Mechanosensitive expression of lamellipodin promotes intracellular stiffness, cyclin expression and cell proliferation

The extracellular matrix (ECM) provides structural and mechanical support for cells in living tissue. Cells exhibit pathological behaviors in response to increased ECM stiffness, including accelerated cell proliferation and migration (Klein et al., 2009; Bae et al., 2014; Mui et al., 2015; Liu et al., 2015; Peyton and Putnam, 2005; Liu et al., 2010; Razinia et al., 2017; Rickel et al., 2020), which are correlated with increased intracellular stiffness and changes in actin cytoskeleton dynamics (Bae et al., 2014; Mui et al., 2015; Yeung et al., 2005; Solon et al., 2007; Califano and Reinhart-King, 2010). The mechanotransduction of ECM stiffness into molecular signals and resultant cellular processes is mediated through integrins, other focal adhesion proteins and small Rho GTPases, together causing stiffening of the actin cytoskeleton (Bae et al., 2014; Mui et al., 2015; Yeung et al., 2005; Solon et al., 2007; Oakes et al., 2018). However, the exact molecular mechanisms regulating cellular responses to ECM stiffening and their link to cell cycle control and proliferation remain unclear. Lamellipodin (symbol RAPH1), Rac GTPase and the Scar/WAVE complex control lamellipodial actin dynamics and cell migration (Krause et al., 2004; Quinn et al., 2008; Law et al., 2013; Hansen and Mullins, 2015; Bae et al., 2010; Carmona et al., 2016; Dimchev et al., 2020). The Scar/WAVE complex is a key activator of the actin filament nucleating Arp2/3 complex (Krause and Gautreau, 2014). Our previous work described that lamellipodin promotes stiffness-mediated induction of cyclin D1 expressed in G1 phase during cell cycle progression and proliferation (Bae et al., 2014). In addition, our atomic force microscopy (AFM) analysis showed that lamellipodin depletion reduces intracellular stiffness.

Here, we show that the levels of lamellipodin expression in mouse embryonic fibroblasts (MEFs) are significantly increased by a stiff ECM, and that this stiffness-mediated lamellipodin upregulation stimulates cyclin expression and cell proliferation as well as intracellular stiffening throughout the cell cycle, from the early G1 phase to M phase. Finally, we provide evidence that Rac activation, Arp2/3 complex activity, an intact actin cytoskeleton, and intracellular stiffening support the mechanosensitive induction of lamellipodin. We thus propose that an integrin-associated Rac–lamellipodin signaling module supports mechanosensitive cell cycle progression and intracellular stiffening.

ECM stiffness controls the expression and activation of critical effector proteins vital to ECM stiffness-dependent cellular functions, including cell cycling and proliferation (Klein et al., 2009; Bae et al., 2014; Mui et al., 2015; Liu et al., 2015). We previously showed that lamellipodin transduces ECM stiffness into early G1-phase cell cycling by upregulating cyclin D1 in MEFs (Bae et al., 2014). To now determine whether the levels of lamellipodin expression are stiffness-dependent, we used fibronectin-coated polyacrylamide hydrogels (Fig. 1A) to create culture matrices that mimic the biologically relevant stiffness of normal and pathological microenvironments in vivo. Hydrogels of low stiffness (2–4 kPa; denoted ‘soft’) approximate normal stiffness, and hydrogels of intermediate stiffness (10–12 kPa; denoted ‘medium’) and high stiffness (20–24 kPa; denoted ‘stiff’) approximate pathological stiffness (Klein et al., 2009; Bae et al., 2014; Mui et al., 2015; Liu et al., 2015, 2010; Razinia et al., 2017; Rickel et al., 2020; Yeung et al., 2005; Solon et al., 2007). Serum-starved MEFs, which are synchronized in G0 of the cell cycle, were cultured on fibronectin-coated hydrogels (Fig. 1A) with 10% fetal bovine serum (FBS) for 9 h and 24 h. We found that lamellipodin mRNA expression was moderately increased at both 9 h and 24 h on stiff hydrogels (Fig. 1B). We also confirmed that cyclin A (an S/G2-phase marker; Fig. 1C) and cyclin D1 (a G1-phase marker; Fig. 1D) mRNA expression are stiffness-sensitive, as previously shown in other studies (Klein et al., 2009; Bae et al., 2014). We observed lamellipodin protein levels in MEFs to be progressively upregulated on stiff hydrogels at 9 h and 24 h (Fig. 1E–G), and with a greater relative increase at 24 h. We then confirmed the temporal relationship of lamellipodin induction on high stiffness and found modestly increased expression at 3 h and 9 h and an ∼4-fold increased expression at 24 h (Fig. 1H–I). Importantly, we found that cyclin A is equally strongly upregulated on stiff ECM at 24 h (arrow in Fig. 1H). Taken together, our results indicate that both cyclins A and D1 and lamellipodin are upregulated by stiff ECM suggesting that the stiffness-sensitive expression of lamellipodin may be associated with cell cycle progression.

Rap1-GTP-interacting adaptor molecule (RIAM; also known as APBB1IP) is a lamellipodin paralog, which together constitute the mammalian MIG-10/RIAM/lamellipodin protein family (Lafuente et al., 2004; Krause et al., 2004). Both RIAM and lamellipodin share significant overlap in their domain structure and molecular mechanisms. We, therefore, examined the effect of ECM stiffness on RIAM expression. When we cultured serum-starved MEFs on hydrogels for 24 h with FBS, we found that RIAM mRNA (Fig. S1A) and protein levels (Fig. S1B,C) were upregulated in response to high ECM stiffness. This suggests that both RIAM and lamellipodin may have complementary functions in ECM stiffness sensing.

Lyulcheva et al. have previously shown that knockdown of pico (the sole Drosophila ortholog of lamellipodin and RIAM) in Drosophila wing disk tissue reduces cell proliferation. However, the germline mutation of pico in Drosophila is larval lethal (Lyulcheva et al., 2008), while embryonic knockout of lamellipodin in mice is perinatally lethal (Law et al., 2013). Furthermore, lamellipodin knockdown in HeLa cells cultured on tissue culture plates reduces EGF-induced cell proliferation (Lyulcheva et al., 2008). We have previously shown that lamellipodin knockdown prevents an increase of cyclin D1 levels in MEFs cultured on stiff hydrogels, suggesting that lamellipodin may be required for cell cycling (Bae et al., 2014). However, the degree to which lamellipodin mediates the mechanosensitive cell cycle progression (or specific cell cycle phases) that contributes to cell proliferation remains unknown. Thus, we determined the ability of lamellipodin to affect specific cell cycle stages. MEFs were transfected with siRNA to transiently knockdown lamellipodin. Lamellipodin knockdown significantly reduced cyclin A (S/G2 phase), cyclin B (G2/M phase) and DNA synthesis (Fig. 2A–C; Fig. S2) when MEFs were cultured on stiff hydrogels for 24 h. We confirmed that cyclin D1 (G1 phase) induction at 9 h was decreased by lamellipodin knockdown (Fig. S3A,B), as previously shown (Bae et al., 2014). Taken together, these results suggest that lamellipodin promotes cell proliferation and that cyclin expression levels may depend on lamellipodin.

To further explore the potential dependence of cyclin expression on lamellipodin, we serum-starved control and lamellipodin overexpressing (Lpd^o/x) MEFs and seeded them on soft hydrogels with FBS for 24 h. We observed a 9-fold upregulation of cyclins A and B (Fig. 2D–F) and a more than 2-fold increase in DNA synthesis in Lpd^o/x MEFs when compared to control MEFs (Fig. 2G; Fig. S2B). Additionally, we asked whether the observed lamellipodin-mediated effects on cell cycling and proliferation were bona fide changes and not the result of indirect effects on cell survival. Lamellipodin knockdown in MEFs did not affect the levels of PTEN (Fig. S3C–E), a potent inducer of apoptosis and cell cycle arrest (Lu et al., 2016). Consistent with this observation, intrinsic survival defects have also not been observed as a consequence of reduced pico levels in Drosophila and lamellipodin knockdown in HeLa (Lyulcheva et al., 2008). Collectively, our results suggest that lamellipodin promotes mechanosensitive cell cycle progression and cell proliferation in a manner that is independent of cell survival.

Cells modulate their intracellular stiffness in response to the stiffness of the surrounding ECM (Yeung et al., 2005; Solon et al., 2007; Fabry et al., 2011), and this intracellular stiffness is critical for cell cycling and thus for proliferation (Klein et al., 2009; Bae et al., 2014; Mui et al., 2015). More specifically, intracellular stiffening is the dynamic result of spatiotemporally organized actin polymerization, stress fiber formation, and is greatest at the cell periphery (Bae et al., 2014; Hsu et al., 2015) where lamellipodin contributes to F-actin-dependent leading-edge formation (Krause et al., 2004; Quinn et al., 2008; Law et al., 2013; Carmona et al., 2016; Dimchev et al., 2020). We confirmed that F-actin in MEFs on stiff hydrogels is mostly located at the cell periphery (Fig. 3A), and these MEFs also exhibit a large increase in cell area compared to MEFs on soft hydrogels (Fig. 3B). When knocking down lamellipodin expression, we observed that cell spreading (Fig. 3C,D) was reduced but focal adhesion density and size were unchanged (Fig. 3E).

AFM was used to measure the stiffness of MEFs on hydrogels. Serum-starved MEFs were seeded on hydrogels with FBS for 3, 9 and 24 h. As previously reported (Bae et al., 2014), stiff ECM increased the intracellular stiffness of MEFs (Fig. 3F, graph columns 2, 5 and 8 from left). We tested whether this increase in intracellular stiffness on stiff hydrogels could be the result of upregulated lamellipodin expression. We found that lamellipodin knockdown significantly reduced intracellular stiffness after being stimulated with FBS (Fig. 3F, graph columns 3, 6 and 9 from left). Conversely, we observed that cells transfected with GFP-tagged lamellipodin (GFP–Lpd) moderately increased intracellular stiffness compared to cells transfected with GFP on soft hydrogels (Fig. 3G). Taken together, these results imply that lamellipodin contributes to mechanosensitive intracellular stiffening potentially by promoting F-actin-dependent cell spreading.

A FAK–Rac (FAK is also known as PTK2) signaling pathway is critical for the transmission of ECM stiffness into mechanochemical cues and the resultant changes to cell cycling and intracellular stiffness (Klein et al., 2009; Bae et al., 2014; Mui et al., 2015). Upon activation of this pathway by ECM stiffness, FAK activates Rac, which stimulates an increase in intracellular stiffness (Bae et al., 2014) and directly binds to lamellipodin (Law et al., 2013). We previously found that Rac overexpression induces the translocation of lamellipodin to the cell periphery of MEFs plated on stiff hydrogels (Bae et al., 2014) and that lamellipodin expression is sensitive to ECM stiffness (Fig. 1B–I). To further dissect the regulatory pathway by which ECM stiffness is translated into lamellipodin induction through stiffness-sensitive Rac activation, we first confirmed that Rac is persistently activated in MEFs in response to increased ECM stiffness at 24 h (Fig. 4A), a time point at which we also observed significantly increased levels of lamellipodin, cyclin D1 and cyclin A expression (Fig. 1B–I). Since cell cycle progression in MEFs occurs over 24 h, we expected persistent Rac activation (Fig. 4A) and intracellular stiffening (Fig. 3G) to be necessary for lamellipodin-mediated cell proliferation. We then tested whether Rac is a potent upstream regulator of stiffness-dependent lamellipodin expression. MEFs treated with Rac1 siRNAs or adenovirus encoding dominant-negative Rac^N17 were plated on stiff hydrogels and stimulated with FBS for 24 h. Immunoblot analysis showed that both siRNA- or adenovirus-mediated Rac inhibition significantly decreased lamellipodin induction (Fig. 4B–E). In contrast, the overexpression of constitutively active Rac^V12 dramatically rescued lamellipodin induction in MEF cultures on soft hydrogels for 24 h (Fig. 4F,G, columns 1 and 2).

Intracellular forces are generated by Rac-induced actin polymerization (Bae et al., 2014; Mui et al., 2015). Thus, we tested whether both Rac activation and intracellular stiffening are required for the increase in lamellipodin induction. MEFs infected with Rac^V12 adenovirus were seeded on soft hydrogels for 24 h with either DMSO (vehicle control) or latrunculin B (an actin polymerization inhibitor), which was previously shown to reduce intracellular stiffness (Bae et al., 2014). Interestingly, we found that the overexpression of Rac in MEFs on soft hydrogels was not sufficient to completely rescue Rac-mediated lamellipodin expression upon the decrease in intracellular stiffness (Bae et al., 2014) induced by latrunculin B treatment (Fig. 4F,G, column 4), compared to the levels of lamellipodin expression in the Rac^V12-infected cells treated with DMSO (Fig. 4F,G, column 2). These findings suggest that although Rac activity is required for the induction of lamellipodin by ECM stiffness, the integrity of actin-dependent intracellular stiffening is necessary.

Cortical branched actin networks are unique to the molecular architecture of lamellipodia (Molinie et al., 2019; Krause and Gautreau, 2014). Their formation and stabilization allow for the generation of forward pushing forces as a result of dynamic actin polymerization. The nucleation and polymerization of actin branches require the recruitment and activation of both Scar/WAVE and Arp2/3 complexes to the cell leading edge as a result of Rac1 and lamellipodin collaboration (Law et al., 2013). Recently, cortical branched actin was implicated in regulating cell cycle progression via the Arp2/3 complex. Interestingly, inhibition of Arp2/3 using CK-666 (an Arp2/3 complex inhibitor) blocks G1/S phase transition (Molinie et al., 2019). Mirroring our Rac^V12 and latrunculin B experiment (Fig. 4F,G), we treated MEFs with CK-666 instead of latrunculin B. We observed CK-666 to partially block the ability of Rac to rescue stiffness-sensitive lamellipodin induction (Fig. 4H,I).

We have previously shown that FAK and p130Cas (Cas; also known as BCAR1) are upstream regulators of stiffness-dependent Rac1-mediated signaling (Bae et al., 2014). We, therefore, asked whether FAK and Cas specifically modulate lamellipodin expression. MEFs were transfected with either FAK or Cas siRNAs, serum-starved, and subsequently seeded on stiff hydrogels with FBS for 24 h. We indeed found lamellipodin levels to be greatly reduced upon FAK (Fig. 4J–L) and Cas (Fig. 4M–O) knockdown, as was also the case when cells were treated with PF562271 (a FAK/Pyk2 inhibitor) and PF573228 (a FAK inhibitor) further supporting our findings (Fig. S4A,B).

Integrin activation is upstream of stiffness-mediated FAK and Cas phosphorylation and Rac activation (Bae et al., 2014; Klein et al., 2009). However, it remains unclear whether stiffness-sensitive lamellipodin induction is mediated through integrin-dependent signaling and adhesion. To determine whether lamellipodin induction is integrin dependent, we seeded serum-starved MEFs on poly-L-lysine (PLL)-coated glass coverslips with FBS. Previous experiments have shown that cells adhere to, but do not spread on, PLL in an integrin-independent manner (Price et al., 1998). Global activation of integrins using Mn²⁺ on PLL enhances cell spreading and FAK phosphorylation (Swaminathan et al., 2017; Lin et al., 2013). Upon seeding MEFs on PLL-treated coverslips, we immediately treated MEFs for 30 min with either 1 mM Mn²⁺, to activate inside-out integrin signaling, or with 3 mM sucrose, an inert osmolarity control. Strikingly, lamellipodin (Fig. S4C,D) and FAK phosphorylation (Fig. S4E) were increased upon integrin activation with Mn²⁺, and cell spreading also paralleled lamellipodin induction (Fig. S4F,G).

We conclude that integrin-associated FAK–Cas–Rac activation, the actin cytoskeleton and intracellular stiffening are necessary for the mechanosensitive induction of lamellipodin (Fig. 4P).

Our previous work described that lamellipodin promotes stiffness-mediated induction of cyclin D1 expressed in early G1 phase during cell cycling, and our AFM analysis showed that lamellipodin depletion reduces intracellular stiffness (Bae et al., 2014). Here, we found that lamellipodin expression levels themselves depend on the stiffness of the ECM environment, and that this lamellipodin induction coincides with the induction of cyclins. In turn, cyclin expression and DNA synthesis are promoted by lamellipodin expression. Mechanistically, increased cell spreading and intracellular stiffening on stiff ECM is promoted by lamellipodin. Taken together, we uncover a signaling pathway of activation of integrins, FAK, Cas, Rac and the Arp2/3 complex that may mediate the intracellular stiffening and cell proliferation induced by induction of lamellipodin expression upon an increase in ECM stiffness.

Spontaneously immortalized mouse embryonic fibroblasts (MEFs; a gift from the Assoian Laboratory, University of Pennsylvania, USA) and MEFs stably overexpressing lamellipodin (Lpd^o/x MEFs; a gift from the Gertler Laboratory, Massachusetts Institute of Technology, USA) were grown in low glucose Dulbecco's modified Eagle's medium (DMEM; Corning, 10-014-CV) supplemented with 50 μg/ml gentamicin solution (Corning, 30-005-CR) and 10% (v/v) fetal bovine serum (FBS; Sigma-Aldrich, F2442). Prior to seeding cells on fibronectin (Calbiochem, 341631)-coated polyacrylamide hydrogels for experimentation, MEFs were synchronized in the G0 cell cycle phase (quiescence), in which near-confluent cells (∼ 80%) were serum-starved by incubation for 24 h in DMEM with 1 mg/ml heat-inactivated, fatty-acid-free bovine serum albumin (BSA; Tocris, 5217). Thereafter, serum-starved MEFs were seeded on fibronectin-coated polyacrylamide hydrogels with 10% FBS and incubated for 3, 9 and 24 h.

Near-confluent MEFs were transfected with 150–200 nM lamellipodin, Rac1, FAK, or p130Cas (Cas) siRNAs using Lipofectamine 2000 reagent (Invitrogen, 11668019) in reduced serum Opti-MEM (Gibco, 31985070) as previously described (Klein et al., 2007, 2009; Bae et al., 2014). MEFs were transfected with siRNA for 4–5 h, immediately followed by serum starvation in serum-free DMEM with 1 mg/ml BSA for ∼20 h. Cells were then trypsinized and reseeded on fibronectin-coated hydrogels with fresh DMEM containing 10% FBS, as described above. A non-targeting scrambled siRNA (Silencer negative control siRNA, Ambion, AM4611) was used as an experimental control. Lamellipodin (RAPH1), Rac1, FAK (PTK2), and Cas (BCAR1) siRNAs were obtained from Ambion as follows: Lamellipodin (ID 509012) siRNA #1, 5′-GGAUUUUCUCUGAAGUAAAtt-3′; Lamellipodin (ID 509013) siRNA #2, 5′-GCUGGAUCAUGUAAAUGUAtt-3′; Rac1 siRNA #1 (ID 214457), 5′-CCGUCUUUGACAACUAUUCtt-3′; Rac1 siRNA #2 (ID 214461), 5′-CCAGUGAAUCUGGGCCUAUtt-3′; FAK siRNA #1 (ID 157448), 5′-CCUAGCAGACUUUAACCAAtt-3′; FAK siRNA #2 (ID 61352), 5′-GGCAUGGAGAUGCUACUGAtt-3′; Cas siRNA #1 (ID 161328), 5′-GCCAAUCGGCAUCUUCCUUtt-3′; Cas siRNA #2 (ID 161329), 5′-GCUGAAACAGUUUGAGCGAtt-3′.

For adenoviral infection, MEFs were incubated in serum-free DMEM with 1 mg/ml BSA for ∼9 h. Thereafter, adenoviruses were added to their respective culture medium and incubated for an additional 15–24 h. Adenoviruses encoding Rac^V12 and Rac^N17 (gifts from the Assoian Laboratory) were added at a multiplicity of infection of 900 and 100, respectively. Additionally, adenoviruses encoding LacZ (a gift from the Assoian Laboratory) and GFP (Vector Biolabs, 1060) were used as experimental controls. Adenovirus-infected MEFs were seeded on fibronectin-coated hydrogels with DMEM containing 10% FBS.

In some experiments (Fig. 4F–I), adenovirus-infected cells were cultured on fibronectin-coated hydrogels for 24 h with DMSO (vehicle, Sigma-Aldrich), 0.5 μM latrunculin B (Actin polymerization inhibitor; Calbiochem, 428020), or 100 μM CK666 (Arp2/3 inhibitor; Cayman Chemical, 29038). Additionally, in other experiments (Fig. S4A,B), serum-starved MEFs were seeded on fibronectin-coated stiff hydrogels for 24 h with 10 μM PF562271 (a FAK/Pyk2 inhibitor; Sigma-Aldrich, PZ0387-5MG) and 10 μM PF573228 (a FAK inhibitor; Sigma-Aldrich, PZ0117).

The general protocol for fabricating stiffness-tunable polyacrylamide hydrogels has been previously described (Klein et al., 2007; Cretu et al., 2010). Briefly, glass coverslip surfaces were etched uniformly with 0.1 M sodium hydroxide solution (Thermo Fisher Scientific, SS266-1), followed by the addition of (3-aminopropyl)trimethoxysilane (APTMS; Sigma-Aldrich, 281778) to introduce amine (-NH₂) groups onto the glass surface. Reactive coverslips were then incubated with 0.5% glutaraldehyde solution (Sigma-Aldrich, G5882) to cross-link the APTMS and the polyacrylamide hydrogels. Hydrogels of varying stiffness were prepared using various ratios of 40% acrylamide (Bio-Rad, 1616140) to 1% bis-acrylamide (Bio-Rad, 1610142) in a mixed solution with water, ammonium persulfate (Sigma-Aldrich, A3678), TEMED (Bio-Rad, 1610800) and acrylic acid N-hydroxysuccinimide ester (NHS; Sigma-Aldrich, A8060) in toluene (Sigma-Aldrich, 244511). Notably, the polyacrylamide hydrogels were functionalized using NHS solution to induce the conjugation of fibronectin (Calbiochem, 341631) to the polyacrylamide hydrogels. The elastic moduli of the hydrogels with various stiffnesses were used for the experiment: low (2 to 4 kPa), medium (10 to 12 kPa), and high (20 to 24 kPa). The hydrogels were then coated with 3 μg/ml fibronectin and incubated overnight at 4°C. The fibronectin-coated hydrogels were extensively washed with 1× DPBS (Corning, MT21031CV) to remove any remaining monoacrylamide. Unreactive cross-linkers were blocked with 1 mg/ml heat-inactivated BSA prior to cell seeding for 1 h at 37°C.

MEFs cultured on polyacrylamide hydrogels were first washed twice with ice-cold DPBS and then collected using TRIzol reagent (Invitrogen, 15596018). Total RNA was extracted according to the manufacturer's instructions as described (Klein et al., 2007). Briefly, TRIzol extracts were incubated with chloroform (Thermo Fisher Scientific) for 3 min at room temperature (RT) and then centrifuged at 17,136 g for 15 min at 4°C. Thereafter, the upper aqueous phase containing RNA was transferred to a new Eppendorf tube, and the RNA fraction was incubated with 100% isopropanol (Thermo Fisher Scientific) for 10 min at RT. Samples were then centrifuged at 17,136 g for 15 min at 4°C. The supernatant was discarded, and the RNA pellet was resuspended in 75% ethanol (Thermo Fisher Scientific). The samples were centrifuged at 7616 g for 5 min at 4°C. RNA was eluted with RNase-free water and then incubated at 55 to 60°C for 10 min. Total RNA was reverse transcribed and analyzed by real-time quantitative PCR (RT-qPCR) as previously described (Klein et al., 2007). TaqMan probes were used for lamellipodin (Invitrogen, Mm01254983_m1), cyclin A (Invitrogen, Mm00432337_m1), cyclin D1 (Invitrogen, Mm00432539_m1), RIAM (Invitrogen, Mm01329408_m1) and GAPDH (Invitrogen, Mm99999915_g1). The relative changes in mRNA expression for each target gene were determined by the ΔΔCt method using GAPDH as the reference gene.

Protein extraction was performed as follows. Adherent MEFs on fibronectin-coated polyacrylamide hydrogels were first washed twice with ice-cold DPBS, and total cell lysates were then extracted by placing the hydrogel face-down on a predetermined volume of 5× sample buffer (250 mM Tris-HCl pH 6.8, 10% SDS, 50% glycerol, 0.02% bromophenol blue, and 10 mM 2-mercaptoethanol) and incubated for 2 min at RT, as previously described (Klein et al., 2007; Bae et al., 2014). Equal amounts of extracted protein were fractionated on 6 to 10% SDS-PAGE gels, and the fractionated proteins were transferred electrophoretically on PVDF membranes. These immunoblot membranes were probed with antibodies against lamellipodin (Santa Cruz Biotechnology, rabbit, sc-68380, 1:500), lamellipodin (Novus Biologicals, rabbit, NBP1-90859, 1:500), cyclin D1 (Santa Cruz Biotechnology, mouse, sc-20044, 1:200), cyclin A [a gift from the Assoian Laboratory (Kothapalli et al., 2003), rabbit, 1:500], cyclin B1 (BioLegend, mouse, 647902, 1:500), Rac1 (Santa Cruz Biotechnology, rabbit, sc-95, 1:200), vinculin (Sigma-Aldrich, mouse, V9131, 1:10000), RIAM (Novus Biologicals, rabbit, NBP1-40722. 1:1000), FAK (Invitrogen, mouse, 39-6500, 1:5000), pFAK (Invitrogen, rabbit, 700255, 1:500), p130Cas (Bio-Rad, mouse, VMA00132, 1:1000), PTEN (Santa Cruz Biotechnology, mouse, A2B1, 1:500), GAPDH (Proteintech, mouse, 60004-1-Ig, 1:10,000), and GAPDH (Proteintech, rabbit, 10494-1-AP, 1:10,000). Antibody signals were detected using Clarity Western ECL Substrate (Bio-Rad, 1705061) or Clarity Max Western ECL Substrate (Bio-rad, 1705062).

For immunostaining, MEFs cultured on polyacrylamide hydrogels were fixed in 3.7% formaldehyde (Thermo Fisher Scientific) in DPBS for 1 h, permeabilized with 0.4% Triton X-100 (Sigma-Aldrich, T8787) for 30 min, and then blocked with 2% BSA and 0.2% Triton X-100 in DPBS for 1 h at RT. After being incubated with anti-vinculin antibody (Sigma-Aldrich, mouse, V9131, 1:50) for 2 h at RT, cells were washed with DPBS containing 2% BSA and 0.2% Triton X-100. Cells were then incubated with Alexa Fluor 594-conjugated goat anti-mouse-IgG antibody (Invitrogen, A11032, 1:100) and Alexa Fluor 488–phalloidin (Invitrogen, A12379, 1:100) for 1 h at RT. Subsequently, cells were then washed with DPBS containing 2% BSA and 0.2% Triton X-100 and then with distilled water before finally mounting with DAPI-containing mounting medium (Electron Microscopy Sciences, 17985-50) for fluorescence microscopy. Fluorescence images of cells were acquired with the 10× or 40× objective of a Leica DM 6B upright fluorescence microscope.

Serum-starved MEFs transfected with control and lamellipodin siRNAs were plated on stiff fibronectin-coated hydrogels with 10% FBS. Independently, serum-starved control and lamellipodin overexpressing MEFs were plated on soft hydrogels with 10% FBS. To examine DNA synthesis, MEFs were incubated with 20 μM 5-ethynyl-2′-deoxyuridine (EdU) for 22 h. EdU incorporation was visualized using Click-iT EdU Alexa Fluor 594 Imaging Kit (Invitrogen, C10339) as per the manufacturer's instructions. Thereafter, coverslips containing cell-attached hydrogels were mounted with DAPI-containing mounting medium on the microscope slides for visualization. For each sample, three to eight fields of view were counted to determine the percentage of EdU-positive cells relative to DAPI-stained nuclei using a ZOE Fluorescent Cell Imager (Bio-Rad).

Glass coverslips were incubated with 0.1 mg/ml of PLL solution (Invitrogen, p4707) for 1 h at room temperature. Coverslips were then washed three times with ddH₂O for 10 min, and allowed to air dry. PLL-coated glass coverslips were then incubated with 1 mg/ml heat-inactivated BSA for 1 h at 37°C to block reactive cross-linkers prior to cell seeding.

AFM was used to measure intracellular stiffness as previously described (Klein et al., 2009; Bae et al., 2014, 2016). Briefly, to measure intracellular stiffness, cells cultured on soft and stiff polyacrylamide hydrogels in Phenol Red-free DMEM with 10% FBS were indented with a silicon nitride cantilever (Bruker, DNP-10; spring constant, 0.06 N/m) with a conical AFM tip (40 nm in diameter) (Fig. 3G) or a silicon nitride cantilever (Asylum, BL-AC40TS-C2; spring constant, 0.09 N/m) with a three-sided pyramidal AFM tip (8 nm in radius) (Fig. 3H). AFM in contact mode was applied to single adherent MEFs using a BioScope Catalyst AFM system (Bruker) mounted on a Nikon Eclipse TE 200 inverted microscope (Fig. 3G) or a Park NX12 AFM system (Park Systems) mounted on a Nikon ECLIPSE Ti2 inverted microscope (Fig. 3H). To analyze the stiffness, the first 300–600 nm of horizontal tip deflection was fit with the Hertz model for a cone (Fig. 3G) or a four-sided pyramid (Fig. 3H). A total of 4 to 10 measurements of intracellular stiffness from each experiment were acquired near the periphery of each cell (6 to 11 cells per experimental condition). AFM curves were quantified and converted to Young's modulus (stiffness) using AFM analysis software, including NanoScope Analysis (Bruker) and XEI (Park Systems).

Data are presented as the mean±s.e.m. of the indicated number of independent experiments. The data were analyzed using paired, two-tailed distribution Student's t-test. Results with P-values lower than 0.05 (*), 0.01 (**), or 0.001 (***) were considered to be statistically significant.

We acknowledge Amanda Krajnik for her technical assistance with this project.