The neuropathic pain triad: neurons, immune cells
Joachim Scholz & Clifford J Woolf
Nociceptive pain results from the detection of intense or noxious stimuli by specialized high-threshold sensory neurons
(nociceptors), a transfer of action potentials to the spinal cord, and onward transmission of the warning signal to the brain.
In contrast, clinical pain such as pain after nerve injury (neuropathic pain) is characterized by pain in the absence of a stimulus
and reduced nociceptive thresholds so that normally innocuous stimuli produce pain. The development of neuropathic pain involves
not only neuronal pathways, but also Schwann cells, satellite cells in the dorsal root ganglia, components of the peripheral immune
system, spinal microglia and astrocytes. As we increasingly appreciate that neuropathic pain has many features of a neuroimmune
disorder, immunosuppression and blockade of the reciprocal signaling pathways between neuronal and non-neuronal cells offer
new opportunities for disease modification and more successful management of pain.
Pain has long been regarded as the unpleasant sensory consequence of
neuronal activity in specific nociceptive pathways that is triggered by
noxious stimuli, inflammation, or damage to the nervous system. This
iscertainly trueforacute nociceptive pain,suchasthepainelicitedbya
pinprickorexcessiveheat.However,itisnowclear that neuronsarenot
the only players that drive the establishment and maintenance of
common clinical pain states. In this review we focus on immune and
glial cell responses to peripheral nerve injury and how they alter
neuronal function in the peripheral and central nervous systems.
Recognition of the critical involvement of immune cells and glia in
the pathophysiological changes after nerve injury offers a completely
new treatment approach, one that is certainly needed because most
analgesic drugs lack satisfactory efficacy for neuropathic pain and
produce undesirable side effects1. Neuropathic pain management is
currently aimed only at reducing symptoms, generally by suppressing
neuronal activity. In contrast, modulating the immune response to
nerve injury and targeting glia may provide opportunities for disease
modification by aborting neurobiological alterations that support the
development of persistent pain.
Peripheral nerve injury provokes a reaction in peripheral immune
Schwann cellsfacilitate the wallerian degeneration of axotomizednerve
fibers distal to a nerve lesion; an immune response in the dorsal root
ganglia (DRGs) is driven by macrophages, lymphocytes and satellite
the CNS to peripheral nerve injury, whichis followed by activation and
proliferation of astrocytes (Fig. 1). Macrophages, which are derived
from circulating monocytes, and microglia, the resident mono-
nuclear phagocytes of the CNS, share numerous similarities in their
immunological and functional properties2. Microglial cells are gene-
rally considered to stem from a monocytic cellular lineage of meso-
dermal (myeloid) origin and enter the CNS during fetal development3.
And although the satellite cells that surround the cell bodies of DRG
neurons originate from the neural crest, they can be regarded as the
for primary sensory neurons, express transporters that regulate neuro-
transmitter levels in the extracellular space and share some astroglial
markers, such as glial fibrillary acidic protein (GFAP)4.
Peripheral inflammatory reactions to nerve lesions
In contrast to primarily immune-mediated neuropathies, such as
Guillain-Barre ´ syndrome and chronic inflammatory demyelinating
polyneuropathy, which are triggered by T cell activation, macrophages
predominate in the initial inflammatory reaction to peripheral nerve
injury. Neutrophil granulocytes participate in the very early immune
response to nerve injury, potentially attracted by the release of nerve
leukotriene-B4. Although neutrophil infiltration is limited to the
immediate vicinity of the lesion site and relatively short-lived5,
chemoattractants and cytokines released from neutrophils may play
an important role by reinforcing the recruitment of macrophages,
particularly during the first 24 h after injury5.
Immediately after nerve injury, resident macrophages, which
account for up to 9% of the cell population in intact peripheral nerves,
rush tothe lesion site like a rapid-response team6. The recruitment and
activation of resident macrophages and the invasion of further mono-
cytes from the peripheral blood are orchestrated by the chemokine
(C-C motif) ligands (CCLs) 2 and 3 acting on the receptors CCR2,
Schwann cells secrete matrix metalloproteases that attack the basal
lamina of endoneurial blood vessels, leading to an interruption of the
blood-nerve barrier8. Vasoactive mediators including calcitonin gene–
related peptide (CGRP), substance P, bradykinin and nitric oxide are
released from injured axons to cause hyperemia and swelling9. These
GLIA AND DISEASE
Published online 26 October 2007; doi:10.1038/nn1992
Neural Plasticity Research Group, Department of Anesthesia and Critical Care,
Massachusetts General Hospital and Harvard Medical School, 149 13thStreet,
addressed to J.S. (email@example.com).
NATURE NEUROSCIENCE VOLUME 10 [ NUMBER 11 [ NOVEMBER 20071361
© 2007 Nature Publishing Group http://www.nature.com/natureneuroscience
vascular changes support the invasion of circulating immune cells, so
that within two days of injury a dense cellular infiltrate, mainly
composed of macrophages, T lymphocytes and mast cells, forms at
the lesion site (Figs. 1b and 2a). Upregulation of lysosomal markers
and an abundant inclusion of lipid droplets indicate that macrophages
transform into active phagocytes after nerve injury. Removal of
distal degenerating axons and myelin debris by phagocytosis enables
a reorganization of Schwann cells and lays the foundation for the
regrowth of injured axons10.
Within minutes of a nerve lesion, neuregulin, a growth and differ-
entiation factor present on the axonal membrane, induces activation of
the tyrosine kinase receptor ERBB2, which is constitutively expressed
on Schwann cells11(Fig. 2b). Blocking this early neuregulin–ERBB2
signaling pathway disrupts demyelination. Later in the course of
wallerian degeneration, ERBB2 and ERBB3
receptor upregulation and activation are asso-
ciated with Schwann cell proliferation12. In
the reverse direction, Schwann cells release
chemical signals that promote axonal growth
and remyelination13. These include NGF and
glial cell line–derived neurotrophic factor
(GDNF), which are retrogradely transported
to the cell bodies of primary sensory neurons
where they act as potent regulators of gene
expression. NGF and GDNF also directly
activate and sensitize nociceptors14, contri-
buting to the initiation of pain in response
to nerve injury (Fig. 2b).
Signaling pathways between primary sen-
are highly intertwined, and cytokines and
chemokines are central components in this
complex network. Schwann cells, active resi-
dent and infiltrating macrophages, neutrophil
granulocytes, and mast cells release prosta-
glandins15, proinflammatory cytokines—
including the interleukins (ILs) 1b, 6, 12
and 18, interferon-g, tumor necrosis factor (TNF) and leukemia
inhibitory factor (LIF)16—and cytokines with anti-inflammatory or
regulatory function such as IL-10 and transforming growth factor-b1
(TGF-b1) (Fig. 2a,b). Chemokine receptors are present on Schwann
cells and satellite glia. Subpopulations of sensory neurons also express
the chemokine receptors CCR1, CCR4, CCR5 and CXCR4 and are
activated by CCL3 (ref. 17), CCL5 and CCL22 (ref. 18). In sensory
neuropathies associated with human immunodeficiency virus (HIV),
HIV may use CCR5andCXCR4 toenter nerve cells and produce direct
Proinflammatory cytokines contribute to axonal damage20, but they
also modulate spontaneous nociceptor activity and stimulus sensitiv-
ity21–23. Activation of TNF receptors in sensory neurons and recruit-
mentofTNFreceptor–associatedfactors (TRAFs),animportant group
a Peripheral nerve injury
b Immune and glial cell reactions
Distal to the nerve lesion site
Dorsal root ganglion
Dorsal root ganglion
Schwann cell proliferation
Regenerating nerve fibers
Figure 1 Immune and glial cell responses to
peripheral nerve injury. (a) Nerve injury provokes
recruitment and activation of immune cells at the
site of a nerve lesion, in the DRG, and in the
ventral and dorsal horns of the spinal cord.
(b) Top, macrophages, T lymphocytes and mast
cells invade the lesion site and spread around the
distal stumps of injured nerve fibers. Schwann
cells begin to proliferate, dedifferentiate and form
bands of Bu ¨ngner, which serve as guiding tubes
for regenerating axons. Middle, macrophages and
a few T lymphocytes reside in the DRG before
injury. Their numbers increase sharply after injury.
Macrophages also move within the sheath that
satellite cells form around the cell bodies of
primary sensory neurons. Satellite cells begin
to proliferate and increase the expression of
glial fibrillary acidic protein. Bottom, one week
after nerve injury, dense clusters of microglial
cells occur in the ventral horn of the spinal
cord, surrounding the cell bodies of motor
neurons. Massive microglial activation is also
found in the dorsal horn, in the projection
territories of the central terminals of injured
primary afferent fibers.
1362 VOLUME 10 [ NUMBER 11 [ NOVEMBER 2007 NATURE NEUROSCIENCE
© 2007 Nature Publishing Group http://www.nature.com/natureneuroscience
of intracellular adaptor proteins, leads to phosphorylation of the
mitogen-activated protein (MAP) kinases p38 (ref. 24) and Jun N-
terminal kinase (JNK), potentially activating the nuclear factor-kB
(NF-kB) and Jun oncogene transcription pathways25. In macrophages,
anti-inflammatory cytokines26. TNF-mediated signaling also promotes
further macrophage invasion by inducing the release of proteases and
upregulation of adhesion molecules. IL-1 regulating the increase in
NGF synthesis and release by Schwann cells27is another example
illustrating the network of signaling pathways among immune cells,
peripheral glia and primary sensory neurons. Communication among
these cells favors axonal growth and survival (adaptation) but also
triggers the development of persistent pain (maladaptation). Distin-
guishing the signals necessary for regeneration from those involved in
the establishment of neuropathic pain will provide important clues for
the development of drugs that reduce maladaptive changes in response
to nerve injury without necessarily impairing the adaptive ones.
Mutant C57BL/6OlaHsd-Wld mice express a chimeric nuclear
protein that interferes with ubiquitination and proteolysis and protects
from wallerian degeneration28. These mice show a delayed immune
response to nerve injury at the lesion site with a decreased and slowed
upregulation of TNF and IL-1b in macrophages and Schwann cells29.
Their pain responsiveness after nerve injury is reduced29, highlighting
the importance of the inflammatory reaction intrinsic to wallerian
degeneration for the early onset of peripheral neuropathic pain.
Macrophage and T lymphocyte invasion of the DRG
Macrophages and a few T lymphocytes are normally present in the
DRG, alongside satellite glial cells. Though remote from the actual
lesion site, these resident immune and glial cells in the DRG react to
nerve injury, and their response is reinforced by invading macrophages
and T cells. In contrast, neutrophil granulocytes are only found in the
DRG in experimental nerve lesions that involve inflammation, such as
the chronic constriction injury model. Injury-induced macrophage
invasion appears to be triggered by a release of chemokine (C-X3-C
motif) ligand 1 (fractalkine)30and the chemokine CCL2 from
DRG neurons31–33. CCL2 may also act directly in a paracrine
fashion on subpopulations of DRG neurons that begin to express
CCR2 after injury32.
The density of macrophages immunoreactive for major histo-
compatibility complex II increases in the DRG 1 week after a nerve
transection and remains elevated for at least 3 months. By that time,
macrophages move from an initially rather diffuse distribution
in the DRG to surround the cell bodies of injured sensory neurons
(Figs. 1b and 3). Two months after a peripheral nerve transection, a
substantial proportion of macrophages in the DRG turn into active
phagocytes, presumably removing debris from injured sensory neu-
rons34, many of which begin to degenerate after axotomy. In the rat, a
decrease, predominantly in small unmyelinated neurons, is detectable
8 weeks after nerve transection35. In the mouse, 24% of DRG neurons
are lost within 7 d of a nerve injury; after 4 weeks, their number is
decreasedby more than50%(ref. 36).Ongoing loss of sensoryneurons
number of macrophages and T lymphocytes in the DRG. The
persistence of the immune response in the DRG certainly
contrasts with the inflammatory reaction distal to the nerve lesion
site, which essentially ends with the removal of myelin debris during
A marked upregulation of genes in the DRG that are related to
immune cell function reflects the extent of both the recruitment and
activity of macrophages and T cells and underscores the substantial
changes that occur in the local environment of primary sensory
neurons after axonal injury37. Increased synthesis and release of
activity and elicit spontaneous action potential discharges. Deletion of
IL-1 receptor antagonist, for example, inhibits the development of
spontaneous (ectopic) sensory neuron firing23, and blocking IL-1- or
TNF has an enhanced direct effect on sensory neurons after nerve
lesions, because both injured and neighboring uninjured nerve fibers
become more sensitive to this cytokine22. Through a signaling pathway
that involves TNF receptor 1 (TNFR1) and p38 MAP kinase, TNF acts
toincreasethedensityoftetrodotoxin (TTX)-resistant sodium channel
currents in nociceptors39(Fig. 3a). In addition, cytokines such as LIF
and IL-6 modulate the synthesis of neuropeptide transmitters. The
CGRP, substance P,
bradykinin, nitric oxide
Macrophages and mast
cells release prostaglandins
and the cytokines IL-1, IL-6,
IL-18, TNF and LIF
factors, prostaglandins and
cytokines including IL-1, IL-6, IL-8,
IL-10, TNF, LIF and TGF-β
Figure 2 Inflammatory changes associated with wallerian degeneration.
(a) Macrophages and Schwann cells produce matrix metalloproteases that
interrupt the blood-nerve barrier. CGRP, substance P, bradykinin and nitric
oxide released from the proximal stumps of injured nerve fibers induce
hyperemia and swelling, promoting the invasion of further monocytes and
T lymphocytes. The chemokines CCL2 and CCL3 attract and guide monocytes
to the lesion site. Macrophages and mast cells release prostaglandins and the
cytokines IL-1b, IL-6, IL-18, TNF and LIF. TNF has an autocrine effect on
macrophages that is mediated through TNFR1 activation and enhances
cytokine synthesis and release. TNF also promotes further macrophage
infiltration. (b) Within minutes of the injury, neuregulin, a growth factor
constitutively expressed on the axonal membrane, binds to a heteromeric
receptor composed of ERBB2 and presumably ERBB3 on Schwann cells.
Early ERBB2 activation is involved in demyelination, whereas late signaling
through ERBB2 and ERBB3 supports Schwann cell proliferation. In the
reverse direction, Schwann cells release the neurotrophic factors NGF and
GDNF, prostaglandins, and cytokines; these sensitize nociceptors and
modulate sensory neuron gene expression.
NATURE NEUROSCIENCE VOLUME 10 [ NUMBER 11 [ NOVEMBER 20071363
© 2007 Nature Publishing Group http://www.nature.com/natureneuroscience
resultant change in the phenotype of sensory neurons is likely to alter
the efficacy of their synaptic input to the spinal cord.
A signaling pathway that modulates both neuronal and macrophage
function in the DRG involves purinergic P2 receptors. Homo- or
heteromers of P2RX subunits form unselective cation channels with a
low affinity (in the micromolar range) for ATP but fast activation. In
contrast, P2Y receptors are slow G protein–coupled receptors with a
high affinity (in the nanomolar range) for ATP that trigger a variety of
second-messenger signaling pathways, resulting in a prolonged signal
duration40,41. Nociceptors predominantly express P2RX2 and P2RX3
subunits42, whereas only few primary sensory neurons express P2Y
receptors43. Cells of myeloid origin, including macrophages and
B lymphocytes, express P2RX7 and several P2Y receptor subunits
such as P2RY1, P2RY2 and P2RY4. T lymphocytes express P2RX1,
P2RX4 and P2RX7 but lack P2Y receptors44. Satellite cells are immu-
noreactive for P2RX2 and P2RX7 (ref. 45) and express P2RY12
and P2RY14 receptor subunits43, indicating that ATP may directly
modulate their activity. Passive release of ATP from severed nerve
fibers and surrounding damaged tissue will activate the purinergic
receptors expressed by sensory neurons and immune cells, but
intact primary sensory neurons themselves also release ATP upon
stimulation. Involvement of P2RX2 and P2RX3 (refs. 46,47), P2RX4
(ref. 48) and P2RX7 (refs. 49,50) in the development of nerve injury–
induced mechanical hypersensitivity has been demonstrated in several
animal models. P2RX (and P2RY) agonists provoke an increase in
nociceptor activity and sensitivity that is mediated in part by the
TTX-resistant voltage-gated sodium channel NaV1.9 (ref. 51) (Fig. 3a).
Blocking P2RX2 and P2RX3 reduces Ad and C fiber activity after
nerve injury52. Mice deficient in P2RX7 have a substantially blunted
hypersensitivity to mechanical and thermal stimuli in models
of neuropathic pain, presumably because of a reduced release of
inflammatory cytokines, including IL-1b, from macrophages and
Several months after nerve injury, macrophages and T cells are also
closely associated with sympathetic nerve fiber terminals that sprout
into basket-like structures around large-diameter sensory neurons.
Sprouting of noradrenergic terminals is reduced in mice lacking IL-6,
indicating that cytokine release by macrophages is one of the signals
that trigger the sympathetic nerve fiber invasion53. The formation of
sympathetic fiber baskets furthermore depends on satellite cell–derived
NGF and neurotrophin-3, suggesting
cation between macrophages and satellite cells (Fig. 3b). In addition,
satellite cells may be the origin of neural crest progenitors
for neurons and Schwann cells that could help repopulate the DRG
Microglial and astrocyte activation in the spinal cord
Microglial cells have a key role in the response to direct injuries of the
diseases such as multiple sclerosis, and in neurodegenerative disorders.
Giventhat microglial cellsshare a myeloidlineage andmany functional
features of peripheral macrophages, it is not too surprising that after a
nerve lesion, microglial cells form dense clusters around the cell bodies
of injured motor neurons in the ventral horn of the spinal cord, similar
to the macrophages that surround injured sensory neurons in the
DRG55. However, it is perhaps far less expected to find a massive
the central terminals of injured sensory nerve fibers56–59(Fig. 1b).
Nerve injury-induced microglial activation is characterized by phos-
Lyn)60–63. Spinal microglial activation in both dorsal and ventral horns
peaks 1 week after injury, followed by a slow decline over severalweeks.
This temporal pattern differs from that of the very early inflammatory
reaction distal to a nerve lesion site and the sustained infiltration of
macrophages and lymphocytes in the DRG, suggesting that the central
immune response to peripheral nerve injury is independently orga-
nized and has distinct functional consequences.
Three signaling pathways mediate the recruitment of resident spinal
microglia and probably also circulating monocytes to the dorsal horn.
These involve the chemokine fractalkine acting on the CX3CR1
receptor64, CCL2 signaling through CCR2 (ref. 65), and Toll-like
Fractalkine is a neuronal transmembrane glycoprotein from which a
soluble chemokine domain can be cleaved by proteolysis; the chemo-
kine is, however, active in both its membrane-bound and soluble form.
As microglial cells64and astrocytes68express CX3CR1, signaling
through fractalkine could help explain the topographic specificity of
microglial recruitment and astrocyte proliferation in the territory of
injured afferent fiber terminals (Fig. 4a). Fractalkine-mediated signal-
ing between neurons and glia seems to contribute to the development
of neuropathic pain. Intrathecal injection of fractalkine produces
mechanical allodynia and thermal hyperalgesia30,69, whereas adminis-
tration of a neutralizing antibody against CX3CR1 delays the develop-
ment of mechanical allodynia after chronic constriction of the sciatic
nerve and spinal nerve ligation30,69.
CCL2 is expressed by primary sensory neurons and Schwann cells
after peripheral nerve injury33and chronic compression of the DRG,
a model of spinal stenosis32. Nerve injury does not induce CCL2
expression in dorsal horn neurons, but CCL2 is transported to the
central terminals of primary afferents in the dorsal horn33. Microglial
NGF, neurotrophin 3
IL-1β, IL-6, LIF
p38 MAP kinase
Figure 3 Immune response in the DRG.
(a) Macrophages potentiate TTX-resistant voltage-
gated sodium channel currents through a signaling
pathway that involves TNF acting on TNFR1 and,
further downstream, phosphorylation of p38 MAP
kinase. Nociceptor activity is further modulated
through activation of the purinergic P2RX2 and
P2RX3 dimer. Activation of P2RX7 expressed by
macrophages regulates the release of the
cytokines IL-1b, IL-6 and LIF. (b) IL-6 is a
macrophage-derived signal that triggers the
sprouting of sympathetic nerve fibers into the
DRG. The formation of noradrenergic fiber baskets
around predominantly large-diameter sensory
neurons is also dependent on NGF and neuro-
trophin-3, which are released from satellite cells.
1364 VOLUME 10 [ NUMBER 11 [ NOVEMBER 2007 NATURE NEUROSCIENCE
© 2007 Nature Publishing Group http://www.nature.com/natureneuroscience
cells express CCR2, the receptor for CCL2 (ref. 70), suggesting a
potential pathway of direct communication between injured primary
afferent fiber terminals and dorsal horn microglia (Fig. 4a); it is,
however, unknown if (and how) CCL2 is released from afferent
terminals. Mice lacking CCR2 show substantially less hypersensitivity
to mechanical stimulation after partial sciatic nerve ligation70, but this
by the absence of CCR2 in microglia. Instead, prevention of peripheral
macrophage and lymphocyte recruitment to the nerve lesion site and
the DRG may partly explain the effect of CCR2 knockout70. Further-
more, because primary sensory neurons start to express CCR2 and are
depolarized by CCL2 after injury32, CCR2 knockout interrupts signal-
ing among sensory neurons, between denervated Schwann cells and
sensory neurons, and potentially also between sensory neurons
Mammalian Toll-like receptors (TLRs) are a family of twelve
evolutionarily conserved membrane proteins that are fundamental in
the initiation of innate immunity against invading pathogens71. TLRs
recognize lipid, carbohydrate, peptide and nucleic acid structures
expressed by different groups of microorganisms. TLR signaling
involves five adaptor proteins that bind to downstream protein kinases
including p38MAP kinase and JNK. Ultimately, TLR signaling leads to
activation of the transcription factor NF-kB, upregulation of inter-
ferons and increased expression of proinflammatory cytokines71.
Recent findings support a link between neuropathic pain and the
innate immune response mediated through TLR activation (Fig. 4a).
In mice lacking TLR2 (ref. 66) or TLR4 (ref. 67), and rats treated with
antisense oligonucleotides to produce TLR4 knockdown67, microglial
activation and the induction of proinflammatory cytokines after a
peripheral nerve lesion are substantially diminished. Moreover, these
animals show less neuropathic pain-like behavior66,67. Although
microglial cells express both TLR2 and TLR4, the endogenous ligands
that activate these TLRs are unclear. TLRs recognize nucleic acids and
proteins released after cell damage—for example, the heat-shock
proteins 60 and 70 (ref. 72)—so that they may be activated by cellular
response to the nerve injury–induced transsynaptic apoptosis of dorsal
horn neurons73. However, the extent of central terminal degeneration
seems insufficient to explain the scale of microglial activation after
Microglial recruitment and activation in the dorsal horn is accom-
panied by an invasion of T lymphocytes55. Nerve injury also leads to
increased proliferation and activation of astrocytes in the ipsilateral
spinal cord. Compared with the microglial response, astrocyte proli-
feration begins relatively late and progresses slowly, but is sustained for
a longer period (more than 5 months)33,63,74. The signal(s) that trigger
and sustain astrocyte proliferation and activation are unknown, and it
is intriguing to speculate that the astroglial response occurs secondary
to the microglial activation (Fig. 4a).
Talk to me, listen to me
Recruitment and activation of different glial cells in complex temporal
patterns requires well organized reciprocal communication between
neurons and glia and among glial cells themselves. A prominent
signaling pathway in the development of neuropathic pain involves
ATP acting on microglial purinergic receptors. Microglial cells express
a potent stimulator of microglia in vitro48. When injected intrathecally,
ATP and ATP-stimulated microglia provoke sustained mechanical
hypersensitivity in the rat48,75. Tonic stimulation of spinal microglia
with ATP causes a P2RX-mediated release of brain-derived neuro-
trophic factor (BDNF), which produces a depolarizing shift in the
anion reversal potential of dorsal horn lamina I neurons75(Fig. 4b).
This shift prompts an inversion of inhibitory GABA currents that
contributes to mechanical allodynia after nerve injury76. On the other
hand, not all GABAergic effects are inverted after peripheral nerve
injury73, and this change in GABA signaling may be limited to a
subpopulation of lamina I neurons projecting to the brain. Peripheral
nerve injury leads to an increase in microglial P2RX4 expression, and
pharmacological blockade of spinal P2RX4 reduces mechanical allo-
dynia after nerve injury without affecting acute pain-responsive beha-
vior in uninjured rats48. However, the source of ATP in the dorsal horn
Microglial cytokines such
as IL-1β, IL-6, IL-10, TNF and
TGF-β may act directly
on the central terminals of
primary afferent neurons and
on dorsal horn neurons
Figure 4 Recruitment and activation of spinal microglia and astrocytes.
(a) Microglial recruitment depends on signaling pathways involving TLR2 and
TLR4, and on the chemokine CCL2 acting on CCR2. The neuronal protein
fractalkine has a chemokine domain that can be cleaved from its membrane-
bound portion. Both bound and soluble fractalkine have chemokine function
and may attract microglia as well as astrocytes by acting on CX3CR1.
Because the microglial response to nerve injury precedes the proliferation of
astrocytes, a direct path of communication may exist between these two glial
cell types to coordinate their sequential temporal patterns of activation.
(b) ATP binding to the purinergic receptor P2RX4 triggers microglial
activation after nerve injury. Active microglia releases BDNF, which induces
in a subpopulation of dorsal horn lamina I neurons an inversion of inhibitory
GABAergic currents. In addition, microglial cytokines are likely to act
directly on the central terminals of primary sensory afferents and on
dorsal horn neurons.
NATURE NEUROSCIENCE VOLUME 10 [ NUMBER 11 [ NOVEMBER 2007 1365
© 2007 Nature Publishing Group http://www.nature.com/natureneuroscience
is unknown. Potentially, ATP is actively released from injured primary
afferents and dorsal horn neurons, or increases as primary afferent
terminals degenerate. Elevated ATP may not only stimulate microglia
but also modulate synaptic transmission among neurons, as presynap-
tic P2X receptors are present on primary afferent terminals and
inhibitory interneurons, and postsynaptic P2X receptors on dorsal
horn neurons41. In slice preparations of the spinal cord, activation of
P2X receptors enhances spontaneous and evoked excitatory post-
synaptic currents and glutamate release in the dorsal horn77.
Microglial activation leads to increased synthesis of the lysosomal
cysteine protease cathepsin S (ref. 78) and the cytokines IL-1b, IL-6,
IL-10, TNF and TGF-b. Direct modulation of dorsal horn neuron
activity by these cytokines may be involved in the development of
neuropathic pain79(Fig. 4b). However, cytokines also provide impor-
tant autocrine feedback signals to microglial cells themselves. In the
brain, IL-1b, IL-6, TNF and interferon-g induce components of the
complement cascade in microglia. We have recently identified several
complement factors and receptors as some of the most prominently
regulatedgenesinthe dorsalhornacross severaldistinct animalmodels
of neuropathic pain80. These include complement-5 (C5) and its
receptor, both of which are upregulated in microglial cells within 3 d
of peripheral nerve injury. C5 is cleaved and activated by C3, a central
complement component that represents the point of convergence for
injection of a synthetic peptide equivalent to active C5 induces cold
hypersensitivity in naive rats, C5 receptor blockade reduces cold-
evoked pain-like behavior in rats after spared nerve injury80, indicating
that complement signaling converging on C5 activation modifies
the painful response to cold. The C5 receptor is exclusively expressed
in microglia; therefore, signaling pathways between microglia and
neurons are required for mediating the effect of its activation on
pain-like behavior. In vitro, C5 induces expression of C5 receptor
on microglial cells, indicating that this complement factor is also
involved in a feedforward mechanism that enhances microglial sensi-
tivity to C5.
Immune and glial modulation: new treatment opportunities
Current therapeutic strategies for neuropathic pain aim to reduce the
excitability of neurons in the peripheral nervous system or the CNS by
modulating the activity of ion channels (gabapentin, pregabalin,
carbamazepine, lidocaine and capsaicin) or by mimicking and enhan-
cing endogenous inhibitory mechanism (tricyclic antidepressants,
duloxetine and opioids). Considering that the involvement of
immune cells and glia in the development of neuropathic pain is
now well established, and given the enormous need for therapeutic
progress, surprisingly few clinical studies have tested immunosuppres-
sive drugs or drugs interfering with glial functions for neuropathic
pain. Epidural and intrathecal corticosteroid injections to prevent or
treat postherpetic neuralgia have been used based on the assumption
that inflammation during the reactivation of herpes zoster virus
contributes to persistent pain in this condition81,82. Epidural corticos-
teroid injections are commonly used to treat sciatica, a condition of
mixed etiology including inflammation at the site of an intervertebral
pain relief but not sustained improvement83.
Preclinical studies have explored several routes of immune and glial
modulation84. Global inhibitors of glial metabolism such as fluoro-
cytokine release and attenuate pain-responsive behavior in several
animal models of neuropathic pain. Fluorocitrate inhibits aconitase,
which leads to blockade of the citric acid cycle. Although fluorocitrate
preferentially acts on glial cells, concerns about its toxicity prevent its
clinicaluse90. Propentofylline reducesproliferationand activityof both
microglia and astrocytes by inhibiting extracellular adenosine trans-
porters and phosphodiesterases, which results in an increase in cyclic
nucleotides, including cyclic AMP (ref. 91). However, trials examining
the treatment effect of propentofylline in Alzheimer’s disease, which
also involves microglial activation, were unsuccessful. Minocycline is a
member of the tetracycline class of broad-spectrum antibiotics that
diffuses into the central nervous system. Apart from its antibiotic
properties, minocycline inhibits matrix metalloproteases, reduces
microglial activity by suppressing the expression of inducible nitric
oxide synthase (iNOS) and the phosphorylation of p38 MAP kinase,
and has neuroprotective function as an inhibitor of neuronal necrosis
and apoptosis92. Minocycline is now under clinical investigation as a
treatment for multiple sclerosis and amyotrophic lateral sclerosis.
Teriflunomide is the active metabolite of leflunomide, an immuno-
suppressive drug approved for the treatment of rheumatoid arthritis.
Leflunomide and teriflunomide block the de novo synthesis of pyrimi-
dines in rapidly dividing cells such as lymphocytes by binding to
dihydroorotate dehydrogenase; leflunomide and teriflunomide also
inhibit antigen presentation. Prolonged treatment with leflunomide
is, however, associated with an increased risk of sensory and
motor neuropathy93, limiting its use for immunomodulation in
More targeted interventions have been aimed at purinoceptors,
cannabinoid receptors, MAP kinases, TNF and interleukins. Recently
developed selective antagonists of P2RX2 and P2RX3 heteromultimers
and inhibitors of P2RX3 (refs. 52,94) and P2RX7 (refs. 50,95) reduce
spontaneous discharges and evoked responses of primary sensory
neurons, decrease cytokine release and attenuate mechanical hypersen-
sitivity after nerve injury. Type 2 cannabinoid receptors (CB2) are
primarily expressed by peripheral immune cells, including macro-
phagesandlymphocytes, andby microglia and astrocytes in the central
nervous system96. CB2-selective agonists reduce pain-like behavior in
animal models of peripheral nerve injury97.
The prominent involvement of p38, JNK and ERK in the activation
of microglia and astrocytes makes MAP kinases promising tar-
gets60,62,63. However, to specifically block spinal microglial activation,
a p38 inhibitor would have to be directed against the b-isoform of this
MAP kinase, sparing the a-isoform that is present in dorsal horn
neurons62. Furthermore, MAP kinases are important in many cellular
processes, such as proliferation anddifferentiation, stress response,and
apoptosis. They are involved in development, learning and memory, so
that systemic application of such inhibitors carries the risk of inter-
fering with these important functions. Nevertheless p38 MAP kinase
may well be useful for neuropathic pain management.
Etanercept, a soluble TNF receptor fusion protein, and anakinra, a
recombinant form of human IL-1receptorantagonist, have beentested
in animal models of peripheral nerve injury and reduce neuropathic
pain-like behavior24,79. Induced expression of anti-inflammatory IL-10
in the DRG and spinal meninges provides a prolonged analgesic effect
in rats after chronic constriction of the sciatic nerve98. Thalidomide,
which primarily inhibits TNF synthesis but also modulates the expres-
sion of other cytokines, including IL-10 (ref. 99), decreases mechanical
and thermal hypersensitivity in rats after nerve injury when the
treatment starts at the time of the injury; thalidomide treatment has
no analgesic effect once neuropathic pain hypersensitivity is estab-
lished100. Likewise, minocycline does not reverse existing hyper-
sensitivity after nerve injury88, perhaps indicating that peripheral
immune cells and microglia have an important but transient role in
1366VOLUME 10 [ NUMBER 11 [ NOVEMBER 2007 NATURE NEUROSCIENCE
© 2007 Nature Publishing Group http://www.nature.com/natureneuroscience
the development of neuropathic pain. In contrast, propentofylline,
which inhibits microglial and astrocyte activation, attenuates pain-
responsive behavior when administered early after nerve injury and
also decreases established hypersensitivity86.
The contribution of immune cells and glia to the development and the
persistence of pain after nerve injury challenges conventional concepts
that are biased toward neurons being responsible for the pathophysio-
logical changes that drive neuropathic pain. Yet this shift in our
understanding provides an exceptional opportunity to progress to a
new disease-modifying therapeutic approach. Therapeutic interven-
tions must, however, take into account the different temporal patterns
of immune and glial cell activation at the nerve lesion site, in the DRG,
linked, and identify the pathways of such a connection. Finally, just as
the participation of peripheral immune cells in wallerian degeneration
facilitates the regeneration and healing of injured axons after a nerve
lesion, the highly coordinated spinal glial response to nerve injury may
not be exclusively maladaptive. Differentiating between the good, the
bad and the ugly aspects of immune and glial responses to nerve injury
will be essential for developing targeted new treatment strategies for
Supported by grants from the US National Institute of Neurological Disorders
and Stroke (J.S., C.J.W.) and the US National Institute of Dental and Craniofacial
Published online at http://www.nature.com/natureneuroscience
Reprints and permissions information is available online at http://npg.nature.com/
1.Dworkin, R.H. et al. Advances in neuropathic pain: diagnosis, mechanisms, and
treatment recommendations. Arch. Neurol. 60, 1524–1534 (2003).
Streit, W.J. Microglia as neuroprotective, immunocompetent cells of the CNS. Glia 40,
Chan, W.Y., Kohsaka, S. & Rezaie, P. The origin and cell lineage of microglia: new
concepts. Brain Res. Rev. 53, 344–354 (2007).
Hanani, M. Satellite glial cells in sensory ganglia: from form to function. Brain Res.
Brain Res. Rev. 48, 457–476 (2005).
Perkins, N.M. & Tracey, D.J. Hyperalgesia due to nerve injury: role of neutrophils.
Neuroscience 101, 745–757 (2000).
Mueller, M. et al. Rapid response of identified resident endoneurial macrophages to
nerve injury. Am. J. Pathol. 159, 2187–2197 (2001).
Perrin, F.E., Lacroix, S., Aviles-Trigueros, M. & David, S. Involvement of monocyte
chemoattractant protein-1,macrophage inflammatory protein-1a and interleukin-1b in
Wallerian degeneration. Brain 128, 854–866 (2005).
Shubayev, V.I. et al. TNFa-induced MMP-9 promotes macrophage recruitment into
injured peripheral nerve. Mol. Cell. Neurosci. 31, 407–415 (2006).
Zochodne, D.W. et al. Evidence for nitric oxide and nitric oxide synthase activity in
proximal stumps of transected peripheral nerves. Neuroscience 91, 1515–1527
10. Stoll, G., Jander, S. & Myers, R.R. Degeneration and regeneration of the peripheral
nervous system: from Augustus Waller’s observations to neuroinflammation.
J. Peripher. Nerv. Syst. 7, 13–27 (2002).
11. Guertin,A.D.,Zhang, D.P., Mak,K.S., Alberta, J.A. & Kim, H.A.Microanatomy of axon/
glial signaling during Wallerian degeneration. J. Neurosci. 25, 3478–3487 (2005).
12. Carroll, S.L., Miller, M.L., Frohnert, P.W., Kim, S.S. & Corbett, J.A. Expression of
neuregulins and their putative receptors, ErbB2 and ErbB3, is induced during
Wallerian degeneration. J. Neurosci. 17, 1642–1659 (1997).
13. Esper, R.M. & Loeb, J.A. Rapid axoglial signaling mediated by neuregulin and
neurotrophic factors. J. Neurosci. 24, 6218–6227 (2004).
14. Malin, S.A. et al. Glial cell line-derived neurotrophic factor family members sensitize
nociceptors in vitro and produce thermal hyperalgesia in vivo. J. Neurosci. 26,
15. Ma, W. & Eisenach, J.C. Cyclooxygenase 2 in infiltrating inflammatory cells in injured
nerve is universally up-regulated following various types of peripheral nerve injury.
Neuroscience 121, 691–704 (2003).
16. Tofaris, G.K., Patterson, P.H., Jessen, K.R. & Mirsky, R. Denervated Schwann cells
attract macrophages by secretion of leukemia inhibitory factor (LIF) and monocyte
chemoattractant protein-1 in a process regulatedby interleukin-6 and LIF. J. Neurosci.
22, 6696–6703 (2002).
17. Zhang, N. et al. A proinflammatory chemokine, CCL3, sensitizes the heat- and
capsaicin-gated ion channel TRPV1. Proc. Natl. Acad. Sci. USA 102, 4536–4541
18. Oh, S.B. et al. Chemokines and glycoprotein120 produce pain hypersensitivity by
directly exciting primary nociceptive neurons. J. Neurosci. 21, 5027–5035
19. Melli, G., Keswani, S.C., Fischer, A., Chen, W. & Hoke, A. Spatially distinct and
functionally independent mechanisms of axonal degeneration in a model of HIV-
associated sensory neuropathy. Brain 129, 1330–1338 (2006).
20. Keswani, S.C. et al. Schwann cell chemokine receptors mediate HIV-1 gp120 toxicity
to sensory neurons. Ann. Neurol. 54, 287–296 (2003).
21. Cunha, T.M. et al. A cascade of cytokines mediates mechanical inflammatory hyper-
nociception in mice. Proc. Natl. Acad. Sci. USA 102, 1755–1760 (2005).
22. Schafers, M., Lee, D.H., Brors, D., Yaksh, T.L. & Sorkin, L.S. Increased sensitivity of
injured and adjacent uninjured rat primary sensory neurons to exogenous tumor
necrosis factor-alpha after spinal nerve ligation. J. Neurosci. 23, 3028–3038 (2003).
signaling attenuates neuropathic pain, autotomy, and spontaneous ectopic neuronal
activity, following nerve injury in mice. Pain 120, 315–324 (2006).
24. Scha ¨fers, M., Svensson, C.I., Sommer, C. & Sorkin, L.S. Tumor necrosis factor-a
induces mechanical allodynia after spinal nerve ligation by activation of p38 MAPK in
primary sensory neurons. J. Neurosci. 23, 2517–2521 (2003).
25. Aggarwal, B.B. Signalling pathways of the TNF superfamily: a double-edged sword.
Nat. Rev. Immunol. 3, 745–756 (2003).
26. Myers, R.R., Campana, W.M. & Shubayev, V.I. The role of neuroinflammation in
neuropathic pain: mechanisms and therapeutic targets. Drug Discov. Today 11,
of nerve growth factor in non-neuronal cells of rat sciatic nerve. Nature 330, 658–659
28. Mack, T.G. et al. Wallerian degeneration of injured axons and synapses is delayed by a
Ube4b/Nmnat chimeric gene. Nat. Neurosci. 4, 1199–1206 (2001).
29. Sommer, C. & Schafers, M. Painful mononeuropathy in C57BL/Wld mice with delayed
wallerian degeneration: differential effects of cytokine production and nerve regenera-
tion on thermal and mechanical hypersensitivity. Brain Res. 784, 154–162 (1998).
30. Zhuang, Z.Y. et al. Role of the CX3CR1/p38 MAPK pathway in spinal microglia for the
development of neuropathic pain following nerve injury-induced cleavage of fractal-
kine. Brain Behav. Immun. 21, 642–651 (2007).
31. Morin,N.etal. Neutrophilsinvadelumbar dorsalrootgangliaafter chronic constriction
injury of the sciatic nerve. J. Neuroimmunol. 184, 164–171 (2007).
32. White, F.A. et al. Excitatory monocyte chemoattractant protein-1 signaling is up-
regulated in sensory neurons after chronic compression of the dorsal root ganglion.
Proc. Natl. Acad. Sci. USA 102, 14092–14097 (2005).
33. Zhang, J. & De Koninck, Y. Spatial and temporal relationship between monocyte
chemoattractant protein-1 expression and spinal glial activation following peripheral
nerve injury. J. Neurochem. 97, 772–783 (2006).
34. Hu, P. & McLachlan, E.M. Distinct functional types of macrophage in dorsal root
ganglia and spinal nerves proximal to sciatic and spinal nerve transections in the rat.
Exp. Neurol. 184, 590–605 (2003).
35. Tandrup, T., Woolf, C.J. & Coggeshall, R.E. Delayed loss of small dorsal root ganglion
cells aftertransection of the ratsciatic nerve.J.Comp.Neurol.422, 172–180(2000).
36. Shi, T.J. et al. Effect of peripheral nerve injury on dorsal root ganglion neurons in the
C57 BL/6Jmouse:markedchanges both incellnumbers and neuropeptideexpression.
Neuroscience 105, 249–263 (2001).
37. Costigan, M. et al. Replicate high-density rat genome oligonucleotide microarrays
reveal hundreds of regulated genes in the dorsal root ganglion after peripheral nerve
injury. BMC Neurosci. [online] 3, 16 (2002).
38. Arruda,J.L.,Sweitzer,S.,Rutkowski,M.D.& DeLeo,J.A.Intrathecalanti-IL-6antibody
and IgG attenuates peripheral nerve injury-induced mechanical allodynia in the rat:
possible immune modulation in neuropathic pain. Brain Res. 879, 216–225 (2000).
39. Jin, X. & Gereau, R.W. Acute p38-mediated modulation of tetrodotoxin-resistant
sodium channels in mouse sensory neurons by tumor necrosis factor-a. J. Neurosci.
26, 246–255 (2006).
40. Burnstock, G. Physiology and pathophysiology of purinergic neurotransmission.
Physiol. Rev. 87, 659–797 (2007).
41. Khakh, B.S. Molecular physiology of P2X receptors and ATP signalling at synapses.
Nat. Rev. Neurosci. 2, 165–174 (2001).
42. Kobayashi, K. et al. Differential expression patterns of mRNAs for P2X receptor
subunits in neurochemically characterized dorsal root ganglion neurons in the rat.
J. Comp. Neurol. 481, 377–390 (2005).
43. Kobayashi, K. et al. Neurons and glial cells differentially express P2Y receptor mRNAs
in the rat dorsal root ganglion and spinal cord. J. Comp. Neurol. 498, 443–454
44. Di Virgilio, F. et al. Nucleotide receptors: an emerging family of regulatory molecules in
blood cells. Blood 97, 587–600 (2001).
45. Zhang, X., Chen, Y., Wang, C. & Huang, L.Y. Neuronal somatic ATP release triggers
neuron-satellite glial cell communication in dorsal root ganglia. Proc. Natl. Acad. Sci.
USA 104, 9864–9869 (2007).
46. Chen, Y., Li, G.W., Wang, C., Gu, Y. & Huang, L.Y. Mechanisms underlying enhanced
P2X receptor-mediated responses in the neuropathic pain state. Pain 119, 38–48
NATURE NEUROSCIENCE VOLUME 10 [ NUMBER 11 [ NOVEMBER 2007 1367
© 2007 Nature Publishing Group http://www.nature.com/natureneuroscience
47. Jarvis, M.F. et al. A-317491, a novel potent and selective non-nucleotide antagonist of Download full-text
rat. Proc. Natl. Acad. Sci. USA 99, 17179–17184 (2002).
48. Tsuda, M. et al. P2X4 receptors induced in spinal microglia gate tactile allodynia after
nerve injury. Nature 424, 778–783 (2003).
49. Chessell, I.P. et al. Disruption of the P2X7 purinoceptor gene abolishes chronic
inflammatory and neuropathic pain. Pain 114, 386–396 (2005).
50. McGaraughty, S. et al. P2X7-related modulation of pathological nociception in rats.
Neuroscience 146, 1817–1828 (2007).
51. Amaya, F. et al. The voltage-gated sodium channel Nav1.9 is an effector of peripheral
inflammatory pain hypersensitivity. J. Neurosci. 26, 12852–12860 (2006).
52. Sharp, C.J. et al. Investigation into the role of P2X3/P2X2/3receptors in neuropathic
pain followingchronic constriction injury in the rat: anelectrophysiological study.Br. J.
Pharmacol. 148, 845–852 (2006).
53. Ramer, M.S., Murphy, P.G., Richardson, P.M. & Bisby, M.A. Spinal nerve lesion-
induced mechanoallodynia and adrenergic sprouting in sensory ganglia are attenuated
in interleukin-6 knockout mice. Pain 78, 115–121 (1998).
from adult dorsal root ganglia. Stem Cells 25, 2053–2065 (2007).
55. Hu, P., Bembrick, A.L., Keay, K.A. & McLachlan, E.M. Immune cell involvement in
dorsal root ganglia and spinal cord after chronic constriction or transection of the rat
sciatic nerve. Brain Behav. Immun. 21, 599–616 (2007).
response to peripheral nerve injury. Brain Behav. Immun. 21, 624–633 (2007).
57. Marchand, F., Perretti, M. & McMahon, S.B. Role of the immune system in chronic
pain. Nat. Rev. Neurosci. 6, 521–532 (2005).
58. Tsuda, M., Inoue, K. & Salter, M.W. Neuropathic pain and spinal microglia: a big
problem from molecules in ‘‘small’’ glia. Trends Neurosci. 28, 101–107 (2005).
59. Watkins, L.R. & Maier, S.F. Beyond neurons: evidence that immune and glial cells
contribute to pathological pain states. Physiol. Rev. 82, 981–1011 (2002).
60. Jin, S.X., Zhuang, Z.Y., Woolf, C.J. & Ji, R.R. p38 mitogen-activated protein kinase is
activated after a spinal nerve ligation in spinal cord microglia and dorsal root ganglion
neurons and contributes to the generation of neuropathic pain. J. Neurosci. 23,
61. Katsura, H. et al. Activation of Src-family kinases in spinal microglia contributes
to mechanical hypersensitivity after nerve injury. J. Neurosci. 26, 8680–8690
62. Svensson, C.I. et al. Spinal p38b isoform mediates tissue injury-induced hyperalgesia
and spinal sensitization. J. Neurochem. 92, 1508–1520 (2005).
63. Zhuang,Z.Y.,Gerner,P.,Woolf,C.J.& Ji,R.R.ERKissequentiallyactivatedinneurons,
microglia, and astrocytes by spinal nerve ligation and contributes to mechanical
allodynia in this neuropathic pain model. Pain 114, 149–159 (2005).
64. Verge, G.M.etal.Fractalkine(CX3CL1) and fractalkine receptor (CX3CR1) distribution
in spinal cord and dorsal root ganglia under basal and neuropathic pain conditions.
Eur. J. Neurosci. 20, 1150–1160 (2004).
65. White, F.A., Bhangoo, S.K. & Miller, R.J. Chemokines: integrators of pain and
inflammation. Nat. Rev. Drug Discov. 4, 834–844 (2005).
66. Kim, D. et al. A critical role of toll-like receptor 2 in nerve injury-induced spinal cord
glial cell activation and pain hypersensitivity. J. Biol. Chem. 282, 14975–14983
67. Tanga, F.Y., Nutile-McMenemy, N. & DeLeo, J.A. The CNS role of Toll-like receptor 4 in
innate neuroimmunity and painful neuropathy. Proc. Natl. Acad. Sci. USA 102,
68. Dorf, M.E., Berman, M.A., Tanabe, S., Heesen, M. & Luo, Y. Astrocytes express
functional chemokine receptors. J. Neuroimmunol. 111, 109–121 (2000).
69. Milligan, E.D. et al. Evidence that exogenous and endogenous fractalkine can induce
spinal nociceptive facilitation in rats. Eur. J. Neurosci. 20, 2294–2302 (2004).
70. Abbadie, C. et al. Impaired neuropathic pain responses in mice lacking the chemokine
receptor CCR2. Proc. Natl. Acad. Sci. USA 100, 7947–7952 (2003).
71. Trinchieri, G. & Sher, A. Cooperation of Toll-like receptor signals in innate immune
defence. Nat. Rev. Immunol. 7, 179–190 (2007).
72. Marshak-Rothstein, A. Toll-like receptors in systemic autoimmune disease. Nat. Rev.
Immunol. 6, 823–835 (2006).
73. Scholz, J. et al. Blocking caspase activity prevents transsynaptic neuronal apoptosis
and the loss of inhibition in lamina II of the dorsal horn after peripheral nerve injury.
J. Neurosci. 25, 7317–7323 (2005).
74. Echeverry, S., Shi, X.Q. & Zhang, J. Characterization of cell proliferation in rat spinal
cord following peripheral nerve injury and the relationship with neuropathic pain. Pain,
published online 8, June 2007 (doi:10.1016/j.pain.2007.05.002).
75. Coull, J.A. et al. BDNF from microglia causes the shift in neuronal anion gradient
underlying neuropathic pain. Nature 438, 1017–1021 (2005).
76. Coull, J.A. et al. Trans-synaptic shift in anion gradient in spinal lamina I neurons as a
mechanism of neuropathic pain. Nature 424, 938–942 (2003).
77. Nakatsuka, T. & Gu, J.G. ATP P2X receptor-mediated enhancement of glutamate
78. Clark, A.K. et al. Inhibition of spinal microglial cathepsin S for the reversal of
neuropathic pain. Proc. Natl. Acad. Sci. USA 104, 10655–10660 (2007).
79. Winkelstein, B.A., Rutkowski, M.D., Sweitzer, S.M., Pahl, J.L. & DeLeo, J.A. Nerve
injury proximal or distalto the DRG inducessimilar spinalglial activation and selective
cytokine expression but differential behavioral responses to pharmacologic treatment.
J. Comp. Neurol. 439, 127–139 (2001).
80. Griffin, R.S. et al. Complement induction in spinal cord microglia results in anaphy-
latoxin C5a-mediated pain hypersensitivity. J. Neurosci. 27, 8699–8708 (2007).
81. Kotani, N. et al. Intrathecal methylprednisolone for intractable postherpetic neuralgia.
N. Engl. J. Med. 343, 1514–1519 (2000).
82. van Wijck, A.J. et al. The PINE study of epidural steroids and local anaesthetics to
prevent postherpetic neuralgia: a randomised controlled trial. Lancet 367, 219–224
83. Arden, N.K. et al. A multicentre randomized controlled trial of epidural corticosteroid
injections for sciatica: the WEST study. Rheumatology (Oxford) 44, 1399–1406
84. Watkins,L.R.& Maier,S.F.Glia:anoveldrugdiscoverytargetforclinicalpain.Nat.Rev.
Drug Discov. 2, 973–985 (2003).
85. Clark, A.K., Gentry, C., Bradbury, E.J., McMahon, S.B. & Malcangio, M. Role of spinal
microglia in rat models of peripheral nerve injury and inflammation. Eur. J. Pain 11,
86. Tawfik, V.L., Nutile-McMenemy, N., LaCroix-Fralish, M.L. & DeLeo, J.A. Efficacy of
propentofylline, a glial modulating agent, on existing mechanical allodynia following
peripheral nerve injury. Brain Behav. Immun. 21, 238–246 (2007).
87. Ledeboer, A. et al. Minocycline attenuates mechanical allodynia and proinflammatory
cytokine expression in rat models of pain facilitation. Pain 115, 71–83 (2005).
88. Raghavendra, V., Tanga, F. & DeLeo, J.A. Inhibition of microglial activation attenuates
the development but not existing hypersensitivity in a rat model of neuropathy.
J. Pharmacol. Exp. Ther. 306, 624–630 (2003).
89. Sweitzer, S.M. & DeLeo, J.A. The active metabolite of leflunomide, an immunosup-
pressive agent, reduces mechanical sensitivity in a rat mononeuropathy model. J. Pain
3, 360–368 (2002).
90. Goncharov, N.V., Jenkins, R.O. & Radilov, A.S. Toxicology of fluoroacetate: a review,
with possible directions for therapy research. J. Appl. Toxicol. 26, 148–161
91. Si, Q., Nakamura, Y., Ogata, T., Kataoka, K. & Schubert, P. Differential regulation of
microglial activation by propentofylline via cAMP signaling. Brain Res. 812, 97–104
92. Zemke, D. & Majid, A. The potential of minocycline for neuroprotection in human
neurologic disease. Clin. Neuropharmacol. 27, 293–298 (2004).
93. Umapathi, T. & Chaudhry, V. Toxic neuropathy. Curr. Opin. Neurol. 18, 574–580
94. McGaraughty, S. & Jarvis, M.F. Antinociceptive properties of a non-nucleotide P2X3/
P2X2/3 receptor antagonist. Drug News Perspect. 18, 501–507 (2005).
95. Donnelly-Roberts, D.L. & Jarvis, M.F. Discovery of P2X7receptor-selective antagonists
offers new insights into P2X7receptor function and indicates a role in chronic pain
states. Br. J. Pharmacol. 151, 571–579 (2007).
96. Stella, N. Cannabinoid signaling in glial cells. Glia 48, 267–277 (2004).
97. Valenzano, K.J. et al. Pharmacological and pharmacokinetic characterization of the
cannabinoid receptor 2 agonist, GW405833, utilizing rodent models of acute and
chronic pain, anxiety, ataxia and catalepsy. Neuropharmacology 48, 658–672
98. Milligan, E.D. et al. Repeated intrathecal injections of plasmid DNA encoding
interleukin-10 produce prolonged reversal of neuropathic pain. Pain 126, 294–308
99. George, A., Marziniak, M., Schafers, M., Toyka, K.V. & Sommer, C. Thalidomide
treatment in chronic constrictive neuropathy decreases endoneurial tumor necrosis
factor-a, increases interleukin-10 and has long-term effects on spinal cord dorsal horn
met-enkephalin. Pain 88, 267–275 (2000).
100. Sommer, C., Marziniak, M. & Myers, R.R. The effect of thalidomide treatment on
vascular pathology and hyperalgesiacausedby chronic constriction injury of rat nerve.
Pain 74, 83–91 (1998).
1368VOLUME 10 [ NUMBER 11 [ NOVEMBER 2007 NATURE NEUROSCIENCE
© 2007 Nature Publishing Group http://www.nature.com/natureneuroscience