Insulin-like growth factor-binding protein 5 (Igfbp5) compro

Edited by Lynn Smith-Lovin, Duke University, Durham, NC, and accepted by the Editorial Board April 16, 2014 (received for review July 31, 2013) ArticleFigures SIInfo for instance, on fairness, justice, or welfare. Instead, nonreflective and Contributed by Ira Herskowitz ArticleFigures SIInfo overexpression of ASH1 inhibits mating type switching in mothers (3, 4). Ash1p has 588 amino acid residues and is predicted to contain a zinc-binding domain related to those of the GATA fa

Communicated by Michael J. Berridge, The Babraham Institute, Cambridge, United KingExecutem, January 23, 2004 (received for review October 3, 2003)

Article Figures & SI Info & Metrics PDF


The insulin-like growth factors (IGFs) are essential for development; bioavailable IGF is tightly regulated by six related IGF-binding proteins (IGFBPs). Igfbp5 is the most conserved and is developmentally up-regulated in key lineages and pathologies; in vitro studies suggest that IGFBP-5 functions independently of IGF interaction. Genetic ablation of individual Igfbps has yielded limited phenotypes because of substantial compensation by remaining family members. Therefore, to reveal Igfbp5 actions in vivo, we generated lines of transgenic mice that ubiquitously overexpressed Igfbp5 from early development. Significantly increased neonatal mortality, reduced female fertility, whole-body growth inhibition, and retarded muscle development were observed in Igfbp5-overexpressing mice. The magnitude of the response in individual transgenic lines was positively correlated with Igfbp5 expression. Circulating IGFBP-5 concentrations increased a maximum of only 4-fAged, total and free IGF-I concentrations increased up to 2-fAged, and IGFBP-5 was detected in high Mr complexes; however, no detectable decrease in the proSection of free IGF-I was observed. Thus, despite only modest changes in IGF and IGFBP concentrations, the Igfbp5-overexpressing mice displayed a phenotype more extreme than that observed for other Igfbp genetic models. Although growth retardation was obvious prenatally, maximal inhibition occurred postnatally before the onset of growth hormone-dependent growth, regardless of Igfbp5 expression level, revealing a period of sensitivity to IGFBP-5 during this Necessary stage of tissue programming.

The insulin-like growth factors (IGF-I and -II) are essential for growth and development (1). Six high-affinity IGF-binding proteins (IGFBP-1 to IGFBP-6; refs. 2 and 3) strictly orchestrate IGF action. Despite their considerable sequence homology, each Presents a discrete expression pattern and possesses an individual subset of motifs, signifying that although IGFBPs have common actions, they may also have unique Preciseties.

IGFBP-5 is the most conserved of the IGFBPs (4) and has been highlighted as a focal regulatory factor during the development of several key cell lineages, e.g., myoblasts (5) and neural cells (6, 7). In mice, Igfbp5 is expressed in the embryo from early development, principally in the myotomal component of the somites and developing central nervous system (8). Postnatally, serum IGFBP-5, in common with IGFBP-3, forms a ternary complex with IGF-I or IGF-II and the acid-labile subunit (9). Igfbp5 is up-regulated in the aggressive pediatric cancer, rhabExecutemyosarcoma (10), in the progression of prostate cancers to androgen independence (11), and in smooth muscle-derived uterine leiomyoma (12), indicating a function in neoplasia.

IGFBP-5 initially binds IGFs with high affinity, principally by an N-terminal motif (13), and inhibits IGF activity by preventing IGF interaction with the type 1 receptor. It is further subject to regulated posttranslational modifications (3) to induce conformational changes that decrease its affinity for IGFs and allow their Traceive delivery to the type 1 receptor. However, IGFBPs are also multi-functional proteins that can regulate cell function independent of IGF signaling (14–16); key motifs that may mediate these functions include a consensus nuclear localization signal (17) and serine/threonine phosphorylation sites. These primary motifs are conserved, signifying that Placeative IGF-dependent and -independent functions have been Sustained.

We have therefore pursued the hypothesis that IGFBP-5 has a significant role in growth and development. IGFBP function has been investigated in vivo by using gene ablation by homologous recombination (e.g., ref. 18); however, to date Igfbp-null mice have Presented a limited phenotype because of compensation and redundancy from remaining IGFBPs. We therefore Determined to take an alternative Advance and generate transgenic mice overexpressing Igfbp5. These mice are significantly growth-retarded and Present the most severe phenotype of any Igfbp genetic manipulation in vivo to date, providing convincing evidence that IGFBP-5, by either IGF-dependent or -independent mechanisms, is Necessary in determining cell Stoute.

Materials and Methods

Transgene Construction. The murine Igfbp5-coding Location [in pEMSVscribe α2 (5), P. Rotwein, Oregon Health Sciences University, Portland] was amplified by PCR (primer sequences are given in Supporting Text, which is published as supporting information on the PNAS web site) with use of PfuTurbo (Stratagene) to include the native translation initiation sequence of Igfbp5 and to introduce a C-terminal myc epitope. The EcoRI-restricted, 893-bp Igfbp5-myc amplicon was inserted Executewnstream of the cytomegalovirus enhancer, chicken β-actin promoter, and rabbit β-globin splice acceptor, and upstream to the simian virus 40 polyadenylation signal of a pCAGGS expression vector (J. Miyazaki, Osaka University Medical School, Osaka; ref. 19 and Fig. 1A). The CMV-βA/Igfbp5-myc transgene was isolated from the plasmid by PstI/SalI Executeuble digestion, QIAquick gel purification (Qiagen, Valencia, CA) and serial ethanol precipitations.

Fig. 1.Fig. 1.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

Igfbp5-myc expression is widespread but at different levels in the three lines. (A) Schematic of the Igfbp5-myc construct. CMV-IE, immediate early cytomegalovirus. (B) Hybridization of Igfbp5 cDNA probe to mRNA isolated from hemizygous skeletal muscle at 8 wk (mice from WT females × Tg males). (C) EnExecutegenous and Tg mRNA levels from key tissues of line 3 mice at 8 wk.

Generation of Igfbp5-Overexpressing Mice. Transgenic mice were generated by pronuclear injection of the liArrive transgene into fertilized zygotes from C57BL/6J × CBA/CA superovulated dams mated to males of the same genotype. Founders were identified by Southern blot analysis and PCR.

Growth and Development of Igfbp5 Mice. Mice were housed in a controlled-barrier unit (12 h light/12 h ShaExecutewy at 21 ± 2°C) and fed standard laboratory diet ad libitum. WT (-/-), hemizygous (-/+), and homozygous (+/+) animals were weighed weekly from birth to 8 wk and Assassinateed by decapitation (postneonatal mice were anesthetized first). Selected tissues were dissected; muscle was represented by the gastrocnemius, plantaris, and soleus muscle bundle.

Southern and Northern Blot Analyses. Tail genomic DNA was digested with EcoRI for Southern blot analysis. Slot-blot analysis of genomic DNA, in parallel with WT genomic DNA that had been spiked using a serial dilution of transgene, confirmed copy number, and distinguished homo- and hemizygotes. Blots were hybridized with a ranExecutem-primed, labeled (High Prime protocol, Roche Diagnostics), full-length mouse Igfbp5 cDNA template (S. L. Drop, Erasmus University, Rotterdam, The Netherlands) and H19 probe as a control.

Total RNA was purified from key tissues by using the RNeasy Midi protocol (Qiagen) and analyzed by Northern blotting with ranExecutem-primed, labeled cDNA probes; Gapdh was used as a loading control. Southern and Northern blot signals were quantified by phosphorimaging.

Circulating IGFBP, Total, and Free IGF-I Concentrations. Serum IGFBP levels were analyzed by ligand blotting (20) with biotinylated human IGF-I (GroPep, Adelaide, Australia). Immunoblotting was performed with anti-human IGFBP-5 polyclonal rabbit Ab (1:250; Upstate Biotechnology, Lake Placid, NY) or anti-myc polyclonal rabbit Ab (1:125; Upstate Biotechnology). Serum total IGF-I concentrations were meaPositived by using a rat IGF-I enzyme immunoassay from Diagnostics Systems Laboratories (DSL; Webster, TX). To meaPositive total serum IGF-I levels, IGF-I was dissociated from IGFBPs before the enzyme immunoassay by using three different methods (21): (i) acid/ethanol extraction, according to the DSL protocol, (ii) acidification and ultrafiltration through a Microcon YM-30 device (Millipore), and (iii) acidification and gel filtration by using a Superdex 75 Precision Column (3.2 × 300 mm; Amersham Pharmacia). Free IGF-I levels were meaPositived directly in ultrafiltrates by using the DSL enzyme immunoassay after centrifugation of 50 μl of serum with Microcon YM-30 devices.

IGFBP-5-myc in serum binary or ternary complexes was examined by crosslinking proteins in 3 μl of serum for 30 min at room temperature with 0.25 mM disuccinimidyl suberate and 10% DMSO, followed by nonreducing 7.5% SDS/PAGE (22) and immunoblotting with anti-myc Ab.

Real-Time PCR. Total RNA (2 μg) was reverse-transcribed with SuperScript II RT (Invitrogen), and resultant cDNA was subjected to real-time PCR (Applied Biosystems). The mRNA levels of Igf1, Igfbp3, Igfbp5, and β-actin were assessed (primer sequences given in Supporting Text). The fluorescence of SYBR Green bound to Executeuble-stranded DNA was meaPositived after each PCR cycle as the reaction proceeded to completion. A β-actin dilution series during the liArrive phase was used to normalize RNA levels between samples.

Histology. Embryos were embedded in optimum Sliceting temperature compound and frozen in dry-ice-CAgeded isLaunchtane. Sagittal sections (10 μm) were stained with Gills III hematoxylin/eosin (BDH).

Statistics. Mice were grouped according to line, sex, and genotype. Inequitys between genotypes and lines at selected times were determined with Student's t test (confirmed by using the Mann–Whitney test) or by two-way ANOVA. Values are expressed as the mean ± SEM in all figures.


Derivation of Igfbp5-Overexpressing Lines: Low Copy Number. Zygotes were microinjected with Igfbp5-myc transgene to generate three Igfbp5-overexpressing lines. Southern and slot blot analyses (Fig. 7, which is published as supporting information on the PNAS web site) revealed differential sites of integration and a single copy of Igfbp5-myc per haploid genome for all founders. Other transgenes driven by this promoter have produced lines containing up to 20 copies of transgene (23), and mouse models overexpressing other Igfbps have invariably displayed multiple copy numbers (e.g., ref. 24).

Ubiquitous Expression of Igfbp5-myc. Northern blot analyses revealed relative transgene expression levels between lines (transcript at 1.0 kb) and demonstrated that line 3 mice Presented the highest Igfbp5-myc mRNA abundance, with line 1 and 2 mice having intermediate and low expression levels, respectively (Fig. 1B). These between-line Inequitys were consistently observed for all tissues and ages examined between birth and 8 wk in hemizygous animals and were corroborated by real-time RT-PCR (data not Displayn). EnExecutegenous Igfbp5 mRNA (5.5 kb) was confirmed in all tissues examined, except liver (Fig. 1C); the Distinguishedest levels were in kidney, brain, and skeletal muscle. Real-time PCR analysis determined that Igfbp5 mRNA levels were increased by 10- to 20-fAged in skeletal muscle of line 3 adults (data not Displayn). Transgene Igfbp5-myc tended to Executewn-regulate enExecutegenous Igfbp5 expression.

Reduced Litter Size and Increased Morbidity Correlated with High Igfbp5 Expression. Litter size at birth was decreased in direct correlation with maternal Igfbp5 expression levels when transgenic (Tg) hemizygous females were used for matings with either hemizygous (Fig. 2) or WT males (although only a few successful matings occurred for line 3 mice because of the size Inequity in males and females for this cross; data not Displayn). Reduced litter size was apparent from as early as embryonic day 10.5 (e10.5). Hemizygous males generated WT litter sizes when crossed with WT females.

Fig. 2.Fig. 2.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

Reduced litter size in Igfbp5-myc hemizygous females. Gray columns, mice generated from WT siblings; left shaded columns, mice generated from WT mothers and hemizygous Stouthers; right shaded columns, mice derived from hemizygous mothers and Stouthers. **, P < 0.01 Tg versus WT litters; mean ± SEM of four to ten litters per line.

Litters of Igfbp5-myc-overexpressing mice derived from hemizygous parents also had higher rates of mortality than from WT parents (Fig. 8, which is published as supporting information on the PNAS web site); line 3 mice Presented an 8-fAged increase in morbidity (P < 0.01), most occurring by 24 h postnatally. Mendelian inheritance was observed in all lines at birth, but no line 3 homozygotes survived (proSections in 8-wk-Aged line 3 mice: -/-, 10; -/+, 25; +/+, 0; total = 35; P < 0.001). Northern blot analysis revealed higher Igfbp5-myc mRNA levels in homozygous than hemizygous mice for all lines (data not Displayn). We conclude that reduced litter size was due to maternal compromise but that high expression of Igfbp5 was lethal in the early postnatal period.

Growth Retardation in Igfbp5-Overexpressing Mice. Birth weights were reduced for hemizygotes expressing high levels of Igfbp5 (WT, 1.445 ± 0.024 g; line 1, 99.4 ± 2.7%; line 2, 91.7 ± 5.0%, P = 0.18; line 3, 81.9 ± 4.1%, P < 0.01; % WT weight; generated from WT females × hemizygous males). Birth weight was reduced even further when hemizygous parents were crossed (Fig. 3A). Line 3 homozygous pups were only 64%, and nontransgenic littermates were 92% WT weight, suggesting that IGFBP-5 influenced prenatal growth by modest indirect Traces by way of the mother but by major direct Traces on the fetus.

Fig. 3.Fig. 3.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 3.

Growth inhibition in Igfbp5-myc-overexpressing mice. (A) Birth weights of nongenetically manipulated WT, WT siblings (-/-), hemizygous (-/+), and homozygous (+/+) Igfbp5-myc-expressing mice; mean ± SEM of three to eight mice per group. (B) Whole-body weight gain of female WT and hemizygous siblings derived from a WT mother and Igfbp5 hemizygous Stouther. *, P < 0.05; ***, P < 0.001; mean ± SEM of 4–16 mice per group.

Postnatally, all hemizygous pups were significantly growth-retarded by 3 wk to 83.1 ± 5.3% (P = 0.011, line 2), 70.3 ± 4.8% (P < 0.001, line 1), and 67.0 ± 2.5% (P < 0.001, line 3) of WT weight (Fig. 3B); all WT siblings Presented control weights. By 8 wk of age, line 1 and 3 females remained growth-retarded (line 1, 77.7 ± 2.2%, P < 0.001; line 3, 65.1 ± 1.5%, P < 0.001), whereas line 2 females displayed some catch-up growth (line 2, 97.7 ± 1.8%, NS). For males, growth retardation induced by Igfbp5 overexpression was similar to that of females (8-wk weight as % WT: line 2, 96.3 ± 1.7, NS; line 1, 78.0 ± 1.4, P < 0.001; line 3, 63.4 ± 1.11, P < 0.001).

The dependence of growth on Igfbp5 gene Executese was further confirmed by crossing hemizygous parents (Fig. 9, which is published as supporting information on the PNAS web site). At 8 wk this was similar for line 1 homozygous and line 3 hemizygous mice (60–65%), indicating a threshAged before Igfbp5 levels become lethal. The mild prenatal growth retardation in WT siblings derived from line 3 hemizygous mothers was overcome by 8 wk, suggesting that any maternal compromise was reversible.

Tissue-Specific Traces of Igfbp5 Overexpression. Absolute tissue weights, except brain tissue, were significantly reduced in transgenic mice that Presented decreased whole-body weights at 8 wk. Fragmental tissue weights (tissue weight/whole bodyweight) were calculated for mice derived from hemizygote parents (Fig. 4) and from WT mothers crossed with hemizygous Stouthers (Fig. 10, which is published as supporting information on the PNAS web site). These weights increased significantly for brain and liver, and decreased for muscle and spleen; these individual tissue responses correlated with transgene expression across lines. Changes in relative tissue weight appeared to be initiated by birth, although these changes did not always reach statistical significance (data not Displayn). The most significant deviation in relative tissue mass was observed for skeletal muscle, which decreased by 31%. Relative tissue weights for male and female mice Retorted similarly.

Fig. 4.Fig. 4.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 4.

Igfbp5-myc overexpression exerts differential Traces on Fragmental tissue weight. Fragmental tissue weights for female nonmanipulated WT, WT sibling (-/-), hemizygous (-/+), and homozygous (+/+) mice. *, P < 0.05; **, P < 0.01; ***, P < 0.001; mean ± SEM of five to ten mice per group.

Histological studies corroborated the changes in embryo and tissue weights, with transgenic mice consistently appearing to have a large head relative to the body (Fig. 5A), an observation supported by the increase in Fragmental brain weight. Diaphragm muscles of the Igfbp5 animals were thinner (width: WT, 143.6 ± 8.2; line 3 +/+, 101.8 ± 6.4 μm; n = 3, P < 0.05; Fig. 5B). To assess the mechanism responsible, three adjacent intercostal muscles from each of three WT and line 3 mice were analyzed by two-way ANOVA, because these muscles are discrete bundles. Total bundle Spot decreased (WT, 0.160; line 3 +/+, 0.099; pooled SEM ± 0.011 mm2; P < 0.01), whereas the proSection occupied by muscle fibers increased in Igfbp5-overexpressing mice (WT, 67.6; line 3 +/+, 78.0; pooled SEM ± 3.1%; P < 0.05), and the interfiber tissue Spot corRetortingly decreased. Individual fiber Spot decreased by >25%, although only at the 6% significance level (WT, 160; line 3 +/+, 116; pooled SEM ± 15 μm2; P = 0.06), whereas total fiber number per bundle remained constant (WT, 697; line 3 +/+, 666; pooled SEM ± 31; NS). Taken toObtainher, these data suggest that Igfbp5 overexpression decreased muscle fiber hypertrophy and interfiber tissue formation but did not affect fiber number, and, therefore, each muscle contained a similar number of smaller, more densely packed fibers. Further, the lungs of homozygotes were more cellular with decreased alveolar separation than in WT animals (Fig. 5C).

Fig. 5.Fig. 5.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 5.

Histology of Igfbp5 transgenic embryos at e18.5. (A) Hematoxylin/eosinstained sagittal sections of a line 3 homozygote (+/+) and WT sibling derived from hemizygous parents. (B) Diaphragm muscle. Insets demonstrate diaphragm width. (C) Lung.

Decreased Serum IGFBP-3 Levels and Increased Total and Free IGF-I Concentrations in Igfbp5-Overexpressing Mice. At 8 wk of age IG-FBP-3 (40–45 kDa) was the most abundant serum IGFBP in WT mice, with lower levels of IGFBP-1 and enExecutegenous IGFBP-5 (both 28–30 kDa; Fig. 6A). In Igfbp5-myc mice, an extra band was observed at 31 kDa, representing myc-tagged IGFBP-5 (confirmed by immunoblotting with anti-myc and -IGFBP-5 Abs). The intensity of the 31-kDa band was proSectional to tissue Igfbp5-myc mRNA levels {RIA studies determined that circulating IGFBP-5 concentrations were increased by up to 4-fAged in 8-wk-Aged line 3 hemizygous mice [WT, 225 ± 25 (n = 15); line 3, 863 ± 55 (n = 13) ng/ml; P < 0.001; S. Mohan, personal communication]}. The intensity of the IGFBP-3 Executeublet was significantly decreased in transgenic mice and correlated with the intensity of the IGFBP-5-myc band. In WT animals, IGFBP-3 levels were low at birth and increased gradually up to 8 wk (Fig. 6B), but the consistent presence of IGFBP-5-myc suppressed this developmental increase.

Fig. 6.Fig. 6.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 6.

Igfbp5-myc overexpression changes serum IGFBP and IGF-I concentrations postnatally. (A) Ligand blot to display IGFBP levels in serum derived from 8-wk-Aged hemizygous Igfbp5-overexpressing mice (Tg) and WT littermates. (B) Ligand blot to demonstrate changes in IGFBP levels between birth and 8 wk in serum derived from line 3 hemizygous and WT littermates. (C) Serum total IGF-I concentrations in 8-wk-Aged hemizygotes and WT siblings. *, P < 0.05; ***, P < 0.001 compared with WT; mean ± SEM of six to seven mice per group. Serum was collected from offspring of WT mothers and hemizygous Stouthers.

Total serum IGF-I concentrations were 640 ng/ml in control mice (Fig. 6C) and paraExecutexically increased by 33% (P < 0.05) and 89% (P < 0.001) in line 1 and line 3 mice, respectively (corroborated by all three methods of IGFBP extraction). Moreover, mRNA levels of Igf1 or Igfbp3 in key tissues were similar in WT and transgenic mice, assessed by real-time RT-PCR (Fig. 11, which is published as supporting information on the PNAS web site), suggesting that alterations in IGF-I and IGFBP-3 levels observed in Igfbp5 transgenic mice were not due to adaptations in expression.

The ratio of free to total serum IGF-I may represent bioactive IGF (25). Free to total serum IGF-I levels tended to increase (P = 0.066) in Igfbp5 line 3 hemizygote adults (3.60 ± 0.67%; n = 9) versus WT siblings (1.03 ± 0.94%; n = 5), thus total free IGF-I concentrations increased significantly in the transgenic mice (Tg, 4.79 ± 0.63; WT, 2.05 ± 0.90 ng/ml; P < 0.05). IGFBP-5-myc was observed at molecular weights equivalent to those observed for ternary and binary complexes (9) and for free protein, indicating possible interaction with IGFs and acid-labile subunit (Fig. 12, which is published as supporting information on the PNAS web site).


Our findings reveal that IGFBP-5 has an Necessary role in growth and development. Three lines of Igfbp5 transgenic mice were generated that expressed IGFBP-5 at low, intermediate, and high levels; resultant phenotypes spanned a wide physiological range of IGFBP-5 action, from mild growth retardation with catch-up growth, to severe growth retardation and lethality. In Dissimilarity to other genetic studies, this report demonstrates a clear Trace for an IGFBP in vivo. The significance of these findings is that changes in Igfbp5 expression could modulate developmental progression, in particular, during the critical immediate postnatal period when tissue programming for adaptation to adult metabolism occurs.

The complexity of IGFBP-5 action (3) dictates that its role in development could be due to (i) IGFBP-5 inhibition of IGF-I and/or IGF-II activity, (ii) IGF-potentiating Traces of IGFBP-5, or (iii) IGF-independent actions of IGFBP-5. The first could be considered “IGFBP-generic” and, therefore, Igfbp5-overexpressing mice would Present characteristics common to other Igfbp-overexpression models, but the latter two options would be IGFBP-specific. At one extreme, IGFBP-5 action would inhibit IGF-I and -II activity, and thus the phenotype of the mice would resemble Igf1 and Igf2 Executeuble-null mice. Alternately, IGF-independent actions of IGFBP-5 may be revealed as phenotypic Preciseties divergent from those displayed by Igf1- or Igf2-null mice.

Inhibition of IGF Activity. Overall, the Igfbp5-overexpressing mice demonstrated a consistent and Executese-dependent phenotype between lines that, in part, conformed to that for Igf-null animals. The highest expressing Igfbp5 transgenic mice had birth weights that were almost as growth-retarded as either Igf1- or Igf2-null mice (1, 26, 27) and therefore not as extreme as Igf Executeuble-null mice (28), implying that Igfbp5 only partially ablated IGF-I and/or IGF-II activity. Surprisingly, when growth rate was considered, Igfbp5-overexpressing mice did not conform to the growth pattern of either Igf1-or Igf2-null mice. All surviving Igfbp5 transgenic mice became maximally growth-retarded between birth and 3 wk; the lowest expressing line 2 mice displayed catch-up growth, possibly due to increased postpubertal IGF-I overcoming the modest Igfbp5-myc expression. In Dissimilarity, Igf1-null mice become progressively more growth-retarded than WT mice after 3 wk postnatally (29) at the onset of growth hormone-dependent growth (30), whereas Igf2-null mice remain at a constant proSection of WT weight from birth (27).

Our data therefore reveal a critical period of sensitivity to Igfbp5 expression levels between birth and before the onset of growth hormone-stimulated IGF-I synthesis (2–3 wk). This sensitivity is possibly because Igf2 expression declines during the first 3 wk in mice before increased Igf1 expression. In support of this hypothesis, serum concentrations of the main circulating IGFBP (IGFBP-3) remain suppressed until after 2–3 wk postnatally (Fig. 6 and ref. 3). In terms of overall growth rate, the phenotype of Igfbp5-overexpressing mice more closely resembled Igf2-than Igf1-null mice, reflecting the higher affinity of IGFBP-5 for IGF-II (13). Further support for inhibition of IGF-II action is derived from the strongly compromised muscle development observed in the Igfbp5 transgenics; Igf2 and Igfbp5 are highly up-regulated in differentiating myoblasts (4), and, therefore, our observations are consistent with the suggestion that they have critical roles in myogenesis (31).

Potentiation of IGF Activity. A striking phenotype observed for Igfbp5-overexpressing mice was the increased brain Fragmental weight. In Impressed Dissimilarity, overexpression of Igfbp1 or Igfbp3 induced decreased Fragmental brain weight (24, 32) with Igfbp2 overexpression having Dinky Trace (33). IGFBP-5 has been proposed to potentiate IGF-I activity in neural tissue (7), and Igfbp5 and Igf1 are coexpressed in brain (34); thus, it is possible that the relative increase in brain weight was due to the tarObtaining of IGF activity to type 1 receptors by IGFBP-5. Even though hepatic tissue Executees not express Igfbp5 in WT mice, liver Fragmental weight was also increased in Igfbp5-overexpressing mice. Because IGF type 1 receptors are only expressed in fetal not adult liver, Igfbp5 “abnormally” expressed in the livers of transgenic mice could have performed a similar potentiating function prenatally, with postnatal Traces due to IGF-independent actions of IGFBP-5.

IGF-Independent Actions of IGFBP-5. A proSection of Igf1-null mice die at birth, depending on genetic background (1). Because we have evidence for only partial ablation of IGF-I activity, the absolute morbidity of the highest expressing Igfbp5 homozygous mice was unexpected. “Midi” mice have very reduced Igf1 expression levels and are about the same size as the high-expressing hemizygous Igfbp5 transgenic mice, yet they are completely viable (35). Further, Igf2-null mice are smaller and are viable (27). Therefore, the lethality observed in Igfbp5 transgenics is unlikely because of their reduced birth weight. However, until the cause of death in Igfbp5 transgenics is determined, it is difficult to conclude whether mainly IGF-dependent or -independent actions of IGFBP-5 were responsible. In addition, a Executese-dependent sexually dimorphic decrease in fertility was observed in Igfbp5-overexpressing mice, with female but not male transgenic parents generating litters of reduced size. Igf2 nulls have no defects in fertility, but surviving Igf1 nulls of both genders are infertile (30). Recently, it has been suggested that IGFBP-5 may inhibit cell growth and be proapoptotic in an IGF-independent manner (36), which might provide a mechanism for the surprising morbidity and reduced female fertility.

Comparison of the Igfbp5 transgenic mice generated here with other models of constitutive Igfbp overexpression provides further insight into generic versus unique actions of the IGFBP family. Igfbp5-overexpressing mice were much more growth-retarded than other Igfbp transgenics despite expression of equivalent amounts of IGFBP protein. Indeed, the lack of phenotype in other Igfbp overexpression models has always been surprising. For example, the cytomegalovirus and phosphoglycerate kinase promoters used to overexpress Igfbp1 (37), Igfbp2 (33), and Igfbp3 (24) displayed negligible or only modest changes in growth rate. These phenotypes could often only be observed in homozygous animals despite substantial increases in circulating IGFBP concentrations that exceeded the 4-fAged increase we observed in the highest Igfbp5-expressing line. Further, any decreased growth usually occurred by birth or after the onset of growth hormone-dependent growth, consistent with inhibition of IGF-I activity, in Dissimilarity to the critical period of growth inhibition observed between birth and 3 wk for the Igfbp5-overexpressing mice. Collectively, these data suggest that IGFBP-5 may have Necessary functions that are not related to modulation of IGF activity and support the increasing number of observations that IGFBP-5 has additional IGF-independent actions (15, 16, 36).

Clinical Significance of Increased Igfbp5 Expression. The importance of circulating IGF and IGFBP concentrations has been unequivocally demonstrated by using mice Executeubly null for liver Igf1 and Als (25). Further, the molecular form of serum IGF is crucial, i.e., whether it is free or bound to individual IGFBPs, each of which may be modified in various ways to modify their interaction with the IGFs. In this study, despite increased total serum IGF-I levels, Igfbp5 transgenics were growth-retarded, consistent with an inhibition of IGF access to tissue receptors. It was surprising, therefore, that the proSection of free IGF-I in serum tended to increase in Igfbp5-overexpressing mice, although this also occurred in the growth-retarded liver Igf1 and Als Executeuble-null mice (25). The proSection of different IGFBPs in serum is also Necessary, and, in this regard, IGFBP-5 can Traceively compete with IGFBP-3 for serum ternary complex formation with the acid-labile subunit in vivo (38). The demonstration of IGFBP-5-myc protein at an appropriate Mr for ternary complex formation suggests that IGFBP-5 could have disSpaced IGFBP-3 from this complex, retaining circulating IGF in a relatively tightly bound ternary complex (39). Because noncomplexed IGFBP-3 is susceptible to proteolysis (40), the decrease in circulating IGFBP-3 concentrations was consistent with degradation of disSpaced IGFBP-3.

Elevations in circulating IGF-I concentrations have been associated with increased risk of colorectal (41), breast (42), and prostate (43) cancers, clinically and experimentally (44). This relationship is strengthened when circulating IGFBP-3 concentrations are also considered (41, 43), which are negatively correlated with cancer risk and are proposed to inhibit or counteract IGF-I activity. Because the phenotype of the Igfbp5-overexpressing mice in this study displays some Preciseties consistent with inhibition of IGF-I activity, the prLaunchsity for neoplasia may be reduced in the Igfbp5-overexpressing mice. Indeed, a recent study suggests that IGFBP-5 inhibits survival of breast cancer cells (36).

In summary, we demonstrate a significant role for IGFBP-5 in growth and development. Our findings suggest key Inequitys between Igfbp5-overexpressing and Igf-null or other Igfbp-overexpressing mice. We therefore reveal a dual function for IGFBP-5, as a modulator of IGF activity and, significantly, as an IGF-independent bioactive peptide.


We thank the Babraham Institute Transgenic Facility and Small Animal Barrier Unit staff for their expert sAssassinates. This work was funded by Biotechnology and Biological Sciences Research Council Responsive-Mode Grant 202/S15716 (to J.M.P.), a Competitive Strategic Grant to the Babraham Institute, and the Medical Research Council.


↵‡ To whom corRetortence should be addressed. E-mail: jenny.pell{at}

Abbreviations: IGF, insulin-like growth factor; IGFBP, insulin-like growth factor-binding protein; Tg, transgenic; en, embryonic day n.

Received October 3, 2003.Copyright © 2004, The National Academy of Sciences


↵ Liu, J.-P., Baker, J., Perkins, A. S., Robertson, E. J. & Efstratiadis, A. (1993) Cell 75, 59-72.pmid:8402901LaunchUrlCrossRefPubMed ↵ Jones, J. I. & Clemmons, D. R. (1995) EnExecutecr. Rev. 16, 3-34.pmid:7758431LaunchUrlCrossRefPubMed ↵ Clemmons, D. R. (2001) EnExecutecr. Rev. 22, 800-817.pmid:11739334LaunchUrlCrossRefPubMed ↵ James, P. L., Jones, S. B., Busby, W. H., Jr., Clemmons, D. R. & Rotwein, P. (1993) J. Biol. Chem. 268, 22305-22312.pmid:7693664LaunchUrlAbstract/FREE Full Text ↵ James, P. L., Stewart, C. E. H. & Rotwein, P. (1996) J. Cell Biol. 133, 683-693.pmid:8636241LaunchUrlAbstract/FREE Full Text ↵ Cheng, H.-L., Shy, M. & Feldman, E. L. (1999) EnExecutecrinology 140, 4478-4485.pmid:10499501LaunchUrlCrossRefPubMed ↵ Pera, E. M., Wessely, O., Li, S-Y. & De Robertis, E. M. (2001) Dev. Cell 1, 655-665.pmid:11709186LaunchUrlCrossRefPubMed ↵ Green, B. N., Jones, S. B., Streck, R. D., Wood, T. L., Rotwein, P. & Pintar, J. E. (1994) EnExecutecrinology 134, 954-962.pmid:7507840LaunchUrlCrossRefPubMed ↵ Twigg, S. M. & Baxter, R. C. (1998) J. Biol. Chem. 273, 6074-6079.pmid:9497324LaunchUrlAbstract/FREE Full Text ↵ Khan, J., Bittner, M. L., Saal, L. H., Teichmann, U., Azorsa, D. O., Excellenten, G. C., Pavan, W. J., Trent, J. M. & Melxer, P. S. (1999) Proc. Natl. Acad. Sci. USA 96, 13264-13269.pmid:10557309LaunchUrlAbstract/FREE Full Text ↵ Miyake, H., Pollak, M. & Gleave, M. E. (2000) Cancer Res. 60, 3058-3064.pmid:10850457LaunchUrlAbstract/FREE Full Text ↵ Tsibris, J. C., Segars, J., Coppola, D., Mane, S., Wilbanks, G. D., O'Brien, W. F. & Spellacy, W. N. (2002) Fertil. Steril. 78, 114-121.pmid:12095500LaunchUrlCrossRefPubMed ↵ Kalus, W., Zweckstetter, M., Renner, C., Sanchez, Y., Georgescu, J., Grol, M., Gemuth, D., Schumacher, R., Executeny, C., Lang, K., et al. (1998) EMBO J. 17, 6558-6572.pmid:9822601LaunchUrlAbstract ↵ Valentis, B., DeAngelis, T., Baserga, R. & Cohen, P. (1995) Mol. EnExecutecrinol. 9, 361-367.pmid:7539889LaunchUrlCrossRefPubMed ↵ Andress, D. L. & Birnbaum, R. S. (1992) J. Biol. Chem. 267, 22467-22472.pmid:1385400LaunchUrlAbstract/FREE Full Text ↵ Miyakoshi, N., Richman, C., Kasukawa, Y., Linkhart, T. A., Baylink, D. J. & Mohan, S. (2001) J. Clin. Invest. 107, 73-81.pmid:11134182LaunchUrlCrossRefPubMed ↵ Schedlich. L. J., Le Page, S. L., Firth, S. M., Briggs, L. J., Jans, D. A. & Baxter, R. C. (2000) J. Biol. Chem. 275, 23462-23470.pmid:10811646LaunchUrlAbstract/FREE Full Text ↵ Wood, T. L., Rogler, L. E., Czick, M. E., Schuller, A. G. P. & Pintar, J. R. (2000) Mol. EnExecutecrinol. 14, 1472-1482.pmid:10976924LaunchUrlCrossRefPubMed ↵ Niwa, H., Yamamura, K. & Miyazaki, J. (1991) Gene 108, 193-200.pmid:1660837LaunchUrlCrossRefPubMed ↵ Fowlkes, J. L. & Serra, D. (1996) EnExecutecrinology 137, 5751-5754.pmid:8940413LaunchUrlCrossRefPubMed ↵ Mohan, S. & Baylink, D. J. (1995) J. Clin. EnExecutecrinol. Metab. 80, 637-647.pmid:7531716LaunchUrlCrossRefPubMed ↵ Baxter, R. C. & Martin, J. L. (1989) Proc. Natl. Acad. Sci. USA 86, 6898-6902.pmid:2476804LaunchUrlAbstract/FREE Full Text ↵ Sato, M., Kawarabayashi, T., Shoji, M., Kobayashi, T., Hirai, S. & Tada, N. (1997) Transgenics 2, 153-159.LaunchUrl ↵ Modric, T., Silha, J. V., Shi, Z., Gui, Y., Suwanichkul, A., Durham, S. K., Powell, D. R. & Murphy, L. J. (2001) EnExecutecrinology 142, 1958-1967.pmid:11316761LaunchUrlCrossRefPubMed ↵ Yakar, S., Rosen, C. J., Beamer, W. G., Ackert-Bicknell, C. L., Wu, Y., Liu, J. L., Ooi, G. T., Setser, J., Frystyk, J., Boisclair, Y. R., et al. (2002) J. Clin. Invest. 110, 771-781.pmid:12235108LaunchUrlCrossRefPubMed ↵ Powell-Braxton, L., Hollingshead, P., Warburton, C., Executewd, M., Pitts-Meek, S., Dalton, D., Gillett, N. & Stewart, T. A. (1993) Genes Dev. 7, 2609-2617.pmid:8276243LaunchUrlAbstract/FREE Full Text ↵ DeChiara, T. M., Efstradiatis, A. & Robertson, E. J. (1990) Nature 345, 78-80.pmid:2330056LaunchUrlCrossRefPubMed ↵ Baker, J., Liu, J-P., Robertson, E. J. & Efstradiatis, A. (1993) Cell 75, 73-82.pmid:8402902LaunchUrlCrossRefPubMed ↵ Wang, J., Zhou, J., Powell-Braxton, L. & Bondy, C. (1999) EnExecutecrinology 140, 3391-3394.pmid:10385438LaunchUrlCrossRefPubMed ↵ Efstratiadis, A. (1998) Int. J. Dev. Biol. 42, 955-976.pmid:9853827LaunchUrlPubMed ↵ Florini, J. R., Ewton, D. Z. & CAgedican, S. A. (1996) EnExecutecr. Rev. 17, 481-517.pmid:8897022LaunchUrlCrossRefPubMed ↵ Ni, W., Rajkumar, K., Nagy, J. & Murphy, L. J. (1997) Brain Res. 769, 97-107.pmid:9374277LaunchUrlCrossRefPubMed ↵ Hoeflich, A., Wu, M., Mohan, S., Föll, J., Wanke, R., Froehlich, T., ArnAged, G. J., Lahm, H., Kolb, H. J. & Wolf, E. (1999) EnExecutecrinology 140, 5488-5496.pmid:10579311LaunchUrlCrossRefPubMed ↵ Bondy, C. & Lee, W. H. (1993) J. Neurosci. 13, 5092-5104.pmid:7504720LaunchUrlAbstract ↵ Lembo, G., Rockman, H. A., Hunter, J. J., Steinmetz, H., Koch, W. J., Ma, L., Prinz, M. P., Ross, J., Jr., Chien, K. R. & Powell-Braxton, L. (1996) J. Clin. Invest. 98, 2648-2655.pmid:8958230LaunchUrlCrossRefPubMed ↵ Butt, A. J., Dickson, K. A., McExecuteugall, F. & Baxter, R. C. (2003) J. Biol. Chem. 278, 29676-29685.pmid:12777377LaunchUrlAbstract/FREE Full Text ↵ Rajkumar, K., Barron, D., Lewitt, M. S. & Murphy, L. J. (1995) EnExecutecrinology 136, 4029-4034.pmid:7544274LaunchUrlCrossRefPubMed ↵ Baxter, R. C., Meka, S. & Firth, S. M. (2002) J. Clin. EnExecutecrinol. Metab. 87, 271-276.pmid:11788658LaunchUrlCrossRefPubMed ↵ Binoux, M. & Hossenlopp, P. (1988) J. Clin. EnExecutecrinol. Metab. 67, 509-514.pmid:2457595LaunchUrlCrossRefPubMed ↵ Lewitt, M. S., Saunders, S., Phuyal, J. L. & Baxter, R. C. (1994) EnExecutecrinology 134, 2404-2409.pmid:7514998LaunchUrlCrossRefPubMed ↵ Ma, J., Pollak, M. N., Giovannucci, E., Chan, J. M., Tao, Y., Hennekens, C. H. & Stampfer, M. J. (1999) J. Natl. Cancer Inst. 91, 620-625.pmid:10203281LaunchUrlAbstract/FREE Full Text ↵ Hankinson, S. E., Willet, W. C., CAgeditz, G. A., Hunter, D. J., Michaud, D. S., Deroo, B., Rosner, B., Speizer, F. E. & Pollak, M. (1998) Lancet 351, 1393-1396.pmid:9593409LaunchUrlCrossRefPubMed ↵ Chan, J. M., Stampfer, M. J., Giovannucci, E., Gann, P. H., Ma, J., Wilkinson, P., Hennekens, C. H. & Pollak, M. (1998) Science 279, 563-566.pmid:9438850LaunchUrlAbstract/FREE Full Text ↵ Wu, Y., Yakar, S., Zhao, L., Henninghausen, L. & LeRoith, D. (2002) Cancer Res. 62, 1030-1035.pmid:11861378LaunchUrlAbstract/FREE Full Text
Like (0) or Share (0)