(296g) Combinatorial Model Building and Selection of Hypothesized Mechanisms Influencing Bone Health

Introduction: Bone health is important for everyday function and mobility. Bone health is regulated by the bone remodeling cycle that replaces older bone with new bone in a balanced process of bone resorption and formation. When this cycle is chronically thrown out of balance, diseases such as osteoporosis, Paget's disease, and osteoarthritis occur. The bone remodeling cycle has been shown to be influenced by several factors generated outside the bone compartment, such as estrogen levels, dietary changes, and immune responses [1,2]. Our previous work focused on the influence of a dietary change and immune response to bone remodeling using a mechanistic model of T regulatory cells and Wnt-10b. However, the results indicated that other interactions were also occurring [3,4]. This has motivated our lab to explore the multifaceted mechanisms influencing overall bone health.

Methods: In this work, we are exploring how factors generated outside the bone compartment influence bone remodeling and bone health. We are creating a literature curated database of interactions from experimental studies that identify a connection between a specific external factor and a corresponding bone remodeling function or cell type. For example, one entry in our database includes the influence of external factor short-chain fatty acid (SCFA) on bone remodeling points to a direct connection to osteoclasts with no changes to osteoblasts [5].

From this database, we are generating multiple hypotheses centered around mechanisms that control bone homeostasis and bone diseases. For example, using the previously mentioned study, we hypothesize that SCFAs impair osteoclast development only; however, another study shows that SCFAs can influence osteoblast activity [6]. So we are testing the hypothesis that states that SCFAs increase osteoblast activity. In addition, we have a new hypothesis stating that SCFAs can influence both osteoclasts and osteoblasts. These hypotheses will be used to compose mechanistic candidate submodels that can be easily incorporated into a computational model of bone remodeling. To test our hypotheses, we follow the hypothesis testing method presented in a work focusing on the combinatory action of macrophages during bone healing, a closely related topic to bone remodeling [7]. We will select specific combinations of the ODE based candidate submodels to incorporate into our previously developed remodeling model and then determine if the hypothesis provides a promising mechanism. This method combines residuals and the Akaike information criterion to compare the hypothesized mechanism to the experiments and other hypothesized mechanisms [7]. This comparative process helps identify the most explanatory mechanisms for a particular external influence.

We are focusing a large portion of this exploration on osteocyte interactions with external influences, as osteocytes have recently been identified as a main controller of the remodeling cycle [8]. Currently, the osteocyte mechanism most often considered in computational models is the relationship between the Wnt activation pathway and the Wnt antagonist sclerostin. So, there is a need for an updated model that considers more osteocyte mechanisms, hence our focus on this cell type.

Results: We are motivated to complete this study using the described methodology because it will give us multiple desirable outcomes. First, we will have compiled a large database of mechanistic candidate submodels that can be strategically combined to explore a variety of bone remodeling factors and bone diseases. Second, we will identify the most promising mechanisms that explain some of the conflicting experimental relationships to different external factors. Lastly, we will provide a foundation for an updated model of bone remodeling that considers more osteocyte-based mechanisms. All of these outcomes support bone health research using systems engineering model identification and discrimination techniques.

Acknowledgment: This work was supported by the National Institutes of Health grants R35GM133763 and R15AT010725 and the University at Buffalo.

References:

[1] Pacifici, et al., Role of T cells in ovariectomy induced bone loss—revisited, J Bone Miner Res (2012).

[2] Tyagi et al., The microbial metabolite butyrate stimulates bone formation via T regulatory cell-mediated regulation of WNT10B expression, Immunity (2018).

[3] Islam, et al., Mathematical modeling of the gut-bone axis and implications of butyrate treatment on osteoimmunology, Ind Eng Chem Res (2021).

[4] Cook, et al., Mathematical modeling of the effects of Wnt-10b on bone metabolism, AIChE J (2022).

[5] Lucas, et al., Short-chain fatty acids regulate systemic bone mass and protect from pathological bone loss, Nat Commun (2018).

[6] Kondo, et al., Short-chain fatty acids, acetate and propionate, directly upregulate osteoblastic differentiation, Int J Food Sci Nutr (2022).

[7] Baratchart, et al., Integrated computational and in vivo models reveal key insights into macrophage behavior during bone healing, PLOS Comp Bio (2022).

[8] Creecy, et al., Control of bone matrix properties by osteocytes, Front Endocrinol (2021).