Produced by the Science/AAAS Custom Publishing Ofce

Purinergic signaling in acupuncture

T
Author:
Geoffrey Burnstock
FIGURE 1. Acupuncture
he proposed role of puriner-
and purinergic
gic signaling in the physiological
signaling. Insertion
basis of acupuncture was Brst presented in 2009.
and twisting of the
Data showing that ATP is released from kerati-
needles employed
nocytes and other skin cells during acupuncture
in acupuncture
treatments lends weigh to this hypothesis. ATP
in turn activates P2X3 receptors on the sensory mechanically deforms
nerves in the skin, which then transmit those the skin, leading to
messages to motor neurons in the brain stem the release of ATP by
that control autonomic functions and modulate skin keratinocytes (1).
nociceptive activities. Here, we review and de- ATP binds to specifc
scribe the recent evidence for purinergic signaling receptors located on
underlying acupuncture effects and propose ways sensory nerve endings
to further test this hypothesis. in the skin, P2X3
and P2X2/3 (2). The
Introduction signaling message is
It has been well established that adenosine then relayed via dorsal
5’-triphosphate (ATP) is an intracellular energy root ganglia to the
source in cellular biochemistry. In 1970, Burnstock spinal cord (3) and
et al. suggested that ATP acted as a nonadrener- subsequently through
gic, noncholinergic neurotransmitter in the gut (1), interneuronal pathways
and in 1972 he named the extracellular actions of (4) to the brain stem
ATP, “purinergic signaling” (since ATP is a purine
(5) that contains motor
nucleotide), and formulated the purinergic signal-
neurons, which control
ing hypothesis (2).
the functions of gut,
In 2009, Burnstock proposed that purinergic
lung, heart, arteries,
signaling could be involved in the physiological
and reproductive
mechanisms mediating acupuncture effects.
This hypothesis suggested that mechanical organs—all major
deformation of the skin by needles or application targets for acupuncture.
of heat or electrical current leads to the release Signals also travel to
of large amounts of ATP from keratinocytes, the pain centers of
Bbroblasts, and other cell types in skin (Figure 1). the cortex, delivering
The released ATP then activates P2X3 ion channel a message to inhibit
receptors on sensory nerves within the skin pain (6). (Reproduced
and tongue that transmit messages via sensory with permission from
ganglia and the spinal cord to the brain stem reference 36.)
and hypothalamus. These brain regions contain
motor neurons that control autonomic functions,
including cardiovascular, gastrointestinal,
respiratory, and urinogenital activities—common
targets of acupuncture treatments. These sensory
neuron messages also modulate the pathways
that lead to centers in the cortex responsible
for conscious awareness of pain and other
central nervous system activities, including sleep
regulation (3). A number of subsequent studies have been Supporting evidence for the hypothesis
published that also implicate purinergic signaling in various Studies that have established the components involved
aspects of acupuncture, detailed below. in the purinergic signaling pathway include: (i) release of
ATP (in response to mechanical or chemical stimulation)
Materials that appear in this section were not reviewed or
assessed by Science Editorial staff, but have been evaluated by
Autonomic Neuroscience Centre, University College Medical School, London, UK,
an international editorial team consisting of experts in traditional and Department of Pharmacology and Therapeutics, The University of Melbourne,
medicine research. Melbourne, Australia
g.burnstock@ucl.ac.uk

S23
Produced by the Science/AAAS Custom Publishing Ofce

FIGURE 2. Schematic illus-

from keratinocytes (4–6) and possibly from Merkel cells, the central nervous system have been reported (15).
which contain high levels of ATP (7, 8); ATP has also been Subsequent papers have built upon and extended
shown to be released from keratinocytes upon heating (9); evidence in support of purinergic signaling underlying
(ii) immunohistochemical data demonstrating the presence acupuncture effects. Several studies have associated the skin
of P2X3 receptors on sensory nerve fbers in the skin (10–12) cells affected by acupuncture techniques with purinergic
and tongue (13); (iii) in an isolated tongue/lingual nerve signaling. For example, ATP has been shown to be released
preparation, mechanical activation of the tongue with De from human keratinocytes in response to mechanical
Frey hairs was shown to result in a discharge in the lingual stimulation by hypo-osmotic shock (16), as well as from
sensory nerve fbers that was mimicked by ATP activation keratinocytes in response to heat (17). Additionally, mast
and blocked by P2X3 receptor antagonists (14); and (iv) both cells, which accumulate around the acupuncture needles,
presynaptic inhibition via adenosine A1 and P2Y receptors, also release ATP in response to mechanical stimulation (18).
and enhancement via P2X and A 2A receptors at synapses in Another skin cell type, human subcutaneous fbroblasts, can

S24
 Produced by the Science/AAAS Custom Publishing Ofce

also release ATP in response to bradykinin and histamine (19, increase or decrease in heart rate or blood pressure, and
20). Tsutsumi et al. demonstrated that mechanical stimulation identify specifc neurons that are activated in the brain using
can evoke the propagation of calcium waves between noninvasive scanning techniques. If acupuncture-induced
human keratinocytes, induced by ATP and activation of effects can be identifed and quantifed, researchers could
P2Y2 receptors (21, 22), which is consistent with the earlier then test whether ATP mimicks the responses and if P2X3
results from Koizumi et al. (5). Tuina (traditional therapeutic receptor antagonists block the effects. Moreover, we
massage) and moxibustion (a traditional Chinese medicine suggest that researchers conduct experiments recording
therapy using a moxa, often made from dried mugwort, responses from sensory neurons in the skin and tongue
either used as a Muff or processed into a cigar-shaped stick; it in animal models and distinguish between low-threshold
can be used indirectly, with acupuncture needles, or burned fbers involved in acupuncture and high-threshold fbers
on to the patient’s skin) may also act via the purinergic that mediate nociception, as well as recordings from the
signaling pathway (23). Papers describing the release of motor nerves in the brainstem responsible for autonomic
ATP from human epidermal keratinocytes via connexin functions.
hemichannels and vesicles involving vesicular nucleotide
transporter have recently been published (24–26). A 2010 References
study has claimed that adenosine, following breakdown of 1. G. Burnstock, G. Campbell, D. Satchell, A. Smythe, Br. J. Pharma
released ATP during acupuncture, can act as a prejunctional col. 40, 668 (1970).
inhibitor of neurotransmission via A1 receptors, resulting in 2. G. Burnstock, Pharmacol. Rev. 24, 509 (1972).
anti-nociceptive actions (27). Valuable reviews are available 3. G. Burnstock, Med. Hypotheses 73, 470 (2009).
describing the neural pathways from different skin regions to 4. N. Mizumoto, M. E. Mummert, D. Shalhevet, A. Takashima, J.
Invest. Dermatol. 121, 1066 (2003).
structures in the brain stem and higher brain centers. These
5. S. Koizumi et al., Biochem. J. 380, 329 (2004).
pathways are important because different acupuncture sites 6. H. E. Burrell et al., J. Biol. Chem. 280, 29667 (2005).
may activate different neural pathways impinging on specifc 7. R. Crowe, M. Whitear, Cell Tissue Res. 190, 273 (1978).
nuclei in the brain stem that control autonomic functions 8. M. Silberstein, Med. Hypotheses 75, 272 (2010).
potentially modulated by acupuncture (Figure 2) (28, 29). 9. S. Mandadi et al., Pfugers Arch. 458, 1093 (2009).
10. E. J. Bradbury, G. Burnstock, S. B. McMahon, Mol. Cell. Neurosci.
Purinergic signaling and electroacupuncture 12, 256 (1998).
Electroacupuncture is a form of acupuncture where a 11. G. Burnstock, Br. J. Anaesth. 84, 476 (2000).
small electric current is passed between pairs of acupuncture 12. M. Taylor, J. C. Peleshok, A. Ribeiro-da-Silva, J. Comp. Neurol.
514, 555 (2009).
needles. This is thought to augment traditional acupuncture
13. X. Bo et al., Neuroreport 10, 1107 (1999).
and is believed to be particularly helpful in treating pain.
14. W. Rong, G. Burnstock, K. M. Spyer, J. Physiol. 524, 891 (2000).
The supraspinal antinociception effect of 15. G. Burnstock, Physiol. Rev. 87, 659 (2007).
electroacupuncture has been associated with P2X3 receptor 16. N. Azorin et al., Exp. Dermatol. 20, 401 (2011).
activation in the midbrain periaqueductal gray region (30). 17. J. R. Gifford, C. Heal, J. Bridges, S. Goldthorpe, G. W. Mack, J.
Moreover, the analgesic effect of electroacupuncture on Physiol. 590, 6403 (2012).
chronic neuropathic pain has been shown to be mediated 18. L. Wang et al., Evid. Based Complement. Alternat. Med. 2013,
by P2X3 receptors in rat dorsal root ganglion neurons (31). 350949 (2013).
Following these studies, electroacupuncture was shown to 19. R. Pinheiro et al., Cell Commun. Signal. 11, 70 (2013).
result in a reduced expression of P2X3 and P2X2 receptors in 20. R. Pinheiro et al., J. Biol. Chem. 288, 27571 (2013).
21. M. Tsutsumi et al., Cell Tissue Res. 338, 99 (2009).
the dorsal root ganglion of rats with chronic neuropathic pain
22. M. Tsutsumi et al., Skin Res. Technol. 16, 146 (2010).
(32) and visceral hypersensitivity (33). Electroacupuncture 23. L. Fan, L. M. Yin, J. Acupunct. Tuina Sci. 12, 125 (2014).
at He-Mu points can also reduce P2X4 receptor expression 24. T. P. Barr et al., PLOS ONE 8, e56744 (2013).
in colon and spinal cord in visceral hypersensitivity (34). 25. T. Takahashi et al., J. Invest. Dermatol. 133, 2407 (2013).
Moreover, in a review by Lin et al., the neuroprotective 26. K. Inoue et al., J. Invest. Dermatol. 134, 1465 (2014).
effects of acupuncture were reported to act via increasing 27. N. Goldman et al., Nat. Neurosci. 13, 883 (2010).
brain derived neurotrophic factor (BDNF) expression via 28. Z. J. Zhang, X. M. Wang, G. M. McAlonan, Evid. Based
stimulation of ATP (35). Complement. Alternat. Med. 2012, 429412 (2012).
29. Q. Q. Li et al., Evid. Based Complement. Alternat. Med. 2013,
267959 (2013).
Conclusions
30. Z. Xiao et al., Brain Res. 1330, 31 (2010).
Evidence in support of the hypothesis of purinergic
31. W. Z. Tu et al., Neurochem. Int. 60, 379 (2012).
signaling mediating the physiological mechanisms 32. R. D. Cheng et al., Chin. J. Integr. Med. 19, 374 (2013).
underlying acupuncture effects has been accumulating over 33. Z. Wang et al., Neural Regen. Res. 8, 802 (2013).
recent years. To help further test this hypothesis, I propose 34. X. Guo et al., Neural Regen. Res. 8, 2069 (2013).
that experienced acupuncturists focus on acupuncture 35. D. Lin et al., Int. J. Mol. Sci. 15, 3234 (2014).
sites that induce effects that can be quantifed, such as an 36. G. Burnstock, The Scientist 25, 24 (2011).

S25