Ablation of the PTHrP gene or the PTH/PTHrP receptor gene leads to distinct abnormalities in bone development (original) (raw)
Early bone development in PTH/PTHrP-R–/– and PTHrP–/– mice. In previous reports, we characterized the phenotypic abnormalities in the tibial growth plates of PTHrP–/– and PTH/PTHrP-R–/– mice late in gestation (E18.5). In both cases, these include the shortness of proliferative columns of chondrocytes, the diminished synchrony of chondrocyte differentiation across the growth plate (9, 20, 27, 28), and glycogen accumulation in epiphyseal chondrocytes (19). To understand better how these abnormalities develop, we examined bones at earlier gestational time points. Figure 1 compares phalanges of PTH/PTHrP-R–/– and wild-type mice at E15.5. In the normal phalanx, chondrocytes in the center of the bone have hypertrophied, and synthesize type X collagen mRNA. In contrast, only the earliest hint of differentiation has occurred in the center of the phalanx of the PTH/PTHrP-R–/– mouse, and no type X collagen mRNA can yet be detected (Figure 1). This delay in chondrocyte differentiation is accompanied by a partial loss in polarity of the axis of chondrocyte differentiation across the length of the bone; this loss in polarity is demonstrated best at a somewhat later stage of differentiation, at E16.5 (Figure 2). As in Figure 1, the hypertrophic chondrocytes in the PTH/PTHrP-R–/– phalanx are small, and now produce small amounts of type X collagen mRNA. Strikingly, whereas the hypertrophic chondrocytes are concentrated in the center of the wild-type phalanx, the moderately hypertrophic chondrocytes in the PTH/PTHrP-R–/– phalanx extend from the center to encompass most of the bone. In the same bone from a PTHrP–/– mouse, the same partial loss of polarity of differentiation is observed. In contrast to the PTH/PTHrP-R–/– bone, however, there is no obvious delay in initial chondrocyte hypertrophy or reduction of type X collagen mRNA in the PTHrP–/– bone.
Chondrocyte differentiation at E15.5. Hematoxylin/eosin staining (a and b) and type X collagen mRNA in situ hybridization (c–f) of wild-type and PTH/PTHrP-R–/– phalanges. PTH/PTHrP receptor–ablated bones show a delay in chondrocyte differentiation and type X collagen expression (b, d, and f) when compared with wild-type bones (a, c, and e). Bright-field, c and d; dark-field, e and f.
Chondrocyte differentiation at E16.5. Hematoxylin/eosin staining (a–c) and type X collagen mRNA in situ hybridization (d–f) of wild-type, PTH/PTHrP-R–/–, and PTHrP–/– phalanges at E16.5. In contrast to PTHrP–/– bones, which exhibit a spatial and temporal progression in chondrocyte differentiation (c and f), PTH/PTHrP-R–/– bones show a delay in chondrocyte differentiation, as shown by the decrease in type X collagen expression (b and e), when compared with wild-type bones (a and d). However, the spatial distribution of the abnormally differentiated hypertrophic cells is identical (b, c and e, f).
Chondrocyte hypertrophy is normally followed by blood vessel invasion and deposition of bone by osteoblasts. Figure 3a illustrates this process in a normal (wild-type) phalanx at E18.5. In contrast, the hypertrophic chondrocytes in the center of the PTH/PTHrP-R–/– phalanx have been incompletely replaced by blood vessels and bone (Figure 3b). Figure 3, d and e compare tibiae at a roughly analogous developmental time (E16.5). Again, the center of the normal tibia demonstrates vascular invasion and bone deposition, while the PTH/PTHrP-R–/– tibia shows no vascular invasion or bone deposition. Figure 3, c and f also show that vascular invasion and bone deposition are not delayed in the PTHrP–/– bones; in fact, these processes appear more advanced than in normal bones. Analogous studies in the xiphoid across the genotypes revealed the same patterns found in the long bones (data not shown).
Delay in replacement of cartilage by bone. Methylmethacrylate sections of phalanges at E18.5 (a–c) and paraffin sections of tibiae at E16.5 (d–f) of wild-type, PTH/PTHrP-R–/–, and PTHrP–/– animals. Blood vessel invasion and deposition of bone by osteoblasts are delayed in PTH/PTHrP-R–/– bones (b and e) when compared with wild-type bones (a and d). In contrast, in PTHrP–/– bones, these processes are advanced (c and f).
These studies of early bone development demonstrate that PTH/PTHrP-R–/– bones exhibit a partial loss in polarity of the normal gradient of cellular differentiation from the center to the end of the bone. They also exhibit a delay in chondrocyte differentiation, followed by a delay in blood vessel invasion and bone deposition. The PTHrP–/– mice share the loss in polarity of differentiation, but exhibit no delay in chondrocyte differentiation, blood vessel invasion, or bone deposition.
Abnormal bone formation in PTH/PTHrP-R–/– and PTHrP–/– mice. The abnormalities in chondrocyte development in the PTH/PTHrP-R–/– mice were accompanied by abnormalities in osteoblast development. In the intramembranous bone adjacent to hypertrophic chondrocytes (illustrated by the metaphyseal region of the tibia [Figure 4, a and b] and the diaphyseal region of a phalanx [Figure 4, d and e], both at E18.5), the PTH/PTHrP-R–/– bones unexpectedly revealed multiple layers of osteoblasts and woven bone matrix instead of the normal 1–2 layers of osteoblasts. Also, the matrix of the PTH/PTHrP-R–/– bone demonstrated a delay in mineral deposition (black areas, von Kossa staining), whereas the PTHrP–/– mice show the opposite — an accelerated mineral deposition at the same site (Figure 4, c and f). The increase in metaphyseal intramembranous bone in the PTH/PTHrP-R–/– mice persists as a strikingly thickened diaphyseal cortex in long bones such as the tibia and fibula, shown in cross-section in Figure 5, and in longitudinal section in Figure 6b (both at E18.5). In contrast, the intramembranous bone of the PTHrP–/– bones (Figure 4, c and f) showed no consistent abnormality in osteoblast number, matrix deposition, or mineralization. Analogous patterns were shown in the xiphoid across the genotypes (data not shown).
Increase in osteoblast number and cortical bone in PTH/PTHrP-R–/– animals. Von Kossa staining on methylmethacrylate sections at the level of the metaphyseal region of a tibia (a–c) and at the diaphyseal region of a phalanx (d–f) in wild-type, PTHrP–/–;PTH/PTHrP-R–/–, and PTHrP–/– animals at E18.5. PTH/PTHrP receptor mutant bones reveal an abnormal augmentation in osteoblast layers accompanied by an increased bone matrix (b and e) that does not mineralize, as demonstrated by the lack of von Kossa staining. In contrast, PTHrP–/– bones (c and f) look indistinguishable or somewhat advanced in terms of mineralization and replacement of cartilage by bone when compared with wild-type bones (a and d).
Increase in cortical bone in PTH/PTHrP-R–/– embryos. Transverse section of tibia and ulna of E18.5 wild-type (a) and PTH/PTHrP-R–/– embryos (b). The abnormal increase in cortical bone in the mutant is clearly indicated by the black brackets.
Decrease in trabecular bone in PTH/PTHrP-R–/– tibia. Von Kossa staining of a wild-type (a) and a PTH/PTHrP-R–/– (b) tibia at E18.5. Increase in cortical bone and dramatic diminution in trabecular bone (arrow) are shown in b.
In contrast to the intramembranous bone, the bone in the primary spongiosa, which is laid down on the matrix produced by hypertrophic chondrocytes after vascular invasion, is dramatically diminished in the bones of PTH/PTHrP-R–/– mice (Figure 6). Correspondingly, all of the long bones of these mice exhibit little trabecular bone throughout the metaphyseal and diaphyseal regions (data not shown). In contrast, the primary spongiosa and trabecular bone of the PTHrP–/– bones differ little from wild-type bones (data not shown).
Bones of double-homozygous PTHrP–/–;PTH/PTHrP-R–/– embryos. The contrasting patterns of chondrocyte differentiation, vascular invasion, and bone deposition in the PTH/PTHrP-R–/– and PTHrP–/– mice probably derive from the complexity of the signaling network in which the PTH/PTHrP receptor and PTHrP participate. It may be that PTH or other ligands compensate for the lack of PTHrP in PTHrP–/– mice, but could not compensate for the lack of the PTH/PTHrP receptor. Alternatively, PTHrP might act on other receptors to generate phenotypes distinct from those controlled by the PTH/PTHrP receptor. To evaluate these alternatives, we examined the paws and xiphoids of 5 double-homozygous mice at E18.5, and compared them in blinded fashion to the bones of 5 PTHrP–/– and 5 PTH/PTHrP-R–/– mice. Consistently, the double-homozygous mice exhibited a partial rescue of the delay in blood vessel invasion observed in the PTH/PTHrP-R–/– mice (Figure 7). In contrast, the enhanced intramembranous bone formation with multiple layers of cortical osteoblasts seen in the PTH/PTHrP-R–/– mice was also seen in the double-homozygous mice.
Partial rescue of vascular invasion in bones of double-homozygous mice. Hematoxylin/eosin staining of a PTH/PTHrP-R–/– (a) and a PTHrP–/–;PTH/PTHrP-R–/– double-homozygous (b) phalanx at E18.5. The additional ablation of PTHrP from the PTH/PTHrP receptor gene knockout bone leads to a partial rescue of the delay in vascularization. Note the red blood cells amidst the chondrocytes in the double mutants (b).