3D nanomechanical mapping of subcellular and sub-nuclear structures of living cells by multi-harmonic AFM with long-tip microcantilevers

Abstract

Recent developments such as multi-harmonic Atomic Force Microscopy (AFM) techniques have enabled fast, quantitative mapping of nanomechanical properties of living cells. Due to their high spatiotemporal resolution, these methods provide new insights into changes of mechanical properties of subcellular structures due to disease or drug response. Here, we propose three new improvements to significantly improve the resolution, identification, and mechanical property quantification of sub-cellular and sub-nuclear structures using multi-harmonic AFM on living cells. First, microcantilever tips are streamlined using long-carbon tips to minimize long-range hydrodynamic interactions with the cell surface, to enhance the spatial resolution of nanomechanical maps and minimize hydrodynamic artifacts. Second, simultaneous Spinning Disk Confocal Microscopy (SDC) with live-cell fluorescent markers enables the unambiguous correlation between observed heterogeneities in nanomechanical maps with subcellular structures. Third, computational approaches are then used to estimate the mechanical properties of sub-nuclear structures. Results are demonstrated on living NIH 3T3 fibroblasts and breast cancer MDA-MB-231 cells, where properties of nucleoli, a deep intracellular structure, were assessed. The integrated approach opens the door to study the mechanobiology of sub-cellular structures during disease or drug response.

Figure 1

Cantilevers with long carbon tips. (A) SEM image (left inset—a fit with parabola to estimate the tip radius; right inset—a regular tip of the same cantilever type). (B) Force vs. Indentation curve obtained on the hydrogel fit well with the Hertz’s contact model (paraboloid). (C) Normalized amplitude vs. distance from the surface (glass) in the liquid (PBS). (D) Phase vs. distance from the surface (glass) in liquid (PBS).

Figure 2

Cantilevers with long tip reduce topography convolution observed in phase images. (A). Phase image of a tall (~ 5 µm) MDA-MB-231 cell. Phase increase beyond the cell is seen in the top part of the image caused by the hydrodynamic interaction between the cell surface and the cantilever beam behind the tip. (B) This effect is not presented when using the long tips; the contrast is higher, cytoskeleton fibers and nucleoli could be distinguished. Insets: SEM image of the tip. (C) Cross-section of the phase signal along the black line indicated in (A) showing the effect of the phase signal affected by the topography for the standard tip. The phase value is lower over the highest part of the cell (lower hydrodynamic drag). (D) Cross-section of the phase signal over the marked line in (B), long tip. The phase does not noticeably correlate with the topography.

Figure 3

The NIH 3T3 fibroblast, AFM observables and nanomechanical maps. (A) and SDCM color-coded height images (B,C). Both perinuclear actin cap fibers and nucleoli could be detected in the amplitude and phase images, and in the E_storage map. Only the actin fibers, but not nucleoli, are presented in topography and deflection images. (B,C) Corresponding images of the F-actin cytoskeleton (SiR-actin staining, B) and nucleoli (SYTO85 staining, C), color-coded Z-stacks. (D) Vertical cross-section along the marked line shows the size and depth of the nucleoli (red) and cell shape (F-actin, green). Scale bars are 10 μm in the horizontal direction and 2 μm in the vertical direction.

Figure 4

MDA-MB-231 breast cancer cell, AFM observables and nanomechanical maps (A) and SDCM color-coded height images (B,C). Some peripheral actin stress fibers and nucleoli could be detected in the amplitude and phase images, and in the E_storage map. Only the actin fibers, but not nucleoli, are presented in topography and deflection images. (B) Corresponding images of the actin cytoskeleton (SiR-actin staining) and nucleoli (SYTO85 staining), color-coded Z-stacks. (D) Vertical cross-section along the marked line shows the size and depth of the nucleoli (red) and cell shape (F-actin, green). Scale bars are 10 μm in the horizontal direction and 2 μm in the vertical direction.

Figure 5

NIH 3T3 fibroblast, nanomechanical maps obtained over the nucleus region. Cross-section of the E_storage moduli over the two nucleoli (line on the map). E_storage above the nucleoli is about twice higher than above the surrounding area. Scale bars are 10 µm for the SDCM image, 5 µm for the nanomechanical maps.

Figure 6

FEM analysis for estimating the mechanical properties of nucleoli. (A) Scheme of the FE model showing the dimensions extracted from the SDC images (the cantilever tip is not in scale). (B) Distribution of effective von Mises stress during the indentation over the nucleolus. (C) Simulated force-indentation curves, the legend shows E_nuc/E_m ratio. The curves went steeper as the ratio increased. (D) The effective local stiffness k_eff (solid line) measured at the experimental value of force as a function of the E_nuc/E_m ratio. It approaches the experimental k_eff value (dashed line, 0.016 ± 0.001 nN/nm, shaded area) at a high E_nuc/E_m ratio (> 100). The 10% error (shaded area) is based on variation in k_eff value over nucleoli and surrounding nucleoplasm.