Pilot & Feasibility Awards

PI: Gaurav Swarnkar Ph.D.

IκBζ as a New Molecular Regulator of Osteoblast Inflammatory Phenotype

Specific Aims

Normal skeletal development requires a precise coordination among osteoblasts (OB), osteoclasts (OC), and osteocytes (OCy), with the latter indirectly influencing OB and OC1-4. However, under pathological conditions like chronic inflammation, aging, and metabolic disorders, the equilibrium of skeletal health is disrupted5-12. Inflammatory signals such as Tumor Necrosis Factor Alpha (TNFα) and Interleukin-1β (IL-1β) trigger bone breakdown through heightened OC-mediated resorption while simultaneously suppressing bone formation, resulting in net bone loss13-15. Nevertheless, the exact mechanisms through which inflammation impacts bone formation, as well as the differentiation and functions of OB and OCy, remain incompletely understood due to the redundancy of pro-inflammatory cytokines and a scarcity of pertinent genetic models16,17. This knowledge gap offers a distinctive opportunity to identify novel targets for interventions aimed at enhancing bone health.

The transcription factor Nuclear Factor kappa B (NF-κB) is recognized as a pivotal mediator of immune and inflammatory responses and plays a crucial role in skeletal development18-23. In this context, we have demonstrated that constitutive activation of IKK2 (IKK2ca) within the mesenchymal compartment using Collagen type II Cre recombinase (Col2-Cre) establishes unresolved inflammation, that disrupts the differentiation of OBs and adversely affects skeletal development 24. In the current proposal, preliminary findings using OB-specific Cre recombinases (Sp7 and Col1 Cre) to induce the expression of IKK2ca in mice have resulted in significant bone abnormalities characterized by impaired OB differentiation and increased expression of osteocyte markers. Using RNAseq and qPCR, we observed that primary calvarial OBs (cOB) expressing IKK2ca, or those exposed to low concentrations of the inflammatory cytokines IL1β or TNFα exhibit a significant reduction in genes associated with OB differentiation. Strikingly, these cells have adopted an osteocytic/dendritic morphology within 3-6 days and expressed bona fide osteocyte markers, such as Podoplanin (Pdpn/E11), Fibroblast Growth Factor 23 (Fgf23), Sclerostin (Sost), Dikkopf-1 (Dkk1), and TNF Receptor Superfamily Member 11a (Tnfsf11a, also known as Rankl). Additionally, these osteocyte-like (OCy-L) cells exhibited heightened expression of inflammatory and senescence markers, including Il1β, Tnfα, Il6, Cyclin Dependent Kinase Inhibitor 2A (Cdkn2/p16), p21, and Tumor Protein P53 (p53). This cellular profile is reminiscent of observations made in inflammatory and metabolic diseases, wherein key factors associated with OCy, particularly elevated levels of DKK1, SOST, RANKL, FGF23, and senescence-associated secretory phenotype (SASPs) were observed 25-37. Therefore, we designated these cells as atypical inflammatory OCy-like (aiOCy-L) cells. Our new concept presents a paradigm shift in understanding differentiation of OB under pathologic conditions, whereby inflammatory responses disrupt and truncates the normal OB differentiation process and decision-making, leading to their premature and accelerated transformation into aiOCy-L cells. These aiOCy-L cells fail to adequately integrate into the bone matrix and predominantly concentrate at the bone lining interface. The heightened expression of DKK1, SOST, and SASPs further hampers OB function. Concurrently, RANKL and SASPs stimulate the differentiation of OCs and bone loss.

From a mechanistic standpoint, our transcriptomic data has identified NF-ĸB Inhibitor Zeta (IĸBζ) as a distinctive inflammatory conduit of NF-ĸB. IĸBζ regulates the transcription of a specific subset of genes, including inflammatory cytokines and SASPs, exclusively during pathological conditions, rather than under normal physiological circumstances. This phenomenon is observed in various immune and bone cell types such as macrophages, chondrocytes, and osteoblasts38,39. In summary, our findings suggest that in the presence of inflammation, IĸBζ plays a role in accelerating the transformation of OBs into aiOCy-L cells. These cells were not reported previously and hence need further investigation to characterize their phenotype and function. Consequently, we hypothesize that inflammatory responses, acting through the NF-κB-IĸBζ axis, alter the OB differentiation program and fate decision, leading to their rapid transformation into aiOCy-L catabolic cells. In this research application, our primary focus is to elucidate the role of NF-κB→IĸBζ in the unconventional differentiation of OB to aiOCy-L cells. To address this, we propose the following specific aims:

Specific Aim 1. Determine the role of IĸBζ in accelerated transformation of OB to aiOCy-L cells and subsequent bone loss in animal model of chronic inflammation.

PI: Aaron Johnson Ph.D.

Mechanisms of Systemic Muscle Repair

Specific Aims

Duchenne Muscular Dystrophy (DMD) is the most common inherited muscle disease, and is characterized by chronic, systemic muscle injury1. However, muscle regeneration has largely been studied with acute, localized muscle injury models. Local injury models have been essential tools for characterizing the signaling pathways proximal to the stem cell niche that regulate the activation, proliferation, and differentiation of the Muscle Stem Cells (MuSCs) that direct muscle repair. Genetic mutations have been used to generate chronic systemic muscle injury models, which cause life-long cycles injury and repair 2, but since these models are not inducible it is
difficult to associate the induction of repair mechanisms with the onset of systemic muscle injury.
To identify systemic muscle repair mechanisms, we developed an inducible systemic injury platform in zebrafish. We used our systemic injury model for genetic screens and whole-organism single cell RNA sequencing (scSeq) to identify cellular and molecular pathways of systemic muscle repair. Our preliminary data support the hypothesis that local MuSC activation is only one facet of a broad organism-wide response that coordinates systemic muscle regeneration. To uncover additional mechanisms that regulate systemic muscle repair, we will pursue the following aims:

Specific Aim 1. Validate injury-induced changes in gene expression

Specific Aim 2. Map alleles that block systemic repair

PI: Rita Brookheart Ph.D.

Novel Aspects of Skeletal Muscle Regeneration

Specific Aims

Skeletal muscle regeneration is an essential process to restore muscle integrity after injury due to acute physical trauma or disease 1. Several diseases are associated with impaired skeletal muscle regeneration 2–5. Dysfunctional regeneration can result in decreased muscle integrity and strength 6–10, and in humans, is associated with immobility that can increase the risk of mortality 2,11–17. These observations highlight a need to understand the factors controlling normal muscle regeneration after injury and a need for therapeutic strategies to alleviate dysfunctional regeneration in human disease.

Satellite cells are adult myogenic stem cells that are required for normal muscle regeneration 6,7. In uninjured muscle, satellite cells are generally quiescent; however, in response to injury, they become activated, proliferate, and differentiate into myocytes that fuse with myofibers to restore muscle structure and strength 1. There are several gaps in our knowledge about the molecular mechanisms that regulate satellite cell function, such as (1) why are regulators of lipid/sterol metabolism inducted during satellite cell activation? and (2) why is the unfolded protein response (UPR) activated during muscle regeneration?

Site-1 Protease (S1P) is a Golgi-resident protease required for the activation of transcription factors that drive lipid/sterol metabolism and the UPR; these same transcription factors are strongly implicated in skeletal muscle regeneration 18–21. However, a mechanistic link between the activation of these transcription factors and satellite cell function during regeneration has not been established. Our preliminary data demonstrate that activation of the S1P substrates that drive lipid/sterol metabolism and the UPR is increased during muscle regeneration and that satellite cell-specific depletion of S1P impairs skeletal muscle regeneration following cardiotoxin-induced muscle injury.

These data implicate a function for S1P in satellite cell biology and muscle regeneration; however, much remains unclear about the potential function of S1P in this context – specifically (1) is S1P required for satellite cell differentiation? and (2) does S1P regulate muscle regeneration after injury?

Based on our preliminary data and expertise in S1P biology, we hypothesize that (1) S1P is required for the differentiation of satellite cells into myocytes and (2) S1P is required for muscle regeneration and restoration of muscle integrity and strength after injury. We will test these hypotheses through the Specific Aims below:

Specific Aim 1: Determine the role of S1P in satellite cell differentiation. Characterize satellite cell-specific S1P knockout (S1PSC-KO) mice and investigate myocyte differentiation in satellite cells from S1PSC-KO and floxed (control) littermates.

Specific Aim 2: Define the function of S1P in regeneration and restoration of muscle strength post-injury. S1PSC-KO and control mice will undergo cardiotoxin-induced muscle injury. Muscles will be harvested at designated time-points to track the stages of regeneration and restoration of muscle strength.

This proposal is designed to advance our understanding of how S1P impacts skeletal muscle regeneration and will enable us to generate the essential preliminary data for our upcoming R01 application.

PI: Cecilia Pascual-Garrido, MD, PhD

Establishing a Critical Interplay between Cartilage and Synovium through PPARy
Suppression, Chondrocyte Autophagy Dysregulation and Macrophage Polarization during Progression of Hip OA

Specific Aims

Idiopathic hip OA (osteoarthritis) has been historically overestimated. Mild hip malformations are observed in 90% of patients with hip OA, with Femoroacetabular Impingement (FAI) seen in 50% of these cases. Still, the molecular mechanism that leads to OA remains unknown. Over the past 4 years, we have focused on epigenetic regulation of OA susceptible genes during progression of hip OA. Using tissue cartilage samples from human hip FAI (early stage) and hip OA (late stage), we observed an altern expression pattern of DNMTs (DNMT3B vs. DNMT1/3A) resulting in abnormal suppression of OA susceptible genes. Specifically, we noticed suppression of PPARγ (Peroxisome proliferator-activated receptor gamma) secondary to hypermethylation of PPARγ promoter. PPARγ has been reported to be critical in cartilage homeostasis via mTOR signaling, critical in chondrocyte autophagy. It has been reported that mTOR activation enhances the proinflammatory macrophage in OA synovium.

This proposal is based on our key preliminary data showing that: i) there is an altered expression pattern of DNMTs (DNMT3B vs. DNMT1/3A) with progression of hip OA, ii) DNMT3A interacts with the PPARγ promoter (CpG island) resulting in suppression of this gene, iii) cartilage autophagy dysregulation is observed with OA progression and iv) we develop a small rabbit animal model of Hip OA secondary to FAI deformity that could provide, for the first time, a platform to study early and late mechanism of hip OA disease. Thus, our central hypothesis is that the inflammation associated with FAI results in altered DNA methylation (epigenetic change), suppression of PPARγ and induces chondrocyte autophagy dysregulation via PI3K/Akt/mTOR signaling resulting in macrophage pro-inflammatory polarization (M1).

Aim 1 (Human Tissue): Investigate the role of PPARγ in chondrocyte autophagy and synovial macrophage polarization during progression of human hip OA and confirm an interplay between cartilage and synovium. Rationale: i) PPARγ is suppressed in articular chondrocytes (ACs) during progression of hip OA6, ii) DNMT3A binds and methylates PPARγ CpG promoter area resulting in PPARγ suppression (unpublished data), iii) PPARγ has been reported to be critical in cartilage homeostasis via mTOR signaling, essential in chondrocyte autophagy3,4, 5 and iv) severe synovitis is observed in late-stage FAI disease.7 Aim 1a will investigate the dysregulation of chondrocyte autophagy in hip ACs (markers: LC3 and Beclin1) and synovial macrophages polarization during hip FAI OA progression with flow-cytometry and immunofluorescence. Aim 1b will investigate the therapeutic efficacy of PPARγ preservation via DNMT3A inhibitor on cartilage degeneration, chondrocyte autophagy, and synovial macrophages polarization, while exploring the paracrine interaction between synovial macrophages and ACs via PI3K/AKT/mTOR pathway with cartilage/synovium co-culture assay.

Aim 2 (Rabbit Animal Model): Investigate cartilage degeneration, epigenetic dysregulation, chondrocyte autophagy, and synovial macrophages polarization during rabbit hip FAI OA progression, while focusing on PPARγ pathway and PI3K/AKT/mTOR signaling. Rationale: i) we have established a hip FAI OA rabbit animal model 8. Aim 2a will test the hypothesis that hip ACs from the impingement zone show overexpression of catabolic markers, abnormal DNMTs expression, and suppression of PPARγ. Aim 2b will test the hypothesis that inflammation associated with FAI suppress PPARγ, dysregulates mTOR signaling and decrease cartilage autophagy markers and induces rabbit synovial macrophages polarization into a pro-inflammatory phenotype (M1).
The proposal will investigate the effect of PPARγ pathway suppression on cartilage autophagy and macrophages
polarization during progression of hip OA. Additional characterization of this pathway in the animal model will be
critical to test future interventional therapies modulating this pathway (R01 submission).

PI: Jennifer Zellers, PT, DPT, PhD

Human Diabetic Tendon Composition: A Data-driven Approach

Specific Aims

Diabetes impacts 34.2 million people in the United States.1 Individuals with diabetes are at 3x greater risk of tendon injury2 contributing to pain, loss of function, and impaired foot health initiating the cascade to foot ulceration and amputation.3–5 Despite the magnitude of this health problem, the effect of diabetes on tendon homeostasis and healing is not well-understood. Additionally, personalization of tendonspecific treatment is limited by a lack of biomarkers and adequate non-invasive tools to improve patienttreatment alignment and clinical decision-making.

The current theoretical framework is that glycation of diabetic tendon limits collagen sliding. 6–8 Impaired sliding alters mechanosensitive cellular function, including reduced transcription of pro-tenogenic factors (e.g., Mohawk). Combined with a hyperglycemic environment, this leads to chronic fibrosis, characterized by shifts in collagen content (increased type III collagen, decreased type I collagen) and alterations in the extracellular matrix (i.e. reductions in proteoglycans and elastin). However, no study to date has comprehensively characterized tendon composition in human patients with and without diabetes. Our central hypothesis is that presence of diabetes results in a characteristic tendon fibrotic signature, which is also affected by aging and high body mass index (aim 1). Further, transcription of tenogenic factors (Mohawk) will be reduced in diabetic tendons with regional dependence (aim 2). The findings of this study will inform potential biomarkers/treatment targets for diabetic tendon injury and non-invasive means to assess tendon structure to aid clinical decision-making. We will test our central hypothesis with the following aims:

Aim 1. Distinguish protein profile composition and degenerative changes in diabetic and non-diabetic tendons. Hypothesis. Alterations in tendon proteome will distinguish diabetic from non-diabetic tendons (H1.1). This is anticipated to include shifts in collagen content and reductions in elastin; however, we will take an unbiased, hypothesis generating approach. Diabetic tendons will also show increased degenerative changes compared to non-diabetic tendon on histology (H1.2). This aim will estimate strength of associations between tendon composition with patient characteristics known to influence tendon homeostasis and healing (age and body mass index). We hypothesize that tendon fibrosis will be associated with older age and higher body mass index (H1.3). Approach. Foot and ankle tendon specimens (posterior tibialis tendon) collected from individuals with (n=10) and without (n=10) diabetes undergoing lower extremity amputation will be assayed with proteomics. Principal component analysis and k-means cluster analysis will reduce tendon components to identify robust protein profile characteristics that differentiate groups. All tendons will be assessed with histology and scored using the Movin9 score to quantify magnitude of degeneration. Differences between diabetic and non-diabetic groups will be tested with one-way ANOVA. Relationships between tendon protein profile (defined as principal component 1 from proteomics analysis) and Movin score with patient characteristics (age, body mass index) will be tested with multivariable regression to identify if tendon structure is related to patient-level characteristics. These findings will comprehensively define the proteome shift in diabetic tendon and identify robust biomarkers/therapeutic targets to evaluate disease severity/restore tendon function in people with diabetes.

Aim 2. Compare spatial distribution of tenogenic transcription leveraging spatial transcriptomics (RNAscope). Hypothesis. Mohawk is a tenogenic marker that has been previously suggested to be downregulated in the presence of hyperglycemia,10,11 and could have implications for the ability for the tendon to maintain homeostasis. Mohawk expression will be more pronounced in the deep compared to superficial portion of the tendon in all donors (H2.1), with expression in diabetic tendons lower than in non-diabetic tendons (H2.2). Approach. Tendon specimens will be assessed with spatial transcriptomics probing Mohawk expression. Mohawk expression will be visualized and quantified. Regional differences (superficial to deep) in Mohawk expression will be examined a using mixed effects model. Should critical proteins distinguishing diabetic and non-diabetic tendons be identified in Aim 1, upstream transcription for these proteins will be interrogated in addition to the Mohawk probe included in Aim 2.

The findings of this study will provide a foundation for future studies competitive for R-level funding. Plans for future R-level proposals building on this work include taking a data-driven approach to understanding diabetesrelated tendon complications. Potential treatment targets identified in Aim 1 requiring additional, focused investigation can be queried in future studies. Aim 2 will provide insights into tenogenic gene expression that will improve our understanding of regional differences in capacity for tendon to maintain homeostasis. These are steps toward the ultimate goal of this research line in promoting person-centered care by using patient and tissue-specific characteristics to inform clinical decision-making, including prescription of biologics and tendon loading treatment to maintain homeostasis and optimize tendon healing. This proposal leverages the Musculoskeletal Research Core for histology and supports the expansion of Histology Core services to include spatial transcriptomics in tendon tissue (and, more broadly, to collagenous, soft tissues).

PI: M. Farooq Rai, PhD

Novel Role of Kif26b in Cartilage Tissue Engineering

Specific Aims

Current treatments for repairing cartilage defects rely heavily on surgical intervention to restore articular surface, but the long-term outcomes are often unpredictable. Therefore, the discovery of novel mechanisms and factors that promote cartilage regeneration is clearly needed. The field of cartilage tissue engineering, which aims to repair, regenerate, and/or improve injured or diseased cartilage functionality, has great promise for improving cartilage therapy but is still evolving. Several efforts have been made to identify factors that promote chondrogenic differentiation, inhibit hypertrophy and protect constructs from inflammation with only limited success thus far. Among others, two factors have emerged as critical for the success of tissue engineering approaches: selection of a promising stem cell source; and the identification of factors that not only help improve chondrogenesis but also provide protection of implanted constructs from the inflammatory environment within the diseased joint. Adult stem cells have a number of limitations that preclude their use in cartilage engineering such as significant donor to donor variability, poor homing and retention, uncontrolled differentiation in the absence of cues, limited therapeutic responses and serious adverse effects e.g., tumor formation. Recently, there is an increasing interest in the use of induced pluripotent stem cells (iPSCs) as a promising cell source for cartilage tissue engineering due to their chondrogenic potential, high homogeneity, negligible immune rejection, and great ability to differentiate into chondrocytes following chondrogenic induction. However, the factors that improve chondrogenic potential in these cells remain elusive. Relevant to this, we have identified kinesin family member 26b (Kif26b), a member of the kinesin superfamily of proteins, as a candidate gene whose knockdown may improve cartilage regeneration. Specifically, we have shown that Kif26b knockdown promotes chondrogenesis of progenitor/stromal cells by inhibiting the canonical Wnt/b-catenin pathway, although mechanistic details have yet to be determined. As the tissue-engineered constructs designed to treat cartilage defects or osteoarthritic lesions are implanted in a diseased joint, they are very vulnerable to the aggressive inflammatory environment within the joint. Therefore, there is a need for biological factors that can restore and maintain functions of engineered cartilage constructs in an inflammatory joint environment. Wnt/b-catenin signaling is also implicated in arthritis-related inflammation with conflicting results as some studies demonstrate that Wnt/b-catenin inhibition exacerbates osteoarthritis, while its activation improves chondrocyte matrix synthesis and lubricin production. A large body of work illustrates that activation of Wnt/b-catenin increases chondrocyte hypertrophy, elevates MMP13 expression, and increases the susceptibility to osteoarthritis. Altogether, these findings suggest that Wnt/b-catenin activation is detrimental to articular cartilage, possibly by inducing inflammation. Given that Kif26b controls canonical Wnt/b-catenin, we posit that Kif26b knockdown will protect chondrogenic pellets from inflammatory conditions, as expected in osteoarthritis. Therefore, we hypothesize that Kif26b loss-of-function will: (a) enhance chondrogenic potential of iPSCs by inhibiting the Wnt/b-catenin pathway, and (b) protect chondrocyte pellets from IL-1-induced inflammation. To address these hypotheses, we will assess the mechanistic role of Kif26b in promoting chondrogenic potential of murine iPSCs (Aim 1) and investigate the effect of Kif26b loss-of-function on chondrocyte constructs under inflammation (Aim 2). By performing this study, we will gain new knowledge about the mechanistic role of Kif26b in cartilage tissue engineering and will provide novel therapeutic targets for improving cartilage regeneration.

PI: Erica Scheller, PhD

Characterization and Function of a New p75-NTR+ Cellular Network in Bone.

Specific Aims

During embryogenesis, vertebrates develop a fold on the neural plate where the neural and epidermal ectoderms meet; this is called the neural crest. This fold is the origin of neural crest cells that contribute to tissues and organs throughout the body. Approximately 5% of the adult human skeleton forms from the neural crest while the rest develops from the mesoderm. Neural crest-derived bones include most of the bones of the skull and portions of the clavicle and scapula. Neural crest cells also give rise to portions of the nervous system, melanocytes, the thyroid gland, and the adrenal medulla. A study in 2010 discovered that the regenerative potential of neural-crest derived bone is 5- to 7-fold higher than adjacent mesodermderived
bone and, furthermore, that this difference is persistent even in adulthood. However, the fundamental molecular and cellular mechanisms underlying this result remain unclear. This represents a key gap in knowledge that has important implications for development of novel treatments to accelerate skeletal regeneration throughout the body.
We recently identified a unique p75 neurotrophin receptor-positive (p75-NTR+) periosteal cell population that forms a dense network specifically over the surface of the neural crest-derived bone in adult animals. This network is absent in bones of mesodermal origin. P75-NTR is a 427-amino-acid transmembrane receptor that forms heteromeric complexes with other receptors and binds to a broad array of neurotrophins. Due to its plasticity, the function of p75-NTR is versatile, context-dependent, and poorly defined in vivo. P75-NTR has also been identified as a marker of neural-crest derived stem cells. Our central hypothesis is that this robust p75-NTR+ cellular network contributes directly to the superior regenerative properties of neural crest-derived bone, either through endogenous stem/progenitor activity or by providing pro-regenerative signals to surrounding cell populations. The work in this pilot application will begin to test this hypothesis while generating preliminary data and models in support of a full R01 application.

Aim 1: Characterize the p75-NTR+ cells.

Aim 2: Assess the functional role of p75-NTR in the regenerative response.

Aim 3: Generate the p75-NTR-CreERT2 model for future lineage tracing and genetic studies.

PI: Silvia Jansen, PhD

Determining the Role of Plastin-3 in Osteoblast Differentiation and Mineralization.

Specific Aims

Osteoporosis debilitates millions of people each year and is a major health care burden. The characteristic brittle bone phenotype of osteoporosis is the result of imbalances between osteoblasts that deposit bone, osteoclasts that remove bone, and osteocytes that translate mechanical load into chemical signals that regulate
osteoblast and osteoclast function. To date, osteoporosis is mainly treated by curbing bone resorption by osteoclasts, however the severe side effects and concerns about the long-term efficacy of these compounds call for novel targets to treat osteoporosis. Moreover, healthy bone maintenance requires new bone formation, and thus there is increasing interest in proteins that can stimulate osteoblast function. The calcium-sensitive actinbundling protein, Plastin-3 (PLS3, also called Fimbrin), has emerged as a promising new target to regulate bone formation. This is supported by its predominant expression in osteoblasts and osteocytes, as well as by the evergrowing list of mutant PLS3 variants in children with early onset X-linked osteoporosis. Additionally, evidence is now mounting that PLS3 expression is also changed in adult patients diagnosed with osteoporosis. In line with this, whole genome PLS3 knockout mice and zebrafish show reduced bone mass, however the underlying mechanisms have not been characterized, including whether this requires the actin-regulatory function of PLS3.
Thus, the goal of this proposal is to determine how PLS3 mechanistically contributes to osteoblast mineralization, as well as to establish the relationship between PLS3 mutation and osteoporosis.

Aim 1: Elucidate how PLS3 mechanistically controls osteoblast mineralization by regulating focal adhesion activity.

Aim 2: Investigate the role of PLS3 mutation in bone formation and maintenance.

PI: Alexander Chamessian, MD, PhD

A Cellular and Molecular Census of Musculoskeletal Nociceptors.

Specific Aims

Chronic pain poses a tremendous burden on society, affecting more than 116 million Americans and costing up to $635 billion annually1. The causes of chronic pain are numerous and varied, differing in location, etiology and pathophysiology2–4. Musculoskeletal sources account for a substantial proportion of cases of chronic pain, with an estimated global prevalence of 30%, weighted heavily toward older populations5–7. Pain commonly arises from the activation of nociceptors, which are primary somatosensory neurons (PSNs) that are specialized to preferentially detect noxious stimuli8. Nociceptors innervate tissues across the body including skin, mucosa, viscera, and musculoskeletal tissues (muscle, joint, bone, fascia, ligament, and tendons) and have functional properties that differ based on their target innervation9–20. Despite the wide-ranging distribution of nociceptors, the preponderance of basic and translational research has focused on cutaneous nociceptors, about which now much is known12. In contrast, much less is known about the distinct attributes of nociceptors innervating pain-relevant deep structures such as muscle, joint, bone, tendon and fascia. Recent large-scale single-cell RNA-sequencing (scRNA-seq) of PSNs of the dorsal root ganglia (DRG) in mouse and human have demonstrated molecular and cellular heterogeneity, with at least a dozen broad neuronal types identified21–26. A key limitation of these studies is that they combine all PSNs from multiple DRG levels, irrespective of their target innervation, and thus do not have the ability to identify target-specific subtypes and their features. Moreover, given the predominance of skin-projecting PSNs in the DRG27, these atlases most likely undersample PSNs projecting to deep tissues28. Based on the body of evidence showing neurophysiological, morphological and developmental differences between deep and cutaneous nociceptors11,13–15,29–33, it is highly probable that transcriptomic heterogeneity exists amongst nociceptors innervating different musculoskeletal targets. However, to date, a comprehensive single-cell atlas of musculoskeletal-projecting nociceptors is lacking. If the molecular features of nociceptors projecting to musculoskeletal tissues were known, this information could be exploited to develop projection-specific and mechanism-based therapies for
musculoskeletal pain. Therefore, there is a critical need to determine the molecular and cellular heterogeneity of musculoskeletal-projecting nociceptors in a comprehensive manner to advance basic science and drug development related to musculoskeletal pain.

Our long-term goal is to elucidate the mechanisms of musculoskeletal pain. Our overall objective in this proposal is to identify the transcriptomic signatures of nociceptors innervating musculoskeletal tissues. Our central hypothesis is that musculoskeletal tissues are innervated by subtypes of nociceptor that are molecularly and functionally distinct from those that project elsewhere (e.g. skin, viscera). Our rationale for this proposal is two-fold. First, delineating the subtypes of musculoskeletal-projecting nociceptors and their gene expression profiles will provide the necessary intellectual foundation from which to select candidate targets for the development of novel treatments for musculoskeletal pain. Second, determining the subtypes of
musculoskeletal-projecting nociceptors will provide a roadmap for numerous follow-on studies that evaluate the role and function of these subtypes in diverse models of musculoskeletal pathology. To attain our overall objective, we propose the following specific aim:

Aim 1: Determine the Cellular and Molecular Diversity of Nociceptors Innervating Musculoskeletal Tissues.

PI: Spencer Lake, PhD

Machine Learning Approach for Disease Evaluation of Elbow Post-Traumatic Contracture and Osteoarthritis.

Specific Aims

Elbow trauma causes the development of post-traumatic joint contracture (PTJC) and post-traumatic osteoarthritis (PTOA) in ~50% of afflicted joints. While the pathogenesis remains unclear, both conditions have clinical manifestations of capsule fibrosis and cartilage damage that contribute to pain, stiffness, and loss of motion of the elbow. Therapies aiming to prevent these conditions or restore pre-injury elbow function are often unsuccessful and/or associated with complications. To study the pathogenesis of, and to test therapies for, elbow PTJC and PTOA, our group developed the first elbow-specific animal injury model. This model has been used to test the efficacy of various treatments including physical and drug therapies; unfortunately, no strategy to date has resulted in much improvement in elbow motion. A need exists to elucidate cellular mechanisms governing the injury response of elbow soft-tissues that impact motion in our animal elbow injury model.
Histopathology is a gold-standard method to study and evaluate therapies for musculoskeletal diseases preclinically with cellular resolution. Despite serving a critical role in establishing and using our animal model, histopathology has notable limitations. For instance, due to their invasive and manual nature, analysis protocols occur post-mortem and are laborious and time-consuming. Furthermore, protocols require a pathologist’s evaluation using semi-quantitative scoring metrics on two-dimensional slides, which are inherently subjective, prone to human error/bias, and spatially restricted. Recently, the emergence of digital histopathology (DH) has reduced the need for direct pathologist analysis and enabled quantitative evaluation; yet, analyses are still time-consuming and semi-automatic. Image segmentation techniques aimed at fully automating DH have limited accuracy due to various cellular and tissue morphologies and histological artifacts. Importantly, no method exists to link the two-dimensional cellular detail of DH with the three-dimensional tissue level detail of non-
invasive imaging, such as micro computed-tomography (μCT) and magnetic resonance imaging (MRI). Establishing such a link may unlock non-invasive insights into cellular mechanisms of elbow PTJC and PTOA, which could yield clinically relevant image-based biomarkers to assist treatment decision-making and prognosis. Novel approaches are warranted to automate and objectify evaluating post-trauma elbow conditions across multiple length scales and modalities.
Machine learning (ML), a subset of artificial intelligence, may revolutionize the study and treatment of musculoskeletal conditions by accelerating and automating the pipeline of evaluation and uncovering spatial heterogeneity in tissue patterns and cell types. ML has recently been used to assess disease in other musculoskeletal joints; however, the use of ML for elbow conditions is limited, particularly in preclinical models. Furthermore, musculoskeletal studies using ML have not yet consolidated data from multiple modalities (e.g., DH, μCT, MRI) to explore spatially dependent and predictive relationships. We hypothesize that ML, combined with DH and non-invasive imaging, can automatically predict elbow pathology and predict DH features based on non-invasive imaging features. We have partnered with Dr. Ulugbek Kamilov, an ML expert in Computer Science at WashU, to assist in implementing a transformative ML workflow to study and treat post-trauma elbow conditions. Thus, this pilot study aims to develop tissue- and modality-specific ML algorithms for automatic evaluation of disease stage and explore multi-modal/level relationships in our animal elbow injury model.

Aim 1: Develop ML algorithms to automatically quantify DH features of elbow soft-tissues and predict disease state.

Aim 2: Develop ML algorithms to explore feature relationships between DH, μCT, and MRI of the elbow

PI: Conor McClenaghan, PhD

Investigations of musculoskeletal pathology in KATP channelopathies

Specific Aims

KCNJ8 and ABCC9 encode the Kir6.1 and SUR2 subunits of ATP-sensitive potassium (KATP) channels, hetero-octameric nucleotide-gated ion channels which function to couple cellular metabolism to electric signaling and are expressed to varying extents in a diverse range of musculoskeletal tissues, including skeletal muscle, chondrocytes, and osteoblasts1-6. Gain-of-function (GoF) and loss-of-function (LoF) mutations in KCNJ8 and ABCC9 result in Cantu Syndrome (CS) and ABCC9-related intellectual disability and myopathy syndrome (AIMS), respectively, each of which include multiple unexplained muscular and skeletal pathologies. In this project I propose to build on my preliminary studies to develop and exploit an armory of animal and cell-based models to address these unexplained pathologies.

Aim 1: To generate genetically-faithful models of ABCC9-related intellectual disability and myopathy syndrome.

Aim 2: To determine the musculoskeletal effects of KATP gain- and loss-of-function mutations in mice

PI: Feini Qu, PhD

Mechanisms of Skeletal Morphogenesis During Digit Tip Regeneration

Specific Aims

Limb loss resulting from disease or trauma affects an estimated 2 million Americans, significantly affecting quality of life. While some animals can regenerate complex body structures, even into adulthood, humans (and most mammals) have limited regenerative potential of their musculoskeletal tissues. To this end, successful attempts to regrow missing digits or limbs could significantly improve the prognosis for amputees. Using the murine digit, we and others showed that local stem/progenitor cells, known as the ‘blastema,’ naturally regenerates digit tissues after distal amputation, but fail to do so after amputation at more proximal levels. The
mechanisms regulating regeneration versus fibrotic scarring are currently unknown and may define new avenues to therapies that will help restore the lost tissues. Past studies suggest that bone outgrowth by osteoblasts is a critical step, but the process by which skeletal elongation and proper spatial patterning occur remains to be
determined. Successful regeneration depends on the recapitulation of a limb-specific developmental program, including the transient upregulation of HoxA cluster genes, which encode transcription factors that coordinate limb development. Of these genes, Hoxa13 is involved in autopod (hand/foot) formation, and Hoxa13 mutations
result in a congenital syndrome with malformed digits. Hoxa13 directly regulates the expression of Eph receptors and ephrin ligands, a family of signaling proteins that mediate limb morphogenesis by influencing cell shape and tissue boundary segmentation. However, the role of Eph/ephrin signaling during regeneration is unclear. A potential downstream pathway is RhoA/ROCK signaling, which induces cytoskeletal changes. Therefore, I will examine the role of HoxA genes in activating and organizing osteoblast lineage cells during digit regeneration. I hypothesize that HoxA genes, and specifically Hoxa13, affect skeletal outgrowth and patterning during digit
regeneration via Eph/ephrin and RhoA/ROCK signaling.

Aim 1: Assess the regenerative capacity of osteoprogenitors with HoxA deletion.

Aim 2: Determine the effect of Hoxa13 expression on spatial patterning during
osteogenesis using an in vitro iPSC model.

PI: Nidhi Rohatgi, PhD

Role of epigenetic complex, ASXL2-Bap1 on inflammatory arthritis.

Specific Aims

Inflammatory arthritis (IA) results in joint destruction by periarticular osteolysis which is a reflection of abundant macrophages. Our laboratory has established that these cells serve as osteoclast precursors and secrete
inflammatory cytokines which promote osteoclast formation directly and by stimulating synovial cell RANKL expression. Thus, regulating macrophage infiltration and activation in addition to osteoclastogenesis in inflamed joints is central to arresting IA .  Epigenetics play an important role in the pathogenesis of rheumatoid osteoarthritis (RA)  and epigenetic modifiers have become candidates for drug targeting.  In this regard, we have shown that myeloid deletion of the epigenetic modifier Additional Sex Comb like (ASXL) gene 2 (ASXL2) suppresses osteoclastogenesis in a cell autonomous manner and thus enhances bone mass. While we originally posited this event is mediated by PPARγ, we subsequently determined such is not the case. Thus, the means by which ASXL2 regulates osteoclastogenesis is enigmatic. We have also established myeloid deletion of ASXL2 in macrophages polarizes them to a novel hypo-inflammatory phenotype which prevents diet induced obesity. In particular, macrophages lacking this gene are less prone to adipose tissue infiltration and produce less inflammatory cytokines. In keeping with this posture, loss of its cofactor BRCA1Associated Protein 1 (Bap1) mirrors this phenotype of increased bone mass and protection from dietinduced obesity. Thus, absence of ASXL2 in macrophages arrests osteoclast formation and induces a hypoinflammatory phenotype, properties which may positively impact the destructive properties of inflammatory arthritis.
We therefore hypothesize that inactivation of ASXL2 and/or BAP1 in myeloid lineage cells will arrest inflammatory joint destruction by inhibiting osteoclastogenesis and reducing inflammation.  Therefore, our Specific Aim is to determine the effects of inactivation of ASXL2 and BAP1, alone and in
combination, on inflammatory osteolysis.
These proposed studies will accelerate research, and also assist my path to independence. This award, if funded, will enable me to write my first RO1, focusing on epigenetic regulation in bone diseases.

• Aim 1:Determine the effects of inactivation of ASXL2 and BAP1, alone and in combination, on inflammatory osteolysis.
  • Effect of co-deletion of ASXL2 and BAP1 on osteoclastogenesis.
  • Effect of deletion of ASXL2 and/or BAP1 on inflammatory arthritis.
  • Epigenetic regulation of macrophage and osteoclast function.

PI: Andrew Findlay, MD

The role of DNAJB6 in skeletal muscle.

Specific Aims

Protein chaperones, or heat shock proteins (HSPs) facilitate protein folding and are critical for skeletal muscle health. An emerging group of hereditary muscle disorders are caused by mutations in HSPs. Recently our group discovered mutations in DNAJB6, an HSP40 co-chaperone, cause limb girdle muscular dystrophy 1D (LGMD1D). This is a dominantly inherited, adult onset, progressive myopathy with vacuolar and aggregate myopathology. DNAJB6’s role in normal muscle, and how mutations cause myopathy, is unknown. Most studies on DNAJB6 and LGMD1D pathogenesis suggest it is important for protein homeostasis and myofibrillar integrity. In fact, absence of DNAJB6 is embryonic lethal due to failure of chorioallantoic due to keratin aggregation.

Our preliminary data indicate DNAJB6 has two distinct roles in skeletal muscle: 1) sarcomeric protein homeostasis and 2) suppression of myogenic signaling pathways crucial for myotube fusion and skeletal muscle hypertrophy.  We made several interesting observations regarding these dual functions by comparing our dominant, knock-in, LGMD1D mice (DNAJB6 F90I +/-) with DNAJB6 knockout (KO) myoblasts. DNAJB6 KO myoblasts and myotubes had myofibrillar disorganization and accumulation of sarcomeric proteins. This resembled the myofibrillar, aggregate, and vacuolar pathology seen with dominant disease causing DNAJB6 mutations in patient muscle, and in our knock-in LGMD1D mice.  Because LGMD1D is not due to a KO or loss of DNAJB6, one could speculate that the vacuolar and aggregate myopathology seen with disease mutations results from a loss of DNAJB6’s sarcomeric chaperone activities. In contrast, our preliminary data on DNAJB6’s impact on myogenic signaling pathways does not agree with such a mechanism. Specifically, we found that absence of DNAJB6 in myotubes leads to enhanced myogenic signaling and significantly enlarged, hypertrophic myotubes. Conversely, we found dominant DNAJB6 mutations suppressed these myogenic signaling pathways in cells and skeletal muscle. Furthermore, pharmacologic activation of these pathways improved muscle fiber pathology and strength in LGMD1D mice. These contrasting findings between KO cells and LGMD1D tissue suggest disease mutations may actually enhance DNAJB6 mediated suppression of myogenic signaling pathways.

Based on these preliminary findings, we hypothesize DNAJB6’s role in skeletal muscle involves both sarcomeric protein homeostasis and suppression of myogenic signaling pathways. Additionally, LGMD1D associated mutations cause disease through both a loss of function on sarcomeric protein homeostasis, and enhanced DNAJB6 mediated suppression of myogenic signaling pathways. Our objective is to delineate DNAJB6’s multiple roles in skeletal muscle and their contribution to LGMD1D pathogenesis. We will therefore create (aim 1) and characterize (aim 2) a skeletal muscle-specific DNAJB6 KO mouse. Creation of a DNAJB6 KO mouse will model loss of its chaperone functions. This is an essential experiment to understand the complex disease mechanisms in LGMD1D.

• Aim 1: Create skeletal muscle-specific DNAJB6 KO mouse. This experiment is required to understand DNAJB6’s roles in muscle and delineate pathomechanisms in LGMD1D. We will utilize the Washington University Mouse Genetics core for design and creation of the DNAJB6 fl/fl mouse line. DNAJB6 has significant homology to other DNAJ proteins. We will therefore validate the guide RNAs (gRNA) and screen for off target effects. We will confirm LoxP sites are on the same allele, breed to homozygosity, and ensure homozygous mice are viable. To create skeletal muscle specific KO mice, we will cross to our existing MYL-Cre animals and confirm KO via immuno-blot.
• Aim 2: Define DNAJB6’s roles in skeletal muscle in-vivo. We will characterize muscle histopathology with assistance from the Musculoskeletal Histology and Morphometry core. We will define functional strength deficits in-vivo and ex-vivo using the Aurora Scientific muscle physiology apparatus in the Structure and Strength core. In-vitro characterization of DNAJB6’s impact on myogenesis will be assessed via primary myoblast and myofiber culture.

PI: Kelsey Collins, PhD

A Fat-Free Mouse Model to Study Biomechanical
and Metabolic Contributors to Osteoarthritis

Specific Aims

The purpose of this pilot and feasibility grant is to create designer fat implants as means of delivering specific adipokines in the LD mouse model in vivo. These important pilot studies will provide the platform for Dr. Collins’ upcoming K99/R00 application aiming to strategically rescue the adipokine signaling in the LD mouse model to define the mechanism of adipose-cartilage signaling and, thus, determine the specific adiposederived factors that override cartilage protection in this model. Furthermore, we will simultaneously develop a novel regenerative medicine strategy for therapeutic application. The proposed studies will provide foundational mechanistic information into the relationship between adipose-cartilage signaling, a nearly completely open area of investigation that will provide the platform for Dr. Collins’ future independent career.

• Aim 1:Demonstrate reciprocal adipose-cartilage signaling by re-introducing vulnerability to post-traumatic OA using an adipose implantation approach.
  • Hypothesis: Restoration of adipokine signaling through implantation of a functional fat pad to LD mice will reverse protection from post-traumatic OA
  • Approach: We will implant mouse embryonic fibroblasts (MEFs) that will form a fat transplant, correct metabolic abnormalities and recapitulate adipose tissue signaling to cartilage. We will also implant WT fat pads from donor mice to host LD mice to evaluate an existing fat implant and provide a framework for the development of tunable iPSC transplants that can screen for the mechanism of fat signaling in a specific and dose-dependent manner. Recipient mice will be challenged with destabilization of the medial meniscus (DMM) surgery to induce OA and will be followed for 28-weeks to evaluate if vulnerability to OA was re-introduced by fat transplant when compared to non-transplanted LD mice. The primary outcome of this study will be knee Modified Mankin Score.
• Aim 2:  Generate designer adipose implants using murine iPSCs to provide a platform to deconstruct adipokine signaling and investigate the mechanisms linking adipose tissue and joint health using the LD mice.
  • Hypothesis: CRISPR-Cas9 knockout of leptin in adipocyte-like murine iPSCs delivered as fat transplants to LD mice will protect against OA when compared to WT iPSC-derived implants.
  • Approach: We will begin with targeting leptin signaling as an initial effort to deconstruct the role of adipokine signaling in obesity-associated OA. First, we will use CRISPR-Cas9 genome editing to specifically delete the leptin gene. Once the stable leptin knockout iPSC line is created, lentiviral overexpression of peroxisome proliferator-activated receptor gamma (Ppar-γ), a master regulator of adipogenesis, will be used to efficiently induce the differentiation of iPSCs into adipocytes. These iPSC-derived adipocytes will be cultured and transplanted, by injecting subcutaneously superior to the sternum, or delivered as a tissue construct to mimic the WT fat transplant. The primary outcome of this study will be knee Modified Mankin Score.

PI: Kelsey Collins, PhD

A Fat-Free Mouse Model to Study Biomechanical
and Metabolic Contributors to Osteoarthritis

Specific Aims

The purpose of this pilot and feasibility grant is to create designer fat implants as means of delivering specific adipokines in the LD mouse model in vivo. These important pilot studies will provide the platform for Dr. Collins’ upcoming K99/R00 application aiming to strategically rescue the adipokine signaling in the LD mouse model to define the mechanism of adipose-cartilage signaling and, thus, determine the specific adiposederived factors that override cartilage protection in this model. Furthermore, we will simultaneously develop a novel regenerative medicine strategy for therapeutic application. The proposed studies will provide foundational mechanistic information into the relationship between adipose-cartilage signaling, a nearly completely open area of investigation that will provide the platform for Dr. Collins’ future independent career.

• Aim 1:Demonstrate reciprocal adipose-cartilage signaling by re-introducing vulnerability to post-traumatic OA using an adipose implantation approach.
  • Hypothesis: Restoration of adipokine signaling through implantation of a functional fat pad to LD mice will reverse protection from post-traumatic OA
  • Approach: We will implant mouse embryonic fibroblasts (MEFs) that will form a fat transplant, correct metabolic abnormalities and recapitulate adipose tissue signaling to cartilage. We will also implant WT fat pads from donor mice to host LD mice to evaluate an existing fat implant and provide a framework for the development of tunable iPSC transplants that can screen for the mechanism of fat signaling in a specific and dose-dependent manner. Recipient mice will be challenged with destabilization of the medial meniscus (DMM) surgery to induce OA and will be followed for 28-weeks to evaluate if vulnerability to OA was re-introduced by fat transplant when compared to non-transplanted LD mice. The primary outcome of this study will be knee Modified Mankin Score.
• Aim 2:  Generate designer adipose implants using murine iPSCs to provide a platform to deconstruct adipokine signaling and investigate the mechanisms linking adipose tissue and joint health using the LD mice.
  • Hypothesis: CRISPR-Cas9 knockout of leptin in adipocyte-like murine iPSCs delivered as fat transplants to LD mice will protect against OA when compared to WT iPSC-derived implants.
  • Approach: We will begin with targeting leptin signaling as an initial effort to deconstruct the role of adipokine signaling in obesity-associated OA. First, we will use CRISPR-Cas9 genome editing to specifically delete the leptin gene. Once the stable leptin knockout iPSC line is created, lentiviral overexpression of peroxisome proliferator-activated receptor gamma (Ppar-γ), a master regulator of adipogenesis, will be used to efficiently induce the differentiation of iPSCs into adipocytes. These iPSC-derived adipocytes will be cultured and transplanted, by injecting subcutaneously superior to the sternum, or delivered as a tissue construct to mimic the WT fat transplant. The primary outcome of this study will be knee Modified Mankin Score.

PI: Rajan Sah, MD, PhD

SWELL1-LRRC8 Regulation of Skeletal Muscle Metabolism and Function

Specific Aims

Regular exercise and maintenance of muscle endurance and lean muscle mass are known to be beneficial in the prevention of obesity and obesity-related diseases such as diabetes and heart disease, in addition to promoting overall health of our aging population. Skeletal muscle atrophy is associated with cancer (cachexia), heart failure, chronic corticosteroid use, paralysis or denervation (disuse atrophy) and aging1 and can contribute to poor metabolic health. Accordingly, a deeper understanding of the molecular mechanisms that regulate skeletal muscle maintenance, growth and function is critical for human health.

It is well established that insulin/IGF1-PI3K-AKT-mTOR signaling is a critical regulator of skeletal muscle differentiation and growth. In addition, mechanical loading of muscle, as occurs with regular activity, exercise, and resistance training is well-known to also induce mTOR-mediated skeletal muscle growth, however the cellular mechanosensor(s) responsible remain unclear. b1-integrin and focal adhesion kinase signaling has been proposed as a candidate mechanosensory complex that contributes the skeletal muscle hypertrophic signaling.  Similarly, ion channels are putative mechanosensory membrane proteins that may regulate intracellular
signaling.

We recently identified SWELL1 (LRRC8a) as a swell or stretchactivated volume sensor in adipocytes that regulates glucose uptake,
lipid content, and adipocyte growth via a novel SWELL1-GRB2-PI3KAKT signaling pathway – providing a putative feed-forward amplifier to enhance adipocyte growth and insulin signaling during caloric excess.

The specific aims of this study are as follows:

• Aim 1:Determine the contribution of SWELL1 signaling to skeletal muscle growth, endurance and composition upon aerobic training.
  • In three genetic models we will perform a treadmill exercise capacity: pre- and post-treadmill training and then assess for skeletal muscle hypertrophy, fiber-type composition (histologically, transcriptionally), induction of AKT/mTOR signaling and mitochondrial biogenesis.
• Aim 2:Determine the contribution of SWELL1 to skeletal muscle force generation and intracellular signaling upon strength training.
  • As in AIM#1, we will use the skeletal muscle-targeted SWELL1 loss- and
    gain-of-function mice to perform in vivo strength testing pre-, immediately post and 48 hour recovery from an eccentric contraction bout. We will measure fiber size and fiber type changes (histologically, transcriptionally), and expression of genes involved in myogenic differentiation, and activation of AKT/mTOR signaling pathways.

PI: Gabriel Haller, PhD

Genetic models of Chiari I malformation: Zebrafish models of CM1-associated genes.

Specific Aims

Chiari type I malformation (CM1) is one of the most common pediatric neurological conditions, affecting ~3% of individuals undergoing brain imaging. CM1 is characterized by the herniation of the cerebellum through the foramen magnum into the spinal canal, often leading to obstruction of normal cerebrospinal fluid flow, compression of the brainstem and numerous neurological symptoms ranging from mild headaches to lifethreatening respiratory failure. The formation of CM1 is hypothesized to largely result from defects of musculoskeletal tissues in the base of the skull. There is a strong connection between adolescent idiopathic scoliosis (AIS) and the development of Chiari I malformation in children with approximately 20% of idiopathic CM1 cases presenting with concomitant severe scoliosis. This rate is notably much higher than the 3% population prevalence of scoliosis. Additionally, the rate of CM1 among patients with monogenic connective tissue disorders is significantly higher than in the general population with 12.7% of CM1 patients meeting criteria for Ehlers-Danlos syndrome or a related hereditary disorder of connective tissue.

We recently found a strong connection between genetic variation in genes related to connective tissue disorders (FBN1 and collagen genes) and scoliosis risk. Together with the epidemiological evidence connecting CM1 and scoliosis, this genetic data leads us to believe that CM1 and scoliosis may share underlying, largely musculoskeletal, etiology. We have now identified several genes enriched for deleterious protein-coding genetic variation among our cohort of 250 exome-sequenced CM1 patients compared to our large in-house cohort of control exome sequenced individuals (>3000 exome sequenced individuals). Several of the top CM1-associated genes have potential roles in the formation of the skull base or connective tissue development, including CUL7, ERCC6 and LAMC3. The goal of this proposed investigation is to determine the effects of disrupting these CM1-associated genes in zebrafish to understand their mechanisms of pathogenicity in humans. By creating directed mutations, knock-out of CM1-associated genes and knock-in of specific CM1-associated mutations, we will be able to determine the role of targeted genes in CM1 pathology and uncover the cellular and molecular etiology of CM1. The specific aims of this study are as follows:

• Aim 1: Creation and characterization of zebrafish with genetic disruption of CM1-associated genes.
  • Aim 1A: Create knock-out zebrafish lines for CM1-associated genes and knock-in lines for evolutionarily conserved CM1-associated mutations using the Washington University zebrafish core facility with validation of mutant lines by next generation sequencing at the Genome Technology Access Center (WUSM).
  • Aim 1B: Measure skull and spine abnormalities in zebrafish mutants using both tissue-specific fluorescent zebrafish lines, bone/cartilage staining using the WU zebrafish core facility and by μ-CT imaging of mutant zebrafish at the Musculoskeletal Research Center (WUMRC).

PI: Michael Harris, PhD

Muscle Performance after Periacetabular Osteotomy for Hip Dysplasia.

Specific Aims

Our long-term goal is to identify structural and functional deficiencies that contribute to the pathophysiology of osteoarthritis (OA) in developmental dysplasia of the hip (DDH) and related diseases. Our ongoing work focuses on in-vivo movement patterns and muscle performance in patients with DDH before and after corrective surgery. In this renewal, we propose coupling in-vivo analyses with musculoskeletal and probabilistic modeling to quantify the effect of surgical correction on joint-level loading in the hip.

DDH is a major etiological factor in early hip OA. DDH is characterized by abnormal acetabular and femoral geometry, which cause insufficient coverage of the femoral head, and result in altered joint loading that can damage the articular cartilage and acetabular labrum. Without correction, tissue damage progresses to OA.
Periacetabular osteotomy (PAO) is a common corrective surgery for DDH that involves reorienting the acetabulum, and optionally reorienting the proximal femur. Short-term results from PAO show improved femoral coverage and return to full activity at 4-9 months. However, 30 years after PAO, 71% of hips have progressed to advanced OA or arthroplasty. Furthermore, 20% of patients treated with PAO for unilateral symptoms of DDH develop symptoms on the contralateral hip within 18 months.

A barrier to improving long-term joint survival in DDH cases is the lack of information about how PAO reorientation alters joint loading. While PAO addresses bony structure, muscle function and movement patterns strongly contribute to loading at the hip and are likely factors in the development of articular tissue damage and contralateral symptoms. Specifically, hip joint reaction forces (JRFs) represent the cumulative load that muscle and ground reaction forces place on the hip during dynamic motion, and influence how stresses on the cartilage and labrum will be distributed.9 Joint reorientation during PAO likely alters JRFs, but such alterations have not been quantified, and the sensitivity of JRFs to variability in surgical reorientation is unknown.

In the current proposal, we will use musculoskeletal modeling and probability analysis to quantify JRFs and muscle forces. Musculoskeletal models are useful for estimating variables, such as JRFs, that cannot be measured directly. Our models will be the most specific for DDH to-date by including subject-specific bone and muscle geometry, and will be validated with experimental data from our ongoing studies. We will also apply probabilistic analysis to determine the sensitivity of JRFs to post-PAO acetabular and femoral orientation. Probability analyses quantify the effects of input variability (e.g. PAO reorientation in a given direction) upon outputs (e.g. JRFs) and will provide important information for optimizing PAO planning.

• Aim 1: Develop pre- and post-PAO musculoskeletal models that incorporate subject-specific bone geometry, muscle lines of action, and muscle strength to estimate joint level loading (muscle forces and JRFs). Utilizing in-vivo data collected in our ongoing project (motion data, electromyography, magnetic resonance imaging, muscle strength), we will develop musculoskeletal models at a level of subject-specificity not previously achieved for the DDH population. We will validate model outputs using industry-standard methods including comparison to in-vivo muscle activation patterns and net joint load estimates.
• Aim 2: Compare hip JRF changes after PAO during level walking, inclined/declined walking, and squatting. Pre- and post-PAO motion data from our ongoing work will be applied to the subject-specific musculoskeletal models to quantify how PAO changes hip loads in patients who are returning to high levels of activity. We hypothesize that post-PAO, hip JRFs will be lower than pre-PAO values for each activity, and will be redirected more toward joint compression (i.e. away from the acetabular rim).
• Aim 3: Characterize the sensitivity of JRFs to acetabular and femoral reorientation. Using probabilistic analyses, we will identify the sensitivity of hip JRFs to individual degrees of PAO reorientation (acetabular extension, acetabular adduction, medial translation, and femoral rotation). We hypothesize that hip JRFs will be most sensitive to surgical corrections that affect the lateral coverage of the femoral head.

The proposed project is a natural and important extension of our in-vivo measurements of hip mechanics. By applying our in-vivo results to detailed musculoskeletal models of DDH, we will elucidate important changes in joint loading that occur because of PAO. We will also develop a tool to inform surgeons about the sensitivity of joint loads to specific PAO reorientation parameters. Understanding how PAO changes hip mechanics will provide valuable insights into short term post-surgical outcomes. The proposed and ongoing projects will also lay the groundwork for larger studies to significantly advance our understanding of how contemporary interventions change the mechanical environment of dysplastic hips and how these changes may then alter the potential for OA development.

PI: Gaurav Swarnkar, PhD

The role of NF-κB signaling in Osteogenesis

Specific Aims

Normal skeletal development is a coordinated process of bone formation by osteoblasts (OB) and bone resorption by osteoclasts (OC), which is tightly regulated by metabolic and bio-energetic circuits. However, pathological conditions, such as chronic inflammation and aging, compromise the rate of bone formation and delay healing of bone fractures1-11. Various inflammatory signals such as TNF and IL-1β elicit bone catabolic responses through increased bone resorption via osteoclasts concurrent with reduced bone formation resulting in net bone loss12-14. However, the mechanisms underlying the effect of inflammation on bone formation and on bone cells such as OB and osteocytes (OCy) are not fully understood owing to pro-inflammatory cytokine redundancy and to paucity of relevant genetic models. This gap of knowledge offers a unique opportunity to identify new intervention targets to improve bone health.

Inflammatory responses are bio-energetically costly, hence they impede metabolic homeostasis. Recent reports reveal an intimate connection between NF-κB and metabolism 15-17. The transcription factor NF-κB has been implicated as crucial mediator of immune/inflammatory responses and is required for skeletal development 18-23. We have shown that constitutive activation of IKK2 (IKK2ca), mimicking unresolved inflammation, in the mesenchymal compartment using Col2-cre, disrupts OB and chondrocyte differentiation and impairs skeletal development. In this proposal, our preliminary results using OB specific cre (SP7 and Col1 cre) to drive expression of IKK2ca in mice resulted in marked decrease in serum glucose levels and significant bone abnormalities illustrated by defective OB differentiation and elevated osteocyte markers. Mechanistically, we further found that primary calvarial OBs (cOB) expressing IKK2ca or treated with IL1β or TNFα cultured under osteogenic conditions, displayed significant decrease in OB differentiation-associated genes including ALP, Col1, OCN, OPG and Runx2. Most surprisingly, these cells assumed an osteocytic/dendritic morphology, expressed bona fide osteocyte markers (E11, FGF23, SOST, DKK1, RANKL), and exhibited high metabolic rate evident by accelerated glucose consumption and increased ECAR (extracellular acidification rate) values compared with control cells. Furthermore, Western blot analysis showed an increase in mTOR expression, which is well established as central sensor of cellular metabolism. Intriguingly, mTOR inhibitors significantly reversed the IKK2ca-induced osteocytic and glycolytic gene expression profiles. These observations, suggested that IKK2ca abnormally upregulated mTOR signaling and cellular energy metabolism in OBs leading to reprogramming and accelerated cell differentiation of the highly anabolic OB into terminal OCY. Taken together, these novel observations lead us to hypothesize that inflammatory responses alter osteoblast differentiation program and fate decision leading to their rapid conversion into osteocytes resulting in abnormal bone formation and bone loss. Mechanistically, the process is mediated by IKK2 activation of mTOR pathway leading to abnormal cellular metabolic activity by OBs resulting with accelerated differentiation of OBs to OCy.

• Aim 1: Investigate the role of activation of IKK2/NF-κB on accelerated differentiation of osteoblast to osteocyte, and subsequent bone loss, in animal model of chronic inflammation. We will use genetic and pharmacologically-induced chronic inflammatory models to analyze the accelerated differentiation of OB to OCy.
• Aim 2: Determine if OB-specific attenuation of NF-κB activity, inhibition of mTOR, or inhibition of glycolysis under inflammatory conditions attenuate bone loss and improves fracture healing in mice. Genetic and pharmacologic inhibition of glycolysis, NF-κB and mTOR pathways will be employed to accelerate and improve bone defect healing.

We expect that these approaches will identify novel therapeutic intervention targets to combat inflammation-induced bone loss.

PI: Aaron Johnson, PhD

Mechanisms of Mammalian Muscle Morphogenesis

Specific Aims

Congenital myopathies (CM) are a heterogeneous collection of disorders defined by early onset hypotonia, and many myopathy patients will require lifelong mechanical assistance and even surgical interventions to maintain mobility and respiration. Skeletal myogenesis initiates with the specification of myoblasts, which fuse to form a syncytial nascent myotube. The nascent myotube must then elongate and attach to tendon cells to form a functional contractile unit. Our overarching hypothesis is that genetic perturbation to the early myogenic program causes severe developmental defects that result in neonatal mortality, whereas mutations that disrupt myogenesis downstream of myoblast specification underlie the clinical phenotypes associated with CMs.

There remain critical knowledge gaps in our understanding of skeletal muscle development. In particular, the molecules that guide myotube leading edges to their tendon attachment sites, a process we termed myotube pathfinding, remain largely unknown. In addition, the mechanisms by which myotubes respond to extracellular signals are still unclear. Our forward genetic screens and transcriptional profiling studies in Drosophila identified Fibroblast Growth Factors (FGFs) as key guidance molecules for myotube pathfinding. In vertebrates, nascent myotubes express FGF receptors and transduce FGF signals, but a functional role for FGF signaling during myotube morphogenesis has not been characterized. In fact, the mechanisms by which individual mammalian muscles acquire their unique size, shape, and morphology are largely unknown. The specific hypothesis for this application is that FGF ligands act as guidance cues during mammalian myogenesis that target myotube leading edges to tendon attachment sites.

• Aim 1: Characterize the in vivo role of the FGF pathway during muscle morphogenesis. FGF signaling directs myoblast proliferation and migration, but the role of the FGF pathway during myotube morphogenesis is unknown. Four FGF receptors (FGFRs) have been identified in vertebrates, and nascent myotubes express FGFR1, FGFR2, and FGFR4. As an entry point to characterize the role of FGF signaling during myotube pathfinding in mice, we will use existing FGFR alleles and a muscle-specific CreER to create triple conditional knockout myotubes (FGFR1floxed/floxed, FGFR2floxed/floxed, FGFR4floxed/floxed, MyogCreERT2+/-; hereafter FGFRTCKO). CreER activity will be induced at two embryonic time points, and we will assess muscle size and morphology in FGFRTCKO neonates. These studies will provide the first functional insights into the roles of FGF signaling during myotube morphogenesis.
• Aim 2: Establish an in vitro platform to probe mechanisms of myotube pathfinding. Despite the cellular and molecular parallels between axon guidance and myotube pathfinding, in vitro techniques have yet to be developed to test the function of putative guidance cues in the context of myotube morphogenesis. We will apply microdevice technologies established for the study of axon guidance toward understanding the role of FGF signaling during myotube pathfinding in vitro. These tools will complement our in vivo studies of mammalian myogenesis in the short term, and will serve as a productive screening platform to identify myogenic mechanisms and characterize myopathy disease variants in the long term.

PI: Michael Harris, PhD

Muscle Performance after Periacetabular Osteotomy for Hip Dysplasia

Specific Aims

Our long-term goal is to identify structural and functional deficiencies at the hip, which contribute to mechanical and metabolic pathways of osteoarthritis (OA) development. The current project introduces an innovative approach to understanding the pathogenesis of OA in cases of developmental dysplasia of the hip (DDH). The specific aims of the project will elucidate abnormalities in muscle performance and hip biomechanics in patients with DDH who have undergone hip preservation surgery, but remain at elevated risk for early OA.

DDH is a major etiological factor in OA development and increases an individual’s likelihood of OA by 4.3 fold. DDH is characterized by abnormal acetabular and femoral geometry, which cause altered intra-articular loading that can damage the articular cartilage and acetabular labrum. Without correction, soft-tissue damage progresses to OA and may require total joint arthroplasty.

Periacetabular osteotomy (PAO) is a common surgical technique designed to preserve the articular cartilage of the hip by reorienting acetabular coverage of the femoral head. Short-term results from PAO show improved femoral coverage and return to full activity at 4-9 months. However, 30 years after PAO, 71% of hips have progressed to advanced OA or arthroplasty. Furthermore, 20% of patients treated with PAO for unilateral symptoms of DDH, develop symptoms on the contralateral hip within 18 months.10 Thus, current PAO treatment cannot guarantee elimination of symptoms, long-term restoration of activity, or delay OA.

Beyond the bony structure treated with PAO, factors such as muscle function, strongly contribute to the mechanical environment at the hip and are likely factors in the development of contralateral symptoms and altered joint loading that lead to OA. Joint mechanics after PAO are severely understudied, and the limited results have been inconclusive. Instead, most post-PAO investigations focus on patient reported outcomes (e.g. quality of life, pain) or joint survivorship. No study has considered post-PAO joint mechanics (motion and loading) during activities that are common to young adults returning to activity after PAO. Furthermore, despite six major muscles of the hip being directly affected by PAO surgery, no study has investigated muscle morphology (e.g. atrophy), neuromuscular coordination, and muscle strength at the time of return to full activity after PAO, when patients are expected to perform similar to healthy individuals.
We hypothesize that muscles, as the primary movers and stabilizers of the hip: (Hyp1) have inferior strength and volume, (Hyp2) manifest abnormal neuromuscular coordination, and (Hyp3) contribute to abnormal joint mechanics at the time of return to full activity after PAO. These abnormalities are suspected to cause asymmetries in hip loading and lead to contralateral symptom development and long-term progression to OA.

Quantifying the role of muscle in the setting of DDH is innovative, highly translational, and can provide invaluable insight into the mechanical environment of the dysplastic hip and the pathogenesis of OA. An immediate impact of our results will be to identify areas for improved post-PAO rehabilitation strategies. Knowing motion and loading of both the surgical and nonsurgical limbs, during dynamic activities can provide valuable, macroscale knowledge about potential mechanisms of future damage. As such, results from this project may also inform microscale and animal studies targeting mechanical and metabolic changes to the musculature in response to an array of bone and articular cartilage surgeries.

• Aim 1: Asymmetries in muscle morphology and strength between surgical and nonsurgical limbs in patients after PAO compared to normal limb-to-limb asymmetries in healthy controls. Hyp1 will be tested using bilateral 3D reconstructions of hip musculature to measure muscle volume and isometric strength measures of muscle torque in hip muscles affected by PAO.
• Aim 2: Asymmetries in neuromuscular coordination between surgical and nonsurgical limbs in patients after PAO compared to healthy controls during static and dynamic activities. Hyp2 will be tested using electromyography (EMG) to quantify muscle activity patterns in major hip muscles during static isometric muscle strength tests and biomechanically challenging activities (e.g. gait, running, squatting).
• Aim 3: Differences in hip joint mechanics (motion, loading) between patients after PAO and controls during dynamic activities of increasing biomechanical demand. Hyp3 will be tested using 3D motion capture and inverse dynamics modeling to quantify hip mechanics (motion, loading) during dynamic tasks that are expected of patients who have returned to full activity.

PI: Timothy Peterson, PhD

Subclinical Doses of Nitrogen-containing Bisphosphonates Produce Clinical Response on Bone.

Specific Aims

New therapies for bone are hard to come by. Bisphosphonates, especially the nitrogen-containing species (NBPs), are the standard of care for many diseases involving bone1,2. They are inexpensive and work well. However, because they cause rare yet devastating side-effects such as atypical fractures and osteonecrosis of the jaw many patients avoid taking them1,3,4. This fear over side effects is so great and is growing that it has caused the number of prescriptions to plummet 50% in the last four years5. Moreover, several drug companies and funding agencies have stalled their support for bone research and development6. Thus, there is a pressing need for the bone community to work with what we’ve got. Our preliminary data suggests that NBPs used at concentrations 1,000-fold less than those currently clinically used increase bone density and strength and could be protective against fracture. The impetus for this work revolves around poorly appreciated aspects of NBP drug actions. The widely accepted mechanism for the NBPs is that they inhibit the function of the bone resorptive cell, osteoclasts7. In osteoclasts the targets for NBPs are the cholesterol biosynthetic enzymes, farnesyl pyrophosphate synthase and geranyl-geranyl phosphate synthase, FPPS and GGPS, respectively8. These enzymes are inhibited by μM doses of NBPs in cell culture contexts. Interestingly, in vitro and ex vivo experiments with the bone depositing cell type, osteoblasts, suggest that very low doses of NBPs (in the nM range, i.e., 1,000-fold less than those used to inhibit cultured osteoclasts) can have osteoblast differentiation-potentiating effects9-11. These same very low doses don’t impair osteoclasts in vitro12,13. Lastly, it has also been shown that NBPs can maintain osteocyte viability depending on the dose14. The mechanisms for these effects are unclear. Considering the number of patients at stake (hundreds of millions), it is surprising that few in vivo studies exist to clarify the effects of the various NBPs at varying doses on the major bone cell types: osteoclasts, osteoblasts, and osteocytes. To add clarity to the NBPs mechanism of action, we’ve recently identified two poorly characterized genes15,16 we named TBONE1 and TBONE2 (Target of BisphOsphonate NitrogEnous 1 and 2), that are critical to the bone protective effects of NBPs at clinical doses. We hypothesize that NBPs given at very low doses – up to 10,000-fold lower than those typically given – increase bone density and strength without increasing fragility by potentiating osteoblast function in a TBONE1/2-dependent manner.

• Aim 1: Determine the doses of NBPs that increase bone density and strength without increasing fragility.
  Though NBPs improve bone density and reduce fracture risk in most people, they can cause atypical fractures which is highly undesirable. In mice, the doses that are normally used in people increase bone fragility. Making use of this predictable response in mice, in this Aim we will seek to define doses of NBPs that don’t increase fragility yet still provide robust therapeutic benefit. The implication is that using these doses in people would reduce the occurrence of side effects to a negligible amount. In preliminary work, we have shown that 1,000-fold less Alendronate than that is normally given to mice (0.1μg/kg/wk vs. 100μg/kg/wk) increases bone density and strength while not increasing fragility as higher doses do. In this aim, we will systematically approach NBP dosing by performing titrations over a 10,000-fold concentration range with two commonly prescribed NBPs, Zolendronate and Alendronate17. We will measure bone structure and strength using μCT and three-point bending tests, respectively. Successful completion of this aim will define the NBPs doses that improve bone structure and strength.
• Aim 2: Identify the bone cell type(s) that respond to very low dose bisphosphonates.
  Here we propose that subclinical doses improve bone strength by improving bone formation in the absence of inhibiting bone resorption. At micromolar (μM) doses that are typical used in cell-based assays, NBPs negatively regulate osteoclasts, i.e., inhibit protein prenylation, yet also inhibit osteoblast viability and differentiation18-20. However, at nanomolar (nM) doses, NBPs promote osteoblast function. In preliminary work, we have shown that 10,000- fold less Alendronate than what is normally used to inhibit protein prenylation, potentiates osteoblast differentiation. Successful completion of this aim will define the doses that regulate the major bone cell types: osteoclasts, osteoblasts and osteocytes.

PI: Clarissa Craft, PhD

Extracellular Determinants of Marrow Adipocyte Function in the Skeletal Niche.

Specific Aims

Marrow adipocytes are increasingly becoming recognized as playing an active role in metabolic and skeletal homeostasis – with potential to function as an energy reservoir, endocrine regulator, and thermal insulator (Fig.1). Recent work has revealed that the relationship between marrow adipose tissue (MAT) and bone is far more nuanced than previously appreciated, being subject to regulation not only between disease models but also by skeletal site. Its inverse relationship with bone mass, in particular, has come under increasing scrutiny in recent years. For example, distal, fatty marrow regions are filled with MAT adipocytes, yet they actually have comparable or even higher basal cancellous bone volume and increased trabecular thickness relative to proximal red marrow sites (1–3). Consideration of mouse strains is also informative. C3H/HeJ mice have significantly more MAT and more bone than C57BL/6J, counterintuitive to the assumption that MAT causes bone loss in all contexts (4). We hypothesize that this discrepancy can be explained by sitespecific differences in the biochemical nature of the marrow adipocyte and its extracellular niche. Indeed, previous work has shown that the phenotype of marrow fat varies depending on its location in the skeleton (4) (Fig.2A). Specifically, regulated MAT (rMAT) adipocytes increase gradually throughout life in the proximal and central regions of the skeleton and readily undergo changes in cell number and/or size when acted on by external forces such as prolonged cold exposure (4), induction of sympathetic tone (5,6), and stimulated hemolysis (7). By contrast, constitutive MAT (cMAT) adipocytes in the yellow marrow are refractory to change (4). Reports that correlate MAT accumulation with decreases in bone mineral density or formation and increased bone loss are generally based on rMAT-enriched sites such as the proximal femur, hip and lumbar spine (reviewed in (8)). Conversely, studies demonstrating that MAT protects against bone loss have all selected cMAT-enriched sites as their area of interest (e.g. distal tibia and tail vertebrae) (1–3).

Though there are many potential mechanisms underlying these findings, we have pinpointed two wellsupported, yet understudied, components of MAT function that we will investigate using this pilot funding, thus providing preliminary data for future NIH grant applications. First, we will definitively answer the hotly contested question of whether MAT has the ability to undergo a beige-like transition during adrenergic stimulation. This is important because recent evidence suggests that brown adipocytes may have the capacity to support bone formation (9), thus, this may be one mechanism by which certain populations of MAT adipocytes promote skeletal maintenance while others drive bone loss. Second, we will characterize extracellular matrix components of the MAT adipocyte niche in health and obesity. Our central hypothesis is that degeneration of the MAT adipocyte niche in the obese state predisposes to pathologic marrow fat hypertrophy and phenotypic change, negatively impacting skeletal health.

• Aim 1: Determine whether marrow adipocytes undergo stimulated ‘beiging’ or regulated lipolysis in response to adrenergic stimulation.
  There are currently two highly-contested hypotheses that seek to explain the phenotype and subsequent function of marrow adipocytes within the skeletal niche (reviewed in (11)). In the first, MAT adipocytes are regarded as white adipose tissue-like cells with the capacity for energy storage and release by lipogenesis and lipolysis respectively. In the second, MAT adipocytes are regarded as “beige” with the capacity for stimulated browning, leading to uncoupling of mitochondrial respiration and utilization of bioenergetic substrates for local heat production. Both hypotheses have significant, but unique, implications for bone. In the first situation, the benefit of MAT for bone would stem from its ability to provide energy in the form of lipid to surrounding hematopoietic and skeletal cells. By contrast, a ‘beige’ MAT adipocyte would instead remove energy substrates (i.e. glucose and fatty acids) from the bone marrow and convert them into heat. Both cells also have a unique capacity for cytokine secretion, further diversifying their impact on local and distant cells. Thus, assigning a phenotype to the marrow adipocyte (white vs beige) is crucial to inform future studies pertaining to the role of the MAT adipocyte in skeletal and metabolic disorders. The objective of this aim is to leverage immunohistochemical and genetic techniques, including a novel UCP1-DTA mouse model, to define the ability of MAT adipocytes to undergo stimulated beiging. We hypothesize that MAT in the proximal tibia (rMAT), but not in the distal tibia or tail (cMAT), will undergo stimulated beiging, as evidenced by UCP1 expression, with β- adrenergic stimulation. The rationale for this work is that ability of the MAT adipocyte to either supply or remove energy from the niche has important implications for the function of skeletal and hematopoietic cells.
• Aim 2: Identify regulators of marrow adipocyte phenotype and function by the characterizing the matricellular components of the skeletal niche in health and obesity.
  All adipocytes are not the same. White adipocytes primarily function to store excess energy as lipids. Brown adipocytes store lipids, but are also highly efficient at energy dissipation. A third adipocyte phenotype is termed “beige”. Beige adipocytes store lipids, however, under the right stimulus take on brown-like properties. Expression of a protein called UCP1 is a hallmark of brown and beige adipocytes, and is the primary protein that allows these cells to dissipate energy through heat production. The discovery that brown and beige adipocytes exist in adult humans, and that “browning” combats both obesity and insulin resistance, has led to significant interest in identifying pharmacological mechanisms to support adipocyte “browning” (12,13). It is well established that adipocytes communicate with the skeletal cells. Specifically, obesity and obesity-associated diabetes correlates with skeletal fragility and predisposition to fracture (10,14,15). By contrast, brown and beige adipocytes have been recently proposed to support bone quality (9,16). In this aim, we will use both UCP-1 knock-out and UCP-1-DTA mice to define the ability of the MAT adipocyte to undergo stimulated beiging. Preliminary Data: Expression of Ucp1 and multi-locularity of the lipid droplet are the standard qualifications for identifying brown and beige adipocytes, we treated mice with CL316,243 – a β3 adrenergic receptor (β3AR) agonist – which is known to induce beiging of inguinal white adipose tissue (WAT) depots. Unexpectedly, we observed that a subset of MAT adipocytes took on a ‘multilocular’ appearance after 72-hour β3AR stimulation (Fig.2B). This occurred selectively in regions of rMAT adipocytes in the femur and not in cMAT regions such as the tail vertebrae, suggesting that MAT adipocytes may undergo selective ‘beiging’ at proximal sites. My previous work also provides correlative evidence suggesting that UCP1 may be functionally important in the MAT adipocyte: male UCP1 knock-out mice have increased MAT and both mice at thermoneutrality and MAGP1 knock-out animals have increases in MAT/decreased UCP1 expression.

PI: Alison Snyder-Warwick, MD

Molecular and Genetic Analysis of Terminal Schwann Cell Function in Homeostasis and Injury.

Specific Aims

Functional outcomes following peripheral nerve injuries diminish with the time required for regenerating axons to reach their muscle target. Terminal Schwann cells (tSCs) are supportive glial cells located at the muscle target that are relatively understudied, but are attractive targets for protecting denervated muscle. Over the past 2.5 years, my laboratory has optimized techniques for tSC investigation. This proposal focuses on tSCs and will describe their unique genetic profiles, their impact on homeostasis and neuromuscular junction (NMJ) reinnervation after injury, and the mechanisms by which they respond to motor nerve injury.

Schwann cells (SCs), the principal glial cells of the peripheral nervous system (PNS), are composed of two main subtypes: myelinating and non-myelinating. tSCs are non-myelinating SCs that cover motor nerve terminals and contribute to formation, maintenance, and regeneration of the synapse [1-8]. Relatively little, however, is known about tSC signaling during these events. The goals of this proposal are to: 1) identify unique genetic markers of tSCs to improve tools for investigation, 2) determine the impact of tSCs on (a) synaptic transmission in homeostasis and (b) NMJ reinnervation after motor nerve injury, and 3) determine the mechanisms by which tSCs contribute to NMJ reinnervation after injury.

• Aim 1: To identify genes uniquely expressed in tSCs. Hypothesis: tSCs have unique genetic expression profiles that differ from other SCs. There are no known genetic markers that are unique to tSCs. We will employ microarray analysis of tSCs isolated from S100-GFP mice, whose SCs fluoresce, using: 1) a novel component dissection technique and subtraction analysis and 2) cell culture. Candidate genes will be validated in vivo with in situ hybridization and RT-PCR, and protein expression will be assessed with immunofluorescence. Our ultimate goal is to build the tools to generate a transgenic mouse line with Cre-mediated recombination in tSCs to facilitate further tSC investigation.
• Aim 2: To determine: a) the functional consequences of tSC loss in homeostasis and b) the requirement for tSC contributions to NMJ reinnervation following motor nerve injury. Hypotheses: (a) Efficiency of synaptic transmission decreases after tSC ablation resulting in reduced muscle force. (b) NMJ reinnervation is compromised after tSC ablation due to loss of axon sprouting, guidance, and trophic support, resulting in diminished or absent muscle force. tSCs interact with neurotransmitters in homeostasis [6], implicating importance for synaptic maintenance. Following motor nerve injury, tSC processes extend, induce nerve sprouting, and guide regenerating axons to adjacent NMJs, suggesting importance for reinnervation [1, 2, 9, 10]. We will perform immune-mediated tSC ablation in homeostasis and at multiple time points following nerve transection and repair. The effects of tSC ablation will be assessed functionally via muscle force testing and morphologically with confocal microscopy.
• Aim 3: To determine the mechanisms active during the tSC response after motor nerve injury. Hypotheses: The tSC response to motor nerve injury mimics myelinating SC development and response to injury with p38 MAPK activation. Myelinating SC elongation during development is mediated by the p38 MAPK signaling [11], and multiple MAPK pathways modulate myelinating SC dedifferentiation to a more immature state in response to injury [12-14]. Activation of ErbB2 mimics tSC response to injury [15]. We hypothesize that tSC process elongation after motor nerve injury is mediated via p38 MAPK activation. We will perform immunohistochemical and functional analyses of the MAPK pathways following nerve transection in different transgenic mouse models that facilitate in vivo evaluation of this pathway.

Twenty million Americans suffer from peripheral nerve injury [16], and management of peripheral nerve pathology remains temporally limited by the ability to successfully reinnervate the target muscle. With their location at the nerve-muscle interface and functional plasticity, tSCs provide an opportunity to decipher events at the muscle target. Knowledge of tSC biology may provide innovative strategies to manipulate their molecular signature to protect the muscle target, lengthen the window during which reinnervation may occur, and improve motor function. The proposed experiments unite important questions from glial biology and challenges in clinical management of peripheral nerve injuries to establish an innovative and translationally important investigative focus.

PI: Christine Pham, MD

Peptide-siRNA Nanotherapeutics for Osteoarthritis Treatment.

Specific Aims

The long-term goal of this project is to develop safe, highly specific and readily translational nanomedicine platforms for the treatment of osteoarthritis (OA). OA represents the most common form of arthritis and a major cause of morbidity in the aging population.1 The annual health care burden for OA is 185 billion dollars, reflecting its very high prevalence and negative quality of life impact.2 OA is a progressive disease for which there are limited treatment options and no disease-modifying drugs. Critical barriers to the development of a successful treatment for OA lie in the fact that the pathways governing early cartilage degeneration have remained largely unexplored. Far from being a “wear and tear” process the current paradigm suggests that, in human OA, ongoing inflammation contributes to disease pathogenesis and progression.3 There is also growing awareness that autophagy regulates cartilage homeostasis and dysregulation of this pathway may negatively impact chondrocyte survival and cartilage repair.4 We posit that modulation of inflammation and autophagy will suppress chondrocyte death, prevent cartilage degeneration and eventually halt or retard OA development.

Much interest has focused on the role of inflammatory cytokines, especially TNF-α and IL-1β, in the pathogenesis of OA. These cytokines play a major role in inflammation and cartilage degeneration by modulating events leading to chondrocyte catabolism and, ultimately, apoptosis. NF-κB is a signaling pathway that controls the expression of gene products involved in myriad cellular responses and is essential for the expression of TNF-α, IL-1β in OA synovium and chondrocytes.5 Thus, we posit that the NF-κB pathway represents an attractive therapeutic target for OA. However, the indiscriminate systemic blockade of NF-κB poses significant risks, given its central role in host immune responses.5 Targeted strategies to silence NF-κB only in affected joints, which might avoid systemic unwanted effects, are highly desirable. However, the historical challenge of safely delivering therapeutic siRNA in effective doses to selected pathologies with targeted transport systems is well known. We employed a novel peptidic nanoparticle (NP) structure that features an amphipathic, cationic, cell penetrating peptide as an siRNA carrier that is stable in biological fluids and enables coordinated endosomal escape and release of siRNA into cytoplasm to rapidly engage the RISC complex and simultaneously suppress both canonical and noncanonical NF-κB activities.6,7 The entire complex can be formed within as little as ten minutes by a simple mixing procedure that allows noncovalent, selfassembly of cationic and anionic moieties into a ~55 nm nanocomplex. The self-assembling nanocomplexes suppress inflammation in a preclinical model of rheumatoid arthritis (RA) by down-regulating NF-κB p65 expression specifically in the joints without affecting p65 expression or host immune responses in offtarget organs, even after serial injections.8 These nanocomplexes, when injected intra-articularly (i.a.), suppress p65 activation and protect against early chondrocyte apoptosis and synovitis in a novel OA model that mimics traumatic joint injury.9 Moreover, our data suggest that NF-κB siRNA nanotherapy partially preserved cartilage integrity by protecting against the impact-injury-induced downmodulation of autophagy in chondrocytes.

• Aim 1: In vivo suppression of experimental OA with peptide-NF-κB siRNA nanocomplexes. To examine the long-term outcome of nanotherapy we will use the Destabilization of the Medial Meniscus ligament (DMM), a well-accepted model to generate slowly progressing OA that approximates the primary disease in humans.10 We will use this model to test the hypothesis that periodic intra-articular delivery of peptide-NF-κB siRNA nanocomplexes will halt or retard cartilage degeneration in OA. We will also incorporate chondrocyte/cartilagetargeting ligands to test for enhanced local accumulation/therapeutic benefit.
• Aim 2: Suppression of OA with combination of NPs that silence NF-κB and modulate autophagy. Autophagy is a key cellular pathway in the maintenance of cartilage homeostasis. Recent evidence suggests that autophagic dysfunction may negatively impact chondrocyte survival, contributing to OA development.11 There also exists a complex interplay between autophagy and NF-κB signaling pathways. We hypothesize that combination of nanomedicines that suppress NF-κB activation and induce autophagy may significantly improve the efficacy of OA treatment.

PI: Gretchen Meyer, PhD

Promoting rotator cuff muscle regeneration with paracrine adipose signaling.

Specific Aims

Rotator cuff (RC) tears are one of the most common causes of musculoskeletal pain and dysfunction, estimated to affect more than 10 million people in the US alone (1). Widespread degeneration of cuff tissues, particularly RC muscles, is a hallmark of chronic tears and is associated with both a progressive decline in shoulder function and high re-tear rates following surgical repair (2-5). In the case of a chronic tear, up to 75% of the muscle volume may be replaced with fat, both internally (intramuscular) (6,7) and externally (epimuscular) (8), in a process termed fatty atrophy. Little is known about the influence of fat expansion on RC tissues. It is generally considered undesirable due to its correlation with poor surgical outcomes (4,9), however no studies have directly examined its influence on cuff degeneration, function, or recovery. Given that over 50% of surgical cases include substantial fatty expansion (6,8), addressing this significant gap in knowledge could significantly impact shoulder function and, ultimately, the quality of life for a large patient population.

Adipose depots influence neighboring tissue function through endocrine/paracrine signaling (10,11). The secreted factors, and pathways modulated, vary substantially between depots. White depots (e.g., visceral or subcutaneous fat), are associated with the release of inflammatory factors, many upstream of the muscle atrophy pathway (12). However, brown fat, a specialized thermogenic depot, secretes various growth factors known to positively influence muscle health (13-15). Recently a new adipose type, termed inducible-brown or beige fat, has been identified with a phenotype intermediate between classical white and brown depots (16,17). Beige fat has been touted for its therapeutic promise as it can be pushed along the white-brown spectrum, in a process termed “browning” (18-20).

Interestingly, our preliminary data demonstrates that epimuscular (EM) fat in the human RC is a unique beige depot whose transcriptional resemblance to brown fat is decreased in the torn vs. intact cuff, suggesting that its phenotype can be modulated. This result reveals a potential therapeutic strategy, as paracrine signaling from “browned” beige fat has been shown to regulate bone density via an insulin-like growth factor (IGF) pathway (21). Considering IGF signaling is also a major pathway for muscle hypertrophy (22), we believe the “brownness” of EM fat could act as a regulatory switch for muscle tone. However, a significant gap in understanding the paracrine regulation of muscle function by adipose tissues stands in the way of targeting EM fat therapeutically. To begin to address this critical knowledge gap, we will adapt mouse models of acute regeneration and chronic degeneration in the rotator cuff to test our central hypothesis that paracrine signaling by brown fat is anabolic for muscle and shifting adipose phenotype away from brown exacerbates muscle degenerative changes and blunts regenerative capacity. This hypothesis represents a paradigm shift in thinking about RC pathology as it suggests that the large stores of fat in the cuff, previously considered a negative feature of the disease, present a novel therapeutic opportunity to prevent muscle degeneration or enhance muscle regeneration. We will test our hypothesis via two Specific Aims:

• Aim 1: Evaluate paracrine regulation of muscle regeneration by phenotype-specific fat. Approach: RC muscle regeneration will be stimulated using an injection of toxin into the infraspinatus (IS) muscle. Adipose grafts from brown, beige or white depots of GFP mice will be transplanted adjacent to the injured IS to examine their effect on muscle regeneration. Regenerative efficiency will be assayed via ex-vivo force production and histological measures of hypertrophy, fibrosis, and angiogenesis. GFP expression will be examined histologically to identify direct contribution of adipose-derived cells to myogenesis. The gene expression signature of excised adipose grafts will be evaluated to identify putative signaling mediators. We hypothesize that muscle regeneration will be more robust in the presence of brown fat compared to beige or white grafts and will be correlated with expression of myogenic growth factors.
• Aim 2: Investigate the effects of adipose phenotype in a chronic model of rotator cuff degeneration. Approach: Chronic degeneration of the rotator cuff will be induced using our well-established model of dual tenotomy of the infraspinatus and supraspinatus tendons. Adipose tissue will be isolated and transplanted as in Aim 1 to evaluate its potential for mitigating cuff degeneration. Following 16 weeks of progressive degeneration, muscle quality will be assayed as in Aim 1. Adipose grafts will be excised and their expression signature compared with results from Aim 1 to identify phenotypic shifts as a result of exposure to the chronic degenerative environment. We hypothesize that beige grafts will shift toward a whiter profile in the chronic model and this shift will be associated with increased expression of inflammatory mediators and decreased muscle quality.

We expect that the completion of these Aims will: (1) develop an in vivo system to manipulate and evaluate muscle-fat cross-talk in a pathological environment, (2) define the influence of fat phenotype on skeletal muscle pathology and, and (3) point to novel signaling mechanisms for future exploration. We believe this data will form a springboard for the development of strategies exploiting the large stores of adipose tissue in the torn rotator cuff for therapeutic benefit, not only for muscle but for tendon and bone healing as well.

PI: Spencer Lake, PhD

Region Specific Mechanics and Multiscale Strain of Human Supraspinatus Tendon.

Specific Aims

Rotator cuff tendon tears are a significant source of pain and dysfunction [1-3], and surgical repair of these tendons fail at a high rate (between 20-94%) [4, 5]. The supraspinatus tendon (SST) is the most commonly injured tendon of the cuff, in part because of its highly complex in vivo loading environment [6]. As evidence of this, our work has shown very unique and highly inhomogeneous mechanical, structural, and compositional properties for the human SST [6-9]. For example, the medial SST exhibits properties typical of tendon while the lateral region (i.e., area of likely multiaxial loading) shows planar mechanical isotropy, disorganized collagen fibers and ECM composition similar to fibrocartilage [7, 10]. These unique properties likely help sustain multiaxial loading, but may limit the ability to support tensile loading and may predispose the SST to high rates of damage seen clinically. Unfortunately, much remains unknown regarding how the SST functions under non-tensile loads. In addition, multiscale force transfer in such complex loading environments has important implications in modulating mechanotransduction, tissue adaptation, and disease pathogenesis. For example, the degree to which macroscale deformations are transferred to the microscale will directly affect tissue remodeling, reorganization and repair. Understanding such multiscale relationships is critical towards elucidating structure-function-composition properties in healthy tissues and understanding how these relationships are impaired in injury and disease. This is particularly relevant for the oft-injured, poor healing SST in order to inform techniques to better prevent and treat rotator cuff injuries, as well as replace the SST with engineered constructs.

Our laboratory recently used a custom mechanical test system integrated with multi-photon microscopy to track multiscale deformations of tendon samples subjected to compression and shear [11]. We found strain attenuated at smaller length scales and quantified fiber sliding and realignment under load. Distinct regions of bovine flexor tendons exhibited differences in deformation behavior at the microscale, likely as a result of differences in compositional/organizational properties. Using similar techniques, the objective of this study is to quantify SST mechanical properties under compression and shear and correlate region-specific function with microstructural organization and composition. Further, using selective degradation, we will quantify – and elucidate mechanisms responsible for – contributions of specific constituents to multiscale mechanical behavior of human SSTs. The innovation of this study is multiaxial mechanical testing coupled with image-based multiscale strain analysis to identify microscale mechanisms in an oft-injured tendon.

• Aim 1: Quantify the macroscale mechanics and macro-to-microscale strains of region-specific samples of human SST under compression and shear, and correlate properties with site-matched distributions of key ECM components. Samples from different SST regions will be loaded in compression or shear while measuring local matrix, cellular, and nuclear strains using two-photon microscopy. Parameters describing the elastic/viscoelastic mechanics and thickness-dependent multiscale strain response will be compared across SST regions, and relative quantities and distributions of ECM components (e.g., collagens, proteoglycans (PGs), elastin) will be evaluated in region-matched samples via immunohistochemistry. Hypothesis: Lateral SST will exhibit more fibrocartilage-like ECM (e.g., increased PGs, COL2) that correlates with stiffer mechanics under compression/shear, and decreased microscale deformations (i.e., strains, fiber rotations) compared to medial SST.
• Aim 2: Identify mechanisms governing region-specific mechanics and multiscale strain transfer of SST under non-tensile loading. Although the role of PGs in tensile loading appears to be minimal [12-14], PGs likely contribute in compression/shear. Recent studies have also suggested a significant role for elastin [15, 16]. To elucidate the contribution of these tissue components, samples from specific SST regions will be enzyme-treated to disrupt either PGs or elastin [13, 14], and then subjected to compression or shear testing with two-photon microscopy as in Aim 1. Biochemical assays will assess the efficiency of GAG and elastin removal. Hypothesis: Selective ECM depletion will demonstrate that large PGs and elastin govern the microscale response of SST in compression and shear, respectively; ECM-disruption will alter tendon properties (i.e., mechanics, multiscale strains) more severely in lateral SST than in medial SST due to region-specific differences in ECM composition and structural organization.

The significance of this study is increased understanding of (1) multiscale strain transfer, (2) structurefunction relationships in non-tensile loading regimes, and (3) the presence/role of key constituents and mechanisms at the microscale in tendon. Further, region-specific and thickness-dependent mechanics and deformation behavior of human SST will provide significant insight into how unique properties enable this tendon to function in a complex physiologic loading environment, and how disruptions to tissue constituents may lead to degeneration and damage. Finally, this study will yield fundamental understanding that can better inform approaches to treat SST tears and/or improve the design of engineered tendon replacements.

PI: Conrad Weihl, PhD, MD

Preclinical Studies to Assess Autophagic Flux in Human Skeletal Muscle

Specific Aims

Protein degradation occurs via two principal degradative pathways, the ubiquitin proteasome system (UPS) and autophagy [1]. The importance of autophagic protein degradation in normal cellular metabolism and its dysfunction in disease has become increasingly clear over the past five years [2]. Alterations in autophagic activity have been a proposed pathogenic mechanism in many musculoskeletal disorders including Duchenne Muscular Dystrophy [3], osteoarthritis [4], Paget’s disease of the bone [5], bone and muscle related malignancy (e.g. osteosarcoma and rhabdomyosarcoma) [6] and even sarcopenia [7]. More importantly, correction of the aberrant autophagic activity is therapeutic in some mouse models of these same human diseases [3, 4, 8-10]. Our laboratory is interested in autophagic protein degradation in skeletal muscle. Skeletal muscle is the largest reservoir of free amino acids in the body. Autophagic protein degradation in skeletal muscle is modulated by diverse stimuli including nutritional state, exercise and atrophic signaling pathways [2, 11]. Moreover, many FDA approved drugs have been repurposed as autophagy modulating therapies that may have efficacy in skeletal muscle [12]. However, whether skeletal muscle autophagy is similarly regulated and dysregulated in humans and human disease states is not known.

The greatest limitation to understanding the regulation of autophagic processes in human skeletal muscle and muscle disease is the lack of a reliable analytical method for quantifying autophagic degradation in vivo.

Proteolysis (the degradation of proteins and subsequent liberation of free amino acids) is tightly regulated in cells and tissue; occurring via the UPS and autophagy. Unfortunately, measuring total proteolysis within a tissue cannot distinguish between these two pathways. Therefore, one must identify a protein that is exclusively degraded via the autophagic pathway. Our preliminary studies have identified the autophagic adaptor protein p62/SQSTM1 as a likely candidate autophagic substrate in skeletal muscle. However, static levels of p62 measured via immunohistochemistry or immunoblot, do not accurately correlate with autophagic protein degradation since p62 is both simultaneously synthesized and degraded when autophagy is stimulated.

To circumvent this inherent difficulty in interpreting the degradation of any autophagic substrate, we propose to quantify the in vivo rate of p62 degradation in human skeletal muscle using stable isotope labeling tandem mass spectrometry (MS). Stable isotope labeling has been utilized to quantify skeletal muscle actin, myosin, and mitochondrial protein synthesis and degradation rates [13, 14]. However unlike those highly abundant muscle proteins, most proteins within biologic samples are present at concentrations less than one nM. This concentration approaches the detection limit for gas chromatography-MS necessitating the utilization of newer technologies. Our methodology can detect and quantitate labeling in low abundant (femtomolar) proteins. More importantly, it can be used on small tissue samples such as skeletal muscle needle biopsy specimens from human patients.

This pilot application is a collaboration between Dr. Weihl (a neuromuscular physician interested in protein degradation pathways in skeletal muscle), Dr. Yarasheski (a pioneer in stable isotope labeling and MS analysis of human skeletal muscle proteins), and Dr. Bateman (the developer of human stable isotope labeling technology for Aβ kinetics). The infrastructure for this proposal is in place at Washington University School of Medicine within the Department of Neurology and we are uniquely positioned to address this question.

We hypothesize that the half-life (t½) of p62 in skeletal muscle represents a quantitative biomarker of autophagic kinetic activity in skeletal muscle.

• Aim 1: Define the t½ of p62 and correlate p62 t½ with autophagic activity in mouse skeletal muscle. Our preliminary data demonstrates our ability to detect labeled and unlabeled p62 in normal mice. We will A) define the synthesis rate of p62 in normal mouse skeletal muscle; B) define the t½ of p62 in normal mouse skeletal muscle and C) manipulate and quantify the t½ of p62 after modulating skeletal muscle autophagy using nutrient deprivation, pharmacologic compounds and transgenic mouse lines.

Upon completion of these studies, we will have validated a novel method to quantify the in vivo t½ of p62 in skeletal muscle and have proof of concept that changes in p62 t½ reflect autophagic activity. We have identified p62 via tandem-MS in human skeletal muscle and begun to orally label humans subjects with stable isotopes. The current proposal will be translated to human patients allowing us to be able to quantify the kinetics of autophagic dysfunction in human muscle disease and normal aging.

PI: Simon Tang, PhD

In Vivo Contrast-enhanced MicroCT Imaging of the Murine Intervertebral Disc

Specific Aims

Degeneration of the intervertebral disc (IVD) is a leading contributor towards back pain, an epidemic that costs billions of dollars in the US [1]. The IVD consists of a proteoglycan(PG)-rich nucleus pulposus (NP) surrounded by a collagenous annulus fibrosus (AF) [2] that together provide support and transmit complex loads [3,4]. The IVD degenerative cascade involves a multifactorial progression of biological, biochemical, and structural changes that lead to the collapse of the disc structure and to compromised mechanical function [3-5]. Despite its significant public health impact, the pathophysiology of disc degeneration remains unclear.

The lack of high-resolution imaging tools to monitor disease progression is one of the key hurdles in understanding the pathophysiology of IVD degeneration. Although magnetic resonance imaging (MRI) is commonly used to quantify IVD morphological and compositional changes in vivo for humans and large animal models [6], the relatively low resolution of MRI (hundreds of microns to millimeters) limits its sensitivity towards sub-millimeter and nuanced changes that may occur with the initiation of degeneration. Higher resolution imaging methods may improve the detection fidelity of the early degenerative process and the ability to leverage rodent models to understand the genetic and mechanobiologic interactions in intervertebral disc homeostasis.

Recent investigations have shown that micro-CT in the presence of contrast agents improves the discrimination and enables the 3D evaluation of unmineralized soft tissues. Ionic contrast agents such as ioxaglate (Hexabrix) proportionally binds to the negatively charged glycosaminoglycans (GAGs) in soft tissue such as cartilage [7,8], and the tissues are revealed by the differential uptake of contrast agent as detected by micro-CT. While ionic contrast agents are highly specific for charged tissues, they tend to be hydrophobic [9] and increase the propensity for in vivo endothelial injury [10] and renal nephropathy [11,12]. Nonionic, low osmolar, and hydrophilic contrast agents such as Ioversol (OptiRay) significantly reduce the complications [9,11] without compromising radiodensity, making them more suitable for in vivo imaging. No published studies to date have examined the effectiveness of these nonionic, hydrophilic agents for contrast-enhanced microCT imaging of the intervertebral disc in rodents.

Our laboratory recently explored the use of Ioversol for contrast-enhanced micro-CT imaging of rodent intervertebral discs. Taking advantage of the high-resolution of microCT (~10 microns) and the increased tissue hydration of the NP over the AF, Ioversol proves to be ideal for nondestructive 3D visualization of the IVD. Moreover, we have developed a set of parameters to distinguish the NP and the AF, and to quantify the intervertebral disc structure and proteoglycan content of the disc (NP Volume Fraction, NP Intensity Fraction, etc). Our preliminary ex vivo studies quantitatively captured the nuanced changes in tissue- and structurallevel changes due to trypsin-induced degeneration of the IVD [13,14]. Building on this work, we will investigate the in vivo longitudinal imaging of murine IVD using contrast-enhanced microCT with the overall hypothesis that the IVD degenerative cascade initiates dose-dependently with enzymatic insult.

• Aim 1: Compare and optimize the timing and efficacy of Ioversol for in vivo intervertebral disc structural imaging in the rat. We will directly deliver the contrast agent to the caudal intervertebral disc of Sprague Dawley rats and image these discs in vivo over 48 hours to determine the effects of local uptake and clearance of Ioversol on x-ray attenuation. Delivery of Ioversol will be achieved through direct injection to the caudal disc of interest, and adjacent discs will serve as internal controls.
• Aim 2: Longitudinally monitor spatial-temporal structural and biochemical changes with trypsin-mediated intervertebral disc degeneration and subsequent adaptation over 3 weeks. Trypsin, a serine protease that catalyzes the hydrolysis of peptide bonds such as those on sulfated GAGs, has been used to chemically degenerate the IVD to induce herniation [15] and is commonly used for the ex vivo validation of IVD MRI techniques [13,14,16,17], yet the spatial-temporal progression of IVD degeneration and adaptation remain relatively unknown. Using our developed technique to monitor the effects of low-, medium-, and high- doses of trypsin injected into the IVD, we will capture structural and compositional changes locally within the IVD, giving insight to the degenerative and regenerative response over time due to varying degrees of nucleolytic insult.

The research proposed here, led by a junior investigator, will heavily utilize the Histology and the Structure/Strength cores of the Musculoskeletal Research Center. The successful completion of these aims will validate a powerful new tool for analyzing the progression and treatment of intervertebral disc degeneration in small animal models, as well as provide critical preliminary data towards future NIH submissions.

PI: Michael J. Gardner, MD

The Effects of Systemic Hedgehog Pathway Modulation on Fracture Healing

Specific Aims

Five to ten percent of the 5.6 million fractures that occur annually in the United States show evidence of impaired healing,1 and these place a significant burden on society. Therefore, elucidation of novel strategies to promote osteogenesis is critical and may potentially benefit a large number of patients. BMP-2, BMP-7, and teriparatide have been developed for clinical use as therapeutics that target inherent biological pathways, but each has disadvantages that have limited their widespread use. The BMP’s, for example, require surgery for direct implantation and local delivery, and the optimal dose and delivery remain unclear. The hedgehog (Hh) signaling pathway is critical to developmental osteogenesis2, 3, chondrogenesis4-6, and bone homeostasis7. Further, the Hh pathway is known to regulate angiogenesis, which may in part explain its role in osteogenesis8-10. These observations suggest a possible central role for Hh signaling in fracture healing, which involves coordinated processes of chondro-, angio- and osteogenesis. Although Hh upregulation has been indirectly associated with fracture healing,11-13 the functional role of Hh signaling in fracture healing in not known. Preliminary studies in our lab indicate that inhibition of Hh signaling impairs non-endochondral bone formation in stress fractures, although it is not known whether a similar effect occurs for endochondral bone formation in complete fractures. On the other hand, activation of Hh signaling using SAG, a small molecule agonist of Smoothened (Smo), through which all Hh signals are transduced, upregulates early markers of bone formation in vitro14. Whether activation of Hh signaling enhances fracture healing has not been studied. Understanding the role of Hh signaling in osteogenesis holds substantial promise in providing therapeutic targets for novel drugs to be used by surgeons to enhance bone healing. Our proposed experiment is designed to provide proof-of-concept for developing biological interventions to augment fracture healing using the Hh pathway. As such, this study is ideally suited for a Pilot & Feasibility Study, from which the data may be used as a basis for a comprehensive, federally-funded grant. Our findings will potentially provide a rationale for larger scale and more mechanistic studies targeting the Hh pathway with pharmacologic agents to improve bone healing in challenging clinical situations. Our central hypothesis is that Hh signaling can be modulated to affect fracture healing. Although preliminary studies are promising and published work is supportive, based on current gaps in our knowledge of this pathway, the proposed study is necessary to support or refute the hypothesis. We will test this hypothesis using a murine, stabilized femoral fracture model that is established in our hands.

• Aim 1: Determine the effect of activation of the Hh pathway on fracture healing in a healing-challenged model. In this aim, we will test the hypothesis that fracture healing can be enhanced by systemically activating the Hh pathway in a healing-challenged fracture model. Mice will receive treatment with either vehicle or SAG, a potent Hh agonist. We will create femur fractures in aged (18 month old) mice. Older rodents demonstrate less robust fracture healing, and of particular relevance have a lower and delayed peak of Hh expression in fracture healing15, 16. We hypothesize that SAG will overcome the age-related decline in Hh signaling and improve radiographic, histologic, biomechanical, and vascular aspects of fracture healing.

PI: Lilianna Solnica-Krezel, Ph.D.

Genetic Basis of Musculoskeletal Disease: Emphasis on Zebrafish Models of Late-Onset Scoliosis and Tendon Development

Specific Aims

In humans, adolescent idiopathic scoliosis (AIS) is defined as a lateral curvature of the spine, free of vertebral malformations, with unknown etiology and occurring near the onset of puberty (Weinstein et al., 2008). Analysis of pedigrees and genome wide association studies (GWAS) of AIS patients are suggestive of multiple chromosomal loci exhibiting both dominant and recessive characteristics (Wise et al., 2008). Genetically tractable animal models of AIS will facilitate the discovery of the genetic, cellular, and tissue level factors that contribute to the pathology of late-onset scoliosis. The zebrafish, Danio rerio, is one of the preeminent vertebrate model systems for modeling disease pathology in a variety of organs. For example, the zebrafish is a robust system for both forward- and reversegenetic approaches, allowing for the discovery of genes essential for a process of interest, or for targeted gene disruptions, respectively. In addition, the large collection of zebrafish mutants and transgenic lines, along with the external development of numerous transparent embryos, afford effective dissection of molecular genetic pathways. The goal of this proposed investigation is to further the discovery of the genetic causes of skeletal deformities – with particular emphasis on late-onset scoliosis – using zebrafish. In addition, we propose to develop novel transgenic animals that will allow us to visualize the development and homeostasis of tenocytes/tendons in vivo. Importantly, we generated a novel dominant zebrafish mutant, Druk, displaying many corollaries to AIS in humans, including progressive curvature of the axial skeleton starting in larval fish (~15 days post fertilization), without any malformations of initial vertebrae development. This model of lateonset scoliosis provides us with a defined time window to focus on new screening efforts for additional scoliosis mutants. We already have established modern fish facilities, and robust screening methods to uncover both recessive and dominant mutations causing scoliosis. We also have generated large numbers of highly mutagenized founder fish that will afford an effective screen for new scoliosis phenotypes. In fact, during our pilot screen (~3 months), we have uncovered one dominant mutant exhibiting late-onset scoliosis (Stl24); two dominant phenotypes exhibiting vertebral fusions and dwarfism (Squat 1; Squat 2); and two putative recessive late-onset scoliosis mutations (ep71; ep108). The elucidation of the genetic lesions underlying these mutant phenotypes will illuminate the cellular and molecular etiology of late-onset skeletal defects. The formation of scoliosis is hypothesized to involve structural defects of musculoskeletal tissues; the bone, muscles or connective tissues. Interestingly, the surgical resection of myotendinous connections of the posterior ends of caudal ribs, including the costo-vertebral joints, can induce scoliosis in young rabbits (Langenskiold and Michelsson, 1961). In addition to a potential role in the formation of scoliosis, the development and homeostasis of tendon-bone insertions is a critical topic in developmental biology. We posit that the zebrafish model provides a promising model with which to study the development of tendons in realtime due to the ease of direct, long-term in vivo imaging available in the externally developing and optically transparent zebrafish embryos. The development of fluorescent tendon marker transgenics will greatly facilitate dissecting the normal development of tendons and the mechanisms of tendon-bone interactions during injury and repair. Finally, the establishment of these tendon-marking transgenics will enable effective genetic screening for the pathways and molecular agents that direct the development and maintenance tendons in a vertebrate organism. Here, we propose to build upon our initial investigations of the late-onset scoliosis zebrafish model, as well as to test a hypothesis involving the role of tendons in the development of late-onset scoliosis. We propose to address the following specific aims in this pilot study:

• Aim 1: We will carry out phenotypic and molecular characterization of novel mutants presenting with musculoskeletal defects, with emphasis on models of late-onset scoliosis.
  • Aim 1A: Determine the genetic lesions associated with mutants found in the late-onset scoliosis screen using next generation sequencing at the Genome Technology Access Center (WUSM).
  • Aim 1B: Assay the quality of the vertebral skeleton in late-onset scoliosis mutants using both histological and μ-CT imaging techniques at the Musculoskeletal Research Center (WUMRC).
• Aim 2: Generate tendon-specific transgenic zebrafish lines using BAC recombineering targeting genes that are essential for tendon development and repair. Our pilot study will identify conserved genes controlling the formation of late-onset scoliosis. Moreover, it will establish tools for imaging tendon development in wild-type and late-onset scoliosis mutant animals.

PI: M. Farooq Rai, Ph.D.

Study of Early Responses in Knee Joint Following Compressive Tibial Loading in Genetic Mouse Strains

Specific Aims

This proposal seeks to answer a critical question in osteoarthritis (OA) research: is the susceptibility to posttraumatic OA (PTOA) functionally related to susceptibility to injurious compressive loading in a non-invasive model of knee joint trauma? Here, we aim to utilize a novel non-invasive compressive tibial loading model in genetic mouse strains to understand the short- and long-term consequences of injury on articular cartilage, bone and synovium from a genetic standpoint. In addition, we will get mechanistic insights into the early molecular differences between the two strains towards understanding the pathophysiology of loading on chondrocyte apoptosis, matrix distribution and synovial cell proliferation. To achieve these targets, this proposal has two specific aims:

• Aim 1: Determine the congruence in phenotypic differences in articular cartilage, synovium and bone following impact loading on the joint in genetic mouse strains. Hypothesis 1: The strain protected from PTOA has ability to withstand injurious compressive loading. Rationale: Although the relationship between PTOA and impact of compressive loading in genetic mouse strains is unknown, we assume that a strain that is protected from developing PTOA is able to better withstand the impact of injurious loading as will be determined by changes in cartilage, synovium and bone.
• Aim 2: Test the effects of compressive loading on chondrocyte apoptosis, matrix distribution and synovial cell proliferation across strains. Hypothesis 2: Compressive loading modulates chondrocyte apoptosis, matrix distribution and synovial cell proliferation in the knee joint tissues. Rationale: Using C57BL/6J mice, we have observed that various loading regimens induced chondrocyte apoptosis and cartilage matrix degradation along with disruption of collagen fibril arrangement. We would study these changes in LGXSM-6 and LGXSM-33 strains to stratify the differences in response to compressive loading.

PI: Michelle Hurchla, Ph.D.

The Role of Stromal Senescence in MGUS and Multiple Myeloma

Specific Aims

Multiple Myeloma (MM) results from the growth of malignant antibody-secreting plasma cell clones within the bone marrow (BM) and is associated with debilitating osteolytic bone lesions. Monoclonal gammopathy of undetermined significance (MGUS), the pre-neoplastic condition preceding all myelomas, is a common disease of aging, occurring in 3.2% and 5.3% of individuals >50 or >70 years old1, respectively. Importantly, MGUS is associated with osteoporosis and increased fracture rate2. The biological basis for the transition from relatively benign MGUS to invariably fatal MM remains unknown. Interestingly, nearly all of the tumoral genetic changes identified in MM are also found in MGUS, suggesting that another factor such as the microenvironment plays a role in disease progression. Senescence, the cellular aging process in which cells lose the ability to divide, is recognized to be a driver of aging associated diseases including cancer, arthritis and arthrosclerosis. The Stewart lab and others have found that senescent fibroblasts produce numerous growth factors and cytokines referred to as the senescent associated secretory phenotype (SASP) that can promote tumor development. Notably, IL-6 is one of the most abundant SASP factors and is also a key factor in myeloma growth. We hypothesize that activation of the senescence-associated secretory program in patient bone marrow stromal cells is a critical switch that causes relatively innocuous MGUS to progress to bone-destructive MM.

• Aim 1: Do senescent BMSC support myeloma cell growth and tumor associated bone loss?
• Aim 2: Can engineered senescence in bone enhance myeloma growth and associated bone loss? Bone marrow stromal cells (BMSC) are a phenotypically heterogeneous population that includes mesenchymal stem cells (MSC), the progenitors of osteoblasts (OB). BMSC regulate both osteoclasts (OC) (production of MCSF and RANKL) and support myeloma growth (adhesion and growth factor production).

Dr. Hurchla, recently promoted to Instructor, is developing her independent research program to study microenvironmental control of myeloma bone disease. The Weilbaecher Lab is recognized as an expert in osteolysis induced by metastatic breast cancer and melanoma. The Stewart Lab is a leader in the study of how cellular aging supports tumorigenesis. They have developed novel methods to induce and identify SASP and have developed a mouse in which stromal senescence in bone can be established in young mice (FAAST mice). Using a genetic approach, the Tomasson Lab has generated a short list of candidate genes, including some with roles in bone biology responsible for MGUS/MM development in the genetically uncharacterized KaLwRij mouse strain. Drs. Michael Tomasson and Ravi Vij have established a MM/MGUS tumor bank and in collaboration with the Weilbaecher Lab have produced primary MM patient BMSC that will be used in this proposal. We have established this collaborative group to utilize the unique expertise and resources of each member to interrogate the complex interactions between stromal senescence, myelomagenesis and osteolytic disease. 

PI: Wei Zou, MD, Ph.D.

ASXL Proteins Regulate Bone Resorption

Specific Aims

Rosiglitizone (BRL) is an insulin-sensitizing drug which exerts its effect by activating the transcription factor PPARγ. While BRL is probably the most effective oral treatment of type II diabetes mellitus (DM II) it carries complications including increased fracture risk. This predisposition to fracture is consistent with the fact that PPARγ preferentially promotes formation of adipocytes at the cost of osteoblasts. Surprisingly however, BRLactivated PPARγ also stimulates osteoclast formation. We have discovered that the nuclear protein ASXL2 is essential for PPARγ -stimulated osteoclast formation. Specifically, osteoclastogenesis is arrested in the absence of ASXL2, in vitro, and mice lacking the protein have osteoclast-autonomous osteopetrosis. We have also established that like PPARγ, ASXL2 accelerates osteoclast formation by promoting mitochondrial biogenesis and increasing RANKL-induced c-Fos expression which in turn stimulates NFATc1. Importantly, BRL-enhanced osteoclastogenesis is completely arrested absent ASXL2. Given that in other cells, ASXL2 interacts with PPARγ, BRL-induced bone loss, in the context of DM II, may entail ASXL2 regulation of the nuclear receptor’s transcriptional activity. ASXL1 is a paralog of ASXL2 which also recognizes PPARγ in fat cells. However, whereas ASXL2 promotes BRL-stimulated PPARγ activation, in adipocytes, ASXL1 inhibits it. As BRL exerts its osteoclastogenic properties via PPARγ we hypothesize, that similar to adipogenesis, the effects of ASXL1 deletion on osteoclast formation and function, in vitro and in vivo, are opposite those of ASXL2.

• Aim 1: determine the effects of ASXL1 deletion on basal and PPARγ -mediated osteoclast formation and function in vitro and in vivo.

PI: Craig A. Micchelli, Ph.D.

Establishing A Novel Genetic Model System to Elucidate Conserved Mechanisms Controlling Adult Muscle Stem Cells

Specific Aims

The goal of the proposed investigation is to establish an experimental model to study stem cell-based control of adult muscle homeostasis in the genetically tractable system, Drosophila melanogaster. Many of the key cellular, molecular and physiological hallmarks of muscle biology are conserved between invertebrates and mammals (Augustin and Partridge, 2009). Therefore, if successful, our pilot study will pave the way for subsequent identification of conserved genes controlling the process of adult muscle homeostasis in vivo, using the unsurpassed molecular genetic screening methodologies available only in the fruit fly.

Adult tissue homeostasis often depends on a single cell type with the capacity to act as a reservoir of renewal potential, replacing old cells that are lost through injury or disease (Morrison and Spradling, 2008). Given the hierarchical organization inherent in stem cell lineages and the reiterative use of conserved signaling pathways in a lineage, it is essential to identify and manipulate adult stem cells with single cell resolution, in order to precisely dissect the molecular control of adult stem cells. However, due to their scarcity, adult stem cells in many tissues are often difficult to identify in vivo (Buckingham and Meilhac, 2011).
In the past few years, the adult fruit fly has emerged as a powerful model organism to identify and study a wide range of different adult stem cell types in their native microenvironments or “niches” using cutting edge molecular genetic methodologies (Casali and Batlle, 2009). These now include stem cells of the germ line, brain and gastrointestinal tract with subsequent studies demonstrating striking conservation in their regulation. Remarkably, however, adult muscle stem cells analogous to mammalian “satellite cells” have not yet been identified in adult Drosophila, despite the persistence of significant muscle mass into adulthood (e.g. flight muscles).
Here, we propose to build upon a unique set of resources and expertise that has accrued in my laboratory during the course of successfully identifying stem cells in other adult Drosophila tissues (Micchelli and Perrimon, 2006; Strand and Micchelli, 2011), to now test the hypothesis that myogenic stem cells maintain muscle homeostasis in the adult fruit fly. We propose to address the following specific aim in this pilot study:

• Aim 1: We will determine the extent to which myogenic stem cells support adult muscle maintenance in the adult fruit fly, Drosophila melanogaster.
  • Aim 1A: Determine if adult muscles in Drosophila regenerate following injury.
  • Aim 1B: Distinguish whether or not the reservoir of renewal capacity in adult muscle resides in an undifferentiated myogenic stem cell compartment or among differentiated muscle cells.

PI: William Frazier, Ph.D.

The Roles of CD47 in muscle homeostasis and mitochondrial regulation and in skeletal homeostasis.

Specific Aims

CD47 is a widely expressed receptor for the matricellular protein thrombospondin-1 (TSP1) and for the immunoregulatory counter-receptor SIRPα (signal inhibitory regulatory protein-α) (Brown and Frazier, 2001). While mice genetically knocked out for either TSP1 or for CD47 initially displayed no remarkable phenotypes, my group along with several collaborators, have discovered a number of interesting musculoskeletal phenotypes. With Dr. Katherine Weilbaecher’s group we found that both TSP1 and CD47 play a role in bone homeostasis (Uluckan et al., 2009). In knockouts of either TSP1 or CD47, bones have increased trabecular volume and bone mineral density. CD47 null bone marrow macrophages are severely impaired in their ability to differentiate into osteoclasts (OCs) in vitro, but increasing RANKL levels or reducing NO levels with L-NAME treatment permits their differentiation into OCs. Furthermore, CD47 null mice are protected from tumor metastasis to bone in a model of left ventricle administration of melanoma cells. This data supports the “vicious cycle” model of bone destruction by tumors, however all cells in the host animal are lacking CD47 and therefore a specific role for CD47 on OCs cannot be assumed; recent data suggests that CD47 influences lineage decisions of mesenchymal stem cells as well (Uluckan, Weilbaecher and Frazier, in preparation).

In addition to its role(s) in bone, CD47 affects mitochondrial biogenesis/turnover in both skeletal and cardiac muscle. CD47 limits both the cGMP and cAMP signaling pathways, both of which can regulate mitochondrial biogenesis. Therefore we screened expression of mitochondria-related genes in several tissues of young CD47-null and WT mice. Only skeletal muscle had markedly elevated levels of such genes. Skeletal muscle from CD47 knockouts displayed a markedly increased volume of mitochondria compared to WTs and fast twitch muscle had undergone fiber type switching to a slow twitch phenotype as occurs with exercise training and calorie restriction. Consequently, young CD47-null mice have twice the exercise endurance of wild type mice. The CD47 nulls utilize less oxygen and produce dramatically less ROS than WTs. When we profiled gene expression in tissues as a function of age, we found that, by 1 year of age, CD47 null skeletal muscle mitochondrial density had declined to WT levels. To our surprise, CD47 null hearts of aging mice develop dramatically increased mitochondrial density that correlates with improved heart function. Thus CD47 plays a role in both skeletal and cardiac muscle function as well as in bone physiology.

In order to sort out the complex phenotypes of CD47-nulls in muscle and bone, each of which contain multiple cell types derived from diverse precursor and stem cell lineages, we propose to generate mice harboring a floxed CD47 gene that, when crossed with mice harboring Cre recombinase under the control of tissue-specific promoters, will allow us to delete CD47 expression in a cell- or tissue-specific fashion. The knockout mouse project (KOMP) has successfully produced ES cells harboring a floxed CD47 gene. We will use all three components of the musculoskeletal core in the following Specific Aims:

• Aim 1: Utilize the mouse genetics component of the musculoskeletal core to generate genetically targeted mice harboring the floxed CD47 gene. Floxed CD47 ES cells have been ordered from KOMP.
• Aim 2: Test the functionality of the floxed CD47 gene, floxed CD47 mice will be mated with LysM promoter-Cre mice to disrupt CD47 expression in myeloid cells, including osteoclasts. The LysM-Cre mice are available from the Weilbaecher lab (Morgan et al., 2009). We will access the in situ analysis core to help characterize the tissue-specific CD47 knockouts for CD47 expression, histology and histomorphometry of bone.
• Aim 3: Bone structure and density (DEXA, microCT) as well as strength testing of the conditional knockouts will be performed with the structure and strength core.

Once the floxed CD47 line is established and verified, we will mate the mice with other promoter-Cre lines to generate tissue-specific CD47 nulls in mature osteoclasts, osteoblasts, platelets, skeletal, smooth and cardiac muscle, and reticulocyte/erythrocytes. Using the CD47 global knockouts, we and others have obtained data implicating CD47 in functions of all of the above cell and tissue types. Only targeted knockouts can test the cell or tissue autonomous function of CD47. Therefore, this pilot project will lay the essential groundwork for many additional NIH-funded projects to come, thereby fulfilling the purpose of this pilot program.

PI: Gabriel Mbalaviele, Ph.D.

NLRP3 Expression in Myeloid Cells is Sufficient to Stimulate Bone Resorption

Specific Aims

The Nod-like receptor protein 3 (NLRP3) inflammasome is an intracellular protein complex responsible for the maturation of several members of interleukin (IL)-1 family, including the osteotropic factors, IL-1b and IL-18. NLRP3 is also involved in the inactivation of poly(ADP-ribose) polymerase 1 (PARP1), a negative regulator of osteoclast (OC) development [1-3]. NLRP3 is activated by danger-associated molecular patterns (DAMPs) such as crystals, wear particles and degradation products of endogenous components of the extracellular matrix (ECM) [4].

Trafficking of both organic and inorganic degradation products from bone ECM through bone-resorbing OC to the extracellular milieu is a critical step during bone resorption [5, 6]. This process enables OC to excrete degraded matrix components while digging deep into bone and maintaining an enclosed resorption site. Thus, the products of bone resorption are potential activators of NLRP3 in OC. Interestingly, patients with neonatal-onset multisystemic inflammatory disease (NOMID), a condition linked to NLRP3-activating mutations [7], exhibit skeletal malformations and low bone mass in up to 60% of the cases [8]. Our own preliminary data demonstrates that mice globally expressing a constitutively active NLRP3 mutant, which model the human NOMID syndrome, have severely decreased bone mass and increased bone resorption, leading to stunted post-natal growth. Thus, there is now strong evidence from human and mouse genetics that NLRP3 is critically involved in normal skeletal homeostasis. However, little if any is known about the role of NLRP3 in bone homeostasis. We hypothesize that 1) NLRP3 activation in myeloid cells is important for bone resorption, and 2) bone ECM degradation products function as endogenous DAMPs for NLRP3 in the OC lineage. Accordingly, we propose to test these hypotheses in the following Specific Aims:

• Aim 1:  Determine whether NLRP3 activation in myeloid cells is sufficient to stimulate bone resorption.
• Aim 2:  Identify bone-relevant DAMPs and study the regulatory mechanisms of NLRP3 activation in OC.

The peri-natal lethality of pups globally expressing mutant NLRP3 hampers the efforts to elucidate the role of NLRP3 in bone. To gain further insight into the role of NLRP3 in bone resorption, we propose to use mice in which the expression of mutant NLRP3 is restricted to the myeloid cell lineage. These mice are available to us from our collaborator, Dr. Hoffman (Univ of Calfornia at San Diego). The funding requested will support the characterization of these mice to advance this project to the point where a comprehensive data package is generated for future NIH applications. We anticipate that these studies will unravel a novel communication mode among bone cells and with their environment via NLRP3, and reveal this upstream component of the IL-1 family and PARP1 pathways as a potential new target for therapeutic intervention in bone diseases.

PI: Audrey McAlinden, Ph.D.

Generation of a knock-in mouse model to study the role of type II collagen alternative splicing during chondrogenesis

Specific Aims

Generation and analysis of recombinant knock‐in mice expressing only one isoform of type II collagen.

Rationale: Alternative splicing of the type II collagen gene (Col2a1) occurs in a developmentally‐regulated manner and only in chondrogenic tissues. Chondroprogenitor cells synthesize Col2a1 mRNA isoforms containing exon 2 while differentiated chondrocytes generate Col2a1 isoforms devoid of exon 2. The significance of this alternative splicing switch in skeletal development is not known.

Hypothesis: We hypothesize that Col2a1 alternative splicing is necessary to generate proper cartilage tissue which subsequently results in normal endochondral bone formation.

Approach: We plan to inhibit the Col2a1 developmentally‐regulated splicing switch in vivo by generating knock‐in mice that will only synthesize type IIA collagen (the exon 2‐containing isoform normally produced by chondroprogenitor cells) while production of type IIB collagen (the isoform devoid of exon 2 normally produced by differentiated chondrocytes) will be completely inhibited. This mouse model will directly address the role of Col2a1 alternative splicing in chondrogenesis and will aid in defining functional roles of these distinct protein isoforms. Recently, we have identified and confirmed positive ES clones that will be used for blastocyst injections. Therefore, in this one year Pilot & Feasibility Study Proposal, we aim to generate the knock‐in animals, carry out preliminary analyses of embryos and post‐natal mice and develop an in vitro based ES cell differentiation assay system. Services provided by the Mouse Genetics Models Core D and the in situ Molecular Analysis Core C (affiliated with the Center of Musculoskeletal Research) will be used extensively. We predict the knock‐in mouse model will provide new and important insights into the mechanisms involved in regulating cartilage and bone development. This proposal therefore provides a unique opportunity to permit the progress of this currently unfunded research project and to generate strong preliminary data for a future RO1 application.

PI: Robert Mecham, Ph.D.

MAGP1: An ECM regulator of Bone Remodeling

Specific Aims

In bone, the collagens have been the traditional focus of ECM studies and, as a consequence, the contribution of other ECM elements is less well understood. One such non-collagenous ECM constituent is the fibrillin-containing microfibril. Though microfibrils are less abundant than their collagen counterparts, their significance has been demonstrated clinically. Disruption of fibrillin microfibrils leads to severe musculoskeletal disease affecting the axial skeleton, appendicular skeleton, muscularity and adiposity. Microfibrils are known to regulate cellular activities by controlling the bioavailability of numerous growth factors, particularly those of the TGF-β family. The core components of the microfibril are the fibrillins and microfibril associated glycoproteins (MAGPs), both of which are capable of binding growth factors. Inactivation of the MAGP1 gene in mice (MAGP1Δ) results in osteopenia, slight long bone overgrowth, significantly less trabecular and cortical bone, and altered trabecular microarchitecture. Our preliminary data suggests that MAGP1 antagonizes osteoclast-mediated bone resorption by influencing cellular differentiation, survival and function. The objective of work outlined in this proposal is to obtain more preliminary data to support this hypothesis. The experimental models that will be developed will provide a foundation for future mechanistic studies of the role MAGP1 plays in bone cell biology. Our specific aims are:

• Aim 1:  Determine the temporal and spatial appearance of MAGP1 in developing bone. To fully understand what role MAGP1 might play in the bone requires knowing when and where it appears in bone differentiation. This aim will utilize in situ hybridization, immunohistochemistry, and RNA quantitation to define the spatial and temporal expression profile of the MAGPs and fibrillins during bone development.
• Aim 2:  Explore MAGP1-mediated coupling of osteoblasts and osteoclasts in bone remodeling. Bone marrow macrophages produce insignificant amounts of MAGP1 relative to osteoblasts. The importance of osteoblast-derived MAGP1 for osteoclast differentiation and function will be studied using co-cultures of WT (or KO) BMMs with WT and KO osteoblasts. In vivo experiments using radiation chimeras (RC) will be used to complement the ex vivo co-culture studies. RCs will be generated by engrafting γ-irradiated WT mice with marrow from KO mice (and vice versus). The RCs will then be challenged and bone loss assessed. Finally, we will test whether MAGP1’s presence in the matrix influences homing of hematopoietic stem cells (or more specifically monocytes) to the bone by extracting and labeling marrow cells from donor mice, i.v. injection of labeled cells into both WT and MAGP1Δ mice, and tracking the localization of these cells.
• Aim 3:  Develop a MAGP1 conditional knockout mouse. To confirm that the bone phenotype in the MAGP1-null mouse is due to the absence of MAGP1 from bone and not influenced by a factor extrinsic to bone, our lab has begun the process of generating a conditional MAGP1 mouse with the MAGP1 gene floxed by LoxP sites. Crossing these mice with strains expressing a bone cell-specific Cre will enable us to inactivate MAGP1 only in bone and at different times during bone development.

PI: Younan Xia, Ph.D.
Co-PI: Jingyi Chen, Ph.D.
Co-Investigator: Jingwei Xie, Ph.D.
Co-PI: Stavros Thomopoulos, Ph.D.

Putting Electrospun Nanofibers to Work for Musculoskeletal Research

Specific Aims

This project is a collaborative effort involving investigators from the Departments of Biomedical Engineering and Orthopaedic Surgery at Washington University in St. Louis. The major goal of this project is to develop and validate electrospun nanofibers as a new platform for musculoskeletal research, with an initial focus on the demonstration of hierarchical scaffolds for tissue engineering at the tendon-to-bone insertion site.

Owing to the high porosity and large surface area, a scaffold derived from electrospun nanofibers can mimic the hierarchical structure of extracellular matrix (ECM) critical for cell attachment and nutrient transportation. The nanofibers can be routinely prepared from a broad range of biocompatible and biodegradable polymers (both natural and synthetic), as well as composites containing inorganic solids such as hydroxyapatite (HA). The nanofibers can also be conveniently functionalized via encapsulation or attachment of bioactive species such as ECM proteins, enzymes, DNAs, and growth factors to control the proliferation and differentiation of cells seeded on the scaffolds. In addition, the fibers can be assembled into a variety of arrays or hierarchically structured films by manipulating their alignment, stacking, and/or folding. All these attributes make electrospun nanofibers a class of enabling materials for biomedical research, with notable examples including tissue engineering, targeted delivery, and controlled release of biofactors. Our global hypothesis is that fibrous scaffolds can be engineered with specific structural order, surface chemistry, degradation profile, mineral composition, biomechanical property, and bioactivity for manipulating the attachment, proliferation, and differentiation of cells and thus serve as a new framework for repairing the tendon-to-bone insertion site and other musculoskeletal tissues.

• Aim 1:  Fabricate nanofibers by electrospinning and then optimize their properties for use as scaffolds for tissue engineering by in vitro cell culture studies. We will initially focus on poly(lactic-co-glycolic acid) (PLGA), whose mechanical properties and degradation profiles can be controlled by varying the copolymer composition. In addition to the use of fibers with various diameters, we will compare two different types of scaffolds constructed from randomly oriented and uniaxially aligned nanofibers, respectively. In one set of experiments, fibroblasts isolated from rat rotator cuff tendons will be used to investigate the proliferation and morphology of cells seeded on these scaffolds. In another set of experiments, mesenchymal stem cells (MSCs) derived from rat bone marrow will be used to examine how their proliferation and differentiation can be manipulated by engineering the properties of the scaffolds. We will systematically evaluate the biomechanical and cellular properties of these scaffolds in an effort to develop an optimal system for the in vivo animal study.
• Aim 2:  Evaluate the performance of the engineered nanofibrous scaffolds in vivo using a clinically relevant animal model for rotator cuff repair. The scaffolds will be seeded with the fibroblast cells and then patched onto the tendon-to-bone repair site. The scaffold is expected to serve as a graft to provide mechanical stability and potentially to guide cell activity during the healing process. We will test and compare three different types of scaffolds: randomly oriented, uniaxially aligned, and specially designed with a gradation in the alignment. We will systematically investigate the progress of healing using structural, compositional, and biomechanical assays at the Washington University Core Center for Musculoskeletal Biology and Medicine.