The exploration of the potential of specific recombinant fibronectin fragments to support fibroblast migration in a two- dimensional assay yielded interesting and promising results for the development of a Smart™ matrix for wound healing. The data presented in this thesis confirms previous data from our laboratory (Clark et.al., 2002) that only three functional domains of fibronectin (Cell- binding domain, Hep II domain, IIICS variable region) are necessary and sufficient to support maximal fibroblast migration in vitro. Additionally, the data presented here confirms that this migration is independent of the contiguity of the fibronectin fragments. These results also confirm that our recombinant human FN fragments are functional.

It was predictable that the IIICS fragment would have significant migration potential. Although it has been reported (Clark et.al. 2002) that fibroblasts could not migrate on IIICS alone, this small fragment contains two binding sizes for the a4b1 integrin (Mould et.al.). It therefore provided other mechanisms to the cell to communicate and migrate across the adhered domains. Additionally, it was not completely unexpected that the 1-6 (III1-6) fragment did not support any more migration than the negative control of BSA, because none of the three essential binding domains are located in those repeats. Although a heparin binding site has been reported within type III repeat 5 that can interact with the cellular surface (Moyano et.al., 1997), this site has been described as weak and its ability to support fibroblast migration has not been substantiated in the literature.

It was also not surprising that the addition of the hep II domain to the plate significantly enhanced migration. This is because the heparin domain interacts with cell surface integrins (a4b1) as well as with heparin sulfate and chondroitin sulfate glycosaminoglycans that convey information through proteoglycans receptors on the cellular surface. Although the presented data depicts more robust migration ascribed to the heparin domain than does previously reported data from our lab (Clark et.al., 2002), the difference is merely quantitative, not qualitative.

The results of the cell binding domains, however, were unanticipated. Whereas the presented data records a migration on 110K and C fragments of 55-70% that of intact fibronectin, previous work in our lab has been unable to illicit significant migration out of either the 120K fragment (Clark et.al., 2002), nor the recombinant cell binding domain fragment, even at higher coating concentrations (Nigro, 2001). The differences between the presented results and the previous results from this laboratory are most likely due to differences between the assays used. Whereas I have used an agarose droplet migration assay similar to the one previously described (Nigro, 2001), different types of plates were used, as well as different salt and cell concentrations.

To treat chronic ulcers, we envision a Smart™ matrix, composed of a crosslinked HA scaffold decorated with the three biologically active domains. This should allow for a bioresorbable scaffold that should expedite cellular infiltration into the wound bed by protecting the FN conduit from degradation. HA is of choice, not only because it forms a hydrogel when crosslinked, but it also stimulates cell migration in synergy with FN. To this end we have developed a HA migration assay in which fibroblasts migrate from collagen-coated beads into various crosslinked HA constructs.

The hyaluronan crosslinking reaction seems to have met some early successes, but also some failures. Among the first successes seems to be the relative ease at which critical properties of the crosslinked matrix can be optimized. The intramolecular crosslinks can be set tighter or looser depending on the length and temperature of the crosslinking reaction. Degradation studies need to be done to empirically find the optimal crosslinking conditions for the clinical need.

A qualified success may be the integration of fibronectin onto the hyaluronan scaffold through covalent crosslinks. Through Western blotting technique and probing for fibronectin with a monoclonal antibody, it was apparent that there was no free fibronectin left over from the crosslinking reaction. Additionally, this experiment assured us that the crosslinking and dialysis conditions did not degrade the fibronectin. However, the particular crosslinking reaction was not very specific. The crosslinking conditions may have been such that the fibronectin became overly crosslinked. Too strong of a crosslink may provide a danger that the critical domains of fibronectin will become masked or distorted. Whether this occurred, however, cannot be determined by the outcome of these experiments.

A qualified failure was the first hyaluronan migration assays. In those experiments the fibroblasts did not migrate from the collagen-coated cytodex bead into the HAFN matrix, whereas the cells did migrate out into pure HA, as well as into fibrin with FN and HA with loose FN. It may seem that the concept of a crosslinked FN to HA matrix is flawed because the fibroblasts migrated more robustly on pure HA. However, the cells in the HAFN matrix displayed rounded morphology, indicative of cytotoxicity. The cytotoxicity of the HAFN was confirmed though the indirect contact experiment.

There were then two possibilities as to what caused the cytotoxicity. The first one, which was least likely, was that the crosslinking of the FN to the HA backbone somehow evolved a cytotoxic species, perhaps even the fibronectin itself in a denatured form. The second possibility was the different way of dialyzing the HAFN could account for some residual cytotoxic material that was not present in the pure HA, nor in the HA+FN, which was constructed by mixing FN into dialyzed, crosslinked pure HA.

The dialysis protocol, as developed by the inventor of the process, calls for alternating the dialysis medium between water and 50% ethanol solution. This protocol was performed on the pure HA after crosslinking. However, we had wanted to avoid alcohol in the dialysis medium to protect the active sites on the fibronectin from denaturation, hence we altered the protocol and dialyzed in water exclusively. Perhaps not all of the crosslinking byproducts were safely dialyzed with this altered protocol.

However, even after redialysis of the HAFN construct and the cell toxicity test confirmed that there was no more than baseline cytotoxic leachable chemicals in the HAFN, the cells still rounded up on the beads in the HA migration assay. It is thus likely that either there remains in the scaffold cytotoxic species that affect cells through direct contact, or the crosslinking reaction has denatured the fibronectin.

There are some additional experiments that could be taken to find a way to build the ideal Smart™ matrix. To determine whether any bound toxic agents within the HAFN scaffold exist, a direct contact method for assessing cytotoxicity can be employed. Additionally, it is important to explore other crosslinking schemes, including the use of spacers to provide a buffer between the functional domains and the potentially denaturing crosslinkers. In addition, a more specialized crosslinking reaction can be employed, allowing more control over the nature and placement of the crosslinks.

Once there is a working hyaluronan migration assay, where the migration through HAFN equals or exceeds the migration through HA+FN, the fibronectin fragments should be introduced into the HA scaffold. This is important because when the is moved into the proteolytic environment of the in vivo wound, the functional domains will have to be secured to the backbone, so that they do not get clipped off and become awash in the wound fluid.

The two dimensional agarose droplet migration assay has confirmed important principles in the search for a better treatment for wounds. While the three dimensional hyaluronan migration assay did not yield any positive results to support the notion of robust migration into a crosslinked HA scaffold decorated with the three biologically active domains, it was valuable nonetheless in terms of a first step toward that goal. We were able to better grasp more of the issues involved in a complex three dimensional crosslinked matrix and brings us closer to the goal of developing a construct embodying the ideal characteristics of a Smart™ matrix for the treatment of soft tissue wounds.

Introduction

Product Design

Materials

Methods

Recombinant FN Fragment Purification

Cell Culture