Oliver Wendell Holmes coined the term "anesthesia" in 1846 to describe drug-induced insensibility to sensation (particularly pain), shortly after the first publicized demonstration of inhaled ether rendered a patient unresponsive during a surgical procedure. Two broad classes of pharmacologic agents, local and general, can result in anesthesia. Local anesthetics, such as Novocain, block nerve transmission to pain centers in the central nervous system by binding to and inhibiting the function of an ion channel in the cell membrane of nerve cells known as the sodium channel. This action obstructs the movement of nerve impulses near the site of injection, but there are no changes in awareness and sense perception in other areas.
In contrast, general anesthetics induce a different sort of anesthetic state, one of general insensibility to pain. The patient loses awareness yet his vital physiologic functions, such as breathing and maintenance of blood pressure, continue to function. Less is known about the mechanism of action of general anesthetics compared to locals, despite their use for more than 150 years. The most commonly used general anesthetic agents are administered by breathing and are thus termed inhalational or volatile anesthetics. They are structurally related to ether, the original anesthetic. Their primary site of action is in the central nervous system, where they inhibit nerve transmission by a mechanism distinct from that of local anesthetics. The general anesthetics cause a reduction in nerve transmission at synapses, the sites at which neurotransmitters are released and exert their initial action in the body. But precisely how inhalational anesthetics inhibit synaptic neurotransmission is not yet fully understood. It is clear, however, that volatile anesthetics, which are more soluble in lipids than in water, primarily affect the function of ion channel and neurotransmitter receptor proteins in the membranes of nerve cells, which are lipid environments.
Two factors make obtaining a detailed description of how these agents act difficult. The first is that volatile anesthetics, unlike most of the drugs used in medicine, bind only very weakly to their site(s) of action. As a result, high concentrations, often more than 1,000 times greater than for typical receptor- or protein-targeting drugs, are needed to achieve an anesthetic state. This makes it tricky to obtain structural details of anesthetics bound in a specific manner to a protein. It also affects the function of many proteins in nerve cell membranes, making it challenging to ascertain which of them are the key mediators of anesthetic action. A second problem is that volatile anesthetics tend to partition into lipids and exert their primary effects on synaptic neurotransmission by interacting with proteins in a lipid environment. It is harder to gain detailed structural information for membrane proteins than it is for water-soluble proteins. Such structural data are essential for understanding how anesthetics interact with proteins and, more importantly, alter their function. Because of the lack of structural data for membrane proteins both in the presence and absence of anesthetics, it remains unclear whether anesthetics exert their primary effects by direct interaction with these proteins, or indirectly via interaction with the lipids surrounding them.
Despite these limitations, researchers are taking advantage of a variety of methods to better discern how anesthetic agents induce an anesthetic "state" at the molecular level. The term state is in quotes, because a wide variety of agents--ranging from single atoms such as xenon to polycyclic hydrocarbons--can produce insensibility to pain and loss of awareness. The molecular targets for these different agents do not appear to be the same. Thus the notion that there is a single molecular mechanism of action for all anesthetic agents is probably an oversimplification.
Genetic tools are providing promising results regarding the molecular mechanism of anesthetic action. For example, researchers can alter specific protein function and then determine whether this protein can be linked to sensitivity or resistance to anesthetic action in lower organisms. These approaches identify what proteins are involved in anesthetic action and can be thought of as a way to better define relevant anesthetic targets, which investigators can then focus on for structural studies. Other approaches, including sophisticated structural modeling of anesthetic binding to protein targets in a lipid environment and detailed structural determinations of anesthetic binding to soluble proteins, are also showing promise in further revealing the how of anesthetic action at the molecular level.
Thus the simple answer to the question "How does anesthesia work?" is that, although we know a great deal about the physiologic effects and macroscopic sites of action, we don't yet know the molecular mechanism(s) of action for general anesthetics. Many of the tools necessary to answer these questions now exist and we can look forward to new insights into how this great boon to humanity works at the molecular level.