Fibrin Matrices

Fibrin plays an important role in healing and regeneration in the developed organism, but not in embryonic and fetal development. As a result of this, few of the interesting morphogenetic signals involved in development, whether adhesion factors or growth factors, interact with fibrin in a specific manner. As such, much of the potential regenerative capacity of the body is not put into play in the process of healing and regeneration in fibrin matrices, either as takes place in natural wound healing or when fibrin is used as a therapeutic material. Our research seeks to incorporate developmental signals into fibrin for use in cell biological studies and as a regeneration matrix. Our goals in this regard are two-fold. First, we seek to understand the biology of the developmental signals under study; in Nature, they are active while immobilized in a three-dimensional milieu, whereas in most experimental systems they have been studied in diffusible form in two-dimensional cultures. As such, we apply our materials systems to examine the biology of these signals, and our target in this mechanistic research is a better understanding of the signals involved in angiogenesis. Under study are the growth factors VEGF121, VEGF165, and Ang1, as well as the matrix adhesion factors L1 and ephrin B2. Second, we seek to develop and exploit these matrices toward clinical use in regenerative medicine. A great deal of effort has been placed in understanding the molecular and cell biology of morphogenetic growth factors, but relatively little effort has been directed toward understanding what to do with these factors, e.g. how to integrate these morphogenetic signals with materials systems to arrive at effective therapeutics. Our clinical targets in these efforts are healing of nerve, exploring b-NGF, NT-3, L1 and a-distroglycan, the chronic dermal wound, exploring PDGF-AB, PDGF-BB, VEGF121, VEGF165 and Ang1, and nonunion defects in bone, exploring BMP-2, IGF-1 and PTH1-34.
We have developed a materials system in which the adhesion factors and growth factors mentioned above are immobilized within an elastic milieu, mimicking the manner in which they are presented in Nature. For example, in Nature VEGF165 binds to glycosaminoglycans in the extracellular matrix, and the enzymatic processes associated with cell infiltration cleave the growth factor from this solid phase and liberate it into bioactivity. We have explored two routes by which to incorporate adhesion factors and growth factors into elastic matrices, employing fibrin as a platform. The growth factors, almost all of which bind sulfated glycosaminoglycans, have been immobilized via electrostatic binding to heparin, which itself has been electrostatically bound via a heparin-binding peptide enzymatically incorporated into the fibrin matrix (left), providing an indirect but strong affinity between the factor and the fibrin. Alternatively, mutant forms of the adhesion factors and growth factors have been developed, containing a novel protein domain for enzymatic coupling into the fibrin with an intervening second novel domain for cleavage by the enzymes utilized by the cell during cell infiltration. When mixed with fibrin precursors, these engineered adhesion and growth factors are covalently incorporated into the fibrin matrix by the enzymes involved in coagulation (right). In both cases, the morphogenetic factors remain tethered to the matrix during cell infiltration until locally released under the influence of invading cells, exactly as occurs in Nature.

a2PI1-8 is a substrate for the coagulation transglutaminase factor XIIIa. ATIII121-134 is the heparin binding domain from the protein antithrombin III. Plasmin Site is a substrate for plasmin, and enzyme involved in cell invasion.
When morphogenetic factors are presented in the immobilized state, bound to the fibrin matrix as an analog of the extracellular matrix, the biological activity of the factor can be substantially altered. As an example, when VEGF121 is immobilized in fibrin, endothelial cells grow out from confluent microcarrier beads within fibrin as tiny tubes, as prototypes of capillaries. However, when cultured in fibrin containing bound VEGF121, very large open tubes can be observed, having diameters up to 150 mm (left). Moreover, when the growth factor is immobilized within the ingrowth matrix, very intense and very highly localized angiogenesis can be observed (center), e.g. as compared to that obtained with the free growth factor (right), both in the chick chorioallantoic membrane model.

We have demonstrated that the nature of the biological activity exerted by the angiogenic growth factor can be modulated by the details of its presentation. For example, VEGFs have been demonstrated to induce substantial angiogenesis, but to induce the formation of immature, malformed, and leaky vessels rather than normally functional vessels. We have demonstrated that when VEGF121 is conjugated to fibrin matrices, rather than presented free, that it induces both more angiogenesis and moreover the formation of vessels that are more normal. This is illustrated below, at the level of the light microscope and the electron microscope for fibrin-bound (upper) versus free (lower) VEGF121.

These approaches of bioactive factor immobilization, to mimic the mode of presentation found in Nature, can potentially be employed to render otherwise poorly effective growth factors into effective therapeutics. We have extensively explored the use of bound matrix adhesion molecules and bound growth factors in peripheral nerve regeneration, and our current efforts are directed toward spinal cord repair. Fibrin is penetrated proteolytically invaded by neurons (below), allowing simultaneous intimate contact between the cell surface and the fibrin ingrowth matrix.

This feature permits the incorporation of bound adhesion molecules within the fibrin as morphogenetic signals, and we have explored the use of bound laminin mimetics in peripheral nerve repair. Laminin is known to play a role in neuromorphogenesis, but it is difficult to apply therapeutically because of the complexity of its recombinant expression. We evaluated fibrin cell ingrowth matrices with bound RGD, YIGSR, RNIAEIIKDI and SIKVAV peptides in repair of resected dorsal root nerves in the rat relative to repair with native fibrin. Nerves repaired with the peptide-containing fibrin (below right) were observed to have 84% more myelinated neurons than those repaired with native fibrin (below left).

The preliminary work carried out in peripheral nerve regeneration has set the stage for current work in spinal cord repair. Here, the molecules under study are the cell adhesion ligands L1, specifically the 6th Ig-like domain, where we have formed recombinant fusion proteins for incorporation into fibrin cell ingrowth matrices, and a-dystroglyan.

Collaborative work with the growth factor NT-3 is being carried out with the former graduate student Shelly Sakiyama-Elbert, who is now an assistant professor at Washington University in St. Louis.
As an additional example in which to demonstrate the concept of growth factor immobilization, the bone forming growth factor BMP-2 was bound into fibrin matrices, and the bioactive material system was compared in ability for joint fusion in pancarpal arthrodesis in canine patients suffering from torn tendons and ligaments in the forepaw as a result of a fall; results were compared to those obtained with the clinical standard of care, namely arthrodesis with a bone autograft. Dogs treated with the material/growth factor system healed both statistically faster and to a statistically higher radiographic healing index than those treated with bone autograft, and moreover at a dose that is more than an order of magnitude lower than that used in the currently next most competitive approach.

Analogous work is underway in bone repair with fusion proteins of the potentially bone-forming peptide PTH1-34, and in chronic dermal wound healing with mutant forms of PDGF-BB, which has in its wild-type form previously shown promise in the laboratory, but has led to disappointment in the clinic, as well as with mutant forms of Ang1, which has yet to be extensively investigated. Of high interest to us is the development of approaches by which to enhance the uptake of plasmid DNA from fibrin matrices during cell invasion. Peptide DNA condensing agents are being developed that contain as a part of the design reversible cross-linking sites for stabilization, sites for binding within the fibrin matrix via transglutaminase activity, signals for cellular uptake, and signals for nuclear localization. Work underway considers the example of therapeutic angiogenesis, where the plasmid encodes a mutant form of the transcription factor hif-1a. Hif-1a responds to oxygen concentration, undergoing a structural change under hypoxic conditions that leads to up-regulation of VEGF expression, among other growth factors. The mutant form being considered as an example is less oxygen sensitive, essentially always being in the hypoxic form, leading to transcription under both hypoxic and normoxic states.