Constitutive stimulatory G protein activity in limb mesenchyme impairs bone growth

Abstract

GNAS mutations leading to constitutively active stimulatory G protein alpha-subunit (Gsα) cause different tumors, fibrous dysplasia of bone, and McCune-Albright syndrome, which are typically not associated with short stature. Enhanced signaling of the parathyroid hormone/parathyroid hormone-related peptide receptor, which couples to multiple G proteins including Gsα, leads to short bones with delayed endochondral ossification. It has remained unknown whether constitutive Gsα activity also impairs bone growth. Here we generated mice expressing a constitutively active Gsα mutant (Gsα-R201H) conditionally upon Cre recombinase (cGsα^R201H mice). Gsα-R201H was expressed in cultured bone marrow stromal cells from cGsα^R201H mice upon adenoviral-Cre transduction. When crossed with mice in which Cre is expressed in a tamoxifen-regulatable fashion (CAGGCre-ER™), tamoxifen injection resulted in mosaic expression of the transgene in double mutant offspring. We then crossed the cGsα^R201H mice with Prx1-Cre mice, in which Cre is expressed in early limb-bud mesenchyme. The double mutant offspring displayed short limbs at birth, with narrow hypertrophic chondrocyte zones in growth plates and delayed formation of secondary ossification center. Consistent with enhanced Gsα signaling, bone marrow stromal cells from these mice demonstrated increased levels of c-fos mRNA. Our findings indicate that constitutive Gsα activity during limb development disrupts endochondral ossification and bone growth. Given that Gsα haploinsufficiency also leads to short bones, as in patients with Albright's hereditary osteodystrophy, these results suggest that a tight control of Gsα activity is essential for normal growth plate physiology.

Keywords: G protein; GNAS; Skeletal development; Stimulatory G protein; cAMP.

Figure 1. Generation and validation of transgenic mice conditionally expressing a constitutively active Gsα mutant

(A) Schematic diagram of the construct used for generation of cGsα^R201H transgenic mice. The Gsa-R201H cDNA, as well as the lacZ gene, is expressed upon Cre recombinase activity. (B) Detection of β-galactosidase activity in BMSCs upon adenovirus transduction. Four to six-week-old F1 or F2 generation mice from each of the five founder lines were used for validation. Results are shown for lines #6, #9, and #18 only. Ad-LacZ was used as positive control for the staining. NT, non-transduced. Ad-YFP was used as negative control. Note that lines #6 and #18, but not line #9, were able to express the transgene upon Cre recombinase activity.

Figure 2. qRT-PCR and Western blot analysis of Gsα-R201H expression in BMSCs

(A) qRT-PCR analysis of Gsα-R201H mRNA level in BMSCs upon adenovirus transduction. BMSCs were isolated from 4-6-week-old F1 or F2 generation of 3 founder lines. Primers were designed to detect the transgene-derived mutant Gsα transcript by taking advantage of the sequences of the HA tag. mRNA levels were normalized against the samples transduced with adenovirus-Gsα-HA. Data represent mean ± SEM of two-to-three independent experiments; **, p<0.01. (B) Western blot analysis of protein lysates from BMSCs upon adenovirus transduction. BMSCs were isolated from 4-6-week-old F1 or F2 generation cGsα^R201H line #6 or control littermates. The Gsα-R201H protein was detected using an antibody against the HA tag; note the immunoreactivity above the non-specific (NS) band in the sixth lane from left. Cells transduced with ad-Gsα-HA were used as positive control. Ad-YFP was used as negative control. NT, non-transduced. β-actin immunoreactivity was used to ensure comparable gel loading.

Figure 3. Tamoxifen induced-lacZ expression in offspring of crosses between cGsα^R201H and CAGGCre-ER™ mice

Cre-only and double mutant littermates from #6 and #18 lines were injected with Tamoxifen from postnatal day 14 to 18 for five consecutive days and liver and kidney were harvested at postnatal day 25. Following X-gal staining, the sections were counter-stained with Nuclear Fast Red. Liver sections were captured at 10X and kidney sections at 20X magnification. (A,B) Double mutants from line #6; (C,E,G) single mutants (CAGGCre) from line #18; (D,F,H) double mutants from line #18. Note the lacZ positive cells in liver (D,F) and kidney (H) of line #18 double mutants.

Figure 4. Abnormal limb development in neonatal Prx1-Cre;Tg mice

(A) Gross appearance of Prx1-Cre;Tg and littermate control at birth. (B) Whole-mount alizarin red/alcian blue staining of Prx1-Cre;Tg neonate and littermate control. Note the shortened long bones and the relatively normal length of the axial skeleton in Prx1-Cre;Tg. (C) Alizarin red/alcian blue staining of forelimbs and (D) hind limbs of Prx1-Cre;Tg neonate and littermate control. (E,F,G) Alizarin red/alcian blue staining of radius, ulna and forefeet (E), hind feet (F), and skull (G) of Prx1-Cre;Tg neonate and littermate control. Note the markedly shorter and thinner long bones and the shorter snout in Prx1-Cre;Tg mice. Images are representative of at least three different pups of the Prx1-Cre;Tg and control genotype from two independent litters.

Figure 5. Hypertrophic to columnar proliferating chondrocyte zone length is reduced in Prx1-Cre;Tg compared to control littermates

(A–D) Hematoxylin and eosin staining of control and Prx1-Cre;Tg growth plates at birth. Tibia (A,B) and distal femur (C,D) captured at 10× magnification. Lines indicate the columnar proliferative (cp) and hypertrophic (h) chondrocyte zones. (E) Ratio of the length of hypertrophic to columnar proliferating zones in control versus Prx1-Cre;Tg neonates. Data represent mean ± SEM (n=5-6); *, p<0.05 vs control (two-tailed Student’s t-test).

Figure 6. Prx1-Cre;Tg mice continues to be smaller than control littermates at postnatal day 21, with delayed endochondral ossification and increased c-fos mRNA levels in BMSCs

(A) Gross phenotypes of 3-week old control (on the left) and Prx1-Cre; Tg (on the right) littermates. (B–E) Hematoxylin and eosin stained femur sections of control (B,D) and Prx1-Cre;Tg mice (C,E). Images were captured at 2× (B,C) or 10× (D,E) magnification. Note the absence of secondary ossification center in Prx1-Cre;Tg samples (Panels C and E). (F) Elevated levels of c-fos mRNA in BMSCs from Prx1-Cre; Tg compared to control littermates, measured by using qRT-PCR and β-actin mRNA as a reference control. Data represent mean ± SEM (n=7); *, p<0.05 vs control (two-tailed Student’s t-test).