Thickness sensing of hMSCs on collagen gel directs stem cell fate.

Wen Shing Leong, Chor Yong Tay, Haiyang Yu, Ang Li, Shu Cheng Wu,Duong-Hong Duc, Chwee Teck Lim, Lay Poh Tan.


Introduction:
Recently, it was found that living cells could respond to substrate elasticity. Stem cells will differentiate into specific tissue only on the substrate with similar elasticity as the native tissues, which are relative soft. But if the substrate is too soft, the stem cells will become quiescent.

Cells could also sense the thickness of the substrate. Within the interaction depth, cells would sense higher effective substrate modulus than the intrinsic modulus because of the rigid support beneath the substrate.

Among all the biological polymers, collagen is the most abundant protein in vivo; therefore it could represent ECM the best. In this study, type I collagen is used as substrate; their thickness are varied to modulate hMSCs.

Materials and Methods:
Before the experiment, the bone marrow-derived hMSCs were expanded and maintained in MSCGM. Passages 4 to 8 hMSCs were used during the experiment. They were cultured in low glucose DMEM supplemented with 10% Gold FBS, 1% L-glutamine and 1% penicillin-streptomycin. hMSCs were seeded onto thin (130μm) collagen gel, thick (1440μm) collagen gel, and collagen molecules coated coverglass in 24-wellplate for 14 days, all with a density of 1500cells/cm2. For multipotency study, adipogenic and osteogenic differentiation were induced in adipogenic and osteogenic induction medium for 14 and 21 days respectively.

To get gene expression profile, After 2 and 14 days of culture, total RNA of hMSCs was isolated for each different groups of substrates. Quantitative real time RT-PCR was performed. For immunocytochemistry, after 7 and 14 days of culture, intracellular actin filaments of the cell were stained with rhodamine conjugated phalloidin; collagen molecules were stained with anti-type I collagen, and TRITC anti-mouse IgG. For BrdU incorporation assay, after 48 hours and 14 days, cells were incubated in medium containing BrdU overnight, then stained with BrdU staining kit and cell nucleus was stained with hematocylin.

Results:
After two days of culture, the spreading area was measure for each group. It was found that projected cell area was increased with decreasing gel thickness. In addition, the hMSCs on thin collagen gel showed neurite-like extension. (Figure 1)

Fig. 1. Morphological responses of hMSCs on collagen gel of different thickness and control group at D2. (A) Spreading area of hMSCs decreased with increasing collagen gel’s thickness. (B) Phase images of hMSCs cultured on collagen gel of different thickness and control. Scale bar = 50 μm.

After 7 and 14 days, the immune-labeled cellular actin filament was visualized. It was showed that on the control (collagen molecule coated coverglass) group, the microfilaments were well defined. Actin fiber bundles were thick and aligned in parallel with the cell major axis. On thin collagen gel, actin bundles were thinner. It had similar microfilaments but the integrity and connectivity were decreased. On thick collagen gel, the actin bundle size was the smallest and most disorganized. Thick collagen gel did support the formation of long and integrated actin bundles. (Figure 2)

Fig. 2. Cytoskeleton structure of hMSCs at 7th and 14th day of culturing on thick A, D (1440μm), thin B, E (130μm) collagen gel and C, F control group. Small white dotted square at lower left corner shows enlargement of the cytoskeleton structure at highlighted area. Scale bar = 25μm.

The BrdU incorporation assay showed at the timepoints studied, none of the cells on thick gel had S-phase. Cells on thin gel demonstrated higher proliferation rate than thick gel, and the control group appeared to have to highest proliferation rate. In addition to the absence of proliferation activity, using qPCR it was showed that on thick gel the cells had downregulation in mRNA expression level of all the gene studied. (Figure 3). Non proliferation rate and downregulation of mRNA suggest that the stem cells on thick collagen gel was quiescent.

Fig. 3. Normalized mRNA expression of (A) osteogenic and (B) neurogenic specific markers of hMSCs upon seeded onto collagen gel of different thickness for 2 and 14 days. Specific upregulation of neuronal markers was observed in thin collagen gel at D14, but none of the analyzed genes was expressed in thick collagen gel over the 14 days of culturing period.

Discussion and Conclusions:
Besides the intrinsic modulus, the substrate thickness also plays an important role. Due to lack of reactive force on soft substrate, cell would have smaller spreading area and less-developed actin fiber on the softer substrate. In this study, it was showed cell spreading area and cytoskeleton development decreased with increasing collagen gel thickness. Along with non proliferative activity, it was suggest thick gel suppress gene expression and therefore the stem cell would remain quiescent state on thick gel.

Commentary:
This study studied the mechanical properties on living cells by changing the thickness of the substrate instead of changing the elastic modulus of the substrates. This is better approach because it’s easier. If the collagen concentration was changed to alter the elastic modulus, then there would be two variables in the experiment: thickness and elasticity. Since there would be variation in measure the thickness for each well, so there is no way to get exactly the same value in thickness. Therefore there is no way to determine if the change was caused by thickness effect or elasticity effect. However, in this study the authors didn’t test what is the range of interaction depth of the hMSCs. If I will perform the same experiment, I will compare more values in thickness, and a wider range of the thickness. Therefore it could be determined beyond which value the cell will not fell the difference in thickness.