Bone anabolic agents repair alveolar bone damaged by diabetes in mice

Abstract

Type 2 diabetes (T2D) is a chronic metabolic disorder that significantly impacts skeletal health, including bones of the oral cavity, where alveolar bone is particularly susceptible due to its high remodeling rate and sensitivity to systemic metabolic changes. Using a mouse model of established T2D induced by high-fat diet and streptozotocin, we investigated the effects of the FDA-approved bone anabolic agents parathyroid hormone (PTH) and abaloparatide (ABL) on alveolar bone. In non-diabetic mice, PTH and ABL stimulated both bone formation and resorption, resulting in no net change in alveolar bone mass. In contrast, these agents selectively increased bone formation in T2D mice without further enhancing the already elevated resorption, leading to net alveolar bone gain. This anabolic effect involves suppression of sclerostin and Dickkopf-1 (Dkk-1) in osteocytes and activation of Wnt signaling in periodontal ligament cells. Studies using DMP1-Cre;PTH1R^fl/fl mice demonstrated that osteocytic PTH1R is essential for the anabolic effects but not for the bone resorption-promoting effects of PTH and ABL. These findings reveal a dual mechanism of PTH1R signaling in oral bone remodeling and suggest therapeutic strategies to restore alveolar bone and improve tooth retention and implant osseointegration in diabetic patients.

Introduction

Type 2 diabetes mellitus (T2D) is a prevalent chronic metabolic disorder characterized by hyperglycemia, insulin resistance, and relative insulin deficiency. Globally, an estimated 462 million individuals were affected by T2D in 2017, representing 6.3% of the population, with projected prevalence increasing to 7.1% by 2030.^1,2 Among its systemic effects, T2D significantly compromises skeletal health, leading to reduced bone quality, increased fracture risk, and impaired bone healing.^3,4

In the oral cavity, T2D exacerbates the loss of alveolar bone, the specialized bone structure that supports and anchors the teeth. This bone loss leads to tooth loosening, increased mobility, and eventual tooth loss, severely impairing patients’ quality of life.⁵ Compared to non-diabetic individuals, patients with T2D have a threefold higher risk of developing periodontitis,^6–8 a chronic inflammatory disease that further disrupts bone homeostasis. The pathogenesis of periodontitis involves elevated inflammatory cytokines and receptor activator of nuclear factor kappa-B ligand (RANKL) that promote osteoclast-mediated bone resorption while inhibiting osteoblast function, thereby disrupting bone remodeling.⁹ Collectively, the interplay among diabetes, periodontitis, and alveolar bone loss synergistically exacerbates disease progression and increases the risk of tooth mobility.

Parathyroid hormone (PTH) signaling plays a central role in skeletal homeostasis.^10,11 PTH and parathyroid hormone-related protein (PTHrP) mediate their effects via the PTH/PTHrP receptor (PTH1R), a G protein-coupled receptor expressed in osteoblasts, osteocytes, and bone marrow stromal cells.^10,11 The skeletal actions of PTH are context-dependent, as intermittent administration exerts anabolic effects by favoring bone formation, whereas continuous exposure induces catabolic effects by promoting bone resorption. Clinically, recombinant human PTH (1-34) (teriparatide) is approved for osteoporosis treatment, demonstrating increased bone mineral density and reduced fracture risk through its anabolic actions.¹² Abaloparatide (ABL), a synthetic PTHrP analog, also activates the PTH1R pathway and is an alternative bone anabolic agent.^13,14 Our previous preclinical studies in diabetic mouse models showed that both PTH and ABL effectively restored bone mass and improved bone strength even under hyperglycemic conditions in bones of the axial and appendicular skeleton.^15,16

In dentistry, PTH and its analogs have shown promise for alveolar bone regeneration and periodontal healing. Preclinical models have shown that intermittent PTH administration enhances alveolar bone formation, improves the osseointegration of dental implants, and accelerates orthodontic tooth movement through increased bone turnover.^17,18 Furthermore, PTH and ABL reduce alveolar bone loss and promote periodontal tissue regeneration in a rat model of periodontitis by modulating the balance between bone formation and resorption.¹⁹ However, whether PTH and/or ABL can restore alveolar bone in the context of T2D remains unexplored. In addition, the mechanisms underlying alveolar bone protection and the identity of the target cells mediating the effects of the PTH1R ligands have not been studied.

The current study investigates the effects of PTH and ABL on alveolar bone remodeling in a mouse model of established T2D, which recapitulates key features of human T2D-associated alveolar bone loss.²⁰ We demonstrate that PTH and ABL selectively enhance bone formation in T2D mice without further exacerbating the already elevated bone resorption, resulting in net alveolar bone gain. This anabolic effect is mediated by suppression of sclerostin and Dickkopf-1 (Dkk-1) in osteocytes and activation of Wnt signaling in periodontal ligament (PDL) cells. Using DMP1-Cre;PTH1R^fl/fl conditional knockout mice, we further identify DMP1-Cre-positive cells, primarily osteocytes, as the key cellular mediators of the anabolic effects of PTH1R signaling, while DMP1-Cre-negative marrow cells drive its pro-resorptive effects, revealing a cell-type-specific duality in PTH1R-mediated bone remodeling. We conclude that bone anabolic regimens using PTH1R ligands have the potential to restore alveolar bone, enhance tooth stability, and promote oral health in diabetes.

Results

PTH and ABL restore bone volume and alveolar bone height in T2D mice

To confirm successful induction of T2D, fasting blood glucose was measured throughout the progression of the study (Fig. 1a). Blood glucose ≥250 mg/dL was confirmed 4 weeks after STZ administration and remained elevated until the end of the experiment. Longitudinal blood glucose and body weight data for this cohort are reported in our recent publication.²¹ Consistent with earlier studies with the HFD/STZ model.²² T2D mice exhibited significantly lower serum insulin levels compared to non-diabetic mice. This reduction persisted through endpoint and was unaffected by PTH or ABL treatment (Fig. S1), suggesting that any subsequent skeletal responses to PTH or ABL were not attributable to improved glycemic control.

Fig. 1.

PTH and ABL restore alveolar bone volume and height in T2D PTH1R^fl/fl mice. a The T2D group was fed a high-fat diet (HFD) starting at 3 months of age, followed by streptozotocin (STZ) injection at 4 months of age to induce T2D. Diabetes was confirmed at 5 months of age. At 6 months of age, mice were further divided into treatment groups: vehicle, PTH, or ABL. Treatments were administered for 4 weeks, with tissue collection at 7 months of age. b–g Representative sagittal μCT images of maxillary molars. Dashed yellow outline in b indicates the region of interest (ROI) used for quantification. h Quantitative analysis of bone volume fraction (BV/TV, %) in maxillary molars across experimental groups. i–n Buccal views of the maxillary first molars. Yellow arrows indicate the CEJ-ABC measurement site. o Quantitative measurement of CEJ-ABC distances for buccal surfaces of maxillary first molars in all groups. * P < 0.000 1, ^ P < 0.001, and # P < 0.01. Scale bars: 1 mm

In non-diabetic PTH1R^fl/fl mice, treatment with PTH or ABL did not significantly alter bone volume/total volume (BV/TV) or total bone volume (Fig. 1b-d, quantified in 1h; Fig. S2A), indicating minimal impact on alveolar bone under normal conditions. Conversely, T2D mice exhibited a marked reduction in BV/TV and total bone volume compared to non-diabetic PTH1R^fl/fl mice (Fig. 1b, e, quantified in 1h; Fig. S2A), consistent with our previous findings that diabetes adversely affects alveolar bone microarchitecture and reduces bone quantity.²⁰ Notably, both PTH and ABL significantly increased BV/TV and total bone volume in T2D mice, demonstrating their capacity to counteract diabetes-induced bone loss (Fig. 1e-g, quantified in 1h; Fig. S2A).

We next evaluated cementoenamel junction-alveolar bone crest (CEJ-ABC) distance at the maxillary first molar, an index of alveolar bone height relative to the tooth. In non-diabetic mice, PTH and ABL treatments did not significantly change buccal or lingual CEJ-ABC distances, whereas T2D mice exhibited increased CEJ-ABC distances that were restored to non-diabetic levels by PTH and ABL treatments (Fig. 1i-n, Fig. S2B–G, quantified in 1o and Fig. S2H), indicating attenuation of diabetes‑induced loss of alveolar bone height.

PTH and ABL enhance osteogenesis by modulating Wnt signaling

To elucidate the mechanisms underlying PTH and ABL-induced alveolar bone formation, we performed histological analyses. H&E staining showed morphological changes consistent with μCT bone volume increases (Fig. S3A–F). We then examined key regulators of bone formation in PDL cells, which reside at the PDL-alveolar bone interface and mediate periodontium remodeling.^23,24 In vehicle-treated T2D mice, the levels of the Wnt antagonist sclerostin were elevated in osteocytes of alveolar bone and in the PDL area compared with vehicle-treated non-diabetic mice (Fig. 2a, d). PTH and ABL significantly suppressed sclerostin expression in osteocytes of both non-diabetic (Fig. 2a–c; quantified in 2g) and T2D mice (Fig. 2d–f; quantified in 2g), with PTH showing greater potency. Similarly, Dkk-1, another Wnt signaling inhibitor, was low in osteocytes and PDL cells of vehicle-treated non-diabetic mice, increased in vehicle-treated T2D mice, and reduced by PTH/ABL in both non-diabetic and T2D osteocytes (Fig. S3G–M). To directly assess Wnt signaling activation, Tcf-1 expression was quantified in PDL cells. PTH and ABL treatments increased Tcf-1-positive cells, particularly in the PDL adjacent to the alveolar bone (Fig. 2h–j, quantified in 2n). In contrast, T2D mice showed reduced Tcf-1 expression, which was restored by PTH and ABL treatments (Fig. 2k–m, quantified in 2n).

Fig. 2.

PTH and ABL promote alveolar bone formation by regulating Wnt signaling in PTH1R^fl/fl mice. a–f Immunofluorescence staining of sclerostin. g Quantification of sclerostin-positive osteocytes (%). h–m Immunofluorescence staining of Tcf-1. n Quantification of Tcf-1-positive PDL cells (%). o–t Immunofluorescence staining of RUNX2. u Quantification of RUNX2-positive PDL cells (%). *P < 0.000 1, ^ P < 0.001. ab, alveolar bone; pdl, periodontal ligament; cc, cementum; maxM1, maxillary first molar. Scale bars: 50 µm

RUNX2, a key transcription factor for osteogenic differentiation, was also markedly upregulated following PTH and ABL treatments in non-diabetic mice (Fig. 2o–q, quantified in 2u). Vehicle-treated T2D mice demonstrated significantly diminished RUNX2 expression in PDL cells (Fig. 2r), while both treatments effectively restored RUNX2 expression to that observed in vehicle-treated non-diabetic mice (Fig. 2r–t, quantified in 2u).

Thus, PTH and ABL promote alveolar bone formation by suppressing osteocytic sclerostin and Dkk-1 to activate canonical Wnt signaling and osteogenic differentiation in PDL cells.

PTH and ABL stimulate osteoclastogenesis in alveolar bone

Although PTH and ABL strongly suppressed sclerostin and Dkk-1 to enhance Wnt signaling (Fig. 2 and Fig. S3), μCT showed only limited net bone gain in non-diabetic PTH1R^fl/fl mice (Fig. 1), prompting us to examine bone resorption.

In non-diabetic mice, PTH and ABL treatments significantly increased Cathepsin K-positive osteoclasts compared to vehicle-treated mice (Fig. 3a–c, quantified in 3g). In vehicle-treated T2D mice, Cathepsin K-positive cells were elevated (Fig. 3d, quantified in 3g), consistent with pathological bone resorption associated with diabetes. Interestingly, PTH and ABL treatments did not further increase Cathepsin K-positive cells in T2D mice (Fig. 3e, f, quantified in 3g).

Fig. 3.

PTH and ABL stimulate osteoclastogenesis in non-diabetic but not in T2D PTH1R^fl/fl mice. a–f Immunofluorescence staining for Cathepsin K, a marker of osteoclast activity, at the PDL-bone interface. a A low-magnification overview of the maxillary first molar (m1); the boxed region indicates the PDL-bone interface area shown at higher magnification in a’, with all other panels depicting the same anatomical region. Dashed lines outline the roots of the maxillary first molar. g Quantitative analysis of Cathepsin K-positive cells (N/mm). h–mTRAP staining to visualize osteoclasts at the PDL-bone interface in sagittal sections of maxillary first molars. n Quantitative analysis of TRAP-positive cells (N/mm). ab: alveolar bone; pdl: periodontal ligament; d: dentin. * P < 0.000 1, ^ P < 0.001. Scale bars: 50 µm

TRAP staining confirmed increased osteoclast numbers on bone surfaces in PTH/ABL-treated non-diabetic mice versus vehicle-treated non-diabetic mice, but no additional increase in PTH/ABL-treated T2D mice versus vehicle-treated T2D mice (Fig. 3h–m, quantified in Fig. 3n).

Thus, despite PTH/ABL-induced Wnt signaling enhancement promoting osteogenesis, PTH and ABL also stimulate osteoclast differentiation, leading to increased osteoclast numbers, in non-diabetic mice, counteracting the anabolic effects and limiting net bone gain. In T2D mice, where resorption is already pathologically elevated, PTH/ABL enhances osteogenesis without additional resorption, resulting in net alveolar bone gain.

PTH and ABL modulate bone-marrow interface remodeling in alveolar bone

Beyond the PDL-bone interface, we examined the effects of PTH and ABL in the alveolar bone marrow microenvironment, focusing on endosteal surface remodeling (Fig. 4a). RUNX2 immunofluorescence showed increased osteogenic cells along the bone surface following PTH or ABL treatment in non-diabetic PTH1R^fl/fl mice (Fig. 4b–d, quantified in 4h). In contrast, T2D mice exhibited reduced RUNX2 expression, which was restored by both PTH and ABL treatments (Fig. 4e–g, quantified in 4h).

Fig. 4.

PTH and ABL modulate bone-marrow interface remodeling in non-diabetic and T2D alveolar bone in PTH1R^fl/fl mice. a Low-magnification view of the maxillary first molar. The boxed area schematically indicates the bone-marrow interface region analyzed at higher magnification in this figure. b–g Immunofluorescence staining for the osteogenic marker RUNX2 in the alveolar bone of the maxillary molar 1 furcation area. h Quantitative analysis of RUNX2-positive cells per bone surface (N/mm). i–n TRAP staining to visualize osteoclasts in the alveolar bone in the maxillary molar 1 furcation area. o Quantitative analysis of TRAP-positive cells per bone surface (N/mm). p Schematic summary illustrating the effects of PTH and ABL treatments on bone formation and resorption in non-diabetic and T2D mice. In non-diabetic mice, PTH and ABL stimulate both bone formation and resorption, resulting in no net change in bone mass. In T2D mice, where baseline resorption is elevated, PTH and ABL enhance bone formation without further increasing resorption, leading to an overall gain in bone mass. maxM1, maxillary first molar; ab, alveolar bone; m, marrow. * P < 0.000 1. Scale bars: 25 µm

At the same interface, Cathepsin K-positive osteoclasts were increased in PTH- and ABL-treated non-diabetic mice (Fig. S4A–C, quantified in Fig. S4G). T2D mice showed an elevated number of Cathepsin K-positive osteoclasts, which was not further increased by PTH or ABL treatment (Fig. S4D–F, quantified in Fig. S4G). TRAP staining confirmed a similar pattern within the marrow space (Fig. 4i–n, quantified in 4o ).

Together, these results confirm that the pattern observed at the PDL-bone interface extends to the bone-marrow interface: PTH and ABL enhance osteoblastogenesis in both non-diabetic and T2D mice, but increase osteoclast activity only in non-diabetic mice, favoring net bone gain selectively in T2D (Fig. 4p).

PTH1R deletion from DMP1-Cre-positive cells abrogates the anabolic response to PTH and ABL treatments

To determine whether osteocytes mediate the anabolic effects of PTH and ABL, we generated DMP1-Cre;PTH1R^fl/fl mice, which lack PTH1R in DMP1-Cre-positive cells. Our recent studies demonstrated that these conditional knockout mice are viable and display unaffected tooth eruption, although they exhibit a molar ankylosis phenotype driven by aberrant cementoblast differentiation and excessive cementum deposition.²⁵ Notably, alveolar bone osteoclast activity and baseline bone volume remained largely unaffected,²⁵ indicating that PTH1R deficiency in DMP1-Cre-positive cells selectively disrupts periodontal homeostasis without broadly compromising alveolar bone development.

We applied the same PTH and ABL treatment protocol to DMP1-Cre;PTH1R^fl/fl mice under both non-diabetic and T2D conditions. μCT analysis revealed significantly reduced BV/TV following PTH or ABL treatment in non-diabetic DMP1-Cre;PTH1R^fl/fl mice (Fig. 5a–c, quantified in 5g). T2D further reduced alveolar bone mass (Fig. 5d, quantified in 5g) and neither PTH nor ABL treatment improved it (Fig. 5d–f, quantified in 5g).

Fig. 5.

Effects of PTH and ABL treatments on DMP1-Cre;PTH1R^fl/fl mice under non-diabetic and T2D conditions. a–f Representative μCT images of maxillary molars from a–c non-diabetic DMP1-Cre;PTH1R^fl/fl mice and d–f T2D DMP1-Cre;PTH1R^fl/fl mice treated with (a, d) vehicle, (b, e) PTH, or (c, f) ABL. g Quantitative analysis of bone volume fraction (BV/TV, %). h–m Representative H&E-stained histological sections of maxillary molars. n Quantification of marrow space area (mm²). Scale bars: 1 mm (μCT) and 50 μm (H&E). * P < 0.000 1, ^ P < 0.001, # P < 0.01

Histological analysis supported these findings. Histology confirmed intact trabecular structure in vehicle-treated non-diabetic DMP1-Cre;PTH1R^fl/fl mice (Fig. 5h), which was markedly reduced after PTH/ABL treatment (Fig. 5i, j), and sparse and disrupted microarchitecture in T2D groups (Fig. 5k–m). Consistently, marrow space analysis showed that PTH/ABL enlarged marrow cavities in non-diabetic DMP1-Cre;PTH1R^fl/fl mice compared with vehicle-treated non-diabetic DMP1-Cre;PTH1R^fl/fl mice, and T2D knockouts already exhibited enlarged marrow spaces that were not significantly altered by PTH/ABL treatment (Fig. 5n).

These results demonstrate that deletion of PTH1R from DMP1-Cre-positive cells alters the skeletal response to anabolic treatments, and that PTH1R signaling in these cells is critical for the anabolic response to PTH and ABL in both non-diabetic and diabetic conditions.

PTH1R deletion from DMP1-Cre-positive cells abolishes PTH and ABL-mediated osteogenic responses

To investigate the molecular basis of the defective response of DMP1-Cre;PTH1R^fl/fl mice to PTH/ABL, we quantified sclerostin expression and found that PTH/ABL failed to decrease the number of sclerostin-positive osteocytes in DMP1-Cre;PTH1R^fl/fl mice under both non-diabetic and T2D conditions (Fig. 6a–f; quantified in 6g), in contrast to the suppression observed in PTH1R^fl/fl littermate controls (Fig. 2a–f, g).

Fig. 6.

PTH1R deletion in DMP1-Cre-positive cells abolishes PTH and ABL-mediated osteogenic effects. a–f Immunofluorescence staining for sclerostin with (a–c) non-diabetic DMP1-Cre;PTH1R^fl/fl mice and d–f T2D DMP1-Cre;PTH1R^fl/fl mice treated with (a, d) vehicle, b, e PTH, or c, f ABL. g Quantitative analysis of sclerostin-positive osteocytes (%). h–m Immunofluorescence staining for RUNX2. n Quantitative analysis of RUNX2-positive cells per bone surface (N/mm). Dashed lines outline the dentin surfaces. ab, alveolar bone; d, dentin; m, bone marrow. * P < 0.000 1. Scale bars: 50 μm

Similarly, the expression of the osteogenic transcription factor RUNX2 remained unaltered following PTH or ABL treatment in DMP1-Cre;PTH1R^fl/fl mice (Fig. 6h–m, quantified in 6n), indicating that the osteogenic response observed in PTH1R^fl/fl littermate controls was abolished following PTH1R deletion in DMP1-Cre-positive cells. Analysis of alkaline phosphatase (ALP) expression, a marker of bone formation activity, further corroborated these findings, with no significant differences in ALP-positive bone surface percentages between vehicle-treated and PTH/ABL-treated DMP1-Cre;PTH1R^fl/fl mice under both non-diabetic and T2D conditions (Fig. S5A–F, quantified in Fig. S5G).

Collectively, these findings indicate that PTH1R signaling in DMP1-Cre-positive cells is required for PTH and ABL to mediate key downstream events regulating bone formation, including the suppression of the Wnt inhibitor sclerostin and the induction of osteogenic markers RUNX2 and ALP.

PTH and ABL stimulate osteoclastogenesis despite the absence of PTH1R in DMP1-Cre-positive cells

To determine whether PTH1R expression in DMP1-Cre-positive cells is essential for the catabolic effects of PTH and ABL, we examined osteoclast activity in DMP1-Cre;PTH1R^fl/fl mice. TRAP staining revealed significantly increased numbers of osteoclasts along alveolar bone surfaces in PTH- and ABL-treated non-diabetic DMP1-Cre;PTH1R^fl/fl mice compared with vehicle-treated non-diabetic DMP1-Cre;PTH1R^fl/fl mice (Fig. 7a–c, quantified in 7g). This suggests that PTH and ABL-stimulated osteoclastogenesis persists despite the absence of PTH1R in DMP1-Cre-positive cells. However, in T2D DMP1-Cre;PTH1R^fl/fl mice, PTH and ABL treatments did not further increase TRAP-positive osteoclast numbers compared with vehicle treatment (Fig. 7d–f, quantified in 7g).

Fig. 7.

PTH and ABL stimulate osteoclastogenesis despite the absence of PTH1R in DMP1-Cre-positive cells. a–f Representative TRAP staining images showing osteoclasts in alveolar bone sections from (a–c) non-diabetic DMP1-Cre;PTH1R^fl/fl mice and (d–f) T2D DMP1-Cre;PTH1R^fl/fl mice treated with (a, d) vehicle, (b, e) PTH, or (c, f) ABL. g Quantification of TRAP-positive cells (N/mm). h, i Immunofluorescence staining for PTH1R in bone marrow of h PTH1R^fl/fl littermate controls and I DMP1-Cre;PTH1R^fl/fl mice. j Quantification of RANKL-positive marrow cells (%), OPG-positive marrow cells (%), and RANKL/OPG ratio in the different experimental groups. k Schematic model illustrating the mechanism of PTH1R signaling in PTH1R^fl/fl versus DMP1-Cre;PTH1R^fl/fl mice. In PTH1R^fl/fl mice (left), PTH/ABL binds to PTH1R to regulate both bone formation (via sclerostin/Dkk-1) and bone resorption (via RANKL), maintaining normal alveolar bone remodeling. In DMP1-Cre;PTH1R^fl/fl mice (right), despite the absence of PTH1R in DMP1-Cre-positive cells, PTH/ABL can still stimulate RANKL production through DMP1-Cre-negative cells, promoting bone resorption and resulting in alveolar bone loss. ab: alveolar bone; m, marrow. * P < 0.000 1. Scale bars: 50 μm

While PTH1R expression is effectively ablated in cementoblasts, osteoblasts, and osteocytes of DMP1-Cre;PTH1R^fl/fl mice (Fig. S6),²⁵ the persistent pro-resorptive response in non-diabetic knockouts suggests that other PTH1R-expressing cells drive this effect. Indeed, immunofluorescence analysis revealed abundant PTH1R expression remaining in bone marrow cells (Fig. 7h, i), prompting us to examine whether marrow cell PTH1R signaling drives the pro-resorptive response.

Further analysis showed that PTH/ABL significantly increased RANKL-positive marrow cells in both PTH1R^fl/fl littermate controls and DMP1-Cre;PTH1R^fl/fl mice (Fig. S7A–F), while osteoprotegerin (OPG) expression was markedly reduced (Fig. S7G–l), resulting in an elevated RANKL/OPG ratio that favored osteoclastogenesis (Fig. 7j).

Together, these data indicate that while osteocytic PTH1R mediates the anabolic effects of PTH/ABL (Figs. 5, 6), their pro-resorptive effects are driven by PTH1R signaling in DMP1-Cre-negative marrow cells via RANKL upregulation and OPG suppression (Fig. 7k).

Discussion

Craniofacial bones, like those of the axial and appendicular skeleton, are profoundly affected by T2D. In particular, alveolar bone, which anchors teeth within their sockets, undergoes significant loss in T2D, compromising tooth stability. However, the mechanisms underlying diabetes-induced alterations in the periodontium and whether these changes can be therapeutically reversed remain poorly understood. The current study demonstrates that T2D exacerbates alveolar bone resorption while reducing formation via elevated osteocytic sclerostin and Dkk-1 expression and impaired Wnt signaling, and that PTH/ABL restore bone volume by reversing these molecular defects through actions on DMP1-Cre-positive cells, primarily osteocytes. PTH/ABL stimulate bone formation in both physiological and T2D conditions, but their pro-resorptive effects—mediated by marrow cell PTH1R via RANKL upregulation—are blunted in T2D, where resorption is already elevated, resulting in net bone gain (Figs. 1–4). Using DMP1-Cre;PTH1R^fl/fl mice, we further show that PTH1R signaling in DMP1-Cre-positive cells is essential for the anabolic effects of PTH/ABL but dispensable for their pro-resorptive effects, which are mediated instead by PTH1R signaling in marrow cells via RANKL/OPG modulation (Figs. 5–7). This uncoupling of anabolic and catabolic responses has significant implications for tailoring PTH1R-based therapies for diabetic bone disease.

Diabetes-specific modulation of PTH/ABL signaling in bone homeostasis

Our findings indicate that PTH and ABL have minimal effects on net alveolar bone mass in non-diabetic conditions but significantly promote bone formation in T2D. This aligns with evidence that PTH exhibits limited efficacy on intact alveolar bone but exerts pronounced regenerative effects at sites of injury.^19,26 Although ABL was previously reported to stimulate bone formation even in non-injured alveolar bone,¹⁹ our results differ, likely due to variations in treatment regimens. Despite a consistent 4-week treatment period in both studies, we utilized a daily dose of 47.5 μg/kg ABL,^15,27,28 whereas Li et al. administered a higher ABL dose (80 μg/kg) but with less frequent, every-other-day dosing.¹⁹ Both studies employed a 4-week treatment period.¹⁹ These discrepancies suggest that total dosage and administration frequency are critical determinants of ABL’s anabolic balance. Further investigation into optimized dosing strategies is essential to maximize the regenerative potential of PTH and ABL in alveolar bone.

Although the net anabolic impact of PTH and ABL on non-diabetic alveolar bone was offset by a concurrent increase in osteoclast number, both agents markedly enhanced osteogenesis in T2D mice, where osteoclast number was already elevated at baseline and was not further increased by PTH/ABL. In this setting, the anabolic effects of treatment predominate, leading to net alveolar bone gain. Several factors may contribute to this shift under T2D conditions. Previous work has suggested that PTH can exert anti-inflammatory actions in the bone marrow microenvironment.^27,29 which could, in principle, mitigate inflammation-driven bone resorption in diabetes. Future studies are warranted to directly assess the role of pro-inflammatory mediators in PTH/ABL action in the alveolar bone. In addition, intrinsic negative-feedback mechanisms that constrain osteoclast expansion and survival may impose a functional ceiling on resorptive activation, beyond which PTH/ABL produces no further increase in resorption in the already elevated T2D context. This hypothesis will require direct testing in future studies. Concurrently, the anabolic potency of PTH and ABL appears particularly pronounced in T2D due to their action on dysregulated Wnt signaling. Diabetic bone shows elevated sclerostin and Dkk-1, and PTH/ABL directly suppress these inhibitors, thereby enhancing canonical Wnt signaling and osteogenic differentiation in PDL cells, consistent with our prior finding that anti-sclerostin antibody treatment rescues T2D-induced alveolar bone loss.²⁰ Notably, PTH showed greater potency than ABL in suppressing sclerostin in alveolar bone. This difference is likely explained, at least in part, by the fact that ABL was administered at half the molar dose of PTH (ABL, 47.5 μg/kgperday [12 pmol/gperday]; PTH, 100 μg/kgperday [24 pmol/gperday]), a dose previously shown to be sufficient to restore bone mass in diabetic mice.¹⁵ Additionally, PTH elicits more sustained cAMP/PKA signaling than ABL,^30–32 the pathway primarily linked to sclerostin regulation in osteocytes.^33,34 While this study focused on osteogenic markers in PDL cells, the increased numbers of RUNX2- and Tcf-1-positive cells may reflect both enhanced differentiation and possible expansion of progenitor populations, as ABL has been reported to promote cell proliferation in the alveolar periosteum.³⁵ Because proliferation markers such as Ki67 were not assessed here, whether PTH/ABL increase the proliferation of PDL cells, including potential progenitor populations, remains to be determined in future studies.

In conclusion, our study indicates that the enhanced net anabolic effect of PTH and ABL on T2D alveolar bone results from concurrently potentiated bone formation and contained bone resorption. Nevertheless, future studies using dynamic histomorphometry with fluorochrome double labeling to measure mineral apposition rate and bone formation rate are needed to provide more direct evidence of increased bone formation and to further dissect the relative contributions of these distinct anabolic and catabolic regulatory pathways in diabetic bone disease, which could guide the optimization of PTH pathway-targeting therapies for this patient population.

Cell-type-specific roles of PTH1R in bone remodeling

Our findings reveal that PTH1R signaling in DMP1-Cre-positive versus DMP1-Cre-negative cells plays distinct roles in alveolar bone remodeling. In DMP1-Cre;PTH1R^fl/fl mice, PTH/ABL failed to suppress sclerostin and increase bone volume (Figs. 5–6), confirming that PTH1R signaling in DMP1-Cre-positive cells is essential for anabolic effects. However, they still stimulate RANKL production and osteoclastogenesis (Fig. 7), indicating that DMP1-Cre-negative cells mediate the pro-resorptive effects. This dual mechanism highlights the complex anabolic-catabolic interplay in the alveolar bone.

What are these DMP1-Cre-positive cells that mediate the anabolic effects of PTH/ABL? Primarily, these are mature osteocytes, as the DMP1-8kb-Cre strain was specifically designed to target this population. Extensive research has established that PTH1R signaling in mature osteocytes is essential for the anabolic response, largely through the suppression of sclerostin and the subsequent activation of Wnt signaling.^36,37 However, the cellular targeting of the DMP1-8kb-Cre strain is broader than initially appreciated, encompassing late-stage osteoblasts.^38–40 and bone marrow stromal cells.⁴¹ Previous studies have shown that intermittent PTH administration upregulates marrow-resident Nestin-positive and LepR-positive mesenchymal stem cells, significantly enhancing their proliferation and osteogenic potential.^42,43 Therefore, although mature osteocytes are likely the principal DMP1-Cre-positive cells mediating the anabolic effects of PTH/ABL, the broader targeting pattern of the DMP1-8kb-Cre strain prevents exclusion of additional contributions from DMP1-Cre-positive marrow-associated cells. Such cells may influence the marrow microenvironment and/or provide osteogenic support, but their precise role remains to be determined.

Our findings indicate that the pro-resorptive effects of PTH/ABL are preserved in DMP1-Cre;PTH1R^fl/fl mice, suggesting that these effects are not primarily mediated by DMP1-Cre-positive osteocytes. Instead, the pro-resorptive actions of PTH/ABL are likely driven by DMP1-Cre-negative cells within the bone marrow, as PTH/ABL-treated mice show increased RANKL-positive and reduced OPG-positive marrow cells together with elevated osteoclast numbers. Although TRAP staining and Cathepsin K expression reflect osteoclast presence and activity rather than direct resorptive function, circulating levels of C-terminal telopeptide of type I collagen (CTX), a validated biochemical marker of bone resorption, were measured in the same experimental cohort,²¹ confirming that PTH and ABL increase systemic bone resorptive activity. The bone marrow harbors several PTH1R-expressing, DMP1-Cre-negative cell populations that have been implicated in regulating the RANKL/OPG axis and therefore represent plausible mediators of this response, including bone marrow stromal cells (such as LepR-positive and Nestin-positive subsets)^43,44 and osteoblast precursors.⁴⁵ In our model, immunofluorescence demonstrates persistent PTH1R expression in marrow cells, but the specific identity and relative contribution of these populations were not directly defined and will require lineage-specific approaches using Cre drivers such as LepR-Cre or Prx1-Cre.

The requirement for PTH1R in DMP1-Cre-positive cells in PTH/ABL-induced bone gain in alveolar bone is consistent with our previous findings in the femur and spine.²¹ However, the cellular contribution to the osteogenic response appears to differ among skeletal sites. In alveolar bone, deletion of PTH1R in DMP1-Cre-positive cells abolished the PTH/ABL-induced increases in osteogenic markers, including RUNX2 and ALP (Fig. 6, Fig. S5). In contrast, we recently observed increased bone formation markers in the femur and spine following PTH/ABL treatment even in DMP1-Cre;PTH1R^fl/fl mice.²¹ These findings suggest that the osteogenic response to PTH/ABL in alveolar bone is more strongly dependent on PTH1R signaling in DMP1-Cre-positive cells, whereas additional cell populations may contribute to this response at other skeletal sites.

Together, this dual mechanism underscores the complexity of PTH signaling and suggests that therapeutic strategies targeting the PTH pathway should consider both sites of action and the involved cell populations.

Conclusion

In conclusion, our study reveals that PTH and PTHrP analogs exert dual effects on alveolar bone remodeling through distinct cellular targets: DMP1-Cre-positive cells, primarily osteocytes, mediate bone formation and bone gain via Wnt signaling, while DMP1-Cre-negative marrow cells drive resorption via RANKL/OPG modulation. These insights provide a foundation for optimizing PTH pathway-targeted therapies in diabetic bone disease and imbalanced remodeling conditions.

Materials and methods

Sex as a biological variable

In this study, only male mice were used because previous studies have shown that male mice exhibit a more pronounced response to high-fat diet (HFD) and streptozotocin (STZ) treatment compared with female mice.^46–48 While our findings provide important insights into T2D-induced periodontal changes and PTH/ABL treatment in male mice, they may not be fully generalizable to female mice.

Animal model of T2D

T2D was induced in mice using a combined HFD and low-dose STZ protocol, as previously described.^15,20 Briefly, 12-week-old male C57BL/6 J mice were randomly assigned to receive either an HFD (Research Diets D12492; 60 kcal% fat, 20 kcal% carbohydrate, 20 kcal% protein) or a control low-fat diet (LFD; Research Diets D12450J; 10 kcal% fat, 70 kcal% carbohydrate, 20 kcal% protein). After 4 weeks on the HFD, mice received five daily intraperitoneal injections of STZ (45 mg/kg in 100 mmol/L citrate buffer, pH 4.6), while LFD-fed mice received buffer. T2D was confirmed by fasting blood glucose ≥250 mg/dL (5-h fast, measured using the AlphaTrak 2 Blood Glucose Monitoring System) 4 weeks after STZ administration.²¹ Mice were then maintained on their respective diets for an additional 4 weeks before initiating treatment (Fig. 1a). Serum insulin levels were also measured at multiple time points using the ALPCO Mouse Insulin ELISA kit.

PTH and ABL administration

Intermittent administration of PTH (1-34, Bachem, Torrance, CA, USA) or ABL (Radius Pharmaceutical, Boston, MA) was performed to evaluate their therapeutic effects in T2D mice. PTH was administered subcutaneously at 100 μg/kg per day for 4 weeks, while ABL was administered at 47.5 μg/kg per day for 4 weeks, starting 4 weeks after T2D confirmation.¹⁵ Vehicle-treated mice received vehicle injections (0.9% saline, 0.01 mmol/L β-mercaptoethanol, 0.1 mmol/L acetic acid) under the same protocol. Mice were harvested 24 hours after the last treatment for analysis (Fig. 1a).

Generation of conditional PTH1R knockout mice

Conditional deletion of PTH1R was achieved by crossing PTH1R^flox/flox (PTH1R^fl/fl) mice, in which the E1 exon of PTH1R is flanked by loxP sites, with DMP1-8kb-Cre mice.⁴⁹ The resulting DMP1-Cre;PTH1R^fl/fl mice and PTH1R^fl/fl littermate controls underwent the same T2D induction and treatment protocol as described above. This experimental design allowed for systematic investigation of the role of PTH1R signaling in DMP1-Cre-positive cells (primarily osteocytes) in mediating the therapeutic efficacy of PTH/ABL anabolic regimens under diabetic conditions.

Microcomputed tomographic (μCT) analysis

Mouse head specimens were scanned and reconstructed using previously established methods.^20,50 Specimens were scanned using a Skyscan 1176 micro-CT system (Bruker, Billerica, MA, USA) under the following acquisition parameters: X-ray energy of 59 kV, isotropic pixel size of 9 μm, and a rotation step of 0.3°. Raw projection images were reconstructed using NRecon software (Bruker) with beam-hardening correction set at 20% and ring artifact reduction applied. Three-dimensional reconstructions were performed with CTvox software (Bruker), and all quantitative measurements were performed using CTan software (Bruker). The region of interest (ROI) for alveolar bone volumetric analysis was defined as the interradicular bone in the furcation area of the maxillary first molar, bounded superiorly by the furcation roof and inferiorly by the alveolar bone crest, encompassing all consecutive sagittal slices through the entire furcation area. Root structures, PDL space, and cortical basal bone were excluded from the ROI. Grayscale thresholding was applied using a fixed threshold range of 75–255 (on an 8-bit, 0–255 grayscale scale) to segment mineralized bone from soft tissue. Bone volume (BV, μm³) and bone volume fraction (BV/TV, %) were calculated from the resulting binarized images. The CEJ-ABC distance was measured separately at the buccal and lingual (palatal) aspects of the maxillary first molar on standardized 2D sagittal sections, as two independent indices of alveolar bone height. Specifically, measurements were taken at the distobuccal cusp region for the buccal CEJ-ABC and at the distopalatal cusp region for the lingual CEJ-ABC. For each sample, measurements were performed in triplicate by a single observer blinded to the experimental groups, and the three replicate values were averaged to yield a single value per sample.

Sample preparation

Specimens were fixed in 4% paraformaldehyde (PFA, Sigma-Aldrich) at 4 °C for 48 h, followed by decalcification in 0.5 mol/L EDTA (pH 7.2–7.4) for 3–5 days with gentle agitation on an orbital shaker as previously described.⁵¹ The EDTA solution was prepared by dissolving EDTA (MW 380.16, ACROS 118440100) and citric acid monohydrate (Fisher, A104-3) in distilled water, with pH adjusted to 7.2–7.4 prior to use. A minimum 10:1 solution-to-tissue volume ratio was maintained throughout decalcification. After thorough rinsing, samples were dehydrated through a graded ethanol series, cleared in xylene, and embedded in paraffin at 60 °C. Sagittal sections (6 μm) were obtained using a microtome and mounted on positively charged slides.

Hematoxylin and eosin (H&E) staining

Tissue sections were deparaffinized in xylene and rehydrated through a descending series of ethanol to distilled water. Slides were stained with Mayer’s hematoxylin (7211, Epredia, Kalamazoo, MI, USA) for nuclear visualization, rinsed, differentiated briefly in acid alcohol, blued in running tap water, and counterstained with Eosin-Y (7111, Epredia) for cytoplasmic and extracellular matrix components. Slides were subsequently dehydrated through an ascending ethanol series, cleared in xylene, and mounted using a permanent mounting medium (Permount, Fisher Scientific).

Tartrate-resistant acid phosphatase (TRAP) staining

Osteoclasts were identified using the Leukocyte Acid Phosphatase (TRAP) Kit (386A, Sigma-Aldrich). Deparaffinized and rehydrated sections were incubated at 37 °C with a freshly prepared staining solution. This solution was mixed according to the manufacturer’s instructions using kit components (Naphthol AS-BI phosphate, Acetate, Tartrate solutions) combined with Fast Red Violet solution (F3381, Sigma-Aldrich) and Sodium Nitrite solution (S2252, Sigma-Aldrich). Incubation was carried out for 2 h under light protection. After washing with distilled water, sections were counterstained with 0.5% Methyl Green (Sigma-Aldrich), washed again, dehydrated, and mounted with Permount.

Immunohistochemistry and immunofluorescence

Immunohistochemical and immunofluorescence analyses were performed as previously described.⁵² Sections were deparaffinized, rehydrated, and subjected to antigen retrieval. Sections were then incubated with primary antibodies. Fluorescence detection employed goat anti-rabbit IgG (A32733, Invitrogen, Carlsbad, CA, USA) and donkey anti-goat IgG (A32849, Invitrogen). Fluorescent sections were counterstained with DAPI and mounted. For chromogenic immunohistochemistry, endogenous peroxidase activity was blocked, followed by incubation with goat anti-rabbit IgG (PI-1000-1, Vector Lab, Newark, CA, USA). Immunoreactivity was visualized using the ImmPACT DAB Substrate kit (SK4105, Vector Lab), producing a brown reaction product at antigen localization sites. DAB-stained sections were counterstained with hematoxylin, dehydrated, cleared, and mounted. The primary antibodies used included anti-Sclerostin (AF1589, R&D Systems, Minneapolis, MN, USA), anti-Dickkopf-1 (Dkk-1) (AF1765, R&D Systems), anti-Tcf-1 (2203, Cell Signaling Technology, Danvers, MA, USA), anti-RUNX2 (ab192256, Abcam), anti-Cathepsin K (ab300569, Abcam, Cambridge, UK), anti-PTH1R (PA5-80955, Invitrogen), anti-Alkaline Phosphatase (ALP, AF2910-SP, R&D Systems), anti-RANKL (A17025, Abclonal, Woburn, MA, USA), and anti-Osteoprotegerin (A2100, Abclonal).

Quantification

All quantitative analyses were performed by observers blinded to the treatment groups using ImageJ software (NIH, Bethesda, MD, USA) unless otherwise specified.

Osteoclasts (TRAP-positive or Cathepsin K-positive multinucleated cells) and osteoblasts (RUNX2-positive cells lining the bone) were quantified by counting the number of positive cells along the alveolar bone surface. Results were normalized to the bone perimeter and reported as positive cells per millimeter bone surface.

Tcf-1^- and RUNX2-positive PDL cells and RANKL^- and OPG-positive bone marrow cells were quantified following previously described methods⁵³ and expressed as percentages (%).

The percentages of Dkk-1⁺ and Sclerostin⁺ osteocytes were calculated within the alveolar bone. The number of marker-positive osteocytes was divided by the total number of osteocytes (identified by DAPI-stained nuclei within lacunae in the bone matrix), expressed as a percentage (%).

ALP-positive bone surface area was measured by calculating the ALP-positive area and the total bone surface area separately. The results were expressed as the percentage of ALP-positive bone surface (%).

In H&E stained sections clearly showing both the distal and mesial roots, the marrow area within the furcation area of the maxillary first molar was measured.

To quantify the PTH1R-positive marrow area, Adobe Photoshop (version 26.5.0, San Jose, CA, USA) was used. The marrow area beneath the furcation was first selected using the magnetic lasso tool, and the total number of pixels within this area was recorded. The selected area was then copied to a new file, where the magic wand tool was used to identify the PTH1R-positive region. The number of pixels in the PTH1R-positive region was recorded as positive pixels. PTH1R-positive area was calculated as the ratio of positive pixels to total pixels and expressed as a percentage (%).

Statistics

Data are presented as mean ± standard deviation (SD). Unless otherwise specified in the figure legends, experiments included n = 8 samples per group. Statistical analyses were conducted using GraphPad Prism 10 (GraphPad Software, Boston, MA, USA). Two-way ANOVA followed by Tukey’s post-hoc test was used for multiple group comparisons. A P-value of <0.05 was considered statistically significant.

Study approval

All animal procedures were conducted with approval of and in accordance with guidelines set by the Institutional Animal Care and Use Committee at the University of Arkansas.

Author contributions

H. Turkkahraman, X. Yuan, and T.M. Bellido designed the research. H. Turkkahraman, C.J. Walton, Y. Koizumi, T. Zhu, N. Akel, and X. Yuan performed data acquisition. H. Turkkahraman and X. Yuan analyzed the data. X. Yuan drafted the manuscript. H. Turkkahraman, C.J. Walton, Y. Koizumi, T. Zhu, N. Akel, X. Yuan, and T.M. Bellido critically revised the manuscript. All authors read and approved the final manuscript.

Data availability

Data supporting the findings of this study are available from the corresponding author upon reasonable request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41368-026-00470-3.