AAV-mediated, RPE-specific expression of the high-affinity lactate transporter MCT2 preserves cone photoreceptors and extends cone function in multiple rat and mouse models of retinitis pigmentosa while FLIM biosensors reveal altered RPE lactate and glucose handling.
0:00Welcome to Base by Base, the papercast that brings genomics to you wherever you are. Thanks for listening, and don't forget to follow and rate us in your podcast app. So imagine a stark reality for a second.
0:10You have this genetic condition, and it begins by slowly stealing your night vision. Right, which is terrifying on its own. Yeah, exactly. But then as the years go on, you start to realize your daylight and color vision, they're vanishing too.
0:23Your whole visual world is just steadily shrinking to this tiny pinhole. And that is the reality for about one in 4000 people, globally with retinitis pigmentosa or RP. Right. But the really baffling mystery here at the cellular level is that the genetic mutations causing this, they almost exclusively affect the rods, right?
0:42Exactly, just the rods, the night vision cells. So if the codes only broken in the rods, why do the cones, which, I mean, they have perfectly normal DNA in these patients, why do they eventually die off too?
0:54What really exposes this massive blind spot in how we've traditionally targeted genetic diseases? I mean, we spend decades hunting for a broken code in one specific cell. Like operating under the assumption that cells just exist in a vacuum?
1:08Right, which they don't. They exist in this highly interdependent neighborhood. So in the primary resident goes down, the infrastructure collapses and the bystanders suffer. And that brings us to our deep dive today.
1:18Today, we celebrate the work of a team at Harvard Medical School, who have really advanced our understanding of this bystander effect. Yeah, Laurel Chandler, Apollonia Gardner, and Constant Sepco. Their 2025 PNAS paper approaches this collapsing neighborhood with just, uh, a brutal elegant simplicity.
1:39Because instead of trying to run gene therapy on the 100 plus distinct mutations that break the rods. Which is a complete logistical nightmare, by the way. Right. So instead of doing that, they engineer this gene agnostic workaround.
1:52Like, if the genetically healthy cones are dying because their local ecosystem is failing, We just bypass the rods entirely. We rewire the ecosystem to feed the cones. Exactly. And to understand that rewiring, we really have to look at the severe metabolic constraints of the retina.
2:07I mean, photoreceptors are arguably the most metabolically demanding cells in your entire body. Wow, really? The most demanding. Yeah, they rapidly metabolize glucose and dump out just massive quantities of lactate as a byproduct.
2:21But they're geographically isolated, right? Like the photoreceptor layer doesn't have its own vascular network. Right, because blood vessels would literally cast shadows on the retina and, you know, scatter the incoming light.
2:32So they rely entirely on the retinal pigment epithelium or the RPE. Exactly. It's this single layer of support cells sitting immediately beneath the photoreceptors. It basically acts as a, like a metabolic broker between the blood supply and the retina.
2:47Okay, so let's look at this like an automated power grid. The RPE is the local substation pulling current, which is the glucose from the main grid. like that analogy. And normally the substation passes that power directly to the high demand factories, which are the photoreceptors.
3:02And then the factories send back a localized waste product, the lactate. Right. And the key is what the RPE does with that lactate. It actually imports the lactate produced by the photoreceptors and oxidizes it to fuel its own basal energy needs.
3:16Oh, so it runs on the waste. Exactly. And this arrangement allows the RPE to spare the imported glucose, routing it almost entirely up to the photoreceptors. But in retinitis pigmentosa, this grid experiences a localized blackout.
3:30The rods die off, which means that massive flow of lactage is suddenly dries up. Right. And without that, the substation basically severs the lines to the remaining factories. The RPE stops exporting glucose to the surviving cones and starts hoarding it for itself.
3:46The cones literally starve. Yeah, it's a tragic cascade. But structurally or mechanically, why does the absence of a waste product trigger the support cell to consume the primary fuel? Well, it's an automatic biochemical feedback loop.
4:00When the RPE metabolizes lactate, it converts it to pyrogate. That specific chemical reaction consumes a molecule called NAD+. Okay, NAD plus. Right. And doing that significantly lowers the intracellular pool of this crucial coenzyme.
4:14So without sufficient NAD plus, the RPE physically cannot run its own glycolysis. It forces the glucose to just pass through the cell unconsumed. Exactly. So the waste product is the physical break on the RPE's own metabolism.
4:28When the rods die, the brake comes off. NAD plus levels rebound, glycolysis in the RPE fires up and it oxidizes the glucose before it can ever reach the cones. You got it. And the Harvard team's intervention steps right into this broken supply chain.
4:43If the RPE needs lactate to shut off its own glycolysis and the rods are dead. The RPE has to get that lactate from somewhere else, specifically the systemic blood supply. Right. They needed to force the RPE to pull lactate up from the chorroid, but the RPE's existing machinery just isn't built for that.
4:58Because it uses a transporter protein called MCT3 on its basal membrane to interface with the blood, right? Exactly. But MCT3 is a low affinity transporter, meaning it requires a really steep concentration gradient to function efficiently.
5:11And the baseline lactate concentration in systemic blood is what, roughly 2.5 to 4.6 millimolar. Yeah, which is simply too dilute for MCT 3 to capture meaningful amounts. It's adapted to export high concentrations of lactate out of the retina, not import trace amounts from the blood.
5:28It is like having a fishing net with giant holes. Great when the water is swarming with fish, but completely useless when fish are scarce. That's a great way to put it. So to reverse this flow, you need a transporter with a much higher sensitivity, a finer mesh net.
5:42Which is where MCT2 comes in. Right. MCT2 is a high affinity transporter. It has a Michaelis constant, or come F, of about 0.7 millimolar. Meaning it can efficiently bind and transport lactate, even at those really low concentrations found in systemic circulation.
5:58Exactly. So the researcher's goal was to install MCT2 on the basal side of the RPE. But this requires gene therapy, and precision is everything here. You can't just flood the eye with a virus and hope it lands in the right place.
6:10No, definitely not. You need the right vector and the right promoter. For the vector, they used an AAV 8 vector. Because AAV 8 have a really highly documented efficient tropeism for retinal cells when injected into the subretinal space.
6:23Right. It penetrates the tissue well and transduces the RPE with high efficiency, but the critical safety feature of this whole construct was the promoter, best one. Promoters are basically the ignition switch for the gene, right?
6:36Exactly. If the viral vector accidentally infects a surviving cone or, you know, a nearby blood vessel, We absolutely do not want them suddenly expressing a high affinity lactate transporter. Because that would just ruin their own metabolic balance.
6:49Right. So the best one promoter ensures the MCT2 gene is strictly expressing the RPE cells and nowhere else. It isolates the metabolic reprogramming to the exact support cells that dictate the flow of nutrients.
7:00Exactly. So now we have this targeted high affinity lactate vacuum installed in the RPE, and then they move to the Invivo models. And they didn't just use one standard lab mouse. They use 3 distinct rodent models.
7:12The S334, line 3 rat, the FEB albino mouse, and the P23H pigmented mouse. Yeah, and using 3 models is a really rigorous way to validate the core premise of this paper. These rodents don't just represent different species.
7:27They represent entirely different genetic pathways of retinal degeneration. Right. The S 334der and P23H mutations affect Rhodopsin folding and trafficking in completely different ways. Well, the FVB mouse.
7:39That one suffers from a mutation in the PDE 6B gene. Exactly. So if the therapy works across all three, it basically mathematically proves the approach is gene agnostic. It doesn't matter how the neighborhood caught fire, deploying the metabolic fire blanket just works.
7:55And structurally, it really did. When they looked at the retinas, the treated eyes had significantly higher cone survival. Yeah, I read that in the FEB mice. The untreated retinas form these literal physical craters, where cones die off en masse.
8:08Right. It completely disrupts the entire architecture of the outer nuclear layer. But the MCT2 treatment largely prevented those craters from forming. But, you know, having more surviving cells under a microscope is one thing.
8:21Structural preservation is just the 1st hurdle. Exactly. A structurally intact cell that cannot fire an action potential is clinically useless. A cone might survive starvation by just entering a dormant state.
8:33But vision actually requires active signal transduction. Can these animals actually see? Right. They had to prove functional rescue, which they did using an optimotor assay. That's the test where you place the mouse in the center of a platform surrounded by monitors displaying rotating black and white stripes, right?
8:52A mouse will reflexively track the stripes with its head if it can actually see them. And if you make the stripes thinner, increasing the spatial frequency, you test the limits of their visual acuity. Exactly.
9:02And at day 40, the mice with the MCT2 treatment retains significantly higher spatial frequency tracking compared to the control group. Because the control group's vision was rapidly crashing. Right. So the preserved cones were not just structurally intact.
9:16They were actively transducing light and sending coherent signals to the visual cortex. The functional rescue is clear then. The mice can see the stripes. They can, but, you know, correlation is a dangerous trap in biology.
9:29Right. How do we prove they are seeing the stripes because of the specific glucose lactate shift? And not just because, say, injecting a viral vector triggered a transient immune response that temporarily cleared debris from the eye.
9:43Exactly. And measuring fluctuating intracellular metabolites in real time, inside a microscopic living cell is notoriously difficult. It requires moving beyond standard fluorescence microscopy. Yeah. Typically, if we want to measure a chemical in a cell, we introduce a fluorescent biosensor and measure its brightness.
10:02But intensity-based imaging is highly susceptible to artifacts. Like if the RPE cell changes volume, or the local concentration of the sensor varies. Right. The brightness changes independent of the actual metabolite concentration.
10:15Like, well, brightness is easily fooled. If you snap a glow stick and put it in a small glass of water, it looks incredibly bright, but pour that same water into a large bucket, and the glow looks dim, even though the absolute amount of the glowing chemical is identical.
10:28That is a perfect analogy. And to bypass this, the researchers utilized phlegm fluorescence lifetime imaging microscopy. So flim bypasses the concentration and volume artifacts entirely. Yes. Instead of measuring how bright the emission is, it measures how long the floor floor remains in its excited state before emitting a photon and returning to the ground state.
10:49So we are analyzing decay curves measured in nanoseconds. Exactly. And this lifetime changes exclusively when the sensor grabs its target molecule. So they engineered 2 specific genetically encoded sensors into the RPE, right?
11:01Lilac, which binds lactate, and gluco SNFRTS, which binds glucose. Right. When these sensors bind their respective metabolites, they undergo a structural conformational change, and that physical shape change alters their fluorescent lifetime.
11:16So you pulse the RPE cells with a laser and measure the nanosecond decay. Exactly. And the cells expressing the new MCT2 transporter showed a significantly shorter lifetime on the lilac sensor. And a shorter lifetime for this specific sensor indicates high intracellular lactate.
11:31The RPE was successfully pulling lactate up from the blood. Yes. And even more crucially, they analyze the glucos NFRTS sensor. The MCT2 expressing cells showed a longer fluorescence lifetime there. Which correlates with an accumulation of intracellular glucose.
11:47The RPE was no longer oxidizing it. Exactly. And to seal the mechanism, they ran a control experiment using iodoacetic acid or IAA. IA chemically blocks glycolysis, right? Right. So by treating normal RPE cells with IAA, they forced glucose accumulation, and the phlegm signature of the IAA treated cells perfectly matched the phlegm signature of the MCT2 treated cells.
12:10So they proved, at the molecular level, that the high affinity transporter was shutting down RPE glycolysis just as effectively as a chemical blockade. It is definitive proof of mechanism. The RPE internalized the systemic lactate, the NAD plus levels dropped, glycolysis halted and the glucose was spared for the cones.
12:29The mechanism is airtight. We successfully rewired the local power grid. But, you know, this all sounds almost too good to be true. So what's the catch? Well, biology rarely lets a single intervention solve a cascading failure.
12:41By day 53 in the mouse model, the significant functional advantage of the treatment had actually vanished. The vision crashed anyway. Wow, so it's a temporary rescue. Yeah, it highlights the multifaceted, aggressive pathology of retinitis pigmentosa.
12:56Metabolic starvation is a massive early driver of cone death, but it is definitely not the only one. Right, because when rods die, they don't just disappear quietly. They rupture and leave behind cellular debris.
13:09Which triggers massive microglial activation and chronic local inflammation. And rods consume the vast majority of the oxygen delivered to the outer retina too. Exactly. So take the rods away. And the local oxygen tensions skyrockets.
13:22You suddenly have a hyperoxic environment, subjecting the surviving cones to brutal oxidative stress, fixing the food supply doesn't stop the neighborhood from being radioactive. Exactly. So a true viable human therapy will likely require pairing this metabolic reprogramming with a secondary payload, maybe an antioxidant gene or an anti-inflammatory agent.
13:43So the MCT2 therapy buys the cones a window of time, but it cannot single-handedly hold back the entire degenerative cascade. Right, and that brings us to maybe the most alarming finding in this paper, a detail that really dictates the future of this therapy.
13:56The toxic? Yes. When they injected this AAV8 MCT2 therapy into rats, the rats tolerated it beautifully. The RPE remained healthy, and cone survival was robust. But when they injected it into mice, even at viral doses, a full log unit lower than what the rats received, the mounds RPE cells became enlarged, dysmorphic, and developed dense cytoskeletal stress fibers.
14:20Right. Why is a therapy that rescues a rat actively poisoning a mouse? That is a huge question. It is a critical translational bottleneck, and there are 2 leading hypotheses here. The 1st is immunological.
14:31Mice and rats have divergent immune architectures, particularly in how they present antigens. Right, for instance, mouse T cells, black MHC class 2 molecules, which are highly active on both rat and human T cells.
14:44Exactly. The mouse immune system might be detecting the AAV vector itself or the exogenous MCT2 protein, and launching a localized cytotoxic response that rats simply do not trigger. Which, if it is merely a quirk of the mouse immune system lacking MHC class too, we can somewhat breathe a sigh of relief for future human trials.
15:01Right. Given that human immunology aligns closer to the rat in this specific regard. But what if the toxicity is mechanical? Because MCT transporters don't just move lactate, they are simporters. And that is the far more concerning hypothesis.
15:15MCT transporters move one molecule of lactate across the cell membrane by tightly coupling it to the transport of one proton. And protons are the fundamental currency of cellular acidity. Exactly. MCT2 is a high affinity transporter, meaning it is operating constantly, pulling lactate in even at trace concentrations.
15:35Consequently, it is simultaneously flooding the mouse RPE cells with protons. So you're dropping the intracellular pH, you are slowly acidifying the cell from the inside out. Right. If the mouse RPE has a lower endogenous buffering capacity than the rat RPE, That constant, relentless influx of protons will disrupt enzymatic functions.
15:54It would trigger the unfolded protein response in the endoplasmic reticulum and cause the exact cytoskeletal stress fibers the researchers observed. Exactly. This is the massive, unanswered question for human translation.
16:06Will human RPE cells react like the rat, buffering the acid load without breaking a sweat? Or will they react like the mouse, where the accumulating acid causes catastrophic cellular stress? Right. If it is the latter, this entire elegant metabolic strategy might be dead in the water.
16:22unless you can figure out how to co-express a proton pump to bail the acid back out of the cell. It underscores the delicate tightrope of metabolic engineering. You cannot alter one variable in a complex dynamic system without triggering downstream perturbations.
16:38Exactly. Solving a glucose bottleneck risks creating a lethal pH bottleneck. It is a sobering reminder that we are dealing with deeply integrated systems, not just isolated chemical pathways. Yeah, absolutely.
16:50But even with a toxicity hurdle, the conceptual leap demonstrated here remains staggering. It really does. By stepping back from the broken genetic code of the rods and focusing entirely on the logistics of the support cells, this research team found a way to buy the cones precious time.
17:06It is a masterclass in treating the ecosystem rather than chasing the individual mutations. What does this mean for the broader future of medicine? Well, the applications of this philosophy extend far beyond ophthalmology.
17:18We are increasingly learning to view degenerative disease through the lens of metabolic dependency and local support networks. Which leaves a final lingering thought for you to consider? The retina is not an isolated organ.
17:30Embryologically, it is a direct outcropping of the central nervous system. Right. It is literally brain tissue pushed forward into the eye. Exactly. If we can successfully reprogram the metabolic support staff, the RPE, to save starving neurons and the retina, look at the broader landscape of neurodegenerative diseases.
17:49Right. Look at Alzheimer's or Parkinson's, where cortical neurons are starving, while Gliel cells and astrosites desperately try to regulate their microenvironment. How long until we apply this exact same Trojan horror strategy, dropping high authinity transporters into astracytes, to rescue the collapsing ecosystems of the human brain?
18:07It's a fascinating question. This episode was based on an open access article under the CCBY 4.0 license. You can find a direct link to the paper and the license in our episode description. If you enjoyed this, follow or subscribe in your podcast app and leave a 5 star rating.
18:23If you'd like to support our work, use the donation link in the description. Now stay with us for an original track created, especially for this episode, and inspired by the article you've just heard about.
18:33Thanks for listening, and join us next time as we explore more science base by base.
