Effects of GBT1118, a voxelotor analog, on bone disease in sickle cell disease mice

Abstract

We assessed the effect of GBT1118, a sickle hemoglobin polymerization inhibitor on bone loss in humanized sickle cell disease (SCD) mice. Healthy control (Ctrl) 4-months-old female and male mice were fed Vehicle-chow for 2-months, while SCD mice were fed Vehicle-chow or GBT1118-chow. By micro-CT, GBT1118 significantly increased femur metaphyseal trabecular thickness (Tb.Th) and tissue mineral density (TMD), and significantly decreased trabecular spacing in female SCD mice. In SCD male mice, there was significant reduction in epiphyseal trabecular bone volume fraction (BV/TV), Tb.Th and TMD and GBT1118 significantly increased BV/TV and TMD but not Tb.Th. A significant decrease in cortical area fraction in SCD female mice was rescued by GBT1118 but not SCD males. Markedly decreased mineralized femur trabeculae in SCD females and males was partially rescued by GBT1118. Bone histomorphometry of femurs demonstrated significantly decreased bone formation parameters and increased bone resorption parameters in SCD mice of both sex that were rescued by GBT1118. Significant alteration in bone and hypoxia related genes of SCD mice of both sexes were differentially modulated by GBT1118. We conclude that "a sickle hemoglobin polymerization inhibitor" might be efficacious in improving some parameters of SCD bone loss.

Introduction

Sickle cell disease (SCD) is the most common inherited blood disorder affecting about approximately 20 million individuals worldwide including some people who come from Hispanic, Southern European, Middle Eastern, or Asian Indian backgrounds¹. SCD affects about 100,000 people in the United States who identify as Black or with African ancestry but also affects people of any ethnic background¹. SCD is caused by a single-nucleotide polymorphism in the beta-globin gene that leads to substitution of valine for glutamic acid at the sixth position in the beta-globin chain¹. Following deoxygenation, the mutated hemoglobin polymerizes to form bundles which result in erythrocyte sickling, aggregation of sickle erythrocytes with neutrophils, platelets, and endothelial cells to promote stasis of blood flow, or vaso-occlusion (VOC) resulting in multiple organ damage, including bone.

Among the organ damage observed in SCD, osteoporosis and osteopenia are common bone complication in affected individuals^2–5. In retrospective studies the prevalence of low bone mineral density (BMD) in SCD patients was approximately 80%⁶ with lumbar spine (55%), radius (30%), and femoral neck (15%) sites most affected⁷. Fracture rates are 28% for transfused and 32% non-transfused male and 16% transfused and 17% non-transfused female SCD patients⁸. Although osteoporosis is prevalent in SCD, there are no consensus guidelines for anti-osteoporotic therapies in SCD patients⁵. Potential adverse effects could limit their use in SCD. Bone, joint, muscle pain, osteonecrosis of the jaw may occur at any point after taking a bisphosphonate, which could potentially worsen acute and chronic pain, and osteonecrosis often seen in SCD patients^9,10. The anabolic effect of parathyroid hormone (PTH1-34) decreases with time and repeated treatment¹¹, and high PTH level is often seen in SCD patients^7,12.

There are limited treatment options for SCD. Although hematopoietic stem cell transplant is curative for SCD, there is only a 15% bone marrow donor to bone marrow recipient match rate¹³. The use of cluster-based regularly interspaced short palindromic repeat sequence (CRISPR/Cas9) system gene editing to increase fetal hemoglobin by inhibiting transcription factors that repress its induction has been investigated in clinical trials, and, although the FDA in 2023 approved Casgevy (exagamglogene Autotemcel), which increases fetal Hb¹⁴, as well as gene therapeutic agent Lyfgenia (Lovotibeglogene autemcel)¹⁵, which is a modified HbA^T87Q, the cost of these therapies may prohibit their wide use in the short term. Complications of the genetic therapy especially Lyfgenia’s associated malignancies were reported¹⁶. Besides genetic therapy, only four FDA approved drugs are available to reduce SCD complications¹⁷. Among these drugs is Voxelotor, a sickle hemoglobin (HbS) polymerization inhibitor that has been approved by the FDA for the treatment SCD in the US and approved for the treatment of hemolytic anemia related to SCD in Europe. Voxelotor was reported to inhibit HbS polymerization¹⁸. Since VOC is a major contributor to metabolic bone disease in SCD, we assessed the effect of 2-hydroxy-6-[(2S)-1-(pyridine-3-carbonyl)piperidin-2yl]methoxy (GBT1118), an analog of Voxelotor (Dufu 2021) on bone loss in SCD since GBT1118 has been shown to improve hemolytic anemia and is renoprotective in transgenic sickle cell disease mice¹⁹.

Materials and methods

Experimental mice

Male and female healthy Control (Ctrl) and Townes sickle cell disease (SCD) mice²⁰ on mixed C57BL/6 and 129 genetic backgrounds were purchased from The Jackson Laboratory (Stock number: 013071, Bar Harbor, Maine, USA) and were housed in the Center for Comparative Medicine at UConn Health. As previously reported²⁰ Townes mice harbor human α- and β-globin genes knocked into the mouse locus, and by crossing sickle cell trait (SCT, heterozygous) mice allows for the generation of control (healthy) and SCD (homozygous) littermates. Previous studies have shown that with aging of SCD patients and progression of SCD, chronic kidney disease (CKD) develops²¹. To mitigate the confounding effects of CKD on bone homeostasis mice beyond 6-months-old were not utilized in this study. Mice were fed with Envigo Teklad Diet-2020, which contains 1% calcium, 0.7% total phosphate, and 1.5 IU D3/g of diet. At 4 months of age, healthy control mice were fed with vehicle control chow (Teklad 2020 Diet, Vehicle), while age and sex matched SCD mice were treated with Vehicle or GBT1118 chow (Teklad 2020 Diet with 4 g/kg GBT1118) for 2 months. GBT1118 chow and control chow are provided by Global Blood Therapeutics (South San Francisco, CA which was acquired by Pfizer Inc. in 2022). The dosage and duration of GBT1118 treatment is based on the improvement of hemolytic anemia in SCD mice¹⁹. Age and sex matched Ctrl and SCD mice were genotype housed to prevent gut microbiota exchange between genotypes due to coprophagic behavior and randomly assigned to treatment group to minimize cage effects. Mice were sacrificed with CO₂ after 2 months of treatment. All animal protocols were approved by The UConn Health Institutional Animal Care and Use Committee. All experiments were performed in accordance with relevant guidelines and regulations. Additionally, the studies complied with the recommendations in the ARRIVE guidelines.

End points

Detailed information about methods, antibodies, and ELISA kits is provided online in the Supplementary Methods and Supplemental Table S1 and S2.

Statistical analysis

Statistical analyses were done using SPSS and p values were calculated using a one-way ANOVA (α = 0.05) with Tukey’s post hoc analysis.

Results

DXA analysis of the effect of GBT1118 on bone density in SCD female and male mice

BMD and BMC were determined in excised femur, tibiae and vertebrae of Control-Vehicle, SCD-Vehicle and SCD-GBT111 treated female and male mice (Table 1). Compared with Control-Vehicle, there were significant reductions in BMD in femur of SCD-Vehicle of both sexes that was not rescued by GBT1118. Femur BMC in SCD of both sexes was further significantly reduced by GBT1118. There were no differences in femur BMC based on sex or genotype among the three groups. Compared with Control-Vehicle, there were significant reductions in BMD and BMC in tibiae of SCD-Vehicle of both sexes that was not rescued by GBT1118. There were no differences in vertebral BMD or BMC based on sex or genotype among the three groups.

Table 1.

Body weight, DXA analysis of femur, tibiae, lumbar vertebrae of Control and SCD female and male mice.

	Ctrl-Veh	SCD-Veh	SCD-GBT1118
Female
BW (g)	24.64 ± 0.95	26.60 ± 0.47	26.89 ± 0.84
Femur BMC (g)	0.0267 ± 0.0004	0.0250 ± 0.0006	0.0239 ± 0.0006
Femur BMD (mg/cm2)	92.9704 ± 1.6663	85.6444 ± 2.3783 a	79.0102 ± 1.5484
Tibia BMC (g)	0.0251 ± 0.0005	0.0222 ± 0.0007 a	0.0217 ± 0.0007
Tibia BMD (mg/cm2)	78.3320 ± 0.6925	67.9331 ± 1.5234 a	64.8289 ± 0.9986
L1–L5 BMC (g)	0.0434 ± 0.0018	0.0454 ± 0.0008	0.0469 ± 0.0022
L1–L5 BMD (mg/cm2)	78.6147 ± 2.5804	74.2428 ± 0.6962	72.1462 ± 1.9110
Male
BW (g)	37.40 ± 1.33	33.46 ± 0.57 a	33.15 ± 0.51
Femur BMC (g)	0.0328 ± 0.0010	0.0298 ± 0.0011	0.0288 ± 0.0010
Femur BMD (mg/cm2)	105.3646 ± 2.4477	96.3396 ± 2.0554 a	88.4586 ± 2.5251
Tibia BMC (g)	0.0313 ± 0.0006	0.0280 ± 0.0007 a	0.0279 ± 0.0010
Tibia BMD (mg/cm2)	80.4986 ± 1.7265	72.9437 ± 1.4929 a	72.7391 ± 1.6394
L1–L5 BMC (g)	0.0556 ± 0.0015	0.0541 ± 0.0012	0.0523 ± 0.0017
L1–L5 BMD (mg/cm2)	88.0595 ± 2.0269	85.4289 ± 1.7462	86.3101 ± 1.6966

Data are mean ± SE, n = 10–12 mice/group.

^aCtrl-Veh vs SCD-Veh P < 0.05.

^bSCD-Veh vs. SCD-GBT1118 P < 0.05.

Effect of GBT1118 on structural parameters of metaphyseal, epiphyseal and diaphyseal femurs in SCD female and male mice

We investigated the effects of GBT1118 on bone micro-architecture in female and male SCD mice by micro-CT and whether there were differential effects on the femur metaphysis, epiphysis and diaphyseal compartments. Three-dimensional representative images of the femur metaphyseal trabecular bone from Control-Vehicle, SCD-Vehicle and SCD-GBT1118 treated female mice are shown in (Fig. 1a). Bone volume fraction (BV/TV) was similar between Control-Vehicle and SCD-Vehicle, and although GBT1118 caused a 30% increase in the BV/TV in SCD mice (Fig. 1b), this was not statistically significantly different from SCD-Vehicle treated. There was no significant difference in trabecular number (Tb.N) among groups (Fig. 1c). Although as shown in (Fig. 1d) trabecular thickness (Tb.Th) and trabecular spacing (Tb.Sp) (Fig. 1e) were similar in Control-Vehicle and SCD-Vehicle, treatment with GBT1118 significantly increased Tb.Th and significantly decreased Tb.Sp in SCD. In contrast to the DXA results, micro-CT showed that tissue mineral density (TMD) in SCD-Vehicle was like Control-Vehicle, however GBT1118 significantly increased TMD in SCD (Fig. 1f).

Figure 1.

Micro-CT analysis of femur metaphysis from Control and SCD female and male mice. Three-dimensional images of the femur metaphysis from female Control-Vehicle, SCD-Vehicle and SCD-GBT1118 treated mice are shown in (a), and structural analysis of metaphyseal parameters are shown for BV/TV (b); Tb.N (c); Tb.Th (d); Tb.Sp (e) and TMD (f). Three-dimensional images of the femurs from male Control-Vehicle, SCD-Vehicle and SCD-GBT1118 treated mice are shown in (g), and structural analysis of metaphyseal parameters are shown for BV/TV (h); Tb.N (i);Tb.Th (j); Tb.Sp (k) and TMD (l). Data are shown as box-and-whisker plots, showing all individual data points. *p < 0.05. n = 10–12 mice/group.

Representative micro-CT images of metaphyseal trabecular bone of male Control-Vehicle, SCD-Vehicle and SCD-GBT1118 treated mice are shown in (Fig. 1g). Micro-CT analysis showed no significant differences in BV/TV (Fig. 1h), Tb.N (Fig. 1i), Tb.Th (Fig. 1j), Tb.Sp (Fig. 1k) or TMD in SCD (Fig. 1l) among the three groups of male mice.

Micro-CT analysis of female femur epiphyseal bone (Fig. 2a–f) revealed no differences in BV/TV (Fig. 2b) or Tb.N (Fig. 2c) among the three groups of mice but a significant reduction in Tb.Th (Fig. 2d) and TMD (Fig. 2f) was observed in SCD-Vehicle compared with Control-Vehicle which was not rescued by GBT1118. Interestingly, micro-CT of male femur epiphyseal bone (Fig. 2g–l) revealed a significant reduction in BV/TV (Fig. 2h), Tb.Th (Fig. 2j) and TMD (Fig. 2l) in SCD-Vehicle compared with Control-Vehicle. GBT1118 significantly increased BV/TV and TMD but not Tb.Th in SCD male mice.

Figure 2.

Micro-CT analysis of femur epiphysis from Control and SCD female and male mice. Three-dimensional images of the femur epiphysis from female Control-Vehicle, SCD-Vehicle and SCD-GBT1118 treated mice are shown in (a), and structural analysis of epiphyseal parameters are shown for BV/TV (b); Tb.N (c);Tb.Th (d); Tb.Sp (e) and TMD (f). Three-dimensional images of the femurs from male Control-Vehicle, SCD-Vehicle and SCD-GBT1118 treated mice are shown in (g), and structural analysis of parameters are shown for BV/TV (h); Tb.N (i);Tb.Th (j); Tb.Sp (k) and TMD (l). Data are shown as box-and-whisker plots, showing all individual data points. *p < 0.05. n = 10–12 mice/group.

Representative micro-CT of femur mid-diaphyseal cortical images from female mice are shown in (Fig. 3a). Quantitative analysis showed that GBT1118 did not alter significantly increased total area (Tt.Ar) (Fig. 3b) and marrow area (Ma.Ar) (Fig. 3d) in SCD mice. The drug reversed the decline caused by SCD in cortical area/total area (Ct.Ar/Tt/Ar) (Fig. 3e). There was no significant difference in cortical thickness (Ct.Th) among groups (Fig. 3f). TMD (Fig. 3g) were significantly lower in SCD-Vehicle group vs. Control-Vehicle but were not significantly increased by GBT1118.

Figure 3.

Micro-CT analysis of femur mid-diaphysis from Control and SCD female and male mice. Three-dimensional images of the femur mid-diaphysis from female Control-Vehicle, SCD-Vehicle and SCD-GBT1118 treated mice are shown in (a), and structural analysis of mid-diaphyseal parameters are shown for Tt.Ar (b); Ct. Ar (c); Ma.Ar (d); Ct.Ar/Tt/Ar (e); CT.Th (f) and TMD (g). Three-dimensional images of the femur mid-diaphysis from male Control-Vehicle, SCD-Vehicle and SCD-GBT1118 treated mice are shown in (h), and structural analysis of parameters are shown for Tt.Ar (i); Ct.Ar (j); Ma.Ar (k); Ct.Ar/Tt/Ar (l); Ct.Th (m) and TMD (n). Data are shown as box-and-whisker plots, showing all individual data points. *p < 0.05. n = 10–12 mice/group.

Representative micro-CT femur mid-diaphyseal cortical images from male mice is shown in Fig. 3h. Quantitative analysis showed that GBT1118 did not rescue the significantly reduced Ct.Ar (Fig. 3j), Ct.Ar/Tt/Ar (Fig. 3l) or Ct.Th (Fig. 3m) in SCD mice.

Effect of GBT1118 on mineralized trabeculae in SCD female and male mice

We examined by histology the effect of GBT1118 on mineralized femur trabeculae in SCD mice of both sexes. Von Kossa staining of femurs from Control and SCD female mice (Fig. 4a) and male mice (Fig. 4b) showed that markedly decreased mineralized trabeculae in SCD-Vehicle-treated mice was rescued by GBT1118.

Figure 4.

Effect of GBT1118 on bone mineralization of femurs from female and male SCD mice. Von Kossa staining of femurs from Control and SCD mice showed that decreased mineralized trabeculae in SCD-Vehicle treated mice vs. Control-Vehicle was rescued by GBT1118 in female (a) and male (b).

Effect of GBT1118 on dynamic histomorphometric parameters of bone formation in SCD female and male mice

Analysis of femur metaphysis of female mice (Fig. 5a–e) showed that there was decreased double labeling distance in SCD-Vehicle which was increased by GBT1118 (Fig. 5a). As shown in Fig. 5b–e, there was significantly decreased mineralization surface/bone surface (MS/BS), mineral apposition rate (MAR), bone formation rate (BFR/BS) and inter-label thickness (Ir.LTh) in SCD-Vehicle group vs. Control-Vehicle. GBT1118 significantly rescued MAR, and Ir.L.Th but not MS/BS and BFR.

Figure 5.

Effect of GBT1118 on dynamic histomorphometry parameters of femurs from female and male SCD mice. Representative image of calcein/xlylenol orange double labeling in female (a) and male (f) mice. Analysis of mineralization surface/bone surface (MS/BS) in female (b) and (g) mice, mineral apposition rate (MAR) in female (c) and male (h) mice, bone formation rate (BFR/BS) in female (d) and male (i) mice, and inter-label thickness (Ir.LTh) in female (e) and male (j) mice. Data are shown as box-and-whisker plots, showing all individual data points. *p < 0.05 n = 10–12 mice/group.

Dynamic histomorphometry of femur metaphysis of male mice is shown in (Fig. 5f–j). Similar to female SCD, there was decreased double labeling distance in SCD-vehicle which was increased by GBT1118 (Fig. 5f). As shown in (Fig. 5g) there was no significant difference in MS/BS among groups. There was significantly decreased MAR (Fig. 5h), BFR/BS (Fig. 5i) and Ir.LTh. (Fig. 5j) in SCD-Vehicle group vs. Control-Vehicle, which were significantly increased by GBT1118.

Effect of GBT1118 on histomorphometric parameters of osteoclast in femur of SCD female and male mice

Since bone loss in SCD could be due in part to increased bone turnover²², we assessed the effect of GBT1118 on osteoclast parameters in SCD female and male mice. TRAP staining was performed on femur metaphyseal bones. As shown in Fig. 6a, in female mice, compared with Control-Vehicle, there were more TRAP positive osteoclasts in SCD-Vehicle treated mice which was decreased by GBT1118. Static histomorphometry showed that osteoclast surface/bone surface (Oc.S/BS) (Fig. 6b) and osteoclast number/bone perimeter (N.Oc/B.Pm) (Fig. 6c) were significantly higher in SCD-Vehicle group vs. Control-Vehicle and were significantly reduced by GBT1118. Although the osteoclasts number/total area (N.Oc/T.Ar) was similar in Control-Vehicle versus SCD-Vehicle, GBT1118 significantly decreased this parameter in SCD female mice (Fig. 6d).

Figure 6.

Effect of GBT1118 on TRAP staining and static histomorphometry of femurs from female and male SCD mice. Representative TRAP-stained femurs of female (a) and males (e) revealed markedly increased osteoclasts in SCD-Vehicle that were reduced by GBT1118. Static bone histomorphometry analysis of femur on metaphysis of osteoclast surface/bone surface (Oc.S/BS) in female (b) and males (f); osteoclast number/bone perimeter (N.Oc/Pm) in females (c) and males (g); osteoclast number/total area (N.Oc/T.Ar) in females (d) and males (h). Data are shown as box-and-whisker plots, showing all individual data points. *p < 0.05 n = 10–12 mice/group.

Compared with Control-Vehicle, there were more TRAP positive osteoclasts in SCD-Vehicle treated male mice, which was decreased by GBT1118 (Fig. 6e). Static histomorphometry showed that Oc.S/BS (Fig. 6f) and N.Oc/B.Pm (Fig. 6g) were significantly higher in SCD-Vehicle group vs. Control-Vehicle and were significantly reduced by GBT1118. Although N.Oc/T.Ar was increased in SCD-Vehicle, it was not significantly different from Control-Vehicle. However, GBT1118 significantly decreased this parameter in SCD mice (Fig. 6h).

Effect of GBT1118 on histomorphometric osteocyte markers in SCD female and male mice

Since we observed impaired osteoblast differentiation and function in both Vehicle-treated SCD female and male mice, we also examined by histomorphometry whether there were changes in osteocyte differentiation markers, numbers, and expression. In female SCD-Vehicle, there was reduced staining for early osteocyte marker E11/gp38 (Fig. 7a), and quantitative analysis demonstrated a significant reduction with no increase by GBT1118. There was no significant difference in differentiated osteocyte marker sclerostin (Fig. 7b) among groups. Since osteocytes resides in bone lacunae, hematoxylin staining was performed to examine osteocyte lacunae (Fig. 7c) which revealed markedly increased empty lacunae in SCD-Vehicle compared with Control-vehicle, and quantitative analysis revealed a significant increase in empty lacunae that was not altered by GBT1118. TUNEL staining for apoptotic osteocytes (Fig. 7d) revealed markedly increased apoptotic osteocytes in SCD which by quantitation was not significant and was not modulated by GBT1118. Similar to female SCD, in male SCD-Vehicle, there was markedly reduced staining for E11 (Fig. 7e) and sclerostin/SOST (Fig. 7f) that were significantly reduced and were not increased by GBT1118. Hematoxylin staining revealed no significant difference in empty osteocyte lacunae (Fig. 7g) among groups. TUNEL staining (Fig. 7h) demonstrated significantly increased apoptotic osteocytes in SCD-vehicle compared with Control-vehicle and was not modulated by GBT1118.

Figure 7.

Effect of GBT1118 on osteocyte in femoral cortical bone of SCD mice. Immunohistochemistry and quantitation on early osteocyte marker E11/gp38 in female (a) and male (e), as well as differentiated osteocyte marker sclerostin/SOST in female (b) and male (f). Hematoxylin staining and quantification for empty osteocyte lacuna in female (c) and male (g). Tunel staining and quantification for apoptotic osteocytes in female (d) and male (h). Data are shown as box-and-whisker plots, showing all individual data points. *p < 0.05 n = 10–12 mice/group.

Effect of GBT1118 on serum biomarkers in SCD female and male mice

We examined whether serum markers of liver, kidney and bone function were altered in Control-Vehicle, SCD-Vehicle and SCD-GBT1118 treated female and male mice (Supplemental Table S3). In female mice, there was no difference in serum BUN, creatinine, calcium, phosphate among the three groups. Vitamin D deficiency may contribute to bone loss in SCD², therefore 25(OH)D was measured and was significantly decreased in SCD-Vehicle and was not changed by GBT1118. There was no significant difference in serum 1,25(OH)2D among the three groups. There was no difference in serum PTH among the three groups. Serum bone formation markers Osteocalcin and P1NP were significantly reduced in SCD-Vehicle and were not increased by GBT118. Bone resorption marker CTX-1 was similar in all three groups. Since liver dysfunction is found in SCD¹ we also examined whether liver enzymes reflective of liver function was altered in the three groups of mice by sex, genotype, or GBT1118 treatment. There was no significant difference in serum liver enzymes AST among the three groups. Serum ALT was significantly increased in SCD-Vehicle and was further significantly increased by GBT1118.

Analysis of serum from male mice revealed no differences in BUN, creatinine, or phosphate among the three groups. Although serum calcium was similar between Control-Vehicle and SCD-Vehicle, it was significantly increased by GBT1118 in SCD male mice. Serum 25(OH)D was also significantly increased in SCD by GBT1118. Significantly increased 1,25(OH)2D in SCD-Vehicle was significantly reduced by GBT1118. There was no difference in serum PTH, serum Osteocalcin, P1NP and CTX among the three groups. In contrast to SCD-female mice, serum AST was significantly increased in SCD-Vehicle treated male mice but was not further increased by GBT1118. There was not significant difference in serum ALT between Ctrl-Vehicle and SCD-Vehicle, GBT1118 significantly increased ALT in SCD mice.

Effect of GBT1118 on bone related gene expression in SCD female and male mice

Since we observed impaired bone formation and increased bone resorption parameters in SCD mice of both sexes, we examined the effect of GBT1118 on bone related gene expression in whole tibia from SCD female and male mice. As shown in Supplemental Table S4, in SCD-Vehicle versus Control-Vehicle treated bones from female mice the mRNAs for osteoblast-related-genes, Runx2, Osterix, Col1a1, alkaline phosphatase, liver/bone/kidney (Alpl, TNAP); mineralization related genes, Ank, Enpp1, and Opn were significantly increased while Mepe was significantly decreased. While GBT1118 markedly reduced Col1a1(p = 0.058), TNAP (p = 0.132), only Osterix was significantly decreased by GBT1118. GBT1118 did not alter Mepe level in SCD mice. Although Bsp, Dmp1 and osteocyte related genes E11 and Sost mRNA were not significantly different between Control and SCD-Vehicle treated, GBT1118 caused a significant decrease in Bsp, Dmp1 and Sost but not E11/gp38. Also shown in Supplemental Table S4 relative to Control-Vehicle we observed in SCD-Vehicle significantly increased mRNAs for osteoclast-related-genes, Trap, Rank, Nfatc1, Slc40a1, Ppargc1b, antioxidant system related gene Prx2 and inflammatory and hypoxia related genes Hif2a, and Hif1a, however only Hif2a was significantly decreased by GBT1118. As shown in Supplemental Table S4, the Rankl/Opg ratio Ctsk were similar among all three groups.

As shown in Supplemental Table S5 gene analysis revealed that in SCD-Vehicle treated tibiae from male mice the mRNA for Osterix, Tnap, Bsp, Dmp1, Mgp, Mepe, Sost and E11/gp38 were significantly decreased relative to Control-Vehicle and were not increased by GBT118. In male mice, Rank and Rankl/Opg mRNAs were similar among all three groups and were not modulated by GBT1118. Similar to female SCD-Vehicle treated mice Prx2 mRNA was significantly increased and was markedly but not significantly reduced (p = 0.093) by GBT1118. While Slc4a1mRNA was significantly increased in SCD-Vehicle, its expression was not decreased by GBT1118.

Discussion

Although a variety of contributing factors to bone loss in SCD are known²³, the mechanism of bone loss has not been fully elucidated²⁴. It is well known that VOC leads to bone ischemia that exacerbates local bone loss¹. We therefore tested the efficacy of GBT1118 an analog of the FDA approved drug voxelotor that inhibits the polymerization of the mutated HbS on bone loss in Townes SCD mice that have been shown to recapitulate human SCD symptoms including bone loss^22,25–27.

In this study, we measured bone BMD and BMC in 6-months-old SCD female and male mice and like our previous studies²⁵, we confirmed a significant reduction in both parameters in femur and tibiae of SCD female mice and demonstrated a significant reduction in BMD and BMC in SCD male mice which was not examined in our earlier study²⁵. However, there was no effect of GBT1118 to increase BMD or BMC in SCD mice of either sex. We also examined vertebral BMD and BMC however, although significant vertebral decrease in vertebral BMD has been reported in humans with SCD²⁸, we did not observe a significant decrease in vertebral BMD in SCD mice of either sex.

However, the Micro-CT studies of the structural parameters of the femur compartments of the metaphysis, epiphysis in the Vehicle and GBT1118 treated SCD mice is noteworthy and do not necessarily correlate with the DXA results. There were no significant differences in metaphyseal parameters in SCD of either sex based on genotype which is consistent with our previous report in 6-months-old female mice²⁵ and 5-month-old male mice as reported previously²⁶. Nevertheless, GBT1118 significantly increased Tb.Th, TMD and decreased Tb.Sp in female SCD but not male SCD.

In contrast to the metaphyseal analysis in SCD mice, we observed a significant reduction in Tb.Th and TMD in SCD femur epiphyses of both sexes and while GBT1118 did not significantly increase either parameter in SCD female, it significantly increased TMD in SCD-male. Interestingly, in contrast to SCD female femur epiphyseal BV/TV in male SCD was significantly reduced and significantly increased by GBT1118.

Sex differences in diaphyseal cortical parameters and the effectiveness of GBT1118 were also observed in SCD-Vehicle treated versus SCD-GBT1118 treated mice as shown in Fig. 3. Of note, although CT.Ar/Tt.Ar was significantly reduced in SCD of both sexes, GBT1118 only significantly increased this parameter in female SCD mice. The reason for the differential effect of GBT1118 on femur structural parameters and compartments based on sex is unclear.

Dynamic histomorphometry was performed to assess whether SCD resulted in impaired bone formation due to reduced osteoblast function reflecting impaired osteoblast number, osteoblast activity or impaired mineralization and whether GBT1118 could modulate bone formation due to effects on these parameters in SCD mice of both sexes. Analysis of femur metaphysis was consistent in that we observed significant decrease in parameters of bone formation in SCD mice of both sexes. These data are consistent with our previous report of reduced osteoblast function in SCD mice²⁵. However, in this study we report for the first time impaired osteocytogenesis in SCD mice of both sexes that could reflect not only impaired terminal differentiation of osteoctyes but also increased osteocyte apoptosis.

Although GBT1118 increased dynamic histomorphometric parameters of bone formation in SCD mice of both sex, serum biomarkers of bone formation, osteocalcin and PINP that were significantly reduced in SCD female but not male were not increased by GBT1118. Previous studies have shown a relationship between vitamin D deficiency and bone fragility in sickle cell disease²⁹. However, although significantly reduced serum 25(OH)D was observed in SCD female, GBT1118 did not significantly modulate this parameter in female SCD mice. In contrast, although serum 25(OH)D was similar in Control-Vehicle and SCD-Vehicle treated male mice, its level was significantly increased in SCD-GBT1118 treated male mice. Interestingly, significantly increased serum 1–25(OH)2D was also observed in SCD-Vehicle treated male mice, that was significantly reduced by GBT1118.

We examined the effect of GBT1118 on liver enzymes since impaired liver function known to occur in SCD¹ could alter levels of serum 25(OH)D since conversion of vitamin D to 25(OH)D requires 25 hydroxylase enzyme that is produced by the liver. Specifically, we observed in SCD-Vehicle increased liver enzymes that was further increased by GBT1118. However, we do not believe that impaired liver functions modulating vitamin D levels explains impaired bone formation in SCD mice.

With reference to the further increase in liver enzymes by GBT1118, studies by Global Blood Therapeutics (unpublished) have shown that GBT1118 inhibits several cytochrome P450 (CYP) enzymes which could compromise liver health. However, although GBT1118 is an analog of Voxelotor, there is no evidence that Voxelotor has the same CYP liabilities of GBT1118 and to date there are no published studies^30–32, of transaminitis in patients treated with Voxelotor.

Bone loss observed in SCD mice of both sexes may be caused not only by impaired osteoblast differentiation and function resulting in reduced bone formation but also increased osteoclast function and bone resorption²². In the present study histomorphometry demonstrated significantly increased osteoclast number and activity and GBT1118 mediated reduction in osteoclast numbers and activity in SCD mice of both sexes is highly significant. Interestingly, like the studies of²², and consistent with our previous report in female SCD mice²⁵, we did not observe any differences in serum CTX-1 among the 3 groups by sex or genotype or treatment conditions.

We also examined osteocytes that are known to communicate with osteoclasts and osteoblasts, that also functions as endocrine cells to maintain phosphate homeostasis and sensors of mechanical stimulation to coordinate with effector cells to modulate bone mass, size, and shape based on mechanical demands³³. Our histomorphometry data showed impaired osteocyte differentiation and maturation in SCD mice as demonstrated by significantly reduced osteocyte markers including E11/gp38 a marker of early osteocytes, and late osteocyte maturation marker SOST/sclerostin associated with reduced osteocyte number as shown by significantly increased empty osteocyte lacuna. Thus, impaired osteocytogenesis could contribute to reduced bone mass in SCD mice.

The pattern of significant upregulation or down regulation of osteoblast, osteoclast and osteocyte-related gene expression in female and male SCD-Vehicle treated mice was not consistent as shown in supplemental Tables S4 and S5 nor was the effect of GBT1118 to modulate changes in gene expression in SCD of either sex. Based on the reports of²² in which VOC in SCD bone disease was associated with increased bone turnover, osteoclast activity as determined by increased Rank and Cathepsin-K mRNA, and osteoclast recruitment reflected by increased Rank mRNA and upregulation of IL-6 mRNA, we were particularly interested in the expression of these genes. Although the sex of the SCD mice examined in the studies by Carbonare was not stated²², in this study we observed a significant increase in the mRNA for Rank but not Rankl or IL-6 in bones of SCD-Vehicle-treated female mice while in SCD-Vehicle-treated male mice, we observed a significant reduction in Rankl and no change in Rank, Ctsk or IL-6 mRNA. Whereas the mRNA expression of osterix, Bsp, Dmp1, Sost was modulated in GBT1118 treated female SCD, there was no effect of GBT1118 on genes examined in bones of SCD male mice. The mechanism of the differentially effects on gene expression in female versus SCD male mice is unclear.

The effect of GBT1118, to rescue the impaired bone formation and increased bone resorption in SCD of both sexes was striking although the mechanism is not clear. Of potential relevance, studies in SCD mice by¹⁹ suggests that GBT1118, improved markers of kidney damage by preserving sickle red blood cell (RBC) health at least in part through prevention of RBC membrane hypoxia-induced damage. Invitro studies³⁴ on hypoxia-induced lethal and sub-hemolytic RBC membrane damage using RBC Mechanical Fragility, a metric of existing membrane damage and prospective hemolysis showed that GBT1118 prevents hypoxia-induced membrane damage in sickled RBC, by mechanisms not associated with compound induced changes in hemoglobin-oxygen affinity. Other studies by Dufu et al.³⁵ showed that treatment with GBT1118 increased hemoglobin and hematocrit in SCD mice. Specifically, their studies showed that GBT1118 increased total Hb by increasing RBC half-life and reducing sickling.

Regarding the role of hypoxia in SCD bone disease and based on the above reports on the ability of GBT1118 alleviate hypoxia, we examined whether antioxidant related gene Prx2 and hypoxia related genes were altered in SCD and the effect of GBT1118. We observed significant increases in Prx2 mRNA in bones of both female and male SCD and there was a marked but not significant reduction in their expression by GBT1118 in SCD of both sexes. Other Hypoxia stimulated genes such as Hif1a and Hif2a were significantly increased in female SCD mice and GBT1118 significantly reduced Hif2a mRNA in female SCD mouse bones. The significant reduction in Hif2a mRNA in SCD female is noteworthy since HIF2a is known to increase osteoclast differentiation³⁶, and recent studies suggest an important role for Hif2a signaling in the skeletal system³⁷. In contrast to SCD females we observed no significant differences in Hif2a in bones of male Control-Vehicle versus SCD-Vehicle mice and Hif1a was significantly reduced in SCD-Vehicle treated male mice and was not affected by GBT1118.

Based on the preponderance of data in this study, we postulate that GBT1118, a sickle hemoglobin polymerization inhibitor, indirectly induced increased bone formation in SCD mice. This increase in bone formation may be related to its ability to reduce hypoxia including reducing bone tissue hypoxia resulting in improved osteoblast differentiation and function and reduced osteoclast-mediated bone resorption.

Author contributions

MMH conceived and coordinated the study. MMH designed the study and wrote the paper. WH performed and analyzed the experiments shown. LX provided technical assistance and revised the manuscript. All authors reviewed the results and approved the final version of the manuscript.

Data availability

All datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Competing interests

The authors declare no competing interests.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-69589-9.