Nanomedicine, Volume IIA: Biocompatibility

© 2003 Robert A. Freitas Jr. All Rights Reserved.

Robert A. Freitas Jr., Nanomedicine, Volume IIA: Biocompatibility, Landes Bioscience, Georgetown, TX, 2003 Cell Response to Patterned Surfaces

The microscale surface texture of an implanted nanoorgan may have a significant effect on the behavior of cells in the region of the implant [1491, 6246]. Compared with smooth surfaces, roughened surfaces show improved osseointegration [1560-1562], improved percutaneous implant integration [1563], and reduced fibrous encapsulation with enhanced integration of breast implant materials [1564]. These improvements are due to increased adhesion of connective tissue cells onto roughened surfaces, resulting in closer apposition of tissue to the implant [1565]. Different cells types respond differently to texture. For instance, macrophages, unlike fibroblasts, accumulate preferentially on roughened and hydrophobic surfaces [1566]. Simple surface roughness (e.g., ~0.2 microns vs. 3-4 microns) appears to be one of the most important variables in determining the proliferation, differentiation, protein synthesis, and local factor production in costochondral chondrocytes [1835] and in MG63 osteoblast-like cells [1836].

One important class of surface features is pores, tunnels, and pegs. In one study [1567], porous polymer membranes with pores >0.6 micron and <5-micron fibers or strands were associated with enhanced new vessel growth. Another study found that fibrosarcomas developed with high frequency (up to 50%) around implanted Millipore filters, with tumor incidence increasing with decreasing pore size in the 50- to 450-micron range [1568]. In general, fibrotic and vascular tissue invades pores larger than ~10 microns and the rate of invasion increases with pore size and with the total porosity of a device [1569-1571]. This invasion results in the formation of a capillary network in the developing tissue [1572]. Vascularization of the new adherent tissue may be required to meet its metabolic requirements and to integrate it with the surrounding tissue [1573], although in urologic applications it is useful to have a nonporous luminal implant surface to prevent leakage of urine through the tissue [1572]. In another experiment [1565] involving 0.1- to 3.0-micron diameter pores >~100 nm deep, 0.9-micron and larger pores completely inhibited bovine corneal epithelial tissue outgrowth even when the surface had a chemistry that was adhesive to cells. Pore size rather than pore number density appeared to be the controlling factor [1565]. Migration of cell monolayers and dissociated cells was reduced but was not completely inhibited even on membranes with 3-micron pores, and individual cells could migrate through these largest holes [1565].

As for pegs and pillars, osteoblasts and amniotic cells cultured on polyethylene terephthalate (Dacron) micropatterned with a square array of 15-micron pegs spaced 45 microns apart extruded bridging processes between the pegs [474]. Picha and Drake [1574] used silicone implants with micropillars ~100 microns in diameter and 500 microns in height, and found that this surface texture reduced fibrosis and improved blood vessel proximity around the implants.

The study of the response of cells to steps and grooves began in 1911 when Harrison [1575] described the reactions of cells grown on spider web fibers. In the 1960s, Curtis and Varde [1576] found that cells grown on cylindrical glass fibers would align on the fibers and were very sensitive to curvature. In many cases, cells orient and migrate along fibers or ridges in the surface, a phenomenon that has been called “contact guidance” [1581] (or “topographic reaction” [5726]) originating from the earliest studies on neuronal cell cultures [1577]. It is now known that the behavior of cultured cells on surfaces with edges, grooves, or other features is significantly different from cell behavior on smooth surfaces [1491, 1581, 5725]. Typically a step inhibits the movement of a cell across it [1578, 1583]. Cells possess an internal cytoskeleton (Section and their normal behavior is to avoid movements that bend this cytoskeleton unduly [1580-1583]. One theory holds that contact guidance is caused by mechanical forces on the cells’ filopodia, which induce the cells to reshape their actin filaments to adjust to the substrate topography [1584]. Cells approaching a step tend to withdraw or to proceed along its edge, only rarely crossing the angular surface. For example, fewer than 10% of baby hamster kidney cells will cross a 10-micron-high step [1583]. Different cells react differently to steps, depending on their biological role: white blood cells tend to cross steps more readily, whereas epithelial cells show a marked aversion to sharp angles [1583]. Some cells types (e.g., endothelial cells [5732], fibroblasts [5733], macrophages [5734], and others [5736]) can react to nanoscale roughness and nanotopographies such as steps as shallow as 11-13 nm [5726], and synthetic nanostructured textured surfaces [5738] have been shown to affect cell behavior [5727, 5731]. According to Curtis and Wilkinson [5726], cell reactions to topography are probably due to stretch reactions to the substratum, not to substratum chemistry: “A given cell type reacts in much the same way to the same topography made with different materials; when both chemical patterns and topographic ones are offered to cells, topography tends to have a greater effect than chemical patterns.”

Cells react to grooves in several ways [5725]. They tend to align to the direction of the grooves, they tend to migrate along the grooves, and they tend to elongate more than they would on a flat substrate [1579, 1583]. The degree of alignment and the rate of orientation depend most on the groove depth [600, 1583, 1585] and pitch [1586], and to some extent on the width [1583], with both motile cells and their processes aligning with the grooves [1583]. Human fibroblasts adherent on surfaces with V-shaped grooves exhibit higher levels of fibronectin synthesis and secretion, relative to similar cells grown on smooth surfaces [1587]. Fibroblasts have been observed to orient on grooved surfaces [1588], particularly for texture dimensions of 1-8 microns [1589]. In one series of experiments [1563, 1588], fibroblasts oriented themselves along 3- and 10-micron deep grooves but inserted obliquely into 22-micron deep grooves. Cells cultured on otherwise identical surfaces may vary in their response to grooves much narrower than one cell diameter. BHK (baby hamster kidney) and MDCK (Madin-Darby canine kidney) cells oriented on 100 nm and 300 nm scale grooves in fused quartz, while cerebral neurons did not [600]. Fibroblasts, monocytes and macrophages spread when cultured on silicon oxide with grooves with a 1.2-micron depth and a 0.9-micron pitch, but keratinocytes and neutrophils did not [1586]. Inflammatory cells show little contact guidance compared to fibroblasts [1586].

A primary failure mode of certain implants is “marsupialization” (Section or “expulsion” [1565], due to downgrowth of epithelial tissue along the edge of an implant in the region where the device penetrates an epithelial layer [1590]. Modification of the microtopography of titanium implants can inhibit this downgrowth of skin epithelial tissue. For instance, grooves measuring 10 microns or 19-30 microns [1563] were sufficient to limit epithelial downgrowth and to promote connective tissue integration at the implant surface.

Patterned surfaces with well-defined peaks, valleys, and islands also influence the function of attached cells. For example, PDMS surfaces with 2- to 5-micron topography maximize macrophage spreading [1591]. Similar surfaces with uniformly distributed 4- to 25-micron2 peaks encourage better fibroblast growth than 100-micron2 peaks or 4-, 25-, or 100-micron2 valleys [1591]. In another experiment, micron-scale adhesive islands of self-assembled alkanethiols stamped on gold surfaces confined cell spreading to those islands [1592]. Larger islands (~10,000 micron2) promoted growth of hepatocytes, while smaller islands (~1600 micron2) promoted albumin secretion. Fibroblast cells attach but do not spread on microlithographically-produced ~500-micron2 palladium islands on pHEMA substrate, but attach and spread to the same extent as an unconfined monolayer culture on ~4000-micron2 islands [1593]. Donald Ingber’s group [3965, 6239-6245] has created surfaces with circular and square islands similar in size to a single cell. When the islands are coated with ECM proteins, cells spread out to assume the shape of the island, regardless of whether the island is a circle or a square [4942, 4943]. Round cells extend lamellipodia (variable extensions of the cell membrane, literally “layered feet”) in random directions, but square cells send out extensions primarily from their corners [4942]. Computer simulations of related processes have been attempted [4941].

Surfaces impressed with biological activity gradients have been found useful in cell biology for examining haptotaxis, the directed migration of cells along surfaces with gradients of immobilized factors [1594, 1595].


Last updated on 30 April 2004