The midregion, nuclear localization sequence, and C terminus of PTHrP regulate skeletal development, hematopoiesis, and survival in mice

Abstract

The functions of parathyroid hormone-related protein (PTHrP) on morphogenesis, cell proliferation, apoptosis, and calcium homeostasis have been attributed to its N terminus. Evidence suggests that many of these effects are not mediated by the N terminus but by the midregion, a nuclear localization sequence (NLS), and C terminus of the protein. A knock-in mouse lacking the midregion, NLS, and C terminus of PTHrP (Pthrp(Delta/Delta)) was developed. Pthrp(Delta/Delta) mice had craniofacial dysplasia, chondrodysplasia, and kyphosis, with most mice dying by d 5 of age. In bone, there were fewer chondrocytes and osteoblasts per area, bone mass was decreased, and the marrow was less cellular, with erythroid hypoplasia. Cellular proliferation was impaired, and apoptosis was increased. Runx2, Ocn, Sox9, Crtl1, beta-catenin, Runx1, ephrin B2, cyclin D1, and Gata1 were underexpressed while P16/Ink4a, P21, GSK-3beta, Il-6, Ffg3, and Ihh were overexpressed. Mammary gland development was aberrant, and energy metabolism was deregulated. These results establish that the midregion, NLS, and C terminus of PTHrP are crucial for the commitment of osteogenic and hematopoietic precursors to their lineages, and for survival, and many of the effects of PTHrP on development are not mediated by its N terminus. The down-regulation of Runx1, Runx2, and Sox9 indicates that PTHrP is a modulator of transcriptional activation during stem cell commitment.

Figure 1

Mice lacking PTHrP-(67–137) (Δ/Δ) have impaired growth rate, increased mortality, and gross morphological abnormalities. A) Growth rate in Δ/Δ mice was significantly stunted when compared to control mice (+/+, +/Δ), with no differences between +/+ and +/Δ animals. B) Cumulative mortality rate for Δ/Δ pups was 35% by d 2 and 95% by d 10. C) Δ/Δ mice had a chondrodysplastic phenotype that was present as early as d E18.5 of gestation (i, white arrow), at birth (ii), but was clearly evident at d 3 postpartum (iii). Dwarfism and kyphosis were apparent in pups older than 3 d (iii–v). Craniofacial dysplasia with foreshortening was present as early as E18.5 (i, black arrow). D) Skeletal staining of 3-d-old Δ/Δ mice revealed less calvarial and rib mineralization (black arrows), and chondrodysplasia (white arrows). E) Micro-CT images of 15-d-old mice confirmed cranial dysplasia, spherical skulls, rostrocaudal facial shortening, kyphosis, and chondrodysplasia. F) Gross craniofacial morphology of 2- to 12-d-old Δ/Δ pups revealed a spherical skull with rostral foreshortening, tongue protrusion (black arrows), and delayed dental eruption with malaligned incisors that had a chalky discoloration (white arrow). G) On micro-CT, lateral, coronal, and rostral views showed morphological differences between Δ/Δ and control skulls (i–iv). There were differences in dental structure. Notice the proximity of the tympanic ring (white arrow) to the third molar in the Δ/Δ skull (iii), confirming foreshortening of the maxillae and mandible.

Figure 2

Lack of PTHrP-(67–137) is associated with aberrant osteogenesis and odontogenesis. A–C) Gross and microscopic evaluation of the tibia revealed shorter bone length and red blood cell retention in mutant (Δ/Δ) mice. D–J) Histomorphometric analysis of Δ/Δ long bones demonstrated shorter physeal and growth plate length, with shorter reserve, proliferative, and hypertrophic zones; shorter primary spongiosa (F, G, J); lower numbers of resting, proliferating, and hypertrophic chondrocytes; chondrocyte disorientation; loss of columnar polarity (F, G); and thinner cortical bone (H, I). The Δ/Δ bone marrow had decreased cellularity and lower numbers of hematopoietic precursors, and sinusoids were dilated with red blood cells (H, I; Supplemental Fig. S2). K–N) Osteoblast numbers were lower (K, L), while TRAP⁺ cell numbers were higher in the Δ/Δ primary spongiosa (M, N). O) In vitro, Δ/Δ calvarial osteoblast proliferation was lower from d 1–7, but there was no difference in mineralization (von Kossa staining) and osteoclast differentiation (TRAP) among genotypes. Treatment with PTHrP-(1–141), PTHrP-(1–86), or PTHrP-(67–139) enhanced proliferation but to a lesser level than wild-type osteoblasts. P) Histology of incisors from 2- to 3-d-old mice Δ/Δ mandibles revealed flattened trabeculae, altered cytodifferentiation, abnormal cuspal formation, and decreased bone area. Q) Higher numbers of TRAP⁺ cells. Values are means ± sd; n = 6. *P < 0.05; **P < 0.01. View: ×40.

Figure 3

Absence of PTHrP-(67–137) impairs hematopoiesis, mammogenesis, brain function, and skin development. A) Vascular channels of Δ/Δ of E18.5 fetuses and d 2–6 pups were dilated with erythrocytes, with this difference becoming less evident by d 6. Inset: megakaryocyte (von Willebrand factor-immunopositive cell) numbers were decreased in Δ/Δ specimens. B) Bone marrow of 2-d-old Δ/Δ pups was less cellular and had a higher myeloid: erythroid (M:E) ratio that was less pronounced by d 6, lower numbers but a higher percentage of Lin⁻ Sca-1⁺ c-Kit⁺ hematopoietic stem cells, and higher Il-6 mRNA expression. C) Mammary gland histology revealed that Δ/Δ fetuses lacked the epidermal changes that characterize nipple morphogenesis observed in control fetuses; however, mammary ducts were observed in Δ/Δ fetuses (black arrow; ×100). D) A number of Δ/Δ brains had cerebellar herniation (arrow) and hydrocephalus (asterisk), which likely was associated with the intention tremors (Supplemental Video S3) observed in these mice (×20). E, F) Δ/Δ pups older than 3 d of age had marked seborrhea sicca (E), which histologically revealed epidermal and follicular basket-weave to compact hyperkeratosis (F). G) Δ/Δ pups had subcutaneous and visceral adipose tissue depletion that was evident on histological examination (5-d-old mice). Values are means ± sd. *P < 0.05.

Figure 4

Ablation of PTHrP-(67–137) alters gene expression. A) mRNA expression of genes involved in osteogenesis, chondrogenesis, apoptosis, and hematopoiesis (2-d-old mice; n=6) revealed dysregulation, with several genes underexpressed (Runx2, Ocn, Sox9, Crtl1, β-catenin, Bcl2, Runx1, Gata1, PU.1) and overexpressed (Fgf3, Ihh) in the Δ/Δ tibia compared to controls (+/+; +/Δ). No difference in Pthrp, or PTH1R and PPARγ mRNA expression was found (not shown). Similar expression patterns were found in the calvaria (Supplemental Fig. S3). B) Western blot analysis of tibial and bone marrow protein extracts revealed lesser amounts of RUNX2, SOX9, RUNX1, EPHRIN B2 (EPB2), β-CATENIN (β-CAT), Rb, PAX5, NOTCH1, and GATA1, and increased levels of GSK-3β, P21, and P16/Ink4a. Values are means ± sd. *P < 0.05.

Figure 5

Model for the role of PTHrP-(67–137) in skeletogenesis and hematopoiesis. A) The work presented here indicates that PTHrP-(67–137) increases Runx2 mRNA expression to lead mesenchymal stem cells to the osteogenic and chondrogenic lineages. By inducing Sox9 mRNA expression in the epiphyses and metaphyses, PTHrP commits stem cells to chondrogenesis, while in the diaphysis, PTHrP-driven mRNA expression of Runx2 commits stem cells to osteogenesis. PTHrP-(67–137) also regulates other important skeletogenic genes (β-catenin, Ihh, Fgf3, Crtl1, ephrin B2) (Fig. 4A). B) Underexpression of Runx1 and Notch1, which are required for hematopoiesis, and underexpression of PU.1, Pax5, and Gata1, which are necessary for further hematopoietic differentiation, links PTHrP to functions not previously described. It also suggests that PTHrP participates in the communication between the bone/bone marrow compartments. C) Dysregulation of genes involved in apoptosis and aging (GSK-3β, P16/Ink4a, P21, β-catenin, P63, Bcl-2) supports PTHrP as a regulator of these processes. HSC, hematopoietic stem cell; LMP, lymphoid-myeloid progenitor; GMP, granulocyte-monocyte progenitor; CLP, common lymphoid progenitor; MEP, megakaryocyte-erythroid progenitor; Meg, megakaryocyte.