# Fileset

[Adv Funct Materials - 2025 - Zhou - Highly Def.pdf](https://mdr.nims.go.jp/filesets/c56dcdb9-721e-4539-b5eb-9c26bf1d8fbe/download)

## Creator

[Junhong Zhou](https://orcid.org/0000-0001-6327-6512), [Hongxin Wang](https://orcid.org/0000-0002-8984-0764), [Jun Nakanishi](https://orcid.org/0000-0003-4457-6581)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Highly Deformable Phototunable Viscoelastic Fluid Interface Modulates Cellular Adaptive Wetting Behavior](https://mdr.nims.go.jp/datasets/c7c213d7-be40-46ed-8b50-6ad79de68d7c)

## Fulltext

Highly Deformable Phototunable Viscoelastic Fluid Interface Modulates Cellular Adaptive Wetting BehaviorRESEARCH ARTICLEwww.afm-journal.deHighly Deformable Phototunable Viscoelastic FluidInterface Modulates Cellular Adaptive Wetting BehaviorJunhong Zhou, Hongxin Wang, and Jun Nakanishi*Emerging evidence shows that the viscoelastic cues of the e Ixtracellularmatrix (ECM) regulate cellular functions and fates. However, as cells areviscoelastic, force dissipation occurs within themselves as well as the ECMside, implying the existence of reciprocal viscous regulation between the two.Here, a fluid-based scaffold with tunable viscoelasticity has been developedto investigate its impact on the cell adhesion process. The platform is basedon the water–perfluorocarbon interface decorated with diacetylene-basedphospholipid membranes (IPLMs), whose viscoelasticity can be systematicallymanipulated by photocrosslinking. Further introduction of a cell-adhesivepeptide and fluorescent tag allows cell adhesion at the highly deformable fluidinterface and confocal observation of dynamic cell–model ECM interactions.The viscoelasticity-tunability is confirmed by fluorescence recovery after photo-bleaching, interfacial rheology, and atomic force microscopy nanoindentation.Cells seeded at the IPLM exhibit so-called adaptive wetting, where the interfacefirst deforms toward the out-of-plane direction before cellular dimensionalchanges, followed by cellular flattening and interfacial restoration. Furthermore,the quantification of these parameters reveals a biphasic response against thecrosslinking levels, which indicates that the cell-ECM viscosity balance deter-mines adaptive wetting phenotypes. The platform may enable the predictionof dynamic adhesion responses in physiological and pathological processes.1. IntroductionIn natural tissues, cells reside in a highly dynamic extracellu-lar matrix (ECM), whose composition and mechanics changeduring development and disease progression. Biomaterials withmechanical and biochemical cues can be customized to modu-late cellular behaviors.[1] Synthetic elastic substrates with varyingstiffness induce various mechanobiological processes, includingcellular adhesion behavior, cytoskeletal contractility, and stemJ. Zhou, H. Wang, J. NakanishiResearch Center for Macromolecules and BiomaterialsNational Institute for Materials Science (NIMS)1-1 Namiki, Tsukuba, Ibaraki 305-0044, JapanE-mail: NAKANISHI.Jun@nims.go.jpThe ORCID identification number(s) for the author(s) of this articlecan be found under https://doi.org/10.1002/adfm.202414534© 2025 The Author(s). Advanced Functional Materials published byWiley-VCH GmbH. This is an open access article under the terms of theCreative Commons Attribution License, which permits use, distributionand reproduction in any medium, provided the original work is properlycited.DOI: 10.1002/adfm.202414534cell fate decisions.[2] Since many softtissues are not purely elastic, ratherthan that they exhibit a viscoelastic na-ture with time-dependent mechanicalresponses.[3] The combination of theelastic and viscous properties is impor-tant for regulating cellular activities.Recently, the role of substrate stress re-laxation (viscous behavior) in cell mechan-otransduction was revealed through the ma-nipulation of the bulk viscoelastic prop-erties of hydrogels.[4] It is reported thatviscoelastic hydrogels with the appropri-ate stress relaxation promote the spread-ing and differentiation of mesenchymalstem cells (MSCs).[5] The tuning of vis-cous components in biomaterials can alsobe achieved by engineering molecular ad-sorption, ligand tethering, or ligand dif-fusivity of a solid-supported substrate.[6]To highlight the viscous component, sup-ported lipid bilayers (SLBs) have been uti-lized to control RGD (Arg–Gly–Asp) lig-and mobility depending on the lipid com-positions, which allows us to understandthe impact of interfacial viscosity on celladhesion behaviors.[7] Dissipation occursin either the materials’ bulk or interface, and the viscous na-ture regulates the cellular spreading speed in the framework ofthe molecular clutch scenario in a similar fashion to the elas-tic component.[8] However, the forces generated by the actin ret-rograde flow are applied not only to the extracellular matricesbut also to the cellular interior.[9] By considering that cells alsohave viscoelastic responses,[10] force dissipation can be expectedwithin the cells, altering cellular adhesion dynamics, includingcellular morphological changes and dimensional transitions.J. Zhou, J. NakanishiGraduate School of Advanced Science and EngineeringWaseda University3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, JapanJ. NakanishiGraduate School of Advanced EngineeringTokyo University of Science6-3-1 Niijuku, Katsushika-ku, Tokyo 125-8585, JapanAdv. Funct. Mater. 2025, 2414534 2414534 (1 of 12) © 2025 The Author(s). Advanced Functional Materials published by Wiley-VCH GmbHhttp://www.afm-journal.demailto:NAKANISHI.Jun@nims.go.jphttps://doi.org/10.1002/adfm.202414534http://creativecommons.org/licenses/by/4.0/http://crossmark.crossref.org/dialog/?doi=10.1002%2Fadfm.202414534&domain=pdf&date_stamp=2025-01-16www.advancedsciencenews.com www.afm-journal.deFigure 1. Emergence of cellular adaptive wetting behaviors at highly deformable fluid interfacial phospholipid membranes (IPLMs) depending on photo-tuned interfacial viscoelasticity. a) Chemical structure of the monomeric photocrosslinkable phospholipid (DA–PC). b) UV-induced polymerization re-action of DA–PC molecules. c) The crosslinked DA–PC acting as physical barriers within IPLM to alter the viscoelastic nature of the water–PFCL (FC-70)interface depending on UV dose. The head groups of monomeric and crosslinked DA–PC are shown in gray and pink, respectively. The lipid membranesalso contain DSPC (cyan), cell-adhesive RGD-DSPE (blue), and fluorescent Liss-Rhod DOPE (red). The color bar increments shown below represent theviscoelasticity changes. d) Confocal microscopic observation and image reconstitution strategy of the side view from the stacked images of cells (green)and IPLM (red). e) Schematic drawing of cellular adaptive wetting, where the interface deformed first without cellular flattening, followed by cellularspreading and interfacial restoration. Green, actin. Red, IPLM with RGD ligand. Blue, nucleus. Yellow, integrin. Dark red, focal adhesion.Our group has recently reported a new platform that canhighlight the impact of cellular viscoelastic nature on spreadingbehaviors.[11] The platform is based on water–perfluorocarbonliquid (PFCL) interfaces decorated with phospholipid mem-branes (IPLMs), containing a cell-adhesive peptide ligand andfluorescent tag. The bioinert nature of the phospholipid layerprevents protein adsorption and robust protein nanofilm forma-tion at the interface, maintaining the intrinsically super-soft na-ture of the fluid interface. On this extremely high deformableIPLM toward the out-of-plane direction, the cells exhibit a uniqueadaptive wetting behavior, in which the interface deforms be-fore cellular dimension changes, followed by interfacial restora-tion and cell flattening. The cellular adaptive wetting behavior isclearly detectable because the cell traction-mediated IPLM defor-mation speed is faster than that of cytoskeletal rearrangement,thereby reflecting the cellular viscoelastic nature. In fact, we havedemonstrated that cellular adaptive wetting can be simulated bythe wetting of viscoelastic polymeric liquid drops at air–waterinterfaces.[11] By considering the significant role of ECM vis-coelasticity in cell spreading in general, it is reasonable to assumethat cellular adaptive wetting can also depend on ECM viscoelas-ticity. However, our previous platform was not suitable to system-atically address this issue.In this study, we refined the previous IPLM system to inves-tigate how dynamic viscoelasticity on cell-ECM interactions af-fects cellular adaptive wetting behavior. Our updated platformintegrates diacetylenyl phosphatidylcholine (DA–PC) with otherlipid species that self-assembled at the water–PFCL interface(Figure 1a). DA–PC with triple bonds was covalently crosslinkedwith neighboring DA–PC molecules by 254 nm ultraviolet (UV)light irradiation (Figure 1b). This was utilized to control the de-gree of crosslinking depending on the UV dose in SLBs, chang-ing the speed of mass transport along the membranes by theformation of stationary obstacles.[12] Therefore, we hypothesizedthat DA–PC lipids were suitable for controlling the viscosity ofassembled phospholipid membranes to manipulate the overallviscoelastic properties of super-soft DA–PC-based IPLM systems(Figure 1c). Confocal analysis of the IPLM, together with a recon-stituted cellular side view from the stacked images, allowed us toinvestigate the impact of viscoelasticity on cell-IPLM interactions(Figure 1d). The cellular adaptive wetting behavior was observed(Figure 1e), where the interface deformed first without cellularflattening, followed by cellular spreading and interfacial restora-tion. This clearly demonstrated the potential of this platform instudying cellular adhesion behaviors with a mutually adaptive na-ture for both ECM and cells.Adv. Funct. Mater. 2025, 2414534 2414534 (2 of 12) © 2025 The Author(s). Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adfm.202414534 by Cochrane Japan, Wiley Online Library on [16/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.afm-journal.dewww.advancedsciencenews.com www.afm-journal.deFigure 2. Photocrosslinking reaction of DA–PC within the vesicles and fluid interface. a) Photographic images (left) and UV–vis absorption spectrum(right) of lipid vesicles before and after 254-nm UV irradiation for 6 J cm−2. b) Evaluation of the fluid nature of IPLM by fluorescence recovery afterphotobleaching (FRAP) with irradiation of 545 ± 5 nm light at a polygonal region. Fluorescence images represent the same area before photobleaching(pre-bleach), or after 0- and 120-min intervals from photobleaching. Red color represents rhodamine-labeled lipid. Scale bars represent 50 μm. c)Quantitative analysis of the recovery efficiency in the FRAP experiment shown in (b) and Figure S3 (Supporting Information). The fluorescence intensity(F.I.) values in the bleached regions were normalized to the value before photobleaching. The number of experiments was three for each case. Statisticaldifferences were analyzed using Student’s t-test: *p < 0.05. N.S., no significance, p ≥ 0.05.2. Results2.1. Modulation of Interfacial Phospholipid Membrane ViscosityVia UV DoseIn this study, photocrosslinkable IPLM was prepared by coat-ing the water–PFCL interface with phospholipid membranesvia vesicle fusion (Figure S1, Supporting Information). Wefirst studied the photocrosslinkable nature of the phospholipidvesicles containing DA–PC before and after their assembly atthe water–PFCL interface. The composition of vesicles is crit-ical for the efficient control of membrane viscosity depend-ing on the degree of cross-linking.[13–16] Here, we used DA–PC:distearoylphosphatidylcholine (DSPC):lissamine rhodamine-labeled dioleoylphosphatidylethanolamine (Liss–Rhod DOPE) =85/14.5/0.5 mol.%. Within this lipid composition, the saturatedDSPC species enable the DA–PC molecules to pack tightly to-gether to improve the crosslinking efficiency.[13] The Liss–Rhodtag was introduced for the following fluorescence observation oflipid distribution as well as the evaluation of the viscous natureof the fused membranes by fluorescence recovery after photo-bleaching (FRAP) experiments.Crosslinking of vesicles containing the DA–PC molecule wasstudied by UV–visible (vis) spectroscopy. Figure 2a shows thephotos and the absorption spectra of the phospholipid vesi-cles dispersed in PBS buffer before and after UV irradiation at254 nm. The significant change in the solution color and in-crease in the absorbance of the wavelength less than 550 nm in-dicated the UV-triggered polymerization of DA–PC.[13] Two newpeaks emerged around 480 and 525 nm. It is known that theabsorption peak wavelength depends on the polymer units; themore coupled the polymer units, the longer the absorbance peakwavelength. Accordingly, the absorption peaks around 480 and525 nm corresponded to shorter and longer polymerized units,respectively.[17,18] We did not further investigate the correlationbetween the magnitude of the two absorbance peaks and theincrease in the UV dose for the vesicles because of the differ-ence in curvature between the vesicles and the planar IPLM. Thisfollows the opinion that planar surfaces may have larger well-packed domains than curved vesicles, changing the crosslinkingefficiency.[19]Next, we evaluated UV-induced crosslinking at the water–PFCL interface in terms of the distribution of the Liss–Rhod-labeled fluorophore. Phospholipid molecules with anAdv. Funct. Mater. 2025, 2414534 2414534 (3 of 12) © 2025 The Author(s). Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adfm.202414534 by Cochrane Japan, Wiley Online Library on [16/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.afm-journal.dewww.advancedsciencenews.com www.afm-journal.deamphiphilic structure in the vesicles self-assembled into a con-tinuous planar surface at the water–PFCL interface prior to UV ir-radiation (Figure S1c, Supporting Information). Despite the com-plexity of the lipid composition, we expected all the lipid compo-nents, regardless of their phase transition temperature, to trans-fer to the planar phospholipid membrane at ambient tempera-ture. This is because according to the literature,[20] the planarlipid layer can be formed by supplying a sufficient amount ofvesicles in aqueous phase to ensure efficient mass transfer fromthe vesicles to the surface. In our case, bulk aqueous disper-sion also allowed efficient mass transfer of phospholipids to theinterface compared with the assembled thin IPLM layer. Afterthe given time for assembly, the Liss–Rhod fluorophore was ho-mogenously distributed in the membrane without UV irradiation(Figure S1d, Supporting Information). To assess the morphol-ogy of the interface and presence of polymerized DA–PC in theIPLM, atomic force microscopy (AFM) was performed. Topogra-phy images of IPLM after treatment with 3, 6, 9, and 12 J cm−2UV doses are shown in Figure S2 (Supporting Information). Theroughness of the resulting polymerized IPLM decreased gradu-ally with increasing UV dose. This follows the opinion that dur-ing polymerization the diacetylene lipid layer shrinks in the ver-tical direction.[21]It is known that the polymerization of lipid layers composed ofpolymerizable and nonpolymerizable lipids can cause the poly-merized domains to grow through the diffusion of reactive DA–PC monomers.[22] Therefore, obvious phase separation was ex-pected depending on the UV dose. Here, the distribution oflipids within the IPLM was investigated by fluorescence mi-croscopy. Figure S3 (Supporting Information) (pre-bleach part)represents the changes in the distribution of the Liss–Rhod flu-orophores depending on UV dose. At a low UV dose, such as3 J cm−2, small black dots representing polymerized DA–PC-rich clusters appeared across the membrane surface throughseparation from the mixture of fluorophore lipids and the un-crosslinked monomer DA–PC, showing the majority with red flu-orescence. In addition, large red fluorescent regions that shareda similar fluorescence featuring with the original IPLM beforeUV irradiation (Figure S1d, Supporting Information) indicatedthe capability for further crosslinking of DA–PC. At higher ir-radiation energies of 6 and 9 J cm−2, the polymerized regionswere connected to each other to make the black domain predom-inant on the surface, with the Liss–Rhod fluorophore confinedwithin smaller fluorescent dot regions, indicating that polymer-ized domains grew gradually via the incorporation of reactiveDA–PC monomers.[23] When the irradiation energy increased to12 J cm−2, the crosslinked DA–PC molecule ratio became muchlarger, and the surface was mostly covered by the black domainswith enlarged fluorescent regions in smear-like structures due tothe stronger phase separation between the two domains. There-fore, the growth of polymerized domains was promoted by in-creasing the UV dose.We then investigated the viscosity of the DA–PC-based IPLM atthe water–PFCL interface upon UV irradiation of the planar sur-face, as polymeric domains would serve as obstacles hinderingthe lateral diffusion of fluid lipids (Figure 2b).[12] The lateral dif-fusivity of the fluorescein-labeled phospholipid membrane wasmeasured by FRAP, which is also a critical characteristic of theviscosity of the IPLM based on the molecule diffusion because theinertia of a liquid and its softness make the conventional charac-terization of viscosity a challenging task.[24] Using this method, adefined polygon was selected as a region of interest (ROI), andsubsequently, the fluorophore was initially bleached by expos-ing strong light at the ROI. Time-lapse fluorescence microscopyimages showed bleaching retention or recovery of fluorescencein the bleached region of the entire membrane depending onthe diffusion of the fluorescent molecules from the surround-ing, which correlated with the different crosslinking levels of themembranes.When the irradiation energy was 3 J cm–2, there was a slightdrift in the entire membrane, and the intensity of the bleachedarea recovered to 80% of the original value after bleaching for 2 h(Figure 2c). As low UV dose was supposed to only produce smalldark polymerized islands, they were not able to cover entire re-gions of the membranes (Figure 2b and 3 J cm−2 case). Thus,the entire IPLM was mobile at room temperature (RT), whichenabled fluorophore recovery in the bleached region. Therefore,the rapid lateral diffusion of fluorescent molecules from the sur-roundings for the recovery indicated that the IPLM had lowerviscosity, because of insufficient irradiation. Thereafter, upon in-creasing the UV dose in the range of 6 to 12 J cm−2, there wasalmost no drift of the membrane because the large, crosslinkedDA–PC regions had an oriented structure in the gel phase, whichprevented the entire mobility. It should be noted that there waslittle recovery of intensity in the bleached areas starting from 6J cm−2 (Figure S3, Supporting Information, 120 min part), as theentanglement of polymer clusters acted as a barrier to block lat-eral diffusion of fluid lipids. From the dose-dependency of thepercentage of the fluorescence recovery in the bleached regions(Figure 2c), there was a drastic change in the fluorescence recov-ery from 3 to 6 J cm−2. However, this did not mean the comple-tion of the crosslinking reaction above 6 J cm−2, rather it con-tinued to change the viscoelastic nature with prolonged irradia-tion. In fact, interfacial rheology with double wall ring (DWR) ge-ometry showed the change in the crossover points of G’ and G’’toward a lower frequency (Figure S4, Supporting Information),which demonstrated continuous alteration of the viscoelastic na-ture through increasing crosslinking levels.[25,26]2.2. Characterization of IPLM Mechanical Tunability by AtomicForce Microscopy (AFM)As discussed, the extensively polymerized domains of DA–PC re-sulted in the fluid–gel transition and significantly lowered lipiddiffusion, which reflected the increases in interfacial viscosity inthe lateral direction. Simultaneously, the conjugated backbonein the gel phase conferred high elasticity to the membrane inthe vertical direction. Next, we investigated the impact of thecrosslinking level on the mechanical properties of the IPLM us-ing AFM nanoindentation to measure the apparent modulusat the liquid–liquid interface by using a colloidal probe with a300 nm radius (Figure 3a).The nanoindentation mode allowed us to obtain force-distancecurves on the whole 5 × 5 μm2 scan with a resolution of 64 pixels× 64 pixels. Figure 3b shows a representative force-distance curveas an example from the below stiffness map (Figure 3d, each map= 4 × 4 = 16 force curves). By fitting each force curve with theAdv. Funct. Mater. 2025, 2414534 2414534 (4 of 12) © 2025 The Author(s). Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adfm.202414534 by Cochrane Japan, Wiley Online Library on [16/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.afm-journal.dewww.advancedsciencenews.com www.afm-journal.deFigure 3. Mechanical property characterization of the UV-irradiated IPLMs by AFM nanoindentation. a) Scheme of AFM nanoindentation. b) Represen-tative force-deformation curve (blue dotted line) and fitting curve based on the Hertz model (orange solid line) for the IPLM with 9 J cm−2 UV-irradiation.c) Dose-dependent changes in apparent modulus. Mean ± s.d. of the 16 different positions in d) are plotted. (d) AFM stiffness maps of the UV-irradiatedIPLM in 300 × 300 nm2 (4 pixels × 4 pixels).Hertz model (Figure S5, Supporting Information) to obtain andplot the respective apparent modulus values, we determined howmechanical stiffness correlated with exposure to contrasting UVdoses by varying the degree of DA–PC crosslinking. The resultshowed a dose-dependent increase in the stiffness of membranesabove 6 J cm–2, demonstrating the tunability of mechanical stiff-ness (Figure 3c). However, the apparent modulus of the IPLM didnot increase linearly with the UV dose, presumably because theconjugation length of DA–PC molecules is not sufficiently highto alter the membrane stiffness with the short irradiation.The apparent modulus given by AFM analysis reflects the stateof the IPLM coated interface (the deformability), although it iscomposed of material modulus and interfacial tension, consid-ering the different mechanics between interfacial and conven-tional bulk materials.[27] To understand the contribution of theabove two factors to the IPLM mechanics for the minimum andmaximum crosslinking level, we dismantled the interfacial ten-sion at the fluid interface (pre-stress) and the IPLM crosslinking-derived modulus, based on the informatics-assisted AFM methoddeveloped by our group.[28] In Figure S6 (Supporting Informa-tion), the estimated moduli and the stress (derived from interfa-cial tension) are shown as red and blue plots, respectively. The re-sults demonstrated these two values are comparable for 3 J cm−2,whereas the contribution of materials’ modulus is dominatingfor 12 J cm−2 due to the increased crosslinking levels. Therefore,we can say that the interfacial viscoelasticity changed from theinterfacial tension-dominating state to the materials modulus-dominating one through increasing the crosslinking level.Furthermore, by drawing the combination relationships be-tween elasticity together with mobility and various UV doses(Figure S7, Supporting Information), the elasticity and mobil-ity gradually increased and decreased, respectively. Particularly inthe case of 6 J cm−2, the elasticity did not increase significantly de-spite the generally decreased mobility. In this respect, the IPLMat 6 J cm−2 may be located close to the boundary between thefluid–gel transition. The FRAP experiment reflects the viscousnature of lipids in terms of molecular diffusion against Van derWaals interaction between the lipid molecules, whereas AFMnanoindentation reflects membrane mechanics against force ex-ertion. Therefore, it is no wonder these two methods demonstratedifferent boundaries in the transition between the viscosity- toelasticity-dominant states. Upon 9 J cm−2 exposure, the IPLMtransitioned to a more solid-like state; thus, there was an increasein the mechanics in a dose-dependent fashion. Therefore, mobil-ity and mechanical experiments confirmed the phototunable vis-cous and elastic nature of the membranes. The non-mobile IPLMgenerated from 6 J cm−2 exposure displayed high viscosity, andthe elasticity increase observed from 9 J cm−2 exposure via AFMnanoindentation originated from its solid-like nature in responseto high crosslinking levels.2.3. Impact of the Viscoelastic Nature of IPLM on CellularAdaptive Wetting CharacteristicsOur previous study defined adaptive wetting behavior with theattribution of out-of-plane deformation at the fluid interface anddynamic mechanical energy transfer in quasi-3D dimension.[11]Here, we applied the above-discussed customizable fluid inter-face to examine how IPLM viscoelasticity alters cellular adaptivewetting behaviors. Specifically, we investigated early-stage cellu-lar adhesion behaviors after seeding Madin–Darby canine kidney(MDCK) cells stably expressing lifeact-green fluorescent protein(GFP) as these epithelial cells exhibit unique characteristics inAdv. Funct. Mater. 2025, 2414534 2414534 (5 of 12) © 2025 The Author(s). Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adfm.202414534 by Cochrane Japan, Wiley Online Library on [16/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.afm-journal.dewww.advancedsciencenews.com www.afm-journal.deFigure 4. Representative images of cellular wetting behaviors at the IPLM with different UV doses. a) Cells are seeded to IPLMs with 3, 6, 9, or 12 J cm−2irradiation. b–d) Time-dependent morphological transition and interfacial deformation. Green, lifeact-GFP. Red, rhodamine-labeled lipid. (b) Cells didnot spread at 3 J cm−2 IPLM. (c,d) Cellular adaptive wetting with (c) deep and (d) shallow invasion, at 6 (or 9) and 12 J cm−2 IPLMs, respectively. Thetop views in each case are the cross-sectional images. The side views were reconstituted from stacked confocal images. Arrowheads on the left indicatethe water–PFCL interface level.the interactions between the cell sheets and substrate through atug-of-war mechanism.[29,30] It is reasonable to assume that a sin-gle epithelial cell can also exert force in all directions consideringthat in vivo they are 3D and can sense and respond to the 3Dgeometry in their environment.[31]First, the bioinertness of IPLM in preventing protein adsorp-tion was investigated as it is critical to maintain the intrinsicallysuper-soft nature of the fluid interface.[11] Compared with thepristine perfluorocarbon liquid, the headgroup of the IPLM de-rived from phospholipid molecules formed a hydration layer thatlargely prevented nonspecific protein adsorption regardless ofthe UV irradiation energy (Figure S8, Supporting Information).Therefore, the deformable nature of the fluid interface toward theout-of-plane direction can also be expected in our photocrosslink-able IPLM system. Next, the biocompatibility of the IPLM wasassessed by evaluating cell survival using a live/dead staining as-say after incubation of MDCK cells with the scaffolds for 2–3 h.As shown in Figure S9 (Supporting Information), the percentageof live cells (appeared as green fluorescence) was almost 100%for all the scaffolds, which verified the IPLM scaffolds had goodbiocompatibility.Given that photocrosslinkable IPLM are bioinert and bio-compatible, we next investigated the adhesion behaviors of theMDCK cells expressing lifeact-GFP thereon by confocal laserscanning microscopy (CLSM) (Figure 4). Considering the het-erogeneous adhesion dynamics of individual cells owing to themixed clones of cells expressing different levels of lifeact-GFPand unsynchronized cell cycles, we first analyzed cellular adap-tive wetting processes of different cells at specific time pointsand looked at the impact of the viscoelastic properties of theIPLM. Figure 4 shows representative images of time-dependentbehaviors. At lower crosslinking levels (3 J cm−2), most cellswere attached with almost no downward indentation through-out the observation time (Figure 4b1). In contrast, unique adap-tive wetting behaviors were observed by increasing irradiationenergy to more than 6 J cm−2, in a fashion similar to ourprevious study,[11] where IPLM deformed first followed by in-terfacial restoration and cellular dimensional changes. Specif-ically for 6 and 9 J cm−2 irradiation, almost the entire cellbody went below the original water–PFCL level at the invasionphase (Figure 4c1) but returned to almost the original level andspread after ≈4 h (Figure 4c2). This can be clearly observedAdv. Funct. Mater. 2025, 2414534 2414534 (6 of 12) © 2025 The Author(s). Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adfm.202414534 by Cochrane Japan, Wiley Online Library on [16/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.afm-journal.dewww.advancedsciencenews.com www.afm-journal.deby focusing on the same cells (Figure S10, Supporting Infor-mation). When we further increased the irradiation energy to12 J cm−2, the invasion depth decreased and IPLM deforma-tion was partially retained (Figure 4d1). Subsequently, the de-formed IPLM transitioned from the PFCL phase to the aqueousphase and cells eventually showed a large-spreading morphol-ogy with IPLM restoration at the end (Figure 4d2). By consider-ing FRAP and AFM nanoindentation experiments, the observedchanges in IPLM deformation and cellular dimension for dif-ferent irradiation energies can be attributed to the regulation ofcellular (adaptive) wetting dynamics by the viscoelastic nature ofIPLM.To further check cellular morphology with a prolongedtimescale, the observation was extended up to 15 h. Cellsshowed different dimensions and spreading morphology as if thecells were undergoing epithelial-mesenchymal transition (EMT)(Figure S11, Supporting Information). Like less viscous 6 J cm−2IPLM, the cells showed the epithelial trait with partial spreadingand maintained a larger cell dimension. In contrast, in the caseof more viscous 9 and 12 J cm−2 IPLMs, the cells gained a mes-enchymal morphology with a larger spreading area. It is knownthat cell morphology is correlated with cellular fate decision cor-responding to the mechanical cues. Therefore, the interplay be-tween cell and IPLM could be an indication of changes in cellfunctions.2.4. Quantitative Analysis of the Impact of IPLM ViscoelasticNature on Cellular Dimension and Interfacial Deformationduring Adaptive WettingAs discussed, the cells showed different wetting behaviors de-pending on the interfacial viscoelasticity, with different degreesof cellular invasion into the PFCL phase. However, the trendswere only discussed by a representative profile on IPLM of eachcrosslinking level, and in fact, cellular wetting behaviors were dif-ferent from one cell to another. To further understand the entirepicture of cellular wetting behaviors for single-cell populations,we focused on and quantitatively evaluated cellular morphologyduring a specific time window (1–2 h after cell seeding), when cel-lular invasion into PFCL was largest for nearly all irradiation en-ergies. To simplify this analysis, we first divided cell attachmentphenotypes into two groups: either the invasion depth (d) wassmaller than 2 μm (d<2) or larger (d≥2) (Figure 5a). This depthwas chosen because two planes need to be separated at a certaindistance to discuss cellular incident angle (𝜃) precisely. The for-mer group was further divided into two based on the cellular ap-pearance in the green fluorescence channel: i) no invasion and ii)highly spreading morphology (Figure 5b). By comparing D0 andD2 at the depth of 0 and 2 μm processed with the red fluorescencechannel, which was the diameter of the convergent boundary be-tween cell and IPLM at the cross-section image (Figure 5a), thelatter group was also divided into two: iii) partial invasion withincident angle (𝜃) < 90°and iv) the deep invasion with 𝜃 > 90°(Figure 5b). For a partial invasion, there was a large opening atthe water–PFCL interface with a protrusion structure at the cellperiphery for partial cell body diving (Figure 5b,iii). In contrast,the deep invasion had a small opening at the level of the water–PFCL interface, where the entire cell body was almost wrappedby the IPLM, with a significant degree of deformation observed(Figure 5b,iv).The histogram of single-cell populations involved in the fourgroups is plotted in Figure 5c. The quantification of the cell pop-ulation in each morphology clearly represents the reproducibleobservation of cellular deep invasion (Figure 5b,iv) for 6 and 9J cm−2, that is, at least 30%–50% of the total number of cells.By considering the single-cell invasion process (Figure S10, Sup-porting Information) as well as the cellular heterogeneous na-ture, we cannot exclude the fact that there may be more cells thanthis population; as we fixed the observation time point at 1–2 h,some cells might have already finished diving or were about to gointo the deep dive afterward. In contrast, most cells stayed at theshallow invasion (47%) (Figure 5b,iii) or spreading morphology(37%) (Figure 5b,ii) on the stiffest IPLM at a UV dose irradia-tion of 12 J cm−2. In addition, the cells at the IPLM at a UV doseof 3 J cm−2 were mostly attached at the interface with no inva-sion (70%) (Figure 5b,i) or partial invasion (30%) (Figure 5b,iii).The higher population of deep invasion (iv) (48% versus 34%)and lower population of no invasion (i) (19% vs 26%) for 9 and 6J cm−2, respectively, might indicate more efficient force exertionfor more viscous 9 J cm−2 IPLM, as discussed in the earlier workon supported lipid membranes.[7] This tendency also agreed withthe slight shape changes of the photobleached region at 6 J cm−2in the FRAP measurement (Figure S3, Supporting Information).However, further analysis of the actin retrograde flow speed onthe two interfaces are needed to attribute the small difference inthese populations.The single-cell population density of cellular adhesion pheno-types is correlated with the mechanical properties of IPLM, re-vealing unanticipated relationships between cell type and IPLMviscoelasticity. Although invasion serves as an important deter-minant of adhesion, cellular adaptive wetting revealed the inter-play between cells and the deformation of IPLM. This raises thequestion of what other shape parameters are coupled with IPLMviscoelasticity and how the interplay can be understood quanti-tatively. Here, we relate cellular dimension and invasion depthto the IPLM viscoelasticity changes. The quantification of cellu-lar dimension and invasion depth can be directly read out fromthe above four cellular phenotypes to understand the correla-tion between cellular adaptive wetting and morphology changes.Figure 5d,e summarizes IPLM deformation (Figure 5d) and cel-lular dimension (Figure 5e) at the cellular deepest invasion states.Both interfacial deformation and cellular dimension exhibited abiphasic response. At low crosslinking levels with 3 J cm−2 irra-diation, the invasion depth was small (≈1.8 μm). It seems rea-sonable that the cells were not able to build up contractile forcesowing to force dissipation,[7] considering the lateral mobility ofthe IPLM characterized by the FRAP measurement. However, byincreasing the viscosity to that of the IPLM with 6 and 9 J cm−2irradiation, there were no significant differences at the deep in-vasion depth of ≈21 μm. By proceeding with UV irradiation, theIPLM with strong stiffness presented higher resistance againstthe cellular traction force, which resulted in partial invasion ofcells of ≈12 μm. Therefore, the corresponding cellular adhesionchanged by manipulating the interfacial viscoelasticity. Cellulardimension also showed a similar tendency (Figure 5e). Especiallyfor intermediated crosslinking levels at 6 and 9 J cm−2, cells ex-hibited significantly elongated dimensions (≈34 μm) comparedAdv. Funct. Mater. 2025, 2414534 2414534 (7 of 12) © 2025 The Author(s). Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adfm.202414534 by Cochrane Japan, Wiley Online Library on [16/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.afm-journal.dewww.advancedsciencenews.com www.afm-journal.deFigure 5. Quantification of the impact of IPLM viscoelasticity on cellular wetting behaviors for single-cell populations. a) Schemes used for classificationof cellular adhesion states. The reconstituted side view was used to determine the IPLM indentation depth (d) and the cellular morphology eitherfrom red and green fluorescence channels, or both. The cross-section views of the red fluorescent channel at d = 0 and 2 μm were used to detect theconvergent boundary between cell and IPLM, whose diameter was defined as D0 and D2, respectively. b) Flow chart of the image processing strategy.By considering of the values of d, D0, D2, and cellular morphology, the cell adhesive phenotypes on the IPLM were classified into four groups. If theinvasion depth was smaller than 2 μm, cells were divided into (i) no invasion or (ii) highly spreading by processing the green fluorescence channel.If the invasion depth was higher than 2 μm, they were further divided into another two groups by comparing D0 and D2: (iii) partial invasion and (iv)deep invasion. Representative images are shown at the bottom. 𝜃 represents cellular incident angle in the PFCL phase. c) Distribution percentage ofcellular phenotypes with regard to the UV irradiation dose. The number of analyzed cells is indicated in parentheses below each bar. d,e) Quantificationof changes in adhesion parameters: (d) invasion depth and (e) cellular dimension. The number of analyzed cells is 21, 20, 25, and 6, for 3, 6, 9, and 12J cm−2, respectively. Statistical differences were analyzed using Student’s t-test: *p < 0.05, **p < 0.01, and ***p < 0.001. N.S., no significance, p ≥ 0.05.with those of 3 and 12 J cm−2, presumably owing to the elasticresistance forces of the IPLM squeezing the cell body into theupper aqueous layer.[11] Although the depth and cellular dimen-sion were no different at either 6 or 9 J cm−2 owing to the cellulardimension limitation, the stiffer one may provide higher resis-tance force by improving the efficiency of forces transmission.3. DiscussionOur study demonstrates the application of a new photopolymer-izable IPLM as a cell scaffold, which displays tunable mechanicalproperties and thereby enables the study of dynamic adhesionbehavior changes in cells together with the IPLM in responseto the altered viscoelastic properties of a model ECM. Althoughour previous work reported the physically mixed phospholipidinterface,[11] in which the lower phase of the hydrophobic speciesneeds to be changed to adjust the interfacial tension to investigatethe adhesion behavior, our current approach enabled controllingthe degree of crosslinking by UV shining to tune viscosity andelasticity. Specifically, this platform allows us to fabricate ECMmimics, where the viscosity balance between cell and ECM canbe manipulated, to build fundamental knowledge about the in-fluence of dynamic viscoelasticity of both cells and ECM on theadaptive wetting behavior.The viscosity-tunable nature of the new IPLM was investigatedusing FRAP by measuring the lateral diffusion of fluorescent-labeled phospholipids. At lower irradiation doses, fluidity was ob-served owing to the recovery of the bleached region. With increas-ing irradiation dose up to 6 J cm−2, the lateral fluidity was signif-icantly suppressed. However, this did not mean the completionof the crosslinking reaction above 6 J cm−2, rather it continuedto change the viscoelastic nature with prolonged irradiation ver-ified by interfacial rheology. Interfacial elasticity was character-ized by nanoindentation using AFM, with a significant decreasein the out-of-plane deformability above 9 J cm−2. Further disman-tling the interfacial apparent modulus into material modulusand interfacial tension-derived pre-stress, the mechanical matu-ration of the IPLM was demonstrated, with a transition from theinterfacial tension-dominating state to the materials modulus-dominating one in response to increasing crosslinking levels.Therefore, the successful preparation of a highly deformablefluid interface with varying viscoelasticity was confirmed.Adv. Funct. Mater. 2025, 2414534 2414534 (8 of 12) © 2025 The Author(s). Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adfm.202414534 by Cochrane Japan, Wiley Online Library on [16/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.afm-journal.dewww.advancedsciencenews.com www.afm-journal.deBy observing epithelial MDCK cells seeded on the IPLM usingconfocal microscopy, we found that the cells actively deformedthe fluid interface by switching the dimensions of the IPLM froma 2D- to a 3D-like structure to reshape the adhesive environmentover time. To date, the impact of viscoelasticity on cellular ad-hesion behaviors has been discussed on viscoelastic hydrogels orviscous-supported lipid bilayers. The viscous nature is introducedinto either the materials’ bulk or interface, and its impact on thecellular straightforward transition from spherical morphology tothe spreading state is established via molecular clutch-based ad-hesions in a similar fashion to that of the elastic component. Incontrast to this so-called canonical wetting behavior, we are ableto see the viscoelastic regulation of cellular adaptive wetting interms of the degree of out-of-plane deformation depending onthe balance between cellular and IPLM viscosity in our highly de-formable viscoelastic platform. In our previous report,[11] we haveexperimentally simulated cellular adaptive wetting behaviors us-ing viscous PDMS drops at the air-water interface. We foundthat adaptive wetting emerged only by matching the timescaleof viscous relaxation of the PDMS drops and the observationtimescale. By increasing the PDMS viscosity, the indentation ofthe PDMS-water interface become larger with a smaller deforma-tion of the PDMS drop. Considering these previous simulationresults, the cellular viscosity increase/decrease is expected to re-sult in an increase/decrease in cellular invasion depth and retar-dation/acceleration of the cellular spreading process in our sys-tem. Moreover, we also observed that the interface with mediumviscoelasticity or stiffness allowed significant diving of the cellsinto the PFCL with increasing cellular dimension, which is sig-nificantly different from cell adhesion behaviors on conventionalviscoelastic platforms. This biphasic profile was reminiscent ofan earlier work in which the intermediate viscosity of soft sub-strate regulated optimal cellular spreading speed,[1] emphasizingthe importance of viscosity-based regulation.From an application perspective, this new viscoelastically-tunable IPLM could be further used to engineer cellular fate, asthe platform is capable of manipulating cellular adaptive wetting,where mechanical forces are dynamic depending on the degreeof out-of-plane deformation. During cellular deep invasion, theIPLM switches from a 2D planar surface to a quasi-3D form bywrapping the cell body, thereby enabling the cells to receive bio-chemical and mechanical cues from the majority of the surfacewithout undergoing significant morphological alterations. Thisconfiguration, featuring a 3D contacting area between cells andthe model ECM, closely resembles native tissues. In contrast,conventional 2D substrates require cells to be flattened to maxi-mize their interactions with ECMs. Subsequent reversible defor-mations modulate the rearrangement of the cytoskeleton. By con-sidering the existence of cellular mechanical hysteresis,[32] wherecellular functions and fate are regulated not only by the mechan-ics of their surrounding environment at that moment, but alsoby those to which the cells have been exposed in the past, the dy-namic changes in IPLM mechanics can impact single cell func-tions, like the speed of actin retrograde flow and stem cell fatedecision. The cell traction force-driven 3D cellular environmentsdescribed above might be advantageous over the usual 3D cul-ture platform in terms of precise optical imaging and cell har-vesting. This allows the target cells to be addressable both mi-croscopically and enzymatically at the 2D fluid interface, therebyenabling direct correlation of microscopic analysis data and geneexpression patterns after trypsinization. This will be useful tofurther investigate how cell-matrix bidirectional interaction reg-ulates cellular activities. Consequently, we hypothesize that ourmethod can facilitate the precise control of cell–matrix interac-tions at the single-cell level, which may enable the prediction ofdynamic cellular responses in physiological and pathological pro-cesses, such as immune cell extravasation and EMT in cancerprogression.4. ConclusionThis study reports a new photopolymerizable fluid interfacial cel-lular scaffold, displaying tunable viscoelastic properties depend-ing on the crosslinking level. Our system offers a systematicallymanipulable viscoelasticity platform that can reveal the impactof material viscoelastic properties on cellular adhesion behav-iors, specifically cellular adaptive wetting, with different degreesof out-of-plane deformation; this cannot be observed in conven-tional solid- or hydrogel-based materials. These adaptive wet-ting transitions depending on the balance between cellular andIPLM viscosity may enable the prediction of dynamic adhesionresponses in physiological and pathological processes. Further-more, our new viscoelasticity-tunable interface can be used to en-gineer cellular fate, as the platform is capable of mechanobiologi-cally manipulating cellular 2D-to-3D transition behaviors, wheremechanical forces between the cell and model ECM change dy-namically both in time and space.The apparent modulus of IPLM (up to 10 kPa for 12 J cm−2)can cover the range of softer living tissues, such as brain[2] andbone marrow.[33] Moreover, our IPLM platform shows mechan-ics with stronger modulus in the in-plane direction but smallervalue for the out-of-plane direction. We can find similar mechan-ical anisotropy[34] in living tissues, including the brain tissue,cornea,[35] endothelial layer in the blood vessels,[36] and the bonemarrow niche.[33] Therefore, cellular behaviors in our model sys-tem could mimic cellular mechanical responses in in vivo envi-ronments.During cellular adaptive wetting, cells undergo dynamic di-mensional changes. These share similarities to the EMT pro-cess and extravasation of immune cells. As an epithelial cell type,MDCK cells can transit from a spherical shape with a collectivefeature to single cells with spreading morphology via the EMT-like process. Through this, the cells undergo a dynamic changein dimension, losing the cohesive nature and increasing mo-bility. This phenomenon can be a hallmark of epithelial cancercells undergoing invasion and metastasis.[37] Another example isthe inflammation-induced emigration of immune cells from thebloodstream into the tissue. The immune cells undergo a mor-phological change from a circular to an amoeboid shape to ad-here to the inflamed endothelium to identify an entry point. Thisprocess eventually results in their extravasation into the infectedtissue. This transition is accompanied by a substantial alterationin dimensional parameters.[38] The similarity between cellulardimensional changes at the interface and these in vivo eventsmakes our system a promising tool for studying dynamic cell be-havior and interactions in physiological environments. This ad-vancement has the potential to enable more precise predictionsof cellular responses in real-life biological processes.Adv. Funct. Mater. 2025, 2414534 2414534 (9 of 12) © 2025 The Author(s). Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adfm.202414534 by Cochrane Japan, Wiley Online Library on [16/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.afm-journal.dewww.advancedsciencenews.com www.afm-journal.de5. Experimental SectionReagents: The 1,2-bis(10,12-tricosadiynoyl)-sn-glycero-3-phosphoch-oline (DA–PC), 1,2-Distearoyl-sn-glycero-3-phosphocholine (18:0 PC,DSPC), 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[4-(p-(cysa-rginylglycylaspartate-maleimidomethyl)cyclohexane-carboxamide] (RGD-DSPE), and 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(lissaminerhodamine B sulfonyl) (Liss-Rhod DOPE) were purchased from Avantipolar lipids (Alabaster, AL, USA). FC-70, trichloro(3,3,4,4,5,5,6,6,7,7,8,8,8-tridecafluorooctyl) silane, fluorescein isothiocyanate labeled bovineserum albumin (FITC-BSA), and transforming growth factor (TGF)-𝛽1were obtained from SIGMA-Aldrich (St. Louis, MO, USA). Potassiumhydroxide (KOH), chloroform, ethanol, and phosphate buffered saline(PBS) tablets were purchased from Wako (Tokyo, Japan).Unilamellar Vesicle Preparation: Solid DA–PC, DSPC, RGD-DSPE, andLiss-Rhod DOPE at a ratio 85/12.5/2/0.5 mol.% was dissolved in chloro-form as a stock solution and stored at −20 °C until use. A lipid film in a10 mL glass vial was formed by evaporating the chloroform in the mixedstock solution and then drying the lipids under vacuum for ≈3 h. The mixedlipid film was hydrated in PBS to a final concentration of 1 mm. The sus-pension was heated to 50 °C in an oven with manual shaking from timeto time. Immediately after heating, the lipid suspension was sonicated forclarity at RT using a probe-type ultrasonic disrupter (UD-100; Tomy, Tokyo,Japan) with 10 cycles of intermittent operation (45 s high-power sonica-tion separated by 30 s intervals). Then, the smaller lipid vesicles weresized by 21 extrusion cycles through a 0.1 μm polycarbonate membranein a Mini-Extruder (Avanti Polar Lipids, Alabaster, AL, USA). The tempera-ture of the sample and extruder was maintained at > 43 °C with a heater.Subsequently, the lipid suspension was held in a 50 °C oven until furtherprocessing.Assembling the IPLM Platform: Fluorination of the in-house-developedcell culture chamber (inner wall: 21 mm; outer wall: 25 mm) was per-formed as previously described.[11] Chambers were activated with potas-sium hydroxide (KOH, 10 wt.%) at 60 °C for 1 h, then rinsed with deionizedwater and dried under air flow. The bottom glass was then fluorinated bytrichloro(3,3,4,4,5,5,6,6,7,7,8,8,8-tridecafluorooctyl) silane (0.5 vol.%) inethanol solution followed by a 30-min heating process at 70 °C. It was thenrinsed 2–3 times with ethanol and water for glass cleaning and sterilizedby autoclaving.To form the water-perfluorocarbon interface, the above fluorination al-lowed the FC-70 (250 μL) wetting well at the lower layer, followed by 2 mLPBS. Next, the preheated phospholipid vesicle solution was added to theupper layer of the PBS to reach a final phospholipid concentration of0.2 mm. The IPLM assembly process was carried out at RT for 2 h. Af-ter successful IPLM layer formation, the membranes were slowly washedby repeatedly replacing the upper layer with fresh PBS before lowering thetemperature to 0 °C with ice. The system was cooled to ≈0 °C to pro-duce the oriented reactive DA–PC monomer, avoid the thermal hysteresis,and make the DA–PC light sensitive after warming above its phase tran-sition temperature.[39] Then, the cooled IPLM was allowed to warm to RTfor 35 min to equilibrate and promote the growth of the DA–PC-rich do-mains before UV-initiated polymerization, because the fluid lipid phasewould promote the incorporation of DA–PC-rich domains.[23] Photosensi-tivity was retained even if the membranes were warmed to RT after coolingto ≈0 °C.254 nm UV Irradiation: The equilibrated IPLM was exposed to 254 nmUV light at RT (CL-1000 Ultraviolet Crosslinker, USA; short-wave assembly115 V, 60 Hz, and 254 nm) and a distance of 5 cm from the lamp. The UVirradiation dose was adjusted by varying the exposure time. For each irra-diation cycle, a UV light dose of 1 J cm−2 was used with different numbersof irradiation cycles (3, 6, 9, and 12 cycles, with ≈2.5 min per cycle).Upon irradiation of the unilamellar vesicles, aqueous dispersion wasconducted with 254 nm UV light (prior to UV irradiation, vesicles werealready lowered to near 0 °C, then equilibrated to RT). The appearance ofthe dispersion color was associated with the polymerization reaction ofDA–PC within the liposome.UV–Vis Spectroscopy: Wavelength-dependent absorbance spectra ofthe unilamellar vesicles were obtained using a UV–vis spectrophotome-ter (UV-Vis; UV-2600, Shimadzu, Kyoto, Japan) at RT. UV (365 nm and10 mW cm−2) or vis (436 nm and 9 mW cm−2) light was irradiated fromthe top using a mercury lamp (SX-UI 251HQ, USHIO, Tokyo, Japan). Thewavelength of irradiated light was selected using bandpass filters (Ed-mund Optics). The polymerization reaction was monitored in the wave-length range between 350 and 650 nm.Ligand Mobility Test of the IPLM: FRAP was used to measure the ex-tent of lateral ligand mobility within the IPLM using a BX-51 upright flu-orescence microscope (Olympus) equipped with a 60× water-immersionobjective lens (Olympus, LUMPlanFLN, Rochester, USA) and CMOS cam-era (MD-695, Molecular Devices, San José, CA, USA) operated using theMetaMorph software (MetaMorph Inc., Nashville, TN 37212). A definedpolygonal region was selected as an ROI for photobleaching by photoirra-diation with 545 ± 5 nm light for 3 min at a polygonal region. Fluorescenceimages of the IPLM were obtained after a given time interval.Interfacial Rheology: The rheology of IPLM at the liquid–liquid inter-face was measured using a rheometer (Discovery HR-20, TA Instruments)with double wall ring (DWR) geometry equipped with a Delrin trough witha circular channel. Briefly, FC-70 (19 mL) was added to the Delrin trough.Excess FC-70 solvent was removed to the level of the liquid layer close tothe sidewall step. A lipid suspension (0.2 mm and 20 mL) was carefullyadded on top of the FC-70 layer. After the above same IPLM assemblingprocess (Section Assembling the IPLM Platform) and UV irradiation (Sec-tion 254 nm UV Irradiation), the DWR was lowered to pass through theaqueous layer until its contact with the FC-70 liquid layer by monitoringthe axial force, followed by further lowering by 500 μm. Frequency sweepswere performed with a constant displacement of 1.0 × 10−3 rad at 25 °C.Atomic Force Microscopy (AFM): AFM was performed to image the insitu planar IPLM using a Nanowizard 4 (JPK BioAFM, Bruker Nano GmbH,Berlin, Germany). A gold-coated silicon nitride AFM cantilever (NovascanTechnologies, Ames, IA, USA) and SiO2 particles with a radius of 300 nmand a spring constant of 0.06 N m−1 were used.After the system was well calibrated, AFM imaging was performed at RT(≈25 °C). All images were recorded at a resolution of 64 × 64 pixels. Theforce and scanning speed were optimized to minimize damage. Imageswere analyzed and processed using the JPKSPM software (JPK/Bruker,Berlin, version 8.0.111).Further curve fitting based on the Hertz model was used rather thanthe value from the stiffness map because the measured response shouldrepresent the material deformation only. An explanation for this discrep-ancy could be the influence of the distortion of the cantilever during theindenting process.For dismantling the AFM analysis into the interfacial tension at the fluidinterface (pre-stress) and the IPLM crosslinking-derived modulus, accord-ing to the previous report,[28] the prestress and modulus were fitted usingEquation (1) at the total 16 positions by focusing on the deformation rangeat 100 nm after removing the cantilever distortion. Fitting errors at eachindentation were also calculated to optimize the best prestress and mod-ulus.F = 4𝜋Rd𝜎cos[tan−1(2√Rd)] + 43(1 − 𝛾2)E√Rd1.5 (1)R: tip radius; 𝜎: pre-stress; d: deformation; 𝛾 : Poisson’s ratio; E: modulus.Protein Adsorption Test on IPLM: A mixed protein solution of FITC–BSA and regular BSA at a weight ratio of 1:9 to a final concentration0.5 mg mL−1 acted as the fouling agent for fluxing the IPLM. After incuba-tion at 37 °C for 1 h, the protein was removed by repeatedly replacing thePBS in the upper layer with fresh PBS. The water–PFCL interface withoutphospholipid assembly served as a pristine interface; the test results areshown in Figure S8 (Supporting Information).Cell Viability Assessment: Live/dead cell staining kit (Invitrogen) wasemployed to assess the viability of non-transfected MDCK cells (therebyno fluorescence from the lifeact-GFP) on the IPLM after incubation for2–3 h. The staining solution was prepared by mixing 5 μL of Calcein-AMand 20 μL ethidium homodimer into 10 mL culture medium. After seedingcells on each IPLM case for 2 h, upper medium was replaced by the aboveAdv. Funct. Mater. 2025, 2414534 2414534 (10 of 12) © 2025 The Author(s). Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adfm.202414534 by Cochrane Japan, Wiley Online Library on [16/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.afm-journal.dewww.advancedsciencenews.com www.afm-journal.deScheme 1. Reconstructed 3D stacked image of IPLM together with a single cell. The right shows a side view of the 3D reconstruction. The quantificationof cellular dimension and invasion depth can be directly read from the side view.medium containing the dye. Then the image of cells on the IPLM withdifferent crosslinking level was captured and quantified by using ImageJsoftware.Cell Culture with IPLM: The culture medium for Madin-Darby Ca-nine Kidney (MDCK) cells (RCB0995, RIKEN cell bank) that stably ex-pressed lifeact-GFP was prepared by diluting 10× minimal essentialmedium (MEM, Thermo Fischer Scientific, Waltham, Massachusetts,U.S.) 10 times, and supplementing it with 10% heat-inactivated fetalbovine serum (FBS, EU origin, Biowest, Naullie, France), 100 units mL−1penicillin, 100 mg mL−1 streptomycin (Nacalai, Kyoto, Japan), 1% MEM-nonessential amino acids (Nacalai), 1% sodium pyruvate (Nacalai), 1%L-glutamate (Nacalai), and 2.2 g L−1 sodium hydrogen carbonate (Wako).The cells were cultured in a Petri dish by adding 10 ng mL−1 TGF-𝛽1 to theabove culture medium for 24 h at 37 °C, supplemented with 5% CO2. Be-fore seeding the cells onto the IPLM, the upper aqueous phase (PBS) of theas-fabricated IPLM platform was replaced with the cell culture medium.For seeding cells to the IPLM platform, the lifeact-GFP-expressing MDCKcells were harvested from the culture dish using trypsin-ethylenediaminetetraacetic acid (Wako) and plated onto the IPLM platform at a density of2 × 104 cells cm−2 and incubated at 37 °C with 5% CO2 for ≈1–2 h beforeobservation.Imaging by Confocal Microscopy: CLSM images were obtained usingan inverted microscope (IX-81; Olympus, Shinjuku, Japan) equipped witha disc-scan confocal unit CSU10 (Yokogawa, Tokyo, Japan), an Andor lasercombiner (Oxford Instruments, Oxfordshire, UK), an MD-695 CMOS cam-era, and a 60× water immersion lens (LUMPlanFL N; Olympus) to ob-serve the cells on the IPLM from the bottom of the chamber with a hu-midified 37 °C incubator supplemented with 5% CO2. The observationcould be obtained through two channels: the red channel is the IPLMwith a fluorescence tag and the green channel is actin cytoskeleton ofMDCK cells. The stacked 3D images of a cell together with the IPLM werebuilt by reconstructing the z-axis sequenced scans at the desired thicknessstep (0.5 or 2 μm). By employing orthogonal analysis with the Fiji soft-ware, the top views and side views could be directly read from the stackedimages.Cell and IPLM Morphological and Invasion Depth Analysis: The invasiondepth/IPLM deformation and cellular morphology change were analyzedusing the Fiji software. As illustrated in Scheme 1, by merging the confo-cal images of the green (cell) and red (IPLM) channels, the stacked 3Dstructures for both the cell and IPLM were reconstructed at a 0.5 μm thick-ness of each cross-sectional image. The side view was obtained throughorthogonal processing. Then, the cellular dimensions and invasion depthcould be directly read using the side-view image of the stacked 3Dstructures.Statistical Analysis: Statistical analyses for the two groups were per-formed using a Student’s t-test for the technical replicates indicated ineach figure. All error bars represent standard deviations. Statistical analy-ses were performed using Microsoft EXCEL.Supporting InformationSupporting Information is available from the Wiley Online Library or fromthe author.AcknowledgementsThis study was supported in part by the Japan Society for the Promotion ofScience, KAKENHI (22H00596 and 23K17481 to J.N.: 23KJ2167 to H.W.).Conflict of InterestThe authors declare no conflict of interest.Data Availability StatementThe data that support the findings of this study are available from thecorresponding author upon reasonable request.Keywordsadaptive wetting, adhesion, biphasic, crosslinking, extracellular matrixReceived: August 9, 2024Revised: December 30, 2024Published online:[1] Z. Gong, S. E. Szczesny, S. R. Caliari, E. E. Charrier, O. Chaudhuri,X. Cao, Y. Lin, R. L. Mauck, P. A. Janmey, J. A. Burdick, V. B. Shenoy,Proc. Natl. Acad. Sci. U. S. A. 2018, 115, E2686.[2] A. J. Engler, S. Sen, H. L. Sweeney, D. E. Discher, Cell 2006, 126, 677.[3] O. Chaudhuri, J. Cooper-White, P. A. Janmey, D. J. Mooney, V. B.Shenoy, Nature 2020, 584, 535.[4] O. Chaudhuri, Biomater. Sci. 2017, 5, 1480.[5] O. Chaudhuri, L. Gu, D. Klumpers, M. Darnell, S. A. Bencherif, J. C.Weaver, N. Huebsch, H. P. Lee, E. Lippens, G. N. Duda, D. J. Mooney,Nat. Mater. 2016, 15, 326.[6] C. Müller, A. Müller, T. J. S. M. Pompe, Soft Matter 2013, 9, 6207.[7] M. Bennett, M. Cantini, J. Reboud, J. M. Cooper, P. Roca-Cusachs, M.Salmeron-Sanchez, Proc. Natl Acad. Sci. U. S. A. 2018, 115, 1192.[8] M. Cantini, H. Donnelly, M. J. Dalby, M. Salmeron-Sanchez, Adv.Healthcare Mater. 2020, 9, 1901259.[9] B. Sabass, U. S. Schwarz, J. Phys. Condens. Matter 2010, 22, 194112.[10] Y. Bu, L. Li, C. Yang, R. Li, J. Wang, Acta Mech. Solida Sin. 2019, 32,599.[11] Z. Lu, M. Tenjimbayashi, J. Zhou, J. Nakanishi, Adv. Mater. 2024, 36,2403396.[12] S. K. Hu, S. W. Hsiao, H. Y. Mao, Y. M. Chen, Y. Chang, L. Chao, Sci.Technol. Adv. Mater. 2013, 14, 044408.[13] M. Callens, M. Beltrami, E. D’Agostino, H. Pfeiffer, D. Verellen, G.Paradossi, K. Van Den Abeele, Coll. Surf. A Physiochem. Eng. Asp. 2019,568, 371.Adv. Funct. Mater. 2025, 2414534 2414534 (11 of 12) © 2025 The Author(s). Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adfm.202414534 by Cochrane Japan, Wiley Online Library on [16/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.afm-journal.dewww.advancedsciencenews.com www.afm-journal.de[14] C. F. Temprana, E. L. Duarte, M. C. M. Taira, M. T. Lamy, S. del ValleAlonso, Langmuir 2010, 26, 10084.[15] S. Baek, M. Dinh Phan, J. Lee, K. Shin, Polym. J. 2016, 48, 457.[16] A. Puri, H. Jang, A. Yavlovich, M. A. Masood, T. D. Veenstra, C. Luna,H. Aranda-Espinoza, R. Nussinov, R. Blumenthal, Langmuir 2011, 27,15120.[17] S. Alonso-Romanowski, N. S. Chiaramoni, V. S. Lioy, R. A. Gargini, L.I. Viera, M. C. Taira, Chem. Phys. Lipids 2003, 122, 191.[18] J. J. Schaefer, C. B. Fox, J. M. Harris, J. Raman Spectrosc. 2012, 43,351.[19] B. M. Peek, J. H. Callahan, K. Namboodiri, A. Singh, B. P. Gaber,Macromolecules 1994, 27, 292.[20] T. K. Lind, M. Cárdenas, H. P. Wacklin, Langmuir 2014, 30, 7259.[21] T. Okazaki, T. Inaba, Y. Tatsu, R. Tero, T. Urisu, K. Morigaki, Langmuir2009, 25, 345.[22] N. M. Fonseka, B. Liang, K. S. Orosz, I. W. Jones, H. K. Hall Jr,H. S. Christie, C. A. Aspinwall, S. S. Saavedra, Langmuir 2019, 35,12483.[23] K. Okuno, D. Saeki, H. Matsuyama, Biochim. Biophys. Acta -Biomembr. 2020, 1862, 183377.[24] L. Costa, G. Li-Destri, D. Pontoni, O. Konovalov, N. H. Thomson, Adv.Mater. Interfaces 2017, 4, 1700203.[25] G. Espinosa, I. L. López-Montero, F. Monroy, D. Langevin, Proc. NatlAcad. Sci. U. S. A. 2011, 108, 6008.[26] R. Masrat, M. Maswal, O. A. Chat, G. M. Rather, A. A. Dar, ColloidsSurf. A: Physicochem. Eng. Asp. 2016, 489, 113.[27] W. Megone, D. Kong, L. Peng, J. E. Gautrot, J. Colloid. Interface Sci.2021, 594, 650.[28] H. Wang, H. Zhang, R. Tamura, B. Da, S. A. Abdellatef, I. Watanabe, N.Ishida, D. Fujita, N. Hanagata, T. Nakagawa, J. Nakanishi, Sci. Tech-nol. Adv. Mater. 2023, 24, 2265434.[29] W. R. Legant, C. K. Choi, J. S. Miller, L. Shao, L. Gao, E. Betzig, C. S.Chen, Proc. Natl Acad. Sci. U. S. A. 2013, 110, 881,[30] B. A. Nerger, M. J. Siedlik, C. M. Nelson, Cell. Mol. Life Sci. 2017, 74,1819.[31] S. S. Hur, Y. Zhao, Y. S. Li, E. Botvinick, S. Chien, Cell. Mol. Bioeng.2009, 2, 425.[32] L. Yu, Y. Hou, W. Xie, J. L. Cuellar-Camacho, Q. Wei, R. Haag, Adv.Mater. 2020, 32, 2006986.[33] L. E. Jansen, N. P. Birch, J. D. Schiffman, A. J. Crosby, S. R. Peyton, J.Mech. Behav. Biomed. Mater. 2015, 50, 299.[34] D. Kong, W. Megone, K. D. Q. Nguyen, S. Di Cio, M. Ramstedt, J. E.Gautrot, Nano Lett. 2018, 18, 8811946.[35] N. Khuu, S. Kheiri, E. Kumacheva, Trends Chem. 2021, 3, 1002.[36] P. L. Ko, C. K. Wang, H. H. Hsu, T. A. Lee, Y. C. Tung, Acta Biomater.2022, 145, 316.[37] S. Yamamoto, T. Miyama, T. Komoda, M. Sugawara, M. Nonomura,J. Nakanishi, Anal. Sci. 2020, 36, 263.[38] M. Schnoor, E. Vadillo, I. M. Guerrero-Fonseca, Curr. Opin. Physiol.2021, 19, 119.[39] D. F. O’Brien, T. H. Whitesides, R. T. Klingbiel, J. Polym. Sci. Polym.Lett. Ed. 1981, 19, 95.Adv. Funct. Mater. 2025, 2414534 2414534 (12 of 12) © 2025 The Author(s). Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adfm.202414534 by Cochrane Japan, Wiley Online Library on [16/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.afm-journal.de