Growth hormone protects against ovariectomy-induced bone loss in states of low circulating insulin-like growth factor (IGF-1)

Abstract

Early after estrogen loss in postmenopausal women and ovariectomy (OVX) of animals, accelerated endosteal bone resorption leads to marrow expansion of long bone shafts that reduce mechanical integrity. Both growth hormone (GH) and insulin-like growth factor (IGF-1) are potent regulators of bone remodeling processes. To investigate the role of the GH/IGF-1 axis with estrogen deficiency, we used the liver IGF-1-deficient (LID) mouse. Contrary to deficits in controls, OVX of LID mice resulted in maintenance of cortical bone mechanical integrity primarily owing to an enhanced periosteal expansion affect on cross-sectional structure (total area and cortical width). The serum balance in LID that favors GH over IGF-1 diminished the effects of ablated ovarian function on numbers of osteoclast precursors in the marrow and viability of osteocytes within the cortical matrix and led to less endosteal resorption in addition to greater periosteal bone formation. Interactions between estrogen and the GH/IGF-1 system as related to bone remodeling provide a pathway to minimize degeneration of bone tissue structure and osteoporotic fracture.

Fig. 1

The effects of OVX on body composition and serum levels of IGF-1, IGFBP-3, and GH in 12-week-old (basal) control and LID female mice. (A) BW measured from 12 to 28 weeks of age. (B) Body adiposity assessed by MRI. (C) Serum IGF-1 determined by RIA. (D) Serum levels of GH in sham-operated (S) mice and mice after OVX (O) determined by ELISA. (E) Serum IGFBP-3 levels determined by ELISA. (F) Uterus weight at 4 weeks after surgery, n = 8 mice per group, except sham-operated control mice. where n = 4. For longitudinal measurements (A–D), n = 8 to 47 mice per group. Data are represented as mean ± SEM at p < .05. a = LID versus sham-operated control or basal mice of same age group (significant genotype effects); b = versus sham-operated mice of same genotype and age group (significant OVX effects).

Fig. 2

Femoral midshaft structural and biomechanical properties determined by µCT and four-point bend test, respectively. (A) Total area at the femoral midshaft. Maximum load (B) and stiffness (C) sustained by the femoral midshaft. Data are represented as mean ± SEM. Group sizes were basal: control (n = 7), LID (n = 8); 16 weeks: sham-operated control (n = 4), OVX control (n = 7), sham-operated LID (n = 8), OVX LID (n = 8); 24 weeks: sham-operated control (n = 7), OVX control (n = 7), sham-operated LID (n = 7), OVX LID (n = 8); and 28 weeks: sham-operated control (n = 10), OVX control (n = 15), sham-operated LID (n = 9), OVX LID (n = 19). p < .05. a = LID versus sham-operated control or basal mice of same age group (significant genotype effects); b = versus sham-operated mice of the same genotype and age group (significant OVX effects).

Fig. 3

Histomorphometry at the tibial midshaft of control and LID mice at 4 and 8 weeks postoperatively. Endosteal (A) and periosteal (B) eroded surface and labeled surface expressed as perimeter fractions (Er.Pm/B.Pm and L.Pm/B.Pm). Sample sizes were n = 4 per group, except for 20-week-old (8 weeks after surgery) sham-operated groups, which were n = 3 per group. Data are represented as mean ± SEM at p < .05. a = versus control mice of same surgical and age group (significant genotype effects); b = versus sham-operated mice of the same genotype and age group (significant OVX effects).

Fig. 4

Apoptotic osteocyte density was increased after OVX and was lower in LID mice. Osteocyte apoptosis was assessed by immunohistochemistry using caspase-3 Ab on 5 µm paraffin sections of the femoral midshaft. Sample sizes were n = 4 per group, except for the 20-week-old (8 weeks after surgery) sham-operated groups, that were n = 3 per group. Data are represented as mean ± SEM at p < .05. a = sham-operated LID mice versus sham-operated control mice of same age group (significant genotype effects); b = versus sham-operated mice of same genotype and age group (significant OVX effects).

Fig. 5

OVX increased GH sensitivity in LID mice. (A) Femur length was measured by caliper (n = 4 to 19 per group). (B) Real-time PCR analysis of GHBP expression in liver (n = 4 per group). Data are represented as mean ± SEM at p < .05. a = versus control mice of same surgical and age group (significant genotype effects); b = versus sham-operated mice of the same genotype and age group (significant OVX effects).

Fig. 6

FACS analysis of cells derived from femoral marrow of control (C) and LID mice at the indicated ages after sham operation (S) or OVX (O) surgery at 12 weeks of age. B220+ cells (A) and CD11b+ cells (B). Data are represented as mean ± SEM of n > 6 mice per group at p < .05. a = versus control mice of same surgical and age group (significant genotype effects); b = versus sham-operated mice of the same genotype and age group (significant OVX effects).

Fig. 7

The effects of serum IGF-1 deficiency and estrogen deprivation on bone. Serum IGF-1 deficiency in LID mice results in an increase in serum GH levels. Additionally, reductions in serum IGF-1 lead to a decrease in osteoclast (OC) progenitors through as yet undefined mechanism(s). Estrogen deprivation in the state of IGF-1 deficiency leads to an increase in the sensitivity to GH, as reflected by increased linear growth and decreased production of GHBP by the liver. The increase in GH may protect osteocytes from OVX-induced apoptosis, leading to reduced recruitment of OC, inhibition of resorption at the endosteal surface, and increased formation at the periosteal surface.