Author response: Contrasting effects of Ksr2, an obesity gene, on trabecular bone volume and bone marrow adiposity

Article Figures and data Abstract Editor's evaluation eLife digest Introduction Results Discussion Methods Data availability References Decision letter Author response Article and author information Metrics Abstract Pathological obesity and its complications are associated with an increased propensity for bone fractures. Humans with certain genetic polymorphisms at the kinase suppressor of ras2 (KSR2) locus develop severe early-onset obesity and type 2 diabetes. Both conditions are phenocopied in mice with Ksr2 deleted, but whether this affects bone health remains unknown. Here we studied the bones of global Ksr2 null mice and found that Ksr2 negatively regulates femoral, but not vertebral, bone mass in two genetic backgrounds, while the paralogous gene, Ksr1, was dispensable for bone homeostasis. Mechanistically, KSR2 regulates bone formation by influencing adipocyte differentiation at the expense of osteoblasts in the bone marrow. Compared with Ksr2’s known role as a regulator of feeding by its function in the hypothalamus, pair-feeding and osteoblast-specific conditional deletion of Ksr2 reveals that Ksr2 can regulate bone formation autonomously. Despite the gains in appendicular bone mass observed in the absence of Ksr2, bone strength, as well as fracture healing response, remains compromised in these mice. This study highlights the interrelationship between adiposity and bone health and provides mechanistic insights into how Ksr2, an adiposity and diabetic gene, regulates bone metabolism. Editor's evaluation This study represents an important advance in connecting bone biology and metabolic functions. It implicates Ksr2 as a key regulator of the switch between adipocytes and osteoblasts that arise from a common precursor. Besides being an actionable target for obesity and osteoporosis, the study reaffirms and provides mechanistic data relating to the human genetic findings on KSR2 variants in metabolic regulation. https://doi.org/10.7554/eLife.82810.sa0 Decision letter Reviews on Sciety eLife's review process eLife digest Our bones are living tissues which constantly reshape and renew themselves. This ability relies on stem cells present in the marrow cavity, which can mature into the various types of cells needed to produce new bone material, marrow fat, or other components. Obesity and associated conditions such as type 2 diabetes are often linked to harmful changes in the skeleton. In particular, these metabolic conditions are associated with weight-bearing bones becoming more prone to facture and healing poorly. Mice genetically modified to model obesity and diabetes could help researchers to study exactly how these conditions – and the genetic changes that underlie them – impact bone health. Gomez et al. aimed to address this question by focusing on KSR2, a gene involved in energy consumption and feeding behavior. Children who carry certain KSR2 mutations are prone to obesity and type 2 diabetes; mice lacking the gene also develop these conditions due to uncontrolled eating. Closely examining mutant mice in which Ksr2 had been deactivated in every cell revealed that the weight-bearing bones of these animals were also more likely to break, and the fractures then healed more slowly. This was the case even though these bones had higher mass and less marrow fat compared to healthy mice. Non-weight bearing bones (such as the spine) did not exhibit these changes. Further experiments revealed that, when expressed normally in the skeleton, Ksr2 skews the stem cell maturation process towards marrow fat cells instead of bone-creating cells. This suggests a new role for Ksr2, which therefore seems to independently regulate both feeding behavior and bone health. In addition, the work by Gomez et al. demonstrate that Ksr2 mutant mice could be a useful model to better understand how obesity and diabetes affect human bones, and to potentially develop new therapies. Introduction Obesity is a major public health problem in the United States, afflicting nearly 40% of adults, and has become a prevalent and destructive health disorder linked to some of the major metabolic diseases, including cardiovascular diseases, type 2 diabetes (T2D), and cancer (Devlin and Rosen, 2015; Pagnotti et al., 2019; Shanbhogue et al., 2016; Walsh and Vilaca, 2017). Although obesity may be considered beneficial to bone health, since increased body weight is associated with higher bone mineral density (BMD), the relationship between excess body fat and bone health is complex, given that obesity has been identified as a risk factor for certain fractures (Greco et al., 2015; Ma et al., 2018; Veldhuis-Vlug and Rosen, 2018). The increasing prevalence of obesity and T2D dictates the need for appropriate in vivo animal models to study the mechanisms of action of obesity and T2D on bone metabolism. The effect of obesity and T2D on bone is an active area of investigation. Studies with several animal models and approaches have contributed to our current understanding of this relationship. Mouse models, in particular, have provided invaluable information through controlled genetic, biochemical, cellular, and molecular approaches to understand the pathological relationship between excess body fat and bone fragility. Most diet-induced obesity studies have reported reduced BMD and trabecular bone mass (Bonnet et al., 2014; Doucette et al., 2015; Inzana et al., 2013; Scheller et al., 2016). By contrast, monogenetic models of obesity have provided a broader range of bone phenotypes, including no change, loss, or gain in bone mass or BMD (Ahn et al., 2006; Baldock et al., 2007; Braun et al., 2012; Steppan et al., 2000; Wang et al., 2007a). There are several explanations for the diversity in skeletal phenotypes in these models, including differences in expression of targeted genes in other tissues besides bone, as well as varied effects of endocrine factors produced in other affected tissues such as the brain, fat, and skeletal muscle. Nevertheless, monogenetic studies have informed the molecular underpinnings of feeding regulation at the hypothalamus, which has fortuitously led to the development of pharmaceuticals to treat a particular population of individuals genetically predisposed to diabetes (Yeo et al., 2021). Although the RANKL monoclonal antibody, denosumab, has been used to treat bone disorders in osteoporotic T2D patients with reduced BMD (Abe et al., 2019), whether these agents can also benefit the population with gains in BMD, which are paradoxically also fragile (Burghardt et al., 2010; Ma et al., 2012), remains to be investigated. Also, it is worthwhile to further identify/study animal models with genetic mutations that phenocopy the human condition to study these interventions. The advent of the genomic era has expanded the list of individual genes associated with obesity and T2D (Loos and Yeo, 2022), yet their effects on bone remain vastly understudied. Recently, the scaffold proteins kinase suppressor of ras (KSR1 and KSR2) were identified as mediators of energy consumption, utilization, and adipogenic regulation (Brommage et al., 2008; Costanzo-Garvey et al., 2009; Kortum et al., 2005; Pearce et al., 2013; Revelli et al., 2011). Although these two genes function as paralogs, we previously found that only Ksr2 knockout (KO) mice become obese and diabetic (Brommage et al., 2008; Costanzo-Garvey et al., 2009; Kortum et al., 2005; Pearce et al., 2013; Revelli et al., 2011), suggesting that these paralogs have non-redundant roles, although Ksr1 does have a role in adipogenesis (Kortum et al., 2005). Several mutations at the KSR2 loci in humans have been associated with severe early-onset obesity (Pearce et al., 2013), and studies in Ksr2 KO mice have revealed a centrally regulated mechanism by Ksr2 expression and function in the hypothalamus that results in hyperphagia, changes in metabolic rate, and consequently, obesity and T2D (Costanzo-Garvey et al., 2009; Guo et al., 2017; Henry et al., 2014; Pearce et al., 2013; Revelli et al., 2011). Although there are hundreds of mouse genes reported to lead to obesity when disrupted, Ksr2 gene deletion is associated with a profound obese phenotype and lethality at a young age (Brommage et al., 2008; Revelli et al., 2011). In this study, we set out to investigate whether the deletion of Ksr2, an obesity and T2D gene, bears any effect on bone health, and if so, to evaluate the mechanisms by which KSR2 affects bone metabolism. Our studies show that loss of KSR2 function increases long bone trabecular bone mass while reducing marrow adiposity and that KSR2 acts as a molecular switch that controls the differentiation of bone resident mesenchymal stem cells into osteoblast or adipocyte differentiation via an mTOR-dependent mechanism. Results Ksr2 negatively regulates femoral bone mass To evaluate whether deletion of Ksr2 affects skeletal morphology, femurs of Ksr2 KO (exon 13 deleted) and wild-type (WT) control mice in the C57BL/6J-Tyrc-Brd × 129SvEvBrd hybrid background (Figure 1A) were subjected to micro-computed tomography (microCT) scanning. Distal femoral metaphyseal bones of Ksr2 KO (Ksr2-/-) female mice exhibited increased trabecular bone mass at both 11 and 15 weeks of age (Figure 1B). Quantification of trabecular parameters at the distal femur secondary spongiosa shows that by 11 weeks bone volume fraction (BV/TV), and trabecular thickness (Tb.Th) were significantly increased in female KO mice (Figure 1C and G), while structure model index (SMI), a measure of rod to plate-like trabecular morphology (Hildebrand and Rüegsegger, 1997), was significantly reduced (Figure 1E), overall implying that structural morphological changes elicited by the absence of Ksr2 promote gains in bone mass. Although mean trabecular connectivity density (CONN.D) and trabecular number (Tb.N) were also increased in KO mice at 11 weeks (Figure 1D and F), these differences were not significant until 15 weeks, as was the reduction in trabecular spacing (Figure 1H). Figure 1 Download asset Open asset Ksr2 regulates bone mass in females. (A) Schematic of Ksr2 knocked out in the C57BL/6J-Tyrc-Brd × 129SvEvBrd hybrid strain with exon 13 deleted (X), and accompanying ventral view of genotyped mice at 4 months of age exhibiting differences in weight gain. (B) Representative 3D micro-computed tomography (microCT) reconstruction images of the secondary spongiosa at the distal femoral metaphysis in wild-type (Ksr2+/+, WT) or knockout (Ksr2-/-, KO) females at 11 and 15 weeks, revealing a prominent increase in trabecular bone in KOs. Scale bar: 100 μm. (C–H) MicroCT measurements from the trabecular bone as represented in panel (B) (n = 6–10/group), BV/TV, bone volume/tissue volume; CONN.D, connectivity density; SMI, structural model index; Tb.N, trabecular number; Tb.Th, trabecular thickness; Tb.Sp, trabecular spacing. (I) Representative 3D reconstruction of microCT images of primary spongiosa in WT or KO mice at 15 weeks of age revealing increased bone density in KO mice. Scale bar: 100 μm. (J–M) Quantification of microCT parameters measured in panel (I) (n = 6–10/group). (N) Representative 3D reconstruction of microCT images of cortical bone at the femoral mid-diaphysis (scale bar: 100 μM), where the TV total volume (O) is not affected, while BV/TV and volumetric bone mineral density (vBMD) (P, Q) are increased in KO mice. Statistics analyzed by unpaired two-tailed Student’s t-test, and graphed lines represent the mean ± SEM, *p<0.05, **p<0.005. Figure 1—source data 1 Micro-computed tomography (microCT) measurements of female trabecular bone. Source data for Figure 1C–H. https://cdn.elifesciences.org/articles/82810/elife-82810-fig1-data1-v1.xlsx Download elife-82810-fig1-data1-v1.xlsx Figure 1—source data 2 Micro-computed tomography (microCT) measurements of female primary spongiosa. Source data for Figure 1J–M. https://cdn.elifesciences.org/articles/82810/elife-82810-fig1-data2-v1.xlsx Download elife-82810-fig1-data2-v1.xlsx Figure 1—source data 3 Micro-computed tomography (microCT) measurements of female cortical bone. Source data for Figure 1O–Q. https://cdn.elifesciences.org/articles/82810/elife-82810-fig1-data3-v1.xlsx Download elife-82810-fig1-data3-v1.xlsx New woven bone is actively formed and mineralized at the primary spongiosa while the woven bone is remodeled into mechanically stronger lamellar bone at the secondary spongiosa. To determine whether new bone formation at the primary spongiosa is altered in the Ksr2 KO mice, we measured trabecular bone parameters at the primary spongiosa, limited to within 300 μm of the distal-most femoral metaphyseal bone from the growth plate (Figure 1I–M) and found significant increases in trabecular bone volume. Ksr2 deleted males also exhibited significantly greater BV/TV, Conn. Den, Tb.N, and Tb.Th but reduced Tb.Sp and SMI compared to littermate control mice (Figure 2A–G). Thus, loss of the Ksr2 gene promotes trabecular bone density in both genders of mice. Figure 2 Download asset Open asset Ksr2 negatively regulates femoral bone in males, while Ksr1 deletion does not affect trabecular bone in either gender. (A) Representative 3D micro-computed tomography (microCT) reconstruction images of the distal femoral metaphysis in wild-type (WT) or knockout (KO) male mice at 16 weeks of age revealing increased trabecular bone in KOs. Scale bar: 100 μm. (B–G) MicroCT measurements from the trabecular bone as represented in panel (A) (n = 5–9 mice per group). BV/TV, bone volume/tissue volume; CONN.D, connectivity density; SMI, structural model index; Tb.N, trabecular number; Tb.Th, trabecular thickness; Tb.Sp, trabecular spacing. (H) Representative 3D microCT reconstruction images of cortical bones at the femoral mid-diaphysis revealing that Ksr2 deletion does not affect TV total volume (I) of cortical bone in males, while BV/TV (J) and volumetric bone mineral density (vBMD) (K) are increased in Ksr2 KO mice (n = 5–6/group). Scale bar: 100 μm. (L–Q) Quantification of microCT data from the distal femoral metaphysis of WT and Ksr1 knockout mice at 16 weeks of age, showing minimal changes in trabecular bone parameters between genotypes in either gender. Statistics analyzed by two-tailed Student’s t-test, and graphed lines represent the mean ± SEM, *p<0.05, **p<0.005. Figure 2—source data 1 Micro-computed tomography (microCT) measurements of male trabecular bone. Source data for Figure 2B–G. https://cdn.elifesciences.org/articles/82810/elife-82810-fig2-data1-v1.xlsx Download elife-82810-fig2-data1-v1.xlsx Figure 2—source data 2 Micro-computed tomography (microCT) measurements of male cortical bone. Source data for Figure 2I–K. https://cdn.elifesciences.org/articles/82810/elife-82810-fig2-data2-v1.xlsx Download elife-82810-fig2-data2-v1.xlsx Figure 2—source data 3 Micro-computed tomography (microCT) measurements of Ksr1 knockout mice. Source data for Figure 2L–Q. https://cdn.elifesciences.org/articles/82810/elife-82810-fig2-data3-v1.xlsx Download elife-82810-fig2-data3-v1.xlsx By further characterization of femoral bones, we found increased mid-shaft femoral cortical bones in Ksr2 nulls. While total tissue volume, indicative of bone size, remained unchanged in the KO mice in both genders (Figures 1N–P–2H–K), a significant increase in bone volume fraction was observed, although gains in BMD were more prominent in females. Nonetheless, this evaluation reveals that obese Ksr2 null mice present increased gains in trabecular and cortical mass of femoral bones. Ksr1 is dispensable for the development of femoral trabecular bone mass Ksr1, the only paralog of Ksr2, is highly expressed in skeletal muscle (Costanzo-Garvey et al., 2009), which interacts with and affects bone physiology (Lara-Castillo and Johnson, 2020). Ksr1 was also expressed in osteoblasts (data not shown). To determine whether Ksr1 might also contribute to limb bone mass accretion, we evaluated metaphyseal femoral bones of 16-week-old Ksr1 KO mice and their wild-type littermate controls by microCT. In contrast with the striking differences found in trabecular bone parameters in Ksr2 KO mice at this age, trabecular bone measurements were nearly identical between Ksr1 KO and WT mice, irrespective of gender (Figure 2L and M). These results suggest that the KSR1 protein is highly unlikely to synergize with KSR2 in regulating femoral bone growth. Validation of bone phenotype by Ksr2 deletion in a different genetic background Since genetic background can influence biological effects in mice (Ackert-Bicknell et al., 2009; Bonnet et al., 2014), we evaluated whether Ksr2 deleted in the DBA/1LacJ strain, which also becomes obese (Costanzo-Garvey et al., 2009), might also exhibit alterations in skeletal phenotype. Deletion of exon 4 in this genetic background did not significantly affect the anus to nose body length, although a substantial gain in body weight and percent body fat was observed by 8 weeks of age (Figure 3A–D). Concordant with Ksr2 deletion in the C57BL/6J-Tyrc-Brd × 129SvEvBrd hybrid background, dual-energy X-ray absorptiometry (DXA) measurements of pooled genders revealed substantial increases in total body BMD (Figure 3E) and, in particular, in femoral BMD (Figure 3G), while femur length remained unchanged relative to WT siblings (Figure 3F). Additionally, no changes were observed in the vertebral trabecular bone (Figure 3—figure supplement 1). Figure 3 with 2 supplements see all Download asset Open asset Ksr2 deletion in a different genetic background, histomorphometry, and histology validates that Ksr2 negatively regulates bone formation. (A) Schematic of Ksr2 knocked out in the DBA/1LacJ strain with exon 4 deleted (X). (B) No differences were noted in body length at 8 weeks of age, while gains in body weight (C) and body fat percentage (D) are noted in knockouts (KOs). Bone mineral density (BMD) is increased in total body (E) and femurs (G) of KO mice, while femur length is not changed (F) (n = 7–12 mice/group; genders combined) (D–G reflect dual-energy X-ray absorptiometry measurements). (H) Representative alizarin red images at the distal femoral epiphysis show increased area of mineral staining in KO mice at 11 weeks of age. Scale bar: 100 μm. (I, J) Quantification of alizarin stain reveals an increase in bone area/total area (BA/TA) and a decrease in osteoid area/bone area (OA/BA). (K) Representative histomorphometric images of fluorescent calcein label reveal increased staining in KO mice. Scale bar: 100 μm. (L–N) Quantification of histomorphometric parameters measured, showing increased bone formation rate/bone surface (BFR/BS) and mineral apposition rate (MAR), yet no changes in the number of osteoclasts per bone surface (Oc.S/BS) (n = 4–7 mice/group). (O, P) Serum levels of bone formation marker (PINP) and bone resorption marker (Ctx-1) in 8-week-old female Ksr2 mutant and wild-type mice (n = 5–7 mice/group). (Q, R) Immunofluorescence staining at distal femoral metaphysis for (IBSP, synonym BSP2) or (SPP1, synonym OPN) (both red), counterstained with DAPI (cyan) reveals broader expression of both bone markers in KO mice; growth plate-osteoblast boundary positioned at the top. p.sp, primary spongiosa; ss, secondary spongiosa. Scale bar: 100 μm. (S) RT-qPCR reveals increased expression of osteoblast markers (Alpl, Bglap2, Spp1, and Sp7), while osteoclast markers (Acp5, Ctsk) remain unchanged in femurs of KO mice. Statistics analyzed by two-tailed Student’s t-test, and graphed lines represent the mean ± SEM, *p<0.05, **p<0.005. Figure 3—source data 1 X-ray measurements of Ksr2 knockouts in DBA/1LacJ. Source data for Figure 3B–G. https://cdn.elifesciences.org/articles/82810/elife-82810-fig3-data1-v1.xlsx Download elife-82810-fig3-data1-v1.xlsx Figure 3—source data 2 Distal femur–alizarin red quantification. Source data for Figure 3I and J. https://cdn.elifesciences.org/articles/82810/elife-82810-fig3-data2-v1.xlsx Download elife-82810-fig3-data2-v1.xlsx Figure 3—source data 3 Histomorphometric measurements of Ksr2 knockout mice. Source data for Figure 3L–N. https://cdn.elifesciences.org/articles/82810/elife-82810-fig3-data3-v1.xlsx Download elife-82810-fig3-data3-v1.xlsx Figure 3—source data 4 Serum ELISA measurements of Ksr2 knockout mice. Source data for Figure 3O and P. https://cdn.elifesciences.org/articles/82810/elife-82810-fig3-data4-v1.xlsx Download elife-82810-fig3-data4-v1.xlsx Figure 3—source data 5 RT-qPCR data of 12-week-old Ksr2 knockout femurs versus wild-type. Average fold changes plotted and t-test values are highlighted in yellow. Source data for Figure 3S. https://cdn.elifesciences.org/articles/82810/elife-82810-fig3-data5-v1.xlsx Download elife-82810-fig3-data5-v1.xlsx Histological evaluation of longitudinal distal metaphyseal femur bone sections by alizarin red also revealed increased amounts of calcified bone in Ksr2 KO mice as the ratio of bone area over total area was increased, while osteoid area over bone area was decreased (Figure 3H–J, Figure 3—figure supplement 2). Moreover, qualitative comparisons of bone markers, integrin bone sialoprotein (IBSP) and secreted phosphoprotein 1 (SPP1), by immunofluorescence suggest an increased areal expansion of both markers throughout the metaphysis in KO mice (Figure 3Q and R). Overall, this data provides further supporting evidence that Ksr2 negatively regulates appendicular bone formation, with confirmation in a different genetic background. Gains in bone mass in Ksr2 nulls are a product of increased osteoblast activity Next, we began to address how the deletion of Ksr2 results in increased bone mass. To determine whether increased bone formation is the cause of increased bone mass in Ksr2 KO mice, we performed histomorphometric analysis by pulsed calcein injections in 8-week-old mice. This resulted in increased calcein labeling in KO mice, with quantitative gains detected in bone formation rate and mineral apposition rate (Figure 3K–M). By contrast, the percentage of bone-resorbing acid phosphatase 5, tartrate-resistant (ACP5+) osteoclasts scored per bone surface, did not change (Figure 3N). Moreover, serum levels of osteoblast bone deposition (procollagen type 1 N-terminal propeptide [P1NP]) were elevated in KO mice, while those of osteoclast activity (carboxy-terminal cross-linked telopeptide of type 1 collagen [CTX1]) were not changed (Figure 3O and P). Bulk comparisons between osteoblast and osteoclast markers were then compared by real-time-quantitative PCR (RT-qPCR) from the metaphysis of Ksr2 KO relative to WT littermates at 12 weeks of age. mRNA expression levels of osteoblast markers, Alpl and Bglap2, showed increased, though insignificant, expression in KO mice, while Spp1 and Sp7 were significantly increased. By contrast, markers of differentiated osteoclasts, Acp5 and Ctsk, remained unchanged (Figure 3S). Combined, these results posit that Ksr2 affects osteoblast function, and does not apparently affect osteoclasts. Ksr2 gains bone at the expense of adipocyte differentiation As Ksr2 KO mice are obese, exhibiting increased visceral and subcutaneous adiposity (Figure 4A), we determined whether genetic disruption of Ksr2 influences adipocyte gene expression in white and brown adipose tissues in 28-week-old mice. mRNA levels of key transcription factors, Pparg and Cebpa, were unchanged in both fat depots in Ksr2 KO mice (Figure 4B) at this age. By contrast, the adipokine leptin (Lep) was increased, while complement factor D (Cfd) was decreased in both white and brown fat of Ksr2 KO mice (Figure 4C). This suggests that KSR2 exerts opposite effects on leptin and complement factor D/adipsin expression in fat tissues, which is consistent with changes observed in other models of obesity in mice (Cianflone et al., 2003; Kwon et al., 2012). Figure 4 Download asset Open asset Obesity in Ksr2 null mice paradoxically presents reduced bone marrow adiposity. (A) Representative image of mice at 28 weeks of age dissected to reveal differences in visceral adipose tissue (VAT) between wild-type (WT) and Ksr2 knockouts (KOs). (B, C) RT-qPCR assessing changes in regulators of adipogenesis (B) or adipokine genes (C), in white or brown fat of Ksr2 KO mice relative to WT at 28 weeks of age (n = 3–4/group). (D) Representative 3D micro-computed tomography (microCT) reconstruction images of osmium tetroxide-labeled femurs, revealing reductions in bone marrow adipose tissue in Ksr2 KO mice at 28 weeks of age. Scale bar: 1 mm. (E) Quantification of adipocyte volume (AV) occupied by marrow adipose tissue in femurs of mice as depicted in panel (D), at proximal (prox), middle (mid), and distal (dist) thirds of the femur with position defined in reference to the spinal cord (n = 6–8/group). (F) Representative hematoxylin and eosin-stained longitudinal distal femur sections of 8-week-old mice in which adipocytes (arrows) were compared at the secondary spongiosa, revealing reductions in KO mice. Scale bar: 100 μm. (G) Quantification of sections as represented in panel (F) (n = 5–7/group). RT-qPCR comparisons in adipogenic (H) and Wnt-related (I) genes from the secondary spongiosa of femurs as shown in (F) (n = 3–5/group). Statistics analyzed by two-tailed Student’s t-test, and lines plotted reflect the mean ± SEM, *p<0.05, **p<0.005, ***p<0.0005. Figure 4—source data 1 RT-qPCR data of 28-week-old Ksr2 knockout versus wild-type (adipose tissue). Average fold changes plotted and t-test values are highlighted in yellow. Source data for Figure 4B and C. https://cdn.elifesciences.org/articles/82810/elife-82810-fig4-data1-v1.xlsx Download elife-82810-fig4-data1-v1.xlsx Figure 4—source data 2 Quantification of osmium tetroxide-labeled micro-computed tomography (microCT) of tibia. Source data for Figure 4E. https://cdn.elifesciences.org/articles/82810/elife-82810-fig4-data2-v1.xlsx Download elife-82810-fig4-data2-v1.xlsx Figure 4—source data 3 Quantification of adipocytes from H&E-stained femurs metaphysis. Source data for Figure 4G. https://cdn.elifesciences.org/articles/82810/elife-82810-fig4-data3-v1.xlsx Download elife-82810-fig4-data3-v1.xlsx Figure 4—source data 4 RT-qPCR data of 12-week-old Ksr2 knockout versus wild-type femur (adipocyte markers). Average fold changes plotted and t-test values are highlighted in yellow. Source data for Figure 4H. https://cdn.elifesciences.org/articles/82810/elife-82810-fig4-data4-v1.xlsx Download elife-82810-fig4-data4-v1.xlsx Figure 4—source data 5 RT-qPCR data of 12-week-old Ksr2 knockout versus wild-type femur (Wnt markers). Average fold changes plotted and t-test values are highlighted in yellow. Source data for Figure 4I. https://cdn.elifesciences.org/articles/82810/elife-82810-fig4-data5-v1.xlsx Download elife-82810-fig4-data5-v1.xlsx Bone marrow stem/stromal cells (BMSCs) represent common precursors for adipocytes and osteoblasts, and marrow adipose tissue (MAT) volume is known to be inversely correlated with trabecular bone mass (Ko et al., 2021; Pierce et al., 2019; Tencerova et al., 2018; Yue et al., 2016). To determine whether MAT volume is affected in the Ksr2 KO mice, we evaluated the levels of osmium tetroxide retaining MAT in femurs of 28-week-old mice by microCT (Figure 4D). MAT volume was significantly reduced in all three compartments (proximal and distal metaphysis, and diaphysis) in the tibia of Ksr2 KO mice compared with controls (Figure 4E). Consistent with these data, there was a significant reduction in adipocytes in the distal femoral metaphysis of Ksr2 KO mice (Figure 4 F and G). To determine the potential regulatory molecules that contribute to changes in MAT in Ksr2 KO mice, we compared mRNA expression of adipocyte markers at the trabecular compartment in distal femurs of Ksr2 KO and control mice. Neither markers associated with white adipocytes (Fabp4, Slc7a10) or brown adipocytes (Ucp1, Prdm16) were different in Ksr2 KO femurs, while of the adipogenic regulatory transcription factors evaluated, Pparg was mildly but significantly reduced, and Cebpa was not changed. However, key adipokines (Lep, Cfd) were decreased (Figure 4H). Since Wnt signaling is critically involved in regulating adipocyte differentiation, we also measured mRNA levels of several Wnt-related genes, but only found a decrease in Wnt8b, and an increase in Ccnd1, a Wnt target gene, in the bones of Ksr2 KO mice (Figure 4I). Thus, changes in adipokine gene expression in both body and MAT adipocytes are altered when Ksr2 is deleted globally. Loss of Ksr2 delays femoral fracture healing and results in more fragile bones The pathological obese/T2D condition predisposes bones to compromised fracture healing and increased fracture risks. Since the absence of Ksr2 results in increased appendicular bone deposition, we evaluated whether this increased rate of bone formation would prove beneficial in Ksr2 KO mice. Healing response was compared between WT and KO mice at 16 weeks of age following stabilized closed femoral fractures (Figure 5A). X-ray analysis 3 weeks after fracture showed improved bony union of the callus in WT mice and increased callus size in KO mice (Figure 5B). Figure 5 Download asset Open asset Delayed fracture healing but increased fragility in Ksr2 knockout mice. (A) Schematic of strategy. (B) Representative X-ray images of bones that underwent closed mid-femoral fracture in wild-type (WT) and Ksr2-/- knockout (KO) mice on the day of surgery (day 0) and day 21 post-fracture (fx). Yellow arrows point to induced fracture, while calluses are outlined by dotted yellow lines. Scale bar: 1 mm. (C) Representative 3D micro-com