Figure 1.
Krm2 expression in osteoblasts.
(A) RT-PCR expression analysis of Dkk, Krm and Rspo genes in primary osteoblasts (Obl. d5, non-mineralized, Obl. d25, mineralized) and various tissues of 6 weeks old mice. (B) RT-PCR expression analysis of Krm genes in non-differentiated MC3T3-E1 cells and tissues of newborn mice. (C) Immunohistochemistry on human bone sections reveals that KRM2 is present on osteoblasts lining the trabecular bone surface (arrows, scale bars, 100 µm). The bottom panel shows staining of osteoclasts (scale bars, 20 µm). (D) DNA transfection in MC3T3-E1 cells using expression plasmids for Wnt1, Wnt2 or Wnt3, Dkk1 and/or Krm2 at the indicated combinations. Bars represent mean ± SD of three independent experiments (n = 9). Asterisks indicate statistically significant changes.
Figure 2.
Generation of Col1a1-Krm2 transgenic mice.
(A) Schematic presentation of the construct used for osteoblast-specific over-expression of Krm2 (top) and identification of three transgenic founder animals (arrows) by Southern Blotting (bottom), one of them dying at the age of 10 weeks. (B) Von Kossa/van Gieson staining of non-decalcified vertebral body sections from 10 weeks old female wildtype mice (n = 3), the dead founder animal (#2) and age-matched female transgenic offspring from the two other founders (n = 3, scale bars, 1 mm). The transgene copy numbers, as well as the trabecular bone volume (BV/TV, bone volume per tissue volume) is given below. (C) Transgene-specific RT-PCR expression analysis revealing expression in bone and weak expression in the eye. (D) Northern Blotting comparing Krm2 expression in transgenic mice compared to wildtype littermates.
Figure 3.
Normal skeletal patterning and growth in Col1a1-Krm2 transgenic mice.
(A) Staining of skeletons from one day old wildtype and Col1a1-Krm2 transgenic littermates by alcian blue and alizarin red (scale bars, 5 mm). (B) Von Kossa/van Gieson staining of non-decalcified vertebral body sections from one day old wildtype and Col1a1-Krm2 transgenic littermates (n = 4, scale bars, 500 µm). The histomorphometric quantification of the trabecular bone volume is given below. (C) Contact Xrays of 2 weeks old female wildtype and Col1a1-Krm2 transgenic littermates (scale bars, 1 cm). (D) Determination of the femur and tibia length of female wildtype and Col1a1-Krm2 transgenic littermates at the indicated ages. Bars represent mean ± SD (n = 6).
Figure 4.
Decreased trabecular bone mass in Col1a1-Krm2 transgenic mice.
(A) Von Kossa/van Gieson staining of non-decalcified vertebral body sections from female wildtype and Col1a1-Krm2 transgenic littermates at the indicated ages (scale bars, 1 mm). (B) Determination of the lumbar spine length. (C) Histomorphometric quantification of the trabecular bone volume, trabecular number (Tb.N.) and trabecular thickness (Tb.Th.). All bars represent mean ± SD (n = 6). Asterisks indicate statistically significant differences.
Figure 5.
Decreased biomechanical stability of bones from Col1a1-Krm2 transgenic mice.
(A) µCT scanning of the vertebral bodies L6 from 24 weeks old female wildtype and Col1a1-Krm2 transgenic littermates (scale bars, 1 mm). (B) Microcompression testing revealed reduced biomechanical stability (Fmax, maximal force until bone failure). Bars represent mean ± SD (n = 6). Asterisks indicate statistically significant differences. (C) µCT scanning of the femora showing reduced cortical thickness and impaired mineralization (scale bars, 500 µm). (D) Cortical thickness (C.Th.) and bone mineral density (vBMD) are decreased in Col1a1-Krm2 transgenic mice compared to wildtype littermates. Bars represent mean ± SD (n = 6). Asterisks indicate statistically significant differences. (E) Three-point bending assays reveal reduced biomechanical competence of transgenic femora. Bars represent mean ± SD (n = 12). Asterisks indicate statistically significant differences. (F) Xray analysis of a 30 weeks old male Col1a1-Krm2 transgenic mouse with a spontaneous tibia fracture (arrow) compared to a non-transgenic littermate (scale bars, 2 mm).
Figure 6.
Impaired bone formation in Col1a1-Krm2 transgenic mice.
(A) Toluidine blue staining of non-decalcified vertebral body sections from 6 weeks old female wildtype and Col1a1-Krm2 transgenic littermates revealing that the normal appearance of cuboidal osteoblasts (arrows) covering trabecular bone surfaces is only observed in wildtype controls (scale bars, 50 µm). (B) Histomorphometric quantification showing that the number of osteoblasts (ObN/BPm, osteoblast number per bone perimeter) is not significantly decreased in sections from transgenic mice. (C) Fluorescent micrographs showing that overall calcein labeling is reduced in vertebral bodies of 6 weeks old female Col1a1-Krm2 transgenic mice (top, scale bars, 1 mm), as is the distance between the labeled surfaces at endosteal bone surfaces of the tibia (bottom, scale bars, 20 µm). (D) Histomorphometric quantification of the bone formation rate (BFR/BS, bone formation rate per bone surface) in female wildtype and Col1a1-Krm2 transgenic littermates. (E) Northern Blot expression analysis for Col1a1, Bglap and Ibsp using femur RNA of 6 weeks old female wildtype and transgenic mice. (F) Serum levels of osteocalcin and activities of alkaline phosphatase in 6 weeks old female wildtype and transgenic mice. All bars represent mean ± SD (n = 6). Asterisks indicate statistically significant differences.
Figure 7.
Cell-autonomous defect of Col1a1-Krm2 transgenic osteoblasts.
(A) BrdU incorporation assays revealed a higher proliferation rate in primary calvarial osteoblast cultures from transgenic mice after 2 days of differentiation. (B) Von Kossa staining performed at 10 days of differentiation reveals reduced mineralization of osteoblasts from Col1a1-Krm2 mice (scale bars, 1 cm), despite higher protein content (given below). Values represent mean ± SD (n = 3). Asterisks indicate statistically significant differences. (C) Western Blot analysis of canonical Wnt signaling using primary osteoblasts from wildtype and transgenic mice following stimulation with Wnt3a for 30 minutes. (D) Western Blot analysis (left) and ELISA (right) showing decreased Opg levels in cellular extracts and conditioned medium of osteoblasts from Col1a1-Krm2 mice. (E) Quantitative RT-PCR (left) and ELISA (right) demonstrating that the reduced expression of Tnfrsf11b in bones of 6 weeks old Col1a1-Krm2 mice results in decreased Opg serum levels. All bars represent mean ± SD (n = 4). Asterisks indicate statistically significant differences.
Figure 8.
Increased bone resorption in Col1a1-Krm2 transgenic mice.
(A) TRAP activity staining for osteoclasts (arrows) in decalcified vertebral body sections from 6 weeks old female wildtype and Col1a1-Krm2 transgenic mice (scale bars, 50 µm). (B) Histomorphometric quantification confirmed the increased number of osteoclasts (OcN/BPm, osteoclast number per bone perimeter) in transgenic mice. Bars represent mean ± SD (n = 6). Asterisks indicate statistically significant differences. (C) Xray analysis (top, scale bars, 1 mm) and µCT scanning (bottom, scale bars, 2 mm) demonstrating the presence of osteolytic lesions in 52 weeks old female Col1a1-Krm2 mice. (D) Goldner staining of the tibia showing osteoclasts at sites of cortical bone erosion (arrowheads), but also inappropriate bone formation in the marrow cavity of 52 weeks old female transgenic mice (arrows, scale bars, 100 µm). (E) Von Kossa/van Gieson staining reveals that the inappropriate bone formation in 52 weeks old female Col1a1-Krm2 mice is associated with an accumulation of non-mineralized osteoid (stained in red, scale bars, 100 µm).
Figure 9.
Differentially expressed genes in Col1a1-Krm2 transgenic osteoblasts.
(A) Affymetrix Gene Chip hybridization demonstrates that several well-established osteoblast differentiation markers are expressed at similar levels in osteoblasts from wildtype (mean values indicated by the dotted red line) and transgenic mice, while other genes are expressed at lower levels in the latter ones. Bars represent mean ± SD (n = 3). Asterisks indicate statistically significant differences between the relative signal intensities in wildtype and transgenic samples. (B) Affymetrix Gene Chip hybridization of wildtype osteoblasts following treatment with Dkk1 for 6 hours at day 10 of differentiation (n = 1). Shown are the Affymetrix signal intensities and the signal log ratios (SLR) for the 25 genes displaying the strongest negative regulation by Dkk1. The mean signal ratios of the Gene Chip comparison between wildtype and Col1a1-Krm2 transgenic osteoblasts are given on the right.
Figure 10.
Increased bone formation in Krm2-deficient mice.
(A) Von Kossa/van Gieson staining of non-decalcified vertebral body sections from 24 weeks old female wildtype (Krm2+/+) and Krm2-deficient (Krm2−/−) mice (scale bars, 1 mm). (B) Fluorescent micrographs showing a higher number of calcein-labelled surfaces in Krm2-deficient vertebral bodies (scale bars, 200 µm). (C) Histomorphometric quantification of the trabecular bone volume and osteoclast surface per bone surface (OcS/BS). (D) Histomorphometric quantification of the osteoblast surface per bone surface (ObS/BS) and the bone formation rate. (E) Von Kossa/van Gieson staining of non-decalcified tibia sections from 24 weeks old female wildtype and Krm2-deficient mice (scale bars, 1 mm). (F) Histomorphometric quantification of the trabecular bone volume and the trabecular number. All bars represent mean ± SD (n = 6). Asterisks indicate statistically significant differences.