| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Neuroscience and Cell Biology (Y.N., A.G.P.S., S.B., J.E.P.), University of Medicine and Dentistry of New Jersey, Piscataway, New Jersey 08854; Department of Biochemistry & Molecular Biology (P.R.), Oregon Health & Science University, Portland, Oregon 97201; Department of Pathology (T.L.), Columbia University, New York, New York 10032; and Institute of Experimental Clinical Research (J.F.), Aarhus University Hospital, Aarhus DK-8000, Denmark
Address all correspondence and requests for reprints to: John E. Pintar, Department of Neuroscience and Cell Biology, University of Medicine and Dentistry of New Jersey, Piscataway, New Jersey 08854. E-mail: pintar{at}cabm.rutgers.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
85–90% of wild type) when IGFBP-4 was eliminated. To continue to address the roles of these proteins in whole animal physiology, we generated combinational IGFBP KO mice. Mice homozygous for targeted defects in IGFBP-3, -4, and -5 remain viable and at birth were the same size as IGFBP-4 KO mice. Unlike IGFBP-4 KO mice, however, the triple KO mice became significantly smaller by adulthood (78% wild type) and had significant reductions in fat pad accumulation (P < 0.05), circulating levels of total IGF-I (45% of wild type; P < 0.05) and IGF-I bioactivity (37% of wild type; P < 0.05). Metabolically, triple KO mice showed normal insulin tolerance, but a 37% expansion (P < 0.05) of ß-cell number and significantly increased insulin secretion after glucose challenge, which leads to enhanced glucose disposal. Finally, triple KO mice demonstrated a tissue-specific decline in activation of the Erk signaling pathway as well as weight of the quadriceps muscle. Taken together, these data provide direct evidence for combinatorial effects of IGFBP-3, -4, and -5 in both metabolism and at least some soft tissues and strongly suggest overlapping roles for IGFBP-3 and -5 in maintaining IGF-I-mediated postnatal growth in mice. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Under specific experimental conditions, however, various IGFBPs may instead potentiate the action of IGFs both in vitro and in vivo. For example, IGFBP-3 can enhance the IGF-I-mediated DNA synthesis in breast carcinoma cells (10) and osteoblasts (11) in vitro and also stimulate proliferation in PC-3 prostatic cancer cells by a process that depends on both active proteolysis of IGFBP-3 and the presence of IGF-I (12). In vivo studies have also shown that coadministration of IGF-I with IGFBP-3 was more effective in eliciting weight gain in hypophysectomized rats than an even larger dose of IGF-I alone (13). In addition, coadministration of IGF-I and IGFBP-3 can also accelerate wound healing (14), whereas injection of an IGF-I antibody that mimics the effect of IGFBP-3 in vivo led to increased muscle weight (16). In addition to these studies of IGFBP-3, other IGFBPs can also potentiate IGF action. For example, a single local injection of IGFBP-5 to the outer periosteum of the parietal bone of IGF-I KO mice increased ALP activity and osteocalcin levels of calvarial bone extracts (15). In addition to these specific effects, most IGFBPs also more generally act as carrier proteins in the bloodstream and are thought to regulate the efflux of IGFs from the vascular space (17, 18, 19, 20). The IGF/IGFBP complexes prolong the half-lives of IGFs and thus may buffer the potential hypoglycemic effects that could result from high concentrations of circulating unbound IGFs (21, 22).
The ability of IGFBPs to inhibit and stimulate effects IGF actions, or perhaps act independently, is a subject of ongoing interest, but the precise roles of individual IGFBPs in vivo are still largely unknown. This is due both to the complexity of IGFBP family number and regulation, as well as the fact that the majority of evidence has been derived from in vitro studies (7). Transgenic and knockout (KO) strategies in model organisms have provided unique opportunities to perform gain of function, partial loss of function, and loss of function studies for a particular gene. Several lines of transgenic mice overexpressing different IGFBPs generally show specific organ growth retardation (23), consistent with inhibitory effects on IGF action. Transgenic mice in which IGFBP-4 is overexpressed in smooth muscle, for example, exhibit a tissue-specific smooth muscle hypoplasia (24). In contrast to the transgenic studies, the single IGFBP KO mice reported to date have shown more modest phenotypes. For example, IGFBP-2 KO mice only show a reduction in spleen weight (25), whereas IGFBP-1 KO mice show normal size and metabolism but have deficits in liver regeneration (26). It is reasonable to speculate that functional compensation by other members of the IGFBP family may have prevented the appearance of more dramatic phenotypes in these lines. Thus, it is of interest to examine both the phenotype of additional IGFBP KO mice as well as the phenotype if multiple IGFBPs are deleted.
In this study, we generated individual IGFBP-3, -4, and -5 single null mice as well as IGFBP-3-4-5 triple null mice. We explored several aspects of the IGFBP-3-4-5 KO mouse phenotype, including growth rate, body composition, IGF-I and insulin levels, IGF-I bioactivity, ability to regulate glucose homeostasis, and the consequence on activation of the MAPK and AKT signaling pathways in muscle. Our study demonstrates overlapping crucial roles for IGFBPs in regulating development and metabolism in vivo.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
, A and C). IGFBP-4 null mice were born with a 10–15% decrease in body weight and remained 10–15% smaller than wild-type mice in the postnatal period (Fig. 1B), demonstrating that only prenatal growth was impaired by the IGFBP-4 mutation. Although only the IGFBP-4 single KO mice show a modest growth deficit phenotype, IGFBP-3-4-5 triple KO mice exhibit a significantly greater growth retardation than IGFBP-4 mutant mice, which becomes significant at P20 (78% of normal) and continues through at least 14 wk (Fig. 1D).
|
for 14-wk-old mice and, for 9-wk-old mice, Supplemental Table 1 published on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org). However, weights of all fat pads were significantly decreased in the mutant mice (Table 1 and Fig. 2A), which also showed a decreased weight of the quadriceps muscle (see below). To further explore the alteration in fat pad weight, the abdominal fat pads from both wild-type and mutant mice were histologically examined and the distribution of adipose cell area was determined. The mutant mice contain adipocytes that are much smaller compared with those from wild-type mice (Fig. 2B).
|
|
), which were consistent with the genotype. However, an intermediate band that potentially reflects contributions from both IGFBP-1 and -2 in ligand binding studies was not significantly changed in mutant mice (Fig. 3A). We then directly compared the IGFBP-2 levels in wild-type and triple KO mice using Western blot. The triple KO mice also showed unchanged IGFBP-2 levels (Fig. 3A), which by inference indicated that there is also no significant change in IGFBP-1 levels.
|
). The total IGF-I levels in serum in mutant and wild-type mice were also determined after overnight fast via RIA, and it was found that IGFBP-3-4-5 triple KO mice had a more than 50% reduction in total serum IGF-I compared with wild-type mice (Fig. 3C). In addition to measuring total IGF-I in extracted serum, we also measured IGF-I bioactivity in those samples. This was achieved by determining the ability of serum to phosphorylate (i.e. to activate) the IGF-I receptor in vitro using cells transfected with the human IGF-I receptor. This assay was first presented in Ref. 27 and is also described in detail in Materials and Methods. This cell-based bioassay showed that IGF-I bioactivity was even more markedly suppressed in IGFBP-3-4-5 KO mice (37% of wild-type levels) as compared with total IGF-I (Fig. 3D). Overall, the IGF-I levels were clearly decreased in mutant mice and we suggest that, following loss of IGFBP-3, IGFBP-5 and IGFBP-4, the mutant mice cannot efficiently stabilize IGF-I in circulation. As a consequence, the triple KO mice exhibit a significant decrease in IGF-I bioactivity, which leads to the enhanced growth deficit in those mice.
Enhanced Glucose Homeostasis in IGFBP-3-4-5 Null Mice
Because IGFBPs regulate the bioavailability of IGF-I that, like insulin, can mediate glucose uptake and metabolism (22), we tested whether the deletion of individual or combined IGFBPs may affect glucose disposal by investigating glucose tolerance in mutant mice after an overnight fast. The responses to a glucose challenge were decreased in IGFBP-4 mutant mice at 60 min and 90 min, whereas IGFBP-3 and -5 mutant mice showed a normal glucose responses compared with wild-type mice (see supplemental Fig. 2). However, glucose levels after glucose challenge were significantly decreased in BP-3-4-5 triple KO mice compared with wild-type mice as well as IGFBP-3 and IGFBP-5 KO mice over a broad time range (from 30–120 min after glucose injection). Moreover, the decrease of glucose in triple mutant mice was significantly enhanced compared with IGFBP-4 mutant mice (Fig. 4A). This result indicated that the mutant mice have increased glucose uptake that is probably due to the combined effect of deletion of three IGFBPs.
|
). Therefore, increased insulin secretion after glucose challenge appears to contribute to the enhanced glucose clearance in triple KO mice. This finding was also consistent with the increased islet area observed in triple KO mice (see below).
IGF-I levels were also examined following glucose challenge. The IGF-I levels in wild-type mice showed a slight increase after glucose challenge. The IGF-I levels in IGFBP-3-4-5 triple KO mice reached a peak at 30 min and then dropped back to normal levels by 60 min. It is notable that IGFBP-3-4-5 triple KO mice exhibited lower circulating IGF-I levels compared with wild type at all time points examined in this experiment, similar to the lower levels seen after overnight fast (Fig. 4C). However, when the percentage of increase in IGF-I levels between these strains was compared, it was found that IGFBP-3-4-5 triple KO mice had an significantly larger percentage of increase in total IGF-I compared with wild-type mice at 30 min (43%), whereas the wild-type mice only exhibited a slight percentage of increase in total IGF-I at 30 and 60 min (6% and 8%, respectively) (Fig. 4D). Therefore, this transient increase in circulating IGF-I might also contribute to the enhanced glucose homeostasis, although more studies need to be done to determine whether this increase is reflected by increased free or bioactive IGF-I.
Islet Area Is Increased, whereas Insulin Sensitivity Is Unchanged in IGFBP-3-4-5 Null Mice
Because IGFBP-3-4-5 null mice showed a significantly enhanced glucose uptake and at the same time had higher circulating insulin levels following glucose challenge as compared with wild-type animals, it was valuable to measure insulin sensitivity using the insulin tolerance test. The result revealed that the adult triple KO mice exhibited normal insulin sensitivity compared with wild-type mice (Fig. 5A).
|
). However, ß-cell area in adult IGFBP-3-4-5 null mice was significantly increased about 37% (Fig. 5, C and D). These findings were consistent with the increased insulin secretion seen in the glucose tolerance experiment described above. Therefore deletion of IGFBPs leads to an increased ß-cell area that accompanies the enhanced insulin response to glucose challenge.
Erk Signaling Pathway and Weight of Quadriceps Muscle Are Significantly Decreased in IGFBP-3-4-5 Null Mice
Insulin and IGF-I bind to distinct cell-surface receptor tyrosine kinases that regulate a variety of signaling pathways controlling metabolism, growth, and survival (32, 33, 34). IGFBP-3-4-5 null mice had lower circulating IGF-I levels but higher insulin levels and proportionally increased IGF-I levels after glucose challenge. All of these changes indicate that IGFBP-deficient mice might also have alterations in downstream signaling pathways. We investigated Erk/MAPK and AKT phosphorylation in muscle of triple KO mice because muscle is the one of the major sites for the mitogenic and metabolic effects of IGF-I and insulin (35). IGFBP-3-4-5 triple KO mice showed a slightly increased AKT activation level and a significantly decreased Erk/MAPK activation level (72%) compared with wild type (Fig. 6, A and B). Moreover, the quadriceps femoris skeletal muscle from the triple KO mice was smaller than that of wild type (Fig. 6C), which is consistent with the decreased Erk/MAPK activation, and thus appears to contribute, along with the lower fat pad accumulation (see above), to the lower body weight of the triple KO mice.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
IGFBP-4 single KO mice exhibit a 10–15% modest growth deficit at birth, which demonstrates that IGFBP-4 is required for normal prenatal growth. Importantly, other IGFBPs cannot compensate fully, if at all, for IGFBP-4 absence. When compared with three single KO mice, the IGFBP-3-4-5 deficient mice also exhibited a modest growth retardation at the day of birth, similar to the growth deficit exhibited by IGFBP-4 KO alone. In contrast to the single IGFBP-4 KO, however, the growth deficit became significantly more extensive after P20 and was accompanied by significantly decreased fat accumulation and quadriceps muscle mass, and a significant reduction in circulating levels of IGF-I immunoactivity and bioactivity. Previous work has shown that lack of IGF-I results in both an intrauterine growth retardation and a post-weaning growth deficit (36). Because both IGFBP-3 and -5, which are the only IGFBPs known to combine with ALS and IGF-I to form a ternary complex, are both deleted in the triple KO mice, one possibility that is consistent with the data is that the ability to form the ternary complexes is greatly reduced in these mutant mice, which is reflected by the lower circulating IGF-I levels. This hypothesis is also consistent with the observation that ALS-KO mice have markedly reduced IGF-I and IGFBP-3 levels and exhibit a modest reduction in postnatal growth (37). Moreover, the lower circulating IGF-I levels in the triple KO mice are accompanied by a even greater decrease in IGF-I bioactivity (1/3 of wild-type level), which would be sufficient to lead to the pronounced post-weaning growth deficit observed in these mice. Although liver-specific IGF-I null mice exhibit a 75% decrease in circulating IGF-I levels, these mice, unlike the triple KO mice, continue to exhibit normal levels of free IGF-I (38) and thus normal growth and development (39). Our finding suggests that IGF-I needs to be stabilized by the binding proteins to retain sufficient bioactivity to provide effective growth stimulation. However, we cannot exclude the possibility that the decreased bioactivity of IGF-I in serum from the triple KO mice may result from the combined effects of both decreased IGF-I and deletion of IGFBPs. In addition, it is possible that IGF-I-independent growth effects of the deleted IGFBPs and/or local growth effects of the remaining IGFs contribute to the overall size and tissue development in triple KO mice.
IGF-I serves as a growth-essential peptide by binding to the distinct cell-surface receptor tyrosine kinases that control growth and development (32, 33, 34, 40). One of the most important signaling pathways activated by IGF-I is the MAPK pathway. Tyrosine phosphorylation and activation of MAPK can serve as signal for cell growth, differentiation, and survival (41, 42). In our study, the triple KO mice showed significantly decreased activation levels of MAPK (72% of wild-type levels) in muscle. Moreover, the weight of quadriceps femoris skeletal muscle from IGFBP-3-4-5 KO mice was significantly smaller than that of wild-type mice. Together, those findings indicate that overall muscle weight may also be reduced and contribute to the body weight reduction in mutant mice. This possibility is supported by other in vivo models. For example, enhancement of muscle mass is seen in transgenic mice expressing IGF-I locally in the skeletal muscle (43), whereas IGF-I KO mice display features of severe muscle loss (44). However, more studies are needed to explore the locations and onset of decreased muscle mass in the mutant mice.
The major change in tissue organization noted so far in the triple IGFBP KO mice is that the weights of all peripheral fat pads are significantly reduced and that the adipocytes are smaller in size compared with wild-type mice. Recent in vivo studies have demonstrated a role for IGF-I in adipocyte formation. Transgenic mice overexpressing leptin, which is characterized by a reduced plasma levels of glucose and IGF-I, exhibit a 75% decrease in adipocyte size (45), Therefore, the two thirds decrease in circulating bioactive IGF-I present in IGFBP-3-4-5 KO mice may thus be a critical factor that leads to the significant decrease in adipocyte size and thus total fat accumulation in these mice at the ages analyzed.
Metabolic Abnormalities in IGFBP-3-4-5-Deficient Mice
Initial metabolic analysis indicated that IGFBP-3-4-5 null mice exhibited lower fasting glucose levels, significantly decreased circulating IGF-I levels, significantly decreased IGF-I bioactivity and significantly increased insulin secretion after the glucose challenge. Thus, one goal of this study became to investigate whether IGFBP loss affects glucose metabolism by regulating insulin, regulating the bioavailability of IGF-I, or both.
Insulin has multiple reciprocal interactions with IGF-I and its binding proteins. For example, in vivo studies of IGF-I administration have consistently shown an inhibitory effect on insulin secretion (46). In muscle-specific insulin receptor KO mice, IGF-I levels increase to compensate for the lowered glucose uptake resulting from the lack of insulin signaling (47). Thus, the increased insulin level might be a response to the lower circulating IGF-I level in triple KO mice. In addition to these effects, it has been reported that IGF-I at physiological concentrations can inhibit insulin secretion in vivo (48), so the decreased IGF-I level in triple KO mice may release this inhibition on insulin secretion, thereby explaining the increased insulin secretion. The morphological analysis of the pancreas in triple KO mice showed a significant increase in the islet area, which likely contributes to the compensatory increase in insulin secretion following the glucose challenge. Increased insulin activation of its cell-surface receptor would result in increased activation of phosphatidylinositol-3 kinase and subsequent formation of phosphatidylinositol-3 phosphate, which could serve to increase glucose transport (49, 50). The triple KO mice also showed slightly increased activation levels of AKT in muscle, which is consistent with the increased insulin secretion.
IGF-I has been reported to have potent antiapoptotic activity in islet cells. Exogenous IGF-I blocks the in vitro cytokine-induced apoptosis of these cells (51) and human islets transfected with the IGF-I gene are also resistant to IL-1ß-induced and Fas-mediated apoptosis (52). Our study shows that mutant mice have a significant increase in islet cell area in the pancreas, which suggests that the lower circulating IGF-I level in mutant mice are not sufficient to enhance apoptosis of islet cells.
Whereas insulin is the key short term regulator of glucose homeostasis, emerging evidence suggests that the IGF-I contributes to long-term glucose homeostasis (22, 53, 54). The insulin-like activity of IGF-I has been estimated to be 5–10% that of insulin (55). Based on the high circulating IGF-I concentration, the insulin-like potential of IGF-I far exceeds that of insulin itself (IGF-I: 100- 800 ng/ml vs. insulin: 1–4 ng/ml) and yet does not cause hypoglycemia, apparently because it is sequestered into 150-kDa complexes with either IGFBP-3 or IGFBP-5 and ALS. In our study, although IGFBP-3-4-5 KO mice consistently exhibited a significantly lower circulating IGF-I after glucose challenge, the mutant mice showed a transient large percentage of increase in IGF-I compared with their lowered baseline. Thus, the increase of total IGF-I in the circulation, combined with the depletion of IGFBPs that would sequester IGF-I, could be expected to lead to a transient increase in bioactive IGF-I. This hypothesis is consistent with a previous study in which administration IGFBP-3 to the mouse temporarily depleted free IGF and caused a rise in plasma glucose (3). A similar observation was made when IGFBP-1 was administered as an iv bolus in the rat (56), where plasma glucose levels increased transiently by approximately 10%. Thus, we postulate that deletion of IGFBPs attenuates the ability of IGFBPs to buffer IGF bioavailability. As a consequence, there is a comparatively larger transient increase in bioactive IGF-I when the mice are given a glucose challenge, which, along with insulin, may also contribute to the enhanced glucose homeostasis. However, direct bioactive IGF-I measurement in the glucose challenge is needed to test this hypothesis.
Taken together, we have demonstrated that the significant decrease in circulating IGF-I in mutant mice is accompanied by a significant increase in insulin secretion and islet expansion. As a consequence, the mutant mice exhibit an enhanced glucose clearance. Moreover, because mutant mice also showed a transient increase in IGF-I levels after glucose challenge, altered tissue availability of circulating IGF-I may also contribute to the enhanced glucose homeostasis.
Conclusion
This study indicates that individual absence of IGFBP-3 and -5 do not affect growth, indicating possible compensation by the remaining IGFBPs, whereas the absence of IGFBP-4 cannot be fully compensated prenatally by other IGFBP family members. However, when the two major IGFBPs (IGFBP-3 and IGFBP-5) known to contribute to the ternary IGFBP-ALS-IGF complex are deleted, together with IGFBP4, the remaining IGFBPs appear to unable to prevent development of significant growth and metabolic disruptions. Deletion of these multiple IGFBPs then results in a significant post-weaning decrease in body weight that is more severe than modest growth deficit of the IGFBP-4 KO mice and appears to result from a significant decrease in bioactive IGF-I. The triple KO mice also exhibit impaired fat pad accumulation that is accompanied by a decrease in adipocyte size. Moreover, the triple KO mice exhibit decreased muscle weight that may result from decreased Erk activation. In addition, the deletion of multiple IGFBPs also leads to an increased response to glucose challenge, which appears to result primarily from the significant increase in insulin secretion that reflects islet expansion. Taken together, these genetic findings indicate that IGFBPs have critical but complex roles in growth and metabolism in the mouse.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Production and Characterization of IGFBP-4-Deficient Mice
To generate IGFBP-4 KO mice, a 5-kb HindIII/BstI fragment containing part of intron1 was cloned between neo and HSV-TK gene of the KO vector. Another 2.5-kb BstX1/NruI fragment was cloned into the KO vector (see supplemental data). The rest of procedures are the same as the procedure described above.
Production and Characterization of IGFBP-5-Deficient Mice
To generate the IGFBP-5 KO mice, a 9-kb EcoRI/SalI fragment containing part of intron1 up to part of exon4 was cloned between neo and HSV-TK gene of the KO vector. Another 1.3-kb PvuII/EcoRV fragment was cloned into the KO vector (see supplemental data). The rest of procedures are the same as the procedure described above.
Southern Analysis Genotyping
Mouse tail tip DNA was digested with SacI (for IGFBP-3 genotyping), BamHI (for IGFBP-4 genotyping), and EcoRI (for IGFBP-5 genotyping), respectively, run on 0.7% agarose gels and transferred to Hybond (Amersham). Membranes were prehybridized for 1 h at 65 C in Rapid Hyb (Amersham) and then hybridized overnight at 65 C in the same solution containing 32P-labeled probe derived from fragment, shown in Fig. 1. Membranes were washed at 68 C in 1x saline sodium citrate/0.1% sodium dodecyl sulfate three times for 1.5 h, and then exposed to film. For IGFBP-3 genotyping, both a wild-type 6.9-kb fragment and a mutant 6-kb fragment were observed in SacI-digested DNA from IGFBP-3 heterozygous mice. For IGFBP-4 genotyping, both a wild-type 9-kb fragment and a mutant 5.5-kb fragment were observed in BamHI-digested DNA from IGFBP-4 heterozygous mice. For IGFBP-5 genotyping, both a wild-type 4-kb fragment and a mutant 2-kb fragment were observed in EcoRI-digested DNA from IGFBP-5 heterozygous mice (see supplemental Fig. 1).
RT-PCR Analyses
Total RNA was purified from key tissues (kidney, brain, heart, and liver) (n = 3 per genotype) by using the RNeasy Midi protocol (QIAGEN, Valencia, CA) and analyzed by RT-PCR. The primers were chosen based the sequence of exon1 from IGFBP-5 gene. The 5' primer sequence is GCTCGCCGTAGCTCTTTTC and 3' primer sequence is GGTTCTTTCGTGCACTGTGA. The expected amplified fragment is 261 bp.
Generation of IGFBP-3-4-5 Null Mice by Cross-Breeding
IGFBP-3 null mice on a mixed 129/C57 background were first cross-bred with IGFBP-4 null mice. The double heterozygous mutants were intercrossed to generate IGFBP-3-4 null mice that were no different in size than IGFBP-4 mutant mice (data not shown). The homozygous IGFBP-3-4 null mice were then cross-bred with IGFBP-5 null mice. After the triple heterozygous were crossed for two generations, IGFBP-3(–/–)4(–/–)5(+/–), IGFBP-3-4-5 null mice and genetically similar littermate wild-type mice were obtained. IGFBP-3(–/–)4(–/–)5(+/–) mice were mated to establish the comparison between triple mutant and IGFBP-3-4 double mutant growth curve. The wild-type mice and the IGFBP-3-4-5 null mice arising from this cross-breeding continued to be intercrossed to produce the wild-type control and triple KO mice used in subsequent experiments. All mice used in the subsequent experiments were 2–4 months old and on a C57Bl6/129S6 mixed background.
Measurement of Postnatal Growth, Organ Size, Fat Pad, and Skeletal Muscle
IGFBP-3-4-5 null mice and IGFBP-3-4 null mice were first obtained from the IGFBP3(–/–)4(–/–)5(+/-) mating to examine any effects resulting from combined IGFBP-3 and IGFBP-5 deficiency on IGFBP-4 null background. Litters from IGFBP-3(–/–)4(–/–)5(+/-) mating were then analyzed from postnatal d 0 (day of birth) until 30 d, and then once per week for 2 months. Postnatal litters were ear marked, tail clipped for genotyping and weights of individuals were correlated with genotypes from the southern blotting of tail DNA. The adult mice were dissected and different organs (heart, liver, kidney, and lung) were collected and measured. Fat pad are mainly accumulated in four sites: sc, gonadal, perirenal, and retroperitoneal. Adult mice were dissected and fat pads from those four sites were collected and measured. Results are presented as fat content in percentage of total body weight. The freshly isolated gonadal white adipose tissues were fixed overnight in 10% formalin, dehydrated, embedded in paraffin and stained with H/E. The cell size was analyzed using ImageJ software (National Institutes of Health, Bethesda, MD). At least 300 cells from each animal were measured. The quadriceps femoris skeletal muscle was separated and weighed from both wild-type and mutant mice.
Western Ligand Blotting and Western Blotting
One to 3 µl serum from each animal was run on 12–16% polyacrylamide gels using either the Mini-Protean II or the Protean IIxi system (Bio-Rad Laboratories, Inc., Richmond, CA). Samples were electrophoresed and run through 5- to 6-cm gels at 85 V for approximately 3 h. Proteins were transferred to a polyvinylidene fluoride nylon membrane (Millipore Corp., Bedford, MA) which provided a signal intensity equivalent to that observed after transfer to nitrocellulose. Membranes were preblocked with 1% BSA in Tris-buffered saline (TBS) containing 3% Nonidet P-40 and 0.1% Tween 20 before overnight incubation with 400,000 cpm of 125I-IGF-I at 4 C (to show the remaining IGFBPs) or incubation with IGFBP-2 Ab (Upstate Corp., Lake Placid, NY). After sequential washes in TBS containing 0.1% Tween 20 and in TBS alone, ligand blots were exposed to Kodak (Rochester, NY) XAR-5 film for varying durations ranging from 1 d to 3 wk, Western blots were visualized Blots Western lighting chemiluminescence kit (PerkinElmer Life Science, Inc., Foster City, CA).
Glucose, Insulin, Total IGF-I, and Bioactive IGF-I Measurements
Mice were fasted for 16 h (1700–0900 h) followed by ip injection of glucose (2 g/kg body weight). Whole venous blood was obtained from the tail vein at the indicated time points (0, 1, 2, 5, 10, 15, 30, 60, 90, and 120 min) after the glucose injection. Blood glucose levels were measured using a Glucotrend glucometer (Johnson & Johnson, New Brunswick, NJ). To measure insulin levels and IGF levels during a glucose tolerance test, blood was collected in EDTA-containing tubes and then centrifuged. Insulin levels were measured at 0, 1, 2, 5, 10, 15, 30, 60, and 90 min using an RIA rat insulin kit (Amersham, Arlington Heights, IL). Total IGF-I levels were measured at 0, 15, 30, 60, and 90 min using an RIA rat IGF-I kit (Diagnostic Inc., Houston, TX).
Serum levels of bioactive IGF-I was measured by an in-house IGF-I kinase receptor activation assay, based on human embryonic renal cells (EBNA 293) transfected with the human IGF-I receptor gene (27). In brief, cultured cells are stimulated at 37 C with either IGF-I standards (a serial dilution ranging from 0.3 to 10 ng/ml of recombinant human IGF-I from Austral Biologicals, San Ramon, CA) or mice sera diluted 1 in 10 in Krebs Ringer buffer (KRB). After 15 min, samples are removed and the cells lysed. Then, crude cell lysates were transferred to an assay that detects the concentration of phosphorylated (i.e. activated) IGF-I receptors. This assay uses a monoclonal antibody against the extracellular domain of the IGF-I receptor for coating and an Europium-labeled monoclonal antiphosphotyrosine antibody (PY20) as tracer. The assay is sensitive (detection limit < 0.08 µg/liter), specific (IGF-II cross-reactivity is 12%, whereas proinsulin, insulin, and insulin analogs have a cross-reactivity < 1%), and precise (mean within and in-between assay CVs were < 7 and 15%).
The bioassay was originally developed to measure IGF-I bioactivity in human serum, and therefore, we performed some studies to investigate whether the bioassay behaved similarly in mouse serum. Because the cells are not viable in undiluted serum, it is necessary to dilute serum at least 1:5 in KRB before bioassay. But it is well known that limited dilution of human serum in KRB (from 1:5 to 1:20) does not have any major effects on levels of free and bioactive IGF-I (27, 57). This observation is explained by the buffering effects of the IGFBPs, which dissociate IGF-I during dilution to maintain equilibrium (58). Therefore, we examined the effect of dilution of healthy mouse serum in KRB, which showed that bioactive IGF-I remained relatively stable when comparing pooled mouse serum diluted 1:5 and 1:10 before assay, whereas levels decreased more at a dilution of 1:20: bioactive IGF-I (µg/l): 4.30 ± 0.06 (dilution 1:5) vs. 3.91 ± 0.04 (dilution 1:10) vs. 2.29 ± 0.03 (dilution 1:20) (mean and SEM of duplicate determinations). Based on this observation, we diluted all further mouse samples 1:10 before bioassay.
The bioassay measures the ability of serum to stimulate the IGF-I receptor in vitro. This receptor accessible fraction of IGF-I is believed to be composed of the sum of truly free IGF-I plus IGF-I being dissociated from the IGFBPs during incubation; the latter fraction has been named "readily dissociable IGF-I" and has been suggested to be of clinical relevance (59). Accordingly, in humans levels of bioactive IGF-I are higher than those of free IGF-I in humans (27). Therefore, levels of ultrafiltered free IGF-I vs. bioactive IGF-I were also compared in the same pool of healthy mouse serum, and in accordance with expectations levels of bioactive IGF-I were about 3-fold higher than those of ultrafiltered free IGF-I (data not shown), hereby supporting the concept that bioactive mouse IGF-I is also composed of the sum of free IGF-I plus readily dissociable IGF-I.
Insulin Tolerance Test
Mice were fasted for 6 h (0900–1500) followed by ip injection of recombinant human insulin (0.5 U/kg body weight; Eli Lilly, Indianapolis, IN). Whole venous blood was obtained from the tail vein at the 0, 15, 30, 60, 90, and 120 min after insulin injection. Blood glucose levels were measured using a Glucotrend glucometer (Johnson & Johnson).
Protein Purification, AKT, and MAPK Immunoblotting
Tissues from skeleton muscle were collected from six wild-type and nine triple KO mice after overnight fast. Then all tissues were lysed in modified RIPA butter [1x PBS, 1% Nonidet P-40 (Sigma, St. Louis, MO), 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate]. A total of 10 mg/ml phenylmethylsulfonyl fluoride in isopropanol (10 µl/ml RIPA) was added at time of use. Total tissue lysates (40 µg per sample) were boiled in SDS-PAGE loading buffer, separated on 4–20% gradient gels (Invitrogen, Carlsbad, CA), and then transferred onto polyvinylidene fluoride membranes. The membrane was preblocked with blotto (5% dry milk in TBS) and incubated overnight at 4 C in antiserum against phosphor-MAPK (Cell Signaling, Danvers, MA), which visualized both phosphorated 42- and 44-kDa Erk bands. After washes in TBS and TBS plus 0.5% Tween-20, membranes were incubated in secondary antibody (horseradish peroxidase-conjugated antirabbit IgG; Promega Corp., Madison, WI) for 1 h and washed again in TBS and TBS plus 0.5% Tween 20. Blots were then visualized using Western lighting chemiluminescence kit (PerkinElmer Life Science, Inc.). Then the membranes were stripped and incubated overnight at 4 C in antiserum against MAPK to visualized nonphosphorated Erk bands using the same procedure above. Signals were scanned and band intensities were quantified by NIH software. The same procedures were performed to measure the stimulation of AKT.
Immunohistochemistry of Pancreata and Quantitation of ß-Cells
Pancreata from six wild-type mice and six IGFBP-3-4-5 null mice were examined. For analysis of adult pancreata, animals were killed by overdose of sodium amytal. All pancreata were removed, cleared of fat and spleen, weighed, and fixed overnight in Bouin’s solution. Then the pancreata were embedded in paraffin and consecutive sections (15 µm) were mounted on slides. After rehydration and permeabilization, one of four continuous slides was immunostained using mouse antiinsulin antibodies (Sigma), followed by detection using fluorescein antibodies (Jackson ImmunoResearch, West Grove, PA) to visualize the ß-cell. To assess the morphology of pancreas, the pancreas sections were stained by H/E. For quantitation of islet area, all sections were viewed using fluorescent microscope and the staining areas were digitally photographed at a magnification of x100. The entire sections were viewed by dissection microscope and were digitally photographed at a magnification of x2. Analyses of islet areas and entire section areas were performed using Matlab software (Mathworks Corp., Natick, MA). Average percentage of islet areas per pancreas was measured from the total count of 48 out of 96 sections (totally over 600 islets in all cross sections) for each mouse with six wild-type mice and six IGFBP-3-4-5 mutant mice analyzed.
Statistical Analysis
Data are expressed as means ± SE. For comparison of multiple groups, an ANOVA with repeated measures followed by Student’s t test was used. P value less than 0.05 shows significant difference, indicated by *, whereas P values less than 0.01 are indicated by**.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
First Published Online May 4, 2006
Abbreviations: ALS, Acid-labile subunit; ES, embryonic stem; H/E, hematoxylin and eosin; IGFBP, IGF binding protein; KO, knockout; KRB, Krebs Ringer buffer; TBS, Tris-buffered saline.
Received for publication May 17, 2005. Accepted for publication April 21, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-actin-IGFBP-4 fusion gene induces smooth muscle hypoplasia. Endocrinology 139:2605–2614This article has been cited by other articles:
 |
|
 |
|
  A. Mukherjee and P. Rotwein Insulin-Like Growth Factor-Binding Protein-5 Inhibits Osteoblast Differentiation and Skeletal Growth by Blocking Insulin-Like Growth Factor Actions Mol. Endocrinol., May 1, 2008; 22(5): 1238 - 1250. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Mukherjee, E. M. Wilson, and P. Rotwein Insulin-Like Growth Factor (IGF) Binding Protein-5 Blocks Skeletal Muscle Differentiation by Inhibiting IGF Actions Mol. Endocrinol., January 1, 2008; 22(1): 206 - 215. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Rehage, S. Mohan, J. E. Wergedal, B. Bonafede, K. Tran, D. Hou, D. Phang, A. Kumar, and X. Qin Transgenic Overexpression of Pregnancy-Associated Plasma Protein-A Increases the Somatic Growth and Skeletal Muscle Mass in Mice Endocrinology, December 1, 2007; 148(12): 6176 - 6185. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Lofqvist, J. Chen, K. M. Connor, A. C. H. Smith, C. M. Aderman, N. Liu, J. E. Pintar, T. Ludwig, A. Hellstrom, and L. E. H. Smith From the Cover: IGFBP3 suppresses retinopathy through suppression of oxygen-induced vessel loss and promotion of vascular regrowth PNAS, June 19, 2007; 104(25): 10589 - 10594. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
