UC Davis

Over 9 million fractures occur each year in the US and an estimated 5-10% will develop a non-union defect requiring additional surgical care amounting to an average of $11,000 per patient. While autologous tissue graft remediation for non-unions is an effective strategy, this approach presents significant detrimental side effects including harvest site morbidity and limited tissue supply. Tissue engineering strategies harnessing the beneficial trophic factor secretion and differentiation capacity of mesenchymal stromal cells (MSC) have been a popular approach to overcome current clinical treatments. Furthermore, age is a comorbidity associated with increased non-union incidence and decreased healing effectiveness. Senescence within the MSC pool from the elderly is an important factor in age associated deficits in bone regeneration and is a limiting factor in the effectiveness of autologous MSCs to be used in tissue engineered approaches. We propose that the selective clearance of these senescent MSCs would improve the osteogenic potential of the remaining cell population. The senescent burden from biological donors can be variable even within age matched samples. To study the effects of senescence on MSC function and to reliably target senescent cells, we generated a consistent and robust method for inducing senescence within human MSCs. Furthermore, identification of the senescent phenotype is under much debate as there is not an agreed upon battery of tests to identify senescence. In the process of developing a consistent senescent population, we validated an array of tests to confirm the senescent phenotype in these MSCs. We scoured the literature and arrived upon three of the most common methods to induce a senescent phenotype in fibroblast culture: hydrogen peroxide exposure, treatment with the chemotherapeutic etoposide, and irradiation. Comparison of these different methods to replicative induced senescence, the gold standard, indicated that both etoposide and 10 Gy of irradiation generated a similar senescent phenotype. Robust senescent markers associated with MSCs were also identified including the secretion of pro-inflammatory cytokines, DNA damage as indicated by phosphorylated γH2AX, cell morphology, and increases in CDNK1A expression. Stress induced senescent MSCs treated with Fisetin, a flavonoid utilized for its ability to induce apoptosis in senescent cells, did not exhibit a reduction of senescent markers. However, we did observe improvements in a heterogeneous population of bone marrow derived cells from an aged biological donor when encapsulated in a hydrogel. The application of pharmacologic compounds that mitigate the senescent phenotype, termed senotherapeutics, has gained significant study for their potential to remedy age related detriments. Previous studies establish that the application of senotherapeutic compounds reduce the senescent burden in vitro. Given these findings, we hypothesized that exposure of senescent MSCs to common senotherapeutics would reduce the senescent burden of the population and improve the viability of the remaining MSCs. We demonstrated that application of two common senotherapeutic compounds does not reduce senescent markers in our senescent model of MSCs when altering dosing or temporal application. Furthermore, we found that these treatments have detrimental effects on healthy MSCs, reducing their viability and metabolic activity. Next, we wanted to investigate the ability of substrate surfaces to modulate the senescent phenotype in our stress induced senescent model. MSC sensitivity to substrate stiffness is well documented, with stiffer substrates inducing an osteogenic phenotype and softer substrates supporting adipogenic differentiation. However, the response of senescent MSCs to these substrate cues is not well understood. Bone tissue from aged individuals is softer and has higher proportions of adipose tissue. This observation led us to hypothesize that stiffer substrate could reduce the burden of senescent MSCs. Contrary to our hypothesis, we observed that senescent MSCs cultured on softer substrates had a dramatic reduction in the secretion of pro-inflammatory factors such as IL-6, IL-8, MCP1, and MIP-1β compared to those cultured on tissue culture plastic. Interestingly, we observed an increase in expression of CDNK1A, CDNK2A, and EGFR genes with softer substrates within the senescent population. When using a heterotypic population of healthy MSCs (75%) and senescent MSCs (25%), there were no changes in gene expression of inflammatory protein secretion. These data indicate that while substrates alter the characteristics of a homotypic senescent population, physiologically relevant numbers of senescent MSCs are buffered by their healthy counterparts. Taken together, these studies demonstrate our ability to leverage a model senescent MSC population to better understand the role of senescence within bone regeneration strategies for the elderly. This model can be utilized to investigate how senescent MSCs respond to chemical and material cues. This work demonstrates that senescent MSCs are differentially responsive to substrate stiffness, a foundational mechanism in tissue engineered approaches. Ultimately, the senescent model and subsequent studies offer insight into how to develop tissue engineered strategies that cater to the unique needs of the elderly.