Following optic nerve injury associated with acute or progressive diseases, retinal ganglion cells (RGCs) of adult mammals degenerate and undergo apoptosis. These diseases have limited therapeutic options, due to the low inherent capacity of RGCs to regenerate and due to the inhibitory milieu of the central nervous system. Among the numerous treatment approaches investigated to stimulate neuronal survival and axonal extension, cell transplantation emerges as a promising option. This review focuses on cell therapies with bone marrow mononuclear cells and bone marrow-derived mesenchymal stem cells, which have shown positive therapeutic effects in animal models of optic neuropathies. Different aspects of available preclinical studies are analyzed, including cell distribution, potential doses, routes of administration, and mechanisms of action. Finally, published and ongoing clinical trials are summarized.
Optic neuropathy is an umbrella term encompassing a large number of disorders that cause optic nerve damage. The retrograde degeneration of axons of retinal ganglion cells (RGCs) within the optic nerve can ultimately lead to the death of RGCs, which have their cell bodies in the inner retina, culminating in irreversible visual loss [
In contrast to the progressive nature of glaucoma, acute optic neuropathies are characterized by the acute onset of visual loss and are usually caused by ischemia (ischemic optic neuropathies), traumatic brain injury (traumatic optic neuropathy), and infection or inflammation (optic neuritis). Other causes of optic nerve injury, with varied clinical presentations, are compression, toxic or nutritional causes, infiltration of neoplastic or inflammatory cells, and papilledema secondary to elevated intracranial pressure [
After optic nerve injury, RGCs are unable to regenerate their axons and undergo apoptosis, mostly due to an intrinsic inability to regenerate but also due to the inhibitory environment of the central nervous system (CNS) [
More robust results have been obtained with the stimulation of RCG intrinsic regeneration program through, for example, the deletion of the phosphatase and tensin homolog (PTEN) or the suppressor of cytokine signaling 3 (SOCS3) [
Although these approaches are very promising, they are not easily translated to the clinic. The development of novel molecular tools for gene silencing has created an exciting new field of research, but there is still a long way to go before promising findings are translated into approved therapies, mainly due to safety issues that must be resolved [
Cell therapy has emerged as a promising tool in regenerative medicine. Different research groups have used embryonic stem cells or induced pluripotent stem cells to generate RGCs that could replace the lost cells [
Another line of investigation has indicated that bone marrow-derived cells, such as mesenchymal stem cells (BM-MSCs) and mononuclear cells (BM-MNCs), could increase RGC survival and promote axonal regeneration after optic nerve injury in rodents [
This review summarizes and discusses the main findings of preclinical studies that have investigated the therapeutic action of bone marrow-derived cells in animal models of optic neuropathies. The use of noninvasive imaging methods to assess the distribution of the transplanted cells in the visual system and to investigate the efficacy of cell therapies is also discussed.
The therapeutic potential of BM-MNCs has been extensively investigated in several disorders, including acute myocardial infarction and stroke [
Pang and colleagues [
EPCs are circulating bone marrow-derived cells involved in endothelial repair and postnatal angiogenesis, due to their capacity to differentiate into mature endothelial cells and to secrete soluble factors, such as insulin-like growth factor 1 (IGF-1) and vascular endothelial growth factor (VEGF). They express HSCs and endothelial cell markers and can be cultured and expanded
In addition, BM-MNCs contain 0.001%–0.01% mesenchymal stem cells (MSCs) that can be expanded in culture [
MSCs were first described by Friedenstein and coworkers [
Because they adhere to plastic, MSCs can be easily isolated by plating the mononuclear cell fraction or even the whole bone marrow suspension in tissue culture flasks. All contaminating nonadherent cells are removed after serial medium changes.
MSCs are characterized by the panel of positive and negative cell surface markers proposed by the International Society for Cellular Therapy in 2006 [
MSCs occupy anatomically distinct locations within the bone marrow and are also found in endosteal, stromal, and perivascular niches [
MSCs have the capacity to migrate to sites of injury following their intravascular administration. This process depends on molecules present on the surface of MSCs and endothelial cells, such as P-selectin and integrins [
Interestingly, MSCs are considered to be not inherently immunogenic, as they express low levels of HLA class I antigens and do not express, or express in negligible levels, HLA class II antigens as well as their costimulatory molecules such as CD80, CD86, and CD40 [
The current view is that MSCs can exert neuroprotective and proregenerative effects, mainly by secreting multiple factors that act in a paracrine fashion [
Several studies have described the therapeutic effects of bone marrow-derived cells in animal models of optic nerve disease. The main characteristics and the principal findings of these preclinical studies are summarized in Table
Summary of published preclinical studies.
Animals | Cell injection(s) | Effects, longest time analyzed after injury | Distribution, longest time analyzed after transplantation | Mechanisms | References |
---|---|---|---|---|---|
|
|||||
|
|||||
Adult female Wistar rats (MSCs were from male) | ivit rat BM-MSCs or rat AT-MSCs 1 w after injury |
|
4 w; cells were integrated into GCL and INL | Reduced TNF- |
[ |
|
|||||
Adult female Wistar rats | ivit rat BM-MSCs 2 w after injury |
|
8 w; ILM, NFL, GCL | Increased expression of bFGF and CNTF | [ |
|
|||||
Adult brown Norway rats (sex not specified) | ivit BDNF-expressing rat BM-MSCs; rat BM-MSCs 2 d after injury |
|
6 w; GCL, vitreous | Chronic low dose delivery of BDNF | [ |
|
|||||
Adult male Sprague-Dawley and Lewis rats |
ivit and IV rat live or dead BM-MSCs 1 w before the injury |
|
5 w: majority of cells in the vitreous, few in the ILM, NFL, and GCL | n/a | [ |
|
|||||
Adult male Sprague-Dawley rats | ivit rat BM-MSCs |
|
4 w: vitreous, ILM, GCL | Increased expression of bFGF, CNTF, and BDNF | [ |
|
|||||
Aged male Sprague-Dawley rats | ivit BM-MSCs 6 w after injury |
|
n/a | n/a | [ |
|
|||||
Adult brown Norway female rats | ant/cha murine BM-MSCs |
|
24 h: cells migrated to the damaged area 96 h: cells cleared | Secretion of paracrine factors; recruitment of ocular progenitor cells | [ |
|
|||||
|
|||||
|
|||||
Adult male and female Lister hooded rats | ivit rat BM-MNCs immediately after injury |
|
2 w: very few cells GCL, INL, optic nerve | Müller glia modulation; bFGF, Tax1BP1, and SytIV increased expression | [ |
|
|||||
Adult male Sprague-Dawley rats | ivit rat BM-MSCs |
|
n/a | n/a | [ |
|
|||||
Adult Wistar rats (sex not specified) | ivit NTF-secreting BM-MSCs; human BM-MSCs; rat BM-MSCs 3 d before the injury |
|
24 d: vitreous | Neurotrophic factors secretion (GDNF, BDNF); possible inflammatory reaction to xenotransplantation | [ |
|
|||||
Adult male and female Lister hooded rats | ivit rat BM-MSCs |
|
18 w: vitreous | Increased expression of bFGF and IL-1 |
[ |
|
|||||
Adult male Sprague-Dawley rats | ivit rat DP-MSCs or rat BM-MSCs |
|
3 w: DP-MSCs in the vitreous; BM-MSCs n/a | Neurotrophic factors release (NGF, BDNF, and NT-3); DP-MSCs release more NGF and BDNF than BM-MSCs; reduced scar tissue on the crush site | [ |
|
|||||
|
|||||
|
|||||
Adult Sprague-Dawley rats (sex not specified) | A droplet containing 1,000–5,000 BM-MSCs was placed on the RGCs surface |
|
1 w: adjacent to GCL, not integrated into the retina | Secretion of growth factors, specially PDGF (also tested |
[ |
ant/cha: anterior chamber; AT-MSCs: adipose tissue-derived mesenchymal stem cells; BDNF: brain-derived neurotrophic factor; bFGF: basic fibroblast growth factor; BM-MNCs: bone marrow-derived mononuclear cells; BM-MSCs: bone marrow-derived mesenchymal stem cells; CNTF: ciliary neurotrophic factor; DP-MSCs: dental pulp-derived mesenchymal stem cells; GCL: ganglion cell layer; GDNF: glial cell line-derived neurotrophic factor; IL-1Ra: interleukin-1 receptor antagonist; ILM: inner limiting membrane; INL: inner nuclear layer; IOP: intraocular pressure; IPL: inner plexiform layer; IV: intravenous; ivit: intravitreal; NFL: nerve fiber layer; NTF: neurotrophic factors; PDGF: platelet-derived growth factor; RGCs: retinal ganglion cells; sub: subretinal; SytIV: synaptotagmin IV; Tax1BP1: Tax1-binding protein 1; TNF-
Bone marrow cells have been tested in several animal models of glaucoma. Yu and coworkers [
Johnson et al. [
Emre and coworkers [
BM-MSCs also protected RGCs in a model of glaucoma that used laser pulses directed to the eye in order to block the aqueous outflow [
In a model of retinal ischemia and reperfusion, Li and coworkers [
Several studies have used optic nerve crush or transection as a broad-spectrum model of diseases that affect the optic nerve. Our group, for instance, demonstrated that both rat BM-MNCs and BM-MSCs had therapeutic effects after optic nerve crush. Intravitreally injected BM-MNCs protected RGCs in the first 2 weeks (although the effect was lost at 4 weeks) and increased axonal outgrowth for at least 4 weeks after optic nerve crush. In addition, animals treated with BM-MNCs showed an increased expression of the immediate early gene NGFI-A in the superior colliculus after light stimulation, indicating full-length axonal regeneration and synaptic connection with neurons of the superior colliculus [
On the other hand, in rats, intravitreal injection of BM-MSCs led to sustained neuroprotection of RGCs for at least 4 weeks after optic nerve crush, which was the longest time period analyzed. Axonal outgrowth was also increased in these animals at the two time points analyzed (2 and 4 weeks). The association of sustained RGC survival and axonal regeneration in BM-MSC-treated animals suggests that RGCs may find favorable conditions to regenerate over longer distances and reconnect to their targets, as observed after BM-MNC treatment. This possibility is currently being investigated. Although BM-MNCs and BM-MSCs remained mostly in the vitreous body, treated animals showed increased expression of bFGF in their retinas. The antiapoptotic Tax1-binding protein 1 (Tax1BP1) and synaptotagmin IV gene expression were also upregulated in BM-MNCs treated retinas, while IL-1
In another recent study, BM-MSCs were injected intravitreally after optic nerve crush and promoted an increase in axonal regeneration, in a dose-dependent manner [
MSCs derived from human umbilical cord blood (hUCB-MSCs) were also tested as an alternative source. Zhao and coworkers injected hUCB-MSCs 7 days after optic nerve crush and observed improved RGC survival up to 28 days after the injury. They also observed increased levels of BDNF and glial cell line-derived neurotrophic factor (GDNF) in the retina after transplantation [
Chen and coworkers, on the other hand, reported a transient effect of grafted hUCB-MSCs. Twenty-one days after injury, both RGC survival and GAP-43 expression increased in the retina of treated animals, but these differences were no longer present after 28 days [
A different perspective was provided by Johnson and coworkers [
In addition to MSCs, the bone marrow includes other cell types with potential therapeutic effects on the visual system. Monocytes represent a population of circulating bone marrow-derived cells that play important roles in vascular and tissue homeostasis, as well as in the responses to pathogens, toxins, and other types of insults [
Monocytes are recruited to the ganglion cell layer and the inner plexiform layer, where they differentiate into macrophages, in the first days following retinal intoxication with glutamate, a murine model of RGC death. The intravenous injection of bone marrow-derived monocytes, 1 day after injury, promoted the survival of Brn3a+ RGCs 7 days after the injury. This neuroprotective effect was further confirmed by Fluoro-Gold labeling of surviving RGCs. In contrast, the transfer of IL-10-deficient monocytes had no effect on RGC survival, indicating that the main mechanism of action of the transplanted monocytes was related to the release of this anti-inflammatory cytokine. In addition, transferred monocytes increased the number of proliferating neural progenitors in the ciliary body, although there was no evidence of neurogenesis [
A recent study, however, found that only 0.5–1% of the microglia/macrophages in the damaged retina came from circulating monocytes, 7–14 days after optic nerve transection in mice [
One strategy to stimulate optic nerve regeneration is the induction of intravitreal inflammation by injuring the lens or by intraocular injection of zymosan, a yeast cell wall preparation [
Concerning the route of administration, most groups have injected the cells intravitreally. In one study, the systemic administration of BM-MSCs had no effect after experimental glaucoma induction, while the intravitreal injection protected RGCs and delayed axonal degeneration [
Cell doses were heterogeneous among the studies. Beneficial effects were observed with doses ranging from 30,000 to 500,000 BM-MSCs, while the BM-MNC dose used in our study was 5,000,000 cells (Table
These observations suggest that the effects of MSCs are dose-dependent, but, to our knowledge, no studies have established the maximum tolerated dose for intraocular transplantation. There is a limit to the volume that can be injected into the vitreous body without causing damage to the eye, and therefore the number of injected MSCs cannot be too high.
Since we have observed that the neuroprotection conferred by BM-MNCs decays over time and that most of the transplanted cells are cleared from the vitreous body within 2 weeks, we investigated whether a second administration would change this outcome. We found that, even with a second dose, neuroprotection is lost over time. Axonal regeneration, however, was improved after the second injection, suggesting that BM-MNCs may provide neuroprotection and stimulate axonal outgrowth through different pathways [
In regenerative medicine, it is essential to determine where and how long the cells remain in the host tissue after transplantation and whether this phenomenon contributes to therapeutic effects.
Several studies have tracked bone marrow cells after injection in animal models of visual diseases. Yu and coworkers injected green fluorescent protein (GFP) expressing BM-MSCs into the vitreous body, after inducing glaucoma by ligation of episcleral veins in adult rats. Two weeks after the injection, they found GFP-positive cells along the inner limiting membrane, with few of them integrated into the ganglion cell layer. Transplanted cells were found in the host tissue for up to 8 weeks.
Interestingly, Na and coworkers found that when BM-MSCs were transplanted into normal eyes, they remained in the vitreous cavity. However, in ischemia/reperfusion injured retinas, BM-MSCs were found along the inner limiting membrane, and few of them were integrated into the ganglion cell layer, 4 weeks after injection [
In glaucomatous eyes, induced by photocoagulation of the trabecular meshwork, the majority of the BM-MSCs remained in the vitreous body, occasionally attached to the posterior lens capsule, and a small number of them reached the nerve fiber layer and the ganglion cell layer [
Johnson and coworkers [
The use of histochemistry for cell tracking has many limitations. For example, misleading results can be obtained if the dye used to label transplanted cells is able to diffuse to retinal cells, as observed after labeling BM-MSCs with 4′,6-diamidino-2-phenylindole (DAPI) [
The use of noninvasive
Another method to macroscopically evaluate cell migration is to directly label cells with a radionuclide for nuclear medicine imaging or an exogenous contrast agent for MRI. Radionuclide cell labeling has been used for decades to diagnose infections through evaluation of the migration of labeled leukocytes by Single Photon Emission Computed Tomography (SPECT) [
Contrast agents for MRI may surmount many of the present limitations of radiopharmaceuticals. SPIONs were originally created as intravenous contrasts for hepatic imaging and more recently have been used by different groups for cell labeling. One advantage of SPIONs is that the iron in the particles can be tracked by MRI
To combine the advantages of SPECT/CT and MRI cell tracking, our group has investigated the distribution of BM-MNCs, using cell labeling with SPIONs (FeraTrack, Miltenyi Biotec, Germany) and
Taken together, these results suggest that while BM-MNCs injected into the vitreous body are cleared from this region within the first 2 weeks [
Several studies suggest that bone marrow-derived cell grafts act in a paracrine fashion and that neurotrophic factors play a role in the therapeutic effect. It is possible that the paracrine effect would be enhanced if transplanted cells could remain longer in the damaged tissue, given the progressive nature of several optic neuropathies. However, the length of time that the presence of grafted cells is necessary to sustain neuronal survival and/or to stimulate regeneration remains to be investigated. It is also important to determine whether once these goals are achieved, the permanence of transplanted cells would somehow impair visual function by, for example, eliciting a sustained inflammatory response in the eye.
While intravitreally transplanted MSCs may remain for several months inside the eye, different results were observed when BM-MSCs were injected into the anterior chamber in a model of glaucoma. They were cleared from the tissue within 96 hours, probably phagocytosed by microglial cells, since about 20% of the transplanted cells expressed the microglia/macrophage marker F4/80 on day 2. In spite of the short time they remained, MSCs were able to induce the regeneration of the damaged trabecular meshwork [
In addition to analyzing cell distribution, noninvasive imaging methods also allow the assessment of structural parameters that can be used in the investigation of the safety and efficacy of cell therapies. Fischer and coworkers used cSLO to study ocular integrity and evaluate possible modifications in the anatomy of the cornea, lens, vitreous, and retina after a subretinal injection of MicroBeads containing eGFP-expressing BM-MSCs [
MRI is an important tool for the evaluation of the CNS and may be performed with or without the use of contrast agents such as gadolinium-diethylenetriamine pentaacetic acid (Gd-DTPA) and manganese (Mn2+) [
Haenold and collaborators [
However, the use of MnCl2 has limitations, because overexposure to Mn2+ causes neurological toxicity in humans and animals [
Diffusion tensor imaging (DTI) is another valuable MRI technique that maps the random motion of water molecules and reflects CNS microstructural integrity and pathology [
Different parameters may be used to evaluate the functional effects of cell therapies in animal models of optic nerve injury, including optokinetic response, pupil light reflex (PLR), electroretinography (ERG), and visual evoked potentials (VEPs).
Optokinetic response may be analyzed to quantify the capacity of distinguishing spatial frequency and contrast sensitivity in animal models of glaucoma [
PLR may be analyzed by computerized pupillometry [
VEP analyzes occipital lobe brain wave potential after visual stimuli [
Although the initial studies of cell therapy suggested that bone marrow cells could differentiate into neuronal cells [
In most of the preclinical studies discussed in the previous sections, neuroprotection was attributed to a paracrine effect, and no evidence of neural transdifferentiation of MSCs has been shown. In this respect, the therapeutic effect of transplanted cells could be related to the capacity of MSCs to alter the microenvironment of the injured tissue, through the release of trophic factors and inflammatory mediators [
BM-MSCs can secrete a variety of trophic factors [
In order to increase this paracrine activity, a few studies have engineered the cells before transplantation. Intravitreally injected rat or human BM-MSCs that were stimulated to secrete neurotrophic factors were found in clusters between the lens and the retina and remained in the eye for at least 3 weeks after optic nerve transection. In this study, human BM-MSCs were more neuroprotective than rat BM-MSCs, which is correlated with the higher secretion of BDNF and GDNF by human cells, although it is possible that a beneficial inflammatory reaction could have been elicited by the xenotransplant [
In addition to the release of neurotrophic factors that may act directly on damaged neurons, the interaction between transplanted cells and retinal glial cells cannot be ignored. Indeed, we have shown that BM-MNC transplantation was associated with a reduction in the expression of glial fibrillary acidic protein (GFAP) in radial glial processes throughout the retinal layers, which is a marker of Müller cell reactivity in response to injury [
A large number of studies have demonstrated that BM-MSCs exert beneficial immunomodulatory effects by interacting with cells of the innate and adaptive immune system [
Several studies have suggested that the crosstalk between MSC and the injury microenvironment leads to the secretion of soluble factors by these cells [
Besides the secretion of soluble factors, it has recently been suggested that the therapeutic effect of BM-MSCs may be partially related to the release of extracellular vesicles. These vesicles, named “exosomes” or “microvesicles” according to their size, contain proteins, mRNA, and microRNA that can be transported from one cell to another [
Considering the biological function of proteins and microRNAs transported by extracellular vesicles, it would be interesting to evaluate the therapeutic potential of BM-MSC extracellular vesicles in animal models of optic neuropathies. In a model of stroke, for example, Xin and coworkers [
Jonas et al. published a case report of autologous BM-MNC therapy for a 43-year-old patient with advanced retinal and optic nerve atrophy due to diabetic retinopathy [
Connick and coworkers [
At least three studies registered in clinicaltrials.gov have been designed to carry out bone marrow cell therapy for optic neuropathies. Weiss and collaborators, from a private clinic in Florida, USA, started an open-label study in August 2013 with an estimated completion date in August 2017, where 300 patients are expected to receive autologous BM-MSCs (
De Paula and coworkers, from the University of São Paulo, Brazil, have designed an open-label, single-group phase 1 study to perform intravitreal autologous transplantation of 106 BM-MSCs in patients with retinal degeneration or primary open-angle glaucoma (
Jamadar and colleagues, from Chaitanya Hospital in Pune, India, started in September 2014 an open-label, phase 1/phase 2 study (
Other trials of stem cell therapies for degenerative eye diseases, such as diabetic and ischemic retinopathy, which may secondarily affect the optic nerve, have been recently reviewed by Mead and collaborators [
Studies using different animal models of optic nerve injury, such as optic nerve compression or transection and elevation of the intraocular pressure, have shown that RGC degeneration can be reduced by intravitreal transplantation of BM-MSCs or BM-MNCs, which was the delivery method used most often. Doses varied among the studies, but few of them suggested that higher doses have increased therapeutic potential. Bone marrow-derived cell effects were mostly attributed to the release of soluble factors that can protect RGCs and/or modulate the inflammatory response in the retina. Interestingly, there is evidence of long-term persistence of BM-MSCs at the injection site, although the importance of this phenomenon remains to be elucidated. Such preclinical studies showing the neuroprotective and proregenerative effects of bone marrow-derived cells have encouraged the execution of phase 1 or phase 2 clinical trials for diseases that affect the optic nerve. Several trials are ongoing and a few have been concluded, indicating the feasibility and safety of intravitreal or intravenous administration of autologous bone marrow-derived cells. Most preclinical studies focused on morphological outcomes such as RGC survival and axonal outgrowth, but as these parameters were improved in treated animals, there is a growing need for visual functional analysis in future studies. Further investigations are also necessary to unravel the mechanisms of action of transplanted cells, in order to allow the development of safe and efficient therapies.
The authors declare that there is no conflict of interests regarding the publication of this paper.
The authors are grateful for the financial support from Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Programa Ciência Sem Fronteiras-CAPES, and Departamento de Ciência e Tecnologia do Ministério da Saúde (DECIT). The authors wish to thank Janet W. Reid for revising and editing the language in the text.