Intrinsic Response Towards Physiologic Stiffness is ... - Springer Link

Cell Biochem Biophys DOI 10.1007/s12013-017-0834-1

ORIGINAL PAPER

Intrinsic Response Towards Physiologic Stiffness is Cell-Type Dependent Michael Reimer1 Silviya Petrova Zustiak2 Saahil Sheth2 Joseph Martin Schober ●

●

●

1

Received: 5 July 2017 / Accepted: 13 October 2017 © Springer Science+Business Media, LLC 2017

Abstract In the continuous search for better tissue engineering scaffolds it has become increasingly clear that the substrate properties dramatically affect cell responses. Here we compared cells from a physiologically stiff tissue, melanoma, to cells isolated from a physiologically soft tissue, brain. We measured the cell line responses to laminin immobilized onto glass or polyacrylamide hydrogels tuned to have a Young’s modulus ranging from 1 to 390 kPa. Single cells were analyzed for spreading area, shape, total actin content, actin-based morphological features and modification of immobilized laminin. Both cell types exhibited stiffness- and laminin concentration-dependent responses on polyacrylamide and glass. Melanoma cells exhibited very little spreading and were rounded on soft (1, 5, and 15 kPa) hydrogels while cells on stiff (40, 100, and 390 kPa) hydrogels were spread and had a polarized cell shape with large lamellipodia. On rigid glass surfaces, spreading and actin-based morphological features were not observed until laminin concentration was much higher. Similarly, increased microglia cell spreading and presence of actin-based structures were observed on stiff hydrogels. However, responses on rigid glass surfaces were much different. Microglia cells had large spreading areas and elongated shapes on glass compared to hydrogels even when immobilized laminin density was consistent on all gels. While cell spreading and shape varied with Young’s

* Joseph Martin Schober [email protected] 1

Department of Pharmaceutical Sciences, Southern Illinois University Edwardsville, Edwardsville, IL, USA

2

Department of Biomedical Engineering, Saint Louis University, St Louis, MO, USA

modulus of the hydrogel, the concentration of f-actin was constant. A decrease in laminin immunofluorescence was associated with melanoma and microglia cell spreading on glass with high coating concentration of laminin, indicating modification of immobilized laminin triggered by supraphysiologic stiffness and high ligand density. These results suggest that some cell lines are more sensitive to mechanical properties matching their native tissue environment while other cell lines may require stiffness and extracellular ligand density well above physiologic tissue before saturation in cell spreading, elongation and cytoskeletal reorganization are reached. Keywords Hydrogel Stiffness Actin Laminin Melanoma Microglia ●

●

●

●

●

Introduction Recent research on the effect of substrate stiffness has shown that most cell types spread more and have a more elongated morphology on stiffer substrates [1]. Further, spreading area and shape factor typically follow a saturation behavior, where no additional spreading and elongation is observed above a threshold stiffness or a biphasic behavior or where maximum spreading and elongation is observed at a particular stiffness for a cell type [2–5]. Other research has also shown that spreading and elongation would dramatically drop below a critical threshold stiffness [6]. Substrate stiffness has been shown to affect other cell behaviors such as motility [7], where the authors demonstrated that cells preferentially migrate from soft to stiff substrates. Substrate stiffness and other physical properties have also been shown

Cell Biochem Biophys

to regulate stem cell differentiation [8], cell phenotype [9], tumorigenic potential [10] and cardiomyocyte organization [11]. In particular for melanoma, responses to stiffness have been varied, and cell- and phenotype-dependent. For example, WM35 cells have been shown to exhibit greater spreading and more distinct f-actin fibers on stiff substrates (13 kPa) vs. soft ones (0.6 kPa), while metastatic A375 melanoma cells showed overall smaller spreading and more irregular morphology independent of substrate stiffness [12]. VMM18 melanoma cells proliferated independent of stiffness (0.15–9.6 kPa) [2], and VM115 cells showed a significant decrease in cell and nucleus surface area when cultured on a soft vs. a stiff substrate (0.75–2.92 MPa) [13]. However, one needs to be aware that different studies consider vastly different stiffness ranges. For example, when using PEG hydrogels, the authors termed 0.6 kPa to represent a soft and 13 kPa to represent a stiff substrate [12]. While using a polydimethyl siloxane substrate, a different study referred to 0.75 MPa as a soft and 2.92 MPa as a stiff substrate [13]. Hence, when discussing substrate stiffness, one needs to consider the physiologically relevant stiffness for the cell type of interest. For example, the modulus of skin has been reported in the range of 0.2–20 MPa depending on the type of measurement, tissue processing, direction of measurement and the age of the patient [14, 15], while the modulus of brain is much lower, near 1–2 kPa [16]. It has also become clear that the effects of ligand density in addition to and in conjunction with substrate stiffness need to be considered to properly emulate and understand in vivo cell environments. Aortic smooth muscle cells have been shown to spread more and form more focal adhesions on rigid glass and stiff hydrogel substrates compared to soft hydrogels where maximum cell spreading occurred on intermediate densities of immobilized collagen [17]. In another study of primary hematopoietic and progenitor cells, the authors showed increased cell spreading in response to increased substrate stiffness or increased ligand concentration [18]. Little is known of the relationship between responses, such as cell spreading, elongation and cytoskeletal reorganization, to substrate properties and the native tissue from which the cell originated. For example, do cells originating from a soft tissue type, such as the brain, retain greater responses to softer substrates in vitro? Further, above a particular threshold in substrate stiffness, do cell responses remain constant or decline? In our current studies we compared responses of cells originating from a physiologically stiff tissue, skin melanoma [14, 15], to cells isolated from a physiologically soft tissue, brain [16]. A laminin coating was used for all conditions, since laminins are important extracellular matrix proteins for both brain [19] and melanoma tissue [20]. We show that maximum melanoma cell response was observed on hydrogels near 40

or 100 kPa and a low ligand concentration, and thereafter at higher stiffness responses declined. In comparison, the same cell response was not achieved on rigid glass (50–90 GPa [21]) until the ligand concentration was higher. A much different response profile was observed in brain microglia cells. Although the native tissue environment of microglia cells is much softer (stiffness of brain is 1–2 kPa [16]), cell responses continued to increase and shape elongation was observed at supraphysiologic stiffness. In both cell types, a decrease in laminin immunofluorescence was observed on glass surfaces coated with the highest laminin concentration, suggesting an active cellular process triggered only during supraphysiologic conditions. To achieve maximum response, certain cell types may require a substrate stiffness within the range of the native tissue environment. However, in other cell types, artificial responses may be induced in environments with supraphysiologic stiffness or ligand density. Such knowledge may shed light into matrix conditions that regulate cell spreading, cytoskeletal re-organization, motility and ultimately invasion towards tissues with distinctly different mechanical and biochemical properties.

Materials and Methods Materials Two percent bis-acrylamide solution, 40% acrylamide solution, ammonium persulfate (APS) and N,N,N′,N′-tetramethylethylenediamine (TEMED) were purchased from Bio-Rad Laboratories, Inc. (Hercules, CA). Dulbecco’s Modified Eagle’s Medium (DMEM, with 4.5 g/L glucose, L-glutamine and sodium pyruvate), 18 × 18 mm #2 glass coverslips, phosphate-buffered saline (PBS, without calcium and magnesium) and 0.05% Trypsin/0.53 mM EDTA solution were purchased from Corning Life Sciences (Manassas, VA). Mouse laminin isolated from EngelbrethHolm-Swarm sarcoma, Alexa Fluor-546 phalloidin, Alexa Fluor-488 anti-rabbit antibody, Hoechst 33258, and sulfosuccinimidyl 6-(4′-azido-2′-nitrophenylamino) hexanoate (sulfo-SANPAH) were purchased from Thermo Fisher Scientific, Inc. (Waltham, MA). PlusOne Repel-Silane ES (2% w/v solution of dimethyldichlorosilane in octamethylcyclooctasilane), PlusOne Bind-Silane (γ-methacryloxypropyltrimethoxysilane) and bovine serum albumin (BSA) fraction V lyophilized powder were purchased from GE Healthcare Life Sciences (Pittsburgh, PA). Fetal bovine serum (FBS) used for B16F10 cell culture media was from Atlanta Biologicals, Inc. (Flowery Branch, GA) and FBS for BV2 cell culture media was from Corning Life Sciences (Manassas, VA). The rabbit anti-laminin polyclonal antibody and the mouse anti-vinculin antibody were purchased from Abcam (Cambridge, MA). Triton X-100 and the rabbit


polyclonal anti-myosin IIA antibody were purchased from Sigma-Aldrich (St. Louis, MO), penicillin-streptomycinamphotericin B solution from MP Biochemicals (Santa Ana, CA), Aqua-Poly/Mount from Polysciences, Inc. (Warrington, PA), and 40% para-formaldehyde were [purchased from Electron Microscopy Sciences (Hatfield, PA). The secondary anti-rabbit-TRITC and anti-mouse-Cy5 antibodies were purchased from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA).

Polyacrylamide Hydrogel Preparation For polyacrylamide (PA) hydrogel preparation, a large hydrophobic surface was prepared by covering a glass plate with a thin layer Repel-Silane solution and incubating for 15 min at room temperature. Excess Repel-Silane was removed and the glass was polished with a dry Kimwipe. Working Bind-Silane solution was prepared by mixing 8 mL ethanol, 200 μL glacial acetic acid, 30 μL stock BindSilane and 1.8 mL water. Eighty μL of working Bind-Silane solution was added to the top surface of 18 × 18 mm coverslips and incubated for 15 min at room temperature. Excess solution was removed and coverslips were dried and polished using a Kimwipe. For preparation of 1 kPa hydrogels, coverslips were dipped for 30 s once in water and dried with a Kimwipe. To initiate polymerization, 5 μL 10% APS and 0.5 μL TEMED were mixed with 1 mL of the acrylamide-bis-acrylamide solutions. Eighty microliter of the reaction was immediately added onto the hydrophobic glass surface and the glass coverslips, hydrophilic-side down, were carefully placed on top of the hydrogel precursor solution and incubated at room temperature for 1 h. The coverslips were then removed from the glass plates and placed in water for 30 min to remove any non-polymerized acrylamide from the gels. A working concentration of 500 μM sulfo-SANPAH was prepared by diluting a 100 mM stock solution (solubilized in DMSO) with de-ionized (DI) water. 350 μL working sulfo-SANPAH solution was immediately added onto the hydrogels and exposed to high intensity ultraviolet (UV) light for 5 min (254 nm; CL-1000, UVP, Upland, CA). Excess sulfo-SANPAH was removed and hydrogels were washed in 0.3 M PBS pH 7.5 for 5 min. Table 1 Acrylamide and bisacrylamide concentrations for the preparation of polyacrylamide hydrogels of elastic modulus as measured by compression mechanical testing

Elastic Modulus (kPa)

The working laminin solutions were prepared by dilution with 0.3 M PBS pH 7.5. 350 μL laminin was added to each sulfo-SANPAH-activated hydrogel and incubated at room temperature for 1 h. For laminin absorption onto untreated glass surfaces, coverslips were placed on top of 80 μL working laminin solution for 1 h at room temperature. The laminin solution was removed, coverslips (with and without hydrogels) were washed once in PBS and then placed in 35 mm dishes containing 2 mL growth media prepared with freshly thawed FBS. The Young’s modulus of PA gels was controlled through varying the volumes of 40% acrylamide, 2% bis-acrylamide and water as shown in Table 1. Mechanical Testing Hydrogel slabs (2 mm thick, 9 mm in diameter) were swollen in PBS overnight. Residual buffer from the hydrogel surface was carefully blotted with a KimWipe prior to measurement to avoid slipping. The Mechanical Testing System (MTS Criterion Model 42, MTS Systems Corporation, Eden Prairie, MN) was fitted with a 100 N load and samples of each hydrogel were placed onto the steel plates. Each test had a rate of 10 mm/min, preload of 0.05 N, data acquisition rate of 50 Hz, and preload speed of 1 mm/min. Samples were compressed to 30% for the softer gels (1 and 5 kPa) and 70% for the stiffer gels (15, 40, 100, and 390 kPa). The elastic modulus (kPa) was recorded by the MTS TW Elite software. Cell Attachment and Fluorescence Microscopy B16F10 mouse melanoma cells (ATCC, Manassas, VA) and BV2 mouse microglia cells [22] were maintained in growth media (DMEM with penicillin-streptomycinamphotericin B solution and 10% FBS) in 25 cm2 flasks until 80–90% confluent. The cells were removed by 2 min incubation with a 0.05% Trypsin/0.53 mM EDTA solution at 37 °C and immediately added to the 35 mm dishes containing media and the hydrogels cross linked with laminin or glass coverslips passively coated with laminin. The dishes were placed in a 37 °C, 5% CO2 incubator for 1 h, after which the hydrogels and coverslips were placed in

40% Acrylamide Solution (mL)

2% Bis-Acrylamide Solution (mL)

De-Ionized Water (mL)

1.4 ± 0.9 (1)

3.1

0.3

21.6

4.6 ± 1.8 (5)

3.1

1.3

20.6

15.3 ± 8.5 (15)

5.0

1.3

18.8

39.3 ± 7.4 (40)

5.0

3.1

16.9

99.1 ± 16.7 (100)

7.5

3.1

14.4

389.7 ± 45.5 (390)

9.4

16.3

0.0


fixation buffer containing 4% para-formaldehyde and 0.1% Triton X-100 in PBS for 1 h. After fixation, samples were washed in DI water and then blocked with 2% BSA solution in PBS for 20 min at room temperature. Samples were incubated with primary antibodies for 20 min at 37 °C, washed in PBS, then incubated with Alexa Fluor-488 antirabbit antibody, Alexa Fluor-546 phalloidin and Hoechst 33258 DNA for the laminin quantification data or with Alexa Fluor-488 anti-rabbit antibody and anti-mouse Cy5 for vinculin and myosin IIA data. Coverslips were washed in PBS and water and mounted onto standard glass slides, hydrogel-side up. A drop of Aqua/Poly mount was added onto the hydrogels and then hydrogels were covered with a plain glass coverslip, so that an optical path through the hydrogels on the inverted microscope was avoided. B16F10 cell images were acquired with a Leica DMIRE2 inverted microscope (Leica Microsystems, Wetzler, Germany) using a 63X oil-immersion objective and BV2 cell images were acquired with a Leica DMi8 inverted microscope using a dry 40X objective, equipped with a 12-bit monochrome CCD camera. All images were acquired using Metamorph software.

fluorescence intensity analysis of actin and immobilized laminin, the actin and laminin image channels were combined and a 2.2 × 67 μm line was drawn across a selected section of each cell. The relative fluorescence intensity of actin and laminin vs. line position was plotted for each cell. After analysis, image intensity display was re-scaled, images were sharpened and then converted to 8-bit depth using Adobe Photoshop (Mountain View, CA). Statistical Analysis Results are reported as averages ± SEM. Statistical significance between multiple samples was tested by one-way ANOVA followed by a Tukey’s post hoc test or unpaired ttest comparisons, where p < 0.05 was considered significant. Only selected comparisons were made, indicated by the brackets above bars in each graph. Experiments were repeated three times where one experiment consisted of a single coverslip for each condition. For single cell analysis, a minimum of two cells per image, from a minimum of seven images from the three independent experiments were analyzed for each condition.

Image Analysis

Results To quantify laminin immobilization on hydrogel and glass surfaces, images were analyzed for integrated pixel intensity in the anti-laminin, green-filter channel. Three samples ranging in size from 10,000 to 20,000 pixels were taken from each image and integrated for fluorescence intensity. Background fluorescence measurements were taken from gels where BSA was cross linked instead of laminin. The same procedure was used to quantify laminin passively absorbed onto glass surfaces. For cell area, roundness index and f-actin measurements, cell perimeter was manually traced in the phalloidin, red-filter channel. Using the integrated morphometry functions in Metamorph software, the following primary parameters were acquired inside each cell perimeter: total pixel area, integrated pixel intensity, and object breadth (defined as the two points on the cell perimeter with the greatest distance). From the primary parameters, the following calculated parameters were determined: cell area in μm2 and roundness index. Area in μm2 was calculated from an image scale where 1 pixel = 140 nm for the 63X lens and 220 nm for the 40X lens. Roundness index was taken as the area of the cell divided by the area of a circle with a diameter equal to the object breadth. Thus, the roundness index is a dimensionless parameter where a perfect circle has an index of one and a highly elongated object an index approaching zero. To minimize effects of intensity vignetting within each image, pixels were sampled from a circle on the image center encompassing 75% of the total image area. For the linescan

Laminin Coating of Polyacrylamide Gels and Glass Immobilization efficiency of extracellular matrix proteins may vary between surface types, immobilization mechanisms, and as a result of pore size differences across PA stiffness. Thus, we probed hydrogel and glass surfaces with an anti-laminin antibody to determine protein immobilization. Hydrogels ranging from 1 to 390 kPa stiffness were coated with 5 μg/mL laminin, while untreated glass coverslips were passively coated with 0.2, 1.0, or 5.0 μg/mL laminin (Fig. 1). Within the entire range of hydrogel stiffness, the fluorescence of immobilized laminin did not significantly change, indicating no change in concentration. The concentration of laminin on glass sharply increased at 1.0 and 5.0 μg/mL, where it was also much higher compared to the hydrogels and likely not saturated. Stiffness and Ligand Density-Dependent Cell Responses: Cell Morphology When investigating the stiffness- and ligand densitydependent responses of B16F10 melanoma cells and BV2 microglia cells, we observed that after 60 min the cells responded to the surfaces in both ligand density- and stiffness-dependent manners (Fig. 2). On the 1 kPa hydrogels the B16F10 cells were attached, but did not form any defined actin structures, while BV2 cells formed actin


ruffles. On the 40, 100, and 390 kPa hydrogels the B16F10 cells formed actin ruffles and lamellipodia, while the cells on 5 and 15 kPa hydrogels formed only actin ruffles. BV2

Fig. 1 Immobilization of laminin on polyacrylamide (PA) hydrogels and glass. Hydrogels with varying stiffness (1–390 kPa) were coated via the cross linker sulfo-SANPAH with 5.0 μg/mL laminin (Ln) and glass coverslips were passively coated with 0.2, 1.0, or 5.0 μg/mL laminin. The concentration of laminin immobilized on hydrogel and glass surfaces was quantified using a polyclonal anti-laminin antibody (n = 30–56 samples from a minimum of 20 images for each group, error bars = SEM)

Fig. 2 Stiffness- and ligand density-dependent actin cytoskeleton changes in a B16F10 and b BV2 cells. The cells were plated onto hydrogels with varying stiffness (1–390 kPa) crosslinked with 5.0 μg/ mL laminin (Ln) or plated onto glass (50–90 GPa) passively coated with 0.2, 1.0 or 5.0 μg/mL laminin. Samples were incubated for 60 min to allow time for cell attachment and actin cytoskeleton responses.

cells produced actin ruffles within the entire range of hydrogel stiffness, but formed prominent elongated shapes only on 40, 100, and 390 kPa. The cells on glass surfaces responded to laminin in a concentration-dependent manner. The B16F10 cells on glass coated with 0.2 and 1.0 μg/mL laminin formed actin ruffles and were similar to cells on the 5 and 15 kPa hydrogels. The B16F10 cells did not form actin ruffles or lamellipodia on glass until the concentration of immobilized laminin was significantly higher (by 25fold) than that of laminin on the hydrogels (5.0 Ln glass, Fig. 2a). Thus, when ligand density was controlled, actin ruffling and lamellipodia formation were much greater on hydrogels with stiffness that was within physiologic range compared to the supra-physiologic stiffness of glass. On glass the BV2 cells had a distinctly different response compared to the B16F10 cells. The BV2 cells formed highly elongated actin bundles on all concentrations of laminin immobilized onto glass, but not on laminin immobilized to even the highest stiffness hydrogel. Thus, actin structures may be driven by supra-physiologic stiffness in this particular cell type. Stiffness and Ligand Density-Dependent Cell Responses: Cell Spreading Area and Cell Shape When quantifying cell spreading area and cell shape (i.e., roundness index) we observed that cells seeded on the hydrogels exhibited stiffness-dependent spreading area (Fig. 3) and shape change (Fig. 4).

Samples were fixed for 60 min at room temperature, and stained for actin (green) and DNA (blue). The color combined images show 3–5 representative cells in each condition. a Arrowheads point to examples of actin ruffling and asterisks mark examples of lamellipodia in B16F10 cells, and b arrowheads point to actin ruffling and asterisks mark elongated structures in BV2 cells


Fig. 3 Projected cell area on laminin (Ln) immobilized on hydrogel and glass surfaces. a B16F10 or b BV2 cells were plated onto hydrogels with a range of stiffness (1–390 kPa) coated with 5.0 μg/mL laminin or glass (50–90 GPa) passively coated with 0.2, 1.0, or 5.0 μg/mL

laminin. Samples were incubated for 60 min, fixed and stained for actin and DNA. Actin fluorescence images were analyzed for average cell area expressed in μm2 (error bars = SEM)

Fig. 4 a B16F10 and b BV2 cell shape on laminin (Ln) immobilized on hydrogels and glass. Cells were plated onto hydrogels with a range of stiffness (1–390 kPa) coated with 5.0 μg/mL laminin or glass (50–90 GPa) passively coated with 0.2, 1.0, or 5.0 μg/mL laminin. Samples were incubated for 60 min, fixed and stained for actin and

DNA. Actin fluorescence images were analyzed for average cell roundness, expressed as cell area divided by the area of a circle with a diameter equal to the longest chord across the cell. A circular cell has a roundness index of 1 and a highly elongated cell a roundness index approaching 0 (error bars = SEM)

On the 1 kPa hydrogels B16F10 cells were attached, but remained near the baseline area and had a high roundness index. The B16F10 cells reached maximal spreading area (570 μm2) and minimal roundness index (0.55) on a hydrogel stiffness of 40 kPa; no further change was observed with further increase in hydrogel stiffness. On

glass, B16F10 spreading area and shape change occurred in a laminin concentration-dependent manner. At 0.2 and 1.0 μg/mL laminin on glass, when immobilization was equal to or exceeded the hydrogels, cell area was less and cell roundness was greater than that on the 40–390 kPa hydrogels. Only when the concentration of immobilized


Fig. 5 Actin cytoskeleton content in a B16F10 and b BV2 cells on laminin (Ln) immobilized on hydrogels and glass. Cells were plated onto hydrogels with a range of stiffness (1–390 kPa) coated with 5.0 μg/mL laminin or glass (50–90 GPa) passively coated with 0.2, 1.0, or

5.0 μg/mL laminin. Samples were incubated for 60 min, fixed and then stained for actin and DNA. Actin images were analyzed for total filamentous actin (total f-actin). Total f-actin is expressed in relative fluorescence units (error bars = SEM)

laminin was 10-fold greater on glass compared to the hydrogels (Fig. 1) was cell spreading and shape comparable between glass and the stiffer hydrogels (>40 kPa). BV2 microglia spreading area was low and roundness index was high on 1–15 kPa hydrogels. Spreading area sharply increased at 40 kPa and roundness index sharply decreased at 100 kPa. Laminin immobilized onto glass did not significantly change BV2 cell area compared to the stiffer hydrogels (>40 kPa), but did cause a further decrease in roundness index due to the presence of many elongated actin bundles on glass.

However, the concentration of f-actin in the B16F10 cells on glass coated with 1.0 μg/mL laminin was greater than in the cells on the hydrogels. The concentration of f-actin in BV2 cells on glass coated with 1.0 µg/mL laminin was not significantly different than the 390 kPa hydrogels. Although cell incubation time was relatively short, 60 min, we examined potential modification of the laminin immobilized to hydrogel and glass surfaces (Fig. 6). On the 40 kPa hydrogels and glass surfaces coated with 0.2 μg/mL laminin, no decrease in laminin immunofluorescence was detected; however, an increase in perinuclear laminin was noted particularly in the B16F10 cells. In contrast, we observed a decrease in laminin immunofluorescence underneath and immediately around both melanoma and microglia cells on glass coated with 5.0 μg/ mL laminin (Figs. 6c and f). Laminin immunofluorescence decrease was associated with nearly every BV2 microglia cell and about 50% of the B16F10 melanoma cells. Thus, these cells modify the immobilized laminin, but only on very rigid surfaces at high ligand density. In addition, we performed myosin IIA and vinculin immunofluorescence in both cell types on 40 kPa hydrogels and glass surfaces (Fig. 7). In B16F10 cells, myosin organized along actin stress fibers on the hydrogel and glass surfaces coated with 0.2 and 5.0 μg/mL laminin while large focal adhesions, indicated by the presence of vinculin, were observed only on glass coated with 5.0 μg/mL laminin (Fig. 7a). In BV2 cells, we did not observe myosin organization or large adhesion sites on the hydrogel or glass surface coated with 0.2 μg/mL laminin; however, myosin organization and large

Stiffness and Ligand Density-Dependent Cell Responses: Actin Content, Immobilized Laminin, Myosin IIA, and Vinculin As cells dynamically respond to the extracellular environment, actin is continuously polymerized and depolymerized between the filamentous (f-actin) and globular (g-actin) states. Cell spreading, shape changes, ruffling, and lamellipodia formation are all processes driven by actin cytoskeleton dynamics, thus, the stiffness- and ligand densitydependent changes we observed may result from a shift in equilibrium between the actin states [23]. We measured phalloidin staining on the hydrogels and glass (Fig. 5) to determine whether the spreading and shape changes were accompanied by alterations in the total concentration of factin. For both cell types, the concentration of f-actin across the range of hydrogel stiffness did not change significantly.


Fig. 6 Laminin (Ln) immunofluorescence on hydrogel and glass surfaces with a–c B16F10 and d–f BV2 cells. The cells were plated onto 40 kPa hydrogels coated with 5.0 μg/mL laminin or glass (50–90 GPa) coated with 0.2 and 5.0 μg/mL laminin. Samples were incubated for 60 min, fixed and then stained for actin and laminin. Linescan

intensity analysis was performed on the color combined images of actin (green) and laminin (blue). Arrowheads in the color combined images indicate locations of linsescans for each of the intensity plots of Ln fluorescence (blue) and actin fluorescence (green). Image sets are representative of three independent experiments

focal adhesions were present in BV2 cells on glass coated with 5.0 μg/mL laminin (Fig. 7b).

extracellular matrix proteins for both brain [19] and melanoma tissue [20] and because both cell types responded to laminin within a similar time frame of 45–60 min. To manipulate stiffness in a wide range of Young’s modulus (1–390 kPa), we used PA hydrogels, while the Young’s modulus of glass is 50–90 GPa [21]. Ligand density was manipulated on glass by passively coating coverslips with varying concentrations of laminin. The laminin concentration was kept constant for all PA gels of various stiffness, where equal coating density and uniformity were confirmed by quantifying immunofluorescence. Our results corroborate previous findings that the concentration of proteins bound to the surface of the hydrogels via the sulfoSANPAH crosslinker is independent of PA gel stiffness [24]. Ensuring that concentration of surface laminin was independent of PA gel stiffness confirmed that subsequent differential cell responses on the PA gels were due to substrate stiffness and not due to differences in the number of ligand sites.

Discussion Here, we describe the effect of substrate stiffness and ligand density on cell responses for two different cell types, namely B16F10 melanoma and BV2 microglia cells. The two different cell types were selected as they originate from tissues of significantly different stiffness: the Young’s modulus of skin is in the range of 0.2–20 MPa depending on the type of measurement, tissue processing, direction of measurement and the age of the patient [14, 15], while the modulus of brain is 1–2 kPa [16]. In comparison, the selected polyacrylamide hydrogels range in modulus from 1 kPa, which is close to that of brain tissue, to 390 kPa (~0.4 MPa), which is close to that of skin. A laminin coating was used for all conditions, since laminins are important


Fig. 7 Myosin IIA and vinculin immunofluorescence on hydrogels and glass surfaces with a B16F10 and b BV2 cells. The cells were plated onto 40 kPa hydrogels coated with 5.0 μg/mL laminin or glass coated with 0.2 and 5.0 μg/mL laminin. Samples were incubated for

60 min, fixed and then stained with anti-myosin IIA and anti-vinculin antibodies. Arrows indicate appearance of myosin organization and arrowheads mark focal adhesions

We examined the effect of substrate stiffness and ligand concentration on B16F10 melanoma and BV2 microglia cell responses. The PA gels were 250 μm in thickness prior to swelling in order to ensure that the cells did not feel the stiffness of the underlying substrate [25]. It is also important to note that we elected to image and analyze cells at 60 min post seeding, since at longer time points additional proteins from the serum media could lead to different surface protein coverage and altered cell responses. When examining cell morphology as a function of stiffness for the same ligand density, we made an unexpected observation. While the BV2 cells had the greatest morphology change on glass, the morphology of the B16F10 cells on the 0.2 μg/mL glass resembled that on the 15 kPa gel, but not that on the stiffest 100 and 390 kPa gels. Specifically, B16F10 cells formed only actin ruffles on the lowest ligand density on glass. B16F10 cells formed actin ruffles and lamellipodia when the gel stiffness was increased to 40 kPa or higher and when the laminin concentration on glass was increased to 1 μg/ mL or higher. Similar observations have not been reported before. Due to the higher coupling strength between laminin and PA (covalent bond) vs. laminin and glass (passive adsorption), we could speculate that cells perceive a different stiffness than the substrate modulus would indicate. However, this explanation alone does not account for the distinct behavior between the two different cell types. The BV2 cells were highly elongated on the laminin immobilized onto glass. In contrast, B16F10 cell area increased and roundness decreased with an increase in substrate stiffness up to a threshold Young’s modulus of 40 kPa, upon which no further significant changes were observed. Our data is consistent with other findings that show such saturation of stiffness-dependent cell responses [2, 4, 26]. The difference in response between the B16F10 and BV2 cells lines may be related, at least in part, to their

different tumorigenic potential. The B16F10 cell line is a highly invasive clone selected from mouse melanoma cells [27]. The ability of melanoma and potentially other cancer cell types to form lamellipodia on softer substrates could promote tumor escape and invasion to nearby tissue, where the elastic modulus is usually lower compared to the tumor mass (80 kPa in dermal layer vs. 300 kPa in melanoma) [28]. The BV2 microglia cell line was not derived from cancer, but from viral-transformed cells from mouse brain tissue preparations [29]. Although the native tissue of microglia cells is soft, high ligand concentration may be required for actin cytoskeleton responses because of lack of tumorigenic ability. Also, note that while we discuss our data in terms of a Young’s modulus, the selected PA hydrogels are viscoelastic [30], similarly to skin, while glass is a rigid substrate. Hence, the unusual response of the B16F10 cells to substrate stiffness might also be dictated by differences in substrate flexibility. For example, previous work has shown a higher percent cell elongation on a flexible as compared to a rigid substrate [31]. Upon quantifying the effect of laminin concentration on glass on B16F10 cell morphology, we noted an increase in cell area and decrease in roundness with an increase in ligand concentration. We fitted a straight line to the initial linear increase in cell area (prior to saturation). We then noted that the slope of the line was 11.7 for the cell area as a function of hydrogel stiffness (R2 = 0.99) and 44.7 for the cell area as a function of ligand density on glass (R2 = 0.99). Thus, our data indicated that ligand density on rigid glass substrate had a more pronounced effect on B16F10 cell responses than hydrogel stiffness. In contrast, BV2 cells exhibited a sharp increase in cell spreading area on hydrogels of modulus > 40 kPa, which was not statistically different than the area on glass; the cells did not exhibit cell area dependence on ligand-concentration on rigid glass


substrate. Thus, BV2 cells appeared to be more responsive to substrate stiffness. We further compared the extent of cell spreading and roundness on the PA gels to the glass of equivalent laminin density, namely 0.2 μg/mL. BV2 cell area was near maximum at 40 kPa and no further increase was observed on higher stiffness hydrogels or glass, while cell roundness continued to decrease to a minimum on glass. B16F10 cell spreading area and roundness on the glass was equivalent to that on the 15 kPa gels, or 50% lower cell spreading area and 15% higher roundness factor than the 40–390 kPa gels. Interestingly, literature suggests that the body of the cell responds not only to the glass but also to softer features on the glass surface (e.g., the ECM protein coating); hence, the glass modulus is perhaps exaggerated in the context of cell sensing [17]. Other work has also shown that cell spreading area shifts to lower ligand densities for decreasing stiffness; however, unlike our observation, the overall spreading has been shown consistently higher on higher stiffness [17]. We quantified the local laminin immunofluorescence and total f-actin as a function of substrate stiffness and ligand concentration. We did not observe an increase in total factin on the PA gels of different stiffness, regardless of notable differences in cell spreading, indicating that changes in cell responses were not the result of changes in f-actin content. On the other hand, in B16F10 cells the concentration of total f-actin was significantly higher on the rigid glass, with the highest concentration of f-actin on the intermediate ligand concentration of 1.0 μg/mL laminin, thereby ruling out the possibility that decreased area and increased roundness were a consequence of less cellular factin. F-actin in BV2 cells was not significantly altered when comparing the high stiffness hydrogels to glass coated with 0.2 and 1.0 µg/mL laminin. Cytoskeletal assembly has been shown to correlate well with cell spreading [17]; we noted such correlation between cell morphology (i.e. actin ruffling and lamellipodia) and cell spreading area and roundness, but not in the concentration of total f-actin. The results in both cell types show that the changes in cell area and shape are not governed by the concentration of f-actin. In our experimental conditions, the concentration of f-actin may be sufficient to drive responses across the entire range of hydrogel stiffness, and other intracellular processes such as integrin or Rho kinase signaling may be upregulated [32]. The difference between B16F10 and BV2 cell spreading on glass could be the result of differences in the amount of clutches and motors available in these cells [33]. When the density of immobilized laminin was equal, the B16F10 cell area was lower on glass compared to 40 kPa hydrogels. By contrast, the BV2 cells were more elongated on glass coated with 0.2 μg/mL laminin compared to the 40 kPa hydrogels. According to the clutch model, the amount of motors and clutches in B16F10 cells may not be optimal

for the supraphysiologic stiffness of glass; whereas, the BV2 cells have more motors and clutches, and thus are able to spread and elongate on rigid glass. In our myosin and vinculin staining experiments we observed organization of myosin II along actin fibers in B16F10 cells on both hydrogels and glass. In the BV2 cells myosin organization was observed only on glass coated with 5.0 μg/mL laminin. Large focal adhesions were observed in both cell types, but only on glass coated with 5.0 μg/mL laminin. The difference in spreading and shape between B16F10 and BV2 could be the result of available focal adhesion machinery and myosin II motors [33], and the organization of myosin II along actin fibers is a consequence of an optimal substrate stiffness for each cell type, 40–390 kPa hydrogels for B16F10 cells vs. supraphysiologic stiffness for BV2 cells. We observed a marked decrease in laminin fluorescence signal nearby and underneath cells, but only on glass surfaces coated with 5.0 μg/mL laminin. Cell release of matrix metalloproteinases (MMPs) [34] or other proteins may only occur during supraphysiologic stiffness and high density. The MMP may cleave immobilized laminin thereby destroying the antibody epitope or other secreted proteins may bind laminin and block the epitope. Release of enzymes have indeed been shown to be regulated by the extracellular mechanical environment [35]. In our experimental system the local modification of laminin signal may negatively feedback to regulate cell-matrix interactions. Note that on the 40 kPa hydrogels and glass surfaces coated with 0.2 μg/mL laminin, no decrease in laminin immunofluorescence was detected, but an increase in perinuclear laminin was noted in the B16F10 cells. The increase in laminin is likely a consequence of cell production [36], thus, the perinuclear immunofluorescence signal was avoided in the linescan measurements. In conclusion, we describe the intrinsic response of the B16F10 melanoma cell line and microglia BV2 cell line toward physiologic substrate stiffness. When interrogating B16F10 cell morphology, shape and spreading, we noted a maximum cell spreading on hydrogels of Young’s modulus > 40 kPa. Cell spreading area was larger with more elongated cells on stiff hydrogels (40–100 kPa) than on glass of equivalent laminin concentration. A 5 to 10-fold increase in laminin concentration was required to achieve maximum cell spreading on glass. Overall, cells needed a higher density of adhesion sites on stiff glass to undergo morphological changes and extend lamellipodia, which could be explained in terms of the traction forces that the melanoma cells are able to exert on the substrate and the clustering of adhesive ligands. Similar responses were not noted for the microglia BV2 cell line, which showed an increase in cell elongation and spreading as a function of substrate stiffness, which seemed independent of ligand concentration. Trends in cell spreading area and cell elongation did


not correlate with total f-actin concentration, indicating that a different mechanism was responsible for cytoskeletal cell differences as a function of substrate stiffness and ligand concentration. Finally, both cell types modified the immobilized laminin, but only at glass and at high ligand density. Our results indicated that the observed cell responses to a physiologic stiffness were cell type-dependent. Acknowledgements This work was funded in part by a Research Grant from the Southern Illinois University Edwardsville School of Pharmacy and by start-up funds awarded to Dr. Zustiak by Saint Louis University. Michael Reimer was funded by the Research Grants for Graduate Students program at Southern Illinois University Edwardsville. Compliance with Ethical Standards Conflict of Interest ing interests.

The authors declare that they have no compet-

References 1. Engler, A. J., Sen, S., Sweeney, H. L., & Discher, D. E. (2006). Matrix elasticity directs stem cell lineage specification. Cell, 126, 677–689. 2. Tilghman, R. W., Cowan, C. R., Mih, J. D., Koryakina, Y., Gioeli, D., & Slack-Davis, J. K., et al. (2010). Matrix rigidity regulates cancer cell growth and cellular phenotype. PloS ONE, 5, e12905. 3. Engler, A. J., Griffin, M. A., Sen, S., Bönnemann, C. G., Sweeney, H. L., & Discher, D. E. (2004). Myotubes differentiate optimally on substrates with tissue-like stiffness pathological implications for soft or stiff microenvironments. The Journal of Cell Biology, 166, 877–887. 4. Mih, J. D., Sharif, A. S., Liu, F., Marinkovic, A., Symer, M. M., & Tschumperlin, D. J. (2011). A multiwell platform for studying stiffness-dependent cell biology. PloS ONE, 6, e19929. 5. Engler, A. J., Carag-Krieger, C., Johnson, C. P., Raab, M., Tang, H.-Y., & Speicher, D. W., et al. (2008). Embryonic cardiomyocytes beat best on a matrix with heart-like elasticity: scar-like rigidity inhibits beating. Journal of Cell Science, 121, 3794–3802. 6. Leach, J. B., Brown, X. Q., Jacot, J. G., DiMilla, P. A., & Wong, J. Y. (2007). Neurite outgrowth and branching of PC12 cells on very soft substrates sharply decreases below a threshold of substrate rigidity. Journal of Neural Engineering, 4, 26. 7. Lo, C.-M., Wang, H.-B., Dembo, M., & Wang, Y.-l (2000). Cell movement is guided by the rigidity of the substrate. Biophysical Journal, 79, 144–152. 8. Reilly, G. C., & Engler, A. J. (2010). Intrinsic extracellular matrix properties regulate stem cell differentiation. Journal of Biomechanics, 43, 55–62. 9. Kloxin, A. M., Benton, J. A., & Anseth, K. S. (2010). In situ elasticity modulation with dynamic substrates to direct cell phenotype. Biomaterials, 31, 1–8. 10. Paszek, M. J., Zahir, N., Johnson, K. R., Lakins, J. N., Rozenberg, G. I., & Gefen, A., et al. (2005). Tensional homeostasis and the malignant phenotype. Cancer Cell, 8, 241–254. 11. Carson, D., Hnilova, M., Yang, X., Nemeth, C. L., Tsui, J. H., & Smith, A. S., et al. (2016). Nanotopography-Induced Structural Anisotropy and Sarcomere Development in Human Cardiomyocytes Derived from Induced Pluripotent Stem Cells. ACS Applied Materials & Interfaces, 8, 21923–21932.

12. Tokuda, E. Y., Leight, J. L., & Anseth, K. S. (2014). Modulation of matrix elasticity with PEG hydrogels to study melanoma drug responsiveness. Biomaterials, 35, 4310–4318. 13. Prauzner-Bechcicki, S., Raczkowska, J., Madej, E., Pabijan, J., Lukes, J., & Sepitka, J., et al. (2015). PDMS substrate stiffness affects the morphology and growth profiles of cancerous prostate and melanoma cells. Journal of the Mechanical Behavior of Biomedical Materials, 41, 13–22. 14. Agache, P., Monneur, C., Leveque, J., & De Rigal, J. (1980). Mechanical properties and Young’s modulus of human skin in vivo. Archives of Dermatological Research, 269, 221–232. 15. Pailler-Mattei, C., Bec, S., & Zahouani, H. (2008). In vivo measurements of the elastic mechanical properties of human skin by indentation tests. Medical Engineering & Physics, 30, 599–606. 16. Budday, S., Nay, R., de Rooij, R., Steinmann, P., Wyrobek, T., & Ovaert, T. C., et al. (2015). Mechanical properties of gray and white matter brain tissue by indentation. Journal of the Mechanical Behavior of Biomedical Materials, 46, 318–330. 17. Engler, A., Bacakova, L., Newman, C., Hategan, A., Griffin, M., & Discher, D. (2004). Substrate compliance versus ligand density in cell on gel responses. Biophysical Journal, 86, 617–628. 18. Choi, J. S., & Harley, B. A. (2012). The combined influence of substrate elasticity and ligand density on the viability and biophysical properties of hematopoietic stem and progenitor cells. Biomaterials, 33, 4460–4468. 19. Jucker, M., Tian, M., & Ingram, D. K. (1996). Laminins in the adult and aged brain. Molecular and Chemical Neuropathology, 28, 209–218. 20. Ishikawa, T., Wondimu, Z., Oikawa, Y., Gentilcore, G., Kiessling, R., & Brage, S. E., et al. (2014). Laminins 411 and 421 differentially promote tumor cell migration via α6β1 integrin and MCAM (CD146). Matrix Biology, 38, 69–83. 21. Makishima, A., & Mackenzie, J. (1973). Direct calculation of Young’s moidulus of glass. Journal of Non-Crystalline Solids, 12, 35–45. 22. Bocchini, V., Mazzolla, R., Barluzzi, R., Blasi, E., Sick, P., & Kettenmann, H. (1992). An immortalized cell line expresses properties of activated microglial cells. Journal of Neuroscience Research, 31, 616–621. 23. Le Clainche, C., & Carlier, M. F. (2008). Regulation of actin assembly associated with protrusion and adhesion in cell migration. Physiological Reviews, 88, 489–513. 24. Flanagan, L. A., Ju, Y.-E., Marg, B., Osterfield, M., & Janmey, P. A. (2002). Neurite branching on deformable substrates. Neuroreport, 13, 2411. 25. Buxboim, A., Rajagopal, K., Andre’EX, B., & Discher, D. E. (2010). How deeply cells feel: methods for thin gels. Journal of Physics: Condensed Matter, 22, 194116. 26. Yeung, T., Georges, P. C., Flanagan, L. A., Marg, B., Ortiz, M., & Funaki, M., et al. (2005). Effects of substrate stiffness on cell morphology, cytoskeletal structure, and adhesion. Cell Motility and the Cytoskeleton, 60, 24–34. 27. Nakamura, K., Yoshikawa, N., Yamaguchi, Y., Kagota, S., Shinozuka, K., & Kunitomo, M. (2002). Characterization of mouse melanoma cell lines by their mortal malignancy using an experimental metastatic model. Life Sciences, 70, 791–798. 28. Chen, K., Fu, X., Dorantes-Gonzalez, D. J., Lu, Z., Li, T., & Li, Y., et al. (2014). Simulation study of melanoma detection in human skin tissues by laser-generated surface acoustic waves. Journal of Biomedical Optics, 19, 077007. 29. Blasi, E., Barluzzi, R., Bocchini, V., Mazzolla, R., & Bistoni, F. (1990). Immortalization of murine microglial cells by a v-raf/ v-myc carrying retrovirus. Journal of Neuroimmunology, 27, 229–237. 30. Chaudhuri, O., Gu, L., Klumpers, D., Darnell, M., Bencherif, S. A., & Weaver, J. C., et al. (2016). Hydrogels with tunable stress

Cell Biochem Biophys relaxation regulate stem cell fate and activity. Nature Materials, 15, 326–334. 31. Kumar, S., Maxwell, I. Z., Heisterkamp, A., Polte, T. R., Lele, T. P., & Salanga, M., et al. (2006). Viscoelastic retraction of single living stress fibers and its impact on cell shape, cytoskeletal organization, and extracellular matrix mechanics. Biophysical Journal, 90, 3762–3773. 32. Guilluy, C., Garcia-Mata, R., & Burridge, K. (2011). Rho protein crosstalk: another social network? Trends in Cell Biology, 21, 718–726. 33. Bangasser, B. L., Shamsan, G. A., Chan, C. E., Opoku, K. N., Tuzel, E., & Schlichtmann, B. W., et al. (2017). Shifting the optimal stiffness for cell migration. Nature Communications, 8, 15313.

34. Maity, G., Sen, T., & Chatterjee, A. (2011). Laminin induces matrix metalloproteinase-9 expression and activation in human cervical cancer cell line (SiHa). Journal of Cancer Research and Clinical Oncology, 137, 347–357. 35. Haage, A., & Schneider, I. C. (2014). Cellular contractility and extracellular matrix stiffness regulate matrix metalloproteinase activity in pancreatic cancer cells. FASEB Journal, 28, 3589–3599. 36. Fligiel, S. E., Laybourn, K. A., Peters, B. P., Ruddon, R. W., Hiserodt, J. C., & Varani, J. (1986). Laminin production by murine melanoma cells: possible involvement in cell motility. Clinical & Experimental Metastasis, 4, 259–272.