Cyclic AMP and Protein Kinase A Explained

A peptide hormone arrives at a cell and never goes inside. It binds a receptor embedded in the membrane, and within seconds the interior of the cell is doing something different — genes are switched on, enzymes are phosphorylated, hormones are released. Something has to carry the message across, and in 1957 Earl Sutherland’s laboratory found the molecule that does it: a small, cyclised derivative of ATP, now called cyclic AMP.

That cascade is the single most useful piece of cell biology for reading research-peptide literature, because it is what the phrase “Gs-coupled receptor” actually means. This guide walks through it, names the enzymes and genes involved, and then does the thing most summaries skip: separates which peptide receptors genuinely use this route from which ones use something else entirely. Two compounds in the same product category can work through completely different second messengers, and conflating them is a common error.

Where it came from

The first report was a short 1957 communication in the Journal of the American Chemical Society from Sutherland and Rall, describing the properties of an adenine ribonucleotide produced with cellular particles, ATP, magnesium and — note the title — epinephrine or glucagon. Glucagon is a peptide hormone, so the molecule was tied to peptide signalling from the very first paper. Two fuller papers followed in 1958 in the Journal of Biological Chemistry, one on the formation of the cyclic nucleotide by tissue particles and one on its fractionation and characterisation.

Sutherland received the Nobel Prize in Physiology or Medicine in 1971, and the citation is worth quoting exactly because it is almost always paraphrased wrongly: it reads “for his discoveries concerning the mechanisms of the action of hormones”. Not “for the discovery of cyclic AMP”. The prize was for the concept — that hormones act through intracellular second messengers — rather than for the molecule.

The kinase came eleven years later. Walsh, Perkins and Krebs described an adenosine 3′,5′-monophosphate-dependent protein kinase from rabbit skeletal muscle in 1968. (The published title spells it “dependant”; that is as printed.) Edmond Fischer and Edwin Krebs shared the 1992 Nobel Prize in Physiology or Medicine “for their discoveries concerning reversible protein phosphorylation as a biological regulatory mechanism” — the other half of the idea, that a phosphate group added and removed is how cells switch proteins on and off.

The cascade, and its three off switches

The forward direction is compact. A peptide hormone binds a G-protein-coupled receptor. The receptor activates a heterotrimeric G protein; if that G protein is Gαs, it stimulates adenylyl cyclase. Adenylyl cyclase converts ATP into cAMP. cAMP binds the regulatory subunits of protein kinase A, which releases the catalytic subunits, and those phosphorylate serine and threonine residues on target proteins. For the receptor end of this, see our guide to GPCRs and peptide receptors.

There are ten adenylyl cyclase genes in the human genome, ADCY1 through ADCY10, and the split matters: nine are transmembrane isoforms and one, ADCY10, is soluble. (A naming trap: ADCYAP1 and ADCYAP1R1 are PACAP and its receptor, not cyclases, and should not be counted among them.)

What makes the system a signal rather than a switch stuck on is that it has three independent ways of being turned off:

What PKA actually is

Protein kinase A is not a single protein sitting ready to act. In its inactive state it is a four-subunit complex: two regulatory chains and two catalytic chains. The protein record puts it plainly — the inactive holoenzyme is composed of two regulatory chains and two catalytic chains. cAMP binds the regulatory pair, four molecules in total, two per regulatory subunit, and the complex comes apart, releasing two free catalytic monomers that are then free to phosphorylate substrates.

Human PKA has three catalytic subunit genes (PRKACA, PRKACB, PRKACG) and four regulatory ones (PRKAR1A, PRKAR1B, PRKAR2A, PRKAR2B), which produce the two classical holoenzyme types, PKA type I and type II. There are also dedicated endogenous inhibitors, encoded by PKIA and PKIB.

The part worth dwelling on is location. PKA is tethered around the cell by a family of scaffolds called A-kinase anchoring proteins — more than thirty of them identified, all sharing a short motif that binds the PKA regulatory subunits. AKAPs do not merely hold PKA in place; they also bind adenylyl cyclases, phosphodiesterases, phosphatases and other kinases, assembling local signalling complexes, and depending on what they assemble they can either amplify or dampen the response.

And here is the unresolved tension, stated by the authors of a major review rather than by us: cAMP’s diffusion coefficient is high enough that it ought to equilibrate across a cell rapidly, which conflicts with the observed compartmentalisation. The proposed explanations are cytosolic buffering, viscosity, physical barriers and local phosphodiesterase activity within nanodomains — and the same review says how phosphodiesterases achieve this is still debated. So the honest position is in the middle: “cAMP floods the cytosol uniformly” is outdated, but “cAMP does not diffuse” is equally wrong.

Reaching the nucleus: CREB

Many of the interesting consequences of cAMP signalling are changes in which genes are transcribed, and the classical route there runs through CREB — cAMP response element binding protein. Freed PKA catalytic subunits enter the nucleus and phosphorylate CREB at serine 133. Phosphorylated CREB then recruits coactivators, notably CBP, which strengthens binding at cAMP response elements and drives transcription of target genes. The 1989 paper that established the site is titled, straightforwardly, “Cyclic AMP stimulates somatostatin gene transcription by phosphorylation of CREB at serine 133”. The CBP interaction was reported in 1993.

It is worth noticing which gene that 1989 title names. The first CREB target characterised this way was somatostatin — a peptide hormone whose own receptors inhibit adenylyl cyclase. The pathway regulates its own antagonist, which is a recurring motif in signalling biology.

What the textbook diagram leaves out

Two things, and both are worth knowing because they are where the simple account breaks.

cAMP has effectors other than PKA. In December 1998, two groups independently reported proteins that cAMP activates directly: de Rooij and colleagues in Nature described Epac as a Rap1 guanine-nucleotide-exchange factor directly activated by cyclic AMP, and Kawasaki and colleagues in Science described a family of cAMP-binding proteins that directly activate Rap1. The word “directly” is doing the work in both titles — the point was that cAMP acts without going through PKA. EPAC works by cAMP binding its regulatory domain, relieving autoinhibition and exposing a catalytic exchange domain. Beyond EPAC, cyclic-nucleotide-activated ion channels form a third class of direct effector, some of which respond to cAMP; Popeye-domain proteins have been proposed as a fourth.

And cAMP is not the only second messenger peptide hormones use. This is the correction with the most practical consequence for reading research-peptide claims, so it gets its own section.

Three peptide receptors, three directions

Take three receptor families whose ligands all show up in the same corner of a research catalogue. All three bind peptides. All three are relevant to growth-hormone biology. They do completely different things to cAMP.

Two cautions on how far to push this. First, neither contrast is absolute. The same review that establishes the GHRH–cAMP chain also reports that GHRH activates phospholipase C in parallel, raising inositol trisphosphate and mobilising calcium — so “GHRH works only through cAMP” overstates it. In the other direction, a review dedicated to ghrelin-receptor signalling labels the Gαs/cAMP/PKA route explicitly as the “debated” pathway: there are positive reports, but they come largely from transfected systems and receptor heterodimers, and one cited study found ghrelin alone had no effect on intracellular cAMP. The accurate framing is principal route versus principal route, not “cAMP versus never cAMP”.

Second, this is a receptor-level comparison. We verified how these receptor families signal; we did not verify receptor selectivity, potency or any other pharmacological property of the individual compounds named in the second column, and nothing here should be read as characterising them beyond their membership of a class. Our guides to growth hormone peptides, sermorelin versus ipamorelin and GHRP-6, GHRP-2 and hexarelin go into the compounds themselves.

One claim this section lets us put to rest. It is widely repeated that growth hormone secretagogues raise GH by suppressing somatostatin. The review usually cited for GHS safety and efficacy reports in its abstract that these compounds stimulate pulsatile GH release while preserving negative feedback — which is the opposite of suppressing the brake. Our somatostatin guide is the site’s reference page for that correction.

And a third peptide receptor, for completeness, because it is on the cAMP route: the VIP receptors VPAC1 (VIPR1) and VPAC2 (VIPR2) are both recorded as activating a cAMP-dependent pathway, with different ligand potency orders — VPAC1 prefers VIP and PACAP27 over PACAP38, VPAC2 the reverse. VPAC2 may also couple to phospholipase C, so “purely Gs” would again be too strong. See our VIP research guide.

Laboratory tools, and their caveats

How you manipulate this pathway in an experiment constrains what the experiment can conclude, and every standard tool has a documented limitation.

The broader point sits alongside the one in our guide to reading a peptide study: an inhibitor that hits eight kinases cannot by itself establish which kinase mattered. A 2000 survey of commonly used protein kinase inhibitors tested twenty-eight compounds against a large kinase panel and concluded that although some were relatively selective, they still inhibited two or more kinases with similar potency — and that conclusions drawn from cell-based experiments using several of them are likely to be erroneous.

What is actually approved

No approved medicine that we could find acts as a direct cAMP or PKA activator. What is approved works on the other side of the equation, by blocking cAMP breakdown — and there is a distinction here that gets confused constantly, so it is worth settling with regulator documents.

PDE5 inhibitors act on cGMP, not cAMP. Sildenafil’s approved label describes nitric oxide activating guanylate cyclase to raise cyclic guanosine monophosphate, and sildenafil enhancing that effect by inhibiting PDE5, the enzyme responsible for degrading cGMP. cAMP is not mentioned anywhere in its mechanism-of-action section.

PDE4 inhibitors act on cAMP. Roflumilast’s label describes selective PDE4 inhibition — PDE4 being a major cAMP-metabolising enzyme in lung tissue — leading to accumulation of intracellular cAMP; the label is candid that the specific mechanism of its therapeutic action in COPD is not well defined. Apremilast’s label likewise describes PDE4 inhibition specific for cAMP, with PDE4 inhibition resulting in increased intracellular cAMP.

These are prescription medicines used under medical supervision, and nothing here describes or recommends their use. We state the fact of their approval only because it calibrates the pathway: raising cAMP is a validated drug mechanism, achieved by blocking its destruction rather than by pushing on the cyclase.

What’s established and what isn’t

Sourcing signalling reagents in Canada

If you are buying peptides for receptor-signalling work, the documentation questions are the usual ones, with one that matters particularly for this pathway.

• Identity, not just purity. A cAMP readout cannot tell you which peptide produced it, so an identity check by mass spectrometry is carrying real weight in these experiments. See reading a mass spectrum and reading an HPLC chromatogram.
• Match the batch number on the certificate to your vial — our COA guide covers what else a useful certificate contains.
• Account for net peptide content, since synthetic peptides arrive as salts carrying counter-ions and water, so labelled mass is not peptide mass: net peptide content explained. Our peptide calculator works through the arithmetic.
• Store cold, dry and dark, and keep material lyophilized until needed: the storage guide and what lyophilized means.

SermorelinA GHRH analogue — the receptor family that activates adenylyl cyclaseView product CJC-1295 with DACGHRH-receptor ligand carrying a drug affinity complexView product CJC-1295 No DACThe same receptor family, without the albumin-binding elementView product VIPActs at VPAC1 and VPAC2, both recorded as cAMP-dependentView product

Frequently asked questions