|Home | About | Journals | Submit | Contact Us | Français|
Over 10 million Americans have osteoporosis, and is the predominant cause of fractures in the elderly. Treatment of fractures in the setting of osteoporosis is complicated by a suboptimal bone regenerative response due to a decline in the number of osteoblasts, their function, and survival. Consequently, an osteogenic therapeutic to prevent and treat fractures in patients with osteoporosis is needed. Nel-like molecule-1 (NELL-1), a novel osteoinductive growth factor, has been shown to promote bone regeneration. In this study, we aim to demonstrate the capacity of recombinant NELL-1 to prevent ovariectomy (OVX)-induced osteoporosis in a senile rat model. Ten-month-old female Sprague-Dawley rats underwent either sham surgery or OVX. Subsequently, 50μL of 600μg/mL NELL-1 lyophilized onto a 0–50-μm tricalcium phosphate (TCP) carrier was injected into the femoral bone marrow cavity while phosphate-buffered saline (PBS) control was injected into the contralateral femur. Our microcomputed tomography results showed that OVX+PBS/TCP control femurs showed a continuous decrease in the bone volume (BV) and bone mineral density (BMD) from 2 to 8 weeks post-OVX. In contrast, OVX+NELL-1/TCP femurs showed resistance to OVX-induced bone resorption showing BV and BMD levels similar to that of SHAM femurs at 8 weeks post-OVX. Histology showed increased endosteal-woven bone, as well as decreased adipocytes in the bone marrow of NELL-1-treated femurs compared to control. NELL-1-treated femurs also showed increased immunostaining for bone differentiation markers osteopontin and osteocalcin. These findings were validated in vitro, in which addition of NELL-1 in OVX bone marrow stem cells resulted in increased osteogenic differentiation. Thus, NELL-1 effectively enhances in situ osteogenesis in the bone marrow, making it potentially useful in the prevention and treatment of osteoporotic fractures.
Although frequently unrecognized until fractures occur, osteoporosis is the predominant cause of bone fractures in the elderly. It is estimated that more than 10 million Americans have osteoporosis; one in two Caucasian women and one in five men are expected to experience an osteoporosis-related fracture in the course of a lifetime.1 Treatment and prevention of such fractures are complicated by suboptimal bone regenerative response due to osteoporosis. With the aging global population, the healthcare cost of treating osteoporosis-related fractures is expected to double or triple within the next four decades.2,3 Consequently, there is an increasing need for improved osteogenic therapeutics to treat and/or prevent bone fractures in patients with osteoporosis.
Osteoporosis is a condition characterized by decreased bone mass and microarchitectural deterioration of bone tissue.4,5 It is generally divided into two types—rapid loss of bone mass in postmenopausal osteoporosis due to estrogen deficiency, or the more gradual-onset senile osteoporosis seen in men and women that occurs with aging.6 The underlying biologic conditions in patients with osteoporosis may include not only an increase in bone resorption due to changes in the microenvironment as in postmenopausal women, but also a decline in bone marrow stem cell (BMSC) content as seen with aging.7 In addition, because osteoblasts and adipocytes are derived from the same BMSCs, age-related increased adipogenesis in the bone marrow leads to decreased osteoblastogenesis.8 Therefore, there is a decrease in the number of osteoblasts, and findings have also shown a decrease in their function and survival.5 For these reasons, the biologic responses to even the commonly used bone substitutes are suboptimal in such patients, in terms of efficacy and efficiency of bone regeneration and the frequency and magnitude of unwanted side effects.9
In terms of prevention therapy, parathyroid hormone (PTH) is the sole anabolic therapeutic approved by the Food and Drug Administration (FDA) for osteoporosis treatment, and has been shown to increase the BMSC population postirradiation. PTH, however, is anabolic only when given intermittently, and its use over 2 years has been shown to cause an increase in the development of bone neoplasms in rats.10 Thus, PTH is limited to only once in a lifetime use and only for a limited duration to temporarily reverse osteopenia, soon after which the osteopenic/osteoporotic condition returns.11 A commonly used antiresorptive agent, bisphosphonate, inhibits osteoclast activity in patients with osteoporosis to prevent further bone loss. However, systemic administration of bisphosphonate is associated with adverse effects, including bowel inflammation and erosion of the esophagus when taken orally; possible osteonecrosis of the jaw after high-dose intravenous administration in patients with cancer; severe bone, joint, or musculoskeletal pain; and fluctuation in calcium blood levels that may increase risk of cardiovascular events.12 In addition, bisphosphonate is an anticatabolic agent only, and is not capable of regenerating bone. Therefore, an improved, alternative agent that would help prevent osteoporosis-related fractures without exhibiting unwanted systemic effects would be of tremendous clinical advantage.
Bone morphogenetic protein 2 (BMP2), the most widely used FDA-approved osteoinductive growth factor, is a nonspecific growth factor that contributes to various growth and developmental processes in the body. This functional heterogeneity of BMP2 is known to contribute to clinical complications such as ectopic bone formation13 and promotion of adipogenesis, leading to cystic bone voids14–16 that compromise the quality of regenerated bone. Moreover, increased complications have been reported with use of BMP2 in patients with osteoporosis, leading experts to suggest avoidance of its use in those with osteoporotic bone disease.9
Nel-like molecule-1 (NELL-1), a novel osteoinductive growth factor originally identified in patients with craniosynostosis, has previously been shown to be effective in bone regeneration in various in vitro and in vivo studies.17–20 It has been shown to be osteoblast specific, and importantly, is able to suppress the side effects of cystic bone formation seen in high-dose usage of BMP2.16 In our preliminary in vivo study using an ovariectomy (OVX)-induced osteoporotic mouse model, NELL-1 also demonstrated successful regeneration of bone when locally injected into femurs of 6-month-old mice (data not shown). These findings suggest a role for NELL-1 for bone regeneration in osteoporotic conditions. [Note: NELL-1 and NELL-1 indicate the human gene and protein, respectively; Nell-1 and Nell-1 indicate the animal gene and protein, respectively.]
In this study, we aim to demonstrate the capacity of NELL-1 to prevent OVX-induced osteoporosis in a senile rat model. We hypothesize that NELL-1, when injected locally into the bone marrow space of rat femurs, would effectively potentiate the osteogenic capacity of endogenous BMSCs and endosteal osteoblasts to prevent bone loss.
Twenty-four 10-month-old female Sprague-Dawley rats (Charles River Laboratories) were used. Ten-month-old animals were used, as studies have shown that rats of this strain reach peak bone mass by 9 months; therefore, little-to-no bone modeling occurs that would confound our findings.21 Animals were randomly assigned to the following groups:
Whole femurs were harvested for high-resolution microcomputed tomography (microCT), histology, and immunohistochemistry analyses after euthanasia. Rat uteri were also harvested and weighed wet and dry as confirmation of OVX as previously described.22,23
β-TCP of 10–50-μm particle size was used as a carrier to deliver and allow sustained release of NELL-1 as previously described.24 The injectates consisted of 50μL for each femur of 100mg/mL β-TCP with or without 600μg/mL of recombinant human (rh) NELL-1. The optimal dose of NELL-1 was determined based on our previous intramedullary injection study on healthy rat models (manuscript in preparation).
Animals were housed in a light- and temperature-controlled environment and given food and water ad libitum. All animals were handled in accordance with guidelines of the UCLA Chancellor's Animal Research Committee of the Office for Protection of Research Subjects.
Sprague-Dawley rats were anesthetized by isoflurane inhalation, and buprenorphine injection was administered before surgical incision. For OVX, a 5-mm dorsal incision was made on the flank of each rat between the bottom of the rib cage and the front of the hind limb. The tip of a double-sharp iridectomy scissors was inserted through the muscle layer, and the muscle fibers were separated. The ovary was pulled through the incision with a pair of blunt forceps by grasping the fat pad surrounding it. High-temperature surgical cautery (Bovie Medical) was used to resect the ovary and oviduct, and the fat pad was replaced into the body cavity after confirming hemostasis. The skin incision was closed using 4-0 vicryl sutures. The procedure was repeated on the contralateral side.
For sham surgery, the fat pad was pulled through the incision with a blunt forceps following the same steps as above. The fat pad was replaced into the body cavity without resection of the ovary and oviduct. The skin incision was closed using 4-0 vicryl sutures. The procedure was repeated on the contralateral side.
Knee arthrotomy and intramedullary injection of the femur were performed immediately after OVX. Briefly, a 5-mm-longitudinal incision was made along the medial aspect of the quadriceps–patellar complex. The patella was dislocated laterally to expose the intercondylar groove. A 0.9-mm k-wire on a trephine drill was used to create a trephination defect. A 26-gauge needle was inserted through the defect, and 50μL of NELL-1/TCP injectate was injected into the intramedullary cavity. The trephination was sealed with bone wax after injection. The quadriceps–patellar complex was then repositioned, and the medial arthrotomy was carefully repaired with 5-0 vicryl sutures. About 50μL of PBS/TCP as control was injected into the contralateral femur, following the same surgical procedure as above.
Posteroanterior radiographs were taken preoperation and at harvest using a high-resolution Faxitron LX-60 Cabinet radiography system at a resolution of 10 lp/mm.
Postharvest, whole femurs were scanned using high-resolution microCT (Skyscan 1172; Skyscan) at an image resolution of 27.4μm (50kV and 201 mA radiation source, using a 0.5-mm aluminum filter), and analyzed using DataViewer, Recon, CTAn, and CTVol software provided by the manufacturer. The quantitative and structural parameters follow the nomenclatures described by the American Society for Bone and Mineral Research Nomenclature Committee.25
Femurs were analyzed in three separate regions—proximal femur (PF), mid-shaft (MS), and distal femur (DF)—to comply with published guidelines and to assess site-specific responses to OVX and treatment.26,27 For analysis, well-established quantitative and structural microCT parameters, including bone mineral density (BMD), percent bone volume (BV/TV), trabecular number, thickness, and separation (Tb.N, Tb.Th, and Tb.S), were measured.
Femurs were first reoriented to an exact upright position in three planes of space (coronal, sagittal, and transaxial) using DataViewer software. Then, the highest point of the growth plate on the vertical axis was noted and used as the baseline level from which to divide the femurs into the three regions. Along the vertical axis, the secondary spongiosa at 1mm proximal to the growth plate was set as the lower (distal) border of the DF to avoid inclusion of newly formed bone in the rats.26 From this level, DF is designated as the lower 6.4mm, MS as the middle 8.2mm, and PF as the top 6.4mm of the femur segment analyzed. These dimensions were determined by careful analysis of the anatomy of all femur samples and calculation of the average number of slices that could be universally applied to all samples.
In examining the scanned images of whole femurs to standardize the method of analysis, transaxial microCT images of the right and left femurs of each animal were generated at various levels along the vertical axis (Supplementary Fig. S1A; Supplementary Data are available online at www.liebertpub.com/tea). This was to determine the relative expansion or resorption of the periosteal and endosteal bone between the OVX+PBS/TCP control and OVX+NELL-1/TCP femurs, and also to examine the degree of variance between the right and left femurs of the SHAM and OVX groups. The newly formed bone was eliminated by differential thresholding for clearer assessment. The right femur in SHAM and OVX, as well as the OVX+NELL-1/TCP femurs at 2-, 4-, and 8-week groups were colored red, mirrored horizontally, and superimposed on the contralateral side in green using Adobe Photoshop CS4 (Supplementary Fig. S1B). This is similar to the previously published method of cortical bone analysis in the OVX rat model by Jee et al., where the bones were analyzed in three distinct layers: inner, middle, and outer cortical zones.26
For analysis of trabecular bone, regions of interests (ROIs) were drawn to discreetly select trabecular bone only, maintaining ~0.25mm clearance from the endosteal bone surface. For analysis of cortical bone, ROIs were drawn to fit exactly the outer border on the periosteal surface and the inner border on the endosteal surface.
To determine the optimal threshold, two representative femur pairs (based on radiographic findings) were selected from each group. The threshold for trabecular bone was set to 95–255 to match the normal bone structure in the SHAM and OVX+PBS/TCP control femurs and also partly include less-mature intramedullary bone. The threshold for cortical bone was set to 150–255 to match the normal bone structure in the SHAM and OVX+PBS/TCP control femurs while excluding the trabecular components, and was reassessed in other experimental groups. Three-dimensional images were generated to confirm that the optimized thresholds properly represent the bones. The selected thresholds were then applied to all samples in the experiment.
After microCT analysis, the specimens were decalcified using Cal-Ex solution (Fisher Scientific) for 5 to 7 days, washed with running tap water for 3 to 4h, and then transferred to 75% ethanol solution. The specimens were then processed for paraffin embedding, and 5-μm-sagittal sections of each specimen were obtained and stained with hematoxylin and eosin (H&E).
For immunohistochemistry, paraffin slices were deparaffinized, dehydrated, rinsed, and incubated with 3% H2O2 for 20min and then blocked with 0.1% bovine serum albumin in PBS for 1h. Primary antibodies, including anti-osteopontin (OPN) and anti-osteocalcin (OCN; Santa Cruz Biotechnology), at a dilution of 1:100 were added to each section and incubated at 37°C for 1h and at 4°C overnight. ABC complex (Vector Laboratories) was applied to the sections after the incubation with biotinylated secondary antibody (Dako Corporation). AEC plus substrate in red color (Dako Corporation) was used as a chromagen, and the sections were counterstained with light hematoxylin. PBS substituted for the primary antibody was used as a negative control. Photomicrographs were acquired using an Olympus BX51 microscope and a MicroFire digital microscope camera with PictureFrame software (Optronics).
Relative intramarrow adiposity was quantified using Adobe Photoshop. Relative lipid area was quantified using the Magic Wand tool, normalized to total marrow space within 20× magnification histological photomicrographs of the femurs.
Primary rat bone marrow mesenchymal stem cells (BMSCs) were obtained from the femoral marrow cavity of Sprague-Dawley rats. BMSCs were obtained by marrow flush either 4 weeks after OVX or from age-matched controls. Passage-2 cells only were used for experiments. Cells were expanded in the Dulbecco's modified Eagle's medium at 37°C and 5% CO2 as previously described.28 For in vitro osteogenic differentiation, OVX and control BMSCs were seeded at equal densities (1.3×104 cells per cm2). After attachment, the medium was replaced with an osteogenic differentiation medium (ODM). The medium was supplemented with either (rh)NELL-1 at 300ng/mL29–33 or control PBS. The ODM was replaced every third day. Cells were differentiated for 15 and 24 days for alkaline phosphatase (ALP) and alizarin red (AR) staining, respectively. Cells were fixed with 10% formalin for 10min, and ALP staining (indicating intermediate osteogenic differentiation) and AR staining (indicating terminal osteogenic differentiation) were performed as previously described.34 Photographic quantification was performed, assessing the relative intensity of staining by averaging 3 random microscopic fields using the Magic Wand Tool in Adobe Photoshop.
For statistical analysis, Student's t-test was performed paired to compare the PBS/TCP control and NELL-1/TCP femurs within the same treatment group (*p<0.05 and **p<0.01), and unpaired to compare two different groups (#p<0.05 and ##p<0.01).
Uteri were weighted 1h (wet weight) and 12h (dry weight) postharvest. There was a statistically significant decline in uterine weight of OVX samples versus SHAM, providing a surrogate marker for estrogen loss and osteoporosis induction (Supplementary Fig. S2).
At 2-, 4-, and 8-week time-points, OVX+NELL-1/TCP femurs had significantly increased trabecular bone compared to OVX+PBS/TCP control femurs that showed progressive loss of bone in the PF and MS regions as shown by three-dimensional reconstructions in Figure 1. Quantitative analysis of femurs showed that BV/TV, BMD, and trabecular thickness and number of OVX+NELL-1/TCP femurs were comparable to that of the SHAM group and significantly increased when compared to the OVX+PBS/TCP control at 8 weeks in the PF and MS regions (*p<0.05, **p<0.01 compared to PBS/TCP control; #p<0.05, ##p<0.01 compared to SHAM) as shown in Figures 2 and and33.
Overall, the DF region showed minimal effects of OVX-induced osteoporosis, and also showed the least difference between PBS/TCP and NELL-1/TCP sides. The OVX+NELL-1/TCP femurs, however, did show maintenance of bone over time, whereas OVX+PBS/TCP control femurs showed continuous loss of bone from 2, 4, and 8 weeks in the DF.
H&E staining was performed on 2-, 4-, and 8-week treated femurs, and 8-week SHAM and OVX femurs. At 2 and 4 weeks, OVX+PBS/TCP and OVX+NELL-1/TCP samples showed similar levels of bone loss, mainly as thinning of trabecular structure (data not shown). By 8 weeks, OVX+PBS/TCP control femurs continued to show considerable reduction of lamellar bone, findings which were similar to that of the OVX group. Conversely, OVX+NELL-1/TCP femurs showed recovery of lamellar bone volume (BV), with very evident increase in endosteal-woven bone formation comparable to the SHAM group (Fig. 4A–H).
Immunohistochemistry showed significantly increased distribution and enhanced intensity of OPN (Fig. 4I–L) and OCN (Fig. 4M–P) immunostaining in the OVX+NELL-1/TCP-treated femurs than that of OVX+PBS/TCP- and OVX-alone controls. Moreover, the positive staining patterns of OPN and OCN appeared to be distinctly different; OPN was expressed predominantly along the active bone formation fronts, while the OCN stains relatively mature bone matrix across all the samples. These results indicate that NELL-1 promotes more active woven bone formation and/or remodeling similar to that observed in SHAM control samples.
In addition, relative intramarrow adiposity quantification of histological images of OVX+PBS/TCP and OVX+NELL-1/TCP femurs at 8 weeks showed statistically significant reduction in lipid area, and thus adipogenesis in NELL-1-treated femurs (Fig. 5).
We next sought to confirm these results in vitro, examining the effect of (rh)NELL-1 delivered on BMSCs extracted from healthy (4 weeks post-SHAM) or ovariectomized (4 weeks post-OVX) rats. Osteogenic differentiation was performed and assessed by ALP staining to assess early differentiation (Fig. 6A–D), and AR staining to assess terminal differentiation (bone nodule formation, Fig. 6E–H). Quantification was performed by photographic assessment (Fig. 6I, J). At baseline, OVX was observed to result in a reduction in osteogenic differentiation by all markers. Among control BMSCs, NELL-1 increased osteogenic differentiation by both ALP and AR staining and quantification. This was also observed among OVX BMSCs, where NELL-1 increased all markers of osteodifferentiation. These findings were consistent with our preliminary data using adenoviral transduction of control LacZ or Nell-1 (AdNell-1) to SHAM and OVX rat BMSCs (Supplementary Fig. S3).
In this study, we demonstrated the ability of NELL-1 to prevent osteoporotic bone loss by local intramarrow injection. In the setting of OVX, NELL-1 injection maintained appendicular bone mass, as assessed by radiologic, histologic, and immunohistochemical methods.
Our osteoporotic model was carefully created to limit confounding factors that may affect the findings in this study. At 9 months or older, Sprague-Dawley rats have reached peak bone mass, and no longer undergoes periosteal expansion from growth at the long diaphyses, and any cortical remodeling will take place only on the inner cortical layer/endosteal surface of bone.37,38 Of note, rats undergo menopause at 15 to 18 months of age.39,40 Therefore, 10-month-old senile rats were used to undergo OVX to induce osteoporosis.
At the cellular level, osteoporosis is characterized by increased bone turnover rate, with an imbalance toward osteoclast-mediated bone resorption. NELL-1's capacity to reverse osteoporosis may be due to its modulation of osteoblast, osteoclast, and adipocyte activity. Mechanistically, NELL-1 is a downstream mediator of Runt-related transcription factor 2 (Runx2), the key transcription factor regulating osteogenic differentiation. Runx2 regulates NELL-1 expression by binding to its promoter region.41 In turn, NELL-1 upregulates Runx2 bioactivity by enhancing its phosphorylation through transient activation of mitogen-activated protein kinase signaling,42 leading to promotion of bone formation. NELL-1 overexpression has also been shown to partially rescue Runx2 haploinsufficiency.43 In this study, we showed, in vitro and in vivo, NELL-1's capacity to promote osteogenesis.
With particular relevance to osteoporosis, multiple studies have found that BMSCs demonstrate defective or reduced osteogenesis.44–46 Li et al. reported that BMSCs obtained from 3- and 6-month-old rats showed decreased proliferative and osteogenic potential compared to control BMSCs.47 Our in vitro studies showed similar findings in which we showed reduced osteogenesis in OVX BMSCs as compared to control BMSCs. Although the exact mechanism of this decline is unknown, our studies showed that NELL-1 is able to reverse this effect. Addition of NELL-1 to OVX BMSCs leads to increased ALP and AR staining, to levels comparable to that of control BMSCs that received no NELL-1 treatment.
In vivo, NELL-1 protein is commonly delivered lyophilized onto TCP particles to increase its stability. Our previous publications have shown the release profile and confirmed the in vivo bioactivities of the NELL-1/TCP delivery system.48,49 In this study, we have found that although there was an initial decline in BV and BMD from 2 to 4 weeks post-OVX, NELL-1/TCP treatment resulted in an increase in bone levels similar to that of the SHAM control by 8 weeks. These findings were particularly significant in the proximal and mid-shaft regions of the femur. Overall, the DF showed less response to OVX as well as treatment and may be due to the proximity of the DF to the epiphyseal plate. Failure to show statistically significant differences of cortical bone between groups is not surprising, as previous studies have shown that osteoporosis induced solely by OVX will not produce significant cortical changes in rats until after 6–9 months post-OVX.35,36
Our previous studies have shown that NELL-1 stimulation of Runx2 may be through the canonical (β-catenin dependent) Wnt-signaling pathway, which is essential for differentiation of progenitor cells into osteoblasts.50,51 While the mechanisms by which NELL-1 activates canonical Wnt signaling have not been determined, our previous in vitro studies showed that NELL-1 significantly induced β-catenin nuclear accumulation in M2-10B4 cells (JBMR manuscript in revision). We further confirmed NELL-1's association with Wnt signaling by performing studies co-applying NELL-1 with Wnt inhibitors. Addition of DICKKOPF-1 (DKK-1), a Wnt antagonist, eliminated NELL-1-induced increase in Runx2 expression in M2-10B4 cells (JBMR manuscript in revision).
NELL-1 has also been shown to downregulate peroxisome proliferator-activated receptor-gamma (PPAR-γ), a transcription factor and a master regulator of adipogenesis. This, again, may be mediated through NELL-1 effects on Wnt signaling, as Wnt signaling is known to inhibit PPAR-γ, and subsequently, adipogenesis in multiple cell lines.52,53 Indeed, quantification of intramarrow adiposity of our histological results confirmed decreased adipocyte accumulation in the marrow of OVX+NELL-1/TCP femurs compared to all other groups.
Finally, several lines of evidence suggest that loss of NELL-1 function is in fact associated with the development of bone loss and osteoporosis. We previously examined and reported on the skeletal phenotype of mice deficient in NELL-1. In the study, loss of NELL-1 was produced by E-nitrosourea mutagenesis54; however, complete deficiency in NELL-1 was shown to be neonatal lethal. After this finding, we have then examined adult and aged mice heterozyotes for NELL-1 mutation induced the same way; preliminary results have demonstrated a clear osteopenic/osteoporotic phenotype that develops with age and is most prominent in the thoracic and lumbar spines. In ongoing studies, it appears that this is associated with reduced bone formation, and also inappropriately maintained bone resorption. Our newly discovered osteoporotic, NELL-1-deficient mouse model also has a human correlate. In a 2010 article, Karasik et al. evaluated over 430,000 single-nucleotide polymorphisms (SNPs) in over 2000 women in the Framingham Osteoporosis Study.55 NELL-1 was one of only 10 genes harboring an SNP found to be associated with reduced BMD in the femoral neck and lumbar spine. In aggregate, these observations span both animal research and clinical studies and suggest that the loss of NELL-1 function may in fact contribute to the natural development of osteoporosis.
In the context of the present study, these data suggest that NELL-1 may be not only a causative factor, but also a future therapeutic for osteoporotic bone loss. From this and previous studies, we know that NELL-1 increases local bone formation by increasing osteoblast differentiation and activity, but without a concomitant osteoclast response. These qualities make NELL-1 an appealing future therapy used for local application of NELL-1 in the prevention of fragility fractures in high fracture-risk patients. Further studies using higher mammals such as sheep that have similar cortical bone structure as humans, including the Haversian structures, are needed to evaluate NELL-1's effects on not only the trabecular, but also the cortical bone.
We demonstrated the ability of NELL-1 to successfully prevent the loss of BV, BMD, and trabecular microstructure in an OVX-induced osteoporotic senile rat model. Our in vitro and in vivo findings suggest NELL-1's capacity to harness and further enhance the osteogenic potential of endogenous BMSCs and endosteal osteoblasts in both healthy and osteoporotic bones.
The authors would like to thank the Translational Pathology Core Laboratory (TPCL) and Surgical Pathology divisions of the UCLA Department of Pathology and Laboratory Medicine for technical assistance with histology. This work was supported by the NIH/NIDCR (grants R21 DE0177711-01, DE01607-01, and R01 DE16107-05S1, ARRA Supplement), an UC Discovery Grant Bio07-10677, a T32 training fellowship (5T32DE007296-14) for Dr. A.W. James, and a CIRM Training Grant TG2-01169 for Dr. J. N. Zara.
Drs. X.Z, K.T, and C.S. are inventors of NELL-1-related patents. Drs. X.Z, K.T, and C.S are founders of Bone Biologics, Inc., which sublicenses NELL-1 patents from the UC Regents.