How does acupuncture work?

Puncturing the Myth

Purinergic signaling, not mystical energy, may explain how acupuncture works.

By Geoffrey Burnstock | August 31, 2011

National Library of Medicine
National Library of Medicine
According to traditional Chinese medical theory, acupuncture points are situated on meridians along which qi, the vital energy, flows. However, I have proposed a less mysterious neurophysiological mechanism to explain the beneficial effects of this 2,000-year-old practice (Medical Hypotheses, 73:470-72, 2009). In particular, my hypothesis is based on the surprising finding that a hitherto unknown extracellular signalling system exists between cells, including nerve cells.

Adenosine 5’-triphosphate (ATP) is well established as an intracellular energy source that powers biochemical processes. In 1972 I proposed that ATP has another biochemical role: it acts as an extracellular signalling molecule between cells. The messages carried by ATP are received on the surface of cells by specific receptors, which I termed purinoceptors, because ATP belongs to a group of chemicals known as purines. Six years later, two families of purinoceptors were identified—P1 receptors for adenosine, the breakdown product of ATP, and P2 receptors for ATP. The purinergic signaling concept was rejected by many for two decades. It wasn’t until the early 1990s, when the chemical and molecular structure of the plasma membrane receptors for ATP was characterized and other downstream members of this primitive signalling pathway were identified, that the concept of purinergic signalling between cells became widely accepted, and it is now a rapidly expanding field of physiological and pathophysiological study.

Two intriguing hints prompted me to consider that inserting and twisting a needle might release ATP from the skin and form the physiological basis for the effects of acupuncture: 1) Initially it was thought that the ATP acting as an extracellular signalling molecule was merely a by-product released when cells were damaged or dying. 2) A paper published 34 years ago reported that ATP injected into the human skin stimulated sensory neurons (Pain, 3:367-77, 1977).

It is now clear that ATP can be released from many cell types (e.g., osteoblasts and endothelial, epithelial, and glial cells) in response to gentle mechanical stimulation that does not damage the cells. ATP is also released in response to heat and electrical currents—techniques used today in conjunction with acupuncture to enhance its effect. Recent evidence has also confirmed the 1977 finding that sensory nerve terminals in the skin are activated by ATP. In this way, messages can be relayed from the skin via interneurons in the spinal cord to the brain stem. Furthermore, the well-established reduction of pain by acupuncture may be explained by the possibility that the binding of ATP to purinoceptors on sensory nerve endings in the skin activates a signaling pathway which ultimately modulates pain perception in the brain’s cortex. Acupuncture’s inhibition of pain may also involve the release of endorphins.

Insertion and twisting of the needles employed in acupuncture mechanically deforms the skin, leading to the release of ATP by skin keratinocytes (1). ATP binds to specific receptors located on sensory nerve endings in the skin known as P2X3 and P2X2/3 (2). The signaling message is then relayed via dorsal root ganglia to the spinal cord (3) and subsequently through interneuronal pathways (4) to the brain stem (5) which contains motor neurons that control the functions of gut, lung, heart, arteries and reproductive organs, all major targets for acupuncture. Signals also travel to pain centers in the cortex, delivering a message to inhibit pain (6). Lucy Reading-Ikkanda

The ATP-activated sensory nerves also lead to modulation of the activity of brain-stem neurons controlling autonomic nervous system functions of gut, lung, urogenital, and cardiovascular systems—all of which have been treatment targets for traditional acupuncture procedures. There is published evidence for the release of ATP from keratinocytes, the major cell type in the skin, during mechanical stimulation. Similarly, ATP is released from urothelial cells lining the bladder and ureter in response to stretch, and receptors for ATP are present on suburothelial sensory nerves, ready to relay messages to the pain centers in the central nervous system. In addition, release of ATP in response to mechanical stimulation (changes in blood flow) from endothelial cells that line blood vessels leads to vasodilatation. And further, ATP is released from epithelial cells lining the airways in response to stretch, leading to activation of ATP receptors on sensory nerves, in turn resulting in the activation of reflexes that protect the lung against hyperventilation.

Immunohistochemical studies have shown that the specific ATP receptor subtypes, P2X3 and P2X2/3, are located on sensory nerve endings in the skin. The same subtypes are also especially abundant in the tongue, another site where acupuncture needles are placed. An isolated preparation of tongue tissue showed that the increased electrical activity in lingual general sensory nerves in response to mechanical stimulation could be mimicked by injecting ATP into the preparation and blocked by injecting antagonists to the P2X3 receptor subtype. The cell bodies of the sensory nerve endings that supply the skin are located in sensory ganglia, which then connect with neurons in the dorsal spinal cord. A series of interneurons then mediate modulatory pathways to the brain stem and hypothalamus, which are the nervous control centers for the activities of visceral organs. (See illustration.)

Suggested experiments


Many tools are available to test various aspects of this hypothesis experimentally. Apyrase, a readily available enzyme that breaks down ATP, could be applied to the skin to see whether the enzyme diminishes the benefits of acupuncture. In contrast, inhibitors of ATP breakdown, such as ARL-67156, could be employed to see whether this would enhance the beneficial effects of acupuncture. There are also very sensitive assay methods for measuring ATP release, which could be used in skin subjected to mechanical deformation, heat, and electrical current. Selected blockers (antagonists) of P2X3 and P2X2/3 receptors are available, which should block the beneficial effects of acupuncture. It seems likely from experiments on the bladder and intestine that ATP-sensitive low-threshold sensory fibers mediate physiological events, while high-threshold fibers mediate pain. This will need to be clarified for the sensory nerves supplying the skin and tongue before approaches to enhancing the ATP-related responses to acupuncture are carried out, in case the enhancement results in pain.
I hope that this hypothesis will provoke some scientists interested in acupuncture to investigate further.

Geoffrey Burnstock is Emeritus Professor and President of the Autonomic Neuroscience Centre of University College Medical School in London. He is editor-in-chief of Autonomic Neuroscience, Purinergic Signalling, and the journals Open Neuroscience and Open Pharmacology.