No abstract

Mesenchymal stem cells (MSC) rely on their ability to integrate physical and spatial signals at load bearing sites to replace and renew musculoskeletal tissues. Designed to mimic unloading experienced during spaceflight, preclinical unloading and simulated microgravity models show that alteration of gravitational loading limits proliferative activity of stem cells. Emerging evidence indicates that this loss of proliferation may be linked to loss of cellular cytoskeleton and contractility. Low intensity vibration (LIV) is an exercise mimetic that promotes proliferation and differentiation of MSCs by enhancing cell structure. Here, we asked whether application of LIV could restore the reduced proliferative capacity seen in MSCs that are subjected to simulated microgravity. We found that simulated microgravity (sMG) decreased cell proliferation and simultaneously compromised cell structure. These changes included increased nuclear height, disorganized apical F-actin structure, reduced expression, and protein levels of nuclear lamina elements LaminA/C LaminB1 as well as linker of nucleoskeleton and cytoskeleton (LINC) complex elements Sun-2 and Nesprin-2. Application of LIV restored cell proliferation and nuclear proteins LaminA/C and Sun-2. An intact LINC function was required for LIV effect; disabling LINC functionality via co-depletion of Sun-1, and Sun-2 prevented rescue of cell proliferation by LIV. Our findings show that sMG alters nuclear structure and leads to decreased cell proliferation, but does not diminish LINC complex mediated mechanosensitivity, suggesting LIV as a potential candidate to combat sMG-induced proliferation loss.

The bone deterioration that astronauts experience in microgravity environments is known to occur in response to the lack of gravity-based tissue stress. Mechanical forces are crucial to maintain healthy bone mass by regulating the function of bone-making osteoblasts as well as the proliferation and differentiation of their progenitors, mesenchymal stem cells (MSC) which replenish osteoblastic cells. Regulation of proliferative function of MSCs in response to mechanical force is in part controlled by the "mechanotransducer" protein YAP (Yes-associated protein) which is shuttled into the nucleus in response to mechanical challenge to induce gene expression necessary for cell proliferation. Our group had recently reported that altered gravity conditions under simulated microgravity (SMG) decreases proliferation of MSCs and that application of daily low intensity vibrations (LIV) during SMG reverses this effect on proliferation. While these findings suggest that LIV may be a promising countermeasure for altered loading, the specific SMG and LIV effects on YAP mechanosignaling are unknown. Therefore, here we tested the effects of SMG and daily LIV treatment on basal nuclear YAP levels as well as on the acute YAP nuclear entry in response to both mechanical and soluble factors in MSCs. MSCs subjected to 72h of SMG, despite decreased nuclear YAP levels across all groups, responded to both LIV and Lysophosphohaditic acid (LPA) treatments by increasing nuclear YAP levels within 6hrs by 49.52% and 87.34%, respectively. Additionally, daily LIV restored the basal decrease seen in SMG as well as nuclear YAP levels as well as restored in part the YAP nuclear entry response to subsequently applied acute LIV and LPA treatments. These results show that rescue of basal YAP levels by LIV may explain previously found proliferative effects of MSCs under SMG and demonstrates that daily LIV is capable of alleviating the inhibition caused by SMG of YAP nuclear shuttling in response to subsequent mechanical and soluble challenge.

27Reducing the bone deterioration that astronauts experience in microgravity requires 28 countermeasures that can improve the effectiveness of rigorous and time-expensive exercise 29 regimens under microgravity. The ability of low intensity vibrations (LIV) to activate force-30 responsive signaling pathways in cells suggests LIV as a potential countermeasure to improve 31 cell responsiveness to subsequent mechanical challenge. Mechanoresponse of mesenchymal 32 stem cells (MSC) that maintains bone-making osteoblasts is in part controlled by the 33 "mechanotransducer" protein YAP (Yes-associated protein) which is shuttled into the nucleus in 34 response cyto-mechanical forces. Here, using YAP nuclear shuttling as a measure of MSC 35 mechanoresponse, we tested the effect of 72 hours of simulated microgravity (SMG) and daily 36 LIV application (LIV DT ) on the YAP nuclear entry driven by either acute LIV (LIV AT ) or 37Lysophosphohaditic acid (LPA), applied at the end of the 72h period. We hypothesized that 38 SMG-induced impairment of acute YAP nuclear entry will be alleviated by daily application of 39 LIV DT . Results showed that while both acute LIV AT and LPA treatments increased nuclear YAP 40 entry by 50% and 87% over the basal levels in SMG-treated MSCs, nuclear YAP levels of all 41 SMG groups were significantly lower than non-SMG controls. Daily dosing of LIV DT , applied in 42 parallel to SMG, restored the SMG-driven decrease in basal nuclear YAP to control levels as 43 well as increasing the LPA-induced but not LIV AT -induced YAP nuclear entry over the non-LIV DT 44 treated, SMG only, counterparts. These cell level observations suggest that utilizing daily LIV DT 45 treatments is a feasible countermeasure for increasing the YAP-mediated anabolic 46 responsiveness of MSCs to subsequent mechanical challenge under SMG. 47 48

Deep learning architectures have demonstrated state-of-the-art performance for object classification and have become ubiquitous in commercial products. These methods are often applied without understanding (a) the difficulty of a classification task given the input data, and (b) how a specific deep learning architecture transforms that data. To answer (a) and (b), we illustrate the utility of a multivariate nonparametric estimator of class separation, the Henze-Penrose (HP) statistic, in the original as well as layer-induced representations. Given an N -class problem, our contribution defines the C(N, 2) combinations of HP statistics as a sample from a distribution of class-pair separations. This allows us to characterize the distributional change to class separation induced at each layer of the model. Fisher permutation tests are used to detect statistically significant changes within a model. By comparing the HP statistic distributions between layers, one can statistically characterize: layer adaptation during training, the contribution of each layer to the classification task, and the presence or absence of consistency between training and validation data. This is demonstrated for a simple deep neural network using CIFAR10 with random-labels, CIFAR10, and MNIST datasets.