Content uploaded by Jie Su
Author content
All content in this area was uploaded by Jie Su on Jul 31, 2018
Content may be subject to copyright.
Phenotypic Changes in Dorsal Root Ganglion and
Spinal Cord in the Collagen Antibody-Induced
Arthritis Mouse Model
JIE SU,
1
TIANLE GAO,
1
TIEJUN SHI,
2
QIONG XIANG,
2
XIAOJUN XU,
1
ZSUZSANNA WIESENFELD-HALLIN,
1
TOMAS H€
OKFELT,
2
and CAMILLA I. SVENSSON
1
*
1
Department of Physiology and Pharmacology, Karolinska Institutet, Stockholm, 171 77, Sweden
2
Department of Neuroscience, Karolinska Institutet, Stockholm, 171 77, Sweden
The mechanisms underlying rheumatoid arthritis (RA)-
induced pain are still not fully elucidated, and accumu-
lating data indicate that peripheral inflammation is not
the only factor driving pain in these patients. The focus
of our work is to investigate the molecular basis for
long-term alterations in nociceptive pathways induced
by polyarthritis using the collagen antibody-induced
arthritis (CAIA) mouse model. In this model, mechanical
hypersensitivity outlasts the joint inflammation by
weeks. Here we examined expression levels of neuro-
peptides, ion channels, and nerve injury markers asso-
ciated with neuropathic and/or inflammatory pain in
dorsal root ganglia (DRGs) and spinal cord both during
the peak of inflammation (day 15) and when the inflam-
mation has resolved but the hypersensitivity persists
(days 45–47). No apparent differences were observed
in substance P, calcitonin gene-related peptide, or neu-
ropeptide Y protein expression in DRGs and spinal cord
of CAIA mice. However, the neuropeptide galanin, the
ATP-gated ion channel P2X3, and calcium channel subu-
nit a2d1 were significantly increased in the CAIA DRGs
as compared to controls, both 15 and 47 days after
induction of arthritis. On day 15 there was an increase
in expression of two factors associated with nerve
injury and cell stress, activating transcription factor 3
and growth-associated protein 43 in DRGs, whereby the
latter was still dramatically upregulated after 47 days.
In conclusion, this study suggests that long-term joint
inflammation has an impact on DRG neurons that
resembles both inflammation and nerve injury-induced
pain states. Thus, antibody-driven inflammation gener-
ates a pain state with a unique neurochemical profile. J.
Comp. Neurol. 523:1505–1528, 2015.
V
C2015 Wiley Periodicals, Inc.
INDEXING TERMS: rheumatoid arthritis; pain; neuropeptides; ion channels; nerve injury; dorsal root ganglia
Rheumatoid arthritis (RA) is a chronic autoimmune
disorder that affects 1% of the adult population world-
wide (McInnes and Schett, 2011). Chronic or episodic
pain is a frequent problem for this group of patients,
and it contributes to the loss of joint mobility and
causes both psychological distress and impaired quality
of life (Andersson et al., 2013). Despite significant
advances in RA research over the past two decades,
the mechanisms that maintain joint pain in this disease
are not fully elucidated (Walsh and McWilliams, 2012).
We have recently characterized pain-like behavior in
two antibody-dependent mouse models of RA: the
K/BxN serum transfer arthritis (K/BxN) model (Christi-
anson et al., 2010) and the CAIA model (Bas et al.,
2012). Notably, both models have similar characteris-
tics, with mechanical hypersensitivity that lasts for sev-
eral months despite the transient joint inflammation,
which is resolved within 3–4 weeks. This finding may
have bearings on the clinical situation, as the degree of
pain that the arthritis patients are experiencing is often
not directly correlated with the degree of joint pathol-
ogy (Walsh and McWilliams, 2012), and there are sub-
groups of RA patients in remission who continue to
report joint pain as a problem (Lee et al., 2011; Taylor
Grant sponsor: Swedish Research Council; Grant numbers: CIS
2013-8373 and TH 2012-2273; Grant sponsor: Swedish Foundation
for Strategic Research; Grant number: CIS RBap08-0062; Grant spon-
sor: Ragnar S€
oderberg Foundation (CIS); Grant sponsor: funds from
Karolinska Institutet (to J.S. and T.H.); Grant sponsor: Knut & Alice
Wallenberg Foundation; Grant number: KAW2008.0149.
*CORRESPONDENCE TO: Camilla I. Svensson, PhD, Department of Phys-
iology and Pharmacology, Karolinska Institutet, Von Eulers vag 8, 171 77
Stockholm, Sweden. E-mail: camilla.svensson@ki.se
Received August 1, 2014; Revised December 26, 2014;
Accepted January 24, 2015.
DOI 10.1002/cne.23749
Published online March 26, 2015 in Wiley Online Library
(wileyonlinelibrary.com)
V
C2015 Wiley Periodicals, Inc.
The Journal of Comparative Neurology |Research in Systems Neuroscience 523:1505–1528 (2015) 1505
RESEARCH ARTICLE
et al., 2011). Thus, it is possible that RA pain may be
mediated by diverse mechanisms and that, in addition
to inflammation, other factors, such as structural dam-
age, nerve injury, and/or alterations in central pain
processing may contribute to the persistence of RA
pain.
Following experimental peripheral inflammation and
nerve injury in rodents, both transient and long-lasting
changes in the expression of neuropeptides and ion
channels have been described. For example, after nerve
transection and ligation injuries, expression levels of
neuropeptides like galanin and neuropeptide Y (NPY)
are elevated, and excitatory peptides such as sub-
stance P (SP) and calcitonin gene-related peptide
(CGRP) are decreased in DRG neurons (Hokfelt et al.,
1984; Wakisaka et al., 1992; Nothias et al., 1993; Sol-
way et al., 2011; Zhang et al., 2011). In contrast, stud-
ies in which peripheral inflammation is induced by
subcutaneous injection of, e.g., complete Freund’s adju-
vant (CFA) or carrageenan have shown an increase of
SP and CGRP levels in DRG neurons (Smith et al.,
1991; Donnerer et al., 1993; Hanesch et al., 1993;
Galeazza et al., 1995; Calza et al., 1998), while there is
no evidence for an increase of NPY in DRGs during tis-
sue inflammation (Ji et al., 1994). Also, an increase of
galanin expression in DRGs is predominantly associated
with nerve injury (Hokfelt et al., 1987; Coronel et al.,
2008), although it has been found elevated after immu-
nization with CFA (Calza et al., 1998; Calza et al.,
2000). Further, while voltage-gated sodium channels
(e.g., Nav1.3, Nav1.7, Nav1.8) are upregulated in pri-
mary sensory neurons both after tissue inflammation
and nerve injury (Fukuoka et al., 2008, 2012; Xiang
et al., 2008), upregulation of the voltage-gated calcium
channel subunit a2d1, has been associated primarily
with nerve injury (Luo et al., 2001; Newton et al.,
2001).
Following nerve injury, increased expression levels of
growth-associated protein 43 (GAP43) (Jacobson et al.,
1986) and activating transcription factor 3 (ATF3) (Tsu-
jino et al., 2000) in DRGs are frequently reported and
therefore these factors are often referred to as markers
of nerve injury. Importantly, in our earlier work we
noted an increased expression of ATF3 in DRG neurons
in the late phase of the K/BxN model (Christianson
et al., 2010), and Inglis et al. (2007) found a modest
increased ATF3 expression in DRGs subsequent to
immunization against collagen type II. These findings
led to the hypothesis that the mechanisms driving
arthritis-induced hypersensitivity changes over time,
and may have an impact on sensory neurons partly
resembling neuropathic pain. Thus, the aim of the cur-
rent study was to examine the expression levels of neu-
ropeptides and ion channels associated with
inflammatory and/or neuropathic hypersensitivity, as
well as nerve-injury markers, in the peripheral and cen-
tral nervous system during peak inflammation and in
the late phase of the CAIA model. We found that the
neurochemical profile of antibody-induced hypersensitiv-
ity had characteristics of both inflammation and nerve
injury-induced hypersensitivity, indicating that joint
inflammation induces a unique pain phenotype.
MATERIALS AND METHODS
Animal model
Adult male CBA mice (8 weeks old, Nova-SCB, Stock-
holm, Sweden) were kept under standard conditions on
a 12-hour light/dark cycle. Food and water were avail-
able ad libitum. All experiments were carried out
according to protocols approved by the local ethical
committee for animal experiments in Sweden (Stock-
holms Norra Djurf€
ors€
oksetiska N€
amnd, N556/11).
Arthritis was induced by injection of 1.5 mg anti-CII
arthritogenic monoclonal antibody cocktail (Chondrex,
Redmond, WA) i.v. on day 0, followed by 30 lgof
lipopolysaccharide (LPS, Chondrex) i.p. on day 5. Both
control groups received 100 ll of saline i.v. on day 0.
On day 5, the saline control group received saline i.p.
while the LPS control group received LPS i.p.
Behavioral test and arthritis scoring
Mechanical hypersensitivity was assessed using von
Frey filaments made of optic glass fibers (Optihair,
Schriesheim, Germany) with force values ranging from
0.03–3.30 g and the up–down method (Chaplan et al.,
1994), applied as previously described (Bas et al., 2012).
The 50% probability withdrawal threshold (force of the
von Frey filament to which an animal reacts to 50% of
the presentations) was calculated. Given the systemic
nature of this model, mechanical sensitivity values for
both hind paws were measured and averaged.
In order to monitor the degree of arthritis develop-
ment, the four paws were visually inspected and scored.
Each inflamed (both swollen and red) toe was noted as
one point; if the dorsal side of the paw or ankle joint
was inflamed, 2.5 points were given for moderate inflam-
mation and 5 points for severe inflammation, resulting in
a score between 0 and 15 for each limb and a maximum
of 60 points per mouse (Bas et al., 2012).
Tissue preparation
At specified timepoints (15–16 and 45–47 days after
CII antibody cocktail injection), mice were deeply anes-
thetized with sodium pentobarbital (Mebumal; 50 mg/kg
body weight, i.p.) and perfused through the heart with
20 ml of saline (0.9% NaCl), followed by 20 ml of picric
J. Su et al.
1506 The Journal of Comparative Neurology |Research in Systems Neuroscience
acid-paraformaldehyde (PFA) fixative solution (4% PFA
with 0.2% picric acid in 0.16 M phosphate buffer [PB, pH
7.2–7.4]), both at 37C, and then 50 ml of the same fix-
ative at 4C. The L4-L6 DRGs and the corresponding
lumbar spinal cord segments were dissected and post-
fixed in the same fixative for 90 minutes at 4C (Brumov-
sky et al., 2005). The tissue was cryoprotected with 10%
sucrose in 0.1M PB containing 0.01% sodium azide
(Sigma, St. Louis, MO) and 0.02% bacitracin (Sigma) for
48 hours. The tissue was then embedded with OCT com-
pound (Tissue Tek, Miles Laboratories, Elkhart, IN) and
cut in a cryostat (Microm, Heidelberg, Germany) at
12 lm thickness for DRGs and 20 lm for spinal cord.
Immunohistochemistry
Tissue sections were incubated with primary antibod-
ies against galanin, CGRP, NPY, SP, P2X3, CACNA2D1
(a2d1), ATF3, GAP43, and IB4 (Table 1) at 4C over-
night. Immunoreactivity (IR) was visualized with TSA Plus
kit (Perkin Elmer, Boston, MA) as previously described
(Brumovsky et al., 2005). Briefly, the slides were rinsed
in TNT buffer (0.1M Tris-HCl, pH 7.4; 0.15M NaCl; 0.05%
Tween 20) for 15 minutes, blocked with TNB buffer
(0.5% blocking reagent [Perkin Elmer] in TNT buffer) for
30 minutes followed by 30 minutes incubation with
horseradish peroxidase-labeled secondary antibody
(1:200, Dako, Glostrup, Denmark) diluted in TNB buffer.
After a 15-minute wash in TNT buffer, all sections were
exposed to fluorescein diluted in amplification solution
(1:100) for 15 minutes and washed in TNT buffer for 20
minutes. The sections were counterstained with 0.001%
propidium iodide (PI, Sigma) and mounted with 2.5%
DABCO (Sigma). In the case of double labeling, the
slides were rinsed in PBS for 20 minutes after the TSA
plus immunohistochemistry (IHC) and incubated with
antibody against CGRP at 4C overnight. The slides were
incubated with Rhodamine Red-X (RRX)-conjugated don-
key antirabbit antibody at room temperature (RT) for 2
hours (1:100, Jackson Laboratories, West Grove, PA) and
mounted with DABCO after a 30-minute wash in PBS.
Another set of slides were incubated with IB4 from Grif-
fonia simplicifolia I (GSA I) (Vector Laboratories, Burlin-
game, CA), followed by overnight incubation with goat
anti-GSA I antiserum (1:2,000, Vector Laboratories).
Finally, these sections were incubated with RRX-conjugated
TABLE 1.
Primary Antibodies Used for Immunohistochemical Studies
Antigen Description of immunogen
Source, host species, cat. #,
clone or lot#, RRID Concentration or dilution
Galanin Albumin-conjugated, synthetic peptide
corresponding to amino acids 1–29 of rat
galanin
Theodorsson E. (Link€
oping, Sweden), rabbit
polyclonal
1:4,000 (TSA)
NPY Synthetic peptide corresponding to full-length
of porcine NPY
Wong H. and Walsh JH. (UCLA, CA), rabbit
polyclonal
1:4,000 (TSA)
CGRP Synthetic peptide corresponding to amino acids
23–37 of Tyr-a-r CGRP
Terenius L. (Stockholm, Sweden), rabbit
polyclonal
1:20,000 (Normal)
SP Albumin-conjugated, synthetic peptide
corresponding to full-length of substance P
Terenius L. (Stockholm, Sweden), rabbit
polyclonal
1:2,000 (TSA)
a2d1 Synthetic peptide corresponding to amino acids
528–668 of human a2d1
Sigma, rabbit polyclonal, Cat# HPA008213,
Lot# A40079 RRID: AB_1233654
0.05lg/ml (TSA)
P2X3 Synthetic peptide corresponding to amino acids
383–397 of rat P2X3
Elde R. (Minneapolis, MN), rabbit polyclonal 1:2,000 (TSA)
GAP43 Recombinant protein corresponding to full-
length of rat GAP-43
Chemicon, rabbit polyclonal, Cat# ab5220,
Lot# NG1780675, RRID: AB_2107282
7.5 lg/ml (TSA)
ATF3 Recombinant protein corresponding to amino
acids 194–212 of Human ATF3 (C-terminal)
Santa Cruz, rabbit polyclonal, Cat# sc-188,
clone C-19, RRID: AB_2258513
0.0 5 lg/ml (TSA)
IB4 IB4 from Griffoniia simplicifolia I Vector Laboratories, Cat# L1104 0.05 lg/ml (Normal)
TABLE 2.
CAIA-Induced Changes in Lumbar DRGs
Inflammatory
phase
Postinflammatory
phase
CAIA LPS CAIA LPS
SP 5555
CGRP 5555
NPY 5555
Galanin """5
Nav1.7
1
5555
Nav1.8
1
5555
a2d1"""5
P2X3 """5
ATF3 "555
GAP43 """5
The table summarizes immunohistochemical analysis of neuropepti-
des, ion channels, and nerve injury markers in DRGs in the CAIA and
LPS control group in comparison to the saline control group. “"” rep-
resents increased and “5” represents similar expression levels in the
CAIA and LPS groups compared to the saline control group.
1
Based on mRNA analysis.
Neuronal plasticity in arthritic mice
The Journal of Comparative Neurology |Research in Systems Neuroscience 1507
donkey antigoat antibody at RT for 2 hours (1:200, Jackson
Laboratories) to visualize the IB4 binding.
Antibody characterization
Details describing immunogen, source, and concen-
trations for all antibodies used in this study are listed
in Table 1. Briefly, the galanin antiserum was generated
from rat galanin (amino acids 1–29), and only a single
band at 10 kDa was observed on western blot of
mouse neonatal basal forebrain (Keimpema et al.,
2013). IHC in DRG and spinal cord showed patterns
that were similar to what has previously been reported
(e.g., Villar et al., 1991). The NPY central terminal stain-
ing in the spinal dorsal horn showed a similar pattern
as previously (e.g., Brumovsky et al., 2002). Immunohis-
tochemistry of CGRP antiserum in DRG and spinal cord
was in agreement with previous reports (e.g., Villar
et al., 1991; Brumovsky et al., 2002). The SP antibody
immunostaining showed similar patterns in both DRG
and spinal cord as reported in previous studies (e.g.,
Villar et al., 1991; Lindh et al., 1993). The calcium
channel subunit alpha 2 delta 1 (CACNA2D1, a2d1)
antibody recognized a single band of 134 kDa on west-
ern blot of rat brain lysate (manufacturer’s datasheet),
and stained cytoplasma of peripheral neurons and its
central terminals in the spinal dorsal horn, while no
staining was detected in the A2D1 knockout DRG or
spinal cord tissue in the current study. The P2X3 anti-
serum was generated against the carboxy-terminus of
rat P2X3 (amino acids 383–397) and it recognized a
single band at 57 kDa on western blot of P2X3-
transfected HEK293 cells (Vulchanova et al., 1997).
P2X3 immunostaining showed the same pattern in DRG
as in previous work (Vulchanova et al., 1998), although
we were not able to generate reproducible immuno-
staining in the spinal cord.
The growth-associated protein 43 (GAP43) antibody
was produced against recombinant rat GAP43 (com-
plete sequence), and recognized two bands on western
blot membranes, with the upper band at 43 kDa corre-
sponded to the expected size and western blot data
from mouse brain lysate (manufacturer’s datasheet).
Staining in neuronal cytoplasm was similar as described
in previous publications (Hirata et al., 2002; Gupta
et al., 2006). The ATF3 antibody was raised against
amino acids 163–181 in the C-terminus of human ATF3
and detailed information of characterization was pub-
lished previously (Starkey et al., 2009). It recognized a
single band of 21 kD on western blot membrane of rat
cultured DRGs (Seijffers et al., 2006). ATF3 nuclear
staining patterns in the current study are consistent
with previously work (Tsujino et al., 2000; Yamanaka
et al., 2011).
Image analysis and quantification
Representative pictures were captured in an LSM710
confocal laser-scanning microscope (Carl Zeiss, Jena,
Germany) operated with ZEN2012 software (Zeiss).
Emission spectra for each dye were limited as follows:
Fluorescein (505–540 nm), RRX (560–610 nm), and PI
(590–680 nm). Multipanel figures were assembled in
Adobe Photoshop CS6 software (Adobe Systems, San
Jose, CA). The proportion of neuron profiles (NPs) (num-
ber of immunopositive NPs / total NPs labeled by PI)
binding antibodies to galanin, CGRP, SP, CACNA2D1,
P2X3, ATF3, GAP43, or IB4 in DRGs was counted from
three or four randomly selected sections at different
levels (with a minimum of 50 lm in between).
The relative IR intensity of NPs was measured in sec-
tions of CAIA and control DRGs mounted on the same
slide by ImageJ (NIH, http://rsb.info.nih.gov/ij/index.
html). Neurons with an intensity of more than the mean
plus 2-fold standard deviation (SD) of background were
considered positive. Size distribution of immunopositive
NPs in DRGs was assessed based on the cross-sectional
area of cells with nucleus. Cross-sectional area was meas-
ured in pictures scanned with LSM710 equipped with a
Plan-Apochromat M27 objective (203, N.A. 0.80) with
ImageJ (NIH) software. Size distribution was performed
according to the criterion of Scherrer et al. (2010) (small
<300 lm
2
, medium: 300–700 lm
2
, large >700 lm
2
).
Images of spinal cords of CAIA and control groups
mounted on the same slide and stained at the same
time were captured using a Nikon TE300 fluorescence
microscope (Nikon, Tokyo, Japan) and the integrated
signal intensity was measured subsequent to back-
ground subtraction using ImageJ (NIH). The relative IR
intensity of galanin, NPY, CGRP, SP, or IB4 expression
in the spinal cord was analyzed in three sections per
animal (with a minimum of 50 lm in between) and per-
centage change of CAIA and LPS animals compared to
saline control group was calculated.
For counting the number of galanin positive(
1
)and
NPY
1
interneuron NPs in the spinal dorsal horn, five sec-
tions from each animal were selected and the number of
galanin
1
and NPY
1
NPs of both sides in superficial
layers of spinal cord was counted with a 633oil-
immersion lens, and the total number was calculated.
For assessment of colocalization between GAP43-IR neu-
rons and CGRP or IB4-IR neurons in DRGs, the number
of positive NPs was counted in three to five sections per
animal, and the percentage of overlap was calculated.
Quantitative real-time polymerase chain
reaction (PCR) (qPCR)
The mRNA from frozen lumber DRGs (L4, L5) was
extracted using Trizol (Invitrogen, Carlsbad, CA) according
J. Su et al.
1508 The Journal of Comparative Neurology |Research in Systems Neuroscience
to the manufacturer’s protocol. Complementary DNA was
prepared and quantitative real-time PCR performed with
TaqMan Gene Expression Assays (Applied Biosystems,
Carlsbad, CA) using the GeneAmp 7500 Fast Sequence
Detection system (Applied Biosystems). Predeveloped spe-
cific primers for Nav1.7 (Mm00450762_s1) and Nav1.8
(Mm00501467_m1) were used to detect the targets. Rel-
ative abundance was calculated by using a standard curve
generated by adding a dilution series of cDNA from
10,000 RAW264.7 cells (ATCC, Bethesda, MD) for each
primer and the data were then normalized to Hprt1
(Mm01545399_m1) (all primers from Applied Biosystems)
gene expression and presented as percent change of con-
trol levels.
Statistics
Behavior data are expressed as mean 6the standard
error of the mean (SEM), while IHC and qPCR results
are expressed as mean 6SD and analyzed with prism 5
(GraphPad Software, La Jolla, CA). For comparison of
pain-like behavior, a two-way analysis of variance
(ANOVA) for repeated measures with Bonferroni post-
hoc tests was used, while for comparison of clinical
score the Kruskal–Wallis test, followed by Dunn’s multi-
ple comparison post-hoc test was used. For comparison
of percentages and signal IR intensity, a one-way
ANOVA with Bonferroni post-hoc test was used for
detecting differences between the CAIA, saline, and
LPS groups. All the mRNA changes were analyzed by a
one-way ANOVA with Bonferroni post-hoc test for com-
parison of percentage changes from control groups.
P<0.05 was considered significant.
RESULTS
CAIA induces transient inflammation but
persistent hypersensitivity
Mice subjected to CAIA developed joint swelling and
redness which was evident as an increase in arthritis
scores starting on day 6 after CII antibody cocktail
injection, peaking around day 15 and being resolved by
day 36 (Fig. 1A). In contrast, mechanical hypersensitiv-
ity was detected from day 3 and remained pronounced
throughout day 47 (Fig. 1B). In this study we refer to
the period with signs of inflammation together with
evoked pain-like behavior as the "inflammatory phase"
(days 12–18), and the period with pronounced mechani-
cal hypersensitivity without visual signs of inflammation
as the "postinflammatory" or the "late" phase (days
45–47). The tactile thresholds in the control group that
received LPS, but no CII antibodies, did not differ signif-
icantly from the saline control group at any timepoint
(Fig. 1B). Neither LPS nor saline control mice developed
any signs of inflammation (Fig. 1A).
Expression of neuropeptides
CAIA increases expression of galanin peptide
in DRGs
Galanin has been reported to be present in na
€
ıve
mouse DRGs and spinal cord (Ch’ng et al., 1985; Sko-
fitsch and Jacobowitz, 1985). In agreement, galanin was
predominately expressed in small-sized neurons and
only in a few medium-size and large neurons in control
DRGs. The percentage of galanin
1
NPs, but not the
galanin signal intensity in the neurons, was increased in
the CAIA lumbar DRGs compared to the saline control
Figure 1. CAIA induces transient joint inflammation and persistent
mechanical hypersensitivity. A: Disease severity is assessed by count-
ing the number of inflamed digits and ankles (arthritis score) and shows
that mice injected with collagen type II antibody cocktail i.v. on day 0
and LPS i.p. on day 5 develop transient joint inflammation, while the
control groups, injected with saline i.v. day 0 and saline i.p. or LPS i.p
day 5, respectively, do not. B: Mechanical hypersensitivity is assessed
by measuring withdrawal thresholds with von Frey filaments. Data
show that the CAIA but not the saline and LPS control groups develop
persistent mechanical hypersensitivity. The data are presented as
mean 6SEM, n512 mice/group. *P<0.05 and **P<0.01 vs. saline
controls,
#
P<0.05 and
##
P<0.01 vs. LPS controls.
Neuronal plasticity in arthritic mice
The Journal of Comparative Neurology |Research in Systems Neuroscience 1509
group during the inflammatory (CAIA 17 62% vs. con-
trol 10 61%, P<0.01) and late phases (CAIA 20 61%
vs. control 13 62%, P<0.01) (Fig. 2A,B). The increase
in CAIA DRGs was restricted to small-sized neurons.
Interestingly, the proportion of galanin
1
NPs was signifi-
cantly increased in lumbar DRGs from the control mice
subjected to i.p. LPS without CII antibody administra-
tion in the inflammatory (14 61% vs. control 10 61%,
P<0.01) but not in the late phase (13 63% vs. control
13 62%, P>0.05) (Fig. 2C,D).
Figure 2. CAIA leads to upregulation of galanin expression in lumbar DRGs. A,B: Representative confocal images of galanin immunoreactiv-
ity in DRGs from saline control, CAIA, and LPS control groups during the inflammatory (A, day 15) and late (B, day 47) phases. C,D: Quan-
tification of galanin immunoreactivity. The bar graphs show that (C), the percentage of galanin-positive neurons is significantly elevated in
both the CAIA and LPS lumbar DRGs compared to the saline control group, while no differences in galanin signal intensity (% of control)
and galanin neuronal size distribution are observed between the three groups in the inflammatory phase. (D) In the late phase the percent-
age of galanin-positive neurons is increased in the CAIA group but not the other groups. Data are presented as the mean 6SD, n55
mice/group. **P<0.01 vs. saline controls,
#
P<0.05;
##
P<0.01 vs. LPS controls. Scale bar 550 lm.
J. Su et al.
1510 The Journal of Comparative Neurology |Research in Systems Neuroscience
We also evaluated galanin-like immunoreactivity (LI)
in the lumbar segment (L4–L5) of the spinal dorsal
horn. Here galanin was found in a superficial dense net-
work of nerve endings, particularly in lamina I–II, but no
difference was observed between the CAIA and control
groups in any of two phases (Fig. 6A,B). Furthermore, a
few galanin
1
interneurons were detected in the superfi-
cial layers (lamina I–II), and their number was similar in
all three groups in both phases (inflammatory phase:
CAIA 55 614 vs. control 49 611 and LPS 48 67,
P>0.05; late phase: CAIA 60 622 vs. control 50 619
and LPS 48 610, P>0.05) (Fig. 8E,F).
CAIA does not change expression of
SP, CGRP, NPY, or IB4
Next we examined the levels of NPY, CGRP, and SP in
DRGs and spinal cord. The percentage of SP
1
and
CGRP
1
NPs in saline control DRGs on day 15 was
16 63% and 38 62%, respectively (Figs. 3A, 4A). No sig-
nificant changes (P>0.05) were observed in the CAIA
group, either during the inflammatory (SP, 16 62%;
CGRP, 37 69%) (Figs. 3C, 4C), or the late (SP: control
17 63% vs. CAIA 19 63%; CGRP: control 38 63% vs.
CAIA 37 67%) (Figs. 3D, 4D) phases. Also, the two sig-
nal intensities were unaltered when comparing the con-
trol and CAIA groups (P>0.05), and there were no
differences in soma size distribution of either SP
1
(Fig.
3C,D) or CGRP
1
(Fig. 4C,D) NPs. No NPY
1
DRG neurons
were found in any group or at any timepoint. Further-
more, we stained DRGs for both galanin and CGRP and
found that most galanin-positive sensory neurons colocal-
ized with CGRP. The percentage of galanin-expressing
cells that also expressed CGRP did not change in the
CAIA group compared to saline and LPS control animals
in both phases (inflammatory phase: CAIA 81 66% vs.
control 76 66% and LPS 76 68%, P>0.05; late phase:
CAIA 83 64% vs. control 81 65% and LPS 85 67%,
P>0.05) (Fig. 5C,D). IB4 is used as a marker for non-
peptidergic neurons (McMahon and Priestley, 1995). No
difference in the percentage of IB4
1
NPs was found
between CAIA and control DRGs in the inflammatory
(CAIA 30 67% vs. saline 31 64% and LPS 25 64%,
P>0.05) or the late (CAIA 29 63% vs. saline 29 64%
and LPS 33 62%, P>0.05) phase.
It has been reported that IB4 staining in the spinal
dorsal horn is reduced in models of nerve injury (Bailey
and Ribeiro-da-Silva, 2006; Huang et al., 2013b). We
examined IB4 staining in the spinal cord and found that
IB4-binding protein expression in the primary afferent
central terminals in lamina II in spinal dorsal horn (Fig.
6C,D). However, no difference in the relative intensity
of the IB4 staining was found between the CAIA and
control groups (inflammatory phase: CAIA 92 620% vs.
control 100 620% and LPS 82 610%, P>0.05; late
phase: CAIA 109 624% vs. control 100 610% and LPS
93 610%, P>0.05) (Fig. 6C,D).
With regard to the superficial NPY
1
,SP
1
, and
CGRP
1
fiber networks in the dorsal horn, no apparent
differences were found between the groups (Fig. 7A–
7F). In contrast, the total number of NPY
1
NPs in lam-
ina I–III was increased in the CAIA mice compared to
the saline control group in both the inflammatory phase
(CAIA 11 63 vs. control 6 62, P<0.05) and late
phase (CAIA 13 67 vs. control 5 61, P<0.05) (Fig.
8G,H), and in the late phase the number of NPY
1
inter-
neuron NPs were was also increased compared to the
LPS group (CAIA 13 67 vs. LPS 2 61, P<0.01)
(Fig. 8H).
Expression of ion channels
Nav1.7 and Nav1.8 mRNA expression are
unaltered in DRGs
Sodium ion channels, in particular Nav1.3, Nav1.7,
Nav1.8, and Nav1.9 (Eijkelkamp et al., 2012), are
important components of pain signaling, and their
expression levels are increased in peripheral sensory
neurons in different models of chronic pain (Hains
et al., 2003; Chattopadhyay et al., 2008). Here we
assessed the expression of Nav1.7 and Nav1.8 in
DRGs. Antibodies from two different sources (Novus
Biologicals, Littleton, CO, and Alomone Labs, Israel) and
different protocols (with or without antigen retrieval and
signal amplification) were tested, but we did not obtain
specific staining in mouse DRGs. As an alternative
approach we used a qPCR methodology, but no statisti-
cally significant difference in Nav1.7 (Fig. 9A,B) or
Nav1.8 (Fig. 9C,D) mRNA levels was detected between
groups in DRGs at any timepoint (P>0.05).
Upregulation of a2d1 expression following
induction of CAIA
The a2d1 subunit (CACNA2D1) has received special
attention, as it is upregulated in DRG neurons and asso-
ciated with enhanced pain signaling following nerve
injury (Luo et al., 2001). In fact, the number of a2d1
1
NPs in DRGs was dramatically increased in both the
inflammatory phase (CAIA 35 65%, vs. saline 16 64%,
P<0.01) (Fig. 10A) and the late phase (CAIA 36 69%,
vs. saline 18 63%, P<0.01) (Fig. 10B). There was also
a pronounced increase in signal intensity in the CAIA
group compared to the saline control group in the
inflammatory (CAIA 256 654% vs. saline 100 613%,
P<0.01) (Fig. 10C) and late phase (CAIA 331 671%
vs. saline 100 631%, P<0.01) (Fig. 10D). a2d1 was
predominantly expressed in medium-to-large size neu-
rons. Induction of CAIA did not lead to a shift in size
Neuronal plasticity in arthritic mice
The Journal of Comparative Neurology |Research in Systems Neuroscience 1511
distribution of a2d1 expression in L4/L5 DRG NPs in the
inflammatory phase (Fig. 10C). However, soma size distri-
bution of the a2d1
1
neurons shifted towards the left,
from large to medium-sized neurons in the CAIA group,
and generated significant differences compared to the
control groups in the late phase (P<0.05) (Fig. 10D). The
number of a2d1-expressing neurons and the a2d1signal
intensity were higher in the CAIA group compared to the
LPS control group (P<0.05) (Fig. 10C,D). LPS injection
alone also led to an increase in the number of a2d1
1
NPs (LPS 26 69% vs. saline 18 63%, P<0.05) (Fig.
10D) and in a2d1-signal intensity (LPS 179 659% vs.
saline 100 613%, P<0.05) (Fig. 10D) in the late phase
comparedtothesalinegroup.
Figure 3. SP immunoreactivity is not altered in DRGs from mice subjected to CAIA. A,B: Representative confocal images of SP immunore-
activity in DRGs from saline control, CAIA, and LPS control groups during the inflammatory (A, day 15) and late (B, day 47) phases. C,D:
No differences in the percentage of SP-positive neurons, SP signal intensity, or SP neuronal size distribution are observed between the
groups in the (C) inflammatory and (D) late phases. Data are as the mean 6SD, n55 mice/group. Scale bar 550 lm.
J. Su et al.
1512 The Journal of Comparative Neurology |Research in Systems Neuroscience
Increased protein expression of P2X3 in the
CAIA DRGs
Increasing numbers of reports show that purinergic
receptors play important roles both in inflammation and
nerve injury-driven nociceptive signaling (Bradbury
et al., 1998; Souslova et al., 2000; Tsuda et al., 2010).
Therefore, we examined if induction of CAIA alters the
levels of P2X3 receptor expression in DRGs. On day 15
the number of P2X3
1
NPs in L4/L5 DRG neurons was
increased in the CAIA and LPS group compared to the
saline control group (49 64% and 48 66%, respec-
tively, vs. 30 63%, P<0.01) (Fig. 11C). In contrast, on
Figure 4. CGRP immunoreactivity is not altered in DRGs from mice subjected to CAIA. A,B: Representative confocal images of CGRP
immunoreactivity in DRGs from saline control, CAIA, and LPS control groups during the inflammatory (A, day 15) and late (B, day 47)
phases. C,D: No differences in the percentage of CGRP-positive neurons, CGRP signal intensity, or CGRP neuronal size distribution are
observed between the groups in the (C) inflammatory and (D) late phases. Data are presented as the mean 6SD, n55 mice/group. Scale
bar 550 lm.
Neuronal plasticity in arthritic mice
The Journal of Comparative Neurology |Research in Systems Neuroscience 1513
day 47 the percentage of P2X3 receptor
1
NPs was
only significantly elevated in the CAIA group (CAIA
55 65% vs. LPS control 31 69% and saline control
35 65%, P<0.01) (Fig. 11C,D). No difference between
the groups was found in size distribution or signal inten-
sity (P>0.05). P2X3 immunoreactivity has been shown
in rat spinal cord as a band of axon terminals in the
lamina II (Saeed and Ribeiro-da-Silva, 2012). Despite
using four different antibodies and different protocols,
including the one that is used for P2X3 labeling in
DRGs, we were unsuccessful in generating reproducible
P2X3 staining in the spinal dorsal horn.
Increased expression of ATF3 and GAP43 in
DRGs
Following the report by Tsujino et al. (2000), numerous
studies have shown that ATF3 expression is induced in
DRGs in experimental models of axonal injury. In the cur-
rent study ATF3 protein expression was absent in na
€
ıve
mouse lumbar DRGs, but following induction of arthritis
a small, transient nuclear ATF3 expression was detected
in DRG NPs on day 15 (CAIA 1 60.5% vs. no expression
in saline and LPS controls) (Fig. 12A,B). ATF3
1
immuno-
staining was not seen in any group on day 47.
Increased GAP43 gene and protein expression in DRG
neurons is another marker of nerve injury (Jacobson et al.,
1986). We found that GAP43 is expressed in the cytoplasm
of DRG neurons and fibers. The number of NPs expressing
GAP43 was elevated in lumbar DRGs in the CAIA (CAIA
28 61% vs. saline 20 63%, P<0.01) (Fig. 13C), and the
LPS groups (29 64%, P<0.05) (Fig. 13C) in comparison to
saline controls on day 15. In the late phase, however, there
was an increase only in the CAIA group (CAIA 30 62% vs.
saline 20 63% and LPS 21 62%, P<0.01) (Fig. 13C, D).
Induction of CAIA did not alter signal intensity of neuro-
nal GAP43-LI (Fig. 13C), nor did it lead to a shift in neuronal
size distribution (Fig. 13C,D). In order to determine which
subpopulation of the sensory neurons expresses GAP43,
we performed double-staining of GAP43 with IB4 or CGRP.
In DRGs, GAP43 was predominantly found in the peptider-
gic subpopulation (75 67% of the NPs), while very few,
2%, were colabeled with IB4. No change in the ratio of
GAP43-expressing CGRP
1
and IB4
1
cells was found
between the CAIA and control groups (Fig. 14A,B).
DISCUSSION
In the current study we characterized changes in
markers, known to be associated with nerve injury and
Figure 5. Galanin and CGRP colocalization in the CAIA DRGs. A,B: Representative confocal images depicting double-staining of galanin (green) with
CGRP (magenta) in lumbar DRGs of saline, CAIA, and LPS groups in the inflammatory and the late phases (asterisks show galanin
1
/CGRP
–
neurons).
C,D: The bar graphs show the percentage of galanin-expressing cells that also expressed CGRP did not change in the CAIA group compared to saline
and LPS control animals in both phases. Data are presented as the mean 6SD, n54–5 mice/group. Scale bar 550 lm.
J. Su et al.
1514 The Journal of Comparative Neurology |Research in Systems Neuroscience
inflammation, in DRG neurons and spinal cord at two
stages after antibody-induced joint inflammation. We
found that galanin, a2d1, P2X3, and GAP43 were
upregulated in the lumbar DRGs in the CAIA group both
in the inflammatory (with joint inflammation and pain-
like behavior) and in the late (with pain-like behavior in
the absence of joint inflammation) phases, while
expression levels of SP, CGRP, NPY, IB4, Nav1.7, and
Nav1.8 were similar in all groups. Taken together, these
data point to the possibility that neurochemical plastic-
ity in the peripheral nerves is driven by antibody-
induced joint inflammation and associated with develop-
ment of long-lasting mechanical hypersensitivity in the
CAIA model, and provide new insights to the potential
underlying mechanisms of RA-induced persistent pain.
Antibody-induced inflammation induces
upregulation of "nerve injury markers" in
DRG neurons
In 15% of the cases, nerve injury-induced pain in
humans does not resolve and persists even after the
injury has healed (Bouhassira et al., 2008; Perrot et al.,
2013; Timmerman et al., 2014). In subgroups of RA
patients the degree of joint inflammation and joint pain
are not always paralleled, and sometimes even
uncoupled (Lee et al., 2011; Walsh and McWilliams,
2014). Even though the new treatment strategies effi-
ciently halt disease progression, pain continues to be a
difficult problem for a considerable number of RA
patients with minimal disease activity (Pollard et al.,
2006; Taylor et al., 2011; Walsh and McWilliams,
2012). RA pain is often described as gnawing or aching,
which suggests nociceptive mechanisms induced by
inflammation or joint damage (Roche et al., 2003).
However, descriptors that are more characteristic of
neuropathic pain such as "burning" or "shooting" are
also common for RA pain, indicating possible compo-
nents of nerve damage (Roche et al., 2003; Sokka,
2011). While analgesic drugs such as gabapentin and
buprenorphine attenuate the mechanical hypersensitiv-
ity during both the inflammatory and postinflammatory
phases in our experimental models of RA, the cyclooxy-
genase (COX) inhibitors diclofenac and ketorolac are
antinociceptive only during the inflammatory phase
(Christianson et al., 2010; Bas et al., 2012). This
Figure 6. No change in spinal immunoreactivity for galanin and IB4 in response to CAIA. A–D: Representative confocal images showing
immunoreactivity in the dorsal horn of the spinal cord for (A,B) galanin and (C,D) IB4 during the inflammatory and late phases. Bar graphs
showing no apparent difference in expression intensity between the CAIA and control groups at any timepoint. Data are presented as
mean 6SD, n54–5 mice/group. Scale bar 5100 lm.
Neuronal plasticity in arthritic mice
The Journal of Comparative Neurology |Research in Systems Neuroscience 1515
indicates that while pain-like behavior is driven by pros-
taglandins during ongoing joint inflammation, mechani-
cal hypersensitivity in the phase subsequent to
inflammation is not prostaglandin-dependent. These
findings have led us to hypothesize that the neurochem-
ical profile in DRG and spinal cord is altered by
antibody-driven joint inflammation generating a unique
neurochemical signature. Of note, similar "nontypical"
changes in neurochemical factors in the peripheral
nerve have been reported in models of bone cancer
(Honore et al., 2000), indicating that bone-related
pathology can cause a drift towards mechanisms that
are in part associated with neuropathic pain in the long
term.
GAP43 is involved in axonal regeneration and neuro-
nal structural plasticity (Jacobson et al., 1986; Coggins
and Zwiers, 1991; Mosevitsky, 2005), and nerve injury
results in an increased expression of GAP43 in neuronal
cell bodies and axons (Gonzalez-Hernandez and Rus-
tioni, 1999). ATF3 is another factor that is commonly
used as a marker of nerve injury, as it is induced in
DRG neurons after complete and partial nerve transec-
tion and nerve ligation (Tsujino et al., 2000; Tsuzuki
et al., 2001; Boeshore et al., 2004; Pezet et al., 2006;
Brumovsky et al., 2011), but not after inflammation
(Braz and Basbaum, 2010). Although ATF3 is very useful
as a marker of nerve injury, it should be noted that this
marker is not only increased in response to nerve injury
Figure 7. NPY, SP, and CGRP immunoreactivity in the CAIA spinal dorsal horn. A–F: Representative confocal images showing immunoreac-
tivity in the dorsal horn of spinal cord for (A,B) NPY, (C,D) SP, and (E,F) CGRP. No significant change in relative intensity of immunoreactiv-
ity in the CAIA group compared to the control groups during the inflammatory or the late phase. Data are presented as mean 6SD, n55
mice/group. Scale bar 5100 lm.
J. Su et al.
1516 The Journal of Comparative Neurology |Research in Systems Neuroscience
but also in a variety of tissues to, e.g., cytokines, hor-
mones, DNA damage, and endoplasmic reticulum stress
(Chen et al., 1996; Hai et al., 1999; Hamann et al.,
2006; Uesugi et al., 2011). Thus, ATF3 induction may
occur for other reasons than nerve injury.
We assessed both ATF3 and GAP43 expression in
DRGs from CAIA mice, and found that GAP43 elevation
is robust both in the early and the late phase, whereas
upregulation of ATF3 was modest and only detectable
in the inflammatory phase. Of note, in line with our
work these two factors has been found elevated in
DRGs following induction of monoiodoacetate (MIA)-
induced arthritis (Ivanavicius et al., 2007; Thakur et al.,
2012) and collagen-induced arthritis (CIA) (Inglis et al.,
2007). The later model is based on immunization
against collagen type II and, therefore, is similar to the
CAIA model, driven by collagen II antibodies, but with
the difference that it involves activation of T-cells and
B-cells, and has a pathology that gets progressively
worse. Even though the MIA, CIA, and the CAIA models
Figure 8. Expression of galanin and NPY in spinal dorsal horn interneurons. A–D: Representative confocal images show (A,B) galanin and
(C,D) NPY immunoreactivity counterstained with DAPI in the superficial layer of lumbar spinal dorsal horn in the (A,C) inflammatory and
(B,D) postinflammatory phases, respectively, in saline and LPS control mice and mice subjected to CAIA. E–H: Bar graphs showing that
while no difference in number of (E,F) galanin-immunopositive interneurons is observed between the three groups, the number of (G,H)
NPY-immunopositive interneurons is increased in the spinal cords from the CAIA group both in the inflammatory and late phases. Inter-
neurons are indicated by arrowheads. Data are presented as the mean 6SD, n54–6 mice/group. *P<0.05 CAIA vs. saline controls;
##
P<0.01 vs. LPS controls. Scale bar 5200 lm.
Neuronal plasticity in arthritic mice
The Journal of Comparative Neurology |Research in Systems Neuroscience 1517
differ in terms of degree and length of inflammation,
bone and cartilage remodeling is taking place in all
models. Thus it is possible that the inflammatory pro-
cess in conjunction with bone and/or cartilage remod-
eling in the joint introduce changes in the local
environment that induces stress or is damaging to sen-
sory neurons.
It has been shown that the IB4 staining of nonpepti-
dergic fibers in the spinal dorsal horn is reduced follow-
ing sciatic nerve transection and constriction (Molander
et al., 1996; Bailey and Ribeiro-da-Silva, 2006). We did
not find a reduction in IB4-binding protein expression in
DRGs or spinal dorsal horn from mice subjected to
CAIA, which indicates that the neurochemical changes
taking place after antibody-driven joint inflammation are
not identical to what has been observed after frank
nerve injury. Thus, further studies are warranted in
order to delineate which factors that are altered after
nerve injury are changed in the same direction in mod-
els of arthritis. For example, after nerve injury Reg-2,
which is thought to have proregenerative properties in
motor and sensory neurons after peripheral nerve
injury, is upregulated first in IB4-positive neurons and
later in medium–large diameter cells (Livesey et al.,
1997; Averill et al., 2002). Reg-2 expression is driven
by cytokines of the interleukin-6 (IL-6) family, which
includes, e.g., ciliary neurotrophic factor (CNTF) and
leukemia inhibitory factor (LIF). Interestingly, even
though Reg-2 has been thought to predominantly be
associated with nerve injury, Averill et al. (2008)
showed that also CFA-induced inflammation in the paw
upregulates Reg-2 in IB4/P2X3-positive neurons. Thus,
in future studies it will be important to examine if levels
of Reg-2 and neuropoietic factors such as IL-6, CNTF,
and LIF are altered and involved in nociceptive signal
transmission after antibody-induced joint inflammation.
Neuropeptide changes in DRGs resemble
nerve injury-driven
Galanin (Tatemoto et al., 1983) is a well-known neu-
ropeptide with normally only low detectable levels in
Figure 9. Gene expression of Nav1.7 and Nav1.8 in DRGs. A–D: Bar graphs presenting mRNA levels of Nav1.7 and Nav1.8 as % of control
showing no difference in (A,B) Nav1.7 and (C,D) Nav1.8 mRNA levels between mice subjected to CAIA and control mice in the (A,C)
inflammatory or (B,D) late phase. Data are presented as mean 6SD, n55 mice/group.
J. Su et al.
1518 The Journal of Comparative Neurology |Research in Systems Neuroscience
DRGs (Ch’ng et al., 1985; Skofitsch and Jacobowitz,
1985). Of note, galanin may act both on spinal inhibi-
tory interneurons through galanin receptor 1 (GAL1)
and through excitatory galanin receptor 2 (GAL2) on
DRG neurons and thereby differently regulate nocicep-
tive signal transmission (Liu and Hokfelt, 2002). Follow-
ing nerve injury there is a pronounced upregulation of
galanin in the small-sized DRG neurons, with a typical
Figure 10. CAIA leads to upregulation of the voltage gated calcium channel subunit a2d1 expression in lumbar DRGs. A,B: Representative
confocal images of a2d1 immunoreactivity in DRGs from saline control, CAIA, and LPS control groups during the inflammatory (A, day 15)
and late (B, day 47) phases. C,D: Quantification of a2d1 immunoreactivity in DRGs. The bar graphs show that (C), the percentage of
a2d1-positive neurons and a2d1 signal intensity are significantly elevated in the CAIA lumbar DRGs compared to the saline and LPS con-
trol groups, while no change in neuronal size distribution of a2d1 immunoreactivity is observed. (D) In the late phase the percentage of
a2d1-positive neurons and the a2d1 signal intensity are increased in the CAIA group and the LPS control group. More medium-sized while
less large-sized a2d1 IR neurons are observed in the CAIA DRGs compared to saline group in the late phase. Data are presented as the
mean 6SD, n55 mice/group. *P<0.05; **P<0.01 vs. saline controls,
#
P<0.05;
##
P<0.01 vs. LPS controls. Scale bar 550 lm.
Neuronal plasticity in arthritic mice
The Journal of Comparative Neurology |Research in Systems Neuroscience 1519
shift to expression also in medium/large-sized neurons
(Hokfelt et al., 1987, 2013; Villar et al., 1988; Zhang
et al., 2011). This increase in galanin expression is
assumed to indicate recruitment of programs associ-
ated with neuropathic pain. Of note, while there is no
increase in galanin expression in DRGs during periph-
eral inflammation induced by carrageenan injected into
the plantar side of the paw (Ji et al., 1995), immuniza-
tion with CFA injected to the base of the tail leads to
an increase in galanin expression in DRGs (Calza et al.,
Figure 11. CAIA leads to upregulation of the ATP gated ion channel in lumbar DRGs. A,B: Representative confocal images of P2X3 immu-
noreactivity in DRGs from saline control, CAIA, and LPS control groups during the inflammatory (A, day 15) and late (B, day 47) phases.
C,D: Quantification of P2X3 immunoreactivity. The bar graphs show that (C), the percentage of P2X3-positive neurons is significantly ele-
vated in both the CAIA and LPS lumbar DRGs compared to the saline control group, while no differences in P2X3 signal intensity (% of
control) and P2X3 neuronal size distribution are observed between the three groups in the inflammatory phase. (D) In the late phase the
percentage of P2X3-positive neurons is increased in the CAIA group but not the other groups. Data are presented as the mean 6SD,
n55 mice/group. **P<0.01 vs. saline controls,
##
P<0.01 vs. LPS controls. Scale bar 550 lm.
J. Su et al.
1520 The Journal of Comparative Neurology |Research in Systems Neuroscience
2000). Thus, even though the total number of peptider-
gic subpopulation (CGRP
1
) was not altered, an increase
in the percentage of galanin-positive DRG neurons was
observed in the CAIA model. This mimics the effector
phase of B-cell activation, suggesting that inflammation
driven by the adaptive immune system may cause
damage-like changes in the peripheral branches of the
DRG neurons, or that galanin expression in addition to
nerve injury is induced by certain types of inflammatory
processes.
NPY is absent in na
€
ıve DRGs, but de novo expression
in mainly large neurons has been reported after periph-
eral nerve injury (Wakisaka et al., 1992) and monoar-
thritis (Ferreira-Gomes et al., 2012). Interestingly, while
our data in na
€
ıve mice are in line with previous reports,
induction of CAIA did not alter NPY expression in lum-
bar DRGs. Since nerve injury upregulates NPY mainly in
large DRG neurons (Wakisaka et al., 1992), our findings
indicate that this neuronal population is not affected by
the collagen antibody treatment. Meanwhile, more spi-
nal local neurons expressed NPY in the arthritis group
compared to the controls, which indicates involvement
of NPY in arthritis-induced spinal nociceptive signaling.
Joint inflammation induced by CFA, CIA, and adjuvant-
induced arthritis (AIA) in rats leads to upregulation of
CGRP and SP expression in DRGs and dorsal horn neu-
rons (Hanesch et al., 1993, 1995; Weihe et al., 1995;
Calza et al., 1998). Surprisingly, we did not find any evi-
dence of altered SP and CGRP levels or distribution at
the inflammatory or later timepoint after induction of
antibody-driven joint inflammation in mice. Also, the
number of galanin-expressing neuron profiles that colo-
calized with CGRP remained constant. Although the pos-
sibility that SP and/or CGRP expression is altered at
other timepoints than investigated, or that our observa-
tion is species related, cannot be excluded, the present
result suggests that antibody-driven joint inflammation
generates a neurochemical profile that is different com-
pared to what is observed after innate immune-driven
inflammation or nerve injury-induced pain.
Mechanical hypersensitivity in CAIA:
upregulation of voltage-gated calcium, but
not sodium channels in DRGs
The excitability of neurons is to a large extent depend-
ent on voltage gated sodium channels, as these propa-
gate the nerve impulses. Nav1.3, Nav1.7, and Nav1.8
are expressed on nociceptors and have been implicated
in persistent pain (Liu and Wood, 2011; Dib-Hajj et al.,
2013). Nav1.7 expression is increased in DRG neurons
following peripheral inflammation, mediated, e.g., by
inflammatory factors like prostaglandins, and disruption
of Nav1.7 in nociceptors attenuates pain hypersensitivity
in several inflammatory pain models (Gould et al., 2004;
Nassar et al., 2004). Nav1.7 has not yet been linked to
neuropathic pain (Nassar et al., 2005). In contrast,
Nav1.8 expression is increased in DRGs both by inflam-
mation and nerve transection (Amir et al., 2006), but
while disrupting Nav1.8 abolishes inflammatory pain (Yu
et al., 2011), it only prevents certain types of neuro-
pathic pain (Stirling et al., 2005; Kistner et al., 2010).
These findings suggest that Nav1.7 and Nav1.8 are dif-
ferentially involved in inflammatory vs. neuropathic pain.
Figure 12. CAIA induces transient expression of activating transcription factor 3 (ATF3) in DRGs. A,B: Representative confocal images
showing nuclear ATF3 immunoreactivity in DRGs from CAIA and control mice in the (A) inflammatory and (B) late phases. C,D: Quantifica-
tion of ATF3 immunoreactivity. The bar graphs show an increase in the percentage ATF3 immunopositive DRG neurons in the CAIA group
during the (C) inflammatory but not the (D) late phase. Data are presented as mean 6SD, n55 mice/group. **P<0.01 vs. saline
controls,
##
P<0.01 vs. LPS controls. Scale bar 550 lm.
Neuronal plasticity in arthritic mice
The Journal of Comparative Neurology |Research in Systems Neuroscience 1521
We undertook multiple experiments with the goal to
assess Nav1.7 and Nav1.8 protein expression in the
CAIA and control DRGs. We used antibodies from two
different companies but despite previously published
successful immunohistological detection of these sodium
channels, we failed to generate specific staining patterns
(Huang et al., 2013a; Sadamasu et al., 2014). As an
alternative approach we analyzed mRNA levels of Nav1.7
and Nav1.8 in DRGs. We did not observe a change of
Nav1.7 or Nav1.8 gene expression in DRGs following
Figure 13. CAIA leads to upregulation of the GAP43 in lumbar DRGs. A,B: Representative confocal images of GAP43 immunoreactivity in
DRGs from saline control, CAIA, and LPS control groups during the inflammatory (A, day 15) and late (B, day 47) phases. C,D: Quantifica-
tion of GAP43 immunoreactivity. The bar graphs show that (C), the percentage of GAP43-positive neurons is significantly increased in both
the CAIA and LPS groups compared to the saline control group, while no differences in GAP43 signal intensity (% of control) and GAP43
neuronal size distribution are observed between the three groups in the inflammatory phase. (D) In the late phase the percentage
of GAP43-positive neurons is increased in the CAIA group but not the other groups. Data are presented as the mean 6SD, n55
mice/group. **P<0.01 vs. saline controls;
##
P<0.01 vs. LPS controls. Scale bar 550 lm.
J. Su et al.
1522 The Journal of Comparative Neurology |Research in Systems Neuroscience
induction of CAIA compared to control groups, but fur-
ther experiments are warranted to determine if protein
levels of Nav1.7 or Nav1.8 changes in the CAIA DRGs.
The calcium channel a2d1 subunit (Dolphin, 2013) is
markedly upregulated, both in terms of number of immu-
noreactive neurons and signal intensity in DRGs following
nerve injury (Luo et al., 2001; Newton et al., 2001; Bor-
oujerdi et al., 2011), while a modest increase in a2d1
mRNA level is observed in the DRGs of CFA rats (day 3)
(Lu et al., 2010). We found an increase of a2d1, both in
terms of number of a2d1-positive neurons and a2d1 sig-
nal intensity in the CAIA DRGs during both phases.
Moreover, in the late phase we found increased a2d1-
labeled medium-sized neurons (shift from Abfibered
sensory neurons to Adfibered nociceptors) in the CAIA
DRGs, which is an additional indication of that long-term
joint inflammation induces nerve injury-like phenotype.
Mechanical hypersensitivity in CAIA:
upregulation of ATP-gated ion channels in
DRGs
Purinergic ionotropic P2X receptors, activated by ATP
and permeable to sodium, calcium, and potassium, are
considered to play important roles in nociceptive signal
transduction in different experimental pain models (Burn-
stock, 2013). While the ATP-gated P2X3 receptor is upreg-
ulated in nociceptors in several models of inflammation
(Xu and Huang, 2002a; Pan et al., 2009), reports show
that P2X3 levels may increase, remain unchanged, or
decrease following nerve injury in a model and time-
dependent fashion (Bradbury et al., 1998; Novakovic et al.,
1999; Tsuzuki et al., 2001; Kage et al., 2002; Averill et al.,
2004; Chen et al., 2005). When elevated, P2X3 mediates
enhanced sensitivity to mechanical noxious stimulation fol-
lowing both inflammation and nerve injury (Novakovic
et al., 1999; Xu and Huang, 2002b; Braz et al., 2011). In
our study, P2X3 was upregulated in DRGs during and after
joint inflammation. Thus, while the expression of sodium
ion channels did not change following induction of CAIA,
neuronal excitability appears modified via increased
expression of calcium ion and P2X3 channels.
LPS alters expression of certain
neuropeptides and ion channels without
inducing mechanical hypersensitivity
Noteworthy, although galanin, GAP43, and a2d1protein
expression levels were elevated transiently in DRGs 10 days
after a relatively small dose of systemic LPS, mechanical
thresholds in the saline and LPS control groups were very
similar throughout the study. Thus, LPS-induced galanin and
a2d1 expression does not appear to be coupled to induc-
tion of mechanical hypersensitivity, or are counterregulated
by other factors after LPS injection. Further studies are war-
ranted for a detailed understanding of these findings.
CONCLUSION
Primary sensory neurons represent the interface
between the external stimulus in terminal sensory nerve
Figure 14. GAP43 and CGRP colocalization in DRGs. A,B: Representative confocal images depicting double-staining of GAP43 with (A)
CGRP (peptidergic neurons) and (B) IB4 (nonpeptidergic neurons) in saline and CAIA DRGs (D47). Scale bar 550 lm.
Neuronal plasticity in arthritic mice
The Journal of Comparative Neurology |Research in Systems Neuroscience 1523
endings (innervating the skin and other structures, e.g.,
in the joint) and the central nervous system (spinal
cord). The current view is that different types of under-
lying pathologies evoke specific adaptations that
include changes in expression of a battery of heteroge-
neously expressed neuropeptides, ion channels, and
receptors. These are often classified based on the
underlying pain-generating condition as being predomi-
nantly inflammatory or neuropathic. We have found that
following an antibody-driven inflammation in the joint,
factors like galanin, a2d1, ATF3, and GAP43 are altered
in the same direction as after nerve injury. In contrast,
other nociception-related neuropeptides like NPY,
CGRP, and SP are expressed in equivalent amounts in
all three groups and modulated differently in conditions
of antibody-driven joint inflammation compared to other
models of inflammatory and neuropathic pain.
Taken together, the present results indicate that at
least some of the inflammation-induced changes in pri-
mary sensory neurons become similar to the neuropathic
pain phenotype, and suggest that an episode of joint
inflammation induces stress or damage-related programs
in peripheral sensory neurons. Hence, CAIA-induced noci-
ception shows features of both the classic inflammatory
pain and neuropathic pain and the net sum of changes
point to that antibody-driven inflammation leads to a
unique neurochemical signature. This highlights the
importance of considering the nociceptive process driven
by antibody-induced joint inflammation, representing
pain in conditions like rheumatoid arthritis, to have a
disease-specific neurochemical pain phenotype. Impor-
tantly, this profile may reflect the underlying pain pathol-
ogy, and our data indicates that it may change over time
and persist beyond the episode of joint inflammation.
ACKNOWLEDGMENTS
For the generous donation of antisera, we thank Profs.
Lars Terenius, Karolinska Institutet, Stockholm, and Ingrid
Nylander, Uppsala University, Uppsala, Sweden (SP and
CGRP), Prof. Elvar Theodorsson, Link€
oping University,
Link€
oping, Sweden (galanin), Prof. Robert P. Elde, Univer-
sity of Minnesota, Minneapolis, MN, (P2X3) and the late
Drs. Helene Wong and John H. Walsh, UCLA, Los Angeles,
CA, (NPY) and Prof. Mathias Uhl
en and Dr. Jan Mulder,
both Science for Life Laboratory, Karolinska Institutet,
Stockholm (a2d1). We thank Prof. A. Dolphin, University
College London, London, UK, for generous donation of
A2D1 knockout tissue.
CONFLICT OF INTEREST
The authors declare that they have no conflicts of
interest.
ROLE OF AUTHORS
All authors had full access to all the data in the
study and take responsibility for the integrity of the
data and accuracy of the data analysis. Study concept
design: JS, TJS, TH and CIS. Acquisition of data: JS,
TLG, TJS and QX. Analysis and interpretation of data: JS
and CIS. Drafting of the article: JS, TH and CIS. Critical
revision of article for important intellectual content: TH
and CIS. Statistical analysis: JS and CIS. Obtained fund-
ing: CIS and JS. Administrative, technical, and material
support: JS, TLG, TJS, QX, XJX and ZWH. Study supervi-
sion: CIS, TH, ZWH and XJX.
LITERATURE CITED
Amir R, Argoff CE, Bennett GJ, Cummins TR, Durieux ME,
Gerner P, Gold MS, Porreca F, Strichartz GR. 2006. The
role of sodium channels in chronic inflammatory and
neuropathic pain. J Pain 7(5 Suppl 3):S1–29.
Andersson ML, Svensson B, Bergman S. 2013. Chronic wide-
spread pain in patients with rheumatoid arthritis and the
relation between pain and disease activity measures
over the first 5 years. J Rheumatol 40:1977–1985.
Averill S, Davis DR, Shortland PJ, Priestley JV, Hunt SP. 2002.
Dynamic pattern of reg-2 expression in rat sensory neu-
rons after peripheral nerve injury. J Neurosci 22:7493–
7501.
Averill S, Michael GJ, Shortland PJ, Leavesley RC, King VR,
Bradbury EJ, McMahon SB, Priestley JV. 2004. NGF and
GDNF ameliorate the increase in ATF3 expression which
occurs in dorsal root ganglion cells in response to
peripheral nerve injury. Eur J Neurosci 19:1437–1445.
Averill S, Inglis JJ, King VR, Thompson SW, Cafferty WB,
Shortland PJ, Hunt SP, Kidd BL, Priestley JV. 2008. Reg-2
expression in dorsal root ganglion neurons after
adjuvant-induced monoarthritis. Neuroscience 155:1227–
1236.
Bailey AL, Ribeiro-da-Silva A. 2006. Transient loss of termi-
nals from non-peptidergic nociceptive fibers in the sub-
stantia gelatinosa of spinal cord following chronic
constriction injury of the sciatic nerve. Neuroscience
138:675–690.
Bas DB, Su J, Sandor K, Agalave NM, Lundberg J, Codeluppi
S, Baharpoor A, Nandakumar KS, Holmdahl R, Svensson
CI. 2012. Collagen antibody-induced arthritis evokes per-
sistent pain with spinal glial involvement and transient
prostaglandin dependency. Arthritis Rheum 64:3886–
3896.
Boeshore KL, Schreiber RC, Vaccariello SA, Sachs HH, Salazar
R, Lee J, Ratan RR, Leahy P, Zigmond RE. 2004. Novel
changes in gene expression following axotomy of a sym-
pathetic ganglion: a microarray analysis. J Neurobiol 59:
216–235.
Boroujerdi A, Zeng J, Sharp K, Kim D, Steward O, Luo ZD.
2011. Calcium channel alpha-2-delta-1 protein upregula-
tion in dorsal spinal cord mediates spinal cord injury-
induced neuropathic pain states. Pain 152:649–655.
Bouhassira D, Lanteri-Minet M, Attal N, Laurent B, Touboul C.
2008. Prevalence of chronic pain with neuropathic char-
acteristics in the general population. Pain 136:380–387.
Bradbury EJ, Burnstock G, McMahon SB. 1998. The expres-
sion of P2X3 purinoreceptors in sensory neurons: effects
of axotomy and glial-derived neurotrophic factor. Mol
Cell Neurosci 12:256–268.
J. Su et al.
1524 The Journal of Comparative Neurology |Research in Systems Neuroscience
Braz JM, Basbaum AI. 2010. Differential ATF3 expression in
dorsal root ganglion neurons reveals the profile of pri-
mary afferents engaged by diverse noxious chemical
stimuli. Pain 150:290–301.
Braz JM, Ackerman L, Basbaum AI. 2011. Sciatic nerve trans-
ection triggers release and intercellular transfer of a
genetically expressed macromolecular tracer in dorsal
root ganglia. J Comp Neurol 519:2648–2657.
Brumovsky PR, Shi TJ, Matsuda H, Kopp J, Villar MJ, Hokfelt T.
2002. NPY Y1 receptors are present in axonal processes
of DRG neurons. Exp Neurol 174:1–10.
Brumovsky P, Stanic D, Shuster S, Herzog H, Villar M, Hokfelt
T. 2005. Neuropeptide Y2 receptor protein is present in
peptidergic and nonpeptidergic primary sensory neurons
of the mouse. J Comp Neurol 489:328–348.
Brumovsky PR, Seroogy KB, Lundgren KH, Watanabe M,
Hokfelt T, Gebhart GF. 2011. Some lumbar sympathetic
neurons develop a glutamatergic phenotype after periph-
eral axotomy with a note on VGLUT(2)-positive perineuro-
nal baskets. Exp Neurol 230:258–272.
Burnstock G. 2013. Purinergic mechanisms and pain—an
update. Eur J Pharmacol 716:24–40.
Calza L, Pozza M, Zanni M, Manzini CU, Manzini E, Hokfelt T.
1998. Peptide plasticity in primary sensory neurons and
spinal cord during adjuvant-induced arthritis in the rat:
an immunocytochemical and in situ hybridization study.
Neuroscience 82:575–589.
Calza L, Pozza M, Arletti R, Manzini E, Hokfelt T. 2000. Long-
lasting regulation of galanin, opioid, and other peptides
in dorsal root ganglia and spinal cord during experimen-
tal polyarthritis. Exp Neurol 164:333–343.
Ch’ng JL, Christofides ND, Anand P, Gibson SJ, Allen YS, Su
HC, Tatemoto K, Morrison JF, Polak JM, Bloom SR. 1985.
Distribution of galanin immunoreactivity in the central
nervous system and the responses of galanin-containing
neuronal pathways to injury. Neuroscience 16:343–354.
Chaplan SR, Bach FW, Pogrel JW, Chung JM, Yaksh TL. 1994.
Quantitative assessment of tactile allodynia in the rat
paw. J Neurosci Methods 53:55–63.
Chattopadhyay M, Mata M, Fink DJ. 2008. Continuous delta-
opioid receptor activation reduces neuronal voltage-
gated sodium channel (NaV1.7) levels through activation
of protein kinase C in painful diabetic neuropathy.
J Neurosci 28:6652–6658.
Chen BP, Wolfgang CD, Hai T. 1996. Analysis of ATF3, a tran-
scription factor induced by physiological stresses and
modulated by gadd153/Chop10. Mol Cell Biol 16:1157–
1168.
Chen Y, Li GW, Wang C, Gu Y, Huang LY. 2005. Mechanisms
underlying enhanced P2X receptor-mediated responses
in the neuropathic pain state. Pain 119:38–48.
Christianson CA, Corr M, Firestein GS, Mobargha A, Yaksh TL,
Svensson CI. 2010. Characterization of the acute and
persistent pain state present in K/BxN serum transfer
arthritis. Pain 151:394–403.
Coggins PJ, Zwiers H. 1991. B-50 (GAP-43): biochemistry and
functional neurochemistry of a neuron-specific phospho-
protein. J Neurochem 56:1095–1106.
Coronel MF, Brumovsky PR, Hokfelt T, Villar MJ. 2008. Differ-
ential galanin upregulation in dorsal root ganglia and
spinal cord after graded single ligature nerve constric-
tion of the rat sciatic nerve. J Chem Neuroanat 35:94–
100.
Dib-Hajj SD, Yang Y, Black JA, Waxman SG. 2013. The
Na(V)1.7 sodium channel: from molecule to man. Nat
Rev Neurosci 14:49–62.
Dolphin AC. 2013. The alpha2delta subunits of voltage-gated
calcium channels. Bioch Biophys Acta 1828:1541–1549.
Donnerer J, Schuligoi R, Stein C, Amann R. 1993. Upregula-
tion, release and axonal transport of substance P and
calcitonin gene-related peptide in adjuvant inflammation
and regulatory function of nerve growth factor. Regul
Pept 46:150–154.
Eijkelkamp N, Linley JE, Baker MD, Minett MS, Cregg R,
Werdehausen R, Rugiero F, Wood JN. 2012. Neurological
perspectives on voltage-gated sodium channels. Brain
135(Pt 9):2585–2612.
Ferreira-Gomes J, Adaes S, Sousa RM, Mendonca M, Castro-
Lopes JM. 2012. Dose-dependent expression of neuronal
injury markers during experimental osteoarthritis induced
by monoiodoacetate in the rat. Mol Pain 8:50.
Fukuoka T, Kobayashi K, Yamanaka H, Obata K, Dai Y,
Noguchi K. 2008. Comparative study of the distribution
of the alpha-subunits of voltage-gated sodium channels
in normal and axotomized rat dorsal root ganglion neu-
rons. J Comp Neurol 510:188–206.
Fukuoka T, Yamanaka H, Kobayashi K, Okubo M, Miyoshi K,
Dai Y, Noguchi K. 2012. Re-evaluation of the phenotypic
changes in L4 dorsal root ganglion neurons after L5 spi-
nal nerve ligation. Pain 153:68–79.
Galeazza MT, Garry MG, Yost HJ, Strait KA, Hargreaves KM,
Seybold VS. 1995. Plasticity in the synthesis and storage
of substance P and calcitonin gene-related peptide in pri-
mary afferent neurons during peripheral inflammation.
Neuroscience 66:443–458.
Gonzalez-Hernandez T, Rustioni A. 1999. Nitric oxide synthase
and growth-associated protein are coexpressed in pri-
mary sensory neurons after peripheral injury. J Comp
Neurol 404:64–74.
Gould HJ 3rd, England JD, Soignier RD, Nolan P, Minor LD, Liu
ZP, Levinson SR, Paul D. 2004. Ibuprofen blocks changes
in Na v 1.7 and 1.8 sodium channels associated with
complete Freund’s adjuvant-induced inflammation in rat.
J Pain 5:270–280.
Gupta R, Rummler LS, Palispis W, Truong L, Chao T, Rowshan
K, Mozaffar T, Steward O. 2006. Local down-regulation
of myelin-associated glycoprotein permits axonal sprout-
ing with chronic nerve compression injury. Exp Neurol
200:418–429.
Hai T, Wolfgang CD, Marsee DK, Allen AE, Sivaprasad U.
1999. ATF3 and stress responses. Gene Expr 7:321–
335.
Hains BC, Klein JP, Saab CY, Craner MJ, Black JA, Waxman
SG. 2003. Upregulation of sodium channel Nav1.3 and
functional involvement in neuronal hyperexcitability asso-
ciated with central neuropathic pain after spinal cord
injury. J Neurosci 23:8881–8892.
Hamann W, Abou-Sherif S, Thompson S, Hall S. 2006. Pulsed
radiofrequency applied to dorsal root ganglia causes a
selective increase in ATF3 in small neurons. Eur J Pain
10:171–176.
Hanesch U, Pfrommer U, Grubb BD, Schaible HG. 1993. Acute
and chronic phases of unilateral inflammation in rat’s
ankle are associated with an increase in the proportion
of calcitonin gene-related peptide-immunoreactive dorsal
root ganglion cells. Eur J Neurosci 5:154–161.
Hanesch U, Blecher F, Stiller RU, Emson PC, Schaible HG,
Heppelmann B. 1995. The effect of a unilateral inflam-
mation at the rat’s ankle joint on the expression of
preprotachykinin-A mRNA and preprosomatostatin mRNA
in dorsal root ganglion cells—a study using non-
radioactive in situ hybridization. Brain Res 700:279–284.
Hirata A, Masaki T, Motoyoshi K, Kamakura K. 2002. Intrathe-
cal administration of nerve growth factor delays GAP 43
expression and early phase regeneration of adult rat
peripheral nerve. Brain research 944:146–156.
Neuronal plasticity in arthritic mice
The Journal of Comparative Neurology |Research in Systems Neuroscience 1525
Hokfelt T, Fried G, Hansen S, Johansson O, Ljungdahl A,
Lundberg JM, Schultzberg M. 1984. [Nerve cells can
have more than one signal substance: coexistence of
classical transmitters and peptides.] Lakartidningen 81:
4933–4939.
Hokfelt T, Wiesenfeld-Hallin Z, Villar M, Melander T. 1987.
Increase of galanin-like immunoreactivity in rat dorsal
root ganglion cells after peripheral axotomy. Neurosci
Lett 83:217–220.
Hokfelt T, Zhang X, Villar M, Xu XJ, Wiesenfeld-Hallin Z. 2013.
Central consequences of peripheral nerve damage. In:
McMahon SB, Wall PD, editors. Wall and Melzack’s text-
book of pain, 6th ed. Philadelphia: Elsevier Saunders. p
902–914.
Honore P, Rogers SD, Schwei MJ, Salak-Johnson JL, Luger
NM, Sabino MC, Clohisy DR, Mantyh PW. 2000. Murine
models of inflammatory, neuropathic and cancer pain
each generates a unique set of neurochemical changes
in the spinal cord and sensory neurons. Neuroscience
98:585–598.
Huang CP, Chen HN, Su HL, Hsieh CL, Chen WH, Lai ZR, Lin
YW. 2013a. Electroacupuncture reduces carrageenan-
and CFA-induced inflammatory pain accompanied by
changing the expression of Nav1.7 and Nav1.8, rather
than Nav1.9, in mice dorsal root ganglia. Evid Based
Complement Alternat Med 2013:312184.
Huang W, Calvo M, Karu K, Olausen HR, Bathgate G, Okuse
K, Bennett DL, Rice AS. 2013b. A clinically relevant
rodent model of the HIV antiretroviral drug stavudine
induced painful peripheral neuropathy. Pain 154:560–
575.
Inglis JJ, Notley CA, Essex D, Wilson AW, Feldmann M, Anand
P, Williams R. 2007. Collagen-induced arthritis as a
model of hyperalgesia: functional and cellular analysis of
the analgesic actions of tumor necrosis factor blockade.
Arthritis Rheum 56:4015–4023.
Ivanavicius SP, Ball AD, Heapy CG, Westwood FR, Murray F,
Read SJ. 2007. Structural pathology in a rodent model of
osteoarthritis is associated with neuropathic pain:
increased expression of ATF-3 and pharmacological char-
acterisation. Pain 128:272–282.
Jacobson RD, Virag I, Skene JH. 1986. A protein associated
with axon growth, GAP-43, is widely distributed and
developmentally regulated in rat CNS. J Neurosci 6:
1843–1855.
Ji RR, Zhang X, Wiesenfeld-Hallin Z, Hokfelt T. 1994. Expres-
sion of neuropeptide Y and neuropeptide Y (Y1) receptor
mRNA in rat spinal cord and dorsal root ganglia following
peripheral tissue inflammation. J Neurosci 14(11 Pt 1):
6423–6434.
Ji RR, Zhang X, Zhang Q, Dagerlind A, Nilsson S, Wiesenfeld-
Hallin Z, Hokfelt T. 1995. Central and peripheral expres-
sion of galanin in response to inflammation. Neuro-
science 68:563–576.
Kage K, Niforatos W, Zhu CZ, Lynch KJ, Honore P, Jarvis MF.
2002. Alteration of dorsal root ganglion P2X3 receptor
expression and function following spinal nerve ligation in
the rat. Exp Brain Res 147:511–519.
Keimpema E, Zheng K, Barde SS, Berghuis P, Dobszay MB,
Schnell R, Mulder J, Luiten PG, Xu ZD, Runesson J,
Langel U, Lu B, Hokfelt T, Harkany T. 2013. GABAergic
terminals are a source of galanin to modulate cholinergic
neuron development in the neonatal forebrain. Cereb
Cortex 24:3277–3288.
Kistner K, Zimmermann K, Ehnert C, Reeh PW, Leffler A.
2010. The tetrodotoxin-resistant Na1channel Na (v)1.8
reduces the potency of local anesthetics in blocking C-
fiber nociceptors. Pflugers Arch 459:751–763.
Lee YC, Cui J, Lu B, Frits ML, Iannaccone CK, Shadick NA,
Weinblatt ME, Solomon DH. 2011. Pain persists in
DAS28 rheumatoid arthritis remission but not in ACR/
EULAR remission: a longitudinal observational study.
Arthritis Res Ther 13:R83.
Lindh B, Risling M, Remahl S, Terenius L, Hokfelt T. 1993.
Peptide-immunoreactive neurons and nerve fibres in lum-
bosacral sympathetic ganglia: selective elimination of a
pathway-specific expression of immunoreactivities follow-
ing sciatic nerve resection in kittens. Neuroscience 55:
545–562.
Liu HX, Hokfelt T. 2002. The participation of galanin in pain
processing at the spinal level. Trends Pharmacol Sci 23:
468–474.
Liu M, Wood JN. 2011. The roles of sodium channels in noci-
ception: implications for mechanisms of neuropathic
pain. Pain Med 12(Suppl 3):S93–99.
Livesey FJ, O’Brien JA, Li M, Smith AG, Murphy LJ, Hunt SP.
1997. A Schwann cell mitogen accompanying regenera-
tion of motor neurons. Nature 390:614–618.
Lu SG, Zhang XL, Luo ZD, Gold MS. 2010. Persistent inflam-
mation alters the density and distribution of voltage-
activated calcium channels in subpopulations of rat cuta-
neous DRG neurons. Pain 151:633–643.
Luo ZD, Chaplan SR, Higuera ES, Sorkin LS, Stauderman KA,
Williams ME, Yaksh TL. 2001. Upregulation of dorsal root
ganglion (alpha)2(delta) calcium channel subunit and its
correlation with allodynia in spinal nerve-injured rats.
J Neurosci 21:1868–1875.
McInnes IB, Schett G. 2011. The pathogenesis of rheumatoid
arthritis. N Engl J Med 365:2205–2219.
McMahon SB, Priestley JV. 1995. Peripheral neuropathies and
neurotrophic factors: animal models and clinical perspec-
tives. Curr Opin Neurobiol 5:616–624.
Molander C, Wang HF, Rivero-Melian C, Grant G. 1996. Early
decline and late restoration of spinal cord binding and
transganglionic transport of isolectin B4 from Griffonia
simplicifolia I after peripheral nerve transection or crush.
Restor Neurol Neurosci 10:123–133.
Mosevitsky MI. 2005. Nerve ending "signal" proteins GAP-43,
MARCKS, and BASP1. Int Rev Cytol 245:245–325.
Nassar MA, Stirling LC, Forlani G, Baker MD, Matthews EA,
Dickenson AH, Wood JN. 2004. Nociceptor-specific gene
deletion reveals a major role for Nav1.7 (PN1) in acute
and inflammatory pain. Proc Natl Acad Sci U S A 101:
12706–12711.
Nassar MA, Levato A, Stirling LC, Wood JN. 2005. Neuro-
pathic pain develops normally in mice lacking both
Na(v)1.7 and Na(v)1.8. Mol Pain 1:24.
Newton RA, Bingham S, Case PC, Sanger GJ, Lawson SN.
2001. Dorsal root ganglion neurons show increased
expression of the calcium channel alpha2delta-1 subunit
following partial sciatic nerve injury. Brain Res Mol Brain
Res 95:1–8.
Nothias F, Tessler A, Murray M. 1993. Restoration of sub-
stance P and calcitonin gene-related peptide in dorsal
root ganglia and dorsal horn after neonatal sciatic nerve
lesion. J Comp Neurol 334:370–384.
Novakovic SD, Kassotakis LC, Oglesby IB, Smith JA, Eglen RM,
Ford AP, Hunter JC. 1999. Immunocytochemical localiza-
tion of P2X3 purinoceptors in sensory neurons in naive
rats and following neuropathic injury. Pain 80:273–282.
Pan AH, Lu DH, Luo XG, Chen L, Li ZY. 2009. Formalin-
induced increase in P2X(3) receptor expression in dorsal
root ganglia: implications for nociception. Clin Exp Phar-
macol Physiol 36:e6–11.
Perrot S, Dieude P, Perocheau D, Allanore Y. 2013. Compari-
son of pain, pain burden, coping strategies, and attitudes
J. Su et al.
1526 The Journal of Comparative Neurology |Research in Systems Neuroscience
between patients with systemic sclerosis and patients
with rheumatoid arthritis: a cross-sectional study. Pain
Med 14:1776–1785.
Pezet S, Krzyzanowska A, Wong LF, Grist J, Mazarakis ND,
Georgievska B, McMahon SB. 2006. Reversal of neuro-
chemical changes and pain-related behavior in a model
of neuropathic pain using modified lentiviral vectors
expressing GDNF. Mol Ther 13:1101–1109.
Pollard LC, Choy EH, Gonzalez J, Khoshaba B, Scott DL. 2006.
Fatigue in rheumatoid arthritis reflects pain, not disease
activity. Rheumatology 45:885–889.
Roche PA, Klestov AC, Heim HM. 2003. Description of stable
pain in rheumatoid arthritis: a 6 year study. J Rheumatol
30:1733–1738.
Sadamasu A, Sakuma Y, Suzuki M, Orita S, Yamauchi K,
Inoue G, Aoki Y, Ishikawa T, Miyagi M, Kamoda H,
Kubota G, Oikawa Y, Inage K, Sainoh T, Sato J,
Nakamura J, Toyone T, Takahashi K, Ohtori S. 2014.
Upregulation of NaV1.7 in dorsal root ganglia after inter-
vertebral disc injury in rats. Spine 39:E421–426.
Saeed AW, Ribeiro-da-Silva A. 2012. Non-peptidergic primary
afferents are presynaptic to neurokinin-1 receptor immu-
noreactive lamina I projection neurons in rat spinal cord.
Mol Pain 8:64.
Scherrer G, Low SA, Wang X, Zhang J, Yamanaka H, Urban R,
Solorzano C, Harper B, Hnasko TS, Edwards RH,
Basbaum AI. 2010. VGLUT2 expression in primary affer-
ent neurons is essential for normal acute pain and
injury-induced heat hypersensitivity. Proc Natl Acad Sci
U S A 107:22296–22301.
Seijffers R, Allchorne AJ, Woolf CJ. 2006. The transcription
factor ATF-3 promotes neurite outgrowth. Mol Cell Neu-
rosci 32:143–154.
Skofitsch G, Jacobowitz DM. 1985. Galanin-like immunoreac-
tivity in capsaicin sensitive sensory neurons and ganglia.
Brain Res Bull 15:191–195.
Smith GD, Seckl JR, Sheward WJ, Bennie JG, Carroll SM, Dick
H, Harmar AJ. 1991. Effect of adrenalectomy and dexa-
methasone on neuropeptide content of dorsal root gan-
glia in the rat. Brain Res 564:27–30.
Sokka T. 2011. Morning stiffness and other patient-reported
outcomes of rheumatoid arthritis in clinical practice.
Scand J Rheumatol Suppl 125:23–27.
Solway B, Bose SC, Corder G, Donahue RR, Taylor BK. 2011.
Tonic inhibition of chronic pain by neuropeptide Y. Proc
Natl Acad Sci U S A 108:7224–7229.
Souslova V, Cesare P, Ding Y, Akopian AN, Stanfa L, Suzuki
R, Carpenter K, Dickenson A, Boyce S, Hill R, Nebenuis-
Oosthuizen D, Smith AJ, Kidd EJ, Wood JN. 2000. Warm-
coding deficits and aberrant inflammatory pain in mice
lacking P2X3 receptors. Nature 407:1015–1017.
Starkey ML, Davies M, Yip PK, Carter LM, Wong DJ,
McMahon SB, Bradbury EJ. 2009. Expression of the
regeneration-associated protein SPRR1A in primary sen-
sory neurons and spinal cord of the adult mouse follow-
ing peripheral and central injury. J Comp Neurol 513:
51–68.
Stirling LC, Forlani G, Baker MD, Wood JN, Matthews EA,
Dickenson AH, Nassar MA. 2005. Nociceptor-specific
gene deletion using heterozygous NaV1.8-Cre recombi-
nase mice. Pain 113:27–36.
Tatemoto K, Rokaeus A, Jornvall H, McDonald TJ, Mutt V.
1983. Galanin — a novel biologically active peptide from
porcine intestine. FEBS Lett 164:124–128.
Taylor P, Gartemann J, Hsieh J, Creeden J. 2011. A systematic
review of serum biomarkers anti-cyclic citrullinated pep-
tide and rheumatoid factor as tests for rheumatoid
arthritis. Autoimmune Dis 2011:815038.
Thakur M, Rahman W, Hobbs C, Dickenson AH, Bennett DL.
2012. Characterisation of a peripheral neuropathic com-
ponent of the rat monoiodoacetate model of osteoarthri-
tis. PLoS One 7:e33730.
Timmerman H, Wilder-Smith O, van Weel C, Wolff A, Vissers
K. 2014. Detecting the neuropathic pain component in
the clinical setting: a study protocol for validation of
screening instruments for the presence of a neuropathic
pain component. BMC Neurol 14:94.
Tsuda M, Tozaki-Saitoh H, Inoue K. 2010. Pain and purinergic
signaling. Brain Res Rev 63:222–232.
Tsujino H, Kondo E, Fukuoka T, Dai Y, Tokunaga A, Miki K,
Yonenobu K, Ochi T, Noguchi K. 2000. Activating tran-
scription factor 3 (ATF3) induction by axotomy in sensory
and motoneurons: a novel neuronal marker of nerve
injury. Mol Cell Neurosci 15:170–182.
Tsuzuki K, Kondo E, Fukuoka T, Yi D, Tsujino H, Sakagami M,
Noguchi K. 2001. Differential regulation of P2X(3) mRNA
expression by peripheral nerve injury in intact and
injured neurons in the rat sensory ganglia. Pain 91:351–
360.
Uesugi K, Sekiguchi M, Kikuchi S, Konno S. 2011. The effect
of repeated restraint stress in pain-related behavior
induced by nucleus pulposus applied on the nerve root
in rats. Eur Spine 20:1885–1891.
Villar MJ, Vitale ML, Hokfelt T, Verhofstad AA. 1988. Dorsal
raphe serotoninergic branching neurons projecting both
to the lateral geniculate body and superior colliculus: a
combined retrograde tracing-immunohistochemical study
in the rat. J Comp Neurol 277:126–140.
Villar MJ, Wiesenfeld-Hallin Z, Xu XJ, Theodorsson E, Emson
PC, Hokfelt T. 1991. Further studies on galanin-, sub-
stance P-, and CGRP-like immunoreactivities in primary
sensory neurons and spinal cord: effects of dorsal rhi-
zotomies and sciatic nerve lesions. Exp Neurol 112:29–
39.
Vulchanova L, Riedl MS, Shuster SJ, Buell G, Surprenant A,
North RA, Elde R. 1997. Immunohistochemical study of
the P2X2 and P2X3 receptor subunits in rat and monkey
sensory neurons and their central terminals. Neurophar-
macology 36:1229–1242.
Vulchanova L, Riedl MS, Shuster SJ, Stone LS, Hargreaves
KM, Buell G, Surprenant A, North RA, Elde R. 1998.
P2X3 is expressed by DRG neurons that terminate in
inner lamina II. Eur Neurosci 10:3470–3478.
Wakisaka S, Kajander KC, Bennett GJ. 1992. Effects of periph-
eral nerve injuries and tissue inflammation on the levels
of neuropeptide Y-like immunoreactivity in rat primary
afferent neurons. Brain Res 598:349–352.
Walsh DA, McWilliams DF. 2012. Pain in rheumatoid arthritis.
Curr Pain Headache Rep 16:509–517.
Walsh DA, McWilliams DF. 2014. Mechanisms, impact and
management of pain in rheumatoid arthritis. Nat Rev
Rheumatol 10:581–592.
Weihe E, Nohr D, Schafer MK, Persson S, Ekstrom G,
Kallstrom J, Nyberg F, Post C. 1995. Calcitonin gene
related peptide gene expression in collagen-induced
arthritis. Can Physiol Pharmacol 73:1015–1019.
Xiang Z, Xiong Y, Yan N, Li X, Mao Y, Ni X, He C, LaMotte RH,
Burnstock G, Sun J. 2008. Functional up-regulation of
P2X 3 receptors in the chronically compressed dorsal
root ganglion. Pain 140:23–34.
Xu GY, Huang LY. 2002a. Peripheral inflammation sensitizes
P2X receptor-mediated responses in rat dorsal root gan-
glion neurons. J Neurosci 22:93–102.
Xu GY, Huang LY. 2002b. Peripheral inflammation sensitizes
P2X receptor-mediated responses in rat dorsal root gan-
glion neurons. J Neurosci 22:93–102.
Neuronal plasticity in arthritic mice
The Journal of Comparative Neurology |Research in Systems Neuroscience 1527
Yamanaka H, Kobayashi K, Okubo M, Fukuoka T, Noguchi K.
2011. Increase of close homolog of cell adhesion
molecule L1 in primary afferent by nerve injury and the
contribution to neuropathic pain. J Comp Neurol 519:
1597–1615.
Yu YQ, Zhao F, Guan SM, Chen J. 2011. Antisense-medi-
ated knockdown of Na(V)1.8, but not Na(V)1.9, gen-
erates inhibitory effects on complete Freund’s
adjuvant-induced inflammatory pain in rat. PLoS One
6:e19865.
Zhang YP, Fu ES, Sagen J, Levitt RC, Candiotti KA, Bethea JR,
Brambilla R. 2011. Glial NF-kappaB inhibition alters neu-
ropeptide expression after sciatic nerve injury in mice.
Brain Res 1385:38–46.
J. Su et al.
1528 The Journal of Comparative Neurology |Research in Systems Neuroscience
