Increased serum IGF-1 levels protect the musculoskeletal system but are associated with elevated oxidative stress markers and increased mortality independent of tissue igf1 gene expression

Summary

Although the literature suggests a protective (anabolic) effect of insulin-like growth factor-1 (IGF-1) on the musculoskeletal system during growth and aging, there is evidence that reductions in IGF-1 signaling are advantageous for promoting an increase in lifespan through reduction in oxidative stress-induced tissue damage. To better understand this paradox, we utilized the hepatocyte-specific IGF-1 transgenic (HIT) mice, which exhibit 3-fold increases in serum IGF-1, with normal IGF-1 expression in other tissues; and mice with an IGF-1 null background that exclusively express IGF-1 in the liver, which thereby delivers IGF-1 by the endocrine route only (KO-HIT mice). We found that in the total absence of tissue igf1 gene expression (KO-HIT), increases in serum IGF-1 levels were associated with increased levels of lipid peroxidation products in serum, increased mortality rate at 18 months, of age in both genders. Surprisingly, however, we found that in female mice, tissue IGF-1 plays an important role in preserving trabecular bone architecture as KO-HIT mice show bone loss in the femoral distal metaphysis. Additionally, in male KO-HIT mice increases in serum IGF-1 levels were insufficient to protect against age-related muscle loss. Overall, we conclude that elevations in serum IGF-1 have beneficial role in the aging musculoskeletal system but may have deleterious effects that lead to earlier mortality.

Results and discussion

In mice mutations that decrease secretion of growth hormone (GH) (Prop1, Pit1) lead to significant extensions in life span. The long-lived GH deficient Ames dwarf mice have very low serum IGF-1 levels and increased activity of catalase and Cu/ZnSOD (Hauck & Bartke 2000). These mice show reduced levels of DNA and protein oxidation in liver (Yamamoto et al. 2005), and reduced levels of serum and liver F2-isoprostanes, which are stable lipid peroxidation products (Choksi et al. 2007). Similar resistance to oxidative stress is observed in GH resistant mice and in GH deficient Snell and lit/lit dwarfs (Bartke & Brown-Borg 2004; Salmon et al. 2005). Likewise, female mice heterozygous for the Igf1r gene had extended life span and are resistant to the ROS generator paraquat (Holzenberger et al. 2003). Furthermore, mice heterozygous for the brain Igf1r gene (bIgf1r+/−) had extended median life span (Kappeler et al. 2008) accompanied by severely reduced serum IGF-1 as a result of reduced GH production.

IGF-1 is one of the most powerful regulators of muscle homeostasis, and a significant regulator of bone mineral density (BMD). IGF-1R and IGF-1 null mice show severe retardation in skeletal muscle development and bone accrual (Liu et al. 1993). On the other hand, IGF-1 transgenic mice exhibit a clear increase in body and organ weights and a 30% increase in muscle and bone mass (Mathews et al. 1988). To begin to understand how autocrine/paracrine or endocrine IGF-1 actions affect the different organ/systems during aging we utilized the hepatocyte-specific IGF-1 transgenic (HIT) mice, which exhibit 3-fold increases in serum IGF-1, with normal IGF-1 expression in other tissues and mice with IGF-1 null background that exclusively expresses IGF-1 in the liver and thereby deliver IGF-1 by the endocrine route only (KO-HIT mice) (Figure 1A) (Elis et al. 2010a; Elis et al. 2010b; Wu et al. 2009). We found that increases in serum IGF-1 levels in male and female HIT mice, led to increased body weight throughout 18 months, while body weights of KO-HIT mice were indistinguishable from controls (Figure 1B,C). Whole body assessment of lean and fat mass (by MRI) at 18 months of age did not reveal any differences between the groups (data not shown). However, although not significant, we found that gonadal fat pads of HIT and KO-HIT mice were reduced in females (−43%) and in males (−23%) as compared to controls. Similarly, in KO-HIT male mice, we found a 20% decrease in the relative weights of the gastrocnemius muscle, as compared to controls (Table1). Together with previous observations, our study suggests that muscle-IGF-1 along with other local factors regulate muscle mass, and that increases in serum IGF-1 alone are insufficient for muscle maintenance during aging (at least in males). This was also evident in HIT mice, where despite significant elevations in serum IGF-1 levels, muscle mass was similar to controls (Table1). Nonetheless, future studies will be necessary to determine how these changes affect muscle function.

Figure 1. Physiological traits, oxidation and mortality in HIT and KO-HIT mice.

Generation and detailed characterization of the HIT and KO-HIT mice was previously described (Elis et al. 2010a; Elis et al. 2010b). (A) Serum levels of IGF-1 (n=10 mice per group) were determined by radio-immunoassay as described earlier (Elis et al. 2010a; Elis et al. 2010b). Body weight of female (B) and males (C) were followed from 16-78 weeks of age (n=8 to 10 mice per group). Tabecular bone loss in female (D) and males (E) calculated between 16-52 and 52-78 weeks of age. 8-iso PGF_2a levels in serum were measured by GC/MS and were normalized to sample volume (F) (n=5 mice per group) (Roberts & Morrow 2000). Data presented in panels A-F were tested by one-way analysis of variance (ANOVA) followed by Fisher’s test, considering P<0.05 as significant. (*) Denotes significant difference between control and KO-HIT mice, and (#) denotes significant difference between control and HIT mice.

(G) Compiled mortality and tumor incidence up to 18 months of age (n=22 to 78 mice per group per gender) analyzed by CHI² test. (H) Survival curves of the different groups up to 18 months of age of both genders. Log rank test revealed significant difference between all the curves (p=0.0021).

Table 1.

tissue weights of control, HIT and KO-HIT mice at time of sacrifice (18 months of age)

Gender	Genotype (N)	Body Weight, grams	Gonadal Fat, % of body weight (weight in grams)	Quadriceps % of body weight (weight in grams)	Gastrocnemius % of body weight (weight in grams)
Female	Control (10)	27.08+/−1.27	4.2+/−0.9 (1.23+/−0.31)	0.85+/−0.04 (0.22+/−0.009)	0.76+/−0.02 (0.20+/−0.006)
Female	HIT (8)	28.63+/−2.12	2.6+/−1.1 (0.83+/−0.39)	0.86+/−0.03 (0.24+/−0.012)	0.72+/−0.10 (0.20+/−0.016)
Female	KO-HIT (10)	20.10+/−0.95	2.4+/−0.7 (0.55+/−0.13)	0.85+/−0.04 (0.17+/−0.012)	0.79+/−0.04 (0.16+/−0.008)
Male	Control (10)	37.70+/−2.18	3.1+/−0.4 (1.21+/−0.17)	0.77+/−0.04 (0.28+/−0.003)	0.73+/−0.04 (0.27+/−0.008)
Male	HIT (8)	42.70+/−3.56	2.6+/−0.6 (1.23+/−0.40)	0.68+/−0.04 (0.28+/−0.009)	0.68+/−0.04 (0.28+/−0.011)
Male	KO-HIT (8)	35.70+/−2.40	2.4+/−0.5 (0.92+/−0.22)	0.68+/−0.04 (0.23+/−0.005)	0.62+/−0.04a (0.21+/−0.014)

^astatistically (p<0.05) different from control tested by ANOVA

The bone phenotype of HIT and KO-HIT mice was maintained throughout 18 months of age, such that HIT bones were larger and mechanically stiffer and stronger in both genders, and KO-HIT bones did not differ significantly from controls (Table 2). Unexpectedly, we found that in female HIT mice trabecular bone volume per total volume (%BV/TV) was higher than controls and trabecular thickness (Tb.Th) and tissue mineral density (TMD) increased significantly (Table 2). In contrast, in female KO-HIT mice %BV/TV was lower than controls and trabecular number (Tb.N) decreased significantly, suggesting that in females, trabecular bone IGF-1 is important for preserving normal %BV/TV and this cannot be compensated by elevations in serum IGF-1 (as seen in KO-HIT mice). Trabecular bone loss between 52 to 78 weeks of age did not show significant differences between the groups in both genders (Figure 1D,E). However, KO-HIT female mice tended to loose more bone than control or HIT mice, while in males increases in serum IGF-1 (in HIT and KO-HIT mice) seems to be protective.

Table 2.

Skeletal assessment of 18 months old mice with altered IGF-1 levels in tissues and serum. Femural bone morphology at the mid-diaphysis, and trabecular bone volume fraction and microarchitecture in the excised distal femoral metaphysis were assessed by micro-CT. Mechanical testing was performed on mouse femora from 18 months old control, HIT and KO-HIT mice. Bones were tested to failure by 4-point bending using a servohydraulic materials testing system (Instron Corp., Canton, MA, USA). (Elis et al. 2010a; Elis et al. 2010b)

Trait	Males			Females
	Control (n=10)	HIT (n=8)	KO-HIT (n=8)	Control (n=10)	HIT (n=10)	KO-HIT (n=10)
Bone Length	15.77+/−0.30	15.9+/−0.24	15.59+/−0.88	15.14+/−0.27	15.29+/−0.36	14.24+/−0.44*
Cortical bone (Femural mid-diaphysis)
Total cross- sectional area Tt.Ar. (mm2)	1.89+/−0.06	2.38+/−0.16 *	1.95+/−0.13	1.59+/−0.09	1.98+/−0.10 *	1.40+/−0.11
Cortical bone Ct.Ar. (mm2)	0.91+/−0.04	1.23+/−0.08 *	0.95+/−0.06	0.95+/−0.06	1.16+/−0.06 *	0.83+/−0.04
Relative cortical area (RCA; Ct.Ar/Tt.Ar)	0.60+/−0.02	0.59+/−0.01	0.61+/−0.02	0.48+/−0.01	0.52+/−0.01 (p=0.06)	0.49+/−0.02
Cortical thickness Ct.Th. (mm)	0.20+/−0.01	0.24+/−0.01 *	0.20+/−0.01	0.24+/−0.01	0.26+/−0.01	0.23+/−0.00
Medullary area Ma.Ar (mm2)	0.98+/−0.03	1.15+/−0.08	1.00+/−0.10	0.64+/−0.04	0.82+/−0.06 *	0.57+/−0.07
Polar moment of inertia Jo (mm4)	0.44+/−0.03	0.72+/−0.09 *	0.47+/−0.06	0.35+/−0.04	0.53+/−0.05 *	0.27+/−0.04
Tissue mineral density TMD (mg/cm3)	1388+/−15	1423+/−12	1406+/−16	1449+/−14	1439+/−11	1443+/−9
Trabecular bone (Femoral metaphysis)
Bone volume fraction BV/TV (%)	0.030+/−0.008	0.038+/−0.011	0.028+/−0.004	0.046+/−0.007	0.055+/−0.015	0.023+/−0.004 (p=0.09)
Trabecular thickness Tb.Th. (mm)	0.033+/−0.001	0.034+/−0.002	0.034+/−0.003	0.034+/−0.001	0.042+/−0.003 *	0.031+/−0.002
Trabecular number Tb.N. (1/mm)	0.91+/−0.24	1.06+/−0.26	0.86+/−0.10	1.34+/−0.20	1.20+/−0.27	0.73+/−0.11 *
Trabecular spacing Tb.Sp. (mm)	1.68+/−0.38	1.45+/−0.36	1.26+/−0.16	0.98+/−0.22	1.20+/−0.24	1.70+/−0.32
Tissue mineral density TMD (mg/cm3)	747+/−34	777+/−29	757+/−29	762+/−17	914+/−37 *	831+/−32
Mechanical properties (femora tested to failure by 4-point bending)
Stiffness (N/mm)	163.5+/−10.6	192.9+/−19.0	180.1+/−16.9	183.0+/−14.9	246.8+/−19.1 *	141.6+/−9.5
Max Load (N)	28.6+/−1.6	37.2+/−3.3 *	34.0+/−3.8	34.2+/−2.7	46.2+/−3.7 *	24.7+/−1.7

^*Different from control (p<0.05) by ANOVA, Statistics were run separately for males and females

In the past 10 years it has become evident that insulin/IGF-1 receptor activation leads to generation of ROS (Mahadev et al. 2001a; Mahadev et al. 2001b; Mahadev et al. 2004), however the exact link between insulin/IGF receptors and the anti-oxidant pathways has not been fully elucidated. We found that increased serum IGF-1 in HIT and KO-HIT mice was accompanied by increased isoprostane levels in serum of both genders and this was independent of tissue igf-1 gene expression (KO-HIT mice) (Figure 1F). It is believed that the duration and the magnitude of exposure to reactive oxygen species (ROS) play a critical role in the cellular defense response and ultimately in acceleration/delay of aging.

Cross-sectional pathology at 18 months of age revealed that 4-5% of female HIT and KO-HIT mice developed (hepatic or gastric tract) tumors, while control mice appeared healthy at that age (Figure 1F, i.e. 0.0% tumor incidence). Additionally, we found that 5% of control male mice developed tumors at 18 months of age, while 17% of HIT mice showed tumors. Although tumor incidence data did not differ significantly between the groups of each gender, these data suggest that excess serum IGF-1 tends to increase tumor incidence, irrespective of tissue IGF-1 levels. We also found that mortality (analyzed by CHI² test) increased significantly in both male and female HIT and KO-HIT mice as compared to controls (Figure 1G). Additionally, survival curves up to 78 weeks of age revealed significant decrease (tested by log rank) in life span of HIT and KO-HIT mice in both genders (Figure 1H). Nonetheless, it is important to note that we did not follow the animals longer than 78 weeks and further studies will be needed to determine the cause of death. Additionally we found an increased death rate (up to a year) amongst the KO-HIT male and female mice (Figure 1H) that stabilizes thereafter. This may imply to hormetic effects of increases in serum IGF-1 levels or simply suggest a stochastic process. However, in order to provide a conclusive explanation to this phenomenon, further studies are needed.

In summary, we found that constant elevations in serum IGF-1 levels (from birth) are associated with increased oxidative stress markers in tissues and serum as well as increased mortality, independent of local igf-1 gene expression. Concurrently, we found that constant elevations in serum IGF-1 were insufficient to protect the musculoskeletal system in the absence of tissue igf-1 gene expression in males. Similarly, elevations in serum IGF-1 were insufficient to protect from trabecular bone loss in the absence of tissue igf-1 gene expression in females.