Mesenchymal Stem Cells Sense the Toughness of Nanomaterials and Interfaces

Abstract

Stem cells are known to sense and respond to the mechanical properties of biomaterials. In turn, cells exert forces on their environment that can lead to striking changes in shape, size and contraction of associated tissues, and may result in mechanical disruption and functional failure. However, no study has so far correlated stem cell phenotype and biomaterials toughness. Indeed, disentangling toughness-mediated cell response from other mechanosensing processes has remained elusive as it is particularly challenging to uncouple Youngs' or shear moduli from toughness, within a range relevant to cell-generated forces. In this report, it is shown how the design of the macromolecular architecture of polymer nanosheets regulates interfacial toughness, independently of interfacial shear storage modulus, and how this controls the expansion of mesenchymal stem cells at liquid interfaces. The viscoelasticity and toughness of poly(l-lysine) nanosheets assembled at liquid-liquid interfaces is characterised via interfacial shear rheology. The local (microscale) mechanics of nanosheets are characterised via magnetic tweezer-assisted interfacial microrheology and the thickness of these assemblies is determined from in situ ellipsometry. Finally, the response of mesenchymal stem cells to adhesion and culture at corresponding interfaces is investigated via immunostaining and confocal microscopy.

Keywords: 2D nanomaterials; liquid-liquid interface; protein nanosheet; self-assembly; stem cells; toughness.

Figure 1

Impact of molecular weight on PLL nanosheet assembly at liquid–liquid interfaces. A) Molecular structure of PLL nanosheets and proposed resulting architecture. B) Evolution of the interfacial shear storage modulus of PLL nanosheets forming at Novec 7500‐water interfaces (Novec 7500 containing 10 µg mL⁻¹ PFBC; aqueous solution is PBS with pH adjusted to 10.5; strain of 10⁻³ rad and 0.1 Hz). PLL with different M_w (3, 10, 22.5, 50, 110, 225, and >300 kDa) was introduced (after 900 s of equilibration) to make a final solution with a concentration of 100 µg mL⁻¹. C) Corresponding ln(Γ(t)/Γ₀) plots at two different time intervals following protein injection. D) Adsorption rate constants extracted from corresponding linear fits. E) Interfacial storage moduli as a function of M _w of PLL, measured from frequency sweeps at a strain of 10⁻³ rad and 0.1 Hz. Error bars are s.e.m.; n = 3. F) XPS spectra (F 1s) obtained for nanosheets generated with PLL with different M _w. G) Functionalization levels quantified from corresponding XPS data (error bars are s.e.m.; n = 3). One‐way ANOVA; n.s., non‐significant; *p < 0.05.

Figure 2

Interfacial viscoelasticity is controlled by the molecular weight of PLL. A) Representative stress relaxation profiles of nanosheets assembled from PLL with different molecular weights (strain: 10⁻³ rad). B) Corresponding stress retentions σ _R extracted from the corresponding fits. Error bars are s.e.m.; n ≥ 3. C) Epifluorescence microscopy images of PLL nanosheets assembled with PLL with a range of M _w (all tagged with Alexa Fluor 488). Detail of interfaces: Novec 7500 containing 10 µg mL⁻¹ PFBC; aqueous solution is PBS with pH adjusted to 10.5; PLL with different M _w (3, 10, 22.5, 50, 110, 225, and >300 kDa) at a final concentration of 100 µg mL⁻¹. One‐way ANOVA; n.s., non‐significant; **p < 0.01.

Figure 3

The nanoscale architecture of PLL nanosheets controls interfacial toughness. A) Representative shear stress‐strain curves extracted from amplitude sweep experiments (frequency of 0.1 Hz). The grey area shaded correspond to the range of interfacial stresses expected to be exerted by mature focal adhesions. B) Summary of interfacial toughness calculated from the corresponding shear stress–strain profiles. (error bars are s.e.m.; n ≥ 3). C) Damping functions calculated from strain sweeps. The trend lines correspond to fits with the Soskey–Winter model. D,E) Proposed model of nanosheet fracture, depending on the molecular weight of PLL chains (D, side view; E, top view; only some chains localized at the fracture line are represented for improved visualization). Ellipsometric thickness of selected nanosheets determined F) dry, G) in deionized (DI) water, and H) PBS. Nanosheets were transferred to silicon substrates using a Langmuir–Blodgett liquid–liquid trough, prior to characterization. Error bars are s.e.m.; n = 3. I) Summary of magnetic‐tweezer assisted interfacial microrheology data (shear moduli G ₀, G ₁, and G ₂ extracted using the six elements Burger's model). Detail of interfaces: Novec 7500 containing 10 µg mL⁻¹ PFBC; aqueous solution is PBS with pH adjusted to 10.5; PLL with different M _w (3, 10, 22.5, 50, 110, 225, and >300 kDa) at a final concentration of 100 µg mL⁻¹. Error bars are s.e.m.; n ≥ 7. One‐way ANOVA; n.s., non‐significant; *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 4

Stem cell expansion at liquid interfaces correlates with interfacial toughness. A,B) Impact of PLL molecular weight on cell spreading at Novec 7500 interfaces stabilized by corresponding nanosheets. C) Confocal microscopy images of MSCs spreading (after 24 h) on PLL/FN functionalized Novec 7500 interfaces. Zoom‐in correspond to the dotted boxes. D,E) MSC expansion at PLL‐stabilized Novec 7500 interfaces (D, representative nuclear stainings). F) Highly confluent MSCs remodel and fracture PLL/FN nanosheets assembled at the surface of Novec 7500. Epifluorescence microscopy images of PLL nanosheets 24 h after seeding MSCs at 200 000 cell/well (left). Red, PLL; blue, nuclei. Detail of interfaces: Novec 7500 containing 10 µg mL⁻¹ PFBC; aqueous solution is PBS with pH adjusted to 10.5; PLL with different M _w (3, 10, 22.5, 50, 110, 225, and >300 kDa) at a final concentration of 100 µg mL⁻¹. Error bars are s.e.m.; n ≥ 4. One‐way ANOVA; n.s., non‐significant; *p < 0.05.