The volume knob for pain: one sodium channel and the search for new painkillers

Part 1: A life without pain

1In the early 2000s, doctors studying several related families in northern Pakistan met children who had never felt pain. They walked on hot surfaces, bit their tongues and broke bones without noticing. Their senses of touch, warmth and pressure were normal; they simply did not experience pain. Far from being a gift, the condition was dangerous. Injuries went untreated and became infected, and joints were damaged by repeated injuries that a child who felt pain would have avoided.

2Genetic studies of the families traced the condition to mutations in a gene called SCN9A. The gene codes for a protein called Nav1.7, a voltage-gated sodium ion channel found in the cell surface membranes of sensory neurones. Affected children had inherited two copies of a mutant allele, one from each parent, so they made no working Nav1.7. Their parents, each with one normal and one mutant allele, felt pain normally. Apart from a poor sense of smell, the rest of their nervous system worked normally.

3Pain begins with nociceptors, sensory neurones whose free nerve endings in the skin and other tissues respond to damaging stimuli such as intense heat, crushing pressure or chemicals released from injured cells. A stimulus opens ion channels in the nerve ending, producing a small depolarisation called a generator potential. If the generator potential is large enough to reach the threshold, an action potential is produced and travels along the neurone towards the spinal cord.

How a nerve impulse starts

4At rest, the inside of a neurone is about 70 mV more negative than the outside. This resting potential is maintained by sodium–potassium pumps in the membrane, which use ATP to move three sodium ions out of the cell for every two potassium ions moved in. The membrane is also more permeable to potassium ions than to sodium ions, because some potassium channels stay open, so potassium ions diffuse out of the cell.

5When the membrane is depolarised to the threshold, voltage-gated sodium channels open and sodium ions diffuse rapidly into the axon. The inside of the membrane becomes positive, reaching about +40 mV. The sodium channels then close and voltage-gated potassium channels open, so potassium ions diffuse out and the membrane is repolarised. Too many potassium ions leave at first, so the membrane is briefly hyperpolarised. During the refractory period that follows, another action potential cannot be produced straight away.

6Nav1.7 is found mainly in the nerve endings of nociceptors and in the neurones that carry signals about smell. It opens in response to small depolarisations, earlier than most sodium channels, and so acts as an amplifier. It boosts a weak generator potential until it reaches the threshold for an action potential. Without Nav1.7, the generator potential produced by a painful stimulus usually dies away before the threshold is reached. The stimulus is still detected by the nerve ending, but the information never leaves it.

7Pain arrives in two waves. If you stub your toe, you feel a sharp first pain almost immediately, followed a second or so later by a dull, throbbing ache. The first pain is carried by thin myelinated neurones that conduct impulses at about 5 to 30 metres per second. The ache is carried by even thinner unmyelinated neurones that conduct at only about 1 metre per second. The myelin sheath is made by Schwann cells wrapped around the axon, with small gaps between them called nodes of Ranvier.

8The nociceptors from your toe end in the spinal cord, where they form synapses with relay neurones. Some relay neurones pass impulses straight to motor neurones, producing a reflex that pulls the foot away before you are even aware of the pain. Others carry impulses up the spinal cord to the brain, where regions of the cerebral cortex identify where the damage is and how serious it is. Neurotransmitters released at these synapses can be increased or reduced by other neurones, which is one reason the same injury can feel very different from one day to the next.

Opposite mutations, opposite effects

9Other families show the opposite problem. People with the inherited condition erythromelalgia have attacks of intense burning pain in the hands and feet, triggered by mild warmth or exercise. Their hands and feet turn red during an attack, and some sufferers spend hours with their feet in cold water. This condition is caused by gain-of-function mutations in the same SCN9A gene, and only one copy of the mutant allele is needed to cause it.

10Many of these mutations are substitutions of a single base in the DNA, which change a single amino acid in the Nav1.7 protein. In some cases the altered channel opens in response to a smaller depolarisation than normal. In others it stays open for longer once it has opened. Either way, nociceptors fire action potentials in response to stimuli that would normally be harmless.

11"Nav1.7 behaves like the volume knob on a radio," says Dr Hana Mori, a neuroscientist who works on pain genetics. "Turn it right down and pain disappears. Turn it up and ordinary warmth becomes agony. The fact that turning it down leaves almost everything else working is what makes it such an attractive target for drug companies." She points out that the body’s other sodium channels are different proteins, coded for by different genes, with slightly different shapes.

Part 2: Blocking the channel

12Nature has been blocking sodium channels for millions of years. Tetrodotoxin, found in pufferfish and some newts and octopuses, binds to the outer opening of voltage-gated sodium channels and blocks the flow of sodium ions. Neurones can no longer produce action potentials, and a person who eats a poorly prepared pufferfish may become paralysed and stop breathing. In Japan, chefs who prepare pufferfish must train for years and hold a licence. Less than two milligrams of the toxin can be enough to kill an adult.

13Pufferfish themselves are not harmed by tetrodotoxin, because their own sodium channel genes contain mutations that alter the binding site so the toxin no longer fits. Some snakes that eat poisonous newts have independently evolved similar mutations. In both cases, the animals with resistant channels were more likely to survive and pass on their alleles. Pufferfish do not make the toxin themselves: it is produced by bacteria that live in their tissues, and pufferfish raised in tanks on toxin-free food are not poisonous.

14Local anaesthetics such as lidocaine also block voltage-gated sodium channels, but they enter the neurone and block the channel from the inside. A dentist injects lidocaine close to a nerve, so the sensory neurones there cannot conduct action potentials to the brain. Because lidocaine blocks sodium channels in all nearby neurones, including motor neurones, the lips and tongue also feel numb and are difficult to move until the drug has been removed.

15A drug that blocked all sodium channels throughout the body would be lethal. Humans have nine types of voltage-gated sodium channel, named Nav1.1 to Nav1.9. Nav1.5 is the main sodium channel in cardiac muscle, and Nav1.1 and Nav1.2 are common in the brain. To be safe as a painkiller, a drug taken by mouth must block Nav1.7 strongly while leaving the other eight types almost untouched, which is difficult because their structures are very similar.

16In an illustrative trial, 60 adults with erythromelalgia were given either a selective Nav1.7 blocker called SL-417 or a placebo tablet for three weeks. Neither the patients nor the doctors knew who received which tablet. Patients scored their pain on a scale from 0 to 10. In the SL-417 group, the mean pain score fell from 7.2 to 4.5. In the placebo group, it fell from 7.0 to 6.3.

17Results in people with common types of pain have been far less encouraging. Several Nav1.7 blockers that worked well on isolated neurones did little for patients with back pain or arthritis. One explanation came from mice and people lacking Nav1.7: they produce unusually large amounts of enkephalins, the body’s own opioid painkillers. Losing the channel seems to reduce pain partly by increasing opioid signalling, which a drug that blocks the channel for a few hours may not reproduce. Another possibility is that many of the drugs tested did not reach a high enough concentration around the nerve endings to block most of the channels there.

18Opioid drugs such as morphine bind to opioid receptors on neurones in the spinal cord and brain. Some of these receptors are on presynaptic membranes, where they reduce the release of neurotransmitter, so fewer impulses about pain pass across the synapse. Opioids are very effective, but with repeated use the body becomes tolerant and larger doses are needed. They can also cause addiction and dangerously slow breathing. Opioid receptors are found in the brainstem, in the centres that control breathing, which is why an overdose can be fatal.

19A different approach avoids drugs that circulate around the whole body. In illustrative experiments, researchers used a modified CRISPR system that cannot cut DNA but instead attaches chemical groups that reduce transcription of the SCN9A gene. Injected into the spinal cord of mice, it reduced their sensitivity to painful heat for more than six weeks, without changing their response to touch. Because the DNA sequence itself is not altered, the effect gradually wears off, which could be an advantage if a patient’s condition improves.

20"Pain is a warning system, and we must not switch it off completely," says Dr Mori. "The children who taught us about Nav1.7 show what happens when the warning never sounds. What we want is a way to turn the volume down for people whose pain no longer protects them, and then turn it back up again." The aim, she says, is not to remove pain altogether but to bring it back to a level that is useful.