Salmon fibrin treatment of spinal cord injury promotes functional recovery and density of serotonergic innervation

Abstract

The neural degeneration caused by spinal cord injury leaves a cavity at the injury site that greatly inhibits repair. One approach to promoting repair is to fill the cavity with a scaffold to limit further damage and encourage regrowth. Injectable materials are advantageous scaffolds because they can be placed as a liquid in the lesion site then form a solid in vivo that precisely matches the contours of the lesion. Fibrin is one type of injectable scaffold, but risk of infection from blood borne pathogens has limited its use. We investigated the potential utility of salmon fibrin as an injectable scaffold to treat spinal cord injury since it lacks mammalian infectious agents and encourages greater neuronal extension in vitro than mammalian fibrin or Matrigel®, another injectable material. Female rats received a T9 dorsal hemisection injury and were treated with either salmon or human fibrin at the time of injury while a third group served as untreated controls. Locomotor function was assessed using the BBB scale, bladder function was analyzed by measuring residual urine, and sensory responses were tested by mechanical stimulation (von Frey hairs). Histological analyses quantified the glial scar, lesion volume, and serotonergic fiber density. Rats that received salmon fibrin exhibited significantly improved recovery of both locomotor and bladder function and a greater density of serotonergic innervation caudal to the lesion site without exacerbation of pain. Rats treated with salmon fibrin also exhibited less autophagia than those treated with human fibrin, potentially pointing to amelioration of sensory dysfunction. Glial scar formation and lesion size did not differ significantly among groups. The pattern and timing of salmon fibrin's effects suggest that it acts on neuronal populations but not by stimulating long tract regeneration. Salmon fibrin clearly has properties distinct from those of mammalian fibrin and is a beneficial injectable scaffold for treatment of spinal cord injury.

Introduction

Spinal cord injury (SCI) severs descending motor and ascending sensory tracts, cutting off the lower regions of the spinal cord from the brain. Secondary stages of CNS injury such as hemorrhage and infiltration of immune cells lead to increasing neuronal losses and functional impairment (Tator and Fehlings, 1991). The injured region of the human spinal cord degenerates leaving a cystic cavity surrounded by a glial (astrocytic) scar that prevents regrowth of axons, thereby prohibiting recovery from the motor and sensory deficits accompanying the injury.

One approach to SCI is to implant a bridging, or scaffold, material in the cystic cavity with the goal of providing a substrate for axonal regrowth and limiting glial scar formation along the lesion edges. A variety of scaffold materials have been tested, including natural polymers, biodegradable and non-biodegradable synthetics, and composite conduits composed of guidance tubes with material fills (Novikova et al., 2003, Straley et al., 2010). Several are associated with improved functional recovery, making it likely that a repair-promoting scaffold will be an important component of a multi-pronged approach to SCI treatment. However, questions remain regarding optimal structure and composition of the scaffold.

Scaffolds with linear, aligned axonal guidance channels have been developed to mimic spinal cord architecture and encourage longitudinal axonal extension through the scaffold. These types of scaffolds must be fabricated prior to implantation in the lesion site in order to generate the linear channels and prevent exposure of spinal cord tissue to the toxic chemicals or processes necessary for their formation. Injectable scaffolds provide an alternative to preformed scaffolds. These materials are placed in the lesion site as a liquid and allowed to polymerize in situ. This approach is therefore restricted to scaffolds for which the polymerization process is biocompatible, preventing the use of some synthetic polymers since toxic byproducts are generated during polymerization. One advantage to injectable scaffolds is complete filling of the lesion site and tight association of the scaffold with the edges of the tissue since the material conforms to the shape of the lesion cavity. Scaffolds that polymerize in situ can also be used to embed transplanted stem cells and growth promoting molecules can be incorporated into the polymerized matrix; thus these scaffolds provide specific benefits in combinatorial therapeutic approaches (Straley et al., 2010, Taylor et al., 2006). Importantly, the mechanical properties, or compliance, of the scaffold should match that of the surrounding nervous system tissue to encourage the greatest degree of regrowth and reduce mechanical stresses on the surrounding tissue (Balgude et al., 2001, Elkin et al., 2007, Flanagan et al., 2002, Oudega et al., 2001, Prange and Margulies, 2002). Injectable, non-toxic materials that match the compliance of spinal cord tissue may provide the best scaffolds for repair.

A variety of biomaterials have been tested as injectable scaffolds, including the biopolymer fibrin. In fact, human fibrin is FDA-approved and commonly used as a surgical glue (Tisseel®). Fibrin formation is non-toxic and occurs during the coagulation cascade when fibrinogen is cleaved by thrombin to form fibrin monomers, which then spontaneously polymerize to form a three-dimensional matrix (Mosesson, 2005). Varying the concentration of thrombin used to induce polymerization controls the rate of fibrin gel formation, which is advantageous for maintaining a liquid state during injection while forming a solid scaffold in vivo. Fibrin has been successfully utilized in repair strategies for a variety of in vivo neuronal injury models (e.g. Patist et al., 2004) (Iwaya et al., 1999, Stokols et al., 2006, Taylor et al., 2006, Tsai et al., 2006, Williams, 1987). In spite of these advantages, mammalian fibrin gels degrade rapidly (Bensaid et al., 2003, Novikova et al., 2003, Sieminski and Gooch, 2004) and may be contaminated with blood-borne pathogens such as HIV, hepatitis C, and prion proteins (Fischer et al., 2000). In animal studies, autologous mammalian fibrinogen may contribute to CNS damage and lack of repair by inhibiting neurite outgrowth (Schachtrup et al., 2007) and activating resident astrocytes and microglia (Adams et al., 2007, Schachtrup et al., 2010).

The limitations of mammalian fibrin have led to the use of fibrin from other species, most notably Atlantic salmon (Uibo et al., 2009). Salmon and mammalian fibrin polymerize similarly and salmon fibrinogen and thrombin are in ready supply from the aquaculture industry (Wang et al., 2000). Salmon fibrin may be safer than mammalian fibrin for use in humans since known salmon viruses are not transmissible to mammals (Wolf, 1988); in part because fish are coldwater animals and most viruses that infect them are inactivated at human body temperatures. Fibrin prepared from salmon proteins encourages greater cell motility and neurite extension than mammalian fibrin (Ju et al., 2007, Sieminski and Gooch, 2004), matches the compliance of CNS tissue (at ~ 3 mg/ml fibrin concentrations) (Ju et al., 2007, Prange and Margulies, 2002), and degrades more slowly than human fibrin (Ju et al., 2007, Laidmae et al., 2006). Salmon fibrin treatment does not cause immune-mediated toxicity (guinea pigs, mice, rabbits) (Michaud et al., 2002) but can induce antibodies to salmon fibrinogen and thrombin in host animals (rat, rabbit, and swine) (Laidmae et al., 2006, Laidmae et al., 2010, Rothwell et al., 2009, Rothwell et al., 2010, Wang et al., 2000). However, these antibodies do not cross-react with the host's proteins, alter the levels of host fibrinogen or thrombin, or create coagulation problems (Laidmae et al., 2006, Laidmae et al., 2010, Rothwell et al., 2009, Rothwell et al., 2010, Uibo et al., 2009).

Salmon fibrin's positive effects on neurite extension, slower degradation kinetics, and lack of potentially infectious agents led us to test whether it would be a beneficial injectable scaffold for the injured CNS. We used a dorsal hemisection model of spinal cord injury commonly utilized to test scaffolds and completed a thorough assessment of animals treated with salmon fibrin, human fibrin (Tisseel®), or left untreated (controls). We show that salmon fibrin-treated animals have greater recovery of locomotor and bladder function and more serotonergic innervation caudal to the lesion site as compared to animals treated with human fibrin or untreated controls. Surprisingly, there was no effect of salmon fibrin on glial scar formation or lesion volume.