis, and to regulate cellular proliferation and differentiation [ 9, II]. This concept of local regulation of growth through autocrine/paracrine IGF-I
growth
Please address all correspondence and reprints to: Joel M. Center, 8700 Beverly Blvd. Los Angeles, CA 90048. 0022-2828/92/060631
heavy
factor Kupfer,
I; GH,
growth
Division
hormone;
of Cardiology,
T3, l-triiodothyronine; Cedars-Sinai
@ 1992 Academic
Medical
Press Limited
J. M. Kupfer and S.A. Rubin
632
action has received further sunnort from reports which demonstrate that other hormones and growth factors that exert trophic effects increase target organ IGF-I mRNA levels to a greater extent than GH [1.2-161. This is also apparent in muscle cells, where autocrine IGF-I expression is known to play an important role in smooth muscle cell proliferation and in skeletal muscle cell differentiation [17, 181. In the heart, myocyte IGF-I expression increases substantially during the first 50 post-natal days consistent with an important developmental role [19-211. Thyroid hormone is known to exert a strong trophic influence on the heart and promotes cardiac growth in the absence of a proportional increase in body weight [22-251. In contrast GH promotes proportional body and cardiac growth. Such data suggest that thyroid hormone, directly or indirectly, induces specific growth promoting factors within the heart. Recognizing the important growth promoting actions of IGF-I and its potential to act at the autocrine/paracrine level, we examined whether T3 preferentially regulates cardiac IGF-I expression compared to GH, which may explain at least part of the specific growth promoting action of thyroid hormone on the heart. To address this question we employed an in vivo model of cardiac growth using hypophysectomized juvenile rats treated for two weeks with physiological doses of GH, T3, or GH + T3. At the end of the treatment period left ventricular IGF-I mRNA expression was studied and correlated with the different patterns of growth. Our results demonstrate that cardiac growth resulting from T3 treatment is accompanied by significant induction of cardiac IGF-I expression compared to hypophysectomized placebo controls or chronic GH treatment. Our findings offer a potential mechanism to account for the specific trophic actions of T3 on the heart.
Methods Animal
model
Male Wistar-Furth rats were hypophysectomized by the vendor (Harlan SpragueDawley Laboratories, Indianapolis, IN, USA) immediately after weaning at about three weeks of age by the transpharyngeal
approach. To thoroughly wash pituitary hormones and to observe growth characteristics, animals were held for an additional eight weeks before entry into the study. Rats were housed in a warm environment and received tap water and chow ad libitum. This study was reviewed and approved by the Institutional Animal Care and Use Committee. Criteria for hypophysectomy Rats were considered to be hypophysectomized if their body weight increased less than 5 %/week for each of the 6 weeks prior to entry into the experiment. As a further assay of complete hypophysectomy, TSH was measured in the trunk blood in 7 placebo treated rats at the time of sacrifice with specific rat TSH materials (5153B) provided by the National Hormone Distribution Program (NIADDK, Bethesda, MD, USA). TSH was undetectable in each sample (< 1.5 lU/ml). Hormone treatment Animals received placebo or hormone treatment for 2 weeks by a mini-osmotic pump (model 2002, Alza, Palo Alta, CA, USA) implanted subcutaneously and interscapularly under volatile anaesthesia. Placebo infusion consisted of 0.045% saline. T3 (Sigma, St. Louis, MO, USA) was dissolved in alkaline water, diluted in saline and infused at a dose of 2.O~gllOOg of body weight/day. Recombinant human GH (Genentech, So. San Francisco, CA, USA) was diluted in saline and infused at a dose of 70pgf 100 g of body weight/day. The combination of T3 and GH was prepared so as to deliver through the pump the same amount as when individually metered. In a sample of 11 rats treated with either placebo or growth hormone alone, the serum level of T3 was undetectable ((25 ngfdl; normal >60 ng/dl) in 10 of the rats, and was 57ng/dl in the other. In a sample of 12 rats treated with either T3 or T3 plus GH, the serum level of T3 was 284 f 42 ng/dl. Protocol Animals were weighed in duplicate measurements on a triple beam balance on the day of pump implantation and the day of sacrifice.
Local IGF-I in T3 Regulated Cardiac Growth Fourteen days after pump implantation, the animals were decapitated and trunk blood collected. The pump was removed at the time of sacrifice and separately weighed. Animal weight does not include pump weight. The cardiac ventricles were harvested at the time of sacrifice and frozen in liquid Nz.
RNA extraction Total RNA was extracted from the ventricles by homogenization in guanidium isothiocyanate and pelleted by ultracentrifugation through a cesium chloride cushion [26]. Purity and concentration of the recovered RNA were assessed by ultraviolet spectrophotometry.
IGF-I
gene expression
The template for probe production was a subclone of an IGF-I clone [19] which contains a cDNA sequence of the 5’ end of rat IGF-I mRNA. An antisense cRNA probe was uniformly labeled with 32P UTP (3000 Ci/mM, ICN, Irvine, CA, USA) by transcription with T7 RNA polymerase (Promega, Madison, WI, USA). For the RNase protection assay, the probe and 25pg aliquots of either sample RNA or internal negative control (yeast tRNA) were solution hybridized in 75% formamide and salt buffer at a temperature of 51 OC. After overnight hybridization, the reaction was treated with a combination of RNase A and Tl (640pg/ml and 2pg/ml, respectively) (Pharmacia, Piscataway, NJ, USA) in reaction buffer for h at 30%. The mixture was incubated with sodium dodecyl sulfate and proteinase K, and then extracted with The reaction products phenol-chloroform. were precipitated with carrier glycogen by ethyl alcohol. The protected fragments were solubilized in formamide loading buffer and were separated on an 8% polyacrylamide denaturing gel. The gel was autoradiographed and the bands on the exposed film were quantitated by scanning densitometry. The homology of the probe sequence to that of rat liver RNA results in three protected fragments of 322, 297, and 241 nt, which correspond to class A’, B’ , and C’ alternatively spliced transcripts, respectively, of the 5’-untranslated regions of IGF-I mRNA.
633
Statistical analysis The study consisted of four independent treatment groups: placebo (10 rats), T3 (10 rats), GH (11 rats) and GH + T3 (7 rats). Group measurements of heart and body weight were calculated as mean f SD. Body weight could be measured both before (baseline) and after (final) treatment, whereas heart weight could be obtained only at the time of sacrifice. An estimate of baseline heart weight was made by constructing a regression formula for the placebo treated rats (final body weight vs. final heart weight) and applying the parameters of the formula to the baseline body weight of the hormone treated rats. Gene expression studies were conducted in five rats from each group, and reactions were performed in duplicate. Group measurements of IGF-I were calculated as mean f S.E.M. For each variable, group comparisons were made from a oneway analysis of variance, and, if significant, individual comparisons were made by the Bonferroni correction for multiple comparisons of the “t” test [27]. The significance level for the statistical tests was set at 0.05 (PCO.05).
RcSUltS Boc& and cardiac growth As expected, chronic administration of GH, T3 or GH + T3 had substantial effects on somatic and cardiac growth compared to placebo-treated hypophysectomized control rats. Treatment with GH resulted in significant gains in body weight compared to placebo [Fig. l(a)]. The addition of T3 (i.e. GH + T3) had no additional effect on body growth, whereas T3 treatment alone resulted in a body weight change which was no different than placebo (5% vs. 3 % change in body weight, respectively, p = N.S.). T3 administration did, however, result in significant cardiac growth compared to placebo [30% vs. 5% change in cardiac weight, respectively, PcO.05, Fig. l(b)]. GH and GH + T3 treatment also resulted in significant increases in cardiac weight. When cardiac growth was normalized with respect to accompanying body growth, it was seen that only in the presence of T3 was cardiac growth increased relative to body growth [Fig. l(c)].
634
J. M. Kupfer
PloCeDO
T3
CH
+T3
l
i .!? ; :: a i .c% E :
s .s z B
60
T 1
60 40 20 0 Placebo
T3
GH
GH+T3
Ploieao
T3
GH
CHtT3
500 400
x E 0
300
4: ;
200
F 5 t Ie
and S. A. Rubiu
100 0
FIGURE 1. Effects of placebo, thyroid hormone (T3), growth hormone (GH) or the combination (GH + T3) on body and cardiac growth. (a) Percentage change in body weight; (b) Percentage change in cardiac weight; (c)Final heart weight in milligrams per 100 grams of final body weight. Data are expressed as mean f S.D. for a sample size of 8 to 10 in each group. *, P
