Modulation of cellular mechanics during osteogenic differentiation of human mesenchymal stem cells

Abstract

Recognition of the growing role of human mesenchymal stem cells (hMSC) in tissue engineering and regenerative medicine requires a thorough understanding of intracellular biochemical and biophysical processes that may direct the cell's commitment to a particular lineage. In this study, we characterized the distinct biomechanical properties of hMSCs, including the average Young's modulus determined by atomic force microscopy (3.2 +/- 1.4 kPa for hMSC vs. 1.7 +/- 1.0 kPa for fully differentiated osteoblasts), and the average membrane tether length measured with laser optical tweezers (10.6 +/- 1.1 microm for stem cells, and 4.0 +/- 1.1 microm for osteoblasts). These differences in cell elasticity and membrane mechanics result primarily from differential actin cytoskeleton organization in these two cell types, whereas microtubules did not appear to affect the cellular mechanics. The membrane-cytoskeleton linker proteins may contribute to a stronger interaction of the plasma membrane with F-actins and shorter membrane tether length in osteoblasts than in stem cells. Actin depolymerization or ATP depletion caused a two- to threefold increase in the membrane tether length in osteoblasts, but had essentially no effect on the stem-cell membrane tethers. Actin remodeling in the course of a 10-day osteogenic differentiation of hMSC mediates the temporally correlated dynamical changes in cell elasticity and membrane mechanics. For example, after a 10-day culture in osteogenic medium, hMSC mechanical characteristics were comparable to those of mature bone cells. Based on quantitative characterization of the actin cytoskeleton remodeling during osteodifferentiation, we postulate that the actin cytoskeleton plays a pivotal role in determining the hMSC mechanical properties and modulation of cellular mechanics at the early stage of stem-cell osteodifferentiation.

FIGURE 1

Measurements of cell elasticity with AFM microindentation. Live cell imaging and mechanical testing were conducted in phosphate-buffered saline with normal or drug-treated cells. (A) Typical force-distance curves obtained for normal osteoblasts (B) and stem cells (C). Measurements were performed on the area between the nucleus and the margin of the cell (arrows). The scan size of both AFM images is 40 × 40 μm. Note the higher contrast in hMSCs (C) than in osteoblasts (B) due to stronger, thicker stress fibers in the hMSC cytoskeleton. (D) Effect of cytoskeleton-destabilizing drugs on the average elastic modulus of the cells; ∼600–800 force curves were acquired for each cell type and condition. There is no statistical significant difference at p = 0.05 between nocodazole and their corresponding controls.

FIGURE 2

Membrane tether extraction with laser optical tweezers. (A) Fluorescent polystyrene beads coated with anti-integrin antibodies attached to a normal stem cell surface and pulled away with LOT (arrowhead) to extract a long thin membrane tether (thin arrow). (B) Force applied by LOT to the bead to form a tether. Force is constant for most of tether growth due to membrane pulling from the membrane reservoir (32). Upon reservoir depletion the maximum tether length is reached, the force sharply increases, and the bead escapes from the optical trap. (C) Effect of the membrane-cytoskeleton interaction on average membrane tether length in two cell types. All treatments produced a statistically significant (p < 0.05) tether length increase in osteoblasts, but not hMSCs.

FIGURE 3

Cytoskeleton organization in undifferentiated hMSCs (A and B) and mature osteoblasts (C and D). In these confocal images, actins and tubulins were stained with rhodamine-phalloidin (red) and FITC-conjugated anti-α-tubulin antibody (green), respectively. Stem cells (A) have many thick actin bundles (stress fibers), unlike osteoblasts (C), which have a thinner actin filament meshwork. After treatment with cytochalasin D and nocodazole, the cytoskeleton is fully disintegrated (B and D), and patches of actins are left at the focal adhesion sites. Scale bar, 30 μm.

FIGURE 4

Alteration of cellular mechanical parameters during hMSC osteodifferentiation. Both the cell average Young's modulus (A) and average membrane tether length (B) decrease significantly (p < 0.05) with 7-day incubation in the osteogenetic medium (dashed lines). In control experiments, undifferentiated hMSCs were grown in regular culture medium for 10 days without significant mechanical changes (solid lines).

FIGURE 5

Actin cytoskeleton rearrangement during hMSC osteodifferentiation. Within 10 days of osteogenic differentiation induction, more and more thick stress fibers are replaced with a thinner actin filament meshwork typical for mature osteoblasts. Scale bar, 30 μm.

FIGURE 6

Model for the actin cytoskeleton in cellular mechanics of hMSCs and osteoblasts. Thin and dense actin filaments in osteoblasts are tightly bound to the plasma membrane through multiple linker proteins and focal complexes. The hMSC thick stress fibers are associated with the membrane mostly at focal contacts due to a smaller contact area with the membrane and thus lower availability of protein-linker binding sites. As a result, in hMSCs, the overall membrane-cytoskeleton adhesion is weaker and the extracted membrane tether is longer than in osteoblasts. However, the thick-bundled actin structure in hMSCs provides a higher elastic modulus, but a weaker membrane-cytoskeleton interaction, than those found for osteoblasts.