By: Alan N Glazier, OD, FAAO, Dipl., ABO

When Proteins Go Bad

Like most of my colleagues, I find managing patients with a condition for which there are no curative treatments extremely frustrating. As an eye doctor, the major conditions we run into in this class include Retinitis pigmentosa, Stargardt-like macular degeneration, dry age-related macular degeneration, and progressive glaucoma. All these conditions share a common molecular root cause: protein misfolding. Misfolding causes pathology when proteins essential to normal retinal function fail to fold into their correct three-dimensional shapes, triggering cascades of cellular stress that ultimately lead to the death of photoreceptors, retinal pigment epithelial cells, or retinal ganglion cells.

Quantum computing has significant potential to help break ground in developing curative treatments for these conditions. Predicting how a protein will fold, understanding why a protein misfolds into a mutant protein, and designing molecules to correct or prevent misfolding are among the hardest computational challenges in all of science. Classical computers, even very powerful ones, are fundamentally limited in their ability to solve these problems within practical timeframes. Quantum computers, by exploiting the principles of quantum physics, are poised to transform this landscape.

How Quantum Computing Works

Quantum computing differs from traditional computing in several ways. A conventional computer processes information in bits; tiny switches that can represent values of either 0 or 1 and switch back and forth to generate combinatorial patterns of 0’s and 1’s. Every calculation a conventional computer performs, from running a spreadsheet to analyzing an OCT scan, is built from combinations of these binary switches.

A quantum computer uses qubits (quantum bits). Thanks to a property of quantum physics called superposition, a qubit can exist as 0, 1, or both simultaneously. A helpful analogy: if a classical bit is like a coin lying flat, showing either heads or tails, a qubit is like a coin spinning in the air, embodying both states at once until it is observed. Whatever superposition the qubit is in when it is measured is known as its “state”.

A second quantum property, entanglement, links qubits together so that the state of one instantly influences the state of another, regardless of distance. By calculating superposition and entanglement, a quantum computer can explore an enormous number of possible solutions simultaneously, rather than testing them one by one. For problems that involve searching through vast numbers of possibilities, such as finding the lowest-energy shape of a protein, this represents a fundamental change in what is within range of computing.

 Quantum Computing: The Answer to Protein Folding Disease?

Proteins begin as linear chains of amino acids. To function properly, each protein must fold into a precise three-dimensional shape, similar to yet way more complex than an origami that folds itself. When it folds correctly, the protein is functional. If it folds incorrectly, the protein fails and often results in toxicity and/or pathology.

Predicting how a protein will fold is one of the most formidable problems in science. A single protein chain has more possible conformations than there are atoms in the observable universe. A classical supercomputer approaches this by testing shapes one by one: computing and testing the energy of one proposed configuration, then another. Even with modern supercomputers, simulating the folding of a single complex protein can take months, and there are millions of proteins humans depend on for healthy functioning biology.

Quantum computing offers a fundamentally different approach. Rather than guessing and checking, the quantum computer can model the physics of folding itself. Through the healthy folding process, proteins naturally settle into the shape that requires the least energy to maintain their “ground”, or stable state. A quantum processor simulates then attributes a value to a protein’s energy use by allowing its qubits to settle into their lowest-energy state, mimicking the physical process by which a protein finds its correct fold.

From Simulation to Digital Twins and Drug Discovery

The vision emerging from current quantum computing research is the concept of molecular digital twins. In principle, a quantum processor could take the molecular data from a patient’s specific disease, the exact structure of their mutant protein, the cellular environment in which it misfolds, and create a computational model of that biochemistry. A candidate therapeutic molecule could then be introduced into this simulation to observe how it interacts with the misfolded protein: Does it correct the folding error? Does it stabilize the protein? Does it produce unintended interactions?

This kind of simulation could dramatically accelerate drug development. Currently, bringing a new pharmaceutical agent from laboratory discovery to the clinic takes an average of 12 to 15 years and costs billions of dollars, in large part because the earliest stages depend on animal models that often do not recapitulate human disease. A quantum simulation based on human molecular data enables rapid, molecular-level testing of thousands of candidate compounds before a single dose is ever given to a living organism.

Protein Misfolding and Retinal Disease

One of the most metabolically active tissues in the human body is the retina. Its photoreceptors, retinal pigment epithelial (RPE) cells, and retinal ganglion cells (RGCs) depend on continuous, high-fidelity production and turnover of specialized proteins. Rod photoreceptors alone replace approximately 10% of their outer segment content each day, a process that requires massive ongoing protein synthesis, folding, and transport. When protein folding goes wrong in the retina, the consequences can be devastating.

Retinitis Pigmentosa: The Prototype Protein Misfolding Disease of the Retina

The most thoroughly understood form of RP caused by protein misfolding involves mutations in the gene encoding rhodopsin, the light-sensitive visual pigment that accounts for roughly 90% of all protein in the rod outer segment.

Different classes of rhodopsin mutations produce distinct protein defects, ranging from misfolding and aggregation to instability and altered function. To date there are no conventional treatments that have been developed that provide any positive therapeutic results. Gene therapy and molecular “chaperones” (small molecules that help mutant proteins fold correctly) are being explored, but progress has been limited by the immense complexity of understanding the specific effects of each mutation on a protein’s three-dimensional structure and dynamics.

This is exactly the type of problem where quantum computing might have a transformative impact. By accurately simulating how each rhodopsin mutant misfolds, modelling the full energy landscape of the protein, and the conformational changes induced by specific amino acid substitutions, quantum computing could identify which pharmacological chaperones or small molecules would be effective for each mutation class, moving treatment from one-size-fits-all to precision medicine.

Stargardt-Like Macular Degeneration

Stargardt-like macular degeneration is a form of juvenile-onset macular dystrophy caused by mutations in the ELOVL4 gene, which encodes a membrane protein involved in the synthesis of long-chain fatty acids. The disease is characterized by bilateral, symmetrical macular lesions and progressive central vision loss. No treatment is currently available.

Quantum computing’s capacity for detailed molecular simulation could elucidate the precise structural changes caused by ELOVL4 modifications and screen for molecules that could prevent the mutant protein from sequestering its wild-type counterpart, a novel therapeutic strategy that current computational methods are poorly equipped to explore.

Age-Related Macular Degeneration and Protein Misfolding

A hallmark of early AMD is the accumulation of drusen beneath the retina. Significantly, amyloid β protein, the same misfolded protein central to Alzheimer’s disease, has been identified within drusen, and its presence correlates with complement activation and RPE/photoreceptor degeneration.

For wet AMD, anti-VEGF therapies can reduce vascular leakage but have limited long-term efficacy in halting disease progression. For dry AMD and geographic atrophy, which account for most AMD cases, there are essentially no effective treatments. The molecular complexity of AMD, involving multiple misfolded proteins, disturbed autophagy, mitochondrial dysfunction, and metabolic dysregulation in the RPE, presents a systems-level challenge well-suited to quantum computational modelling. Quantum simulations could model the interactions between amyloid β aggregates, complement proteins, and RPE cellular machinery to identify druggable targets that current approaches have not revealed.

Glaucoma: Protein Misfolding in the Trabecular Meshwork and Retinal Ganglion Cells

Protein misfolding is implicated in glaucoma at two anatomical sites. First, in the trabecular meshwork (TM): mutations in the MYOC gene, which encodes the myocilin protein, are linked to both juvenile open-angle glaucoma and adult-onset primary open-angle glaucoma. Mutant myocilin misfolds and accumulates in the endoplasmic reticulum (ER) of TM cells, leading to ER stress, cell death, and extracellular matrix disruption. These changes reduce aqueous humour outflow, increasing IOP. Chemical chaperones that promote protein folding and reduce ER stress have been shown in animal models to prevent TM cell death and lower IOP, a proof of concept that correcting protein misfolding can treat the root cause of pressure elevation.

Second, retinal ganglion cells: experimental evidence has demonstrated that amyloid β protein, the same misfolded protein implicated in AMD and Alzheimer’s, contributes to RGC apoptosis in glaucoma. Exogenous amyloid β induces significant RGC death in a dose- and time-dependent manner, and agents that target amyloid β formation have shown efficacy in altering the profile of RGC loss in experimental models.

The dual role of protein misfolding in glaucoma, driving both the pressure elevation that initiates damage and the ganglion cell death that causes vision loss, makes it a particularly compelling target for quantum computational approaches. Modelling myocilin misfolding could lead to more effective IOP-lowering strategies, while simulating interactions between amyloid β and RGCs could identify neuroprotective agents that address the disease’s currently untreatable progressive neurodegeneration and help identify new and better targets for gene therapies, leading to more permanent therapies in the future.

Quantum Computing for Retinal Disease

Precision Pharmacological Chaperones

Pharmacological chaperones are small molecules that enter cells and stabilize misfolded proteins into their optimal three-dimensional shape, represent one of the most promising therapeutic strategies for protein misfolding diseases of the retina. Studies have shown that retinoid-based chaperones can help certain mutant rhodopsin’s bind their chromophore and regain partial function. However, the enormous number of distinct mutations (over 200 in rhodopsin alone), each producing different structural defects, makes identifying the right chaperone for each mutation a monumental computational challenge.

Quantum computing could simulate the misfolded structure produced by each individual mutation, model the energy landscape of the mutant protein, and computationally screen vast libraries of candidate chaperone molecules to identify those that best stabilize each specific misfolded variant. This would transform the development of pharmacological chaperones from a laborious, mutation-by-mutation experimental process into a rapid, targeted, computational pipeline.

Accelerated Drug Discovery Through Molecular Simulation

Beyond chaperones, quantum simulation could accelerate the discovery of entirely new classes of therapeutic molecules. By creating high-fidelity computational models of the misfolded proteins involved in each disease, quantum computing could enable researchers to design drugs that interact with these targets at the atomic level. Candidate compounds could be tested against digital models of the diseased molecular environment, identifying therapeutic hits and eliminating toxic candidates before any animal or human testing, potentially compressing the drug development timeline from decades to years.

Patient-Specific Molecular Modelling

Perhaps the most transformative long-term application is the concept of patient-specific molecular modelling, creating a computational replica of an individual patient’s mutant protein in its specific cellular context. Given that the clinical severity of diseases like RP varies enormously even among patients with the same mutation, the ability to model each patient’s unique molecular landscape could enable personalized treatment strategies and lead to the ability to select the optimal therapeutic agent for each individual case. The concept of a “molecular digital twin” represents a paradigm shift from empirical trial-and-error medicine to precise molecular engineering of therapies.

Where Are We Now, and What Lies Ahead?

While all this is fantastic news for patients with these conditions, the technology is not ready for prime time. Current quantum computers are in what researchers call the Noisy Intermediate-Scale Quantum (NISQ) era; they handle hundreds to thousands of qubits, but these qubits are prone to errors from environmental interference from something as small as interference by a cosmic ray or moderate temperature change. Fault-tolerant quantum computers with millions of stable qubits required for large-scale protein simulations, remain years to perhaps a decade or more away.

Investment in quantum computing for healthcare applications is accelerating across both the public and private sectors. Pharmaceutical companies and academic research centers are already developing hybrid classical-quantum algorithms for drug discovery. As hardware matures over the coming decade, the retinal disease community is well-positioned to be among the early beneficiaries and the urgent unmet need for effective treatments.

Conclusion: A Convergence of Need and Capability

In my practice, when I counsel patients diagnosed with these conditions, I make it a point to share that meaningful research is underway while being transparent that these technologies are still years, if not decades, from clinical availability. I find patients appreciate knowing that some of the brightest minds in science are actively working toward solutions, even when I cannot promise those solutions will arrive in time to help them personally. It doesn’t erase the weight of the diagnosis, but for many, there is comfort in knowing they are not forgotten, that their condition is not a closed chapter, but one that researchers around the world are still writing.

The diseases discussed in this article collectively represent the majority of incurable blinding conditions we see in our clinical practice. What they share at the molecular level is a failure of protein folding, a problem of geometry so complex that identification of treatments has lagged as identifying appropriate therapies required technology beyond what researchers could access.  By exploiting the laws of quantum physics to simulate the molecular world at its own level, quantum computers have the potential to teach us how and why specific proteins misfold in diseases, and to identify the precise molecules that can correct those errors. The technology is not mature, but advancing rapidly, and foundational research linking protein misfolding to retinal disease is well established. The scientific foundation is sound, computational tools are advancing, and the unmet need drives innovation, ensuring that one day we will have the tools required to treat these conditions.

Disclosures: The author reports no financial conflicts of interest relevant to the subject matter of this article.

Note: This article is intended as an overview for clinicians. Readers interested in detailed molecular mechanisms are referred to the primary literature cited herein.