654453

research-article2016
NROXXX10.1177/1073858416654453The NeuroscientistTang et al.

Original Article
The Neuroscientist

Acupuncture-Induced Analgesia: A
1­–16
© The Author(s) 2016
Reprints and permissions:
Neurobiological Basis in Purinergic sagepub.com/journalsPermissions.nav
DOI: 10.1177/1073858416654453

Signaling nro.sagepub.com

Yong Tang1, Hai-Yan Yin1, Patrizia Rubini2, and Peter Illes2

Abstract
Chronic pain is a debilitating and rather common health problem. The present shortage in analgesic drugs with a
favorable spectrum but without remarkable side effects furthered the search for alternative therapeutic manipulations.
Increasing evidence from both basic and clinical research on acupuncture, a main alternative therapy of traditional
Chinese medicine, suggests that chronic pain is sensitive to acupuncture procedures. Clarification of the underlying
mechanisms is a challenge of great theoretical and practical significance. The seminal hypothesis of Geoffrey Burnstock
and the astounding findings of Maiken Nedergaard on the involvement of purinergic signaling in the beneficial effects
of acupuncture fertilized the field and led to an intensification of research on acupurines. In this review, we will
summarize the state-of-the-art situation and try to forecast how the field is likely to develop in the future.

Keywords
acupuncture analgesia, purinergic signaling, ATP, adenosine, P2X receptors, P2Y receptors, adenosine receptors

Introduction at acupoints using a variety of techniques (NIH Consensus

Acupuncture (AP; from Latin acu “with a needle”; Latin
punctura, from pungere “to prick”), the word invented
by Wilhelm ten Rhyne in his Latin treatise De
Acupuncture is originally characterized by the insertion
of fine, solid metallic needles into or through the skin at
specific sites, so called acupuncture points (also named
acupoints) (Berman and others 2010; Johnson 2006). It
has probably been used in traditional Chinese medicine
(TCM) as a therapeutic manipulation for the past 3000
to 4000 years, although the first document called The
Yellow Emperor’s Classic of Internal Medicine (known
as the Huangdi Neijing) that unequivocally described
AP as an organized system of diagnosis and treatment is
dating from about 100 BCE (Cheng 2010; White and
Ernst 2004). During the Ming Dynasty (1368-1644),
The Great Compendium of Acupuncture and Moxibustion
(Zhenjiu Dacheng) was published, which is a synthesis
of classic AP and forms the basis for its use in modern
times.
To date, AP has gained popularity since the landmark
NIH Consensus Conference in 1997; it has been employed
in more than 180 countries and regions around the world
based on a report issued by the World Federation of
Acupuncture-Moxibustion Societies (WFAS) in 2013.
Currently, the term acupuncture should refer to a family
of procedures involving physical or chemical stimulation
Conference 1998; see also below).
In traditional AP, needles may be manipulated manu-
ally (Fig. 1A,B) or stimulated electrically, the latter is
called electroacupuncture (EAP). Manipulation of the
needles causes deqi, a sensation of soreness, numbness,
distension, aching, or heaviness, which is thought to be
essential for the therapeutic outcome (Loyeung and
Cobbin 2013). Moxibustion means the burning of a cone-
shaped preparation of moxa (made of dried mugwort) on
or near the skin, often but not always near to an acupoint
(Fig. 1A,B; Fu and others 2016). Cupping therapy is an
ancient Chinese form of alternative medicine in which a
local suction is created on the skin (Fig. 1A,B; Cao and
others 2012). In this publication, the term acupuncture
will be used in its broad sense to include traditional body
needling, moxibustion, electroacupuncture, laser acu-
puncture, and so on (Fig. 1A; World Health Organization
2002). Onetime sessions of AP last for about 30 minutes,
1
Acupuncture and Tuina School, Chengdu University of Traditional
Chinese Medicine, Chengdu, China
2
Rudolf-Boehm-Institut für Pharmakologie und Toxikologie,
Universität Leipzig, Leipzig, Germany
Corresponding Author:
Peter Illes, Rudolf Boehm Institute for Pharmacology und Toxicology,
University of Leipzig, Haertelstrasse 16-18, Leipzig 04107, Germany.
Email: peter.illes@medizin.uni-leipzig.de

Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
2 The Neuroscientist 

Figure 1.  Pain therapy with different acupuncture family procedures. (A) The Chengdu panda bear offers a variety of
acupuncture (AP) procedures for selection. In Chengdu (China), the first 2 authors carry out intensive research on AP at the
University of Traditional Chinese Medicine, and also in this city the famous Chengdu Research Base breeds panda bears. Guasha,
cupping technique; EAP, electroacupuncture; TEAS, transdermal electrical acupoint stimulation. (B) Needle insertion, moxibustion
and cupping techniques. Needles are inserted at specific acupoints and either manually manipulated or electrically stimulated
(EAP). Moxibustion means the burning of cone-shaped preparations of moxa (dried mugwort) on or near the skin, often but not
always near to an acupoint. Cupping therapy means local suction of the skin. Figure B has been modified from Cheng (2010).

are usually repeated twice a week at the beginning, and
continued once a week later, for a minimum of 12 ses-
sions. AP treatment is favored by the relatively low inci-
dence of major adverse effects, which can be ascribed
less to the procedure itself than to the moderate skills of
the acupuncturist (Berman and others 2010; Ernst and
others 2011).
In 1979, the World Health Organization conducted a
symposium in Beijing, China, to identify the conditions
that might benefit from acupuncture treatment. The inter-
national participants drew up a list of 43 suitable diseases.
In 1996, the number was updated to 64 in another meeting
in Milan, Italy. In 2002, the World Health Organization
published a book titled Acupuncture: Review and Analysis
of Reports on Controlled Clinical Trials, in which 29 con-
ditions were identified for which acupuncture has been
proved, through controlled trials, to be an effective treat-
ment. Among the 29 conditions, 12 were associated with
pain. In the PubMed database, we found that in a total of
24,430 acupuncture publications, 2,236 (9.15%) focus on
acupuncture analgesia (AA) and 6,553 (26.82%) on acu-
puncture and pain. Furthermore, increasing evidence from
both basic and clinical research on AP has indicated that
pain is particularly sensitive to acupuncture (Berman and
others 2010; Lu and others 2016; Staud and Price 2006;
Vickers and others 2012; Vickers and Linde 2014; Zhang
and others 2014). In this article, we will summarize the
available data on acupurines, that is, the role of purines
and purinergic signaling in acupuncture, from research
published in the recent years on the mechanism of AA.
Clinical Effectiveness and
Neurobiological Basis of
Acupuncture Analgesia
The clinical effectiveness of AP in various pain condi-
tions is a much debated issue with powerful advocates
and similarly powerful opponents. The opponents’ argu-
ments are basically the following (Colquhoun and
Novella 2013; Madsen and others 2009): (1) Large multi-
center clinical trials conducted in Germany and the
United States consistently revealed that true (verum) and
sham AP (stimulation at nonspecific acupoints) do not
differ in their effectiveness in decreasing pain levels
across multiple chronic pain disorders; (2) the existing
differences between the AP (verum or sham) and non-AP
groups in improvement were only minor and therefore

Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
Tang et al. 3

Figure 2.  The neuronal purinergic system. 1. ATP is released from healthy cells into the extracellular space by vesicular
release (as a transmitter or co-transmitter substance) or via ion channels/hemichannels/transporters. 2. Spontaneous efflux of
purines occurs from injured or dying cells resulting in pathological concentrations of ATP in the extracellular space. 3. Ecto-
enzymes rapidly hydrolyze or interconvert the extracellular nucleotides thereby either terminating their action or producing
an active metabolite of altered receptor selectivity. 4. Postsynaptic ionotropic P2X and metabotropic P2Y receptors mediate
fast and slow synaptic responses, respectively. 5. Extracellular purines activate pre- and postsynaptic P2X and P2Y receptors;
adenosine stimulates its own P1 receptor type. These receptors (further divided into subtypes) modulate the effects of classic
neurotransmitters (e.g., glutamate acting at its ionotropic receptors); thereby, purines represent a complex neuromodulatory
system involved in fine-tuning of neurotransmission. In addition, purines of course also function as signaling molecules between
cells of non-neuronal origin. Modified from Köles and others (2016).

could be due to a placebo effect; (3) it is hard to under-
stand why AP relieves some types of pain while leaving
other types of probably similar etiology unaffected; (4)
interestingly, the strongest evidence for a positive out-
come of AP treatment is in case of postoperative nausea
and vomiting, a condition not related to pain.
However, the arguments of the advocates of AP appear
to be also meaningful. (1) Numerous authors published
meta-analyses of randomized, sham-controlled clinical
investigations or evaluated Cochrane reviews, which are
internationally recognized as the highest standard of evi-
dence. They concluded that AP is effective in the short-term
management of low back pain, neck pain, and osteoarthritis
involving the knee, but ineffective in the management of
dental pain, coloscopy pain, and intraoperative analgesia
(Johnson 2006; Lee and Ernst 2011; Wang and others
2008). It was stated that the majority of early reviews
arrived at negative conclusions (Johnson 2006), while the
majority of recent reviews were positive, probably indicat-
ing a biased or incomplete evaluation of the early studies
(Ernst and others 2011). (2) Neuroimaging techniques such
as positron emission tomography (PET), single photon
emission computed tomography (SPECT), and functional
magnetic resonance imaging (fMRI), all suggested that AP
results in significant activation of multiple brain areas of
healthy human volunteers, including the hypothalamus, pri-
mary somatosensory cortex and rostral anterior cingulate
cortex, important for pain sensation in higher brain centers
(Lewith and others 2005; Staud and Price 2006; Wang and
others 2008). (3) There is convincing evidence that AP by
stimulating sensory nerve terminals in cutaneous/subcuta-
neous and muscle tissue causes the release of endogenous
opioid peptides in the brain and spinal cord. Classic experi-
ments in mice demonstrated that the administration of the
opioid antagonist naloxone blocked the AP-induced analge-
sia (Pomeranz and Chiu 1976). Similarly, in a human
model, AP raised the levels of opioid peptides in the cere-
brospinal fluid of human volunteers (Sjölund and others
1977). Apparently, analgesia by AP originates at all levels of
the central nervous system (CNS). Opioids released in inflamed
tissue from lymphocytes, monocytes/macrophages, and
granulocytes by EAP activated specific receptors (Rs) at

Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
4 The Neuroscientist 
sensory nerve terminals to suppress nociception (Cabot and
others 1997; Zhang and others 2014). In studies, with unin-
jured rats, low frequency EAP released β-endorphin and
enkephalins, while high-frequency EAP released dynor-
phins to suppress nociception as assessed by the tail flick
test (Han 2003). Eventually, it has been shown that EAP
inhibits the sensory dimension of pain by targeting many
sites of the brain such as periaqueductal gray, locus coeru-
leus, habenula, nucleus accumbens, and caudate nucleus
(Zhao 2008). It has been hypothesized that the periaqueduc-
tal gray is neuronally connected with the raphe nuclei which
send descending serotonergic fibres to the spinal cord dor-
sal horn activating enkephalinergic, analgesia-mediating
interneurons (Takeshige and others 1992). (4) AP activates
a number of non-opioid pain-relevant neuronal systems in
the CNS which operate via serotonin, noradrenaline, chole-
cystokinin, glutamate, neuropeptide Y, and so on (Irnich
and Beyer 2002; Zhao 2008).
The Purinergic System
Since the discovery of purinergic neurotransmission in
the mammalian organism (Burnstock 1972), and the
observation that adenine and uridine nucleotides as well
as their enzymatic degradation products (e.g., adenosine)
have a widespread signaling function coordinating the
activities of almost all neuronal and non-neuronal cells
(see Ralevic and Burnstock 1998), research activities of
many scientists have concentrated on purinergic/pyrimi-
dinergic mechanisms. It has been noticed relatively early
that purines play an important role in the sensation of
noxious stimuli in the periphery and the transmission/
modulation of such impulses by neuronal pathways in the
CNS (Burnstock 2006, 2009a, 2016).
In this context, ATP was identified as an autacoid
released into the extracellular space from the interior of
damaged cells via holes in their plasma membrane caused
by noxious stimuli, or from healthy cells in consequence
of shear stress, stretch, osmotic swelling or metabolic
limitation (Fig. 2; Lazarowski 2012). Subsequently, ATP
acts at its receptors or is degraded by the ectonucleotidase
families E-NPTDases (ectonucleoside triphosphate
diphosphohydrolases), E-NPPs (ectonuclotide pyrophos-
phatase and/or phosphodiesterases), alkaline phospha-
tases and ecto-5′-nucleotidase (Yegutkin 2008). The
individual enzymes differ in their substrate specificity
and product formation. E-NPTDases and E-NPPs hydro-
lyze ATP and ADP to AMP, which is further hydrolyzed
to adenosine by ecto-5′-nucleotidase (Abbracchio and
others 2009). Alkaline phosphatases equally hydrolyze
nucleoside tri-, di-, and monophosphates. Adenosine is
generated either extracellularly from AMP by ecto-5′-
nucleotidase or intracellularly by endo-5′-nucleotidase;
The latter can leave the cell by means of an equilibrative
adenosine transporter (Sawynok and Liu 2003).
Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
Eventually adenosine is inactivated by adenosine-deami-
nase to inosine or by adenosine-kinase to AMP.
Nucleotides act at two receptor-types called P2X
(ligand-gated cationic channels; P2X1-7 subtypes) or P2Y
(G protein-coupled receptors; P2Y1,2,4,6,11-14 subtypes)
(Fig. 2; Abbracchio and Burnstock 1994; Köles and others
2007). Adenosine acts at its own receptor-types, which are
all coupled to G proteins (A1, A2A, A2B, A3; Fredholm and
others 2011). P2XRs occur as trimeric constructs of the
same subunit (homomeric) or of different subunits (het-
eromeric) receptors (Köles and others 2008). Each P2X
subunit has two transmembrane domains, a large extracel-
lular loop, and intracellular N- and C-termini. P2Y and
adenosine receptors have typical seven transmembrane
domains, an extracellular N-terminus and an intracellular
C-terminus. P2YRs may also constitute homomeric or
heteromeric assemblies with other P2Y or adenosine
receptors. The heteromeric assemblies add further diver-
sity to purinergic receptors, by combining the original bio-
physical and pharmacological characteristics of their
parent subunits.
Pharmacology of Purinergic
Receptors and Blockers of Purine
Degrading Enzymes Involved in Pain
Sensation
A number of P2X- and P2YR-subtypes as well as all ade-
nosine receptors-types are involved in pain sensation
(Table 1). Homomeric P2X3Rs and heteromeric P2X2/3Rs
are located at the terminals of Aδ and C fibers projecting
from sensory neurons of dorsal root ganglia to the inner-
vated tissues (Chen and others 1995; Lewis and others
1995). P2X3Rs respond to the ATP structural analogue
α,β-methylene ATP (α,β-meATP) and desensitize rapidly
during agonist application. P2X2/3Rs are also sensitive to
ATP/α,β-meATP, but do not exhibit desensitizing proper-
ties. P2X3 and P2X2/3Rs participate in acute, inflamma-
tory, neuropathic and cancer pain as proven for various
experimental pain models (Burnstock 2006; Wirkner and
others 2007). The subcutaneous or intraplantar application
of ATP and α,β-meATP causes nociceptive responses and
hyperalgesia/allodynia as characteristics of acute and neu-
ropathic pain, respectively. Further, non-selective (sura-
min) and P2X3R-selective pharmacological antagonists
(A317491) as well as antisense oligonucleotides all
improve the mentioned pain states in animal models
(Donnelly-Roberts and others 2008; Kennedy and others
2003). Accordingly, P2X3R-deficient mice exhibit
decreased nocifensive behavior in comparison with their
wild-type backgrounds in experimental pain states.
P2X4Rs at spinal microglia mediate neuropathic pain
(Coull and others 2005; Inoue and others 2007). ATP-
induced activation of microglial P2X4Rs leads to the
Tang et al. 5

Table 1.  Purines Involved in Pain Sensation.

Modulation of
Receptor Receptor Inflammatory/
Purines Types Subtypes Neuropathic Pain References
Adenosine P1 A1 ↓ Lima and others (2010); Maione and others (2007);
Sawynok (2015); Zylka (2011)
A2A ↑ or ↓ Sawynok (2015); Zylka (2011)
A2B ↑ or ↓ Sawynok (2015)
A3 ↓ or ↑ Little and others (2015); Sawynok (2015)
ATP P2X P2X1 –  
P2X2 –  
P2X3, P2X2/3 ↑ Burnstock (2006); Donnelly-Roberts and others (2008);
Kennedy and others (2003); Wirkner and others (2007)
P2X4 ↑ Coull and others (2005); Inoue and others (2007)
P2X5 –  
P2X6 –  
P2X7 ↑ Carroll and others (2009); Honore and others (2006);
Inoue and others (2007); Villa and others 2010
ATP, ADP, UTP, P2Y P2Y1 ↑ or ↓ Barragán-Iglesias and others (2014, 2015, 2016); Gerevich
UDP and others (2007)
P2Y2 ↑ Moriyama and others (2003)
P2Y4 –  
P2Y6 ↑ Barragán-Iglesias and others (2014, 2015); Okada and
others (2002); Syhr and others (2014)
P2Y11 ↑ Barragán-Iglesias and others (2014, 2015); Okada and
others (2002); Syhr and others (2014)
P2Y12 ↑ Kobayashi and others (2008); Tozaki-Saitoh and others
(2008)
P2Y13 –  
P2Y14 –  

↑, potentiation of pain; ↓, inhibition of pain; –, no relevance to pain.

secretion of brain-derived neurotropic factor (BDNF),
which was implicated in the hypersensitivity of dorsal
horn neurons that follows sensitization and inflammation.
BDNF causes a depolarizing shift in the Eanion of spinal
lamina I neurons at the dorsal spinal horn, resulting in
conversion of GABAA and glycin receptor-mediated
inhibition to excitation. Indeed intrathecal administration
of recombinant BDNF-sequestering fusion protein
(TrkB-Fc) acutely inhibited allodynia and the shift of
Eanion of lamina I neurons (Coull and others 2005).
P2X7Rs are also involved in both inflammatory and
neuropathic pain (Carroll and others 2009; Inoue and oth-
ers 2007). P2X7R subunits possess a particularly long
intracellular C-terminus which has been implicated in
regulating receptor function, including signaling pathway
activation, cellular localization, protein-protein interac-
tions, and post-translational modification (Costa-Junior
and others 2011). In addition, P2X7Rs function as non-
selective cationic channels after activation by relatively
low concentrations of ATP, but form large diameter pores
on their long-lasting activation by high ATP concentra-
tions or alternatively open pores formed by an accessory
protein (e.g., Pannexin-1 hemichannels) (Sperlágh and
others 2006; Sperlágh and Illes 2014). P2X7Rs also acti-
vate the apoptotic caspase enzyme cascade and induce
rapid maturation and release of the proinflammatory
cytokine interleukin-1β (IL-1β). All these properties as
well as the localization of P2X7Rs at peripheral and cen-
tral immunocytes (lymphocytes, monocytes/macro-
phages, microglia, astrocytes) enable them to become key
players at the neuroimmune interface (Burnstock and oth-
ers 2011; Carroll and others 2009).
Systemic application of P2X7R antagonists such as
A438079 produced dose-dependent antinociceptive
effects in models of neuropathic (Nelson and others
2006) and inflammatory pain (Honore and others 2006).
In the case of neuropathic pain, P2X7Rs located at satel-
lite glial cells (astrocyte-like cells in sensory ganglia)
initiate an inflammatory reaction (Villa and others
2010). Accordingly, deletion of the p2rx7 gene not only
altered inflammatory pain but also reduced pain associ-
ated with nerve injury (Chessell and others 2005). Other
genetic manipulations of the IL-1 system, including tar-
geted gene disruption of the IL-1 type I receptor or the
IL-1 accessory protein (IL-1acp) have generated mice
that show reduced nociceptive responses relative to

Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
6 The Neuroscientist 
Figure 3.  Inhibition of voltage-sensitive Ca2+ channels and
P2X3R currents by ATP and ADP-β-S, respectively. (A)
Ca2+ currents were recorded in cultured rat DRGs using
the whole-cell configuration of the patch-clamp technique.
Individual tracings with a normal GTP (0.3 mM)-containing
pipette solution (a) and a pipette solution in which GTP
was replaced by GDP-β-S (0.3 mM) (b). Depolarizing steps
from −90 to −10 mV were evoked every 20 seconds for
80 ms. Current responses before (control) and during
the application of ATP (100 μM) for 2 minutes. One
representative experiment out of 6 (a) or 7 (b) similar
ones. The inhibition of G proteins by intracellular GDP-β-S
blocked the P2Y1R-mediated opening of voltage-sensitive
Ca2+ channels. Hence, P2Y1 receptors appear to block ICa2+
by a pathway involving the β,γ subunit of a Gq/11 protein.
(B) P2X3R currents were induced by 3 μM α,β-meATP
in HEK293 cells transfected with the human P2X3R. The
whole-cell configuration of the patch clamp technique was
used. The P2X3R agonist α,β-meATP was applied for 1
second every 5 minutes. The P2Y1,12,13R agonist ADP-
β-S (1 μM) was applied for 10 min, for the duration of
two agonist applications either through a normal GTP (0.3
mM)-containg pipette solution (a) or through a GDP-β-S
(0.3 mM)-containing pipette solution (b). One representative
recording out of 7 (a) or 9 (b) similar experiments. Since
intracellular GDP-β-S blocked the inhibitory effect of ADP-
β-S, it was concluded that the activation of P2Y receptors
G protein-dependently facilitates the desensitization of
P2X3 receptors and suppresses the recovery from the
desensitized state. Modified from Gerevich and others
(2004) (A) and Gerevich and others (2007) (B).
Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
wild-type animals (Donnelly-Roberts and others 2008;
Wolf and others 2003).
The contribution of the metabotropic P2YRs to nor-
mal and pathological pain have been less well examined
than that of P2XRs (Burnstock 2009a). P2Y1Rs have
been described to cause analgesia by blocking via G pro-
teins N-type Ca2+ channels responsible for the release of
nociceptive transmitters in the spinal cord (Fig. 3A;
Gerevich and others 2004; Gerevich and Illes 2004) or
via P2X3Rs (Fig. 3B; Gerevich and others 2007). The
analgesic effect of the endogenous P2Y1R agonist ADP
has been demonstrated in acute thermal and neuropathic
pain (Andó and others 2010; Barragán-Iglesias and others
2016). The transient receptor potential vanilloid 1
(TRPV1) is a polymodal detector of pain-producing
chemical and physical stimuli. It has been reported that
extracellular ATP potentiates the TRPV1 currents evoked
by capsaicin or protons and reduces the temperature
threshold for its activation through metabotropic P2Y2Rs
(Moriyama and others 2003). Co-expression of TRPV1
mRNA with P2Y2 mRNA was demonstrated in the rat
lumbar DRG using in situ hybridization histochemistry.
Increasing evidence indicates that UDP-sensitive
P2Y6Rs are involved in pain, even though it has also
been reported that the P2Y6R antagonist (MRS2578) has
no effect on neuropathic pain behavior in mice (Syhr and
others 2014). In the neuropathic pain model induced by
partial ligation of the sciatic nerve of rats, intrathecal
administration of UDP produced significant antiallodynic
effects (Okada and others 2002). In these rats, the expres-
sion of P2Y6,11Rs was increased in the ipsilateral lumbar
spinal cord 7 to 21 days after injury (Barragán-Iglesias
and others 2014). Intrathecal treatment with selective
P2Y6 and P2Y11R antagonists reduced both the injury-
induced allodynia and the accompanying receptor up-
regulation. Spinal application of minocycline damaged
microglia and in consequence, depressed both nerve
injury-induced up-regulation of P2Y6,11Rs and the
accompanying tactile allodynia.
In formalin-induced inflammatory pain, pre-treatment
with P2Y1, P2Y6, and P2Y11R agonists increased the
flinching behavior in the ipsilateral paw (Barragán-
Iglesias and others 2015). At the same time pretreatment
with selective antagonists of these receptors decreased
this behavioral reaction to pain. Western blot analysis
confirmed the presence of P2Y1, P2Y6 and P2Y11Rs in
dorsal root ganglia and sciatic nerve. It has been con-
cluded that these receptors have a pronociceptive role in
formalin-induced pain.
Accumulating evidence implicates that microglial
P2Y12Rs critically participate in the development of neu-
ropathic pain via p38 MAPK activation (Kobayashi and
others 2008). First, up-regulated expression of P2Y12Rs
Tang et al. 7
was observed on microglia after peripheral nerve injury
in rats (Kobayashi and others 2008; Tozaki-Saitoh and
others 2008). Secondly, inhibition of P2Y12Rs by intra-
thecal administration of the selective antagonists
AR-C69931MX or MRS2395, or the oral administration
of clopidogrel (a pro-drug of a P2Y12R antagonistic
intermediary metabolite), the antisense knockdown of
P2Y12R expression, or deletion of the P2ry12 gene, all
prevented the development of tactile allodynia and inhib-
ited thermal hyperalgesia (Kobayashi and others 2008;
Tozaki-Saitoh and others 2008).
The effects of adenosine on the nociceptive system
are quite divergent, because of the existence of 4 differ-
ent adenosine receptor-types and also because of the
localization of these receptors at different hierarchic lev-
els of the nervous system (Sawynok 2015; Sawynok and
Liu 2003; Sawynok and Sweeney 1989). The general
view is that A1R agonists produce pain-alleviating
actions in inflammatory and neuropathic pain models
that exhibit hyper-responsiveness (allodynia, hyperalge-
sia). A1Rs are located at peripheral sensory nerve end-
ings (Lima and others 2010), within the superficial layers
of the dorsal horn of the spinal cord (Ackley and others
2003) and at specific supraspinal sites (Maione and oth-
ers 2007) within the pain signaling neuraxis. A2ARs are
mostly neuronal but occur also at immunocytes; it
appears that with more prolonged inflammation, the loss
of inhibitory A2ARs on inflammatory cells would out-
weigh the loss of pronociceptive A2ARs on nerve end-
ings, such that there was now an overall increase in
nociceptive signaling (Sawynok 2015). A2B and A3Rs
have a preferential location at immune cells, and their
stimulation results in potentiation of many inflammatory
responses (Haskó and others 2009). The peripheral and
central effects may oppose each other and in conse-
quence the systemic application of A2BR ligands can
generate highly unpredictable actions; by contrast, recent
data supply convincing evidence for antinociception by
A3R agonists in neuropathic pain (Sawynok 2015). In
fact, A3R activation generated selective alleviation of
persistent neuropathic pain states and decreased spinal
cord pain processing by reducing the excitability of spi-
nal-wide dynamic range neurons and causing supraspi-
nal inhibition of spinal nociception via activation of
noradrenergic bulbospinal and serotonergic circuits
(Little and others 2015). Eventually, it is important to
note that adenosine itself has a higher affinity to A1 and
A2ARs than to A2B and A3Rs and therefore the modula-
tion of neuronal functions predominate when this purine
is released endogenously.
Adenosine receptors may be activated by the applica-
tion of adenosine itself or by drugs blocking the enzy-
matic degradation or tissue uptake of adenosine
(adenosine deaminase, deoxycoformycin; adenosine
Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
kinase, 5′-iodotubercidin; equilibrative adenosine trans-
porter, dipyridamole) thereby leading to higher endoge-
nous concentrations of the agonist. Positive allosteric
modulators are a further possibility to induce the recep-
tor-mediated effects with possibly lower incidence of
unwanted effects than observed in the case of the appli-
cation of adenosine or its agonistic structure analogues
(Jacobson and Müller 2016). The endogenous levels of
adenosine can be also raised by applying recombinant
ectonucleotidases facilitating the conversion of AMP to
adenosine (Zylka 2011). A single intrathecal injection of
secretory (non-membrane bound) versions of prostatic
acid phosphatase or ecto-5′-nucleotidase has long-last-
ing (2-3 days) antinociceptive effect in naïve mice and in
mouse models of inflammatory pain and neuropathic
pain ( Sowa and others 2010; Zylka and others 2008; see
also next section). In all cases, these antinociceptive
effects were dependent on A1R activation.
There are many selective antagonists for all 4 adenos-
ine receptor types available allowing the use of these
ligands according to necessity. It is interesting that caf-
feine, the stimulatory ingredient of black coffee or tea,
which is regularly consumed by a majority of the adult
population over the world, is a combined A2A/A2BR
antagonist (Sawynok 2011). In preclinical studies, caf-
feine produces intrinsic antinociceptive effects in several
rodent models, and augments the action of non-steroidal
anti-inflammatory drugs.
Acupuncture-Induced Analgesia and
Purinergic Signaling
Since the first report on acupurines demonstrated that (1)
applying weak EAP stimulation at acupoints Yanglingquan
(GB34) and Xuanzhong (GB39) prolonged the latency of
nociceptive hind limb withdrawal reflex and (2) intraperi-
toneal administration of the adenosine receptor antago-
nists theophylline and caffeine blocked the EAP-induced
elevation of the nociceptive threshold (Liu and others
1994), interest on the role of purines in AP analgesia has
continuously increased.
Acupuncture treatment releases ATP/ADP from kera-
tinocytes, the major cell type of the skin and from subcu-
taneous mast cells both during AP and moxibustion
(Fig. 4; Burnstock 2009b; Wang and others 2015; Yao
and others 2014). The outflow of the whole range of ATP
metabolites (ADP/AMP/adenosine) has been measured
in the neighboring interstitium by a microdialysis probe
after AP of the Zusanli point of mice (Goldman and oth-
ers 2010). A very similar spectrum of purines was
observed in the microdialysate after acupuncturing the
Zusanli point of human subjects; the concentration of
purines did not increase when the needle was only
inserted but not rotated (Takano and others 2012).
8 The Neuroscientist 
Figure 4.  Insertion and twisting of the acupuncture needles
releases ATP by mechanical stimulation of keratinocytes, the
major cell type of the skin and from subcutaneous mast cells,
without a direct damage to the respective cell membranes.
Before its rapid degradation to adenosine, ATP may stimulate
P2X3 and P2X2/3Rs localized on nociceptive terminals of
sensory ganglia (e.g., dorsal root ganglion [DRG]) neurons.
The signaling message is then relayed via the DRGs to the
spinal cord and subsequently through ascending pathways to
the brain stem that contains motor neurons which control
the functions of gut, lung, heart, arteries, and reproductive
organs, all major targets for acupuncture. Signals also travel to
certain centers in the cortex that perceive pain and localize
painful stimuli in the body. These centers can be modulated
by locally released adenosine to deliver a message to inhibit
pain. Reproduced from Burnstock (2011; artwork by Lucy
Reading-Ikkanda).
Purines were identified by high-performance liquid
chromatography. The extracellular concentration of ATP
returned soon to baseline after AP, whereas adenosine,
AMP and ADP remained significantly elevated (Goldman
and others 2010). ATP/ADP triggers the increase of cyto-
solic Ca2+ signaling in a cultured human fibroblast cell line
and this leads to the remodeling of its actin cytoskeleton
Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
(Goldman and others 2013). A similar situation may arise
under in vivo conditions and could contribute transient
changes in fibroblast cytoarchitecture that is probably
related to the beneficial local effects of AP.
In addition to the activation of P2Rs at fibroblast, P1/
P2Rs at intracutaneous sensory nerve terminals can also
be stimulated by the released ATP or its enzymatic degra-
dation products ADP and adenosine. In fact, free nerve
endings of both Aδ and C classes of sensory fibers are
located in the dermis and epidermis extending to the stra-
tum granulosum and are activated by ATP (Zhang and oth-
ers 2006). Based on these findings, a hypothesis was
forwarded on the crucial role of purines in AP-induced
analgesia (Burnstock 2009b). It has been suggested that
sensory nerve activity initiated in the skin by AP would
exert an inhibitory modulating effect on the spinoparabra-
chial and spinothalamic tracts to the brain pain centers by
a mechanism which has still to be elucidated. In the fol-
lowing passages we will report on experiments aimed at
supporting this “purinergic hypothesis of acupuncture.”
An exceptionally large number of studies have dealt
with the involvement of P2X3Rs (the structure of this
receptor was only recently clarified after successful crys-
tallization of the zebrafish (zf)P2X4R; Fig. 5A) in
AP-mediated alleviation of neuropathic pain (Fig. 5B).
Neuropathic pain in rats was induced by constricting one
sciatic nerve with a ligature (chronic constriction injury;
CCI). Then, the withdrawal threshold to pressure and the
withdrawal latency to radiant heat, as two parameters of
hyperalgesia and allodynia, respectively, have been mea-
sured for 14 days in total. During the first 7 days of obser-
vation, the two parameters gradually declined and then
levelled off at a minimum value. Starting with the sev-
enth day, EAP was applied to ipsilateral or contralateral
acupoints (Zusanli, ST36; Yanglinquan, GB34) for 30
minutes daily and for 7 days in total (Cheng and others
2013; Tu and others 2012; Wang and others 2014). This
treatment caused a gradual and moderate reversal of the
neuropathy symptoms by about 40%; the recovery was
almost identical for mechanical pain and temperature
sensation, and was practically indistinguishable for the
ipsi- and contralateral EAP. Intrathecal application of the
selective P2X3R antagonist A317491 on the 10th day,
immediately and dramatically facilitated the EAP-
induced analgesia (Wang and others 2014). Unfortunately,
A317491 was applied only in combination with EAP but
never alone and thereby it could not be decided whether
the effects of EAP and A317491 were additive or not. An
additive interaction would indicate that the mechanism of
analgesia is the same, mediated in both cases by the
exclusion of P2X3R function. However, equal efficiency
of ipsi- and contralateral EAP tentatively suggests that
this treatment acts at the spinal or supraspinal level rather
than at the peripheral terminals of DRG neurons.
Tang et al. 9
Figure 5.  Involvement of P2X3Rs in acupuncture (AP)
analgesia. (A) Homology model of the human (h)P2X3R
in its closed state generated on the basis of the published
crystallized zebrafish (zf)P2X4R structure (Kawate and
others 2009). The three subunits of the P2X3R are indicated
in different colors. The intracellular N- and C-termini of
zfP2X4Rs were not crystallized and are therefore missing
also in the homology model. The agonist binding pouches
are located at the border of two neighboring subunits at
the upper part of the receptor. Modified from Stephan
and others (2016). (B) Schematic presentation of the
effect of acupuncture against neuropathic pain symptoms;
the “mechanical withdrawal threshold” is a measure of
hyperalgesia and the “thermal withdrawal threshold” is
a measure of allodynia. Ligation of the right sciatic nerve
(chronic constriction injury [CCI]) caused 7 days later a
manifest hyperalgesia and allodynia in rats on the ipsilateral
side. Sham CCI on the contralateral side did not cause any
pain symptoms. Daily electroacupuncture (EAP) at ipsilateral
Zusanli (ST36) and Yanglingquan (GB34) points caused a
time-dependent improvement of these symptoms both
ipsi- and contralaterally during the following 7 days. When
the P2X3R antagonist A317491 was applied intrathecally, a
more dramatic improvement of the neuropathy symptoms
occurred. The schematic was made on the basis of figures 1
and 2 of Wang and others (2014).
Very similar results were obtained by another group of
researchers who used von Frey filaments and a hot plate
apparatus for measuring pain intensity after CCI and a some-
what different time-schedule; EAP to Zusanli-Yanglingquan
Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
acupoints 3 days after ligating the sciatic nerve again caused
moderate analgesia (Yu and others 2013). A possible weak-
ness of this study is that neuropathic pain may not reach its
maximum at this early time point and therefore the effect of
EAP cannot be reliably evaluated.
All studies equally demonstrated an increase of the
α,β-meATP-induced current amplitudes in acutely disso-
ciated DRG neurons prepared from rats which underwent
CCI, when compared with their non-operated counter-
parts (Cheng and others 2013; Tu and others 2012; Wang
and others 2014). When the DRG cells were prepared
from operated rats which obtained also an ipsi- or contra-
lateral EAP treatment, the α,β-meATP currents were
depressed but still larger than the current amplitudes mea-
sured in non-operated DRGs. Quantitative reverse tran-
scription–polymerase chain reaction (RT-PCR), in situ
hybridization, quantitative immunohistochemistry, and
Western blotting, all showed an increased appearance of
the P2X3R mRNA/protein in DRG neurons and the dor-
sal horn of the spinal cord after CCI to the unilateral sci-
atic nerve (Tu and others 2012, Yu and others 2013; Wang
and others 2014). EAP moderately counteracted this
increase, without any side preference.
Visceral pain is also supposed to react to AP. A model
of the painful irritable bowel syndrome was generated in
few days old rats by inflating a balloon in their terminal
colon/rectum twice daily for 14 days in total (Weng and
others 2013, 2015). These rats developed hypersensitivity
to bowel distension within a further period of 6 weeks,
elapsing without any pressure stimulation. An arbitrary
withdrawal reflex score was used to determine the inten-
sity of pain reaction caused by subsequent colorectal dis-
tension. In pressure-treated rats, the pain withdrawal
score increased in an EAP reversible manner. The expres-
sion levels of P2X3R mRNA (real-time RT-PCR) and
protein (quantitative immunohistochemistry) in DRGs
also responded with increase and decrease to balloon dis-
tension and EAP, respectively. These changes in P2X3R
expression were reflected by similar changes in the spinal
cord, prefrontal cortex and anterior cingulate cortex.
Thus, P2X3Rs appeared to become upregulated along the
neuronal pathways mediating pain to higher brain cen-
ters. Quantitative RT-PCR measurements were performed
in L4-L6 DRGs innervating the colon and rectum.
Nonetheless, we are left with some justified doubt,
because the criteria to prepare the spinal cord for RT-PCR
were not reported in the discussed papers. Of course the
spinal cord segments corresponding to the L4-L6 DRGs
should be prepared (optimally only their dorsal parts con-
taining the dorsal horn) and processed for RT-PCR.
In complete agreement with these findings, a rat neuro-
pathic pain model showed up-regulated P2X3R protein
levels in the midbrain periaqueductal gray (PAG), a crucial
site in endogenous pain modulatory systems (Xiao and
10 The Neuroscientist 
others 2010). EAP at Zusanli and Sanyinjiao (SP6) ipsilat-
eral to the side of the ligated nerve counteracted the
increase of the P2X3R protein, whereas sham-acupuncture
was ineffective. Interestingly, the down-regulated P2X3R
expression in the PAG with an antisense oligonucleotide
for the p2rx3 gene significantly attenuated the antinocicep-
tive effect of EAP. At the same time a mismatch antisense
oligonucleotide had no comparable action.
In conclusion, CCI as a model of neuropathic pain in
rats increased P2X3R expression and function in DRGs
as well as in pain-relevant areas of the spinal cord and the
brain. Both ipsi- and contralateral EAP invariably coun-
teracted these effects. It is reassuring that sham CCI (the
same operation but without nerve constriction) was rou-
tinely performed as a control. A visceral pain model
caused by balloon inflation in the colon/rectum delivered
similar data with respect to EAP. Hence, peripheral and
central P2X3Rs were decided to participate in the attenu-
ation of pain by EAP, although, P2X4 and P2X7Rs could
also contribute (see below). Unfortunately, sham EAP is
almost always missing from the experimental designs.
This is a major deficit, because the electrical stimulation
of acupuncture needles inserted at non-acupoints is a nec-
essary control procedure.
As mentioned earlier, ATP-mediated stimulation of
microglial P2X4Rs causes the secretion of BDNF onto
lamina I neurons of the spinal cord dorsal horn and is
thereby one of the causal factors of neuropathic pain
(Inoue and others 2007). Interferon-γ (IFN-γ) facilitated
the activation of microglia endowed with P2X4Rs (Tsuda
and others 2009). In consequence, excessive release of
IFN-γ after nerve injury may transform quiescent spinal
microglia into an activated state associated with P2X4R
expression. It has been reported that in rats with CCI,
EAP significantly increased the paw withdrawal thresh-
old relative to controls (Chen and others 2015). According
to expectations, EAP down-regulated both P2X4R and
IFN-γ expression in the spinal cord after a preceding up-
regulation of these proteins by CCI.
EAP in rats at the Shangjuxu (ST37) and Tianshu
(ST25) acupoints improved visceral hypersensitivity
after colorectal distension (Guo and others 2013).
Accordingly, P2X4R immunoreactivity in colon and spi-
nal cord was also decreased after EAP. In another study,
moxibustion at Dachangsu (BL25) normalized the
abdominal reflex score after colorectal balloon distension
of rats (Liu and others 2015). The up-regulation of
P2X7Rs in L6-S1 DRGs caused by balloon distension was
also prevented by moxibustion as documented by quanti-
tative immunohistochemistry, real-time RT-PCR, and
western blotting. However, it is unclear whether P2X4 or
P2X7Rs rather than e.g. P2X3Rs are primarily involved
in the analgesic effect of EAP. In other words, the inhibi-
tory effect of pharmacological antagonists, antisense
Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
oligonucleotides or siRNA should document the effec-
tiveness of the blockade of P2X4,7Rs on visceral pain.
The significance of endogenous adenosine, generated
by enzymatic degradation of ATP, released locally after
AP stimulation, is much more clear-cut than the impor-
tance of ATP itself for the induction of analgesia. It has
been convincingly demonstrated by the Nedergaard
group that, in mice, after AP at a Zusanli point, consider-
able amounts of ATP appeared in the microdialysate
which became rapidly converted to ADP, AMP, and ade-
nosine; the rise in the local concentration of adenosine
was larger and longer lasting than that of ATP (see above;
Goldman and others 2010). These authors modelled
inflammatory pain by injecting complete Freund’s adju-
vant into the right paw of mice and neuropathic pain by
spared injury of the sciatic nerve. Injection of a selective
A1R antagonist (2-chloro-N6-cyclopentyladenosine;
CCPA) into the Zusanli point reduced both inflammatory
and neuropathic pain with equal efficacy. AP at the
Zusanli point also reduced both inflammatory and neuro-
pathic pain, apparently through the accumulation of ade-
nosine and the consecutive A1R stimulation. This causal
relationship has been confirmed by experiments showing
that AP failed to reduce pain in A1R-deficient mice.
Adenosine is degraded to the pharmacologically inactive
inosine by adenosine deaminase. Blockade of adenosine
deaminase by deoxycofomycin led to increased and pro-
longed adenosine concentrations in the microdialysate; at
the same time, deoxycoformycin considerably length-
ened the duration of analgesia after AP.
Further support to this finding has been lent by the
injection of the selective A1R agonist N6-
cyclopentyladenosine (CPA) into the popliteal fossa
(Weizhong acupoint; UB40) of mice, while monitoring
thermal and mechanical hypersensitivity in neuropathic
and inflammatory pain models (Hurt and Zylka 2012).
CPA caused analgesia which was antagonized by the selec-
tive A1R antagonist 8-cyclopropyl-1,3-dipropylxanthine
(CPX). In further experiments, the endogenous levels of
adenosine were manipulated by injecting a secretory ver-
sion of human prostatic acid phosphatase (hPAP) into the
popliteal fossa of mice. As already mentioned previously,
hPAP generates adenosine by dephosphorylating AMP and
thereby largely increases the local concentration of adenos-
ine. Interestingly, hPAP injected to the popliteal fossa
inhibited thermal sensitivity for 3 days in the injected leg
of wild-type but not A1R-deficient mice, demonstrating a
critical requirement for A1R activation.
The results of the Nedergaard and Zylka groups
unequivocally confirm that increases in the local adenos-
ine concentration either by increased enzymatic produc-
tion (hPAP) or decreased enzymatic degradation
(deoxycofomycin) both boost analgesia at certain pain-
relevant acupoints (Zusanli, Weizhong). These effects
Tang et al. 11

Figure 6.  Acupuncture family procedures alleviate pain. These techniques (physical or chemical approaches, see Fig. 1) can be
applied, for example, to Zusanli (ST36) to release purines into the interstitial space both locally and probably also in the spinal
cord/brain. ATP, ADP, AMP, and adenosine activate pain-relevant purinoceptors of the P1 and P2 classes at neurons and glial
cells (see Table 1). UTP may also be released, degraded to UDP and stimulate some purinoceptors. We suggest that further
complexity is added to AP-induced analgesia via interactions between these receptors in the plasma membrane or interactions
between their downstream signaling mechanisms. Single nucleotide polymorphisms of the receptors or the generation of splice
variants may result in gain-of-function or loss-of-function variants. CAR1-4, four subtypes of G protein-linked cAMP receptors;
AC, adenylcyclase; PI-PLC, phosphatidylinositol-bisphosphate phosphodiesterase; PLA, phospholipase A; Src, Src tyrosinkinase;
cAMP, cyclic adenosine monophosphate; DG, diacylglycerol; IP3, inositoltrisphosphate; PGs, prostaglandins; MAPKs, mitogen
activated protein kinases; PKC, protein kinase C; PKA, protein kinase A; Akt, protein kinase B; GSK, GS kinase; CaMK, Ca2+/
calmodulin-dependent protein kinase; RhoK, Rho kinase; CREB, cAMP response element-binding protein; STAT3, signal
transducer and activator of transcription 3.

were caused by A1R activation. In addition, A2ARs have
been reported to mediate anti-inflammation by EAP on
synovitis in collagen-induced arthritis (Li and others
2015). EAP treatment resulted in significantly reduced
pathological changes in joint tissue; the anti-inflamma-
tory and tissue protective effects of EAP were reversed by
co-application of the A2AR antagonistic SCH58261.
ATP versus Adenosine Effects:
Which One Is Responsible for
Acupuncture-Induced Analgesia?
In the past couple of years purines have been suggested to
mediate AP-induced analgesia. This is a most important
issue, because it complements a theory according to which
mainly central opioids are responsible for analgesia
caused by AP, and in addition it provides a natural scien-
tific explanation for the observed effectiveness of AP in
human patients. It is hypothesized that the mechanical or
electrical stimulation of acupoints by AP needles causes
the release of ATP, which can be enzymatically degraded
to a whole range of biologically active metabolites (ATP,
ADP, AMP, adenosine). ATP causes pain by acting at
P2X3,4,7Rs, while ADP and/or UTP cause analgesia by
acting at P2Y1,12 or pain by acting at P2Y2Rs. Eventually,
adenosine has an analgesic activity especially by A1R
stimulation, although under special conditions (inflamma-
tory pain, A2ARs; neuropathic pain, A3Rs) other adenos-
ine receptor-types may also participate.
Around the turn of this decade, strong arguments were
forwarded which confirm that A1Rs mediate AP effects
for inflammatory and neuropathic pain. The concentra-
tion of endogenous adenosine at the site of AP could be

Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
12 The Neuroscientist 
raised by a blockade of its metabolic elimination, or by an
enhancement of its production from the precursor AMP;
the intensity of AP-induced pain mirrored these manipu-
lations. Moreover, A1R antagonistic methylxanthines
counteracted the effect of AP.
It was only after establishing the causal relationship
between AP and adenosine-induced analgesia that many
scientists started to elucidate the possible participation of
P2XRs in this process. Numerous authors demonstrated
an increased expression and function of these P2XR-
types during inflammatory, neuropathic and visceral pain
at different levels of the peripheral and central nervous
system. It has also been shown that AP on the one hand
induced analgesia and on the other hand decreased the
facilitated expression and function of P2X3,4,7R-types.
However, the modulation of P2XR-types might be the
consequence rather than the reason for anti-nociception.
In other words, AP may act via A1Rs, and as a result of
the normalized pain sensation, P2XR up-regulation may
return to quasi normal levels.
A real deficit of most studies is that a selective antago-
nist for P2X3Rs has been used only in some cases and
P2X4 and P2X7R antagonists have not been utilized at
all. Of course, the modulation of ATP degradation by
ecto-ATPase inhibitors would not really help to differen-
tiate between ATP and adenosine as executors of
AP-induced analgesia, because an increased level of local
ATP concentration would eventually result in a corre-
spondingly higher level of adenosine.
Perspectives and Open Questions
Further interesting questions also wait for an unequivocal
answer:
1. Purines: So far, the concentration of purines (ATP,
ADP, AMP, adenosine) following AP have only
been detected at the puncture site. Do any changes
occur in the concentration of purines in response
to AP in pain-relevant areas of the brain? This
approach would help to differentiate between
peripheral and central effects of purines by AP.
2. Purinoceptors: Apart from A1, A2A, P2X3,
P2X4, and P2X7Rs, do also other pain-relevant
purine receptors (Table 1) participate in
AP-induced analgesia? In fact, it is still unclear,
how P2XR stimulation by ATP released in
response to AP can directly generate analgesia.
There are two main possibilities: (a) The stimula-
tion of sensory afferents activates the opioid sys-
tem suppressing the nociceptive system. An
attractive alternative is that (b) the stimulation of
sensory afferents by AP excludes the excitation of
pain neurons either in the dorsal horn spinal horn
or in the pain centers of the brain (Burnstock
Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
2009b). In addition, it is most likely that interac-
tions between different purine receptors (e.g.,
P2X4 and P2X7 or P2Y1 and P2X3) (Fig. 6) or
between pain-sensitive purinoceptors and non-
purinoceptors (e.g., TRPV1 and P2X3) can fine-
tune the analgesic effect of AP. Desensitization of
P2X3Rs at the peripheral or central terminals of
sensory ganglia neurons and the consequent inter-
ruption of pain signaling is an additional possibil-
ity. Eventually, single nucleotide polymorphisms
and transcriptional regulation of the involved
receptors might also modify the alleviation of
pain by AP. In contrast to various P2X and ARs,
the involvement of P2YRs in AP-induced analge-
sia is hitherto a black box. Such investigations are
badly needed, especially because combination of
AP with adenosine and ADP agonists acting as
presynaptic inhibitors of transmitter release via
A1 and P2Y1Rs at synapses involved in pain
pathways might be a novel therapeutic strategy
for the treatment of pain.
3. New experimental tools: It might be most helpful
to introduce new experimental tools in order to
elucidate the role of purinoceptors in AP-induced
analgesia such as optogenetics, Designer
Receptors Exclusively Activated by Designer
Drugs (DREADD), and so on.
4. Acupuncture in humans: It is most likely, but by
no way proven without any doubt that ATP/ade-
nosine participate in AP-induced analgesia in
humans. Thus, a rest of uncertainty remains about
the translation of data from laboratory animals to
humans. Almost all animal experiments were car-
ried out by EAP, which is a more powerful stimu-
lus than AP usually applied in clinics. Of course,
it should also be kept in mind that higher concen-
trations of ATP cause pain, while comparable con-
centrations of adenosine relieve pain. The various
mechanisms by which low, endogenously released
concentrations of the excitatory ATP could pro-
duce analgesia have been discussed in the section
“Clinical Effectiveness and Neurobiological
Basis of Acupuncture Analgesia.”
Adenosine has been tested in clinical trials and
showed analgesic effect on intravenous or intra-
thecal application (Hayashida and others 2005).
Injection of prostatic acid phosphatase (termed as
PAPuncture by Zylka and others 2008) into the
acupoint Weizhong exhibited analgesia for an
extended period of time. Along this line of think-
ing enzymatically stable analogs of ATP at low
concentrations and adenosine could be injected
into various acupoints. Finally, the observation
that AP appears to be less efficacious in the
Western population than in China could be due to
Tang et al. 13
the regular consumption of coffee with a high
content of the adenosine antagonistic caffeine by
Europeans and Americans, whereas in China, tea
is the favored stimulant with a much lower con-
tent of caffeine.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with
respect to the research, authorship, and/or publication of this
article.
Funding
The author(s) disclosed receipt of the following financial sup-
port for the research, authorship, and/or publication of this arti-
cle: This work has been supported by grants from Deutsche
Forschungsgemeinschaft (IL 20/21-1), Sino-German Centre
(GZ919), Sichuan Provincial Innovative Research Team
Program (2014TD0018, 2015TD0010), Innovative Research
Team in University of Sichuan Province (16TD0015), and
National Natural Science Foundation of China (81373735).
References
Abbracchio MP, Burnstock G. 1994. Purinoceptors: are there
families of P2X and P2Y purinoceptors? Pharmacol Ther
64:445–75.
Abbracchio MP, Burnstock G, Verkhratsky A, Zimmermann H.
2009. Purinergic signalling in the nervous system: an over-
view. Trends Neurosci 32:19–29.
Ackley MA, Governo RJ, Cass CE, Young JD, Baldwin SA,
King AE. 2003. Control of glutamatergic neurotrans-mis-
sion in the rat spinal dorsal horn by the nucleoside trans-
porter ENT1. J Physiol 548:507–17.
Andó RD, Méhész B, Gyires K, Illes P, Sperlágh B. 2010.
A comparative analysis of the activity of ligands acting
at P2X and P2Y receptor subtypes in models of neuro-
pathic, acute and inflammatory pain. Br J Pharmacol
159:1106–17.
Barragán-Iglesias P, Mendoza-Garcés L, Pineda-Farias JB,
Solano-Olivares V, Rodríguez-Silverio J, Flores-Murrieta
FJ, and others. 2015. Participation of peripheral P2Y1,
P2Y6 and P2Y11 receptors in formalin-induced inflamma-
tory pain in rats. Pharmacol Biochem Behav 128:23–32.
Barragán-Iglesias P, Pineda-Farias JB, Bravo-Hernández
M, Cervantes-Durán C, Price TJ, Murbartián J, and oth-
ers. 2016. Predominant role of spinal P2Y1 receptors in
the development of neuropathic pain in rats. Brain Res
1636:43–51.
Barragán-Iglesias P, Pineda-Farias JB, Cervantes-Durán C,
Bravo-Hernández M, Rocha-González HI, Murbartián J,
and others. 2014. Role of spinal P2Y6 and P2Y11 receptors
in neuropathic pain in rats: possible involvement of glial
cells. Mol Pain 10:29.
Berman BM, Langevin HM, Witt CM, Dubner R. 2010.
Acupuncture for chronic low back pain. N Engl J Med
363:454–61.
Burnstock G. 1972. Purinergic nerves. Pharmacol Rev
24:509–81.
Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
Burnstock G. 2006. Purinergic P2 receptors as targets for novel
analgesics. Pharmacol Ther 110:433–54.
Burnstock G. 2009a. Purinergic receptors and pain. Curr Pharm
Des 15:1717–35.
Burnstock G. 2009b. Acupuncture: a novel hypothesis for the
involvement of purinergic signalling. Med Hypotheses
73:470–72.
Burnstock G. 2011. Puncturing the myth. Purinergic signaling,
not mystical energy, may explain how acupuncture works.
Scientist 25:24–5.
Burnstock G. 2016. Purinergic mechanisms and pain. Adv
Pharmacol 75:91–137.
Burnstock G, Krügel U, Abbracchio MP, Illes P. 2011.
Purinergic signalling: from normal behaviour to pathologi-
cal brain function. Prog Neurobiol 95:229–74.
Cabot PJ, Carter L, Gaiddon C, Zhang Q, Schafer M, Loeffler
JP, and others. 1997. Immune cell-derived β-endorphin.
Production, release, and control of inflammatory pain in
rats. J Clin Invest 100:142–8.
Cao H, Li X, Liu J. 2012. An updated review of the efficacy of
cupping therapy. PLoS One 7:e31793.
Carroll WA, Donnelly-Roberts D, Jarvis MF. 2009. Selective
P2X7 receptor antagonists for chronic inflammation and
pain. Purinergic Signal 5:63–73.
Chen CC, Akopian AN, Sivilotti L, Colquhoun D, Burnstock G,
Wood JN. 1995. A P2X purinoceptor expressed by a subset
of sensory neurons. Nature 377:428–31.
Chen XM, Xu J, Song JG, Zheng BJ, Wang XR. 2015.
Electroacupuncture inhibits excessive interferon-γ evoked
up-regulation of P2X4 receptor in spinal microglia in a CCI
rat model for neuropathic pain. Br J Anaesth 114:150–7.
Cheng RD, Tu WZ, Wang WS, Zou EM, Cao F, Cheng B, and
others (2013) Effect of electroacupuncture on the patho-
morphology of the sciatic nerve and the sensitization of
P2X3 receptors in the dorsal root ganglion in rats with
chronic constrictive injury. Chin J Integr Med 19:374–9.
Cheng X. 2010. Chinese acupuncture and moxibustion. 3rd ed.
Beijing, China: Foreign Languages Press.
Chessell IP, Hatcher JP, Bountra C, Michel AD, Hughes JP,
Green P, and others. 2005. Disruption of the P2X7 puri-
noceptor gene abolishes chronic inflammatory and neuro-
pathic pain. Pain 114:386–96.
Colquhoun D, Novella SP. 2013. Acupuncture is theatrical pla-
cebo. Anesth Analg 116:1360–3.
Costa-Junior HM, Sarmento VF, Coutinho-Silva R. 2011. C ter-
minus of the P2X7 receptor: treasure hunting. Purinergic
Signal 7:7–19.
Coull JA, Beggs S, Boudreau D, Boivin D, Tsuda M, Inoue K,
and others. 2005. BDNF from microglia causes the shift
in neuronal anion gradient underlying neuropathic pain.
Nature 438:1017–21.
Donnelly-Roberts D, McGaraughty S, Shieh CC, Honore P,
Jarvis MF. 2008. Painful purinergic receptors. J Pharmacol
Exp Ther 324:409–15.
Ernst E, Lee MS, Choi TY. 2011. Acupuncture: does it alleviate
pain and are there serious risks? A review of reviews. Pain
152:755–64.
Fredholm BB, IJzerman AP, Jacobson KA, Linden J, Müller
CE. 2011. International Union of Basic and Clinical
14 The Neuroscientist 
Pharmacology. LXXXI. Nomenclature and classification of
adenosine receptors—an update. Pharmacol Rev 63:1–34.
Fu SF, Kun W, Zeng XX, Zhang L, Cheng CW, Song L, and
others. 2016. Urgent need to improve the quality of case
report in traditional Chinese medicine: assessment on
reporting quality of 3,417 cases. Chin J Integr Med [Epub
ahead of print].
Gerevich Z, Borvendég SJ, Schröder W, Franke H, Wirkner
K, Nörenberg W, and others. 2004. Inhibition of N-type
voltage-activated calcium channels in rat dorsal root gan-
glion neurons by P2Y receptors is a possible mechanism of
ADP-induced analgesia. J Neurosci 24:797–807.
Gerevich Z, Illes P. 2004. P2Y receptors and pain transmission.
Purinergic Signal 1:3–10.
Gerevich Z, Zádori Z, Müller C, Wirkner K, Schröder W, Rubini
P, and others. 2007. Metabotropic P2Y receptors inhibit
P2X3 receptor-channels via G protein-dependent facilita-
tion of their desensitization. Br J Pharmacol 151:226–36.
Goldman N, Chandler-Militello D, Langevin HM, Nedergaard
M, Takano T. 2013. Purine receptor mediated actin cyto-
skeleton remodeling of human fibroblasts. Cell Calcium
53:297–301.
Goldman N, Chen M, Fujita T, Xu Q, Peng W, Liu W, and
others. 2010. Adenosine A1 receptors mediate local
anti-nociceptive effects of acupuncture. Nat Neurosci
13:883–8.
Guo X, Chen J, Lu Y, Wu L, Weng Z, Yang L, and others. 2013.
Electroacupuncture at He-Mu points reduces P2X4 recep-
tor expression in visceral hypersensitivity. Neural Regen
Res 8:2069–77.
Han JS. 2003. Acupuncture: neuropeptide release produced
by electrical stimulation of different frequencies. Trends
Neurosci 26:17–22.
Haskó G, Csóka B, Németh ZH, Vizi ES, Pacher P. 2009. A2B
adenosine receptors in immunity and inflammation. Trends
Immunol 30:263–70.
Hayashida M, Fukuda K, Fukunaga A. 2005. Clinical appli-
cation of adenosine and ATP for pain control. J Anesth
19:225–35.
Honore P, Donnelly-Roberts D, Namovic MT, Hsieh G,
Zhu CZ, Mikusa JP, and others. 2006. A-740003 [N-(1-
{[(cyanoimino)(5-quinolinylamino) methyl]amino}-2,
2-dimethylpropyl)-2-(3,4-dimethoxyphenyl)acetamide], a
novel and selective P2X7 receptor antagonist, dose-depend-
ently reduces neuropathic pain in the rat. J Pharmacol Exp
Ther 319:1376–85.
Hurt JK, Zylka MJ. 2012. PAPupuncture has localized and
long-lasting antinociceptive effects in mouse models of
acute and chronic pain. Mol Pain 8:28.
Inoue K, Tsuda M, Tozaki-Saitoh H. 2007. Modification of
neuropathic pain sensation through microglial ATP recep-
tors. Purinergic Signal 3:311–6.
Irnich D, Beyer A. 2002. Neurobiological mechanisms of acu-
puncture analgesia. Schmerz 16:93–102.
Jacobson KA, Müller CE. 2016. Medicinal chemistry of ade-
nosine, P2Y and P2X receptors. Neuropharmacology
104:31–49.
Johnson MI. 2006. The clinical effectiveness of acupuncture for
pain relief—you can be certain of uncertainty. Acupunct
Med 24:71–9.
Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
Kawate T, Michel JC, Birdsong WT, Gouaux E. 2009. Crystal
structure of the ATP-gated P2X4 ion channel in the closed
state. Nature 460:592–8.
Kennedy C, Assis TS, Currie AJ, Rowan EG. 2003. Crossing
the pain barrier: P2 receptors as targets for novel analge-
sics. J Physiol 553:683–94.
Kobayashi K, Yamanaka H, Fukuoka T, Dai Y, Obata K,
Noguchi K. 2008. P2Y12 receptor upregulation in acti-
vated microglia is a gateway of p38 signaling and neuro-
pathic pain. J Neurosci 28:2892–902.
Köles L, Fürst S, Illes P. 2007. Purine ionotropic (P2X) recep-
tors. Curr Pharm Des 13:2368–84.
Köles L, Gerevich Z, Oliveira JF, Zádori ZS, Wirkner K, Illes
P. 2008. Interaction of P2 purinergic receptors with cellular
macromolecules. Naunyn Schmiedebergs Arch Pharmacol
377:1–33.
Köles L, Kató E, Hanuska A, Zádori ZS, Al-Khrasani A, Zelles
T, and others. 2016. Modulation of excitatory neurotrans-
mission by neuronal/glial signalling molecules: interplay
between purinergic and glutamatergic systems. Purinergic
Signal 12:1–24.
Lazarowski ER. 2012. Vesicular and conductive mechanisms of
nucleotide release. Purinergic Signal 8:359–373.
Lee MS, Ernst E. 2011. Acupuncture for pain: an overview of
Cochrane reviews. Chin J Integr Med 17:187–9.
Lewis C, Neidhart S, Holy C, North RA, Buell G, Surprenant A.
1995. Coexpression of P2X2 and P2X3 receptor subunits
can account for ATP-gated currents in sensory neurons.
Nature 377:432–5.
Lewith GT, White PJ, Pariente J. 2005. Investigating acupunc-
ture using brain imaging techniques: the current state of
play. Evid Based Complement Alternat Med 2:315–9.
Li QH, Xie WX, Li XP, Huang KT, Du ZH, Cong WJ, and oth-
ers. 2015. Adenosine A2A receptors mediate anti-inflam-
matory effects of electroacupuncture on synovitis in mice
with collagen-induced arthritis. Evid Based Complement
Alternat Med 2015:809560.
Lima FO, Souza GR, Verri WA., Jr, Parada CA, Ferreira SH,
Cunha FQ, and others. 2010. Direct blockade of inflamma-
tory hypernociception by peripheral A1 adenosine recep-
tors: involvement of the NO/cGMP/PKG/KATP signaling
pathway. Pain 151:506–15.
Little JW, Ford A, Symons-Liguori AM, Chen Z, Janes K,
Doyle T, and others. 2015. Endogenous adenosine A3
receptor activation selectively alleviates persistent pain
states. Brain 138:28–35.
Liu C, Zhao F, Zhu L. 1994. Involvement of purines in analge-
sia produced by weak electro-acupuncture. Zhen Ci Yan
Jiu 19:59–62.
Liu S, Shi Q, Zhu Q, Zou T, Li G, Huang A, and others. 2015.
P2X7 receptor of rat dorsal root ganglia is involved in the
effect of moxibustion on visceral hyperalgesia. Purinergic
Signal 11:161–9.
Loyeung BYK, Cobbin DM. 2013. Investigating the effects of
three needling parameters (manipulation, retention time,
and insertion site) on needling sensation and pain profiles:
a study of eight deep needling interventions. Evid Based
Complement Alternat Med 2013:136763.
Lu KW, Hsu CK, Hsieh CL, Yang J, Lin YW. 2016. Probing the
effects and mechanisms of electroacupuncture at ipsilateral
Tang et al. 15
or contralateral ST36-ST37 acupoints on CFA-induced
inflammatory pain. Sci Rep 6:22123.
Madsen MV, Gotzsche PC, Hrobjartsson A. 2009. Acupuncture
treatment for pain: systematic review of randomised clini-
cal trials with acupuncture, placebo acupuncture, and no
acupuncture groups. BMJ 338:a3115.
Maione S, de Novelis V, Cappellacci L, Palazzo E, Vita D, Luongo
L, and others. 2007. The antinociceptive effect of 2-chloro-
2′-C-methyl-N6-cyclopentyladenosine (2′-Me-CCPA), a
highly selective adenosine A1 receptor agonist, in the rat.
Pain 131:281–92.
Moriyama T, Iida T, Kobayashi K, Higashi T, Fukuoka T,
Tsumura H, and others. 2003. Possible involvement of
P2Y2 metabotropic receptors in ATP-induced transient
receptor potential vanilloid receptor 1-mediated thermal
hypersensitivity. J Neurosci 23:6058–62.
Nelson DW, Gregg RJ, Kort ME, Perez-Medrano A, Voight EA,
Wang Y, and others. 2006. Structure-activity relationship
studies on a series of novel, substituted 1-benzyl-5-phenyl-
tetrazole P2X7 antagonists. J Med Chem 49:3659–66.
NIH Consensus Conference. 1998. Acupuncture. JAMA
280:1518–24.
Okada M, Nakagawa T, Minami M, Satoh M. 2002. Analgesic
effects of intrathecal administration of P2Y nucleotide
receptor agonists UTP and UDP in normal and neuropathic
pain model rats. J Pharmacol Exp Ther 303:66–73.
Pomeranz B, Chiu D. 1976. Naloxone blockade of acupuncture
analgesia: endorphin implicated. Life Sci 19:1757–62.
Ralevic V, Burnstock G. 1998. Receptors for purines and
pyrimidines. Pharmacol Rev 50:413–92.
Sawynok J. 2011. Methylxanthines and pain. Handb Exp
Pharmacol. 311–29.
Sawynok J. 2015. Adenosine receptor targets for pain.
Neuroscience [Epub ahead of print].
Sawynok J, Liu XJ. 2003. Adenosine in the spinal cord and
periphery: release and regulation of pain. Prog Neurobiol
69:313–40.
Sawynok J, Sweeney MI. 1989. The role of purines in nocicep-
tion. Neuroscience 32:557–69.
Sjölund B, Terenius L, Eriksson M. 1977. Increased cerebro-
spinal fluid levels of endorphins after electro-acupuncture.
Acta Physiol Scand 100:382–4.
Sowa NA, Voss MK, Zylka MJ. 2010. Recombinant ecto-5′-
nucleotidase (CD73) has long lasting antinociceptive
effects that are dependent on adenosine A1 receptor activa-
tion. Mol Pain 6:20.
Sperlágh B, Illes P. 2014. P2X7 receptor: an emerging target
in central nervous system diseases. Trends Pharmacol Sci
35:537–47.
Sperlágh B, Vizi ES, Wirkner K, Illes P. 2006. P2X7 receptors
in the nervous system. Prog Neurobiol 78:327–46.
Staud R, Price DD (2006) Mechanisms of acupuncture anal-
gesia for clinical and experimental pain. Expert Rev
Neurother 6:661–7.
Stephan G, Kowalski-Jahn M, Zens C, Schmalzing G, Illes P,
Hausmann R. 2016. Inter-subunit disulfide locking of the
human P2X3 receptor elucidates ectodomain movements
associated with channel gating. Purinergic Signal 12:
221–33.
Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
Syhr KM, Kallenborn-Gerhardt W, Lu R, Olbrich K, Schmitz
K, Mannich J, and others. 2014. Lack of effect of a P2Y6
receptor antagonist on neuropathic pain behavior in mice.
Pharmacol Biochem Behav 124:389–95.
Takano T, Chen X, Luo F, Fujita T, Ren Z, Goldman N, and oth-
ers. 2012. Traditional acupuncture triggers a local increase
in adenosine in human subjects. J Pain 13:1215–23.
Takeshige C, Sato T, Mera T, Hisamitsu T, Fang J. 1992.
Descending pain inhibitory system involved in acupuncture
analgesia. Brain Res Bull 29:617–34.
Tozaki-Saitoh H, Tsuda M, Miyata H, Ueda K, Kohsaka S,
Inoue K. 2008. P2Y12 receptors in spinal microglia are
required for neuropathic pain after peripheral nerve injury.
J Neurosci 28:4949–56.
Tsuda M, Masuda T, Kitano J, Shimoyama H, Tozaki-Saitoh
H, Inoue K. 2009. IFN-γ receptor signaling mediates spinal
microglia activation driving neuropathic pain. Proc Natl
Acad Sci U S A 106:8032–7.
Tu WZ, Cheng RD, Cheng B, Lu J, Cao F, Lin HY, and others.
2012. Analgesic effect of electroacupuncture on chronic
neuropathic pain mediated by P2X3 receptors in rat dorsal
root ganglion neurons. Neurochem Int 60:379–86.
Vickers AJ, Cronin AM, Maschino AC, Lewith G, MacPherson
H, Foster NE, and others. 2012. Acupuncture Trialists’
Collaboration. Acupuncture for chronic pain: individual
patient data meta-analysis. Arch Intern Med 172:1444–53.
Vickers AJ, Linde K. 2014. Acupuncture for chronic pain.
JAMA 311:955–6.
Villa G, Fumagalli M, Verderio C, Abbracchio MP, Ceruti S.
2010. Expression and contribution of satellite glial cells
purinoceptors to pain transmission in sensory ganglia: an
update. Neuron Glia Biol 6:31–42.
Wang L, Hu L, Grygorczyk R, Shen X, Schwarz W. 2015.
Modulation of extracellular ATP content of mast cells
and DRG neurons by irradiation: studies on underlying
mechanism of low-level-laser therapy. Mediators Inflamm
2015:630361.
Wang SM, Kain ZN, White PF. 2008. Acupuncture analgesia:
II. Clinical considerations. Anesth Analg 106:611–21.
Wang WS, Tu WZ, Cheng RD, He R, Ruan LH, Zhang L, and
others. 2014. Electroacupuncture and A-317491 depress
the transmission of pain on primary afferent mediated by
the P2X3 receptor in rats with chronic neuropathic pain
states. J Neurosci Res 92:1703–13.
Weng Z, Wu L, Lu Y, Wang L, Tan L, Dong M, and others.
2013. Electroacupuncture diminishes P2X2 and P2X3
purinergic receptor expression in dorsal root ganglia of rats
with visceral hypersensitivity. Neural Regen Res 8:802–8.
Weng ZJ, Wu LY, Zhou CL, Dou CZ, Shi Y, Liu HR, and oth-
ers. 2015. Effect of electroacupuncture on P2X3 receptor
regulation in the peripheral and central nervous systems of
rats with visceral pain caused by irritable bowel syndrome.
Purinergic Signal 11:321–9.
White A, Ernst E. 2004. A brief history of acupuncture.
Rheumatology (Oxford) 43:662–3.
Wirkner K, Sperlágh B, Illes P. 2007. P2X3 receptor involve-
ment in pain states. Mol Neurobiol 36:165–83.
Wolf G, Yirmiya R, Goshen I, Iverfeldt K, Holmlund L, Takeda K,
and others. 2003. Impairment of interleukin-1 (IL-1) signaling
16 The Neuroscientist 
reduces basal pain sensitivity in mice: genetic, pharmacologi-
cal and developmental aspects. Pain 104:471–80.
World Health Organization. 2002. Acupuncture: review and
analysis of reports on controlled clinical trials. Geneva,
Switzerland: World Health Organization. http://www.who.
int/iris/handle/10665/42414
Xiao Z, Ou S, He WJ, Zhao YD, Liu XH, Ruan HZ. 2010.
Role of midbrain periaqueductal gray P2X3 receptors in
electroacupuncture-mediated endogenous pain modulatory
systems. Brain Res.1330:31–44.
Yao W, Yang H, Yin N, Ding G. 2014. Mast cell-nerve cell
interaction at acupoint: modeling mechanotransduction
pathway induced by acupuncture. Int J Biol Sci 10:511–9.
Yegutkin GG. 2008. Nucleotide- and nucleoside-converting
ectoenzymes: important modulators of purinergic signal-
ling cascade. Biochim Biophys Acta 1783:673–94.
Yu J, Zhao C, Luo X. 2013. The effects of electroacupunc-
ture on the extracellular signal-regulated kinase 1/2/P2X3
Downloaded from nro.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on June 28, 2016
signal pathway in the spinal cord of rats with chronic con-
striction injury. Anesth Analg 116:239–46.
Zhang Q, Zhao Y, Guo Y, Cao DY, Tang XD, Tian YL, and
others. 2006. Activation and sensitization of C and Aδ
afferent fibers mediated by P2X receptors in rat dorsal skin.
Brain Res 1102:78–85.
Zhang R, Lao L, Ren K, Berman BM. 2014. Mechanisms
of acupuncture-electroacupuncture on persistent pain.
Anesthesiology 120:482–503.
Zhao ZQ. 2008. Neural mechanism underlying acupuncture
analgesia. Prog Neurobiol 85:355–75.
Zylka MJ. 2011. Pain-relieving prospects for adenosine
receptors and ectonucleotidases. Trends Mol Med 17:
188–96.
Zylka MJ, Sowa NA, Taylor-Blake B, Twomey MA, Herrala
A, Voikar V, and others. 2008. Prostatic acid phosphatase
is an ectonucleotidase and suppresses pain by generating
adenosine. Neuron 60:111–22.