Lab-Grown Retinas Mirror the Real Thing, but Their Metabolism Lags Behind

The retina is one of the greediest tissues in the human body. Photoreceptors burn through glucose at a staggering rate, shuttling carbon through glycolysis and mitochondrial pathways to sustain phototransduction, synaptic transmission, and the relentless daily renewal of their light-sensitive outer segments. Now, a team of researchers from Radboud University Medical Centre and Leipzig University has asked a deceptively simple question: when scientists grow miniature retinas in a dish, do those lab-made tissues actually run on the same metabolic fuel as the real thing? The answer, published in the journal Metabolomics, is a fascinating mix of yes and not quite.

Retinal organoids, the three-dimensional tissues coaxed from stem cells that self-organize into layered, light-sensitive structures, have become one of the most celebrated tools in modern vision science. They allow researchers to study retinal development and inherited blindness without experimenting on animals, and they can be produced in large batches from a single starting culture. But most assessments of how faithfully organoids replicate the native retina have relied on RNA sequencing and proteomics, techniques that measure molecular blueprints rather than the actual chemistry of life. Genes can be switched on without the corresponding metabolic machinery running at full tilt, and proteins can be present without the small molecules they process flowing through pathways at realistic rates.

To close that gap, the team led by first author Pragati Gupta turned to untargeted metabolomics, a technique that captures the small molecules, sugars, amino acids, and lipids that represent the working currency of cellular metabolism. Using gas chromatography coupled to mass spectrometry, they compared the intracellular metabolite profiles of mouse retinal organoids at day 28 of differentiation with retinas dissected from three-day-old C57BL/6 mice. The choice of comparison points was deliberate: previous transcriptomic work had shown that day-28 organoids most closely resemble early postnatal retinal development, making the P3 retina the fairest developmental benchmark available.

The results were striking in their overall overlap. Of 166 annotated compounds detected across the study, 148 were shared between organoids and native retina, spanning amino acids and their derivatives, lipids and fatty acids, carbohydrates and sugar alcohols, and organic acids. Crucially, the conserved metabolites included the chemical heavyweights of retinal function: glutamate and gamma-aminobutyric acid, the central mediators of retinal neurotransmission; taurine, one of the most abundant metabolites in the vertebrate retina and essential for photoreceptor survival; and the TCA cycle intermediates citrate, succinate, fumarate, and malate that power the cell’s energy factories. The organoids, in other words, had built a genuinely retina-like biochemical identity rather than becoming some metabolically generic blob of tissue.

But the devil, as always, was in the differences. When the researchers applied stringent statistical thresholds, requiring a false discovery rate below 0.05 and a fold change of at least tenfold in either direction, a pattern of selective metabolic immaturity emerged. Several metabolites central to central carbon metabolism were abundant in native retina but fell below detection in organoids, including glucose-6-phosphate, a gateway metabolite of glycolysis; ribulose-5-phosphate, a signature intermediate of the pentose phosphate pathway; glycerol-3-phosphate and dihydroxyacetone phosphate, which link sugar metabolism to lipid synthesis; and phosphoethanolamine and myo-inositol, both tied to membrane phospholipid metabolism. TCA cycle-associated compounds including citric, fumaric, and oxalic acid also showed altered abundance between the two tissues.

Pathway enrichment analysis sharpened the picture further, identifying significant alterations in alanine, aspartate, and glutamate metabolism, along with changes in arginine biosynthesis, glycine, serine, and threonine metabolism, taurine and hypotaurine metabolism, and starch and sucrose metabolism. Multivariate statistics told the same story from a different angle. Principal component analysis and partial least squares discriminant analysis both separated organoids from native retina cleanly, with metabolites such as pantothenic acid, erythritol, phosphate, and glycerol-3-phosphate ranking among the strongest drivers of the distinction. Metabolites including urea, erythritol, phosphoethanolamine, and pyroglutamic acid pointed toward the native retina cluster, while fructofuranose, galactopyranose, and modified galactose derivatives pointed toward the organoids.

The team also tracked what the growing organoids ate and excreted, by analyzing the conditioned culture media collected longitudinally between day 8 and day 28 of differentiation. Here, the data revealed clear stage-dependent remodeling: early, intermediate, and late media samples each formed distinct clusters in multivariate space, with amino acid-associated and carbohydrate-associated metabolites showing the most pronounced temporal shifts. Notably, extracellular glutamate and glycine changed in abundance as the organoids matured, a finding consistent with earlier reports that developing retinal tissue actively consumes glutamate and aspartate. This supports an emerging concept in vision science of the retina as a metabolic ecosystem, in which photoreceptors, Müller glia, and the retinal pigment epithelium exchange lactate, nutrients, and antioxidants in a tightly choreographed economy.

Why does this matter beyond the laboratory bench? The retina’s extraordinary metabolic demands make it uniquely vulnerable to metabolic failure, and defects in energy metabolism are increasingly implicated in inherited retinal degenerations, diabetic retinopathy, and age-related macular degeneration. If organoids are to serve as faithful models of these diseases, or as testing grounds for gene therapies and cell transplants, their metabolic fidelity becomes a critical quality metric. The new findings suggest that current culture conditions produce organoids with the right metabolic vocabulary but an incomplete fluency: the core retinal metabolites are present, yet the energy-generating and biosynthetic pathways that support mature photoreceptor function have not fully caught up with the native tissue.

The authors are careful to frame their conclusions within the study’s limitations. Metabolomics by gas chromatography measures relative abundance rather than metabolic flux, so a metabolite absent from the organoid signal is not necessarily absent from the organoid itself; it may simply fall below the detection threshold or be lost during extraction and derivatization. The technique is also relatively blind to complex lipids, retinoids, and low-abundance signaling molecules, precisely the classes that matter for photoreceptor outer segments. The comparison captures a single developmental snapshot rather than a longitudinal trajectory, and some differences may reflect developmental asynchrony between a dish and a living eye rather than intrinsic organoid deficiencies. The organoids also lack the retinal pigment epithelium interaction and vascular nutrient exchange that shape retinal metabolism in vivo, and the conditioned media changes may partly reflect stage-specific culture supplements rather than biology alone.

Even so, the study delivers a practical roadmap. The specific metabolic gaps it identifies, in glycolysis, the pentose phosphate pathway, phospholipid metabolism, and amino acid neurotransmitter cycling, point directly to candidate interventions: adjusted nutrient compositions, improved oxygenation, co-culture with retinal pigment epithelium, and lipid or antioxidant supplementation of the kind already shown to improve photoreceptor outer segment development in earlier organoid studies. The authors also emphasize that their mouse organoid system, which matures in roughly 28 days compared with around 200 days for human retinal organoids, offers a fast, cost-effective platform aligned with the 3R principles of replacing, reducing, and refining animal use. Future work combining metabolomics with isotope tracing and functional assays should reveal whether these metabolic differences represent delayed maturation that culture optimization can fix, or a more fundamental ceiling of the dish-bound retina. Either way, the message is clear: to build a better retina in a dish, scientists must now feed it not just growth factors, but the right chemistry.

Subject of Research: Metabolomic comparison of stem cell-derived mouse retinal organoids and native mouse retina

Article Title: Metabolic profiling of retinal organoids reveals conserved core metabolites and alterations in glycolytic and amino acid metabolism

Article References: Gupta, P., Wiesner, C., Ziemssen, F., Klevering, J., Kulkarini, P., & Hosseinzadeh, Z. (2026). Metabolic profiling of retinal organoids reveals conserved core metabolites and alterations in glycolytic and amino acid metabolism. Metabolomics, 22(6), Article 171. https://doi.org/10.1007/s11306-026-02523-4

Image Credits: AI Generated

DOI: 10.1007/s11306-026-02523-4

Keywords: retinal organoids, metabolomics, mouse retina, GC-MS, glycolysis, TCA cycle, amino acid metabolism, photoreceptors, taurine, glutamate, pentose phosphate pathway, stem cell models