A second second messenger system was discovered in the late 1960s that used not cAMP but cyclic GMP (cGMP) as the second messenger. The mechanism of synthesis of cGMP turned out to be remarkably similar to that of cAMP, namely that a receptor (usually) activated a G protein that activated guanlylyl cyclase, which converted GTP to cGMP, and then it was off to the races. An exception to this scheme arose when it was realized that the gas, Nitric Oxide (NO), is a transmitter molecule. As I mentioned earlier, NO is produced from arginine by the enzyme NO synthase, and once produced can diffuse through the membrane of the cell that makes it and into nearby cells, because it crosses membranes readily (like O2 or CO2 does). It is very labile and thus is quite rapidly inactivated by reactions with water or O2, so it has a lifetime of at most a few seconds. That means that it can only affect cells in its immediate vicinity. It has recently been shown that the way that NO works on its target cells is to activate the enzyme guanylyl cyclase to make cGMP. So cGMP is a second messenger for NO, but NO acts directly on the synthetic enzyme, bypassing the receptor/G protein mechanism. Another gas, CO, is thought to act similarly as a signalling molecule. The discovery of NO as a signalling molecule in the nervous system, adding a gas to the small organic molecules and peptides that were known to be transmitters, was a surprise to workers in this field, including yours truly. cGMP is also a crucial signalling molecule in the rods and cones of the retina; its normally high levels in these cells is reduced in response to light, which causes a closing of cGMP-gated Na channels, which hyperpolarizes the membrane potential of the rods and cones. If you're interested in learning more about this process, chapter 10 of Purves et al., especially pages 186-7, discusses "phototransduction".

Another second messenger/signalling system is based on the metabolism of polyphosphatidyl inositols, especially phopsphatidylinositol di(or bis)phosphate (PIP2). How does this system work? Just as in the other systems, there are three crucial proteins--a receptor, a G protein, and an enzyme (see Fig. 7.11; the muscarinic ACh receptor). The receptor binds a signalling molecule such as acetylcholine, angiotensin, vasopressin, thyrotropin releasing hormone, etc. The now-activated receptor binds to a G protein, catalyzing the usual GTP/GDP swap. Then the Ga subunit binds to and activates an enzyme. (The G protein is related to but different from the Gs discussed earlier for cAMP). The enzyme activated by this G protein is called phospholipase C (don’t ask me why it's "C"--I think it's because they first found phospholipases A and B). As the name of the enzyme implies, its substrate is phospholipids, particularly the phospholipid phosphatidylinositol diphosphate. The enzyme cleaves the PIP2, which releases inositol triphosphate (InsP3) into the cytosol--it’s soluble because its highly charged, but the diacylglycerol (i.e., glycerol with two fatty acid esters attached) remains stuck in the lipid bilayer.

Now the story gets complicated. The InsP3 can diffuse throughout the cytosol, and bind to an InsP3 receptor protein located in the membrane of the endoplasmic reticulum. This InsP3 receptor is a kind of Ca²⁺ selective transmembrane protein--i.e., it is a ligand-gated calcium channel. Normally it is impermeable to Ca²⁺, but when it binds InsP3, it opens up, allowing Ca²⁺ to pass through the ER membrane. Now inside the ER the Ca²⁺ concentration is around 1 mM, while in the cytosol it is more like 100 nM. That is, the concentration in the cytosol is about 10,000 times lower than in the ER, and if Ca²⁺ can diffuse out of the ER, it does. So InsP3 binding to its receptor causes a rapid increase in the [Ca^2+]] in the cytosol of 10-50 fold. And guess what? Ca²⁺ can act as a second messenger molecule for many different cellular processes.

One of the enzymes in the cell that is Ca²⁺ dependent is protein kinase C (C means Ca²⁺ dependent in this case). Normally inactive and in the cytosol, this enzyme is activated when it is bound by two different signalling substances. One is Ca²⁺, which appears to cause it to move from the cytosol to associate with the inner side of the plasma membrane. The other is diacylgycerol (DG), which PKC binds to if it finds DG available in the membrane. When will DG be available? When PLC has produced it in response to a first messenger molecule. Thus, the effects of PLC on PIP2 are to raise the concentration of both Ca²⁺ and DG, both of which are required to activate PKC. PKC, like PKA, is a serine/threonine kinase that phosphorylates these amino acids on protein side chains. It’s especially active in phosphorylating membrane proteins, which typically, changes the membrane’s permeability to certain ions. Other proteins phosphorlated by PKC appear to be gene regulatory proteins that can move into the nucleus when they are phosphorylated and turn specific genes on or off.

A second effect of the rise in Ca²⁺ in the cytoplasm as a result of the InsP3 is the activation of a calcium binding protein called calmodulin. Calmodulin binds 4 molecules of Ca²⁺ and once that happens, it can interact with a bunch of other proteins. A group of these are also protein kinases, called, ingeniously enough, the Ca²⁺/Calmodulin dependent protein kinases (CaM Kinase). Normally inactive, the CaM kinase is activated by calmodulin (which is activated by elevated Ca²⁺) and can phosphorylate a bunch of other proteins, again on serine or threonine residues. CaM kinase appears to be important in release of transmitter by neurons in response to an action potential and at least some synaptic proteins are phosphorlated both by CaM kinase and PKA, though at different places. Again the effects of CaM kinase are widespread but not well characterized, as is the case for the other kinases we have discussed.

Finally, I want to mention another lipid that seems to serve as a second messenger in the nervous system, arachidonic acid. This lipid is generated from membrane lipids by a process that is remarkably similar to the formation of DG. Binding of transmitters such as norepinephrine, 5-HT, glutamate (to metabotropic glutamate receptors), and some peptides (FMRFamide) to their receptors leads to activation of Phospholipase A2 which degrades phospholipids by cutting the fatty acid free from the glyceride. One such fatty acid, arachidonic acid, is the precursor to a number of biologically active compounds such as eicosanoids, prostaglandins, etc. Again it appears that PLA2 requires both a G protein and Ca2+ for activation.

As if this weren’t complex enough, it’s becoming increasingly clear that these various second messenger systems can interact with each other. The point here is that cell signalling is a complex process and scientists have only just begun to scratch the surface of this complexity. However, it’s already clear that the various second messenger systems that exist (I haven’t mentioned them all) don’t operate independently but interact in ways that can either augment a response when two pathways are activated and enhance each other, or cancel a response when the two pathways inhibit each other. (In some cells Ca activates phosphodiesterase and short circuits the response to cAMP, for example). Can you think of reasons why nature might want to design these complex interactive pathways?

Finally, the systems I’ve been describing must be turned off as well as turned on. How does that happen? Well, you already know from our previous discussions that there are Ca²⁺ "pumps" in the cell that can use ATP to move Ca²⁺ across membranes or that can exchange Na and Ca. Such pumps exist in the ER membrane, the plasma membrane and the mitochondria. Usually the fastest transport is accomplished through the plasma membrane and almost as fast through the ER. The mitochondria seem to be an emergency pathway that only kicks in when the other mechanisms are overwhelmed by high Ca²⁺. The mitochondria take up Ca temporarily, at the expense of the ATP they’ve made, but they are not healthy with high internal Ca²⁺.^. So they must eventually excrete the Ca back into the cytosol so it can be moved into the ER or out of the cell or they will self-destruct. As the Ca²⁺ levels in the cytosol return to the resting level of 100 nM, Ca²⁺ unbinds from the various proteins that it bound to, such as calmodulin, PKC, etc. And these proteins therefore become inactive again. Eventually phosphatases remove the phosphate groups from the various proteins that got phosphorylated by these kinases, returning those proteins to their resting state. (Fig. 7.12)

There are phosphatases that degrade cGMP and InsP3 as well, and so these inactivated, usually pretty quickly. Thus the concentration of these compounds only remains high as long as they are being synthesized, which in turn depends on the continued presence of the extracellular signal--the neurotransmitter. If its concentration falls, the concentration of the intracellular signal also falls fairly quickly as well, which turns off the complex cascade of events that is set in motion by the rise of the intracellular signal.

What do all the following substances have in common?:

They’re all substances that affect second messenger pathways in cells. The first two affect the PIP2 system and the others affect the cAMP system. One way to determine what the second messenger is for a system is to find out whether one or more of these substances can interfere with or simulate the response.

Phorbol esters are called tumor promoters because they increase the likelihood of an animal’s getting cancer from some other cause, though they don’t cause cancer directly. They bind to and activate protein kinase C, in the absence of diacylglycerol, thus bypassing one arm of the second messenger response. If phorbol esters can mimic a response, then it’s a good bet that PKC is part of the response pathway. Similarly the Ca²⁺ ionophores can allow Ca²⁺ out of the ER and thus mimic the effect of raising InsP3 levels in the cell.

All the other substances increase the levels of cAMP in cells. Forskolin directly stimulates adenlylyl cyclase in the absence of a first messenger; caffeine and theophylline, from coffee and tea, respectively, inhibit phosphodiesterase and thus prevent breakdown of cAMP; cholera toxin binds to the Gs protein and prevents it from hydrolyzing GTP, so it’s locked in the on position and continuously activates adenylyl cyclase; and pertussis toxin prevents the activation of Gi, which would normally inhibit cAMP production. Again if these compounds mimic or augment a natural response, one tentatively concludes that cAMP is involved in that response. The use of such activators and inhibitors has proved invaluable in establishing which second messengers are part of which responses; similar strategies are used in all signaling systems that have been studied.