Local anaesthetics stop pain, but block all other sensations too. In rats, one molecular delivery vehicle makes an unusual local anaesthetic specific for pain — provided a little spice is added to the mix first.
Medicine has no shortage of great anaesthetics: they have been making surgery tolerable by eliminating consciousness, or by blocking complete nerve systems, for 160 years. But what we do need more of are good analgesics: drugs that suppress pain without affecting any other sensation. On page 607 of this issue, Binshtok and his colleagues — based at the Massachusetts General Hospital in Boston, where surgical anaesthesia was first demonstrated in 1846 — report a new approach to analgesia1. They have found a way to deliver a local anaesthetic so that it blocks pain sensing alone.
To those in the know, an anaesthetic seems the last place to look for an analgesic effect. When you are numbed for the dentist's drill, the anaesthetic used blocks all the voltage-gated sodium channels in a nerve. These are proteins that conduct action potentials in nerve axons by timing, with submillisecond precision, the flow of current in the form of sodium ions across cell membranes. Blocking all sodium channels blocks all sensation; it blocks movement if the affected nerve includes motor axons, or the responses of internal organs if a controlling 'autonomic' nerve is exposed. Recently, however, certain sodium channels have been identified as being present only in pain-sensing neurons2. Drug companies have thus begun to comb their libraries of uncommercialized local anaesthetics for a 'magic bullet' drug that affects pain alone.
Binshtok et al.1 follow a different path. They study the TRPV1 ion channel, which is opened up by capsaicin — the chemical that makes chilli peppers spicy and creates the burning sensation when they are rubbed on skin — or by temperatures that rise above 42 °C, the skin temperature that we consider unpleasantly warm. The temperature-dependent gating of TRPV1 and its specific expression profile, as demonstrated in a series of animal and human studies3, show that it is a molecular sensor for noxious heat that is expressed only on small nociceptors. Nociceptors are sensory neurons that translate noxious stimuli into action potentials and conduct these electrical signals from the site of stimulation to the spinal cord. Small nociceptors conduct slowly (at about 1 m s−1), and TRPV1-positive neurons therefore mediate slowly developing, persistent pain — anything from the half-minute of smarting after stubbing a toe to the never-ending discomfort of an arthritic joint.
Crucial to the story is that, although the pore of the TRPV1 channel rejects negatively charged anions, it promiscuously allows most positively charged inorganic cations to pass — and even organic cations as large as some local anaesthetics. In most cases, this is irrelevant: local anaesthetics are generally weak acids that occur in both a positively charged (protonated) form and an uncharged, 'free-base' form. As free bases, local anaesthetics readily permeate lipid cell membranes without the need to pass through the pores of an open ion channel. This is why they shut down sodium channels in all neurons; indeed, the potency of an anaesthetic rises in proportion to its lipid solubility.
Once an anaesthetic has passed through the membrane, it blocks the sodium channel from within by docking in a wide hydrophobic vestibule just to the inner side of the sodium selectivity filter, the narrowest part of the channel's pore4. This inner vestibule, whose existence was first deduced in potassium channels5 and has now been seen in their crystal structure6, is common to all voltage-gated ion channels and is the binding site for many different drugs used against various voltage-gated channels7.
But there is one particular anaesthetic, known as QX-314, that cannot infiltrate lipid membranes under its own steam: it has a permanent positive charge, making it lipid-insoluble. As a result, it fails to block sodium channels when applied outside cells, although it succeeds when injected into them. Once inside, QX-314 binds within the hydrophobic vestibule just like any other anaesthetic. In fact, kinetic features of QX-314's blocking mechanism — dependence on the direction of current flow and on membrane voltage — were the initial evidence for the vestibule binding site of anaesthetics in general.
What if QX-314 could be delivered to the insides of cells through the large, cation-selective pores of the TRPV1 channel? By exploiting the fact that TRPV1 resides only on nociceptors, this might inhibit pain without other side effects. And Binshtok et al.1 found that, although neither capsaicin nor QX-314 alone could shut off the sodium channel or action potentials, applied simultaneously to nociceptors they could do just that (Fig. 1). When injected into the foot or perfused onto a nerve of a rat, the mixture inhibited the animal's sensitivity to noxious thermal and mechanical stimuli without causing paralysis. This analgesia lasted about 2 hours.
The significance of this discovery might go beyond just the blockade of sodium channels. First, the same sort of strategy might prove useful for delivering other drugs to other voltage-gated channels. Second, local anaesthetics are more than simple ion-channel blockers. At low concentrations, they exhibit a clinically important selectivity, known as 'use-dependent inhibition', by binding with greater affinity to inactivated sodium channels than to those poised to open8. Low concentrations of anaesthetic thus selectively inhibit sodium channels in cells that are frequently firing action potentials, because those channels cycle more frequently through the inactivated state.
Such selective targeting of hyperactive cells explains why the local anaesthetic lidocaine, when perfused into the entire body at very low concentrations, suppresses certain cardiac arrhythmias without affecting nerves. Pathological pain — pain that is persistent but not caused by an existing injury — is another example of hyperactivity, this time in nociceptors. Nerve blocking by local anaesthetics is sometimes used to treat pathological pain because relief from the pain persists long after the initial numbness wears off; this might be because low, residual levels of local anaesthetic selectively inhibit nociceptors that are too active9. If Binshtok and colleagues1 are right, the humble local anaesthetic may thus prove to have four mechanisms for specificity: through infusion onto a particular nerve for its most basic application; through use-dependence, as a remedy for electrical hyperactivity; through the targeting of distinct subtypes of sodium channel to alter the activity of distinct cells that express them; and now through the TRPV1 channel, a nociceptive drug-delivery vehicle that yields an analgesic effect.
But before we get carried away, the TRPV1 trick must first be shown to work in humans. In addition, the ideal cocktail of capsaicin and TRPV1 modulators must be found, in order to avoid any damage that capsaicin would cause at excessive levels. It will be useful, too, to optimize the use-dependence and TRPV1-permeability of QX-314. This drug, which was until now just an exotic reagent used by ion-channel biologists, will be the focus of a new effort in the search for better analgesics.