As we mammals get older, many of us start to lose our eyesight because the neurons in our retinas degenerate. Our retinal ganglion cells might get attacked by glaucoma, or our rods and cones (photoreceptors) might get eroded by macular degeneration or retinitis pigmentosa. Somewhere in the course of evolution, we lost our ability to regenerate those kinds of cells, just like we lost the ability to regenerate limbs. Once they’re gone, they’re gone.
But we humans did develop some other things really well: the ability to use reason and the desire to sustain ourselves. And those attributes have brought us to the verge of making up for some of our evolutionary shortcomings.
It’s amazing enough that we can now grow human stem cells into retinal “organoids” — little balls that contain all the different types of cells it takes to make a functioning retina, even organized into the right layers.
But now we’ve learned that if we break up the organoid into individual cells, those cells are capable of spontaneously forming signal-communicating connections (synapses) with other retinal cells. That means that a patient could have their own stem cells grown into retinal cells and applied to their own retina, these new cells could functionally replace the old, and vision could be restored. No gene therapy required, thanks very much.
You can read all about this last hurdle being surmounted at the University of Wisconsin labs of Drs. David Gamm and Xinyu Zhao in the January 4 issue of the Proceedings of the National Academy of Sciences.
Just last year, Gamm’s lab had shown that rods and cones (photoreceptors) made from stem cells can respond to light just like healthy ones do. That’s a great development for making individual cells for therapy, but to be part of a functional retina, those rods and cones need to be able to transmit their signals to the rest of the retina. That happens through synapses, ultrathin connections between neurons through which signaling molecules (mostly glutamate) are passed:
Retinal organoids (ROs) gave Gamm and Zhao hope that the defective parts of a retina could be reconstituted for real from stem cells, because not only do all the RO cells form the layers they are supposed to, but they also make connections to each other inside the RO with synapses. You can see how similar the structure of an RO is to a real retina as far as cell types and synapses (dyed green):
So the question is, if we break these RO cells up and apply the appropriate ones to the patient’s retina, will they be able to remake these synapse connections? That’s what Gamm and Zhao’s labs set out to test here.
They broke up some ROs with papain, which is an enzyme from papaya used as a meat tenderizer and digestive aid but which is also known to destroy synapses. (So no injecting papain directly into your eyeballs, OK?)
After the papain treatment, they saw that the proteins that are important for synapse function were fortunately still there, but they had kind of recoiled back into the cells. So it seemed the cells would have a good shot to reestablish synapses with each other if they could just get their bearings again.
They cultured these RO cells together as individuals for 20 days on a plate, in a situation similar to what they’d encounter when applied to a real retina. But how can you tell if neurons have formed these tiny synapses and that those synapses are functioning?
Luckily there’s a slick way to do that called “synaptic tracing”. It turns out that rabies virus can transmit between neurons only through functioning synapses, so we can use it to find out not only whether synapses are present, but also how well they’re working. (This feels like a good time to add rabies virus to the very long, and yet still-growing, list of things to not inject into your eyeballs.)
The way this is done is very cool, and stick with me here because you will get some colorful photos at the end that will make it pretty obvious what happened.
First we have to get rabies virus to infect only a small percentage of our cells without ransacking the entire culture, and we also have to mark those cells as “starters” somehow. So we have to do a little setup first.
We’re going to start with a different virus — lentivirus — into which we have put a gene for green fluorescent protein (GFP) that we’ve aimed at the nucleus. We’ll be able to spot any cells that get infected with our lentivirus, then, because they’ll have a big green dot at the center. We can do some trial-and-error with the amount of lentivirus we use so that we end up with about 5% of our cells infected.
We’ll put two other genes into our lentivirus called TVA and Rgp, and we’ll get to why those are both important in just a second.
Next we’re going to go ahead and infect our cells with rabies virus, but we’re going to change the gene for its coat protein. Usually that is Rgp, but we’re going to replace it with a different one called Env. Viruses that use Env as their coat proteins can only infect cells that have TVA, and that’s exactly why we just put TVA into our green-dot cells. Now we can set rabies virus loose on the culture, and it will only infect the green-dot cells.
We’ll put a gene for mCherry (a red fluorescent protein) into our rabies virus, so any cells infected with it will have a red color throughout the cell, and it will be easy to spot rabies-infected cells. So our green-dot “starter” cells are all going to get infected with rabies because they all have TVA, and that will turn our “starter” cells a festive red and green.
Recall that we had also put the Rgp gene into our lentivirus, so our green-dot cells also make Rgp protein. Once the rabies virus infects our green-dot cells, they will regain their original coat protein, go back to their old selves, and … ohhhhhhh.
So now about 5% of our cells are red-and-green “starter” cells, and they can infect other cells in the culture with rabies (and give them a red color) only if they are connected to other cells by working synapses! If that happens, we should see red cells with no green dot — that is, rabies-infected cells that weren’t starter cells. Bam! There’s your visualization, and now let’s get to it…
A nice control to start with is the whole system we just talked about, but no Rgp in the lentivirus. That means starter cells shouldn’t be able to infect any other cells, because rabies won’t have its normal coat protein. All we should see are starter cells, colored red and green.
So the little graphic at left below shows red-and-green starter cells unable to infect other cells, even if there are active synapses. The bluer pictures on the left have an additional stain called DAPI, which detects DNA with a blue color, so every cell will show up blue. This way you can visualize the percentage of cells that got infected as starter cells. Then on the right side we get rid of the blue DAPI so you just see red and green. Notice everybody who is red also has a green dot.
OK, now let’s do the real test, where Rgp is included in the lentivirus, so that now the rabies virus can infect other cells, but only through active synapses. Same deal on the colors, and now we hope to see some red-only neurons:
We do see a good deal of rabies infection of non-starter cells, which means we have active synapses! And that means we are on to clinical trials!
“We’ve been quilting this story together in the lab, one piece at a time, to build confidence that we’re headed in the right direction,” says Gamm, who patented the organoids and co-founded Madison-based Opsis Therapeutics, which is adapting the technology to treat human eye disorders based on the UW-Madison discoveries. “It’s all leading, ultimately, to human clinical trials, which are the clear next step.”
After they confirmed the presence of synaptic connections, the researchers analyzed the cells involved and found that the most common retinal cell types forming synapses were photoreceptors — rods and cones — which are lost in diseases like retinitis pigmentosa and age-related macular degeneration, as well as in certain eye injuries. The next most common cell type, retinal ganglion cells, are degenerate in optic nerve disorders like glaucoma.
“That was an important revelation for us,” says Gamm. “It really shows the potentially broad impact these retinal organoids could have.”