In this study, we have systematically investigated the effects of SMG on mES cells and we found that mES cells were sensitive to SMG and the major alterations in cellular events were population growth inhibition, decreased adhesion rate, increased apoptosis and delayed DNA repair progression, which are unique to the responses of other types of cells to SMG.
We first detected the effect of SMG on the cell number expansion of mES cells and observed that the cell number was significantly reduced in the SMG group than that in the 1G group (). It seems that SMG impaired the expansion of mES cells throughout the whole process of the 5 days' treatment. However, after we transformed the data to the doubling generation curve and calculated the doubling time, we found that the doubling time of the SMG group was longer than that of the 1G group only from day 0 to day 2 (26.1 hr versus 11.1 hr), and from day 3 to day 4, the doubling time of the two groups were almost identical (13.4 hr versus 12.6 hr), indicating an adaptation of cells to SMG. From day 4 to day 5, the slope of the doubling generation curve of the 1G group was smaller than that of the SMG group because of the reaching of the plateau phase (data not shown here). These data suggest that major effects of SMG on cell expansion were acutely sensed in the first two days and counterbalanced afterwards by mES cells. There were also reports about the effects of SMG on cell expansion in other cell types. For example, in normal vascular smooth muscle cells, human breast cancer cells 
, malignant glioma cells 
and human mesenchymal stem cells 
cell expansion was affected, while the effect on cell expansion was not observed in normal human osteoblasts 
. Thus the effects of SMG on cell expansion vary among different cell types. However, they did not characterize the cell doubling time as we did in this study, so detailed population kinetic information such as proliferation adjustment to SMG in these studies was unknown.
The impaired cell number expansion might be caused by impaired cell proliferation, decreased attachment ability of the cells to the flask wall and/or increased apoptosis. We observed that no specific cell cycle phase was significantly altered under SMG conditions, suggesting that there was no significant difference in cell proliferation between the 1G and SMG groups. This was further verified by the expression of PCNA, a broadly used marker of cell proliferation 
. Interestingly a similar result was reported by Takeda et al 
. They found that cell expansion in malignant glioma cells was inhibited under SMG without evident change in any phase of the cell cycle. However, mitochondrial activity did decrease under SMG. They argued that, under SMG conditions, cells enter a hibernation-like state, without any cell cycle arrest but a slowdown of the whole cell cycle. Another way to check whether there is a cell hibernation-like state was to analyze the BrdU uptake. In principle, slowed cell cycle speed coupled with slowed BrdU incorporation during the limited time of pulse-labeling, resulting in less BrdU uptake. In our study, however, the BrdU uptake was not less in the SMG group than in the 1G group, suggesting the cell cycle progress of mES cells was not slowed down, thus no cell hibernation occurred. However, we observed increased apoptosis and decreased adhesion rate of mouse ES cells under SMG. Taken together, our results indicates that SMG did not affect the proliferation of mES cells and the SMG-induced apoptosis and cell detachment from the flask wall were important factors affecting cell expansion.
Interestingly, we found that mES cells started to detach from the flask between 6 hr and 8 hr of rotation in SMG treatment, and the adhering ability bounced back to normal level quickly at 10 hr of SMG treatment, indicating that cells made quick adjustments to enhance their adhering abilities. As discussed in the above paragraph, cell expansion rate also demonstrates adaptation behavior to SMG. It is likely that many more cell activities adjust to MG. Sarkar et al 
cultured the osteoblast-like ROS 17/2.8 cells under SMG condition, and observed that there were significantly more detached cells in the SMG group than that in the control groups at 24 hr and the cells became confluent in both the SMG group and the control groups at 72 hr. Consistent with our findings, this phenomenon might be due to adaptation of osteoblasts to SMG. However, it seems that the adaptation time needed for mES cells was much shorter than osteoblasts. As to what underlying molecular adjustments have been made need further investigation.
Gelatin-coated or feeder cell-coated plates/flasks are commonly used for culturing mES cells. In this study, we cultured mES cells on feeder cell-free and gelatin-coated flasks. On gelatin-coated flasks, mES cells demonstrated a brief period of detachment from the flask wall and apoptosis. We wondered if these effects were depended on the coating material. The information on other coating conditions than gelatin in the culture of mES cells is rare. Hayashi et al 
compared the effects of different coating conditions on mES cells in a defined serum-free culture medium. They tested seven extracellular matrix (ECM) components for their effects on mES cell differentiation, and they found that poly-D-lysine (PDL) could also maintain the undifferentiated state of mES cells efficiently as gelatin among the ECM components they tested. Thus we chose PDL as another coating condition and evaluated the cell attachment and apoptosis under PDL coating. Compared to the 1G group, the adhesion rate of the SMG group cells was lowest at 6 hr and 8 hr. After that, the adhesion rate bounced back, and this trend of change is similar to what we observed in the cells cultured in gelatin-coated flasks although the overall adhesion rates of the cells cultured in PDL-coated flasks in the SMG group were lower (Figure S1
). However, we did not observed a higher apoptosis level in cells in the SMG group than that in the 1G group in PDL-coated flasks, in contrast to the phenomenon in gelatin-coated flasks (Figure S2
). Therefore, the apoptosis increase is a coating-material dependent phenomenon. Even without the apoptosis increase, mES cells also demonstrated a cell number expansion reduction in the SMG group in PDL-coated flasks similar to that in gelatin-coated flasks (Figure S3
). It is likely that the overall lower attachment rate and a short period of detachment play a major role in cell number expansion reduction of mES in PDL-coated flasks under SMG condition.
The effects of SMG on genomic stability have been investigated by several groups and SMG was found to induce genomic DNA damage in lymphocytes with extended period of exposure (7 days) to SMG 
, but not in human lymphocytes and lymphoblastoid cells exposed to SMG for 8 hr or 24 hr 
. In this study, we did not observe increased DNA lesions in mES cells after two days' exposure to SMG, and our result is in consistent with the findings reported by Degan et al 
, indicating that SMG alone could not induce DNA damage in mES cells after short time of exposure.
Other than SMG, radiation is another important environmental factor during spaceflight, which could cause severe DNA lesions. In the present study, we also investigated the repair of radiation-induced DNA lesions in mES cells under SMG. We found that SMG impaired the repair capacity of mES cells at 1 hr post-irradiation while at 2 hr post-irradiation, there was no significant difference between the 1G and SMG group. Therefore, we speculate that SMG impairs the repair of radiation-induced DNA lesions only at the early stage of the lesion repair. Mognato et al also reported that in human peripheral blood lymphocytes (PBLs), the mutant frequency induced by IR was increased by incubation in SMG 
and the SMG incubation during DNA repair delayed the rate of radiation-induced DSB rejoining 
. However, in their reports the effects could be observed 24 hr after IR and SMG treatment and they did not show the data of more than 24 hr, obviously different from the short effects of SMG on the DNA repair capacity of mES cells observed in this study.
We observed that the attachment of mES decreased and then bounced back after exposure to SMG. Hammond et al 
compared the expression of 10000 genes of primary cultures of human renal cortical cells grown under the culture condition including MG, rotating wall vessel, 3-g centrifuge and static culture. They found that the genes whose expression changed the most include adhesion molecules, cytoskeletal protein genes and apoptosis genes. Furthermore, cytoskeleton abnormalities have been observed in the studies conducted on different types of cells such as osteoblastic ROS 17/2.8 cells 
, glial cells 
, human MCF-7 cells 
. Under SMG, the cytoskeleton of these cells was disorganized at the beginning and after a certain time was reestablished. Interestingly, this accords with our findings of the adaptation of mES cells to SMG (proliferation rate and detachment). Interestingly, cytoskeleton is known to be crucial for numerous cellular processes, including cell adhesion. All the above studies used differentiated cells. Here we showed that stem cells had the bounce-back phenomenon in adhesion to the flask, thus the mechanisms behind the adhesion bounce-back may involve the induction of adhesion and cytoskeleton genes and warrant further investigation.
In this study, we investigated the unique responses of mES cells to SMG. We found that mES cells were sensitive to SMG at the early stage of SMG treatment and the cells could make quick adjustments to adapt to the SMG condition (proliferation rate and detachment). Our results provide new information on the effects of SMG on mES cells. Such information is novel in ES cell biology and may be helpful in stem cell based regenerative medicine.