Main

Cytokines regulate cell growth and differentiation by binding to cell surface cytokine receptors, activating signal transduction cascades, such as the JAK-STAT pathway. The SOCS are a recently discovered group of proteins that have the ability to regulate cytokine signaling. Eight members of this group, including SOCS 1–7 and CIS, have been identified (1). The SOCS transcripts are up-regulated in response to cytokine stimulation, and the corresponding SOCS proteins inhibit the cytokine signal transduction pathways, via a mechanism that involves blocking the JAK activity or STAT binding sites on cytokine receptors, creating a negative feedback loop of cytokine signaling (1). In vivo studies have revealed very diverse functions of the SOCS proteins. Hence, SOCS-1 has been demonstrated to be a regulator of lymphocyte differentiation and interferon γ signaling (2,3), whereas SOCS- 3 is believed to be essential for placental function (4). In vitro studies have shown that SOCS-2 inhibits signal transduction induced by several cytokines including IL-6, leukemia inhibitory factor (LIF), IGF-I, prolactin, and GH (1,5,6). However, only low doses of SOCS-2 inhibited GH signaling, whereas high doses were reported to, in fact, stimulate GH signaling (7). We have previously reported that mice with a targeted gene deletion of the SOCS-2 gene display increased postnatal body growth, resulting in increased long bone length and a proportionate augmentation of most visceral organs (8). These effects were associated with deregulated growth hormone/IGF-I signaling, suggesting that SOCS-2 may have an essential negative regulatory role in the GH/IGF-I signaling pathway. Both GH and IGF-I are involved in the regulation of adult bone mineral metabolism, including BMD (9). The aim of the present study was to determine the role of SOCS-2 in the regulation of BMD, and we demonstrate that SOCS-2 deficiency leads to reduced BMD in both the trabecular and the cortical bone compartment.

METHODS

Animals.

Male mice lacking SOCS-2 (SOCS-2−/−) and WT litter-mate mice were generated and genotyped as previously described (8). All mice were of a pure C57Bl/6 background, and all animal experimentation was performed in accordance with National Health and Medical Research Council of Australia guidelines (Ethics Committee approval number 2003.014). Animals were fed a diet containing 12.8 kJ/g, 0.012 mg calcium/g, 6.6% fat, 22.7% carbohydrates, and 21.7% protein.

DXA.

Measurements of aBMD of the total tibia and of a 0.3 × 0.3 mm large subregion, with a relatively high content of trabecular bone, located in the proximal metaphyseal region of tibia, were performed ex vivo with the Norland pDEXA Sabre (Norland, Fort Atkinson, WI) and the Sabre Research software (v3.6) as previously described (10). The interassay CV for these measurements was below 5%.

pQCT.

Computerized tomography (CT) was performed with the Stratec pQCT XCT Research M (Norland; v5.4B) operating at a resolution of 70 μm (11). Trabecular volumetric BMD was determined ex vivo, with a metaphyseal pQCT scan of the distal femur, and defined as the inner 45% of the total cross-sectional area. Cortical bone parameters were determined ex vivo with a mid-diaphyseal pQCT scan of the femur and tibia. The CV for these measurements was <2%.

Serum parameters.

Serum osteocalcin levels were measured using a MAb raised against human osteocalcin (Rat-MID osteocalcin ELISA, Osteometer Biotech, Herlev, Denmark). The sensitivity of the osteocalcin assay was 21.1 ng/mL and intra- and interassay CV were <10%. Serum IGF-I levels were measured by double-antibody IGF binding protein-blocked RIA (12).

Statistical analysis.

Results are presented as means ± sem. Differences in bone parameters between WT and SOCS-2 deficient mice were calculated using students t test. A p value < 0.05 was considered statistically significant.

RESULTS

Longitudinal bone growth.

The lengths of the two long bones tibia and femur were unchanged in 4-wk-old SOCS 2−/− compared with age-matched WT mice (Table 1). However, in adult mice, at 15 wk of age, SOCS-2−/− mice had both longer tibias (4.4%) and femora (5.6%) than the WT mice (Table 1).

Table 1 Lengths of femur and tibia in SOCS-2−/− and WT mice
Full size table

aBMD.

aBMD of excised tibia was analyzed by DXA. The aBMD was reduced in the total tibia in both 4-wk-old (−8.6%) and in 15-wk-old (−6.0%) SOCS 2−/− compared with WT mice (Fig. 1). These reductions were more pronounced when the aBMD was analyzed in a subregion, with a relatively high content of trabecular bone, located in the proximal metaphyseal region of the tibia. In this region, the aBMD was reduced by 11.4% in 4-wk-old and by 10.7% in 15-wk-old SOCS 2−/− compared with WT mice (Fig. 1).

Figure 1
figure 1

DXA measurements of aBMD in the total tibia (Total) and the proximal metaphyseal region (Metaphyseal; a region with a relatively high content of trabecular bone) of tibia in 4-wk-old (A) and 15-wk-old (B) WT and SOCS-2−/− mice. Values expressed as means ± sem. *p < 0.05, **p < 0.01 vs WT (n = 10 for 4-wk-old, n = 6 for 15-wk-old).

Full size image

Trabecular volumetric BMD.

The results obtained from DXA measurements are a combination of effects on trabecular bone and cortical bone parameters. To be able to distinguish between effects on trabecular and cortical bone, the bones were analyzed by pQCT. The trabecular volumetric BMD was measured in the metaphyseal region of the distal femur using pQCT. Both 4-wk-old and 15-wk-old SOCS 2−/− mice displayed lower trabecular volumetric BMD (−10% and −32%, respectively) than WT mice of the same age (Fig. 2).

Figure 2
figure 2

pQCT measurements of trabecular volumetric BMD (tvBMD) in the distal metaphyseal region of femur in 4-wk-old and 15-wk-old WT and SOCS-2−/− mice. Values expressed as means ± sem. *p < 0.05, **p < 0.01 vs WT (n = 10 for 4-wk-old, n = 6 for 15-wk-old).

Full size image

Cortical bone parameters.

The cortical bone mineral content (BMC) in the diaphyseal region of femur was reduced in 4-wk-old SOCS 2−/− mice compared with WT mice (Table 2). The reduction of cortical BMC was associated with a decrease in cortical volumetric BMD, cross-sectional area, and thickness (Table 2). The reduced cortical thickness was caused by a greater endosteal circumference whereas the periosteal circumference was unaffected in the 4-wk-old SOCS-2−/− mice. In adult mice, at 15 wk of age, a less severe cortical bone phenotype was seen with only the volumetric BMD significantly lower in the SOCS-2−/− mice than in the WT mice (Table 2).

Table 2 Cortical bone parameters measured by pQCT in the diaphyseal region of femur in SOCS-2−/− and WT mice
Full size table

Serum markers.

Serum levels of osteocalcin, a marker of bone formation/turnover, were significantly increased in 4-wk-old but not in 15-wk-old SOCS 2−/− mice compared with WT mice (Table 3). The levels of IGF-I were not significantly different between SOCS 2−/− and WT mice at any age (Table 3).

Table 3 Osteocalcin and IGF-I measured in serum from SOCS-2−/− and WT mice
Full size table

DISCUSSION

We have in the present study confirmed the finding that SOCS-2 is of importance for the regulation of skeletal growth. However, more importantly, we made the novel observations that i) SOCS-2 is a crucial regulator of BMD and ii) that this effect on BMD is not consistent with an expected augmented GH/IGF-I signaling in SOCS-2–deficient mice. GH is the most important regulator of postnatal body growth. A large part of its growth-promoting effect is mediated by GH-stimulated expression of IGF-I in the liver, but also in peripheral tissues (9,13,14). Mice with a targeted gene deletion of the SOCS-2 gene display increased postnatal body growth, resulting in increased long bone length and a proportionate augmentation of most visceral organs (8). These mice have several characteristics of deregulated GH and IGF-I signaling, such as collagen accumulation in the dermis, affected production of major urinary protein, and increased local production of IGF-I in several organs, including the heart, lungs, and spleen. These findings suggest that an important effect of SOCS-2 is to inhibit the GH/IGF-I signaling pathway, resulting in an increased growth in SOCS-2–inactivated mice (8). The high growth (hg) mouse strain displays a 30–50% increment in postnatal growth, without resulting in obesity, compared with WT mice, and has been found to carry a mutation in the SOCS-2 gene (15), further supporting the role of SOCS-2 in regulating growth.

In vitro data have revealed that SOCS-2 inhibits GH signaling at low doses, probably to a lesser extent than SOCS-1 and SOCS-3 (5). Furthermore, SOCS-2 inhibition of GH signaling has been shown effective only in low concentrations, whereas higher concentrations restore and even stimulate GH signaling, suggesting an additional role of SOCS-2 in restoring the sensitivity of inhibited GH signaling caused by other SOCS proteins (7,16). These intriguing in vitro data are supported by our recent in vivo findings, showing that both absence and over-expression of SOCS-2 cause growth enhancement in mice (8,16). The SOCS-2 over-expressing transgenic mice display enhanced growth (16), suggesting an enhanced GH signaling. Therefore, for the regulation of growth, it has been proposed that SOCS-2 has a dual effect on GH signaling. Both inactivation and over-expression of SOCS-2 result in enhanced GH signaling and growth whereas physiologic levels of SOCS-2 reduce GH signaling.

Both GH and IGF-I treatment result in an increased BMD (9,14, 17–19). Thus, the present finding of reduced BMD in SOCS-2–deficient mice is not consistent with an enhanced GH/IGF-I signaling as proposed above for the regulation of growth. One may speculate that SOCS-2 inactivation results in differential effects in the regulation of different physiologic processes, including enhanced GH signaling in the regulation of growth, while it does not affect or reduces GH/IGF-I signaling in the regulation of BMD. Alternatively, the ability of SOCS-2 to inhibit signal transduction induced by other cytokines including IL-6, LIF, and prolactin (1,20,21) might be of importance for the mechanism behind the reduced BMD in SOCS-2–deficient mice. For instance, BMD is affected in prolactin receptor–inactivated mice, and LIF has been reported to stimulate bone resorption (22,23). Gene deletion of IL-6 has been shown to prevent bone loss in mice after gonadectomy (24,25), and increased IL-6 has been found in a multitude of conditions involving increased bone resorption, such as rheumatoid arthritis and Paget's disease (26,27). Furthermore, IL-6 has been shown to directly induce bone resorption in mice calvaria (23). Hence, a reduced inhibition of IL-6 and LIF signaling in SOCS-2–deficient mice could account for increased bone resorption and resulting lower BMD in these animals. Thus, further studies are required to determine which cytokine signaling pathway is involved in the affected BMD in SOCS-2-deficient mice. The reduced BMD in SOCS 2−/− mice, seen in the present study, was demonstrated by extensive x-ray analyses, including both DXA and pQCT. DXA analysis demonstrated that the aBMD was reduced in the total tibia. Initial aBMD measurement could not distinguish whether it was an effect on the size of the bone, the trabecular volumetric BMD, the cortical volumetric BMD or a combination. Subregion analyses in the proximal metaphyseal region of the tibia, with a relatively high content of trabecular bone, indicated that the reduced aBMD, at least partly, was due to a reduced trabecular BMD in SOCS 2−/− mice. pQCT analyses were then performed to clearly distinguish between effects on trabecular and cortical bone, demonstrating that both the trabecular and cortical volumetric BMD was reduced in SOCS 2−/− mice. The cortical cross-sectional area and cortical thickness were reduced in 4-wk-old but not in 15-wk-old SOCS 2−/− mice, suggesting that the main effect on aBMD is a result of reduced trabecular and cortical volumetric BMD and to a lesser extent due to reduced size of the cortical bone.

In conclusion, SOCS-2 inactivation results in reduced trabecular and cortical volumetric BMD. These effects are not consistent with an augmented GH/IGF-I signaling, and, therefore, the mechanism behind the reduced BMD remains to be elucidated. SOCS-2 has been suggested as a possible target for the development of novel treatment strategies for children with growth disturbances (15), but the present study indicates that reduced BMD and an associated increased risk of osteoporosis is a potential side effect when the activity of SOCS-2 is affected.