A receptor is a switch with two faces
In the last guide you met the basic grammar of cell signaling: a [[molbio-signaling-ligand|ligand]] arrives, a receptor catches it, and the news is relayed inward, often through a second messenger that spreads and amplifies the alarm. We left one question open, and it is the most important one. The ligand — adrenaline, insulin, a steroid, a whiff of an odorant — never enters the cell to do its job directly. Almost all of them stop at the door. So everything depends on the door itself: the receptor. Change the receptor and you change the whole reply, even if the message is the same.
Think of a receptor as a switch with two faces. One face sticks out of the cell, into the wet world outside, and is shaped to grip exactly one kind of ligand — the same lock-and-key (really induced-fit) recognition you saw with enzymes and their substrates. The other face hangs inside the cell. Binding on the outside changes the shape on the inside — this is just allostery again, the principle that a protein can carry news from one site to a distant site by changing its fold. That single trick, a shape-change passed across the membrane, is how a message gets from outside to inside without the messenger ever crossing.
GPCRs: the seven-pass switchboard
The [[molbio-g-protein-coupled-receptor|G-protein-coupled receptor]], or GPCR, is the largest receptor family of all — humans have roughly 800 of them, and they are how we sense light, smell, taste, and a great many hormones. A single chain of protein threads back and forth through the membrane seven times, like a thread sewn through cloth, leaving a ligand-pocket facing out and a loop facing in. Bind the ligand, and the seven-pass bundle twists. That twist is felt by a partner waiting on the inside: a [[heterotrimeric-g-protein|heterotrimeric G protein]], a three-part relay that does the actual broadcasting.
Here is the elegant part. The G protein is a molecular timer. At rest it holds a molecule called GDP and does nothing. When the activated receptor nudges it, it swaps GDP for GTP — and now it is "on." It splits into pieces that go switch on other machines: one common target makes the second messenger cyclic AMP, another opens a flood of calcium. Crucially, the G protein then slowly hydrolyses its own GTP back to GDP and shuts itself off — a built-in self-timer that ends the signal even if the ligand is still bound. This same GTP-on, GDP-off logic runs all through signaling; hold on to it.
Notice what GPCRs buy the cell: speed and amplification. One bound receptor can activate many G proteins before its ligand falls off; each G protein switches on an enzyme that churns out thousands of second-messenger molecules. A single photon hitting one rhodopsin GPCR in your eye ends up shutting millions of channels — that is the gain that lets you see a dim star. The price of this branching, broadcast design is that GPCRs are not made for carrying one precise instruction to one precise place; they are made for flooding the cell with a fast, loud, short-lived shout.
RTKs: catch a partner, then write in phosphate
The [[molbio-receptor-tyrosine-kinase|receptor tyrosine kinase]], or RTK, takes a completely different approach — and it is the receptor for growth signals like insulin and the growth factors that tell a cell to divide. An RTK passes through the membrane just once. Its outside catches the ligand; its inside is itself an enzyme, a kinase, whose job (recall phosphorylation from the protein rung) is to attach phosphate groups onto the amino acid tyrosine. But a lone RTK is inert. The trick is that the ligand makes two receptors pair up, side by side.
- A ligand binds and pulls two receptor molecules together into a pair — dimerization. This is the on-switch: alone they were quiet, but together their inner kinase tails can finally reach each other.
- The two kinases phosphorylate each other — cross-phosphorylation — studding their own inner tails with phosphate-tyrosine marks. The receptor has, in effect, written a fresh address label on itself.
- Inside the cell, relay proteins recognize those new phospho-tyrosine marks and dock onto them, like cars parking only in freshly painted spots. The receptor has become a recruiting platform.
- One docked relay flips on [[ras-gtpase|Ras]] — a small GTPase that, like the G protein, is on with GTP and off with GDP — which fires the kinase cascade we meet in the next guide.
So an RTK does not broadcast a flood of small messengers; it builds a precise, addressable platform that hands the signal to named proteins. That precision is exactly what a slow, weighty decision like "divide" demands. It is also why RTKs are dangerous when broken: a mutation that locks an RTK or its Ras in the "on" state tells the cell to grow with no ligand at all, and such stuck-on growth signals are among the most common drivers of cancer. We will return to this in the crosstalk guide.
Ligand-gated channels: the fastest answer
The third surface family skips relays entirely. A ligand-gated ion channel is a receptor and a pore in one. Its several subunits form a ring around a central tunnel that is normally shut. When the ligand binds the outside, the whole ring twists open for a few thousandths of a second, and ions — sodium, potassium, calcium — pour through down their gradients. There is no second messenger, no kinase, no G protein in between. Binding *is* the response.
Because there is nothing to relay, this is the fastest signaling there is — it is how nerve and muscle cells talk. A neurotransmitter released into the gap between two neurons crosses in well under a millisecond, opens these channels, and the rush of ions changes the next cell's voltage almost instantly. The trade-off is the mirror image of the nuclear receptor we meet next: ligand-gated channels are blazingly fast but their effect is local and fleeting. They change a voltage, not a gene. When the ligand leaves, the pore snaps shut and the cell is reset, ready for the next pulse.
Nuclear receptors: when the receptor is also the switch on the gene
Now the exception we promised. Steroid hormones — cortisol, estrogen, testosterone — and a few cousins like thyroid hormone and vitamin D are small and oily. Recall the hydrophobic effect: the lipid membrane is an oily film, and an oily molecule passes through it freely, the way a drop of oil melts into butter. These ligands need no surface receptor at all. They drift straight through the membrane and meet their [[molbio-nuclear-receptor|nuclear receptor]] waiting inside the cell.
And here is the beautiful shortcut. A nuclear receptor is itself a transcription factor — a protein that binds specific DNA sequences and turns genes on or off. So there is no relay chain to build at all. The hormone binds the receptor; the receptor changes shape, moves to the DNA, lands on its target genes, and switches transcription up or down directly. The same outside-in shape-change you saw with GPCRs is happening, but the receptor's inner face does not nudge a G protein — it walks to the genome and acts. Outside signal becomes gene expression in essentially one step.
This explains the timing. A GPCR or channel changes the cell in seconds, but its effect fades when the ligand leaves. A nuclear receptor is slow — it takes hours, because changing which genes are read, then making new RNA and new protein, simply takes time — but its effect is deep and lasting. This is why steroids act over hours to days, not seconds. Honest caveat: the clean split is a useful first picture, not a law. Some nuclear receptors also act in the cytoplasm in minutes, and surface receptors very much do reach the genome — just more slowly, through the relays of the following guides. The reliable rule is the design logic, not a stopwatch.
One picture, four doors
Step back and the whole family portrait clicks into place. The same ligand can mean opposite things in different cells purely because they carry different receptors — adrenaline speeds the heart through one GPCR and relaxes the gut through another. The receptor, not the message, chooses the reply. And notice the recurring theme: a shape-change carries the news across, and a GTP/GDP timer (the G protein, Ras) keeps switching on and off so the cell can both start and stop. Hold these four doors in mind and the relays in the next guides — second messengers, kinase cascades, the MAP kinase pathway — are just the corridors that lead inward from each one.
LIGAND -> RECEPTOR ----------------------> RESPONSE SPEED
adrenaline GPCR (7-pass) -> G protein -> 2nd messenger seconds
insulin RTK (1-pass) -> Ras -> kinase cascade minutes
acetylcholine ligand-gated channel -> ions flow < 1 ms
cortisol nuclear receptor = transcription factor hours
(ligand crosses membrane; receptor binds DNA)