Olfactory ensheathing glia: Repairing injury to the mammalian visual system

Abstract

The visual system is widely used as a model in which to study neurotrauma of the central nervous system and to assess the effects of experimental therapies. Adult mammalian retinal ganglion cell axons do not normally regenerate their axons for long distances following injury. Trauma to the visual system, particularly damage to the optic nerve or central visual tracts, causes loss of electrical communication between the retina and visual processing areas in the brain. After optic nerve crush or transection, axons degenerate and retinal ganglion cells (RGCs) are lost over a period of days. To promote and maintain axonal growth and connectivity, strategies must be developed to limit RGC death and provide regenerating axons with permissive substrates and a sustainable growth milieu that will ultimately provide long term visual function. This review explores the role olfactory glia can play in this repair. We describe the isolation of these cells from the olfactory system, transplantation to the brain, gene therapy and the possible benefits that these cells may have over other cellular therapies to initiate repair, in particular the stimulation of axonal regeneration in visual pathways. This article is part of a Special Issue entitled: Understanding olfactory ensheathing glia and their prospect for nervous system repair.

Introduction

Over the last two decades, olfactory ensheathing glia (OEG) have been identified as a potential transplantation strategy in central nervous system (CNS) disease and after injury. This is due to their role in the mature olfactory system where OEG guide constantly renewing olfactory axons from the basal cells in the olfactory neuroepithelium through the cribriform plate to reach their glomerular targets deep in the olfactory bulb (Barber & Raisman, 1978, Graziadei et al., 1979). This makes OEG unique as they exist both within peripheral neural tissue and the CNS. As a consequence, OEG can be isolated that are either peripherally or centrally derived (Barnett et al., 1993). Studies on these two types of OEG have shown that they share many similar features and characteristics; however there are also significant differences including the ability to migrate and create a permissive regeneration-promoting environment at a lesion site (Richter et al., 2005).

The purity of OEG obtained for transplantation also affects the capability of the cells in a cell-transplantation regime. There are various methods used to purify OEG including differential adhesion method (Wu et al., 2010), immunopanning (Ramón-Cueto et al., 1993, Ramón-Cueto et al., 1998, Plant et al., 2002), fluorescence activated cell sorting (FACS) and surface marker antibodies (Barnett et al., 1993). Immortalized cell lines have also been studied and provide a useful tool for single cloned sources of human or rodent OEG (Moreno-Flores et al., 2006, García-Escudero et al., 2010). Production of some lines requires the introduction of the catalytic subunit of telomerase (TERT) (Lim et al., 2010, Llamusí et al., 2010) which may complicate interpretation over and above those from purified or FACS sorted cell populations, especially when two viral transductions are required.

In experimental studies on repair after spinal cord injury, various OEG populations have been reported to induce axonal regeneration and/or improve functional recovery. Grafted cells were either purified OEG, non-purified cell preparations obtained from the olfactory bulb (Ramón-Cueto et al., 1993, Ramón-Cueto et al., 1998, Ramón-Cueto et al., 2000, Li et al., 1997, Li et al., 2003a, Plant et al., 2003) or nasal olfactory tissue implanted directly after removal (Lu et al., 2001). Other studies using a mix of meningeal cells and OEG have shown improved remyelination after transplantation into an x-irradiated/ethidium bromide model compared to a purer population of OEG obtained using cytosine arabinoside treatment to reduce the level of fibroblastic contamination (Lakatos et al., 2003). Contrasting results have also been found after dorsal root lesion using OEG grafts as a repair strategy (Ramón-Cueto & Nieto-Sampedro, 1994, Li et al., 2004, Riddell et al., 2004, Ramer et al., 2004) that may be attributed to the method of purification or the age of the animals from which the cell grafts were prepared (Barnett and Riddell, 2004).

In the brain and spinal cord many neurological diseases such as Alzheimer's, Parkinson's and amyotrophic lateral sclerosis (ALS) are a result of chronic neurodegenerative changes that result in loss of specific neuronal populations. Thus major therapeutic targets include neuroprotection, and replacing cells and restoring compromised circuitries in the brain by enhancing neurogenesis. Neurotrophic factors have been identified as having positive effects in promoting neurogenesis in the brain, particularly in the hippocampal region. These factors include brain derived neurotrophic factor (BDNF; Lee et al., 2002), basic fibroblast growth factor (bFGF, Palmer et al., 1999), and nerve growth factor (NGF, Frielingsdorf et al., 2007). Due to the different neurotrophic factors that they can secrete, OEG could be a key cell in attempting to not only support the survival and extension of remaining axons, but also in promoting neurogenesis. In culture OEG promote the survival and outgrowth of hippocampal neurons that is comparable to the growth of neurons exposed to different growth factors (Pellitteri et al., 2009). Conditioned media obtained from OEG cultures and used to maintain hippocampal neurons also resulted in improved survival and outgrowth of these neurons (Pellitteri et al., 2009) indicating that there is a secretion mechanism underlying the ability of OEG to promote neuronal growth and survival.

Transplantation studies of OEG into the brain have utilized different lesion models. Smale et al. (1996) isolated rat embryonic OEG and transplanted the cells in the form of cell/collagen matrix graft into a unilateral lesion of the fimbria–fornix pathway. The studies showed that OEG survived in the graft up to 4 weeks and promoted growth of axons in the septal–hippocampal pathway (Smale et al., 1996). Further evidence showing the integration of OEG in the brain was demonstrated with the transplantation of OEG into the normal thalamus (Pérez-Bouza et al., 1998). The study showed that transplanted OEG can span natural barriers within the brain as well as induce axonal growth between independent CNS structures (Pérez-Bouza et al., 1998). In addition, transplanted OEG may determine the direction of host axon growth (Pérez-Bouza et al., 1998), a characteristic that has also been described in culture (Sonigra et al., 1999). Other studies also show that transplantation of OEG can be beneficial in neurodegenerative diseases. In rats, a model for studying Parkinson's disease is by performing a (6-OHDA)-lesion in the nigrostriatal dopaminergic pathway. Transplantation of unpurified OEG obtained from the glomerular layers of the adult olfactory bulb with or without fetal ventral mesencephalic cells (VMC) into the model showed an increase in functional recovery (Agrawal et al., 2004). Although OEG transplantation alone showed positive results in the injury model, co-transplantation with VMC showed significant functional restoration compared to individual transplant groups (Agrawal et al., 2004) indicating that a multifunctional approach could be important in the repair of the injured brain.

OEG have been used as a cellular choice in many models of spinal cord injury (for review see Raisman, 2001). In such models, central ascending and descending pathways are subjected to a trauma that leads to axonal injury and often axotomy, in which the axons are completely severed. Olfactory glia have been transplanted into the spinal cord either alone (Li et al., 1997, Ramón-Cueto et al., 2000, Ruitenberg et al., 2002) or in combination with other cell types such as Schwann cells (Ramón-Cueto et al., 1998, Takami et al., 2002, Barakat et al., 2005). These studies have reported varying levels of success including an increase in tissue sparing, axonal sparing/regeneration and functional behavior. Interesting differences do exist between the spinal cord and brain; regenerating axons are able to leave OEG transplant zones within the injured spinal cord and can re-enter the host spinal tracts. However, within the visual system OEG show no evidence of advancing with regenerating RGC axons and co-extending into the astrocytic territory of the distal stump of the injured optic nerve (Li et al., 2003b). Without showing a true correlation of behavior to anatomical outcomes, it is difficult to ascertain the real mechanism of repair; however there is evidence showing the positive uses of olfactory glia in the injured nervous system and particularly the spinal cord. A more in-depth review of spinal cord transplantation studies using olfactory glia can be found in the other reviews within this special edition.

We have further explored the differences in OEG cell properties when animals of different ages are used to prepare the primary cultures. Rat age groups for cell isolation can be divided into three broad categories — embryonic, postnatal and adult. Embryonic ages are E18–E19, the postnatal age range is between day 1 (P1) after birth up to day 21 (P21) and the adult group can comprise rats aged anywhere from 3 weeks to 6 months. A number of previous studies have used animals of different ages to obtain primary OEG cultures, but there has been a lack of direct comparison between the cell types. While embryonic and postnatal cells are unlikely to be considered in translational studies due to ethical reasons, it is still important to understand the mechanisms underlying their functions in order to gain a better overall understanding of OEG biology. Furthermore, while the technique used to isolate and prepare OEG may not solely explain contrasting results reported in the literature, it remains an important issue that should be taken into consideration in all OEG studies.

Our laboratory has conducted a study on the age-dependent myelination by OEG. Comparison between p75 immunopanned purified embryonic, postnatal and adult OEG showed that only embryonic OEG are capable in vitro of myelinating TrkA-dependent DRG neurons under culture conditions containing serum, ascorbate and progesterone (de Mello et al., unpublished data). In vivo studies comparing embryonic to adult OEG myelinating capacity using a lysolecithin-induced demyelination model in rats also showed differing results. While both cell groups possessed a similar proportion of intact myelin compared to a Schwann cell control group, the proportion of axons surrounded by loose uncompacted myelin was significantly less in embryonic OEG groups, and the proportion of axons that were completely unmyelinated was correspondingly higher (de Mello, unpublished data).

Our observations indicate a developmental shift in the properties of OEG populations, and suggest that age-related differences may yet be found in the bulb in vivo despite other findings (Magavi et al., 2005). This indicates that comparison studies between age groups could yet be used to gain a better understanding of the function of OEG with the aim of promoting regeneration and functional repair of the damaged CNS. There remains more to be elucidated about OEG biology; it has potential therapeutic uses because of the positive pre-clinical results obtained thus far in cellular transplantation studies in animal models, and also because OEG are easily obtained for autologous transplantation in the clinical setting.

The mammalian visual system is a useful model in which to examine the effects of transplanted OEG on the survival and regrowth of CNS tissue. Injury and degeneration in the visual system can be induced by direct trauma or ischemia, but also occurs as a consequence of more chronic ophthalmic diseases such as glaucoma; all can result in the death of a significant proportion of retinal ganglion cells (RGC). It is well-established that after an optic nerve crush or transection, death of these neurons occurs on a wide scale, in part due to trophic factor deprivation. Promoting RGC survival and then inducing the regeneration of their damaged axons are both required in order to repair the visual system after an injury. Just keeping the neurons alive is, in itself, not sufficient to promote axonal regrowth (Goldberg et al., 2002, Cui et al., 2003, Leaver et al., 2006a); it is necessary to induce RGCs to re-acquire a growth state that seems to be lost or otherwise considerably diminished after the embryonic stage (Fischer et al., 2004, Sun & He, 2010). In many ways, the issues and problems associated with inducing the damaged optic nerve to regenerate to fully functional capability are very similar to those that pertain when attempting to regenerate the injured spinal cord. The inhibitory nature of the lesion site needs to be overcome, and permissive growth substrates need to be available in order to bridge the injury site and induce the regrowth of axons. Should sufficient regrowth be achieved it is also important to ensure that remyelination occurs, thus increasing salutatory conduction speeds and aiding in the return of function.

Similar to the rest of the CNS, it has been suggested that appropriate therapeutic manipulation of neurotrophic signaling pathways will promote the survival of injured RGCs (for reviews see Chierzi & Fawcett, 2001, Zhi et al., 2005, Harvey et al., 2006, Johnson et al., 2009). Among the neurotrophic factors known to promote survival of RGCs include BDNF (Rohrer et al., 2001, Leaver et al., 2006b, Peinado-Ramón et al., 1996), neurotrophin 4/5 (Cui et al., 2003, Peinado-Ramón et al., 1996), glial cell-derived neurotrophic factor (GDNF; Jiang et al., 2007) and ciliary neurotrophic factor (CNTF; Cui et al., 2003, Ji et al., 2004, Parrilla-Reverter et al., 2009). Numerous laboratories have tried to improve the efficacy of neurotrophic treatments to the injured optic nerve with varying success (Mey & Thanos, 1993, Chierzi & Fawcett, 2001, Harvey et al., 2006, Berry et al., 2008).

OEG are known to secrete multiple neurotrophic factors such as BDNF and CNTF that could prove beneficial in promoting the survival of RGCs (Lipson et al., 2003) and the cells can be engineered to secrete biologically active neurotrophins in vivo (Ruitenberg et al., 2003, Ruitenberg et al., 2005, Cao et al., 2004). In addition, studies of Schwann cell transplantation into the injured visual system have shown that: 1) there is poor integration of the Schwann cells with host tissue and 2) the number of axons that grow out of the Schwann cell-rich environment into CNS neuropil is limited (Zwimpfer et al., 1992, Harvey et al., 1995, Tan & Harvey, 1999: Symons et al., 2001, Vidal-Sanz et al., 2002, Whiteley et al., 1998, Avilés-Trigueros et al., 2000). This is mainly due to the relative attractiveness of the Schwann cell versus the adult CNS environment, as well as the negative interactions between Schwann cells and CNS glia that can enhance the inhibitory nature of the injury site (Plant et al., 2001, Lakatos et al., 2003) OEG however, are known to be able to overcome this latter obstacle and intermix well with glia in the CNS neuropil (Vukovic et al., 2007). Additionally they do not appear to exacerbate the scar tissue that forms after injury (Ramón-Cueto et al., 1998). In this regard, OEG have been shown to induce less chondroitin sulfate proteoglycan (CSPG) expression compared to Schwann cells (Lakatos et al., 2003). CSPG is found in the scar tissue and contributes to the non-permissive properties of the CNS environment but can be removed or reduced by the use of chondroitinase ABC treatment and if combined with BDNF injections results in significant sprouting of retinal afferents into the denervated superior colliculus (Tropea et al., 2003). One known factor produced from olfactory glia that would reduce the effects of scar inhibition is metalloproteinase 2, as shown by Pastrana et al. (2006) and is thought to be important in the successful regeneration of RGCs when produced in primary OEG. Matrix molecules such as laminin, L1 and N-cadherin are also good candidate molecules known to promote axon outgrowth (Lander et al., 1985, Bixby et al., 1988, Lochter & Schachner, 1993) and interestingly are present in olfactory glia. However, the presence of these molecules alone in RGC regenerative studies was insufficient to induce mature RGC axonal growth (Goldberg et al., 2002).

An important property of OEG needs consideration in the context of visual system repair. OEG do not appear to interfere with appropriate axon–target cell interaction and recognition. In recent studies using tissue transplantation in the visual system, Vukovic et al. (2007) showed that Schwann cells but not OEG interfere with retinal target recognition in a fetal co-graft transplant model. Fetal tectal tissue was mixed with either adult Schwann cells or OEG, each purified glial population having been labeled ex vivo with a lentiviral vector that encoded the sequence for green fluorescent protein (GFP) (Ruitenberg et al., 2002). These mixed tissues were transplanted onto the midbrain of neonatal rats where they established neural connections with the underlying host brain. When assessed several weeks later, the majority of host retinal axons were consistently found to innervate appropriate target regions in the tectal grafts in the presence of OEG, but axons were more scattered in grafts containing Schwann cells and only a few retinal fibers found their target sites in the graft neuropil. Therefore OEG have particular potential as a transplant cell within CNS neuropil because they can support axonal growth but also allow correct target innervation. OEG may also have an advantage of being able to ameliorate the inhibitory nature of the damaged visual system while providing suitable growth matrix and growth-promoting factors to assist injured RGCs to survive and regenerate. Given these initial studies highlighting the potential benefits OEG treatment in the injured visual system, we now summarize some more recent studies that have been published in the context of new data from our laboratory.