Supramolecular crafting of cell adhesion
Introduction
The supramolecular design of artificial extracellular matrices is key to our understanding of cell–matrix interactions, and of great importance for the future of regenerative medicine. The most common target has been the use of a matrix to activate receptor-mediated biological adhesion mechanisms since they are critical to cell survival and function. Biological cell adhesion to the extracellular matrix (ECM) commonly occurs via binding of integrin receptors to specific epitopes present on the surface of ECM proteins such as fibronectin, leading to the formation of focal adhesions (FAs) and related contacts [1], [2]. These adhesions consist of protein assemblies that serve to connect the cell interior to the external environment [3], [4], [5], [6], [7], [8], [9], [10], [11]. Upon binding of an integrin to its ligand, additional integrins, adapter proteins and cytoskeletal components are recruited to the nascent adhesion, forming a structurally defined FA. The cytoplasmic proteins that are recruited are of two major types, namely scaffolding proteins, which provide the mechanical link between integrins and the actin cytoskeleton, and signaling molecules that trigger and transduce adhesion-dependent cues [3], [4], [6], [7], [8], [9], [10], [11], [12], [13], [14]. A common ligand present on several ECM proteins is the peptide sequence arginine-glycine-aspartic acid (RGD) [15], [16], which is sufficient for triggering cell adhesion.
Most previous work on artificial ECMs has focused on the design of polymeric matrices with attributes such as biodegradable backbones [17], [18], ability to form gels [19], [20], [21], [22], or functionalization with bioactive peptides [23], [24] most notably RGD [25], [26]. Massia and Hubbell first showed that RGD ligand density can affect cell adhesion using peptide coated glass surfaces, reporting complete fibroblast spreading on RGD-grafted glass surfaces at a ligand density of 10 fmol/cm2 (440 nm ligand spacing) and FA formation occurring at an average ligand density of 1 fmol/cm2 (140 nm ligand spacing) [27], [28], [29]. In other work, a lateral spacing between epitopes of 50–70 nm has also been shown to be critical for FA formation [30]. Subsequent studies in different cell types have shown that while there is a maximum density above which further enhancement of cell spreading is no longer observed, it is often higher than that observed by Massia and Hubbell (30 fmol/cm2 for myoblasts [31]) [32], [33], [34]. Studies by Griffith et al. and Mooney et al. using polymers grafted with cell adhesion epitopes have shown that at a given ligand density, cell adhesion can be modulated by the spatial arrangement of the ligands [35], [36]. For example, forming arrays of clustered RGD ligands by coupling multiple epitopes to a single polymer chain enhances cell attachment as compared to distributing the same number of RGD ligands by coupling one epitope per polymer chain [35]. These studies indicate that while a sufficient number of adhesion epitopes are required to sustain cell attachment, the nanoscale presentation of those ligands is also important in regulating the attachment of cells to artificial matrices.
The supramolecular details of biological signal display for optimal cell–matrix interactions in vitro and in vivo remain largely unclear. Supramolecular structure in the ECM is directly linked to epitope density and dynamics, and it is relevant in cell–matrix interactions given the importance of receptor clustering in signaling and the mechanical softness of the interface where cells meet their matrix. Previously studied polymeric systems are inherently disordered, which precludes control of parameters such as the local supramolecular structure or the density and spatial orientation of bioactive signals. Self-assembling monolayers, which can present highly ordered ligand arrays, have been studied, [37], [38] but these systems do not mimic the three-dimensional (3D) filamentous environment of extracellular matrices. In ordered monolayers it is also difficult to change signal accessibility without diluting the signal with mixtures of short and long molecules.
In this work we probe the interactions of cells with supramolecular peptide amphiphile (PA) nanofibers in which the display of RGD epitopes can be altered at extremely high ligand density by changing their local dynamics either through the architecture of molecules or dilution of the epitopes. The nanofibers are cylindrical in shape and have a very high aspect ratio and therefore are able to mimic the soft fibrous environment that naturally surround cells (diameters on the order of 6–8 nm and lengths easily approaching microns) [39], [40], [41], [42], [43]. The fibers, as opposed to monolayers, are able to maintain essentially constant epitope densities with different degrees of epitope packing as a result of sterics and therefore, signal accessibility can be examined in three dimensions. This versatility as one varies the sterics of molecular architecture of monomers is a virtue of structures with curvature that are not pinned on a lattice. These nanostructures are formed from molecules containing an aliphatic tail covalently linked to a peptide segment terminated with the bioactive RGD sequence. Therefore every single molecule in the supramolecular assembly displays an epitope, and cells can explore an environment with essentially van der Waals densities of bioactive signals. This approach also presents to cells a well-defined density of epitopes which is difficult to achieve when grafting epitopes to surfaces or substituting polymers. Upon self-assembly from aqueous solution, the cylindrical nanofibers display their bioactive sequences perpendicular to the long axis of the nanostructure, and inter-fiber contacts create a network that functions as an artificial ECM for cells. In the current study, we have utilized PAs with different covalent architectures in the segment that presents the bioactive epitope, allowing us to explore the effects of epitope dynamics through variations in supramolecular packing of PA molecules, while maintaining high epitope density. Dense tight packing of epitopes with little mobility is achieved with PA molecules with linear architecture, and less efficient packing resulting in more mobile epitopes with branched or cyclic architectures (see Fig. 1) [44]. We examine here the behavior of cells on PA nanofibers to determine if cells can sense and respond to differences in epitope dynamics as supramolecular packing is modified.
Section snippets
Preparation of peptide amphiphiles
Details of linear and branched PA synthesis have been reported elsewhere [44], [45].
Cell culture
3T3 fibroblasts were maintained in phenol red-free Dulbecco's modified eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and penicillin/streptomyocin. REF52, B16, and MDA 231 cells were maintained in DMEM media containing phenol red supplemented as for 3T3 fibroblasts.
Cover slip preparation
Ethanol sterilized number 1 glass cover slips were drop cast with 0.01 wt% solutions of PAs, fibronectin, and poly-d-lysine (PDL)
2D cell adhesion and spreading
2D cell adhesion and spreading experiments were carried out using as the substrate supramolecular nanofibers formed by the molecules shown in Fig. 1 cast on glass cover slips. These self-assembling molecules form nanofibers by self-assembly from aqueous solutions. We investigated two types of molecules, linear and branched PAs, with branched ones presenting either one or two RGDS epitopes in linear or cyclic form (See Fig. 2). Cells examined for cell attachment, spreading, and FA formation were
Conclusions
This work has demonstrated the possibility of signaling cells using supramolecular nanofibers displaying cell adhesion epitopes at the extremely high densities associated with van der Waals packing of molecules. Using branched architectures in the molecules that form the nanostructures, greater epitope accessibility to cells at high density can be achieved given their lower packing efficiency and thus additional space for epitope motion. We have observed differential biological cell adhesion
Acknowledgments
S.I. Stupp is supported by the Department of Energy through Grant DOE DE-FG02-00ER45810, and the National Institutes of Health through Grants NIH 1 R01 DE015920 and NIH 5 R01 EB003806.
B. Geiger holds the Erwin Neter Chair in Cell and Tumor Biology and is supported by grants from the Israel Science foundation, the NIH NanoMedicine Center for Mechanical Biology and the Volkswagen Foundation.
References (60)
- et al.
Signalling from adherens-type junctions
Eur J Cell Biol
(2005) - et al.
Biodegradation and biocompatibility of PLA and PLGA microspheres
Adv Drug Del Rev
(1997) - et al.
Hydrogels for tissue engineering: scaffold design variables and applications
Biomaterials
(2003) - et al.
RGD modified polymers: biomaterials for stimulated cell adhesion and beyond
Biomaterials
(2003) - et al.
Polymer networks with grafted cell adhesion peptides for highly biospecific cell adhesive substrates
Anal Biochem
(1994) - et al.
Lateral spacing of integrin ligands influences cell spreading and focal adhesion assembly
Eur J Cell Biol
(2006) - et al.
Modification of surfaces with cell adheion peptides alters extracellular matrix deposition
Biomaterials
(1999) - et al.
Self-assembling peptide amphiphile nanofiber matrices for cell entrapment
Acta Biomater
(2005) - et al.
Influence of stereochemistry of the sequence Arg-Gly-Asp-Xaa on binding specificity in cell adhesion
J Biol Chem
(1987) - et al.
Focal adhesions: a nexus for intracellular signalling and cytoskeletal dynamics
Exp Cell Res
(2000)