Library › DNA & Proteins › Heart and vessels — molecules
Heart and vessels — molecules: 20 3D models
DNA, RNA, nucleic-acid drugs, gene editing, proteins and 300+ real PDB structures (real sizes 1–100 nm). Each one is a ready-made 3D model you can turn, recolour and place in a figure — open BioDiorama and type its name in the Library search.
- Why low-dose aspirin works — One enzyme in two cells, and the difference between them is the whole mechanism. Arachidonic acid comes out of the membrane and up a long channel to the active site of cyclo-oxygenase; aspirin acetylates a serine part way up that channel, and nothing takes the acetyl group off again. A platelet has no nucleus, so it can never make another copy — its thromboxane is gone for its ten-day life. An endothelial cell transcribes a new one within hours and goes on making prostacyclin, which does the opposite. Which is why seventy-five milligrams a day works and a painkilling dose is no better. (COX-1 ≈ 70 kDa per subunit)
- LDL receptor, open and shut — The receptor that clears cholesterol from the blood, drawn at the two pH values that matter. At the cell surface it stands extended with its seven ligand-binding repeats reaching up, each holding a calcium ion, and an LDL particle caught on repeats 4 and 5 by the basic patch of apoB-100. Inside the acidified endosome the beta-propeller folds down onto those same repeats and takes the particle's place, so the particle is let go and the receptor goes back for another. (receptor ≈ 20 nm tall; LDL ≈ 22 nm)
- PCSK9 taking the LDL receptor down — PCSK9 clamps onto one small domain of the LDL receptor — EGF-A — and nothing else. That is enough: the receptor can no longer fold shut to release its cargo, and the whole complex is sent to the lysosome instead of back to the surface. Fewer receptors, more LDL in the blood. An antibody that covers the flat face PCSK9 presents to EGF-A sends the receptor back up instead, which is why blocking one protein-protein contact lowers cholesterol so steeply. (PCSK9 ≈ 75 kDa)
- An LDL particle, modified — What turns a harmless particle into one the plaque will keep. Oxidation clips the polyunsaturated tails short and leaves reactive aldehyde ends sticking out of the surface, and those aldehydes then stick onto the lysines of apoB — so the particle no longer fits the LDL receptor but does fit the scavenger receptors, which have no feedback and will keep eating. Glycation in diabetes does the same job with sugars. Choose the modification. (≈ 22 nm across)
- Lipoprotein(a) — An LDL-like particle with a second protein tied to it: apo(a), a chain of kringle domains ending in an inactive protease domain, joined to apoB-100 by a single disulfide. The kringles carry lysine-binding sites, which is how the particle sticks to fibrin and to the artery wall, and it carries far more oxidised phospholipid than LDL does. The number of kringle repeats is inherited and sets the blood level, which diet and statins barely move. (particle ≈ 25 nm; apo(a) 300–800 kDa)
- The lipoproteins, at one scale — All five classes drawn to the same scale, which is the thing a diagram usually hides: a chylomicron is as much bigger than LDL as LDL is than a large protein. Each has a phospholipid surface with free cholesterol in it and a core of cholesteryl ester and triglyceride, and each carries one copy of apoB — apoB-48 on a chylomicron, apoB-100 on the rest — except HDL, whose apoA-I runs round the rim as two belts. (chylomicron 75–1200 nm down to HDL 8–12 nm)
- HMG-CoA reductase, with a statin in it — The rate-limiting enzyme of cholesterol synthesis: eight helices through the ER membrane at the sterol-sensing end, and a catalytic tetramer above it with four active sites at the dimer interfaces. A statin works by being two things at once — an HMG-like head that sits where the substrate sits, and a bulky hydrophobic group that fills the channel the coenzyme A arm needs. Less cholesterol made, more LDL receptors put out, less LDL in the blood. (tetramer ≈ 97 kDa per subunit)
- NPC1L1 on the brush border — How cholesterol gets out of the gut and into the body: NPC1L1 stands between the microvilli of an enterocyte with a sterol-shaped pocket in its large N-terminal domain, takes a cholesterol molecule out of a bile-salt micelle and passes it down into the cell. Ezetimibe sits in that pocket instead, and the sterol stays in the micelle and leaves in the stool — which is why the drug works on absorption and not on synthesis. (NPC1L1 ≈ 145 kDa)
- ABCA1 making an HDL particle — Reverse cholesterol transport begins here. ABCA1 — twelve membrane helices in two halves, with a tunnel formed by its extracellular domains — spends ATP to push cholesterol and phospholipid out of the cell onto apoA-I, which wraps the growing patch as two belts and lifts away as a disc. The disc is a nascent HDL; only after it fills with cholesteryl ester does it round out into the spherical particle a blood test measures. (disc ≈ 10 × 5 nm)
- CETP, bridging two particles — Cholesteryl ester transfer protein is a boomerang about six nanometres long with a hydrophobic tunnel running through it and a phospholipid plugging each end. It dips one end into an HDL particle and the other into LDL and moves cholesteryl ester one way and triglyceride the other — the reason a high triglyceride level lowers HDL cholesterol. Block the tunnel and HDL cholesterol rises steeply, which turned out not to be the same thing as preventing heart attacks. (≈ 60 Å long)
- Lipoprotein lipase, held out in the flow — Lipoprotein lipase is made by muscle and fat but works in the capillary lumen, and it gets there on GPIHBP1 — a small endothelial protein with a long acidic tract and an LU domain that carries the lipase across the cell and then holds it out into the blood. The lid over its catalytic groove swings open against a triglyceride-rich particle and fatty acids come off for the tissue underneath. Lose either protein and the blood turns milky with chylomicrons. (LPL ≈ 55 kDa)
- Cholesterol crystals — Free cholesterol in the necrotic core comes out of solution as cholesterol monohydrate — thin rhomboid plates with a notched corner, and needles. They are what the clefts in a stained section are the ghosts of, and they are not inert: a needle taken up by a macrophage pierces the phagolysosome from the inside, the contents leak into the cytosol, and ASC gathers below to build an inflammasome. That is one of the ways a plaque keeps itself inflamed. (plates 10–100 µm)
- Von Willebrand factor unfurling — The only protein in the blood that is switched on by being pulled. A multimer circulates as a ball with its platelet-binding A1 domains tucked inside; where flow is fast and the wall is torn it stretches into a long string and those A1 domains come out, so platelets catch on it from the stream. ADAMTS13 keeps the strings short by cutting the A2 domains — and when it cannot, the strings stay long and platelets clump in small vessels. (multimer up to 20 µm when stretched)
- Tissue factor starting the clot — The switch that begins coagulation outside a vessel. Tissue factor sits in the membrane of cells that blood should never touch; when it does, factor VIIa docks on it — its Gla domain pinned flat to the membrane by a row of calcium ions, its protease domain held up at exactly the height that factor X arriving along the same membrane presents its cleavage site. One cut, and factor Xa is made. Everything downstream depends on that geometry. (TF ≈ 47 kDa)
- Fibrin assembling — Fibrinogen is trinodular — two D domains and a central E — and it will not polymerise until thrombin cuts the fibrinopeptides off that central domain. That uncovers knobs which fit holes in the D domains of the next molecule along, and because the fit is end-to-middle the growing protofibril is two strands half-staggered. Factor XIIIa then cross-links it, which is what makes a clot tough rather than merely sticky. (fibrinogen 45 nm long)
- The prothrombinase complex — Thrombin is not made in the blood — it is made on a surface. Factor Xa tethers its Gla domain to the phosphatidylserine of an activated platelet, sits on factor Va as a platform, and cuts prothrombin held alongside it. Assembling the complex speeds the reaction up by five orders of magnitude, which is why a clot forms where the platelets are and nowhere else. A direct oral anticoagulant fills Xa's S1 pocket and the cut is not made. (complex ≈ 15 nm)
- The platelet integrin switching on — Integrin alphaIIb-beta3 is the most abundant protein on a platelet and it spends most of its life bent over and shut. Activation straightens it: the legs extend, the head swings up, and the metal ion at the ligand site becomes able to hold the end of a fibrinogen molecule — which has two such ends, so it bridges one platelet to the next. That bridge is the aggregate. Block the head and platelets activate but cannot stick together. (integrin ≈ 23 nm extended)
- P2Y12 and the drugs that block it — The receptor that makes platelet activation self-sustaining: ADP released from one platelet's dense granules binds P2Y12 on the next, Gi turns cyclic AMP down, and the platelet commits. The two drug classes block it differently and the difference is clinical — a thienopyridine metabolite forms a covalent bond to a cysteine and the receptor is finished for the platelet's life, while a reversible blocker sits in its own pocket and washes out in days. (receptor ≈ 42 kDa)
- Lysis, on the clot and nowhere else — Fibrinolysis is kept local by both of its proteins having to land on the same surface first. tPA and plasminogen each bind lysines on fibrin, which brings them close enough for plasmin to be made right on the fibre it is going to cut — in free plasma the reaction is hundreds of times slower. Plasmin then chews the mesh apart and the fragment that comes away with a cross-link still in it is what a D-dimer test measures. (fibre ≈ 100 nm across)
- The brake: thrombomodulin and protein C — Why a clot stops at the edge of a wound. Thrombin that drifts onto healthy endothelium is caught by thrombomodulin, and docking there turns its active site away from fibrinogen and towards protein C, which EPCR holds up to be cut. Activated protein C then goes back and cuts factors Va and VIIIa apart — so the same enzyme that built the clot is made to dismantle the machinery that was building it. Lose this switch and thrombosis follows. (thrombomodulin ≈ 60 kDa)