| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
(PPAR) Heterozygous Mice: Involvement of PPAR in a Negative Feedback Regulation of Growth Hormone Action
Center for Integrative Genomics (J.R., S.I.A., P.E., L.M., N.S.T., W.W., B.D.), NCCR Frontiers in Genetics, University of Lausanne, CH-1015 Lausanne, Switzerland; Centre Médical Universitaire (J.S.), Département de Physiologie, CH-1211 Geneva 4, Switzerland; and Institut de Génétique et de Biologie Moléculaire et Cellulaire (D.M., P.C.)/Centre National de la Recherche Scientifique/Institut National de la Santé et de la Recherche Médicale/Université Louis Pasteur/Collège de France, 67404 Illkirch cedex, CU de Strasbourg, France
Address all correspondence and requests for reprints to: Professor Béatrice Desvergne, Center for Integrative Genomics, University of Lausanne, Bâtiment de Biologie, CH-1015 Lausanne, Switzerland. E-mail: beatrice.desvergne{at}cig.unil.ch.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
(PPAR) plays a major role in fat tissue development and physiology. Mutations in the gene encoding this receptor have been associated to disorders in lipid metabolism. A thorough investigation of mice in which one PPAR allele has been mutated reveals that male PPAR heterozygous (PPAR +/–) mice exhibit a reduced body size associated with decreased body weight, reflecting lean mass reduction. This phenotype is reproduced when treating the mice with a PPAR- specific antagonist. Monosodium glutamate treatment, which induces weight gain and alters body growth in wild-type mice, further aggravates the growth defect of PPAR +/– mice. The levels of circulating GH and that of its downstream effector, IGF-I, are not altered in mutant mice. However, the IGF-I mRNA level is decreased in white adipose tissue (WAT) of PPAR +/– mice and is not changed by acute administration of recombinant human GH, suggesting an altered GH action in the mutant animals. Importantly, expression of the gene encoding the suppressor of cytokine signaling-2, which is an essential negative regulator of GH signaling, is strongly increased in the WAT of PPAR +/– mice. Although the relationship between the altered GH signaling in WAT and reduced body size remains unclear, our results suggest a novel role of PPAR in GH signaling, which might contribute to the metabolic disorder affecting insulin signaling in PPAR mutant mice. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
, ß or 
, and ) that have specific expression patterns and specific functions. So far, PPAR is the most studied PPAR isotype. Its two isoforms (1 and 2), which differ in their 30 N-terminal amino acids in mice (2), are produced from the same gene by alternative promoter usage and mRNA splicing (2, 3). Whereas PPAR1 expression occurs in several tissues, albeit to various and often low levels, PPAR2 is specifically expressed in fat tissues in which both PPAR1 and PPAR2 are present at high levels (4, 5). PPAR plays a key role in adipocyte differentiation as demonstrated in vitro and in vivo (6, 7, 8, 9), as well as in the maintenance of differentiated mature adipocytes (10). Its natural ligands, on the one hand, comprise polyunsaturated fatty acids and some of their metabolites (11) and, on the other hand, thiazolidinediones (TZDs) are synthetic highaffinity ligands of PPAR, which also trigger its transcriptional activity. The latter are currently used as insulin sensitizers (12), and their ability to bind and to activate PPAR in vitro correlates with their antidiabetic action in vivo, suggesting that PPAR mediates the insulin-sensitizing effect of these molecules (13, 14). In agreement with these observations, mutations in the PPAR gene have been discovered in a few patients with severe insulin resistance (15).
Recently, the use of genetic approaches such as gene targeting has shed additional light on the role of PPAR in adipocyte function and insulin resistance. The PPAR null mutation is embryonic lethal around d 10 of gestation due to a defect in placental development (Refs. 6 and 7 and our unpublished results). A single mouse that was brought to full term by means of a placental rescue technique lacked adipose tissue and showed extreme metabolic defects, similar to those seen in human lipodystrophy (6). PPAR heterozygous mice (PPAR +/–), which are viable and bear no obvious phenotype under normal feeding conditions, have been thoroughly investigated with respect to lipid metabolism. Interestingly, PPAR +/– mice are partially protected from high-fat diet-induced weight gain and insulin resistance when compared with their wild-type (wt) littermates (7). In addition, insulin induction of both glucose disposal and suppression of hepatic glucose production were greater in PPAR +/– mice than in wt mice (16). These surprising results raise the intriguing possibility of a dichotomy in PPAR function with respect to insulin sensitization. On the one hand, PPAR activated by TZDs improves insulin sensitivity in obese mice and, on the other hand, reduced PPAR activity, as in PPAR +/– mice, prevents the development of insulin resistance and improves insulin response. Similarly, a reduction in PPAR activity using a specific antagonist decreasing PPAR/RXR heterodimer activity, also leads to enhanced insulin sensitivity in mice, in support of the latter observation (17, 18).
To further investigate the role of PPAR in metabolic disorders, we analyzed the growth characteristics of PPAR +/– mice, raised under standard conditions as well as neonatal treatment with monosodium glutamate (MSG). MSG induces a hypothalamic lesion, which results in reduced pituitary functions that provoke a reduced body size associated with the development of a marked obesity from age 6–10 wk onward (19). We show herein that male PPAR mutant mice have an impaired GH signaling in white adipose tissue (WAT), which is associated with a defect in postnatal growth, revealing a new and unexpected role of PPAR in growth regulation.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Gene in Mouse gene was disrupted in embryonic stem (ES) cells by homologous recombination with a replacement-type vector, using an approach based on positive-negative selection (Fig. 1A
). In this vector, PPAR genomic sequences containing the exons encoding the A/B domain and part of the DNA-binding domain of the receptor (exon 1 and 5'-region of exon 2) were replaced with a phosphoglycerate kinase- neomycine cassette. Homologous recombination at the PPAR locus in ES cells led to the deletion of both exon 1 containing the translational initiation site of PPAR1 and part of exon 2 encoding the first zinc finger of the DNA-binding domain. This deletion disrupted the production of both the PPAR1 and PPAR2 isoforms. Two independent ES cell clones carrying the mutated allele were injected into blastocysts to generate chimeras. The cross of heterozygous mice obtained from germline transmitter chimeras resulted in the birth of wt and heterozygous animals, but not of null pups (Fig. 1B, right side). PCR-based genotype analysis of the yolk sac of embryonic d 9.5 embryos derived from a PPAR +/– cross, revealed the presence of the expected three genotypes (Fig. 1B, left side), confirming the embryonic lethality of the PPAR null allele mutation (6, 9, 17). In contrast, PPAR +/– pups developed normally, and adult male and female mice were fertile and apparently healthy. As expected, the PPAR +/– mice displayed decreased PPAR mRNA and PPAR protein levels in WAT (Fig. 1, C and D). Some of their phenotypic characteristics have been recently described (17, 20, 21, 22).
|
Heterozygous Mice activity, increased fat accumulation, and body weight gain has been exemplified in mice and human subjects treated with TZDs (23, 24). Based on these findings, we analyzed whether the deletion of one PPAR allele had an effect on body weight gain from the weaning period up to 24 wk of age. Female wt and PPAR +/– mice, raised with standard diet (UT, untreated condition), showed a similar body weight gain and reached an average of 27.9 ± 0.5 g at the age 24 wk (Fig. 2A). In contrast, the PPAR +/– males had a lower body weight than wt mice, over the whole period analyzed, and 24-wk-old PPAR +/– males weighted 2 g less than the wt animals (32.8 ± 0.5 vs. 34.7± 0.5 g). Although modest, this underweight was statistically significant (Fig. 2C, P < 0.05).
|
+/– males, by analyzing body composition of untreated wt and PPAR +/– mice. At 24 wk of age, the absolute amount of lean mass was decreased in PPAR +/– mice compared with wt animals (Fig. 3A). However, this difference was not observed if the lean mass was expressed in percentage of total body weight (Fig. 3B). Total body lipid content, when expressed either in grams or in percentage of body weight, did not differ between wt and PPAR +/– mice under these standard conditions (Fig. 3, C and D). These observations were further confirmed by the weight of the major organs (liver, heart, gastrocnemius muscle, WAT, as well as brown adipose tissue), which was also significantly decreased in PPAR +/– mice compared with wt littermates (Table 1). Although the weight changes of gastrocnemius or of the epididimal fat do not necessarily reflect modifications over the whole muscular or fat tissue, respectively, they show that parts of both lean and fat mass are affected.
|
|
+/– mice were significantly shorter than the wt mice (Fig. 3E
). Thus, deletion of one allele of the PPAR gene results in a general growth retardation of the male mice.
Growth Retardation of PPAR +/– Male Mice Is Aggravated by MSG Treatment
MSG treatment of neonatal mice induces a hypothalamic lesion resulting in reduced pituitary functions that provoke a reduced body size associated with the development of a marked obesity from age 6–10 wk onward (19). Treating PPAR +/– mice with MSG would serve two main purposes: 1) to test whether we could amplify the growth-related phenotype of PPAR +/– mice by using chemical-induced growth alteration; and 2) to appreciate the resistance of PPAR +/– mice to induced obesity. As expected, the body weight of MSG-treated wt mice, for the first 4 wk after birth, was decreased compared with untreated mice, due to the general effect of MSG on body growth. At 12 wk of age, the average body weight of the MSG-treated mice (females and males) began to rise above that of untreated wt mice, and these animals rapidly became morbidly obese. On average, they weighed 8 g more than untreated wt mice. Strikingly, whereas MSG-treated PPAR +/– females followed a similar evolution of their body weight compared with their wt counterparts (Fig. 2B), PPAR +/– males were efficiently protected against MSG-induced weight gain (Fig. 2D). Food intake was evaluated and found to be slightly (90.5 ± 6.7, in percentage of the wt male food intake) but significantly (P < 0.05) decreased in untreated PPAR +/– male mice. However, the difference in food intake is not statistically significant in MSG-treated PPAR +/– vs. MSG-treated wt mice (91.2 ± 4.1 in percentage of the MSG-treated wt male food intake).
Analysis of body composition revealed that MSG treatment in male wt mice was associated with the expected decrease in lean mass (Fig. 3, A and B) and weight of most organs other than adipose tissue (Table 1), increased total body lipid content (Fig. 3, C and D), and reduced body size (Fig. 3E). The same alterations were seen in MSG-treated PPAR +/– mice, thereby aggravating the growth alteration that is already present in untreated PPAR +/– mice. Albeit the total percent decrease in lean mass and body size and percentage increase of body lipids upon MSG treatment are the same for both sets of animals, MSG treatment of PPAR +/– males resulted in a weaker increase in total body lipid accumulation compared with wt mice (Fig. 3, C and D, and Table 1), in keeping with the lower starting body size and weight of PPAR +/– vs. wt mice.
Thus, invalidation of one PPAR allele aggravates the MSG-induced growth retardation, indicating that growth alteration due to the PPAR mutation and to MSG treatment is achieved by additive mechanisms.
Retarded Growth of wt Mice Treated with a PPAR Antagonist
An alternative way to inhibit PPAR activity is to treat mice with a specific antagonist of PPAR. Recently, we have identified and characterized such an antagonist of the phosphonophosphate family (SR202). With respect to both high-fat diet-induced adiposity hypertrophy and insulin resistance, this compound reproduces the effects of the invalidation of one PPAR allele. However, it does not further reduce PPAR activity in PPAR +/– mice (17). This PPAR antagonist effect prompted us to study growth and body composition in male wt mice treated or not with this compound and compare them with those obtained from matched PPAR +/– males. SR-202 was given as food admixture (400 mg/kg/d) starting at 3 wk of age for 10 consecutive wk. SR-202-treated wt mice showed reduced body weight compared with untreated wt mice (Fig. 4A). In addition, body size (Fig. 4B) and lean mass (Fig. 4C) were also decreased, whereas total body lipid content (Fig. 4D) was not modified in treated vs. untreated wt mice. Consistent with this, the weight of most organs of SR-202-treated wt mice was decreased in comparison with untreated wt mice, reaching similar levels of that of PPAR +/– mice (Table 2). These results from 13-wk-old mice also confirmed that the retarded growth phenotype of the PPAR +/– mice described in Fig. 3 (24-wk-old mice) is already present at a younger age.
|
|
activity, either by treatment with a specific antagonist or by invalidation of one allele of the PPAR gene, leads to retarded growth in male mice. These data indicate that the altered growth of PPAR +/– mice is a bona fide consequence of the reduction of PPAR activity in vivo.
Circulating GH and IGF-I Levels in PPAR +/– and wt Mice
GH, released by the pituitary, and its downstream effector, IGF-I, are essential for normal growth and development. Because of the growth alteration in mutant mice, we measured serum levels of these peptides in wt and PPAR +/– mice treated or not with MSG. We found the predicted decrease of both serum GH and IGF-I levels in wt MSG-treated mice compared with wt untreated mice (Fig. 5), confirming the validity of the assays. However, circulating GH (Fig. 5A) and IGF-I (Fig. 5B) levels in PPAR +/– mice were similar to those of their wt littermates, either in untreated condition or after MSG treatment. First, this demonstrated that MSG treatment had the same impact on the hypothalamo-pituitary axis in both genotypes, excluding the hypothesis that the reduced growth of MSG-treated PPAR +/– mice is due to an altered MSG action on the hypothalamus. Second, it showed that invalidation of one PPAR allele had no effect on the overall levels of circulating GH and IGF-I.
|
+/– Mice expression and also expresses high level of IGF-I under the control of GH signaling (27, 28). RNase protection assay (RPA) using liver and adipose tissue of 24-wk-old mice, treated or not with MSG, revealed that IGF-I mRNA levels in liver of untreated wt and PPAR +/– mice were similar. As expected, IGF-I mRNA levels were decreased in the liver of MSG-treated mice compared with untreated mice. However, this decrease was similar in both genotypes (Fig. 6A), and thus does not explain their difference in growth. In contrast, IGF-I mRNA levels in adipose tissue of PPAR +/– mice were lower compared with those of wt mice (Fig. 6B, left panel). This decrease was even more pronounced after MSG treatment, which also affected the IGF-I mRNA levels in adipose tissue of wt mice (Fig. 6B, left panel). This RPA evaluation, which requested the pooling of samples from five to six animals, was confirmed by using quantitative RT-real time PCR (RT-qPCR), on five additional separated tissues prepared from adult male mice (3 months old). The results are expressed as a ratio with hypoxanthine phosphoribosyltransferase-1 (HPRT) mRNA, an internal control the expression of which was not different between wt and PPAR +/– mice (30 ± 2 and 28 ± 3 attomol/µg total RNA, P > 0.05). This method confirmed the significant decrease in IGF-I mRNA levels in adipose tissue of PPAR +/– mice compared with age-matched wt mice (Fig. 6B, right panel). Most importantly, we found a corresponding decrease in IGF-I protein levels in adipose tissue of PPAR +/– mice compared with wt mice (Fig. 6C).
|
Effect of GH on IGF-I and PPAR mRNA Levels in WAT
Lower levels of IGF-I expression in the WAT of PPAR +/– mice might be either directly related to the decrease of PPAR activity or a consequence of an altered GH action. A direct effect of PPAR activity on IGF-I expression is unlikely because treatment with troglitazone (a PPAR ligand, 150 mg/kg/d during 3 d) of wt mice failed to increase IGF-I mRNA levels, suggesting that IGF-I is not a direct target gene of PPAR (data not shown). We then investigated IGF-I gene expression in WAT, in response to acute administration of GH. Recombinant human GH (rhGH, 1 mg/kg) was injected sc in 4-wk-old wt and PPAR +/– mice. The mice were killed 4 h later, and the WAT was removed for the determination of IGF-I mRNA levels. As expected, rhGH increased IGF-I mRNA level in the WAT of wt mice (Fig. 7A). In contrast, the levels of IGF-I mRNA in the WAT of PPAR +/– mice were low, as shown above (Fig. 6B), and remained unchanged after GH stimulation (Fig. 7A). These results indicate that acute GH action is altered in the WAT of PPAR +/– mice and suggest that the decrease of IGF-I mRNA levels in these animals might be related to an altered GH-signaling pathway. It must be noted that rhGH also binds and activates the mouse prolactin receptor. However, there are no reports of prolactin-induced IGF-I expression in adipose tissue.
|
and GH signaling pathways, we investigated the acute effect of rhGH on PPAR expression levels in the WAT. Interestingly, GH injection induced a decrease of the PPAR mRNA levels in the WAT of wt mice (Fig. 7B). However, the already low PPAR mRNA levels in the WAT of PPAR +/– mice were not further decreased by the injection of rhGH (Fig. 7B). Thus, these data demonstrate that GH decreases PPAR expression in the WAT and confirm that GH signaling is altered in the WAT of PPAR +/– mice.
Expression of Suppressors of Cytokine Signaling (SOCS)-2 in WAT of PPAR +/– Mice
SOCS are a group of related proteins implicated in the negative regulation of cytokine action through inhibition of the transduction pathway of both janus kinase-signal transducers and activators of transcription (STAT). Among the eight proteins belonging to the SOCS family (SOCS1–7 and cytokine-inducible SH2 protein), SOCS-2 was reported to be induced by GH in vitro and in vivo and, at least in some tissue culture assays, to have a dose-dependent feedback inhibitory effect on GH (29, 30, 31). The essential role of SOCS-2 in growth control is underscored by the phenotype of SOCS-2 –/– mice, which grow significantly larger than their wt littermates (32, 33). Interestingly, SOCS-2 mRNA levels were significantly increased in the WAT of untreated PPAR +/– mice compared with their wt littermates (Fig. 8A). MSG treatment did not significantly modify SOCS-2 mRNA levels in the WAT of wt mice compared with untreated wt mice. However, MSG-treated PPAR +/– mice showed higher levels of SOCS-2 mRNA compared with MSG-treated wt mice. Thus, the pattern of SOCS-2 expression in WAT, both in MSG-treated and untreated mice, was inversely correlated to PPAR levels, indicating that PPAR might indeed be responsible for the SOCS-2 pattern of expression and for the subsequent effects on GH signaling.
|
activity in PPAR +/– mice or to a consequence of the lack of GH inhibition on PPAR expression, we investigated the effect of both rhGH and troglitazone on SOCS-2 mRNA levels in the WAT. One hour after rhGH administration (34), SOCS-2 mRNA levels in the WAT of wt mice were increased, as expected (Fig. 8B, left panel). However, the already high levels of SOCS-2 in PPAR +/– mice were not anymore sensitive to GH action (Fig. 8B, left panel). RT-qPCR in individual tissues confirmed the results obtained by RPA and established a statistically significant increased expression of SOCS-2 mRNA and a lack of SOCS-2 response to GH stimulation in the adipose tissue of PPAR mutant mice compared with their wt littermates (Fig. 8B, right panel). Most interestingly, troglitazone treatment administered to wt mice decreased SOCS-2 mRNA levels in the WAT (Fig. 8C), indicating that PPAR activity mediates the regulation of SOCS-2 expression.
In summary, the decreased PPAR activity in PPAR +/– mice is responsible for a high expression of SOCS-2 in the WAT, which results in an alteration of GH signaling in that tissue, possibly contributing to the mild alteration of the body growth of the PPAR +/– mice.
Expression of IGF-I and SOCS-2 in Females
To determine whether the proposed mechanism concerning altered growth in males was also present in females, we investigated IGF-I and SOCS-2 expression, by RT-qPCR, in WAT of 3-month-old female mice of both genotypes (n = 5 per group). As in males, there was more than a 2-fold reduction of IGF-I mRNA levels, concomitant with a more than 2-fold increase in SOCS-2 mRNA levels, in PPAR +/– females compared with their wt female littermates (Fig. 8D). These results might argue against a role of SOCS-2 and IGF-I regulation in the differences observed between the response to the altered PPAR activity of the males and the females. However, adipose tissue of wt females express lower basal levels of SOCS-2 and IGF-I than wt males, whereas the absolute values of HPRT, used as a control, do not differ between males and females (30 ± 2 and 29 ± 3 attomol/µg total RNA in wt males and females respectively, P > 0.05).
Altogether, these results indicate that, compared with wt males, wt females exhibit a significant decrease of SOCS-2 and IGF-I mRNA levels, albeit their sensitivity to PPAR gene dosage is similar to that seen in males.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
in postnatal growth and development of male mice. In two models of decreased PPAR activity, the growth of mice is altered with a decrease of body size and the corresponding reduction in the weight of several organs and, consequently, total lean mass. Although serum levels of GH and IGF-I are similar in wt and mutant mice, IGF-I mRNA levels are decreased in the WAT of PPAR +/– mice and remain insensitive to acute GH treatment, which suggests an alteration of GH signaling in this tissue. Importantly, and consistent with the above mentioned observations, we demonstrate that the mRNA levels of SOCS-2, a protein known to be an essential negative regulator of GH signaling, is markedly increased in the WAT of PPAR +/– mice.
Sexual Dimorphism of the PPAR +/– Mice Phenotype
Sexual dimorphism of GH action is well established if not well understood (35). One main mechanism contributing to this gender dimorphism stems from intrinsic differences in the GH secretion patterns which act independently as regulator of GH bioactivity (36). It has also been suggested that gender differences in insulin sensitivity are, at least in part, linked to the particularities of GH action (37). The sexual dimorphism of the PPAR +/– phenotype includes both the body size (herein) and the susceptibility to develop insulin resistance upon high-fat diet (Rieusset, J., B. Desvergne, and W. Wahli, manuscript in preparation), suggesting a link between the two observations. There are so far few clues to explain the molecular mechanism of the gender differences. Our study demonstrates that females express lower levels of SOCS-2 and IGF-I mRNA than males. Although PPAR +/– females present similar alterations as PPAR +/– males in the expression of these genes, the difference of expression levels of SOCS-2 and IGF-I between males and females is likely to contribute to the metabolic sexual dimorphism, particularly with respect to insulin sensitivity in WAT. Regarding the lack of growth phenotype in PPAR +/– females, it is interesting to note that in SOCS-2 mutant mice, the gigantism phenotype is less marked in –/– female than in –/– male mice and is absent in +/– females although giving an intermediate phenotype in +/– males (33). These observations suggest that the growth of females is less sensitive to the change in mRNA levels of SOCS-2. Consequently, the 2-fold change in SOCS-2 gene expression between wt and PPAR +/– females might not be sufficient to trigger an altered response in terms of growth in PPAR mutant females.
Action of PPAR on the GH/IGF-I Axis
According to the somatomedin (i.e. IGF-I) hypothesis proposed three decades ago, GH (i.e. somatotropin)-mediated somatic growth is dependent on the endocrine form of IGF-I that is mainly produced and secreted by the liver (38). In agreement with this hypothesis, IGF-I null mice are growth retarded and fail to respond to GH treatment. The surprise came from the phenotype of mice with a liver-specific knockout of the IGF-I gene (26). This animal model confirmed that most of the circulating IGF-I is produced by the liver. However, the liver IGF-I-deficient mice grew normally, suggesting that somatic growth is controlled via a GH-dependent local production of IGF-I, which would act as a paracrine/autocrine signal. Whereas in some tissues, such as lung, heart, or testis, the expression of IGF-I is GH independent, expression of IGF-I in liver and adipose tissue is GH dependent (39, 40). In our study, male PPAR +/– mice presented an altered body growth associated with a normal serum IGF-I concentration, consistent with little if any PPAR activity in normal liver. In contrast, local production of IGF-I in WAT is strongly decreased in PPAR +/– mice. This decrease is associated with a marked increase in SOCS-2 gene expression. The role of SOCS-2 in growth control is best exemplified in SOCS-2 –/– mice, which grow significantly larger than their wt littermates (32, 33). It is known that SOCS-2 is induced by GH and subsequently has a reciprocal dual effect on GH signaling, as shown in vitro and in vivo (29, 30, 31, 41). Mild increased levels of SOCS-2 inhibit approximately half of all GH-induced STAT5 activity, whereas a high SOCS-2 concentration would increase activity. The mechanism of this dual effect is not presently known, but the moderate increase (2-fold) in SOCS-2 expression levels observed in PPAR +/– mice is consistent with an inhibition of the GH signaling, resulting in decreased IGF-I expression. In addition, the increase of SOCS-2 expression in the WAT of PPAR +/– mice may also impair IGF-I signaling itself. Indeed, SOCS-2 has been shown to interact with the IGF-I receptor (42). Thus, our data suggest that the decrease in IGF-I expression in WAT and its lack of response to acute GH administration in PPAR +/– mice are due to an increased expression of SOCS-2 (Fig. 9).
|
is directly involved in controlling SOCS-2 gene expression is not only revealed by the invalidation of one PPAR allele in mice but is also corroborated by the negative regulation of SOCS-2 expression in wt animals treated with a PPAR ligand. In further support of our observation, a recent work demonstrated that in cell culture, either PPAR or PPAR was able to repress STAT5b transcriptional activity (43). STAT5b is an important mediator of GH signaling, and dominant-negative STAT5b inhibits expression of SOCS-2 (44). It is tempting to speculate that the induction of STAT5 activity associated with decrease of PPAR expression in PPAR +/– mice might be responsible, at least in part, for SOCS-2 increased expression. However, the elevated levels of SOCS-2 in PPAR+/– mice are constitutive and independent of GH, raising the possibility of an additional PPAR-dependent mechanism.
Body Size and PPAR Action on the GH/IGF-I Axis
A careful dissection of the GH/IGF-I pathway has been performed via analyses of compound mutations of the GH receptor –/– and IGF-I –/– (40). It revealed that whole-body growth reflects the sum, in various tissues/ organs, of variable effects of 1) the endocrine and GH-dependent action of IGF-I, 2) the overlapping activity of GH and IGF-I in a few extrahepatic tissues (GH-dependent), and 3) some GH- and IGF-I-independent activities. At the present time, it is therefore difficult to evaluate how much of the mild but proportionate growth defect in PPAR +/– mice is due to the decreased expression of IGF-I in the adipose tissue. This would require evaluating whether the decreased expression of IGF-I is present in all adipose tissue depots. Further studies, using for example conditional WAT-IGF-I-deficient mice generated by a Cre/loxP system, will allow the evaluation of the importance of WAT in postnatal growth.
It must be noted that in both genetic and chemical models of reduced PPAR activity, the defect in growth was measurable after the weaning period. The retardation is mild, but it must be taken into account that the study was conducted with PPAR +/– animals only, because the null allele mutation of PPAR is embryonic lethal. The difficulty in genotyping and individualizing pups precluded, at this time, a thorough analysis of growth before weaning. However, the fact that the weight of PPAR +/– mice was already lower at the age of 3 wk when weaning takes place suggests that the growth defect is not related to the food transition but is initiated earlier. Given the fact that GH does not act before weaning, this implies that a GH-independent action needs to be considered for the early growth defect in PPAR +/– mice. In this respect, it remains possible that slight general metabolic alterations in PPAR +/– mice are sufficient to explain the relative growth defect that we observe. However, the change in body weight in PPAR +/– mice is probably not secondary to increased insulin sensitivity associated with the loss of one allele of PPAR because in standard nutritional condition, insulin sensitivity in PPAR +/– mice is not enhanced at this early age (7, 17).
Metabolic Consequences of the Reciprocal Interactions of PPAR and GH Signaling
We (17) and others (7, 16, 18) have shown recently that PPAR +/– mice are more sensitive than wt to insulin and failed to develop an insulin resistance upon high-fat diet. Although the molecular mechanism of this metabolic action remains unclear, the link between PPAR and GH that we are demonstrating here (Fig. 9) suggests interesting hypotheses. In contrast to their coordinated role in growth control, GH and IGF-I seem to have opposing effects on insulin action. The mice with a loss of hepatic-derived IGF-I (liver IGF-I-deficient mice) develop insulin resistance via insulin desensitization in muscle (45). In contrast, GH is known for counteracting insulin action on lipid and glucose metabolism (46, 47, 48). Thus, the negative feedback of IGF-I onto GH secretion is one means by which liver-secreted IGF-I might regulate glucose homeostasis. In our PPAR +/– model, circulating GH levels are normal but are inefficient in the WAT due to an increase in SOCS-2 expression. A thorough exploration of the interaction of the GH/IGF-I system and glucose homeostasis in PPAR +/– mice needs now to be undertaken.
Our results also reveal that GH decreases PPAR mRNA levels in vivo, in the WAT of wt mice. These observations are in agreement with previous studies showing that GH-activated STAT5b inhibits PPAR and PPAR transcriptional activity in vitro (43, 44, 49). This action on PPAR expression may contribute to the antiadipogenic activity of GH. Indeed, GH inhibits adipocyte differentiation, reduces triglyceride accumulation, and increases lipolysis, all processes that reduce adipose tissue mass (50) and are opposite to PPAR functions.
Concluding Remarks
PPAR was shown to have a key role in adipogenesis and be a master controller of the gene response leading to efficient energy storage (51). The exploration presented herein of the mechanism that involves PPAR in postnatal growth points to a complex regulatory network that involves the GH/IGF-I axis and its modulation via SOCS-2 expression (see Fig. 9). These effects were analyzed in WAT, where PPAR is highly expressed. However, the defect in linear body growth that we observed in PPAR +/– mice suggests that alteration of GH action might also occur in the developing bone.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Gene cDNA by PCR using the primers 5'-CCAACTTCGGAATCAGCTCT-3' and 5'-CAATCGGATCGTTCTTCGGA-3' and labeled by random primer labeling (High Prime, Roche, Indianapolis, IN) using [32P]dCTP (3000 Ci/mmol, Amersham Pharmacia Biotech, Arlington Heights, IL). Twelve positive genomic clones from a 129/Sv mouse genomic library established in Lambda FixII (Stratagene, La Jolla, CA) were characterized. The 5'-homology region located between exon B1 and exon 1 was PCR-amplified using the primers 5'-CCATCGATGGATTGCTTTCTCAGGCACCAC-3' and 5'-CCATCGATGGATGAGTCAAGTTCTGGAGGG-3' and subcloned into a unique restriction site ClaI of the replacement vector (52) containing a phosphoglycerate kinase-neomycine cassette (53) as a positive selection marker and a herpes simplex virus-thymidine kinase cassette (54) as a negative selection marker. The 7.7-kb 3'-homology region fragment corresponding to a region between exon 2 and exon 3 was obtained by PCR amplification with the primers 5'-CAATCGGATCGTTCTTCGGA-3' and 5'-GACCCAGCTCTACAACAGGC-3', followed by NheI digestion. After an intermediate subcloning step into a SmaI site of a modified pBluescript vector (Stratagene), the 3'-homology region was introduced as a NotI fragment in the replacement vector. Electroporation of cultured D3 ES cells (55) and screening of clones modified by homologous recombination were as described elsewhere (56). Positive D3 clones were microinjected into the blastocele of 3.5-d-old blastocyst stage embryos isolated from C57BL/6 females (10–15 ES cells per blastocyst). Five to seven injected blastocysts were reimplanted into each uterine horn of CD1 pseudopregnant foster mothers. The sex of the offspring was determined 10 d after birth, and the extent of agouti coat color was evaluated. Male chimeric animals were mated for germline transmission with both 129/Sv and C57BL/6 females.
Animals
wt and PPAR heterozygous (PPAR +/–) mice, on a mixed background (Sv129/C56BL/6), were maintained at 20 C with a 12 h light-dark cycle. Animal experimentations were approved by the relevant commission of the canton of Vaud (Switzerland).
MSG Protocol.
Newborn mice, from d 1 to d 7, received a daily sc injection of MSG (2 mg/g body weight/d) or vehicle (NaCl 0.9%). At 3 wk of age, animals were separated, genotyped for the PPAR-mutated allele, and fed with a standard diet. All mice were weighted on a weekly basis, starting at 3 wk of age. At 24 wk of age, animals were killed by cervical dislocation, and tissues were removed, weighed, and frozen.
SR202 Protocol.
At weaning, wt and PPAR +/– male mice were fed with a standard diet supplemented or not with a specific antagonist of PPAR, SR-202 (400 mg/kg/d) (17) for 10 wk. Mice were killed at the end of the treatment, and the tissues were removed, weighed, and frozen. For measurement of food intake, animals were housed by groups of three per cage for minimizing interindividual variation. After the groups were acclimated, food intake was measured daily over a period of 5 d, taking spillage into account.
Growth and Body Composition
The growth rate of mice was determined on a weekly basis by measuring the body weight from the weaning period until the mice were killed. For measurement of body size, animals were lightly anesthetized with 3% of isoflurane and extended to their maximal length to determine the nose-to-anus distance. For measurement of body composition, the whole carcasses of mice were incised, dried to a constant weight in an oven at 70 C, and then homogenized. Total body fat content was determined by the Soxhlet extraction method using petroleum benzine (57). The fat-free dry mass (lean mass) was obtained by subtraction of body fat content from dry weight.
Hormone Measurements
Blood was withdrawn from the orbital sinus using heparinized microcapillary tubes and immediately centrifuged. The serum was removed, placed in a fresh tube, and frozen immediately. The serum concentration of GH (Amersham Pharmacia Biotech) and IGF-I (Diagnostic System Laboratories, Webster, TX) was determined by enzyme immunoassay.
Acute GH Administration
wt and PPAR +/– male mice (4-wk-old) received sc injection of rhGH (1 mg/kg, Calbiochem, La Jolla, CA) or vehicle (sodium bicarbonate) and killed 1 or 4 h later as indicated. Tissues were removed and immediately frozen for RNA preparation.
RPA
Mouse IGF-I and SOCS-2-specific antisense riboprobes were obtained by in vitro transcription with the T7 RNA polymerase, using as template partial cDNAs cloned in pGEM-T easy vector (Promega Corp., Madison, WI). The amplicon of 300 bp for IGF-I was obtained using IGF-If: 5'-AAGCTTACCAAAATGACCGCACCTGC-3' and IGF-Ir: 5'-TCTAGAAACACTCATCCACAATGCCTGTC-3'; the amplicon of 225 bp for SOCS-2 was obtained using SOCS-2f: 5'-CTGCAGAAGACGTCAGCTGGACCGAC-3', SOCS-2r: 5'-TCTAGATCTTGTTGGTAAAGGCAGTCCC-3'. The PPAR and L27 probes have been described previously (20, 58). The RPA III assay was carried out on total RNA isolated from the epididymal adipose tissue, as described by the manufacturer (Ambion, Inc., Austin, TX).
RT-Real-Time PCR
Total RNA from WAT was prepared using the RNeasy kit (QIAGEN, Courtaboeuf, France). The absolute concentrations of SOCS-2 and IGF-I mRNAs were determined by RT-qPCR using a Light-Cycler (Roche Diagnostics, Meylan, France), as previously described (59). The list of the primers and PCR assay conditions are available upon request. The results were expressed as a percentage of the HPRT mRNA concentration, a housekeeping gene used as an internal control and measured in each samples by RT-real-time PCR.
Western Blot Analyses
For measurement of PPAR protein levels, tissues were lysed in ice-cold lysis buffer (20 mM Na2H2PO4; 250 mM NaCl; Triton-100, 1%; sodium dodecyl sulfate, 0.1%) supplemented with Complete protease inhibitors (Roche). Protein concentration was determined using the Bradford method. Equal amounts of protein extracts were resolved by SDS-PAGE and electrotransferred onto polyvinylidine difluoride membranes. Membranes were blocked 1 h at room temperature with 5% BSA in TBST (10 mM Tris, pH 7.6, 150 mM NaCl, 0.1% Tween-20) and incubated overnight at 4 C with primary antibodies. Membranes were washed in TBST, incubated 1 h at room temperature with horseradish peroxidase-conjugated secondary antibodies, and detected by chemiluminescence (Pierce Chemical Co., Rockford, IL). Equal loading/transfer was verified by Coomassie staining of blots. The antimouse PPAR, full-length protein, and polyclonal antibody were kindly provided by Parke-Davis (Ann Arbor, MI). For measurement of IGF-I protein levels, protein extracts were resolved by Tricine SDS-PAGE (15%) due to the small size of protein (60). Membranes were blocked with 4% milk in TBST and incubated overnight at 4 C with IGF-1 antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). After washing, membranes were incubated 1 h with horseradish peroxidase-conjugated secondary antibodies and detected by chemiluminescence (Pierce).
Statistical Analysis
Values are reported as mean ± SEM. Statistical significance was determined by the unpaired Student’s t test.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Present address for J.R.: Institut National de la Santé et de la Recherche Médicale U449, Faculté de médecine R.T.H. Laennec, rue Guillaume Paradin, 69372 Lyon cedex 08, France.
Present address for P.E.: Department of Physiology/ Neurobiology, University of Basel, Klingelbergstrasse 50/70, CH-4056 Basel, Switzerland.
Abbreviations: ES, Embryonic stem; HPRT, hypoxanthine phosphoribosyltransferase-1; MSG, monosodium glutamate; PPAR, peroxisome proliferator-activated receptor ; rhGH, recombinant human GH; RPA, RNase protection assay; RT-qPCR, quantitative RT-real time-PCR; RXR, retinoid X receptor; SOCS, suppressor of cytokine signaling; STAT, signal tranducer and activator of transcription; TBST, Tris-buffered saline-Tween 20; TZD, thiazolidinedione; WAT, white adipose tissue; wt, wild-type.
Received for publication August 27, 2003. Accepted for publication June 25, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
(mPPAR) gene: alternative promoter use and different splicing yield two mPPAR isoforms. Proc Natl Acad Sci USA 92:7921–7925 gene. J Biol Chem 272:18779–18789 in humans: no alteration in adipose tissue of obese and NIDDM patients. Diabetes 46:1319–1327[Abstract]
is required for placental, cardiac, and adipose tissue development. Mol Cell 4:585–595[CrossRef][Medline]
mediates high-fat diet-induced adipocyte hypertrophy and insulin resistance. Mol Cell 4:597–609[CrossRef][Medline]
is required for the differentiation of adipose tissue in vivo and in vitro. Mol Cell 4:611–617[CrossRef][Medline]
is required in mature white and brown adipocytes for their survival in the mouse. Proc Natl Acad Sci USA 101:4543–4547: binding and activation correlate with antidiabetic actions in db/db mice. Endocrinology 137:4189–4195[Abstract]
agonism and the antihyperglycemic activity of thiazolidinediones. J Med Chem 39:665–668[CrossRef][Medline]
associated with severe insulin resistance, diabetes mellitus and hypertension. Nature 402:880–883[Medline]
deficiency. J Clin Invest 105:287–292[Medline]
antagonist with antiobesity and antidiabetic activity. Mol Endocrinol 16:2628–2644 ameliorates dietinduced obesity and type 2 diabetes. J Clin Invest 108:1001–1013[CrossRef][Medline]
and PPARß mutant mice. J Cell Biol 154:799–814 (PPAR) heterodimer. A basis for new therapeutic strategies. J Exp Med 193:827–838 activation by thiazolidinediones induces adipogenesis in bone marrow stromal cells. Mol Pharmacol 50:1087–1094[Abstract]
