Micropost arrays to model ECM fiber obstacles during cell migration in confinement

Abstract

Micropost arrays are a valuable platform for studying cell migration in controlled microenvironments. These arrays enable researchers to mimic how migrating cells deform their nuclei in reaction to physical constraints presented by extracellular matrix (ECM) fibers during processes such as wound healing and cancer metastasis. By tuning micropost patterns and sizes, researchers can not only explore the impact of confining spaces created by such patterns on cell migration, signaling and nuclear deformation, but also simultaneously quantify cell-generated forces and forces of nuclear deformation. This paper outlines methods to design and fabricate ECM-fiber mimicking micropost arrays for use in fundamental studies of confining cell migration.

1. Introduction

Biological processes such as wound healing and cancer invasion involve the migration of cells through confining spaces in the interstitium. Extracellular matrix (ECM) fibers present obstacles that cells must overcome in order to migrate [1]. In particular, cells must squeeze the nucleus, which is the largest organelle in the cell, into these spaces. Being composed of structures such as the nuclear lamina and chromatin [2], the nucleus mechanically resists deformation, and imposes a limit on the smallest cross-section that cells can move through [3, 4]. It has been shown that cells cannot migrate through confinement if the confining space is less than 10% of the cross-sectional area of the nucleus [5]. Thus, understanding how cells deform their nuclei to navigate through these tight spaces, and how the nucleus might limit migration, is essential for a complete understanding of cell migration in the tissue interstitium.

In vitro model systems have proved to be powerful in visualizing nuclear deformation during confining cell migration. One type of model system involves PDMS (Polydimethylsiloxane) devices that provide smooth, contiguous surfaces of narrow cross-section, through which the nucleus must be squeezed for cells to migrate [3, 6]. While such model systems mimic contiguous surfaces that may be present in tissue, they do not effectively mimic any fibrous obstacles that may impede cellular motion in tissue. An alternative approach is to present cells migrating on a 2D surface with a ‘forest’ of upright, slender, soft microposts [7]. Choosing the micropost dimensions and patterns in such a way that the inter-post distance is smaller than nuclear sizes, enables the study of nuclear deformation as it is squeezed through slender obstacles like ECM fibers. Adjusting the height, diameter and the Young’s modulus of the material that the microposts are made of, allows the tuning of the stiffness of the microposts to be of the same order of magnitude as individual ECM fibers. This paper outlines the methods for the fabrication of micropost arrays, and for the use of these arrays for cell migration and nuclear deformation studies.

2. Materials

2.1. Micropost Design and Fabrication

1. Designing software (Autodesk, AutoCAD 2018).

2. Maskless aligner (Heidelberg Instruments, MLA-150 DWL 66 laser writer).

3. Contact aligner (Karl Suss, MA-6 contact aligner).

4. P-type silicon wafer (100 mm diameter) (MTI Corporation, SIBa101D0525C1R1US).

5. AZ1512 photoresist (MicroChemicals, 1A001512).

6. Chrome-coated glass mask (University Wafer, 272336).

7. (Tridecafluoro-1, 1, 2, 2-tetrahydrooctyl)-1-trichlorosilane (T2492, UCT Inc., Bristol, PA, USA) for Chemical Vapor Deposition.

8. Vacuum Desiccator (SP Bel-Art, B002VBW9SM).

9. Deep Reactive Ion Etching equipment (Oxford DRIE).

10. Oxygen plasma etcher (PlasmaEtch, PE-25).

11. 4:1 AZ-300MIF solution (developer) (MicroChemicals, 1004145).

12. SYLGRAD 184 Silicone Elastomer Kit (Dow Corning, Midland, MI, USA) for PDMS (dimethylpolysiloxane).

13. Scanning electron microscope.

14. Glass-bottom dishes (No. 1.5, 35 mm, FD35–100, WPI Inc., Sarasota County, FL, USA).

15. 22 mm × 22 mm glass coverslips (No. 1.5) or glass bottom dish.

16. 100% ethanol.

17. Silane solution (T-2492, UCT Inc.)

2.2. Preparing Microposts for Cell-Culture

1. 200-proof ethanol.

2. 70% ethanol.

3. 40% ethanol.

4. DI water.

5. Rhodamine-fibronectin (r-FN).

6. 1X PBS (Phosphate-buffered saline).

7. 4% paraformaldehyde (PFA).

8. Suitable cell culture media.

9. Pipettes (20–200 μl and 1–10 μl)

10. Vacuum Aspirator.

11. Cell culture incubator.

12. Live-cell or fixed cell fluorescent probes (optional).

2.3. Deflection Measurements

Image Analysis software (Mathworks, Matlab 2019a).

3. Methods

3.1. Micropost Design and Fabrication

We designed the micropost patterns in AutoCAD 2018.

3.1.1. Designing Micropost Patterns

1. Choose the side length of the hexadecagon so that it encloses an area equivalent to the average cell spreading area. (see Note 1)

2. Ensure the center-to-center distance between microposts results in a minimum edge-to-edge distance that is less than the average nuclear diameter. (see Note 2)

3. Array the hexadecagon pattern onto an area of 1 cm by 1 cm.

3.1.2. Patterning and Etching

1. Coat a P-type silicon wafer (diameter 100 mm) with a 1-micron layer of AZ1512 photoresist.

2. Pattern the AutoCAD design onto a chrome-coated glass mask using a Maskless aligner.

3. Mount the glass mask on the silicon wafer using a contact aligner and expose the photoresist-coated wafer to ultraviolet (UV) light through the glass mask.

4. Use a UV source with a wavelength of 365 nm (i-line), an intensity of 10 mW/cm², and an exposure duration of 5–10 seconds, ensuring optimal transfer of the mask pattern onto the photoresist layer.

5. Develop the wafer by immersing it in a 4:1 AZ-300MIF solution for 60–90 seconds at room temperature.

6. Monitor the process visually or use a microscope to ensure that the unexposed areas of the photoresist layer are fully removed, leaving behind the desired micropost pattern.

7. Adjust the development time slightly if necessary based on the photoresist thickness and exposure conditions.

8. Place the wafer in the Deep Reactive Ion Etching equipment and perform etching using the Bosch process. Use SF₆ (50 sccm) for etching, C₄F₈ (90 sccm) for passivation, at a chamber pressure of 20 mTorr, platen temperature of −20°C, and RF power of 200 W.

9. Etch for 10–15 minutes depending on the desired depth, monitoring intermittently for uniformity using a Scanning Electron Microscope and repeat the development and etching steps iteratively until the desired micropost height is achieved.

10. After etching, transfer the wafer to an oxygen plasma cleaner. Perform oxygen plasma treatment with an oxygen flow rate of 80 sccm, pressure of 100 mTorr, RF power of 150 W, for 10 minutes, or until the photoresist is fully removed. Inspect the wafer to ensure no residue remains.

11. Apply a ~50 nm thick coating of (tridecafluoro-1,1,2,2-tetrahydrooctyl)-1-trichlorosilane on the wafer using Chemical Vapor Deposition (CVD). Place the clean wafer in a vacuum desiccator with 1 mL of the undiluted silane solution, and maintain a static vacuum for 1–2 hours to ensure uniform coating. (see Note 3)

3.1.3. Preparing Negative Molds

1. To prepare a negative mold from the patterned master silicon wafer using PDMS (dimethylpolysiloxane), mix the silicon elastomer kit at a base-to-cross-linking agent ratio of 10:1 (w/w).

2. Cure the PDMS mixture at 60°C for 2 hours.

3. Peel the negative mold from the wafer and cut it into blocks containing the patterned area.

3.1.4. Fabricating PDMS Microposts

1. Evenly spread a thin layer of uncured PDMS, mixed at a base-to-cross-linking agent ratio of 10:1 (w/w), over a block of the negative mold. (see Note 4)

2. Carefully press the block onto a glass-bottom dish or a glass coverslip.

3. Cure the dish with the negative mold at 60°C for 2 hours to form the micropost topology on the glass surface.

4. Carefully peel off the negative mold after curing (see Figure 1a and 1b).

5. Submerge the newly cast molds in 100% ethanol for further processing.

Figure 1:

(a) Schematic illustrating the setup of PDMS microposts (red), formed after peeling away the negative PDMS mold (blue) from the glass coverslip (grey). (b) SEM image of PDMS microposts. (c) Schematic showing cells seeded on the microposts. (d) High-resolution confocal image of a cell expressing GFP-BAF and microposts labeled with r-FN (top panel), with the bottom panel showing the x-z reconstruction of the same nucleus.

3.2. Imaging for Cell-Culture with microposts

3.2.1. Washing micropost

1. Begin by replacing the 200-proof ethanol with 70% ethanol through serial dilutions, followed by a subsequent dilution to 40% ethanol. (see Notes 5–8)

2. Pipette and aspirate with DI water 3 times to remove 40% ethanol. (see Note 9)

3.2.2. Coating with Rhodamine-Fibronectin (r-FN)

1. Add ~20 μl of 1λ r-FN solution to the remaining DI water without physically disturbing the dish.

2. Mix gently using a 1–10 μl pipette.

3. Incubate the microposts with r-FN for 2 hours.

4. Wash the r-FN solution three times with 1X PBS.

5. Aspirate all the PBS from the corner of the dish or well. (see Note 10)

6. Add cell media to the dish, aspirate, and repeat this process three times to ensure thorough removal of any residual PBS.

3.2.3. Seeding Cell

1. Before seeding cells, add at least 2 ml of cell media to the 35 mm dish.

2. Seed cells drop-by-drop over the glass area.

3. Maintain cell confluence below 50% to facilitate optimal cell migration.

4. Mix the cell suspension by pipetting gently using a 20–200 μl pipette.

5. Incubate the cells. (see Note 11)

3.2.4. Live-Cell Imaging

1. Wash the sample with 1X PBS and replace the media.

2. Add appropriate live-cell stains and incubate.

3. Make sure the microposts are within the working distance of the objective of choice else flip the coverslip upside down for imaging.

4. The sample is now ready for imaging. (see Figure 1c, 1d and Note 12)

3.2.5. Fixed-Cell Imaging

1. Aspirate all media.

2. Wash the sample three times with 1X PBS.

3. Fix the cells with 4% paraformaldehyde (PFA) for 15–18 minutes at room temperature.

4. After fixation, wash the sample three times with 1X PBS.

5. The sample is now ready for staining.

3.3. Deflection Measurements

1. As migrating cells deform their nuclei against the labeled microposts, capture confocal Z-stacks in a time-lapse format using the desired microscopy.

2. Utilize MATLAB 2019a or a similar image processing software for automated analysis.

3. Implement centroid-detection-based particle-tracking routines to track micropost centroids.

4. Process raw images with bandpass filters and brightness thresholds to isolate the cross-sections of the top and bottom of the microposts from background noise.

5. Fit 2-D Gaussians to pixel intensities around the top and the bottom micropost section to determine the centroid positions respectively.

6. Identify the peak of the Gaussian fit as the micropost centroid.

7. Calculate the deflection vector as the difference between the centroids of the bottom and top of each micropost.

8. Subtract the deflection vector just before nuclear contact from subsequent time points.

9. Draw a line connecting the two protruding nuclear surfaces adjacent to the micropost.

10. Rotate this line 90° to determine the normal direction.

11. Determine the differential deflection vector component along the normal direction in successive frames by calculating cosines of the angle between the normal vector and the differential deflection vector. (see Figure 2)

12. Use $\omega =\frac{8EI}{l^4}\delta$ to quantify the uniformly distributed force exerted by microposts on the nucleus. (see Note 13 and 14)

13. Multiply the uniformly distributed load force per unit length by the micropost length (5 μm) to determine the total force exerted during nuclear contact.

Figure 2:

The schematic illustrates the calculation of a differential deflection vector ($\overrightarrow{r}$) as the change in the micropost’s deflection after contacting the nucleus, followed by its projection onto the normal to the deformed nuclear surface (see Note 15).

4. Notes

1. For example, we placed the circles of diameter 1 μm on the corners of an equilateral hexadecagon with a side length of 7.71 μm to customize the micropost patterns for NIH 3T3 fibroblasts which have an average spreading area of ~1200 μm².

2. For example, in the design we tested, the edge-to-edge distance between adjacent micropost was 6.6 μm which is less than the average nuclear diameter of ~10 μm for NIH 3T3 fibroblasts.

3. This treatment facilitates minimal adhesion of PDMS to the wafer, ensuring ease of peeling the negative mold.

4. Spreading 20 microliters of PDMS over an area of approximately 1 cm by 1 cm on the negative PDMS mold will result in a thin film with a thickness of 50–70 microns. This thickness is within the working distance of high numerical aperture, high magnification objectives such as 40x or 60x, facilitating high-resolution imaging.

5. PDMS is hydrophobic. Place the dish or well plate on a perfectly flat surface before aspirating to avoid drying out the microposts. If dried out, the microposts may bend and stick to the bottom surface.

6. Add solutions drop by drop along the sides of the dish to minimize strong convective currents.

7. Do not add solutions directly onto the glass area, rather, add it towards the edge of the well or dish.

8. Leave at least 1 ml of solution or always maintain a minimum liquid depth of around 1 cm in the dish during aspiration.

9. During the last DI water wash, leave the microposts submerged in 2–3 ml of DI water for at least 15 minutes to allow adsorbed ethanol to diffuse out followed by pipetting more water and aspirating

10. If the r-FN coating is effective, PDMS will retain a thin layer of water despite the aspiration of all liquid.

11. To facilitate the clear visualization of nuclear shape, it is recommended that cells express a nuclear envelope GFP-tagged protein like GFP-laminA or GFP-BAF. Additionally, F-actin can be visualized using Lifeact or similar approaches, allowing for the study of cytoskeletal dynamics together with nuclear dynamics.

12. For optimal imaging quality and effective optical sectioning of microposts, it is essential to use a high numerical aperture (NA) objective. We utilized a 60x/1.4 NA oil immersion objective to capture high-resolution images. The pinhole size was set to 1 Airy Unit (A.U.) with a 488 nm laser, corresponding to an optical section thickness of approximately 600 nm. To satisfy the Nyquist criterion, a z-step size of 250 nm was used, with a time resolution set to either 5 or 10 minutes.

13. In $\omega =\frac{8EI}{l^4},\omega$ represents the uniformly distributed force exerted by the nucleus across the length of the micropost. $E$ is the Young’s modulus of PDMS, measured to be 2.21±0.04 MPa for PDMS mixed at a 10:1 base-to-cross-linking agent ratio and cured at 60°C for 2 hours, $I=\frac{{\pi \text{r}}^4}{4}$ is the area moment of inertia of the circular cross-section of the micropost of radius $\text{r}$ and $l$ is the length of the micropost. $\delta$ is the measured deflection of the top of the micropost relative to micropost bottom.

14. Flexural rigidity ($EI$) is calculated using the formula $EI=E\times I$, where $E$ represents Young’s modulus and $I$ denotes the second moment of inertia for cylindrical geometry. Collagen fibers have a characterized Young’s modulus ranging from 100 MPa to 360 MPa and an average diameter of 0.41 ± 0.09 μm[8], yielding calculated flexural rigidity ($EI$) between $1.37\times {10}^{-16}{\text{Nm}}^2$ and $4.94\times {10}^{-16}{\text{Nm}}^2$. The PDMS microposts fabricated using the methods described here, with a diameter of 1 micron and a Young’s modulus of 2.21±0.04 MPa, have an equivalent $EI$ of $~1.11\times {10}^{-16}{\text{Nm}}^2$, similar to that of the collagen fibers.

15. All deflection vectors are calculated by subtracting the centroid of the top of the micropost from that of the bottom. The vector $\overrightarrow{x}$ represents the deflection of the microposts before nuclear contact, while $\overrightarrow{y}$ denotes the deflection vector after nuclear contact. The unit vector $\overrightarrow{n}$ is directed normal to the nuclear surface after contact with the micropost. The vector $\overrightarrow{r}$ is the difference between $\overrightarrow{x}$ and $\overrightarrow{y}$.