The purpose of this study is to investigate the behavioral change of ADSCs in various degrees of aging with respect to self-renewal and differentiation induced by chemical cues. Another focus of this study is to explore improvement of the declined functions in aged ADSCs by mechanical loading. That ADSCs responded to 10% strain was previously reported [23, 24], therefore, we used 10% strain in most experiments as the starting loading magnitude. We demonstrated that the result of mechanical loading in ADSCs depends on magnitude of the applied strain in which high strain (2-10%) significantly reduces adipogenesis and low strain (0.5%) has no effect (Figure 9). It should be noted that ADSCs in the mouse gonadal fat pad are not likely to experience 10% stretching in vivo, however, this magnitude of stretching is within the physiological range of loading for MSCs in muscular and peri-vascular tissues. We conducted our current study in ADSCs to take advantage of the ease of isolating large quantities of these cells from adult and aged animals. Further studies will be in need to verify that the observed results can be extrapolated to MSCs from other tissues as well as to the study of other forms of mechanical loading, using strain of various magnitudes, frequency, loading patterns, and durations.
Most current studies on MSCs have used established clones in which MSCs were selected through at least several rounds of passage and expansion. Aging (senescence) of MSCs in in vitro culture and loss of differentiation potential after the sixth passage has been demonstrated previously . In this study, we chose to study primary ADSCs culture in very early passages (P0-P3) in order to explore functional differences in ADSCs of various ages in the absence of artifacts likely contributed by clonal selection and expansion in long-term cultures. SVF from adipose tissue is known to contain primary cells with high degree of heterogeneity. We found that more than 50% of P0 cells after in vitro culture for one week exhibited the CD34-/CD45-/CD105+/CD73+/CD90+ surface markers, which are the most accepted surface markers for MSCs (data not shown). It should also be noted that > 95% of P6 young ADSCs were CD105+, CD73+, and CD90+ (data not shown). We failed to obtain ADSC colonies beyond P10 from 21-month-old mice (data not shown), indicating that passage-related senescence increases rapidly in old ADSC culture.
Short-term mechanical loading that simultaneously activates many mitotic signaling pathways is a strong inducer for cell proliferation . But, long-term mechanical loading (more than 24 h), which has not been as extensively studied as short-term loading, is less mitogenic [23, 27]. We purposely investigated the cellular responses to long-term stretching (48 h) to distinguish the responses as the true adaptation of cells to mechanical loading from transient reactions to sudden changes in mechanical environment. Immediately following mechanical stretching, instead of an increase, a slight decrease (less than 5%) in the number of ADSCs was found (data not shown). Thus, it is possible, also as a limitation for similar membrane stretching systems that mechanical stretching selects for cells that firmly adhere to the elastic membrane and allows the detached cells to die of anoikis. As a result, firmly adhered cells would be enriched following mechanical loading. Because the cell loss after strain was not significant when compared to the whole population, we believe that anoikis does not play an important role in our result. Yet, cell adhesion to substrate is an inherent cellular property that may be associated with self-renewal and differentiation potential. Indeed, BMMSCs of various sizes and different morphologies have been shown to exhibit different adhesion property to substrate . In a separate report, these heterogeneous cells also display different self-renewal abilities and express distinct surface markers . In addition, the role of adhesion to substrate in dictating MSC differentiation potential was recently demonstrated . Therefore, a change in ADSC adhesiveness to the substrate following mechanical loading should be investigated further. The finding that 48 h of pre-strain results in increased ADSC proliferation after replating (Figure 3) is intriguing to us. We did not find significant changes in the surface marker expression in ADSCs after applying strain (data not shown), therefore, the increase in proliferation following strain was not likely due to a selection for CD105+/CD73+/CD90+ cells during mechanical loading.
Because of a clear increase of calcium deposition in pre-strained ADSCs as well as in young cells, we originally hypothesized that aging would reduce, and mechanical stretching would enhance the osteogenesis capacity of ADSCs. An increase in calcium deposition can result from an increase in the number of total colonies without increasing the percentage of contained osteoblast colonies, or from an increase in both total colony number and the ratio of contained osteogenic colonies. While the former only results from an increase in proliferation, the latter represents a true increase in osteogenic capacity. By using adult ADSCs as a model, we demonstrated that in pre-strained cells, calcium deposition, when normalized to the DNA content, was not increased by strain as compared to non-strained cells. This result corroborates with the real time PCR data, which showed that bone marker gene expression did not increase as a result of pre-strain. The conclusion that the osteogenic potential of ADSC colonies was not increased by pre-strain is analogous to the result that no loss of osteogenic potential occurred in old ADSCs. Our finding in murine ADSCs is in agreement with previous reports that the osteogenic potential of human ADSCs from old donors was not reduced [13, 31]. In the literature, conflicting reports on the relative osteogenic potential of BMMSCs and ADSCs exist, and more findings indicate that ADSCs have an inferior osteogenic potential relative to BMMSCs [32–34]. Thus, it is also possible that we failed to detect changes in osteogenic potential due to aging or mechanical loading in ADSCs, while mechanical loading induces osteogenesis in BMMSCs in other reports, may result from the intrinsically lower osteogenic capacity of ADSCs.
The adipogenic program is regulated by multiple signaling pathways and involves the activation of numerous transcription factors. PPARγ serves as the pivotal transcription factor in adipogenesis. Several previous reports demonstrating that mechanical stretching inhibits PPARγ signaling in 3T3-L1 cells , BMMSCs  and skeletal myoblasts  may provide insight into the mechanism that underlies inhibition of adipogenesis by mechanical loading. We also found that a change in AP2 expression, previously considered a late marker of adipocytes, precedes the change in PPARγ expression induced in ADSCs by both aging and mechanical loading (10% strain). This result suggests that changes in PPARγ gene expression may not adequately reflect changes in PPARγ signaling, which can be better measured by the expression of its target genes. Indeed, AP2 is a PPARγ target gene [37, 38]. We are currently examining the molecular basis of the inhibition of PPARγ signaling and reduction in adipogenesis as a result of mechanical strain. Two mechanical loading methods, pre-strain (PS) and simultaneous adipogenic induction and strain (AS), were compared in this study. It is not surprising to observe a stronger inhibitory effect of AS on adipogenesis, considering that the mechanical loading activates known signaling pathways leading to adipogenesis inhibition. We have found activation of non-canonical Wnt signaling by mechanical strain, involving calcium/calmodulin-dependent kinase II and nemo-like kinase, in various precursor cell lines and primary cells . Therefore, inhibition of adipogenesis as a result of mechanical loading may be mediated by both canonical (β-catenin dependent)  and non-canonical Wnt pathways.