Michael Young, Ph.D., is Associate Scientist and Director, Ocular Regenerative Medicine Institute at Schepens Eye Research Institute, Massachusetts Eye and Ear Infirmary and Associate Professor of Ophthalmology, Harvard Medical School and Affiliated Faculty member at the Harvard Stem Cell Institute. He completed his Ph.D. at the University of Cambridge and did postdoctoral work at University College London and M.I.T.
StemBook editor, Lisa Girard, spoke with Michael recently about the work in his lab on stem cells, retinal repair, and what the future might hold translating findings into the clinic. Below is an edited transcript of the conversation.
Can you tell me a little about what your lab here at Schepens works on?
We work on regeneration of the retina, and we do that with drugs and we also do that with stem cells. So were focused on trying to rebuild the photoreceptor layer, mostly, of the degenerating retina.
What are, a bit more specifically, some of the projects ongoing in your lab?
We have two main projects. One is using retinal stem cells to try to replace lost photoreceptors in the degenerating retina, and that’s really targeted at retinitis pigmetosa initially, but we hope things like macular degeneration, diabetic retinopathy, diseases like those, will also be treatable someday with stem cells. The other project involves drug delivery to the eye. We are trying to use drugs that mimic or induce the expression of GDNF (glial cell-derived neurotrophic factor), a growth factor, in the retina and it has been shown to be one of the most potent growth factors, rescuing all cells really, photoreceptor cells, ganglion cells, all these cells can be protected by GDNF so we’re trying to come up with new strategies to deliver these drugs into the back of the eye.
Do you know what is the mechanism by which GDNF acts?
So we know that the Muller glia of the retina, sort of the radial glia of the eye, are very much involved in the process. The signaling likely goes through these cells and the Muller glia have a whole host of mechanisms to save, or in cases of disease, actually target for death other neurons in the retina. So what we think is that Muller glia have receptors for GDNF that signals through GDNF Receptor Type 1 and Ret and then it gets complicated and we don’t know what comes next. But there are multiple pathways that could all converge on rescue of other neurons.
One thing that’s a big challenge in stem cell therapeutics is allogeneic response in stem cell transplantation, so one of the things that seems interesting about the eye as a model for stem cell therapeutics is immune privilege and I was wondering if you could talk a little bit about that?
Sure. Immune privilege was discovered many years ago by my mentor, Wayne Streilein, in the anterior chamber of the eye and he called it ACAID (anterior chamber immune deviation). What that means is that, in general, the eye has a deviant immune response, a distinct immune response that is antigen specific. So the first time the eye sees an antigen it’s recognized and that generates suppressor T cells, meaning the second time that antigen comes around, the second time the body sees that antigen, there is an active downregulation of the immune response, so it can induce tolerance to tissues. But, a couple of things, it’s not complete, and not all cells will survive in the eye, and we don’t know what happens in the diseased eye. So that’s a big question, a lot of disease states get rid of immune privilege and we don’t know in the diseased eye whether it’s going to be present. It is a really big question right now whether the altered immunity of the eye is a good thing or a bad thing for graft survival.
So has this kind of altered immunity been seen in some of the diseases that are more commonly known such as retinitis pigmentosa and macular degeneration?
We know only about this from animal studies, and we know that laser injury gets rid of it. A model of RPE cell death, sodium iodate toxicity, also gets rid of it, so it seems like things that break down the blood-retinal barrier, which is analogous to the blood-brain barrier, might impact this. Some diseases have the broken down barrier and some do not. It depends on the disease, I think, but more to the point I think its important to note that the cell type that’s being transplanted is very important also. RPE cells can present antigen, so my feeling is that that just using allogeneic RPE cells might not be enough, you’re going to need immunosuppression. Something’s going to have to happen for those cells to survive, I predict. For cells that target the photoreceptor cell layer, or ganglion layer, these are neurons that do not present antigen and if they’re derived from stem cells they should be pure and not contain passenger leukocytes that could induce immunity. So, I think for our photoreceptor work, ganglion cell work, that’s in its early stages. I think neuron replacement in the eye is going to be very good in terms of acceptance of allogeneic tissue without systemic immune suppression. That’s what we are working on, a platform that would be an allogeneic graft and we are not proposing systemic immunosuppression, but local anti-inflammatory drugs that might get you over that hump of initial cell transplantation and cell survival but not do the kinds of things that chronic immunosuppression can do, which can be very bad especially in an elderly population, it’s not a good idea.
I know you mentioned GDNF, but are there any other drugs or compounds that could potentially protect or facilitate regeneration of retinal pigment epithelium cells (RPE) or retinal ganglion cells?
So the molecule that we like best is GDNF. It works great for ganglion cells. We’ve published stuff in mice, rats, and pigs showing ganglion cell protection with GDNF. CNTF (cilliary neurotrophic factor) is a drug that is already in the clinic for retinitis pigmentosa. There is an interesting paper that just came out showing CNTF works very well as an adjuvant therapy to gene therapy, but its got this weird controversial quality. CNTF temporary dedifferentiates cones and it’s controversial what it is doing but certainly if you put CNTF into the eye, the ERG (electroretinogram), test of retinal function, goes down, and then it gets better and some groups have described this response as a dedifferentiation of the cones and they think that’s because the cones get a collapse of their outer segments the structures that actually capture light, so they can’t capture much light anymore, but this is temporary and they come back.
That’s fascinating. So have they seen that morphologically?
Yes, exactly. Now it looks like we can look very carefully at this with noninvasive stuff, like OCT (optical coherence tomography), that can look very carefully at some of these 4-5 micron-thick layers of the retina and quantify them noninvasively over time and that is where this data has come from. So it looks like CNTF, which was initially messing up the eye, it was feared, might actually be temporary and then get better. So that’s a drug that’s being used by Neurotech in a slow release vehicle. We just had a conference here and I asked a real thought leader in the field about neuroprotection in glaucoma and he was not very optimistic. His opinion was that everything that has been tried has failed, that is basically true I would say, not great drugs have been tried so far but the reality is that it is a very difficult clinical trial for a long progressive degeneration like in glaucoma. So it’s very expensive to design a clinical trial and what are the endpoints? There are many unanswered questions and that is why glaucoma drugs have died on the vine before getting to clinical trials, because well what would it look like? It’s scary. And drug companies back away from that.
So aren’t there other diseases that are thought to be similar in mechanism to long- term diseases such as glaucoma but progress more quickly?
Yes, a surrogate, and that’s what we’ve tried to do but have been less than successful at that so far. There’s a mitochondrial disease, labors hereditary optic neuropathy (LHON), that happens very quickly, that’s the best example of an optic neuropathy that happens very rapidly. What we know from animals basically comes from crush injury, which doesn’t really model anything except for something that happens occasionally, traumatic optic neuropathy. To create crush injury you just use tweezers and squeeze the nerves and you get an injury. I am not a fan of that, I don’t do that. We actually elevate the intraocular pressure. You can do that in a mouse, a rat, a pig. It’s not that easy in a mouse but it can be done, it’s just designing the experiment correctly, you can get these sort of models that tell us, “yes this would work in glaucoma, but it would take a 10 year study to prove that in a phase 1 trial”. So, a surrogate disease is what we’ve been pushing forward. You mentioned RPE, and that not something I’m terrible familiar with, since they are not neurons, not something I have worked a lot on, but saving the RPE cells is something that is tempting to think about but hasn’t really been attempted, we’ve gone straight to replacement without the neuroprotective step in between. The RPE cell therapy now in the clinic is an RPE cell transplant and that’s ACT work and you mentioned Masayo Takahashi from RIKEN in Japan using iPS cells to generate RPE cells as well . So, one is in the clinic now, and we’ll see if the iPS work is going to be in the clinic as fast as people think. I think it’s early days to be putting iPS cells in a patient’s eyes. The eye is a great place to do cell therapy but it doesn’t mean we can do anything in the eye. There are lots of places to get RPE cells that aren’t iPS so the question is, do we need, as a group in Wisconsin is doing, a bank of immune tissue types present? These “super donors”, these people that are homologous for some of these loci. They might be great donors. That’s one way to go but I’m not convinced we need to go all the way to syngenic or autologous transplantation. That brings up the issue of maybe those RPE cells are in fact diseased and that’s why they’re dying, that’s one possibility. There are some diseases for which that is clearly not the case and you can target the RPE, but again, finding RPE cells is not a hard task. You can make them from ES cells, get them from fetal tissue, we don’t need to reinvent the wheel to obtain RPE cells.
You mentioned the ACT trial as well as the anticipated RIKEN trial in Japan. So what do you think subsequent trials are going to look like?
I predict, and I think its not an old prediction, that we will move beyond injecting a bolus of stem cells, or stem cell progeny, in this case and move toward an organized sheet, and depending on which group is working on it, a degradable or nondegradable polymer scaffold. I think it is clear from the early work in the ACT trial that most of the cells they are injecting die because RPE cells need a substrate and they don’t get one for a few days after they inject them into the subretinal space in a bolus, so that’s a bad way to deliver cells. A great first step and it’s easier to make a needle injection than place a scaffold in the eye, its much more difficult surgically, but RPE cells are not going to survive in that microenvironment. So for next steps, a group in London and a group in California are both working on scaffold approaches, delivering sheets of cells as a monolayer and then they’ll survive. What they will do we can then address but they will be involved in the next trial which is going to be more in cell survival.
This is a degradable scaffold?
The two groups are using different nondegradable scaffolds, different materials. Pete Coffey’s group on one hand and Mark Humayun, Dennis Clegg and these people working on the CIRM funded project. And they work together but their clinical trials will be different.
Michael, thanks so much for your time