Integrins as sensor of substrate’s biochemical composition and surface patterns
FAs play the role of transducers of physico-chemical properties of the ECM. In particular, the formation of FAs depends on the surface density of epitopes and on their distribution (patterns) at nanoscale up to micrometer scale. Various techniques of surface functionalization with or without patterning have allowed precise spacing at the molecular level of adhesion proteins and peptides (colloidal patterning (Malmström et al., 2010), photolithography + grafting from chemistry (Cheng et al., 2013), nanoimprint + self-assembled monolayers (Gallant et al., 2005)). A representative and comprehensive investigation in this field has been conducted by Spatz et al. who used block copolymer/gold nanoparticle nanolithography (Figure 3). Results established that the density of adhesive sites, or of corresponding peptide/proteins must be increased above a threshold to induce cell adhesion (Arnold et al., 2004; Hudalla and Murphy, 2009; Lagunas et al., 2012). Furthermore, this threshold density depends on several additional factors (chemical composition of coatings, adhesive cues, cell line, …). For example, Spatz et al. worked on glass substrates bearing a quasi-hexagonally ordered array of gold nanoparticles (Arnold et al., 2004; Huang et al., 2009) (Figure 3A). Glass was passivated by PLL-g-PEG while cyclic(RGDfK)-thiol was used to functionalize the array of nanoparticles. These authors report a drastic decrease in cell adhesion upon increasing the interparticle distance above 73 nm (Figure 3B). Moreover, as an integrin receptor has a large radius than the one of nano-adhesive particles, only one integrin can bind to each nanoparticle: this enabled authors to study the integrin clustering phenomenon. They showed that the window of inter-receptor distances ranging between 58 and 73 nm is a universal length scale for integrin clustering and activation, since these properties are shared by a variety of cultured cells. Based on comparison between micro-patterned and homogeneous substrates, they argued that the development of stable focal adhesions, their number, and their size as well as the cellular adhesion strength are influenced by the local (length scale ≤ o e i o ete ) rather than the average ligand density (Deeg et al., 2011). Similarly, FA properties and compositions vary above and below a threshold area of patterned RGD-presenting surfaces (Gallant 2005).
Figure 3 – (A) Scheme of biofunctionalized nanopattern to control integrin clustering in cell membranes: Au dots are functionalized by thiol-coupled c(RGDfK) and areas between cell-adhesive Au dots were passivated by PEG. No integrin clustering below 73 nm (i.e. no focal adhesion). (B) Pictures of MC3T3-osteoblasts on substrates of different distances between nanoparticles (28 nm, 58 nm, 73 nm and 85 nm) (Arnold et al., 2004).
Cells sense gradients of adherent molecules, and migrate towards higher concentrations of adhesive cues. Spatz’s team investigated the cell sensitivity to nanoscale variation in adhesion density on their Au-nanoarrays (Arnold et al., 2008). Examination of cell behavior on such surfaces indicated that the weakest gradient to which cells responded was 15 nm.mm-1. Using the estimate of a typical cell length of 60 µm, this implies that cells respond to an inter-particle distance difference of 1 nm between its front and rear parts. Finally, the distance between the surface and the epitope may also matter. The presence of a molecular tether and its flexibility has been reported to modify the time of adhesion. As the length of the tether increases, the time required for cell spreading increases. The penetration depth of integrin-containing filopodia into gel-like surface layers has also been discussed, and epitopes below the surface by significantly more than~ 10 nm are generally considered as inaccessible (Halperin and Kröger, 2012).
Integrins as sensor of substrate’s mechanical properties
The majority of cells within multi-cellular organisms experiences forces that are highly regulated in space and time. Mechanical forces contribute to control cellular fate, including changes in gene expression, proliferation, and differentiation (Vogel and Sheetz, 2006). Cells sense external stiffness or forces through FAs, by intracytosolic association of the cytoskeleton to the membrane adhesion spots, which exerts a pulling force on FAs that triggers or not kinase activation (mechano-transduction signaling pathway) (Ingber, 2003). Cell responses to either local application of a force, or to substrate rigidity have been investigated on various model conditions that are briefly illustrated below.
Effect of mechanical strain on cells
Sniadecki et al. (Sniadecki et al., 2007) designed an original device to apply forces onto living cells. External forces were applied to cells thanks to micro-fabricated magnetic posts (pillars) containing cobalt nanowires interspersed among an array of nonmagnetic PDMS elastomeric posts (Figure 4). In a magnetic field, nanowire-containing posts are bended, which transduce a force to cells bound on their top. Deflection ( ) of nonmagnetic posts are measured to characterize the response to traction forces of the cells. Using this system, local FA growth was observed only at magnetically actuated posts and not at nonmagnetic posts. They also recorded a loss in traction forces upon local force application that was non-uniformly distributed across the cells. These data suggest that cells actively adjust their internal tension to mechanical forces arising from their microenvironment.
Figure 4 – Microfabricated fibronectin-coated PDMS arrays of magnetic and nonmagnetic posts for applying external forces and measuring traction force responses. (E) External force FMag is applied to the adherent cell through magnetic posts that bend under the influence of a magnetic field. Nonmagnetic posts report local traction forces through post deflections . (A) Representative immunofluorescent micrograph of FAs (green), microposts (red), and nucleus (blue) after force actuation. The cell is outlined, and the location of the magnetic post is marked by the asterisk (*). (B) Vector plot of traction forces at each post are shown with white arrows. (C, D) Plot of average FA area for all posts underneath cells (white bars) and average FA area at magnetic posts (blue bars) when cells are subjected to no actuation and single or multiple actuation.
Effect of rigidity on cells
Stiffness of the different parts of the body ranges from ~ 0 Pa (blood) to > 109 Pa (bone). In soft tissues/organs, stiffness varies from 100 Pa to MPa as a consequence of complex reticulation of networks of collagen and elastin fibers.
Engler et al. first showed that adhesion strength of myotubes increased monotonically with substrate stiffness and eventually reached the strongest adhesion on rigid substrates such as glass (Engler et al., 2004). Then, Engler et al. reported the influence of substrate rigidity in stem cell differentiation (Engler et al., 2006) (Figure 5). Mesenchymal stem cells seeded on soft matrices, whose rigidity is similar to that of brain, formed branched shapes. When they were seeded on rigid surfaces, closer to bone matrix rigidity, polygonal shapes were observed. These results suggest that rigid matrices are osteogenic while soft ones are neurogenic.
Figure 5 – Tissue elasticity and differentiation of naive mesenchymal stem cells (MSCs). (A) Solid tissues exhibit a range of stiffness, as quantified by the elastic modulus, E. (B) In vitro polyacrylamide gel system allows for control of E through crosslinking, control of cell adhesion by covalent attachment of collagen-I, and control of thickness. MSCs of a standard expression phenotype are initially small and round but develop increasingly branched, spindle, or polygonal shapes when grown on matrices of increasing elasticity. (E) Paxillin-labeled cells showing diffuse o ta ts o softer gels (11 kPa) and long and thin adhesions on the stiffer gels (34 kPa). (F) F-actin labelling showing similar trend, from diffused on soft gels to progressively organized on stiffer substrates. Scale bar is 20 μm (Engler et al., 2006).
By seeding motile 3T3 fibroblasts on collagen-coated polyacrylamide substrates with a rigidity gradient, Lo et al. showed that cells can sense stiffness gradient and consistently migrates in the direction of increasing stiffness (Lo et al., 2000). This kind of migration is called « durotaxis ». Although its traction mechanism is still unclear, it appears that FAs are considered as local rigidity sensors, allowing discernment of differences in the extracellular matrix rigidity with a high spatial resolution(Plotnikov and Waterman, 2013).
Integrins as sensor of substrate’s geometrical properties
Into their natural niche, cells adopt a shape by anchoring their cytoskeleton on the surrounding 3D environment. Up to now, the use of 2D scaffolds with micro-patterns represent a convenient way to study effects of cell shape, although the third dimensionality is entirely missing. For example, Peng et al. (Peng et al., 2011) investigated the differentiation of mesenchymal stem cells (MSCs) from rats on adhesive patterns of the same adhesive area but differing by their shapes (Figure 6A). The osteogenic and adipogenic differentiations exhibited different trends depending on the cell shape. In this case, they found that extents of both adipogenic and osteogenic differentiations were linearly related to cell perimeter, which reflects the non-roundness or local anisotropy of cells. On Figure 6B, we can see that the mean alkaline phosphatase (ALP) activity per cell, a marker of osteogenic differentiation, increases with the average cell perimeter. The opposite trend is noticed for adipogenic differentiation.
Figure 6 – Differentiation of single stem cells on different micro-adhesive patterns. (A) Optical micrographs of MSCs; Upper panel shows bright-field images; lower one shows the corresponding fluorescent micrographs with cellular nuclei labeled by DAPI. Scale bars: 25 µm. (B) Extent of differentiation (ALP expression stained by Fast Blue shown in A, upper panel) as a function of measured perimeter of single cells on globally isotropic micropatterns (Peng et al., 2011).
A few studies take into account the three dimensionality of ECM, reveling different cell behavior from flat surfaces. Liu et al.(Liu et al., 2015a) studied migration of different kind of cells with or without confinement. On low adhesive substrates, cells don’t migrate without confinement whereas they acquired a fast migration mechanism (« amoeboid migration ») under confinement by imposing a cell thickness of 3 µm.
Micrometer and/or nanometer scale topographies affect different aspects of cell behavior as illustrated here by considering roughness and porosity (Anselme and Bigerelle, 2011, 2014; Bigerelle et al., 2011; Zinger et al., 2004).
Roughness is recognized to influence cell behavior. Moreover, different regimes of cell response to roughness were identified, depending on the amplitude of reliefs. The minimal height threshold at which substrate nanogroove dimensions may influence filopodial guidance and subsequent whole-cell alignment is around 35 nm (Loesberg et al., 2007). Zinger et al. studied osteoblastic cell adhesion and proliferation on titanium surfaces of different roughness(Zinger et al., 2004). They seeded human bone-derived cells (MG63 cells) on almost only flat surfaces differing at the nanoscale roughness, smooth surfaces with micrometer features (disks of 10, 30 and 100 µm), and surfaces combining the two roughness scales (Figure 7). The cells responded to nanoscale roughness by a higher cell thickness and a delayed apparition of the focal contacts. On surfaces with micrometric hemispherical cavities, the MG63 cells preferably adhered and proliferated inside cavities of 30 or 100 µm in diameter, whereas they did not recognize the 10 µm diameter cavities. On surfaces combining micrometer-size cavities and nanometric roughness, cell proliferation exhibited a marked synergistic effect of microscale and nanoscale topographies.
Figure 7 – SEM pictures of MG63 cells cultured on surfaces composed of cavities (30 µm in diameter) with two different nanoscale roughness. (a-d) correspond to pictures taken respectively after 4 h, 24 h, 3 days or 7 days for one nanoscale roughness while (e-h) correspond to one other nanoscale roughness. Scale bar: 15 µm (Zinger et al., 2004).
Porosity is the equivalent of surface roughness in 3D and is known for decades to be critical for cell proliferation in scaffolds. Many reviews and reports have been published on benefits of controlled porosity, pointing in particular sizes of ~ 100 micrometers for tissue reconstitution and preservation of cell viability (Akhmanova et al., 2015; Di Cio and Gautrot, 2016; Ventre et al., 2012). Part of the effect may be due to facilitation of the circulation of nutrient and gases. Micrometer curvatures are now clearly identified as an important signal that cells can interpret. Similarly to cultures on flat surfaces, thinner topographies matter in 3D scaffolds. For instance, TiO2 nanotubes with different diameters were studied to direct cell vitality (Figure 8) (Park et al., 2007). Self-assembled layers of vertically oriented TiO2 nanotubes with defined diameters between 15 and 100 nm can regulate adhesion, spreading, growth, and differentiation of mesenchymal stem cells. The nanotubes with a spacing < 30 nm and a diameter ~ 15 nm provided an effective length scale for accelerated integrin clustering/focal contact formation and strongly enhanced cellular activities compared to smooth TiO2 surfaces. Cell adhesion and spreading were severely impaired on nanotube layers with a tube diameter > 50 nm, resulting in dramatically reduced cellular activity and a high extent of programmed cell death. Thus, on a TiO2 nanotube surface, a lateral spacing geometry with openings of 30 – 50 nm represents a critical borderline for cell fate.
Figure 8 – Focal contact formation, differentiation, and apoptosis of mesenchymal stem cells on 15 and 100 nm nanotubes. (A) Nanoscale spacing directs cell fate: hypothetical model showing the lateral spacing of focal contacts on nanotubes of different diameters. A spacing of 15 nm seems optimal for integrin assembly into focal contacts, thus inducing assembly of actin filaments and signaling to the nucleus. Nanotubes larger than 70 nm diameter do not support focal contact formation and cell signaling, thus leading to apoptosis. (B) Analysis of focal contacts by SEM using immunogold staining with paxillin, and β -integrin. (C) Schematic representation of the influence of porosity on cell fate(Park et al., 2007).
Conclusion on the relevant surface parameters
With the goal to better understand the early events of cell adhesion and downstream consequences on cell behaviors, exploration of the detailed temporal and spatial pattern of FA proteins is nowadays very active. In this context, a generic approach consists in perturbing experimentally specific association or interaction of cells with a polymer layer. This approach is based on molecular tools as well as engineered materials enabling to adjust the cell environment. Dynamic modulation of this environment is emerging. Namely, versatile and accurate tools are desirable to study the role of the presentation of epitopes on biomaterial surfaces, and to investigate the effect of:
1- Surface density and distribution of epitopes at nanometer resolution. Indeed, the formation of stable cell adhesions involves grouping of integrins, which depends on epitope organization (density, random/regular distributions) and mobility (attachment to deformable or fluid surfaces),
2- Nanoreliefs: as it guides cell spreading/migration, and/or commitment in specific lineage. Indeed, cell perceives grooves of > 30 nm depth, a size of the order of the radius of long polymer chains,
3- Micrometer-large reliefs and patterns: both geometry and deepness affect spreading, division, …
4- Rigidity: requiring deposition of epitopes on various substrates whose rigidity ranges from soft elastomers to solid-like materials,
5- Synergetic effects: the simultaneous presentation of several signals are likely key triggers (Ventre et al., 2012).
Direct comparisons between different epitope attachment/presentation systems are challenging. Indeed, changing the surface chemistry introduces alterations in concentration, spacing, substrate rigidity, or affinity of integrins for the epitopes. A close mimic of the plasticity and constant remodeling of ECM is provided by surfaces able to trigger in situ variations of surface properties in contact of living cells. Dynamic switchable systems are thus desirable for fundamental purposes for direct comparison of the effect of external parameters. This objective can be reached by the design of surface-bound functional layers combining excellent anti-biofouling properties and dynamic presentation of specific epitopes. Promising candidates are reviewed in the next paragraph.
Stimuli-responsive biomaterials to modulate in situ the effects of adhesive cues
A few reviews report on specific cases of bioswitchable interfaces, including optically controlled hiding/presentation of epitopes, and thermally responsive layers used for triggered cell detachment/harvesting (Cole et al., 2009; Nakanishi, 2014; Uhlig et al., 2014). Representative and robust approaches developed up to now are detailed in this section. Special attention is accorded on the ones providing good control on specific adhesion and detachment. We introduce the topic with additional indications about the need for absence of non-specific binding, that may not be easy to achieve with some technologies.
The study of cell responses to specific immobilized biomolecules at solid–liquid interfaces needs an antibiofouling background. Indeed, once immersed into biological media, biomolecules (proteins, …) spontaneously adsorb on non-passivated surfaces, causing non-specific cell adhesion or responses.
Chemically, nonadhesive molecules can be grouped as carbohydrates (agarose(Nelson et al., 2003), mannitol (Luk et al., 2000)), synthetic polymers (poly-(ethylene glycol), polyacrylamide), or proteins (albumin (Nelson et al., 2003)). Physically, nonadhesive regions are either monolayers, swollen hydrogels, lipid bilayers, or polyelectrolyte multilayers. They all appear to resist adhesion of cells by preventing nonspecific protein adsorption. The functional lifespan of these nonadhesive materials varies widely in the literature (24 h to 60 days).
Supported phospholipid bilayers constitute an effective class of cell/protein repellent adlayers(Andersson et al., 2003). To explain the passivation of such adlayers, a cooperativity between the following factors has been mentioned: polarizability, electrical neutrality (when using zwitterionic lipids), strongly bound/hydration water, and fluidity. Surface-immobilized lipid bilayers suffer from one major drawback they cannot be dried and must therefore always be stored in aqueous solution.
Polyelectrolyte multilayers (deposited by the layer-by-layer technique) have been reported as cell resistant materials under specific pH conditions(Berg et al., 2004). Weak polyelectrolytes offer the ad a tage of ei g a le to ha ge the fil ’s ole ula a hite tu e simply adjusting the assembly pH. Using this approach, it was found that highly swellable films prohibit cell adhesion while tight ionically stitched films support cell attachment. They are now widely studied as substrates for tissues engineering (Boudou et al., 2010; Kerdjoudj et al., 2011; Lee et al., 2011; Monge et al., 2013, 2015).
Poly-(ethylene glycol) (PEG) with the monomeric repeat unit [-CH2-CH2-O-] is the most widely used system to passivate surfaces. Many different strategies have been successfully designed to immobilize PEG on substrates relying on « grafting to » methods.
Chemisorption : On solid surfaces such as glass or silicon, PEG can be grafted covalently thanks to the silane coupling reaction (Blümmel et al., 2007). Silanes are nowadays less popular because of their hygroscopic behavior, their tendency to polymerize and form island-like domains leading to heterogeneities. Another widely used method relies on PEG-modified alkanethiolates to form self-assembled monolayers (SAMs) on metal surfaces (mainly on gold). Due to their tendency to oxidation under ambient conditions, Bearinger et al. reported an alternative method for gold passivation based on PEG-PPS-PEG triblock copolymers with poly(propylene sulfide) (PPS) as the central anchoring block (Bearinger et al., 2003).
Physisorption: Poloxamers or Pluronics® have been extensively used to modify hydrophobic surfaces. These polymers are triblock co-polymers PEG-PPO-PEG with a hydrophilic part (PEG) and a hydrophobic part (PPO polypropylene glycol) which are able to spontaneously assemble on hydrophobic surfaces via hydrophobic-hydrophobic interactions. The adsorbed layer of copolymer is thought to form a comblike structure at the surface, with hydrophobic PPO bound to the surface, while hydrophilic PEG extremities are exposed to the surrounded solution. These polymers must be used for short-term experiments because they are prone to detach from the surface(Nelson et al., 2003). Poly(L-Lysine)-g-Poly(ethyleneglycol) is similarly used to passivate anionic surfaces (glass, metal oxides, tissue culture polystyrene), affording a stable adlayer (Huang et al., 2001; Kenausis et al., 2000). The better layer stability stems from electrostatic interactions between the polycationic PLL backbone and the anionic surface under appropriate pH and ionic strengths.
Among the numerous model proposed to explain the antibiofouling properties of PEG coatings, steric stabilization and excluded-volume effects are the most commonly cited (Israels et al., 1995; Jeon et al., 1991). Hence, Sofia et al.(Sofia et al., 1998) shown that non-overlapping linear PEG brushes was not efficient to significantly prevent protein adsorption. By increasing the density to obtain overlapping PEG brushes, PEGylated surfaces become protein-repellent. The degree of overlap can be evaluated by the expression , where L and Rg are respectively the distance between two PEG chains and the radius of gyration of the PEG side chains. Considering this expression, degrees of overlap inferior to unity imply that the neighboring PEG chains overlap and that the surface is densely packed. Conversely, values superior to unity correspond to non-overlapping PEG chains. According to the results of Sofia et al., this degree of overlap appeared to be independent of PEG molecular weight and protein size. Of course, short PEG chains need a grafting density higher than long PEG chains to guarantee a dense surface with efficient PEG packing. Kenausis et al. gather data from different surface-immobilized PEG systems (Figure 9), including PLL-g-PEG on metal oxides, SAMs of PEG-alkanethiolates on gold and PEG covalently linked to silanized silica. On Figure 9, we can see that surfaces with a low decrease considerably the quantity of adsorbed proteins, thus confirming that systems with the lowest degree of overlap are the most protein-repulsive. Moreover, this behavior seems to be independent from the surface-immobilized PEG system. In the case of PLL-g-PEG, optimizations based on the polymer architecture (Huang et al., 2001) shown that PLL-g[3.5]-PEG(2000) was strongly effective to passivate anionic surfaces.
Figure 9 – dependence of serum adsorption over different surface-immobilized PEG systems with the degree of overlap (Kenausis et al., 2000).
For physically stable PEGylated surfaces (i.e. surfaces obtained from chemisorption or passivated with PLL-g-PEG), the functional lifetime is generally less than ten days and strongly depends on cell line (Nelson et al., 2003). Indeed, PEG chains are prone to degradation in water via oxidation mechanism (Pidhatika et al., 2012). Moreover, specific cell lines (3T3-L1 pre-adipocytes) express enzymes able to catalyze this oxidation (3T3-L1 pre-adipocytes), and consequently accelerate the decrease in repellency of the PEG layer. Further information on PLL-g-PEG is given on the last part of this chapter (see section 3.1).
The choice of the passivation chemistry is dictated by different factors:
the type of substrate (glass, silicium, metal oxides, gold, …)
the complexity of the chemistry involved in the layer formation the functional lifetime
the compatibility with patterning techniques.
Basically, chemical-sensitive materials respond to changes in pH, ionic strength, introduction of co-solvents or solutes. Every chemical stimulus isn’t however, by far, compatible with manipulation of living cells. Variation of pH, ionic strength and co-solvent can’t be used in general because mammalian cells need a specific culture medium. Introduction of non-toxic solutes that can spontaneously adsorb, or bind by bio-orthogonal chemistry to the surface is the only biocompatible chemical trigger. These solutes can covalently react or non-covalently interact with surfaces. When they are used as a trigger (in situ introduction and not pre-stamping), the whole surface is exposed to solutes. A local control is thus achieved only on pre-patterned surfaces. Representative examples are listed below with a focus on systems allowing to control sequentially adhesion/detachment of cells, and revealing a high degree of control of specific cell-substrate interactions.
Covalent surface modifications
Here, the idea consists in adding to the culture medium a reactant able to modulate a ligand accessibility on the surface. The reaction must be rapid, quantitative and biocompatible (i.e. involving non-toxic reagents, and no side reaction with biological components such as proteins). These requirements limit drastically the choice of reactions.
Azide-containing surfaces fulfill these conditions. Indeed, azides can undergo a series of chemoselective reactions such as the Curtius rearrangement, Staudinger ligations, and Huisgen 1,3-dipolar cycloadditions. Azide-based cycloaddition reactions have recently received significant attention and the strain-promoted azide-alkyne cycloaddition is a copper-free non-toxic reaction. It employs a strained cyclooctyne to react with an azide in a spontaneous, catalyst-free cycloaddition. Van Dongen et al. used this reaction to introduce an adhesive peptide onto an cell-repulsive azide-functionalized surface (van Dongen et al., 2012; van Dongen et al., 2013). Via this reaction, the stimulus isn’t reversible.
Another solution consists in using enzymes to modify artificial surfaces as it is the case in ECM remodeling. Todd et al. (Todd et al., 2009) reported the use of elastase, a protease associated with tissue remodeling, switching the surface from bioinert to adhesive (Figure 10). They worked on a poly(ethyleneglycol) monolayer coupled to glass substrate. On the top of the layer, a RGD cue is protected by a cleavable steric group (Fmoc-alanine-alanine-RGD-PEG). Before hydrolysis of the specific alanine-alanine sequence, RGD isn’t accessible while after cleavage cells can recognize and interact with RGD via their integrins. This stimulus isn’t reversible.
Figure 10 – (A) Schematic representation of the enzyme responsive system. (B) Structural representation of the system.
Each surface bound group consists of the Fmoc lo ki g group, a i o a ids whose ide tity depe ds o the ature of the side groups (R), and a PEG linker attached to glass via an epoxy silane. (Todd et al., 2009)
Non-covalent surface modifications
Non-covalent methods, based on reversible interactions between a bioinert surface and adhesive factors, represent a promising approach to realize dynamic cell adhesion. Current attempts rely on host-guest supermolecular chemistry, electrostatic interactions, complexation, and hydrogen bonding.
Boekhoven et al. reported a method based on host-guest supermolecular chemistry (Boekhoven et al., 2013). They worked on an alginate matrix covalently functionalized with β-cyclodextrin (host), which is cell repellent. Addition of guest molecules containing a RGD peptide to culture medium induced focal adhesion formation and spreading of 3T3 fibroblasts due to the host-guest interaction. The spreading was reversed by adding competitive guest molecules of higher affinity for β-cyclodextrin and bearing the mutated non-bioactive sequence RGE.
The major drawback of this method is related to the weak non-covalent link between the surface and adhesive cues which can be broken by adherent cells. Indeed, cells exert important forces via integrin on the adhesive cues (Jurchenko et al., 2014), leading to the dissociation of streptavidin-biotin tethered ligands in focal adhesions. Pan et al. (Pan et al., 2014) suggest that multivalency non-covalent interactions between an adhesive cue and surface could make it possible to maintain the stability of the assembly. To a layer of poly(hydroxyethyl methacrylate)-graft-(phenylboronic acid) (PHEMA-graft-PBA) polymer brushes, they exposed a solution of RGD-poly(3-gluconamidopropyl methacrylamide) (RGD-PGAPMA), forming a stable polymeric complex via PBA/cis-diol interactions (Figure 11). Stable focal adhesions give a proof of the stability of the polymeric assembly. The introduction of high concentrations in fructose makes it possible to destroy this polymeric assembly. This system shows a reversible switching between adhesive/repulsive state simply by changing the fructose concentration of the culture medium. Cell detachment has been studied after 3 h of incubation in DMEM (i.e. allowing cell spreading) by adding fructose (final concentration: 60 mM). As clearly shown in Figure 11c, gradual transition of the cell morphology from a spread-out shape to a round shape was observed. About 64% of the cells were released in fructose-containing DMEM after 30 min of incubation. No longer incubations in this medium were reported to investigate a quantitative cell release.
Table of contents :
Chapter 1 – Introduction
1 Integrin binding to extracellular matrix, a model for adhesion on biomaterials
2 Stimuli-responsive biomaterials to modulate in situ the effects of adhesive cues
3 Coating based on PLL-g-P(NIPAM-co-RGD)
Chapter 2 – Synthetic route to PLL-g-PNIPAM derivatives
1 PLL functionalization
2 Synthesis of -Ci,-NHS-P(NIPAM-co-B) grafts
Chapter 3 – Characterizations of surfaces coated with PLL-g-PNIPAM derivatives
1 Adsorption of PLL-g-PNIPAM on flat SiO2 surfaces
2 Thermal transition of PLL-g-PNIPAM coatings
Conclusions and discussion on thermal transition of PLL-g-PNIPAM adlayers
Chapter 4 – Temperature-controlled ligand accessibility in PLL-g-PNIPAM coatings
1 Effect of ligand position in PLL-g-PNIPAM
2 Effect of ligand density
3 Effect of the size of Streptavidin probes
Chapter 5 – Cell adhesion on PLL-g-PNIPAM coatings
1 Pure (100 %) PLL-g-[0.15](NIPAM-co-RGD)(6) adlayers
2 Cell adhesion on mixed PLL-g-[0.2]PNIPAM(6)/PLL-g-[0.15]P(NIPAM-co-RGD)(6) coatings at 27
and 37 °C
3 Cell detachment from mixed PLL-g-[0.2]PNIPAM(6)/PLL-g-[0.15]P(NIPAM-co-RGD)(6) coatings
4 An example of dynamic experiment on mixed thermo-responsive PLL-g-[0.15]P(NIPAM-co-RGD)(6)/PLL-g-[0.2]PNIPAM(6) coatings
Chapter 6 – Studies of poly(N-isopropylacrylamide) polymers functionalized with azobenzene to develop light-responsive brushes
1 State of the art on thermo- and photo-responsive polymers
2 Our strategy to make our PNIPAM grafts sensitive to light
3 Studies of PNIPAM copolymers with azobenzene pendant groups
4 Studies of PNIPAM copolymers with azobenzene at one extremity
5 Use of azobenzene-containing PNIPAM as light-responsive polymer coatings
Conclusion & perspectives
Appendix 1 – Protocols for synthesis
1 Materials and methods
2 Synthesis of HOOC-Azo-NH2
3 Synthesis of -Ci,-NHS-P(NIPAM-co-B) and -Ci,-COOH-P(NIPAM-co-B)
4 Grafting of NHS-compounds on PLL
Appendix 2 – Coating and patterning with PLL derivatives
1 Coating of coverslips with PLL derivatives
2 Photolithography to obtain micrometric patterns of PLL derivatives
Appendix 3 – PNIPAM phase transition
1 PNIPAM phase transition in aqueous solution
2 Phase transition of PLL-g-PNIPAM adlayers
Appendix 4 – Captures of fluorescent probes on PLL-g-PNIPAM patterns
1 Generic protocol used for captures of fluorescent probes
2 Adsorption of Streptavidin-Cy3 on mixed PLL-g-[0.1]P(NIPAM-co-biotin)(6):PLL-g-PEG adlayers
3 Non-specific protein adsorption on PLL-g-[0.2]PNIPAM(6)
Appendix 5 – Cell experiments
1 Cell seeding
2 Qualitative experiments of cell adhesion
3 Quantification of cell adhesion
4 Cell detachment
5 Reversible temperature-triggered change in cell shape
Appendix 6 – Thermal relaxation