Title: Which of these 6 candidates is Pain1
1Mapping a gene for neuropathic pain in the Pain1
QTL Tina Elahipanah1, Shihong Zhang2,3, Merav
Yarkoni-Abitul3, Ruslan Dorfman4, Yan Lu2, Edith
Gershon2, Jessica Livneh5,6, Yoram Shir7, Raphael
Pfeffer6, Tamar Peretz8, Jean-Jacques Vatine9,
David Tichauer2 and Zeev Seltzer1,2 1Faculty of
Medicine, Univ. of Toronto 2Faculty of
Dentistry, Toronto, ON, Canada 3School of
Medicine, Zhejiang University, Hangzhou, PR of
CHINA 4Hospital of Sick Children, Toronto, ON,
Canada 5Faculty of Dental Medicine, Hebrew
University, Jerusalem, Israel 6Sheba Medical
Center, Tel Hashomer, Israel 7McGill University,
Montreal, Canada 8Hadassah University Hospital,
Jerusalem, Israel 9Sackler Medical School, Tel
Aviv University, Israel
INTRODUCTION The hindpaw in mammals is
innervated by two peripheral nerves, the sciatic
and saphenous nerves. Transection of these nerves
causes total denervation and loss of sensibility
towards the stimuli applied to the hindpaw. The
axonal segments that became disconnected from
their cell body in dorsal root ganglia (DRGs)
degenerate (dotted lines in Fig. 1), while the
proximal segments try to regenerate by sprouting
nerve extensions that grow distally in an attempt
to reinnervate the original targets. If their
access is blocked, such as in human limb
amputees, the axonal sprouts of sensory, motor
and autonomic nerve fibers get caught in a
neuroma, a bulbous growth composed of blood
vessels and scar tissue. Fig. 1 The
Autotomy model of neuropathic pain Sprouts
of injured sensory fibers develop spontaneous
firing of action potentials (Fig. 1), that
produce pain in humans (painful neuroma). Such
firing also develops in the DRG cell body of
injured sensory fibers (Fig. 1). These input
sources drive rats and mice of certain genetic
backgrounds to express an abnormal behavior of
licking, scratching and self-mutilation of the
denervated hindpaw (autotomy). Wall et al.
(1979) suggested autotomy is an attempt to
relieve neuropathic pain that is referred to the
denervated paw. Wall et al. developed a scoring
system that assigns 1 point to injury of the
nails and an additional point to any injured
phalanx, up to a maximal score of 11 points.
Autotomy is an accepted model of neuropathic
pain. In 2001, Seltzer et al phenotyped the
incidence of autotomy scores in 23 recombinant
inbred (RI) lines of mice of the AXB-BXA set,
that were produced by crossing mice of the A/J
(A) and C57BL6/J (B) lines. As shown in Fig.
2, A line mice express significantly higher
levels of autotomy than B line mice. Fig. 2
Strain distribution pattern (SDP) of autotomy
scores 2 in 23 AXB-BXA RI lines and their A and
B parental lines. These lines were mapped
with 400 microsatellite genetic markers, each of
which identifies whether the marker locus
originated from the A or B parental lines. In a
linkage analysis between the phenotypic SDP and
the genetic map, Seltzer et al. identified an
interval on mouse chromosome 15 (Fig. 3) that
harbors gene(s) having a major effect on the
incidence of autotomy. In 2005,
Darvasi et al. replicated this finding using two
different mice lines and a different genetic
mapping method. However, their study did not
further refine the interval length of Pain1.
Which gene controls the
variance in autotomy levels at Pain1?
Human Pain1 position refines its map in mouse
OBJECTIVE
3. Changes in expression levels following nerve
injury
Fig. 9 Fold changes in mRNA levels expressed for
6 candidate pain genes in the DRGs of
sham-operated and denervated mice. Shown are the
levels in denervated mice. Expression levels are
relative to those of sham mice.
Cacng2 PValb MPST KCTD17 TMPRSS6 PBR
In addition to suggesting that Pain1 is a
conserved locus, mapping it to human chromosome
22 refined its confidence length in the mouse as
well. Fig. 6 Shaded in green is the refineed
confidence length of Pain1 on mouse chromosome
15 that corresponds to the position of Pain1 on
human chr. 22q12.
RESULTS
A new genetic map of the AXB-BXA RI set
became available in 2005. It is based on 8,500
SNPs, rather than on 400 microsatellite markers.
Using this map we hoped to refine the interval
length of Pain1. We tested the significance of
linkage between these SNPs and the same autotomy
trait as used by Seltzer et al. in 2001 (i.e., RI
line incidence of scores 2), as well as with
several other traits that were not used in that
study, including (i) the RI line incidence of
autotomy scores 1, 3, and 5, (ii) the average
first day that autotomy scores of 1, 2, 3, and 5
were observed in each RI line, and (iii) the
average autotomy score that each RI line
expressed on day 36 postoperatively (PO), the
last day of the experiment. Fig. 4 Significance
of the linkage (i.e., LOD scores) between the
incidence of RI line autotomy scores 2 and SNPs
on Chr. 15. While these analyses confirmed
the existence of a QTL for various aspects of
autotomy in the location of Pain1 on chromosome
15, as well as identifying novel QTLs on other
chromosomes, no further refinement of the
interval length of Pain1 was achieved. This may
result from several reasons (i) the genetic
diversity in 23 AXB-BXA RI lines available for
this set is not detailed enough, (ii) Pain1
harbors more than one causative gene that affect
autotomy.
1. In-silico refinement of the map of Pain1
Refined position of Pain1 in mouse
The refined confidence length of Pain1 in mouse
harbors the following six candidate genes
Cacng2, Mpst, Kctd17, Pvalb, Pbr, and Tmprss6 .
Only Kctd17 meets the 3 criteria that a gene must
meet in order for it to be an autotomy-regulating
gene 1. The expression levels in the
high-autotomy subgroup of the A mice must
contrast those of the low-autotomy subgroup of
this line. 2. The expression levels in the
high-autotomy subgroup of the A mice must
contrast those of the low-autotomy B mice. 3.
The expression levels in the low-autotomy
subgroup of the A mice must not be significantly
different from the low-autotomy B mice.
1
Which of these 6 candidates is Pain1? Levels of
mRNA expression of these 6 genes were determined
in intact A and B mice, and in mice of the same
lines following sham-operation and hindpaw
denervation by transection of the sciatic and
saphenous nerves. Table 1 The compared groups
and number of mice per line/group . Fig. 7
Mice were scored for autotomy on a daily basis.
Average (SEM) autotomy levels of the compared
groups. When reaching score 11 or on PO day 14
all mice were sacrificed, the L3-6 DRGs were
removed, stored in RNAlater at 20oC, mRNA was
extracted, and expression levels of these 6 genes
and additional reference genes were determined
with real-time PCR coupled with MASS-ARRAY.
4. Do Cacng2 and Kctd17 control neuropathic pain
in humans as well? Multiple SNPs in a number of
candidate genes, including Cacng2 and Kctd17,
were genotyped along Pain1 in our two human
cohorts (men leg amputees and women
postmastectomy). SNPs only in these candidate
genes significantly associated with chronic pain
levels.
There is a remarkable genetic homology in
mammals. For example, 98 of all mouse genes are
present in the human genome. Fig. 5 shows that
Pain1 on mouse chromosome 15 corresponds to two
chromosomes in humans 8 22. Fig. 5
Orthologous regions of mouse Pain1 on human
chromosomes 8 and 22. To find out
whether the same genes in mouse Pain1 that
control the variance in autotomy levels also play
a role in neuropathic pain in humans, we
genotyped the orthologous regions of mouse Pain1
in a cohort of humans that underwent nerve
injury, some developed neuropathic pain and
others were pain-free. We tested whether pain
levels associate significantly with specific
genotypes on these markers. We used 9
microsatellite markers to genotype 8q24, but none
of the genotypes showed significant association
with human neuropathic pain. However, two of the
markers on 22q12 significantly associated with
pain levels (p0.001-0.004), suggesting that
Pain1 is conserved and maps to human chr. 22q12.
2. Comparative approach to mapping Pain1 Is
Pain1 conserved in humans?
1
Fig. 10 Pain1 in the mouse (blue) and Pain1 on
human chr. 22q12 (green shade), showing the
location of Cacng2 and Kctd17. P values indicate
the significance of association found for some
SNPs and their genotypes in these two genes.
Constitutive mRNA levels in intact A and B mice
Conclusions 1) Autotomy in rodents is a valid
model for human neuropathic pain, since it
correctly predicted that genes controlling human
pain exist on human chr. 22q12. 2) Comparative
pain genetics is a rapid, efficient and robust
method for identifying conserved pain genes. 3)
Cacng2 and Kctd17 are promising candidate genes
that should be further studied as targets for RD
of novel analgesics that could prevent the
outbreak of chronic pain in humans and alleviate
it in those already in pain.
Fig. 8 Constitutive mRNA levels of
six candidate genes in intact A and B mice.
Cacng2 is the only gene that shows significant
contrast between these mice lines, suggesting a
protective role against neuropathic pain for
higher expression levels of this gene in B mice
compared to A mice.
1-
Fig. 3 Significance of linkage between the
incidence of autotomy scores 2 and loci markers
on chromosome 15 reveals a quantitative trait
locus (QTL) on chromosome 15 peaking at
marker D15Mit28 (plt0.0003).
Pain1
3.6 Mbp
Having higher levels of this gene expressed in
DRGs of intact B6 mice may protect
pain-suppressing neurons (GABA/Gly) in their CNS
against excitotoxicity of Injury discharge
Intact A mice have lower expression levels of
Cacng2, therefore, their unprotected spinal pain
gate opens up widely to ectopic inputs from the
injured nerve
References 1) Wall PD et al., Pain 1979
7109-113 2) Shir et al, Pain 2001 9075-78 3)
Seltzer Z et al., Pain 2001 93101-106 4) Devor
et al, Pain, 2005 116, 294-301
D