A Micropatterning Strategy to Study Nuclear Mechanotransduction in Cells

Micropatterning techniques have been widely used in biology, particularly in studies involving cell adhesion and proliferation on different substrates. Cell micropatterning approaches are also increasingly employed as in vitro tools to investigate intracellular mechanotransduction processes. In this report, we examined how modulating cellular shapes on two-dimensional rectangular fibronectin micropatterns of different widths influences nuclear mechanotransduction mediated by emerin, a nuclear envelope protein implicated in Emery–Dreifuss muscular dystrophy (EDMD). Fibronectin microcontact printing was tested onto glass coverslips functionalized with three different silane reagents (hexamethyldisilazane (HMDS), (3-Aminopropyl)triethoxysilane (APTES) and (3-Glycidyloxypropyl)trimethoxysilane (GPTMS)) using a vapor-phase deposition method. We observed that HMDS provides the most reliable printing surface for cell micropatterning, notably because it forms a hydrophobic organosilane monolayer that favors the retainment of surface antifouling agents on the coverslips. We showed that, under specific mechanical cues, emerin-null human skin fibroblasts display a significantly more deformed nucleus than skin fibroblasts expressing wild type emerin, indicating that emerin plays a crucial role in nuclear adaptability to mechanical stresses. We further showed that proper nuclear responses to forces involve a significant relocation of emerin from the inner nuclear envelope towards the outer nuclear envelope and the endoplasmic reticulum membrane network. Cell micropatterning by fibronectin microcontact printing directly on HMDS-treated glass represents a simple approach to apply steady-state biophysical cues to cells and study their specific mechanobiology responses in vitro.


Calculation of Theoretical Contact Angle for a Monolayer Formation with HMDS
The Cassie equation [1] describes the expected contact angle of a solute on a bifunctional surface (in our case the piranha-treated coverslip glass surface and its HMDS layer) as follows: where Ф is the equilibrium contact angle of water on the bifunctional surface (glass coverslip + HMDS), f1 is the molecular fraction coverage of HMDS on the bifunctional surface, θ1 is the expected contact angle of water on a pure HMDS surface (without coverslip support), f2 is the molecular fraction of the bifunctional surface not covered by HMDS, and θ2 is the expected contact angle of water on the pure glass coverslip surface (without HMDS).
f1 and f2 depend: (i) On the circular cross-sectional area taken up by the trimethylsilyl group (Si-(CH3)3) of HMDS, which has been estimated by Herzberg et al. [2] at 27.7 Å 2 and (ii) On the number per surface area of hydroxyl silanol groups (Si-OH) that are available to react with HMDS on the coverslips. For a fully hydroxylated silica surface, this number has been estimated at 5 per 100 Å 2 by different groups [3][4][5], and it is assumed to be unchanged for glass.
Thus, for a perfectly modified glass coverslip substrate where coverage and reaction of HMDS with silanol groups is optimal (monolayer), the surface molecular fraction coverage of trimethylsilyl should be f1 = (100/27.7)/5 = 0.722. Consequently, we have a glass surface molecular fraction noncovered by HMDS of f2 = 1 − 0.722 = 0.278.
The contact angle value of water on the piranha-treated glass coverslip surface (without HMDS) was measured and reported in Fernandez et al. [6] as θ2 = 3.5°. There is, however, no known θ1 value for a pure HMDS surface (without substrate support) because HMDS cannot polymerize on its own into a solid surface. However, the contact angle value of water on PDMS can be used as an appropriate estimate of θ1 for HMDS because this polymer is extremely rich in methyl groups similar to those found in HMDS. Here, we used a contact angle of water on PDMS of 108° as determined by molecular simulation [7]. This value is in good agreement with reported experimental estimates between 98° and 112° (see references in Ismail et al. [7]). Thus, with values: f1 = 0.72, θ1 = 108°, f2 = 0.28 and θ2 = 3.5° we can evaluate the expected equilibrium contact angle of water (Ф) for a perfect and monolayer coverage of HMDS on a glass coverslip, from Cassie's equation. We obtain cos Ф = 0.054 and Ф = 87°. This value is in excellent agreement with our measured value of 87 ± 1° after 90 min of reaction, thus indicating that vapor coating provide a homogenous monolayer deposition of HMDS on glass coverslips.
We note that multilayer polymerization of HMDS at the coverslip surface (although it is improbable considering the nature of the chemical reaction) would result in water contact angles that would increase toward 108° overtime. Similarly, sub-monolayer coverage would result in water contact angles that would decrease from the theoretical 87° value towards the 3.5° angle value measured for piranha-treated glass.

Measurements of Contrast Values to Compare the Quality of Fibronectin Microcontact Printing across Coverslips Coated with Different Organosilanes
To quantify the quality of the fibronectin micropatterns, fluorescence intensities of Cy3Bfibronectin along microcontact printed strips (n = 15-25 rectangular 10 × 210 µm strips) on multiple coverslips were measured for the three organosilane coatings, before and after PF-127 treatments. Specifically, the mean contrast fluorescence values ± standard deviation were calculated as previously described [8] as: where C is the mean contrast value, Ion is the mean fluorescence intensity within the rectangular fibronectin strips where fibronectin transfer occurs, and Ioff is the mean fluorescence intensity inbetween strips, where no transfer should occur. These results are reported in Supporting Table S1.