Thylakoid membranes are the internal membrane system inside chloroplasts where the light-dependent reactions of photosynthesis take place. They capture sunlight, split water molecules, shuttle electrons through a chain of protein complexes, and build up the proton gradient that drives the production of ATP. In vascular plants, these membranes form an elaborate network of stacked and unstacked regions, and their architecture is far more dynamic than textbook diagrams suggest.
The Basic Layout
If you could peer inside a plant chloroplast at very high magnification, you would see two distinct but connected membrane domains. One consists of flattened, disc-like sacs stacked on top of each other like coins in a roll; these stacks are called grana. The other consists of longer, unstacked sheets that wind through the surrounding fluid (the stroma) and connect one granum to the next; these are the stroma thylakoids, sometimes called stroma lamellae. Electron tomography has confirmed that grana thylakoids connect to stroma thylakoids through tubular junctions, meaning the entire membrane network is continuous. A single chloroplast can contain dozens of grana stacks, each typically holding five to twenty discs, all linked into one vast, folded surface.
This stacked-versus-unstacked arrangement is not just for compactness. The two domains house different sets of protein machinery, a feature known as lateral heterogeneity. Grana thylakoids are enriched in the protein complex that captures light and splits water (Photosystem II, along with its associated light-harvesting antenna), while stroma thylakoids are enriched in Photosystem I and the enzyme that synthesizes ATP (ATP synthase).1PubMed Central. Lateral heterogeneity of plant thylakoid protein complexes: early reminiscences This spatial segregation has real functional consequences: it separates two photosystems that absorb slightly different wavelengths of light and prevents them from competing for energy in unproductive ways.
What the Membrane Is Made Of
Thylakoid membranes have an unusual lipid composition that sets them apart from most other biological membranes. Rather than being built primarily from phospholipids (as your cell membranes are), they are dominated by galactolipids, sugar-bearing fat molecules rarely found outside of photosynthetic organisms. The two most abundant are monogalactosyldiacylglycerol (MGDG) and digalactosyldiacylglycerol (DGDG), which together can account for roughly three-quarters of all thylakoid lipids. DGDG plays a particularly important structural role: neutron diffraction studies show that hydrogen bonds between the sugar head groups of DGDG in adjacent bilayers help stabilize the tight stacking of grana discs.2FASEB J. Contribution of galactoglycerolipids to the 3-dimensional architecture of thylakoids Adjusting the ratio of MGDG to DGDG can shift the membrane between different physical states, giving the plant a way to tune membrane geometry.
The remaining fraction includes two negatively charged lipids that balance each other in an interesting way. One is a phospholipid called phosphatidylglycerol (PG), and the other is a sulfur-containing glycolipid called sulfoquinovosyldiacylglycerol (SQDG). When phosphorus is plentiful, PG dominates, but under phosphate-starved conditions SQDG can partially substitute for PG, keeping the total negative charge of the membrane roughly constant.3PubMed Central. Role of membrane glycerolipids in photosynthesis, thylakoid biogenesis and chloroplast development PG itself cannot be entirely replaced, however, because it has direct roles in photosynthetic function that SQDG cannot fully cover.4PubMed. Phosphatidylglycerol and sulfoquinovosyldiacylglycerol: anionic membrane lipids and phosphate regulation This lipid swapping is a quiet survival trick that lets photosynthetic organisms keep working even when soil nutrients run low.
Splitting Water and Moving Electrons
The core job of the thylakoid membrane is to host the chain of protein complexes that capture light energy and convert it into chemical energy. The process begins at Photosystem II (PSII), embedded in the grana membranes. When PSII absorbs light, it uses that energy to pull electrons from water molecules at a special cluster of manganese and calcium ions known as the oxygen-evolving complex (OEC).5PubMed Central. Structural changes of the oxygen-evolving complex in photosystem II during the catalytic cycle This reaction releases oxygen as a by-product and feeds the freed electrons into a mobile carrier molecule called plastoquinone, which ferries them through the lipid bilayer to the next complex.
That next stop is the cytochrome b6f complex, a dimeric protein that acts as the bottleneck of the whole electron transport chain. It catalyzes the rate-limiting step, meaning it sets the maximum pace at which electrons can flow.6Biochimica et Biophysica Acta (BBA) – Bioenergetics. Cytochrome b6f – Orchestrator of photosynthetic electron transfer Inside b6f, a mechanism called the Q-cycle splits each pair of electrons from plastoquinol into two pathways, one high-potential and one low-potential, while simultaneously dumping protons into the thylakoid interior (the lumen). The complex also functions as a sensing hub, adjusting electron flow and even influencing gene expression depending on the redox state of its carriers.6Biochimica et Biophysica Acta (BBA) – Bioenergetics. Cytochrome b6f – Orchestrator of photosynthetic electron transfer
From b6f, electrons travel via a small copper-containing protein to Photosystem I (PSI) in the stroma thylakoids, which absorbs a second photon of light and boosts the electrons to an energy level high enough to reduce the final acceptor, NADP+, producing NADPH. Together, PSII and PSI form what is called linear electron transport, a chain that begins with water and ends with NADPH while pumping protons along the way.
Building the Proton Gradient That Makes ATP
Every proton pumped into the thylakoid lumen during electron transport creates an electrochemical difference across the membrane. This gradient has two components: a difference in proton concentration (the pH component) and an electrical voltage (the charge component). Together, they form the proton motive force, or pmf.7PubMed Central. Proton motive force in plant photosynthesis dominated by ΔpH in both low and high light ATP synthase, a rotary enzyme embedded in the stroma thylakoids, lets protons flow back out of the lumen and harnesses that flow to attach a phosphate group onto ADP, producing ATP.8PubMed. Rethinking the existence of a steady-state Δψ component of the proton motive force across plant thylakoid membranes In plant thylakoids, the pH component tends to dominate the proton motive force under both low and high light, which matters because the pH drop in the lumen also triggers protective mechanisms (discussed below).7PubMed Central. Proton motive force in plant photosynthesis dominated by ΔpH in both low and high light
Ion channels in the thylakoid membrane fine-tune how the proton motive force splits between its two components. Potassium and chloride channels help dissipate the electrical component faster, pushing more of the driving force into the pH gradient.9PubMed. Effect of ion fluxes on regulating the light-induced transthylakoid electric potential difference Computational modeling suggests that tweaking these ion flow rates could improve photosynthetic efficiency by speeding up the formation of the pH gradient that both powers ATP synthesis and activates photoprotection.10PubMed Central. Hacking the thylakoid proton motive force for improved photosynthesis: modulating ion flux rates that control proton motive force partitioning into Δψ and ΔpH
How Plants Deal with Too Much Light
Sunlight is essential for photosynthesis, but too much of it damages the machinery. Excess light energy that is not used for chemistry can generate reactive oxygen species that destroy proteins and lipids. Plants have evolved a rapid defense called non-photochemical quenching (NPQ), which safely converts surplus absorbed light energy into harmless heat before it can cause trouble.
The fastest component of NPQ depends on two things working together. First, the lumen must become acidic enough, which happens naturally when light drives protons inward faster than ATP synthase can let them out. A protein called PsbS acts as the pH sensor: it has two specific amino acids facing the lumen that get protonated when acidity rises, triggering a conformational change that helps reorganize the light-harvesting antenna.11PubMed. Regulation of photosynthetic light harvesting involves intrathylakoid lumen pH sensing by the PsbS protein Second, the xanthophyll cycle converts a pigment called violaxanthin into zeaxanthin, which directly participates in quenching. Under moderate light, both PsbS and zeaxanthin contribute to NPQ, but under very high light zeaxanthin becomes the dominant player.12PubMed Central. Non-photochemical Quenching Plays a Key Role in Light Acclimation of Rice Plants Differing in Leaf Color When zeaxanthin production is impaired, for instance through disruptions in fatty acid processing in the endoplasmic reticulum that ripple through to the thylakoid lipid profile, NPQ drops specifically under high light.13PubMed. Inter-Organellar Effects of Defective ER-Localized Linolenic Acid Formation on Thylakoid Lipid Composition, Non-Photochemical Quenching of Chlorophyll Fluorescence and Xanthophyll Cycle Activity in the Arabidopsis fad3 Mutant This is a good reminder that the thylakoid is not a self-contained unit; changes in lipid metabolism elsewhere in the cell can ripple in and impair its protective responses.
State Transitions and Energy Rebalancing
Beyond dumping excess energy as heat, thylakoid membranes also redistribute energy between their two photosystems on a timescale of minutes. If PSII is getting more light than PSI can handle, the imbalance causes the plastoquinone pool to become overly reduced. This triggers a kinase enzyme that phosphorylates part of the light-harvesting antenna (LHCII), loosening its grip on PSII and increasing its affinity for PSI.14PubMed. State transitions at the crossroad of thylakoid signalling pathways The reverse happens when PSI is over-excited: a phosphatase removes the phosphate groups from LHCII, and it returns to PSII.
Modeling work has refined this picture in a surprising way. Rather than imagining antenna complexes physically migrating long distances from grana to stroma lamellae, simulations indicate that most of the rebalancing happens through more local rearrangements. Phosphorylation causes LHCII proteins already sitting in the stroma thylakoids to be more efficiently recruited into functional complexes with PSI, while dephosphorylation helps LHCII already in the grana form tighter working clusters around PSII.15PubMed Central. Modeling the Role of LHCII-LHCII, PSII-LHCII, and PSI-LHCII Interactions in State Transitions Phosphorylation also weakens the lateral and stacking interactions that hold grana together, causing a measurable decrease in grana diameter. So the physical shape of the thylakoid network shifts subtly with every change in light quality.
Repairing Photodamaged Machinery
Even with photoprotection running at full tilt, PSII inevitably sustains damage, especially to a core protein subunit called D1. D1 sits at the heart of PSII and bears the brunt of the reactive chemistry; it has one of the fastest turnover rates of any protein in the plant. The repair process is a carefully choreographed cycle: damaged PSII complexes are phosphorylated, then the dimeric supercomplex is taken apart into monomers that migrate laterally from the tightly stacked grana core out to the more accessible stroma thylakoids, where the wrecked D1 protein is removed by specific proteases and replaced with a freshly synthesized copy. The repaired monomer then migrates back into the grana, re-dimerizes, and reassembles with its antenna proteins.16PubMed Central. Revisiting the photosystem II repair cycle
This cycle depends on the physical architecture of the membrane. The narrow gap between stacked grana discs can expand under high light, creating enough room for the bulky PSII monomers to slide out toward the edges.17PubMed. Close Relationships Between the PSII Repair Cycle and Thylakoid Membrane Dynamics If the grana were permanently rigid, damaged complexes would be trapped and the whole system would grind to a halt. This is a case where the dynamic, flexible quality of the thylakoid is not just a structural curiosity but a survival requirement.
Proteins That Shape the Membrane
What determines how many discs a granum stacks, or how wide each disc is? A family of small proteins called CURT1 (CURVATURE THYLAKOID 1) turns out to be a major player. These proteins concentrate at the margins of grana discs, exactly where the flat surface of a disc curves sharply to connect to the stroma thylakoid below or above. In Arabidopsis plants engineered to lack CURT1 proteins, grana had fewer but much wider layers. Overexpression of CURT1 produced the opposite phenotype: taller stacks of smaller-diameter discs.18PubMed Central. Arabidopsis CURVATURE THYLAKOID1 Proteins Modify Thylakoid Architecture by Inducing Membrane Curvature The dose of CURT1 essentially dials the geometry of the grana, and because geometry influences how tightly protein complexes are packed and how easily molecules diffuse, it ultimately affects photosynthetic efficiency.
Beyond CURT1, the grana-stroma system behaves like a set of bellows, swelling and shrinking in response to osmotic water fluxes driven by the ion and proton gradients that photosynthesis itself generates. This swelling and shrinking changes diffusional path lengths for mobile electron carriers, adjusts the probability of protein collisions, and even creates separate pH environments in the grana lumen versus the stroma lumen.19PubMed Central. Granal thylakoid structure and function: explaining an enduring mystery of higher plants In other words, the architecture is not a static scaffold; it is an active participant in regulating photosynthesis.
How Thylakoid Membranes Are Built
Chloroplasts are bounded by a double-membrane envelope, and the thylakoids sit inside, separate from both envelope membranes. So how do lipids and proteins get from their site of synthesis to the thylakoid? This remains one of the bigger open questions in chloroplast biology. Three models have been proposed: direct physical connections between the inner envelope membrane and the thylakoid, soluble transfer proteins that shuttle lipid molecules through the stroma, and a vesicle transport system in which small membrane bubbles bud off the envelope and fuse with the thylakoid.20PubMed Central. A brief history of thylakoid biogenesis There is evidence for all three, and they are not mutually exclusive; different components may take different routes. Vesicles have been observed inside chloroplasts under electron microscopy, and genetic screens have identified proteins that seem to be part of a vesicle-trafficking pathway, but the system is not nearly as well understood as vesicle trafficking in the rest of the cell.
Evolutionary Origins and Structural Diversity
Thylakoid membranes are not unique to plants. They first appeared in cyanobacteria, single-celled organisms that invented oxygenic photosynthesis over two billion years ago. Cyanobacterial thylakoids are relatively simple, often running as parallel sheets or concentric rings just inside the cell membrane. The elaborate grana-and-stroma architecture of vascular plants evolved much later, after a cyanobacterium was engulfed by an ancestral eukaryotic cell in the endosymbiotic event that gave rise to chloroplasts.21PubMed. Biogenesis and origin of thylakoid membranes Algae sit somewhere in between, with thylakoid complexity varying widely across different lineages.22PubMed Central. Structure, biogenesis, and evolution of thylakoid membranes
Even within a single plant, thylakoids can look very different depending on which cell type you examine. In C4 plants like maize and sugarcane, which concentrate carbon dioxide through a two-step biochemical relay, the chloroplasts of mesophyll cells and bundle-sheath cells are structurally distinct. Bundle-sheath chloroplasts in many C4 species have reduced or absent grana stacking and dramatically lower levels of the water-oxidizing components of PSII.23PubMed. Differential biogenesis of photosystem-II in mesophyll and bundle-sheath cells of ‘malic’ enzyme NADP(+)-type C4 plants This structural dimorphism reflects the different energy demands of the two cell types: bundle-sheath cells primarily run cyclic electron transport around PSI (producing ATP without NADPH), so they do not need the full PSII apparatus or the grana that house it.24PubMed Central. Electron Microscopy Views of Dimorphic Chloroplasts in C4 Plants
When Thylakoids Disappear
Chloroplasts are one type of plastid, and plastids can change form depending on the tissue and stage of development. One of the most visible examples is fruit ripening. As a tomato turns from green to red, its chloroplasts transform into chromoplasts, specialized organelles that accumulate large amounts of carotenoid pigments. During this transition, the thylakoid membrane system is dismantled: chlorophyll degrades, photosynthesis-related proteins drop, and the orderly stacks of grana break down.25PubMed Central. Maintenance of Chloroplast Components during Chromoplast Differentiation in the Tomato Mutant Green Flesh Electron microscopy of pepper fruits during ripening shows that grana and some connecting thylakoids undergo outright lysis, with the structured membrane losing its particulate appearance before dissolving entirely.26American Journal of Botany. ULTRASTRUCTURE OF CHLOROPLASTS AND CHROMOPLASTS IN CAPSICUM ANNUUM I. THYLAKOID MEMBRANE CHANGES DURING FRUIT RIPENING This is a deliberate developmental program, not random decay; the cell redirects the plastid’s resources toward producing the pigments that attract seed-dispersing animals.
Harvesting Thylakoids for Technology
The ability of thylakoid membranes to generate electron flow in response to light has attracted interest from engineers designing bio-hybrid solar cells and biosensors. Researchers have isolated thylakoid membranes from several plant species and immobilized them on carbon electrodes to measure photocurrent output. Among the species tested, spinach thylakoids produced the largest photocurrent, roughly 80 microamps per milligram of protein.27Electrochimica Acta. Comparative study of thylakoids from higher plants for solar energy conversion and herbicide detection The same electrodes doubled as herbicide detectors: because many herbicides work by blocking electron transport through PSII, exposing a thylakoid-coated electrode to contaminated water causes a measurable drop in photocurrent. The approach is far from replacing silicon solar panels on a rooftop, but it hints at niche applications in environmental monitoring and as proof-of-concept platforms for studying how biological photosynthesis might inform the design of artificial light-harvesting systems.
How Scientists Visualize Thylakoid Architecture
Much of what we know about thylakoid structure comes from a succession of microscopy techniques, each revealing a new layer of detail. Transmission electron microscopy, developed in the mid-twentieth century, first showed that grana consist of stacked, appressed membrane discs. Freeze-fracture and freeze-etch methods in the 1970s and 1980s verified that stroma thylakoids spiral around grana stacks as right-handed helices and confirmed the non-random distribution of protein complexes between the two domains.28PubMed Central. A brief history of how microscopic studies led to the elucidation of the 3D architecture and macromolecular organization of higher plant thylakoids Electron tomography, a technique that reconstructs three-dimensional volumes from a series of tilted two-dimensional images, was what finally confirmed the tubular junctions connecting grana to stroma lamellae.29PubMed Central. Three-Dimensional Architecture of Grana and Stroma Thylakoids of Higher Plants as Determined by Electron Tomography More recently, cryo-electron tomography and atomic force microscopy of hydrated, unfixed membranes have begun to resolve individual protein complexes in their native positions within the membrane, getting us closer to a molecular-resolution map of a working thylakoid.
Each of these advances has corrected earlier oversimplifications. The textbook image of grana as neat, perfectly aligned coin stacks connected by flat stroma sheets has given way to a messier, more biological reality: helical connections, variable stacking heights, irregular disc diameters, and a system that flexes and remodels constantly in response to light, temperature, and developmental signals. That dynamism is not noise in the data. It is how the thylakoid membrane does its job.