Content uploaded by Alan J Burns
Author content
All content in this area was uploaded by Alan J Burns
Content may be subject to copyright.
Proc.
Natl.
Acad.
Sci.
USA
Vol.
93,
pp.
12008-12013,
October
1996
Physiology
Interstitial
cells
of
Cajal
mediate
inhibitory
neurotransmission
in
the
stomach
ALAN
J.
BURNS*,
ALAN
E.
J.
LOMAx*,
SHIGEKO
TORIHASHIt,
KENTON
M.
SANDERS*,
AND
SEAN
M.
WARD*:
*Department
of
Physiology
and
Cell
Biology,
University
of
Nevada
School
of
Medicine,
Reno,
NV
89557;
and
tDepartment
of
Anatomy,
Nagoya
University
School
of
Medicine,
Tsurumai-cho
65,
Nagoya
466,
Japan
Communicated
by
C.
Ladd
Prosser,
University
of
Illinois
at
Urbana-Champaign,
Urbana,
IL,
June
11,
1996
(received
for
review
March
29,
1996)
ABSTRACT
The
structural
relationships
between
inter-
stitial
cells
of
Cajal
(ICC),
varicose
nerve
fibers,
and
smooth
muscle
cells
in
the
gastrointestinal
tract
have
led
to
the
suggestion
that
ICC
may
be
involved
in
or
mediate
enteric
neurotransmission.
We
characterized
the
distribution
of
ICC
in
the
murine
stomach
and
found
two
distinct
classes
on
the
basis
of
morphology
and
immunoreactivity
to
antibodies
against
c-Kit
receptors.
ICC
with
multiple
processes
formed
a
network
in
the
myenteric
plexus
region
from
corpus
to
pylorus.
Spindle-shaped
ICC
were
found
within
the
circular
and
longitudinal
muscle
layers
(IC-IM)
throughout
the
stom-
ach.
The
density
of
these
cells
was
greatest
in
the
proximal
stomach.
IC-IM
ran
along
nerve
fibers
and
were
closely
associated
with
nerve
terminals
and
adjacent
smooth
muscle
cells.
IC-IM
failed
to
develop
in
mice
with
mutations
in
c-kit.
Therefore,
we
used
W/WV
mutants
to
test
whether
IC-IM
mediate
neural
inputs
in
muscles
of
the
gastric
fundus.
The
distribution
of
inhibitory
nerves
in
the
stomachs
of
c-kit
mutants
was
normal,
but
NO-dependent
inhibitory
neuro-
regulation
was
greatly
reduced.
Smooth
muscle
tissues
of
W/Wv
mutants
relaxed
in
response
to
exogenous
sodium
nitroprusside,
but
the
membrane
potential
effects
of
sodium
nitroprusside
were
attenuated.
These
data
suggest
that
IC-IM
play
a
critical
serial
role
in
NO-dependent
neurotransmis-
sion:
the
cellular
mechanism(s)
responsible
for
transducing
NO
into
electrical
responses
may
be
expressed
in
IC-IM.
Loss
of
these
cells
causes
loss
of
electrical
responsiveness
and
greatly
reduces
responses
to
nitrergic
nerve
stimulation.
Ramon
y
Cajal
observed
cells
at
the
terminus
of
the
autonomic
nervous
system
in
the
acini
of
salivary
glands,
in
the
connective
tissue
of
the
pancreas,
between
the
glands
of
Lieberkuhn,
in
the
intestinal
villi,
and
within
the
tunica
muscularis
of
the
gastrointestinal
(GI)
tract
(1).
The
processes
of
cells
in
the
GI
tract,
which
became
known
as
interstitial
cells
of
Cajal
(ICC),
form
a
network
that
is
intercalated
between
nerve
terminals
and
effector
cells.
Cajal
believed
that
the
structures
he
iden-
tified
were
essential
elements
in
peripheral
neurotransmission
(2),
and
this
hypothesis
has
been
investigated
for
the
past
century
(3).
ICC
were
later
recognized
to
be
nonneural
(4),
but
their
anatomical
locations
in
the
GI
tract
continue
to
suggest
a
role
for
these
cells
in
neurotransmission
(3,
5).
Ultrastructural
studies
have demonstrated
very
close
apposi-
tion
between
some
classes
of
ICC
and
varicose
nerve
terminals
(5).
ICC
are
also
commonly
coupled
via
gap
junctions
to
neigh-
boring
smooth
muscle
cells.
Thus,
the
structural
features
neces-
sary
for
neurotransmission
and
spread
of
electrical
signals
from
ICC
to
smooth
muscle
cells
suggest
that
ICC
could
play
a
role
in
reception,
transduction,
and/or
conduction
of
inputs
from
enteric
motor
neurons.
Functional
support
for
the
intercalation
theory
has
been
more
difficult
to
obtain
because
ICC
are
small
cells
integrated
into
complex
tissues.
Although
it
has
been
shown
that
ICC
are
responsive
to
neurotransmitters
(6-9)
and
innervated
by
inhibitory
motor
neurons
(9),
these
studies
have
not
clarified
whether
neurotransmission
depends
upon
the
functional
prop-
erties
of
ICC.
A
parallel
arrangement
could
exist
in
which
ICC
and
smooth
muscle
cells
are
jointly
innervated,
or
neurotrans-
mission
could
be
a
serial
process
in
which
ICC
provide
the
critical
link
between
neurons
and
muscle
cells.
Recently,
it
was
shown
that
mutations
in
c-kit,
a
protoon-
cogene
located
at
the
W
locus
on
chromosome
5
in
the
mouse
that
encodes
a
receptor
tyrosine
kinase
(10),
or
in
stem
cell
factor,
the
natural
ligand
for
c-Kit
receptors,
result
in
devel-
opmental
defects
in
some
classes
of
ICC
(11-14).
For
example,
ICC
in
the
myenteric
plexus
region
(IC-MY)
of
the
small
intestine
are
greatly
decreased
in
numbers
in
these
mutants
and
slow
waves
are
abolished,
confirming
a
role
for
IC-MY
as
pacemakers
(3,
15,
16).
ICC
in
the
region
of
the
deep
muscular
plexus,
however,
were
unaffected
by
defects
in
the
c-Kit
signaling
pathway
(14).
Loss
of
IC-MY
did
not
affect
neuro-
transmission
to
the
circular
muscle,
and
major
excitatory
and
inhibitory
motor
inputs
were
preserved
(11,
14).
These
obser-
vations
suggest
that
either
IC-MY
are
not
essential
for
neu-
rotransmission
in
the
small
bowel
or
a
division
of
labor
exists,
in
which
IC-MY
are
primarily
involved
in
pacemaker
activity,
while
ICC
in
the
region
of
the
deep
muscular
plexus
mediate
neural
inputs.
This
hypothesis
is
difficult
to
test
in
the
small
bowel,
since
ICC
in
the
region
of
the
deep
muscular
plexus
are
difficult
to
remove
without
disrupting
neural
elements.
It
is
possible,
however,
that
defects
in
the
c-Kit
signaling
pathway
may
affect
different
classes
of
ICC
in
other
regions
of
the
GI
tract,
and
some
of
these
ICC
could
be
critically
involved
in
neurotransmission.
In
the
present
study,
we
have
tested
the
involvement
of
ICC
in
neurotransmission
in
the
stomach
using
W/Wv
mutant
mice.
The
results
provide
the
first
functional
support
for
the
intercalation
theory
of
Cajal
(2).
METHODS
Heterozygotes
(W/+)
and
(WV/+)
derived
from
C57BL/6
mice
were
paired
to
obtain
+/+,
W/+,
WV/+,
WIWv,
and
WV/Wv
offspring
(17).
Animals
(20-30
days
post
partum)
were
anesthesized
by
chloroform
inhalation
and
exsanguinated
by
decapitation
following
cervical
dislocation.
The
use
and
treatment
of
animals
was
approved by
the
Institutional
Animal
Use
and
Care
Committee
at
the
University
of
Nevada.
Morphological
Studies.
The
mucosa
was
removed
from
samples
of
gastric
antrum
(greater
curvature)
and
fundus
and
the
muscles
were
fixed
in
acetone
(4°C;
10
min).
After
fixation,
muscles
were
exposed
to
10%
rabbit
serum
to
reduce
nonspe-
cific
antibody
binding.
Tissues
were
incubated
overnight
at
4°C
with
monoclonal
antibodies
to
c-Kit
(ACK2;
5
,ug/ml
in
PBS;
Abbreviations:
GI,
gastrointestinal;
ICC,
interstitial
cells
of
Cajal;
IC-MY,
ICC
in
the
myenteric
plexus
region;
EFS,
electrical
field
stimulation;
L-NAME,
L-N-nitro
arginine
methyl
ester;
VIP,
vasoactive
intestinal
peptide;
c-Kit-LI,
c-Kit-like
immunoreactivity;
IC-IM,
intramuscular
ICC;
NANC,
nonadrenergic,
noncholinergic;
IJP,
inhibitory
junction
potential;
SNP,
sodium
nitroprusside;
NOS,
NO
synthase.
ITo
whom
reprint
requests
should
be
addressed.
12008
The
publication
costs
of
this
article
were
defrayed
in
part
by
page
charge
payment.
This
article
must
therefore
be
hereby
marked
"advertisement"
in
accordance
with
18
U.S.C.
§1734
solely
to
indicate
this
fact.
Proc.
Natl.
Acad.
Sci.
USA
93
(1996)
12009
GIBCO/BRL,
Gaithersburg,
MD).
Immunoreactivity
was
de-
tected
with
fluorescein
isothiocyanate-conjugated
secondary
antibody
(fluorescein
isothiocyanate-anti-rat
1:100
in
PBS).
Controls
were
prepared
in
a
similar
manner,
omitting
ACK2.
Whole
mounts
were
examined
by
confocal
microscopy
(Bio-
Rad
MRC
600).
The
micrographs
presented
are
composites
of
Z-series
scans
(10-15
optical
sections,
20-35
,um
in
depth).
Ultrastructural
studies
were
performed
on
tissues
fixed
with
glutaraldehyde
(2.5%)
and
postfixed
with
OS04
(1%),
as
previously
described
(13).
Ultrathin
sections
were
examined
by
electron
microscopy
(model
CM10;
Phillips
Electronic
Instru-
ments,
Mahwah,
NJ).
For
immunocytochemistry,
tissues
were
fixed
with
paraformaldehyde/acetone
(2:1)
and
processed
for
immunohistochemistry
as
outlined
above.
However,
instead
of
fluorescein
isothiocyanate-labeled
secondary
antibody,
the
diaminobenzidine
technique
was
used
to
demonstrate
specific
antibody
binding
(11).
NADPH-diaphorase
histochemistry
was
performed
on
whole
mounts
fixed
with
paraformaldehyde,
as
described
(9).
Electrophysiological
and
Isometric
Force
Studies.
Strips
of
muscle
from
the
fundus,
corpus
and
antrum
were
prepared
from
wild-type
and
W/WV
animals.
Parallel
platinum
elec-
trodes
were
used
for
electrical
field
stimulation
(EFS).
Smooth
muscle
cells
of
the
circular
layer
were
impaled
with
glass
microelectrodes
to
measure
transmembrane
potential,
and
simultaneous
measurements
of
isometric
force
were
obtained.
EFS
consisted
of
square
pulses
of
current
(0.5
msec,
supra-
maximal
voltage).
Data
are
means
±
standard
errors
of
the
mean,
and
differences
were
evaluated
by
Student's
t
test
(P
<
0.05
was
considered
significant).
Solutions
and
Drugs.
Muscles
were
maintained
in
Krebs-
Ringer
buffer
(KRB)
(37.5
+
0.5oC;
pH
7.3-7.4)
as
described
B
c
+/+
FUNDU'
D
0
s
-44
l10sac
-39
mV
I
-49
10
sec
FIG.
1.
Cells
with
c-Kit-LI
and
electrical
activities
of
the
antrum
and
fundus
of
the
stomach.
(A)
IC-MY
(arrows)
and
IC-IM
(arrow-
heads)
expressed
c-Kit-LI
in
the
antrum
of
wild-type
(+/+)
animals.
Only
IC-IM
were
found
in
the
circular
and
longitudinal
muscle
layers
of
the
fundus
(C;
cells
running
at
90°
to
each
other).
IC-MY
were
normal
in
W/WV
animals
(B),
but
IC-IM
were
absent
throughout
the
stomach
(B
and
D).
Typical
electrical
activities
recorded
from
circular
muscle
cells
of
the
antrum
(see
traces
below
A
and
B)
and
fundus
(below
C
and
D)
from
wild-type
and
W/WV
animals
are
shown.
Slow
waves
were
apparent
in
the
antrum
of
wild-type
and
W/WV
animals.
The
fundus
was
electrically
quiescent.
(11).
Atropine
sulfate,
propranolol
hydrochloride,
tetrodotoxin
(Sigma),
and
phentolamine
mesylate
(CIBA-Geigy)
were
dis-
solved
in
distilled
water
and
diluted
to
10-6
M
in
KRB.
Apamin
(Sigma)
was
diluted
to
(3
x
10-7
M).
L-N-Nitro
arginine
methyl
ester
(L-NAME;
Sigma)
was
diluted
to
2
x
10-4M.
Vasoactive
intestinal
peptide
(VIP)
and
VIP10-28
(Sigma)
were
dissolved
to
10-4M
in
distilled
water
and
further
diluted
to
10-7M
and
10-6M
in
KRB.
Sodium
nitroprusside
(SNP;
Sigma)
was
dissolved
in
distilled
water
(10-3
M),
and
diluted
to
10-8-10-6
M
in
KRB.
Lemakalim
(SmithKline
Beecham)
was
prepared
in
dimethyl
sulfoxide
and
diluted
to
10-6
M.
Oxyhemoglobin
was
prepared
as
described
(18).
RESULTS
ICC
in
the
stomachs
of
wild-type
(+/+)
and
mutant
(WIWv)
mice
were
examined
with
antibodies
to
c-Kit,
a
specific
marker
for
ICC
in
murine
GI
muscles
(11-14).
Two
types
of
cells
expressed
c-Kit-like
immunoreactivity
(c-Kit-LI)
in
wild-type
animals.
One
population,
IC-MY,
formed
a
highly
branching
network
in
the
corpus
and
antrum
in
the
space
between
the
circular
and
longi-
tudinal
muscle
layers
(Fig.
1A-B).
These
cells
were
adjacent
to
the
myenteric
plexus,
and
their
density
increased
distally
along
the
stomach,
as
observed
in
humans
(19).
Stomachs
of
W/WV
animals
contained
similar
numbers
and
distributions
of
IC-MY
as
control
animals
(Fig.
1
A
and
B).
This
contrasts
with
the
small
intestine,
where
IC-MY
were
sparsely
distributed
in
W
mutants
(refs.
11
and
12;
see
Table
1).
A
second
type
of
cell
expressing
c-Kit-LI
was
found
within
the
muscle
layers
of
wild-type
animals
from
the
fundus
through
the
distal
antrum.
These
cells
were
bipolar
with
long
spindle-
shaped
processes
extending
along
the
main
axes
of
the
muscle
fibers
(Fig.
1C).
Electron
microscopy
revealed
dark,
spindle-
shaped
cells
within
the
circular
and
longitudinal
muscle
layers
that
corresponded
to
the
location
and
features
of
the
cells
expressing
c-Kit-LI
(Fig.
2A-C).
These
cells
were
identified
as
ICC
on
ultrastructural
criteria,
including
an
electron
dense
nucleus
with
heterochromatin
distributed
toward
the
periph-
ery
of
the
nuclear
envelope,
a
dense
granular
cytoplasm,
numerous
mitochondria,
an
incomplete
basal
lamina,
and
occasional
caveolae.
Intramuscular
ICC
(IC-IM)
were
closely
associated
with
smooth
muscle
cell
and
nerve
varicosities
(<20
nm)
within
the
muscle
layers
(Fig.
2
D
and
E).
Smooth
muscle
cells
formed
junctions
with
IC-IM.
Immunocytochemistry
confirmed
that
cells
with
c-Kit-LI
were
IC-IM
(Fig.
2F).
These
cells,
which
were
most
abundant
in
the
fundus
of
wild-type
animals,
were
absent
in
WIWv
mice
(Fig.
1D;
Table
1).
Numerous
studies
have
suggested
that
ICC
may
act
as
pace-
makers
in
the
GI
tract
(e.g.,
see
refs.
20-23),
and
loss
of
ICC
in
the
small
intestines
of
W/WV
animals
abolishes
pacemaker
activity
(11,
12).
Generation
of
pacemaker
activity
may
not
be
a
ubiquitous
role
for
all
classes
of
ICC.
Intracellular
electrical
recordings
were
made
from
gastric
muscles
of
wild-type
and
W/WV
animals
to
determine
whether
IC-IM
play
a
role
in
generation
of
electrical
rhythmicity
in
the
stomach.
Cells
within
the
circular
muscle
layer
had
resting
potentials,
averaging
-51
±
2.9
mV
(n
=
5)
and
-64
±
1.2
mV
(n
=
12)
in
the
corpus
and
antrum
of
wild-type
animals
and
-56
±
1.4
mV
(n
=
6)
and
-61
±
1.1
mV
(n
=
10)
in
the
corpus
and
antrum
of
W/WV
animals.
These
values
were
not
significantly
different
in
wild-type
and
W/W-vanimals
(P
>
0.05).
Regular
slow
wave
activity
was
also
recorded
from
corpus
and
antrum
regions
of
wild-type
and
mutant
animals
(Fig.
1
A
and
B).
These
data
suggest
that
IC-IM
have
no
role
as
pacemakers,
and
the
loss
of
the
IC-MY
network
and
absence
of
electrical
rhythmicity
in
the
small
bowel
that
have
been
attributed
to
defects
in
c-Kit
signaling
(11-13)
are
not
generalized
phenomena
throughout
the
GI
tracts
of
these
mutants
(Table
1).
Consistent
with
the
idea
that
IC-IM
are
not
pacemaker
cells,
fundus
muscles
of
wild-type
mice
were
not
electrically
rhythmic
and
had
steady
resting
potentials
averaging
-44.1
±
1.1
mV
(n
=
18).
Loss
of
IC-IM
in
W/WV
animals
had
no
obvious
effect
on
the
Physiology:
Burns
et
aL
Proc.
Natl.
Acad.
Sci.
USA
93
(1996)
Table
1.
Phenotypes
of
wild-type
(+/+)
and
mutant
(W/W")
animals
NO-dependent
IC-MY
IC-IM
IC-DMP
RMP,
mV
Slow
waves
neurotransmission
Gastric
corpus/antrum
(+/+)
++
+
-
-51
±
3/-64
±
1
+
Not
tested
Gastric
fundus
(+/+)
-
+
+ -
-44
±
1
-
X
Gastric
corpus/antrum
(W/WV)
++
-
-
-56
±
1/-61
±
1
+
Not
tested
Gastric
fundus
(W/Wv)
_
_
-
-50
±
1
Small
intestine
(+/+)*
+
+
-
+
+
-68
±
4
+
X
Small
intestine
(W/WV)*
-/+
-
++
-57
±
2
-
Xt
IC-DMP,
ICC
in
the
deep
muscular
plexus;
RMP,
resting
membrane
potential;
X,
present;
-,
absent.
*Data
taken
from
ref.
11;
-/+
means
that
few
of
these
cells
were
observed.
tIC-DMP
may
serve
a
similar
function
in
the
small
bowel
as
that
served
by
IC-IM
in
the
stomach.
basal
electrical
activity
of
the
fundus,
and
these
tissues
had
resting
potentials
averaging
-49.5
±
1.3
mV
(n
=
10;
Fig.
1
C
and
D).
M
_L
''s,~~~~~~~~~~i
E,'
_
X = _
=~~~~~~~~~~~~Am
FIG.
2.
Morphology
of
ICC
of
the
fundus.
Cells
with
c-Kit-LI
had
ultrastructural
features
that
identified
them
as
interstitial
cells.
(A)
An
IC-IM
(ic)
with
an
elongated
electron
dense
nucleus
and
cytoplasm,
numerous
mitochondria,
and
processes
extending
in
the
axis
of
the
circular
muscle
(cm).
IC-IM
made
close
associations
with
smooth
muscle
fibers
(arrowheads),
although
definitive
gap
junction
structures
were
difficult
to
identify.
(B
and
C)
Ultrastructural
details
of
IC-IM.
These
cells
possessed
a
dense
granular
cytoplasm,
numerous
mito-
chondria
(m),
cisternae
of
endoplasmic
reticulum,
Golgi
apparatus,
and
free
ribosomes.
Close
contacts
between
IC-IM
and
smooth
muscle
cells
were
frequently
observed
(arrowhead).
(D
and
E)
The
morpho-
logical
basis
for
the
role
of
IC-IM
in
neurotransmission.
A
close
relationship
was
observed
between
IC-IM
(ic),
nerve
bundles
(nb),
varicose
axons
(a),
and
smooth
muscle
cells
(cm).
Nerve
fibers
and
smooth
muscle
cells
ran
in
the
same
axis
as
IC-IM
and
short
lateral
processes
of
IC-IM
were
closely
associated
with
varicose
axons
(ar-
rowhead)
and
smooth
muscle
(arrows).
(F)
Immunocytochemical
staining
of
IC-IM
in
the
fundus.
Cells
expressing
c-Kit-LI
(arrows
indicate
diaminobenzidine
labeling)
were
observed
within
the
circular
and
longitudinal
muscle
layers
of
the
fundus.
These
cells
possessed
an
electron-dense
nucleus
and
long
processes
extending
within
the
muscle
layers.
An
abundance
of
mitochondria
(m)
was
also
evident.
The
observations
demonstrate
that
cells
with
c-Kit-LI
are
ICC.
IC-IM
might
be
involved
in
neurotransmission
because
regions
of
these
cells
were
found
to
be
<20
nm
from
nerve
varicosities
(Fig.
2
D
and
E).
In
the
presence
of
atropine,
phentolamine,
and
propranolol
[all
10-6
M;
to
demonstrate
nonadrenergic,
noncholinergic
(NANC)
neurotransmission]
EFS
activated
intrinsic
nerves
and
caused
frequency-
dependent
hyperpolarization
of
membrane
potential
[inhibi-
tory
junction
potentials
(IJPs)]
and
relaxation
of
muscle
strips
(Fig.
3).
IJPs
and
mechanical
relaxations
were
reduced by
L-NAME
(2
x
10-4
M;
P
<
0.05
at
all
frequencies;
see
Fig.
3
B
and
E),
suggesting
that
a
significant
portion
of
the
response
to
inhibitory
neurotransmission
depends
upon
synthesis
of
NO,
as
in
other
regions
of
the
GI
tract
(24,
25).
Under
identical
conditions,
stimulation
of
enteric
neurons
in
W/WV
muscles
produced
significantly
attenuated
IJPs
and
little
or
no
relax-
ation
in
tone
(Fig.
3
C
and
F;
P
<
0.05
at
all
frequencies).
L-NAME
had
no
significant
effect
on
IJPs
or
mechanical
responses
in
the
muscles
of
W/W
V
animals
(Fig.
3
D
and
F;
P
>
0.1
at
all
frequencies),
demonstrating
that
NO-dependent
neurotransmission
is
greatly
attenuated
in
animals
lacking
IC-IM.
IJPs
in
wild-type
and
mutant
animals
were
not
further
reduced
by
oxyhemoglobin
(2%)
in
the
presence
of
L-NAME.
VIP
(10-7
M)
hyperpolarized
fundus
muscles
of
wild-type
(11.3
±
1.2
mV;
n
=
12)
and
W/WV
(11.0
±
2.5
mV;
n
=
6)
animals
before
and
in
the
the
presence
of
L-NAME
(2
x
10-4M;
e.g.,
11.5
±
1.2
mV
in
wild-type
(n
=
11)
and
9.8
±
2mVin
W/Wv
animals
with
L-NAME;
n
=
6;
P
>
0.05
when
comparing
VIP
hyperpolarization
before
and
after
L-NAME
in
both
groups).
These
data
suggest
that
the
hyperpolarization
elicited
by
VIP
was
independent
of
NO
synthesis
and
transduction
by
IC-IM.
Tetro-
dotoxin
(10-6M)
did
not
block
VIP-dependent
hyperpolariza-
tions.
The
residual
IJPs
present
after
blocade
of
NO
synthesis
in
control
animals
or
the
IJPs
observed
in
W/Wv
mutants
could
be
due
to
VIP,
but
IJPs
were
not
blocked
by
the
VIP
antagonist
(VIP1o-28)
at
10-6
M.
It
should
also
be
noted
that
hyperpolar-
ization
responses
to
lemakalim
(10-6
M)
were
also
unaffected
in
W/-W4
mutants
[23
±
2.3
mV
in
+/+
animals
(n
=
3)
and
19
±
5
mV
in
W/Wv
mutants
(n
=
4)].
Apamin
(3
x
10-7
M)
reduced
IJPs
in
wild-type
and
mutant
animals
before
and
in
the
presence
of
L-NAME.
The
block
by
apamin
was
frequency-dependent;
IJPs
were
reduced
by
75%
at
5
Hz
and
59%
at
20
Hz
in
wild-type
animals
(n
=
3).
In
W/Wv
mutants
IJPs
were
reduced
by
46%
at
20
Hz
(n
=
2).
Apamin
did
not
block
hyperpolarizations
to
VIP.
To
determine
whether
the
defect
in
NO-dependent
neuro-
transmission
in
W/WV
animals
was
due
to
a
reduction
in
the
numbers
of
nitrergic
nerves,
muscles
were
stained
with
the
NADPH
diaphorase
technique.
This
method
has
been
shown
to
quantitatively
label
cells
containing
neural
NO
synthase
(NOS)
and
to
serve
as
a
marker
of
inhibitory
innervation
in
GI
muscles
(26-28).
There
was
no
difference
in
the
density
or
distribution
of
NADPH
diaphorase-stained
neurons
or
pro-
cesses
within
the
muscle
layers
in
wild-type
and
W/WV
animals
(Fig.
4).
The
density
of
NADPH
diaphorase-positive
nerve
cell
bodies
averaged
26.2
±
1.4
per
mm2
in
wild-type
and
27.6
±
2.4
per
mm2
in
W/Wv
mutants
(n
=
20
fields
from
three
12010
Physiology:
Burns
et
al.
Proc.
Natl.
Acad.
Sci.
USA
93
(1996)
12011
(WANv)
A.
Control
B.
L-NAME
(2x104
M)
C.
Control
Electrical
ll44~
10Hz
-40
1-80
~ ~
Mechanical
5
Hz
30
>
25
a,
20
-o
.L
15
E
10
0L5
0
_/
Iioo00
mg
E.
(+/+)
30
-
I--
>
25-
0
20-
1:315
I
~
~
~
.
0
5
10
15
20,'
z~~
~
~~~~~~~
/
10-
o
5
10
15
20
c
Stimulus
Frequency
(Hz)
F.
(
Stir
D.
L-NAME
(2x104M)
FIG.
3.
Responses
to
NANC
nerve
stim-
ulation
in
muscles
of
wild-type
and
W/WV
animals:
Responses
were
elicited
by
electrical
field
stimnulation
(EFS;
0.5
msec
duration,
1-20
Hz,
1-sec
trains
for
electrical
recordings
and
30-sec
trains
for
mechanical
recordings,
supramaximal
voltage;
dots
at
1
Hz,
black
bars
at
5
and
10
Hz
in
A-D)
and
blocked
by
tetrodotoxin
(3
x
10-7
M).
Control
responses
of
wild-type
animals
(A,
in
NANC
solution)
-44
consisted
of
a
frequency-dependent
hyperpo-
1mV
larization
of
membrane
potential
(IJPs
elic-
ited
by
1,
5,
and
10
Hz;
top
three
traces)
and
sec
relaxation
of
tone
(response
to
5
Hz;
bottom
trace).
Electrical
and
mechanical
responses
shown
are
from
different
muscles.
Little
re-
bound
response
was
observed
after
a
few
pulses
of
EFS
(see
electrical
responses),
but
100
mg
after
cessation
of
long
trains
of
stimuli
(e.g.,
.1.
30
sec
in
mechanical
traces),
rebound
con-
-
|
traction
was
observed.
IJPs
were
greatly
re-
1
min
duced
by
L-NAME
(2
x
10-4
M;
B),
suggest-
ing
that
a
significant
portion
of
these
re-
'W/W)
sponses
depended
upon
NO
synthesis.
The
inhibitory
phase
of
the
contractile
response
was
abolished
by
L-NAME
(bottom
trace
in
B),
and
NANC
contraction
was
observed
during
stimuli.
IJPs
and
NANC
relaxations
were
always
significantly
attenuated
in
W/Wv
--
muscles
(C).
L-NAME
had
little
or
no
effect
on
these
responses
(D),
suggesting
that
NO-
dependent
neurotransmission
was
compro-
mised
in
W/WV
animals.
(E
and
F)
Summa-
l
l
ries
of
frequency-response
relationships
for
5
10
15
20
IJP
amplitudes
in
wild-type
and
WIWv
mus-
cles
before
(-)
and
in
the
presence
of
L-
miulus
Frequency
(Hz)
NAME
(a).
Data
points
are
means
±
SEM.
wild-type
and
three
W/WV
animals;
P
>
0.05).
The
number
of
varicose
nerve
trunks
within
the
circular
muscle
layer
running
B
4
;
-
t~t
-.
C-
.-
*s
-
F
wS!
~
-
;^
W
-}
s~~~~.
....
..
..s-
(arrows)
(B
and
E).
Ahigher
power
micrograph.
..o
d
ugg
Jt
diaphorasepo;sitive
v
ancs
fib
er
(arweas
within
the
ciuca
mus-e
cle.
laye
ScAleH
barsapplhoralefpstanright
panelns
in
teac
rudsofw.W
(A<
ad
++
D-)
aimls
(Aan
D
Th
dstibuio
o
myntri
gagla
otanignuerusNDP
dahoas-pstie
elAode
(aros)
(
ad
)
hghr
owr
icogap
o
idiidalgagla
diphrsepoiiv
nuon
(row),ad
ntranlinc
roese
()
(Cad
Hghpwr
irgrpsshwngdnit.finevtinb
diaphorase-positive
vancose
fibers
(arrowheads)
within
the
ciruclar
mus-~~~~~~~..........
clelaer.Sclebar
apl
fo
lft
ndrilltpael
ineahTow
across
an
arbitrary
100-,um
line
averaged
14.8
±
0.82
and
14.1
±
0.85
(P
>
0.1)
for
wild-type
and
mutant
animals,
respectively.
These
observations
suggest
that
the
number
of
inhibitory
motoneurons
and
the
pattern
of
innervation
of
gastric
muscles
are
not
altered
in
WIWv
mutants.
We
also
tested
whether
postjunctional
responses
to
NO
are
altered
in
WIWv
mutants.
Fourteen
muscle
strips
(from
seven
wild-type
and
seven
WIWv
animals)
were
exposed
to
the
NO
donor,
SNP.
Wild-type
muscles
hyperpolarized
to
SNP
in
a
concentration-dependent
manner
(Fig.
SA),
but
the
hyperpo-
larization
response
of
W/WV
muscles
was
significantly
atten-
uated
[Fig.
SB;
e.g.,
responses
to
SNP
(10-6
M)
averaged
16.6
±
1.4
mV
in
wild-type
and
2.0
±
0.9
mV
in
W/Wvmuscles,
respectively;
P
<
0.0001].
Although
the
electrical
response
to
SNP
was
almost
absent
in
WIW"
animals,
the
fundus
from
wild-type
and
W/WV
animals
relaxed
in
response
to
exogenous
SNP
to
the
same
extent
as
wild-type
muscles
(Fig.
5
C
and
D).
Wild-type
and
W/WV
muscles
relaxed
by
0.57
±
0.04
g
and
0.61
±
0.03
g,
respectively,
to
10-6
M
SNP
(P
>
0.05).
DISCUSSION
The
proximal
stomach
relaxes
to
accommodate
the
increase
in
volume
resulting
from
intake
of
food
(29).
This
reflex,
known
as
adaptive
relaxation
(30),
depends
primarily
upon
the
synthesis
and
release
of
NO
from
NANC
enteric
inhibitory
motoneurons
(31).
The
present
study
suggests
that
adaptive
relaxation
depends
upon
motor
units
composed
of
enteric
inhibitory
motoneurons,
IC-IM,
and
smooth
muscle
cells.
Our
data
suggest
that
either:
(i)
IC-IM
express
the
ion
channels
and/or
second
messenger
system
necessary
for
transducing
NO
signals
into
electrical
responses,
or
(ii)
IC-IM
indirectly
regulate
the
expression
of
the
NO-dependent
effectors
responsible
for
electrical
responses
in
smooth
muscle
cells.
If
the
former
is
true,
then
this
study
provides
the
first
(+/+)
1
Hz
5
Hz
Physiology:
Burns
et
aL
fo
Proc.
Natl.
Acad.
Sci.
USA
93
(1996)
functional
evidence
for
the
intercalation
hypothesis,
proposed
by
Cajal
(2)
and
supported
by
numerous
morphological
studies
in
which
ICC
have
been
shown
to
be
physically
interposed
between
nerve
terminals
and
smooth
muscle
cells
(e.g.,
refs.
5
and
32).
If
either
hypothesis
is
true,
then
the
results
suggest
a
fundamental
role
for
some
classes
of
ICC
in
enteric
neurotransmission.
Based
on
morphology
and
anatomical
location,
there
appear
to
be
several
discrete
classes
of
ICC
in
the
GI
tract.
The
present
report
demonstrates
that
different
classes
of
ICC
have
different
functional
roles
as
well.
ICC
have
been
suggested
as
pacemakers
(3,
33),
and
recent
evidence
supports
this
idea
by
demonstrating
that
some
ICC
express
ionic
conductances
compatible
with
pace-
making
(22).
Further
support
comes
from
studies
on
animals
with
mutations
in
the
c-Kit
signaling
pathway.
IC-MY
of
the
small
intestine
fail
to
develop
in
these
animals,
and
electrical
rhythmicity
is
lost
in
the
small
bowel
(11,
12,
14).
Mutations
in
c-kt
affect
different
types
of
ICC
in
other
regions
of
the
GI
tract.
IC-MY
of
the
colon
were
not
affected
by
Wlocus
(c-Eit)
mutations
(34),
and
the
present
study
shows
that
IC-MY
of
the
stomach
were
also
unaffected
by
these
mutations.
Gastric
rhythmicity,
which
appears
to
depend
upon
IC-MY
(35),
was
also
unaffected.
Anatomical
(e.g.,
refs.
2
and
5)
and
functional
studies
have
also
suggested
that
ICC
could
be
involved
in
neurotransmission
(6-9).
We
found
that
ICC
within
the
circular
and
longitudinal
muscle
layers
(IC-IM)
of
the
stomach
were
absent
in
Wmutants,
and
loss
of
this
class
of
ICC
caused
disruption
in
nitrergic
neurotransmission.
These
data
pro-
vide
the
first
functional
support
of
the
idea
that
some
classes
of
ICC
are
important
mediators
of
enteric
neurotransmission.
NOS
is
constitutively
expressed
by
enteric
inhibitory
motor
neurons
and
distributed
within
cell
bodies
and
the
fine
varicose
axons
that
innervate
the
smooth
muscle
of
the
GI
tract
(26-28).
The
neural
isoform
of
NOS
is
activated
by
Ca2+
(36).
Therefore,
although
nNOS
immunoreactivity
is
observed
throughout
enteric
neurons,
the
enzyme
may
be
activated
only
in
discrete
locations,
such
as
varicosities,
in
which
intracellular
Ca>2
increases
during
activation.
Thus,
release
sites
of
NO
from
neurons
are
likely
to
be
restricted
to
point
sources
in
GI
muscles,
just
as
traditional
transmitters
are
released
from
specialized
sites.
We
found
no
difference
in
the
distribution
of
cells
or
processes
stained
with
NADPH
diaphorase
in
wild-type
and
WlYWv
animals.
NADPH
diaphorase
staining
is
an
indicator
of
NOS
expression
in
enteric
neurons
(27,
28);
therefore,
there
appeared
to
be
no
defect
in
nitrergic
innervation.
In
wild-type
muscles,
SNP
caused
hyperpolarization
of
the
membrane
potential
and
relaxation,
but
in
W/Wv
mutants
lacking
IC-IM,
SNP
relaxed
muscles
but
failed
to
hyperpolarize
these
tissues.
These
data
might
suggest
that
the
relaxation
response
to
nitrosovasodilators
is
unrelated
to
membrane
potential,
but
it
is
also
possible
that
the
significance
of
membrane
potential
is
underestimated
by
bath
application
of
these
compounds.
Several
mechanisms
appear
to
contribute
to
inhibitory
responses
of
smooth
muscles
to
NO.
These
include:
inhibition
of
L-type
Ca
2
current
(37,
38),
activation
of
K+
channels
(18,
39),
reduction
in
intracellular
Ca>2
concentration
by
enhanced
uptake
into
intra-
cellular
stores
(40),
and
reduction
in
the
Ca2+
sensitivity
of
the
contractile
apparatus
(41).
Most
of
these
cellular
mechanisms
are
thought
to
be
mediated
by
cGMP-dependent
mechanisms;
how-
ever,
direct
regulation
of
ion
channels
has
also
been
suggested
(39,
42).
Bathing
muscles
in
solutions
containing
NO
exposes
much
of
the
biochemical
apparatus
that
is
responsive
to
NO.
In
contrast,
release
from
nerve
terminals
may
produce
a
heteroge-
nous
distribution
of
NO
characterized
by
locally
high
concentra-
tions
around
varicosities
and
a
decrease
in
concentration
as
a
function
of
distance
from
release
sites.
The
pattern
of
distribution
of
NO
depends
upon
the
number
of
release
sites
per
unit
volume,
the
rate
of
diffusion
(which
is
fast),
and
the
half-life
of
NO
(which
is
short,
but
poorly
understood)
in
smooth
muscle
tissues.
It
is
possible
that
NO
released
during
nerve
stimulation
reaches
effective
concentrations
within
a
limited
volume
very
close
to
release
sites.
Processes
of
IC-IM,
which
are
closely
associated
with
varicosities,
may
be
exposed
to
relatively
high
concentrations
of
NO,
and
smooth
muscle
cells,
especially
the
parts
of
these
cells
remote
from
varicosities,
may
be
exposed
to
much
lower
con-
centrations.
Thus,
many
of
the
cellular
effectors
responsive
to
NO
may
not
experience
effective
concentrations
during
stimulation
of
enteric
inhibitory
neurons.
Relaxation
responses
to
inhibitory
nerve
stimulation
may
be
mediated
via
electrical
mechanisms
activated
in
IC-IM.
This
concept
is
consistent
with
the
observa-
tion
that
NO-dependent
hyperpolarization
and
relaxation
were
lost
in
WIWv
mutants
that
lacked
IC-IM.
The
loss
of
NO-dependent
LUPs
and
hyperpolarization
re-
sponses
to
SNP
in
W/W'mutants
could
also
be
due
to
a
deficit
in
A.
(+/+)
_~~~~~~~~~.....
-54
I0
Mr
-74
C.
(+/+)
SNP
10M
Electrical
_
,
-0
1
~10
o
...............
.
--40
\
~~~mV
Mechanical
L55
t
0.5g
B.
(WIWv)
SNP
10-M
la'M
ia6M
20
sec
D.
(W/WV)
SNP
la6M
WM-,-
w
WWT-wr
--44m
'
to0.5g
I
min
FIG.
5.
Postjunctional
effects
of
SNP
in
wild-type
(A
and
C)
and
W/WV
muscles
(B
and
D).
All
responses
were
recorded
in
the
presence
of
L-NAME
(2
x
10-4
M)
and
under
NANC
conditions
(see
text).
SNP
(10-8-10-6
M)
caused
concentration-dependent
hyperpolarization
of
membrane
potential
in
wild-type
animals
(A).
Responses
to
SNP
were
greatly
reduced
in
W/WV
muscles
(B);
however,
WIWv
muscles
hyperpolarized
normally
to
lemakalim
(10-5
M),
an
ATP-dependent
K
channel
agonist
and
VIP.
(C
and
D)
Simultaneous
electrical
and
mechanical
recordings.
Hyperpolarization
caused
by
SNP
in
wild-type
animals
was
associated
with
relaxation
of
muscle
tension
(C).
In
WIWV
muscles
hyperpolarization
response
to
SNP
was
attenuated;
however,
the
relaxation
response
was
not
affected
(D).
These
data
suggest
that
the
transduction
mechanism
responsible
for
the
electrical
effects
of
SNP
(and
NO)
may
be
localized
in
IC-IM.
-
-45mV
12012
Physiology:
Burns
et
al.
Physiology:
Burns
et
al.
smooth
muscle
function.
For
example,
either
the
second
messen-
ger
systems
necessary
for
transducing
NO
or
the
specific
K+
channels
activated
could
be
absent
in
these
mutants.
The
fact
that
other
NO-dependent
responses,
such
as
relaxation,
are
preserved
in
W/W'V
mutants
(see
above
paragraph),
suggests
that
second
messenger
systems
needed
to
transduce
NO
signals
are
intact
in
smooth
muscle
cells.
Resting
membrane
potentials
and
responses
to
VIP
and
lemakalim,
two
agonists
that
activate
K+
channels
and
cause
hyperpolarization
of
fundus
muscles,
were
not
significantly
altered
in
W/Wvrmutants.
Therefore,
if
the
loss
of
response
to
SNP
is
due
to
defects
in
smooth
muscle
cells,
then
this
would
require
a
highly
specific
lesion
in
the
complement
of
K
channels
expressed
by
smooth
muscle
cells.
Smooth
muscle
cells
do
not
express
c-kit
in
the
murine
GI
tract.
Therefore,
if
the
W/WV
mutation
affects
the
phenotype
of
smooth
muscle
cells,
this
would
have
to
occur
indirectly,
possibly
by
secretion
of
a
trophic
factor
from
IC-IM.
Although
this
is
possible,
it
would
seem
that
loss
of
a
trophic
factor
in
W/Wv
mutants
would
result
in
a
broader
spectrum
of
pheno-
typic
changes
in
smooth
muscle
cells
than
loss
of
a
specific
conductance
activated
by
NO.
Based
on
this
reasoning,
we
believe
the
most
likely
explanation
for
the
defect
in
NO-dependent
neurotransmission
in
W/WV
mutants
is
the
loss
of
IC-IM.
It
is
not
unreasonable
to
suggest
that
different
ion
channels
or
second
messenger
systems
may
be
expressed
in
ICC
than
in
smooth
muscle
cells.
We
have
observed
at
least
three
ionic
conductances
in
ICC
of
the
canine
colon
that
differ
from
the
complement
of
channels
found
in
neighboring
circular
muscle
cells.
ICC
express
a
low
threshold
Ca2+
conductance,
a
voltage-
dependent
K+
conductance
that
inactivates
at
relatively
negative
potentials
(22),
and
a
K+
conductance
activated
by
lemakalim
(43).
The
latter
suggests,
as
is
apparent
in
the
present
study,
that
some
agonist
responses
may
be
mediated
by
ICC.
It
has
also
been
reported
that
cGMP
levels
increase
in
ICC
in
response
to
NO
(8,
9)
suggesting
that
these
cells
have
readily
available,
active
second
messenger
systems
responsive
to
NO.
Although
neurotransmission
to
smooth
muscle
cells
has
generally
been
characterized
as
an
outpouring
of
transmitter
that
diffuses
throughout
the
smooth
muscle
syncytium
(44),
recent
studies
in
which
autonomic
varicosities
were
serially
sectioned
suggest
there
may
be
greater
structure
to
the
auto-
nomic
neuromuscular
junction
than
previously
thought
(45).
Therefore,
neurotransmission
may
be
far
more
focused
at
"nerve
terminals,"
and
widespread
diffusion
of
transmitter
may
be
less
likely
to
occur.
This
organization
is
supported
by
the
present
study
in
which
NO-dependent
neurotransmission
was
lost
when
IC-IM
were
gone.
If
NO
diffusion
was
wide-
spread,
then
the
effects
of
NO
release
from
nerves
should
have
been
mimicked
by
bath-applied
SNP
(i.e.,
relaxation
should
occur
in
spite
of
the
loss
of
electrical
effects).
Other
forms
of
neurotransmission
remained
in
W/WV
animals,
suggesting
that
the
spatial
influence
of
various
transmitters
may
differ.
In
summary,
a
different
class
of
ICC
is
affected
by
mutations
of
c-kit
in
the
stomach
than
in
the
small
intestine
(11,
12).
IC-MY
provide
pacemaker
input
in
the
small
intestine
(11,
12),
and
the
observation
that
gastric
IC-MY
were
unaffected
in
W/Wv
ani-
mals
is
consistent
with
the
fact
that
normal
electrical
rhythmicity
was
recorded
from
stomachs
of
these
mutants.
IC-IM
are
dis-
tributed
within
the
muscle
layers
and
form
close
associations
with
nerve
fibers
and
smooth
muscle
cells.
Loss
of
these
cells
in
W/WV
animals
was
associated
with
a
loss
of
NO-dependent
inhibitory
nerve
responses.
These
observations
suggest
that
IC-IM
provide
a
critical
link
in
enteric
inhibitory
neurotransmission.
These
cells
may
selectively
express
the
ion
channels
and/or
second
messen-
ger
systems
necessary
to
transduce
NO
signals
into
electrical
responses
in
postjunctional
cells.
Proc.
Natl.
Acad.
Sci.
USA
93
(1996)
12013
The
authors
are
grateful
to
Dr.
William
Shuttleworth
for
helpful
suggestions
with
the
manuscript
and
to
V.
Margaret
Jackson
and
Yulia
Bayguinov
for
technical
assistance.
This
work
was
supported
by
a
National
Institute
of
Diabetes
and
Digestive
and
Kidney
Diseases
Program
Project
Grant
41315
and
DK-40569.
1.
Cajal,
S.
R.
(1893)
C.
R.
Soc.
Biol.
45,
217-223.
2.
Cajal,
S.
R.
(1911)
Histologie
du
Systeme
Nerveux
de
l'Homme
et
des
Vertebres
(Maloine,
Paris),
Vol.
2,
pp.
891-942.
3.
Thuneberg,
L.
(1982)
Adv.
Anat.
Embryol.
Cell
Biol.
71,
1-130.
4.
Taxi,
J.
(1952)
Arch.
Anat.
Microsc.
Morphol.
Exp.
41,
281-304.
5.
Daniel,
E. E.
&
Posey-Daniel,
V.
(1984)
Am.
J.
Physiol.
246,
G305-G315.
6.
Publicover,
N.
G.,
Horowitz,
N.
N.
&
Sanders,
K.
M.
(1992)
Am.
J.
Physiol.
262,
C589-C597.
7.
Publicover,
N.
G.,
Hammond,
E.
M.
&
Sanders,
K.
M.
(1993)
Proc.
Natl.
Acad.
Sci.
USA
9,
2087-2091.
8.
Young,
H.
M.,
McConalogue
K.,
Furness,
J.
B.
&
de
Vente,
J.
(1993)
Neuroscience
55,
583-596.
9.
Shuttleworth,
C.
W.,
Xue,
C.,
Ward,
S.
M.,
de
Vente,
J.
&
Sanders,
K.
M.
(1993)
Neuroscience
56,
513-522.
10.
Chabot,
B.,
Stephenson,
D.
A.,
Chapman,
V.
M.,
Besmer,
P.
&
Bernstein,
A.
(1988)
Nature
(London)
335,
88-89.
11.
Ward,
S.
M.,
Burns,
A.
J.,
Torihashi,
S.
&
Sanders,
K.
M.
(1994)
J.
Physiol.
(London)
48,
91-97.
12.
Huizinga,
J.
D.,
Thuneberg,
L.,
Kluppel,
M.,
Malysz,
J.,
Mikkelsen,
H.
B.
&
Bernstein,
A.
(1995)
Nature
(London)
373,
347-349.
13.
Torihashi,
S.,
Ward,
S.
M.,
Nishikawa,
S.-I.,
Nishi,
K.,
Kobayashi,
S.
&
Sanders,
K.
M.
(1995)
Cell
Tissue
Res.
280,
97-111.
14.
Ward,
S.
M.,
Burns,
A.
J.,
Torihashi,
S.,
Harney,
S.
C.
&
Sanders,
K.
M.
(1995)
Am.
J.
Physiol.
269,
C1577-C1585.
15.
Suzuki,
N.,
Prosser,
C.
L.
&
Dahms,
V.
(1986)
Am.
J.
Physiol.
250,
G287-G294.
16.
Hara,
Y.,
Kubota,
M.
&
Szurszewski,
J.
H.
(1986)
J.
Physiol.
(London)
372,
501-520.
17.
Russell,
E.
S.
(1979)
Adv.
Genet.
20,
357-458.
18.
Thornbury,
K.
D.,
Ward,
S.
M.,
Dalziel,
H.
H.,
Carl,
A.,
Westfall,
D.
P.
&
Sanders,
K.
M..
(1991)
Am.
J.
Physiol.
261,
G553-G557.
19.
Faussone-Pellegrini,
M.
S.,
Pantalone,
D.
&
Cortesini,
C.
(1989)
J.
Sub-
microsc.
Cytol.
Pathol.
21,
439-460.
20.
Thuneberg,
L.,
Johansen,
V.,
Rumenssen,
J.
J.
&
Andersen,
B.
G.
(1983)
Gastrointestinal
Motility,
ed.
Roman,
C.
(MTP,
Lancaster,
U.K.),
pp.
495-502.
21.
Langton,
P.,
Ward,
S.
M.,
Carl,
A.,
Norell,
M.
A.
&
Sanders,
K.
M.
(1989)
Proc.
Natl.
Acad.
Sci.
USA
86,
7280-7284.
22.
Lee,
H.
K.
&
Sanders,
K.
M.
(1993)
J.
Physiol.
(London)
460,
135-152.
23.
Sanders,
K.
M.
(1992)
Annu.
Rev.
Physiol.
54,
439-453.
24.
Bult,
H.,
Boeckxstaens,
G.
E.,
Pelkmans,
P.
A.,
Jordaens,
F.
H.,
Van
Maercke,
Y.
M.
&
Herman,
A.
G.
(1990)
Nature
(London)
345,
346-347.
25.
Dalziel,
H.
H.,
Thornbury,
K.
D.,
Ward,
S.
M.
&
Sanders,
K.
M.
(1991)
Am.
J.
Physiol.
260,
G789-G892.
26.
Belai,
A.,
Schmidt,
H.
H.
H.
W.,
Hoyle,
C.
H.
V.,
Hassall,
C.
J.
S.,
Saffrey,
M.
J.,
Moss,
J.,
Fostermann,
U.,
Murad,
F.
&
Burnstock,
G.
(1992)
Neurosci.
Lett.
143,
60-64.
27.
Ward,
S.
M.,
Xue,
C.,
Shuttleworth,
C.
W.
R.,
Bredt,
D.
S.,
Snyder,
S.
H.
&
Sanders,
K.
M.
(1992)
Am.
J.
Physiol.
263,
G277-G284.
28.
Young,
H.
M.,
Furness,
J.
B.,
Shuttleworth,
C.
W.
R.,
Bredt,
D.
S.
&
Snyder,
S.
H.
(1992)
Histochemistry
97,
375-378.
29.
Paton,
W.
D.
M.
&
Vane,
J.
R.
(1963)
J.
Physiol.
(London)
165,
10-46.
30.
Meyer,
J.
H.
(1987)
in
Physiology
of
the
Gastrointestinal
Tract,
ed.
Johnson,
L.
R.
(Raven,
New
York),
pp.
613-630.
31.
Desai,
K.
M.,
Sessa,
W.
C.
&
Vane,
J.
R.
(1991)
Nature
(London)
351,
477-479.
32.
Imaizumi,
M.
&
Hama,
K.
(1969)
Z.
Zellforsch.
Mikrosk.
Anat.
97,
351-357.
33.
Faussone-Pellegrini,
M.
S.,
Cortesini,
C.
&
Romagnoli,
P.
(1977)Arch.
Ital.
Anat.
Embriol.
82,
157-177.
34.
Burns,
A.
J.,
Torihashi,
S.,
Harney,
S.
C.,
Sanders,
K.
M.
&
Ward,
S.
M.
(1995)
Gastroenterology
108,
A.719
(abstr.).
35.
Prosser,
C.
L.
(1995)
Am.
J.
Physiol.
269,
G386-G395.
36.
Kowles,
R.
G.,
Palacios,
M.,
Palmer,
R.
M.
J.
&
Moncada,
S.
(1989)
Proc.
Natl.
Acad.
Sci.
USA
86,
5159-5162.
37.
Clapp,
L.
H.
&
Gurney,
A.
M.
(1991)
Pfluegers
Arch.
418,
462-470.
38.
Koh,
S.
D.
&
Sanders,
K.
M.
(1996)
Am.
J.
Physiol.,
in
press.
39.
Koh,
S.
D.,
Campbell,
J.
D.,
Carl,
A.
&
Sanders,
K.
M.
(1995)
J.
Physiol.
(London)
489,
735-743.
40.
Cornwell,
T.
L.
&
Lincoln,
T.
M.
(1989)
J.
Biol.
Chem.
264,
1146-1155.
41.
Nishimura,
J.
&
van
Breemen,
C.
(1989)
Biochem.
Biophys.
Res.
Commun.
163,
929-935.
42.
Bolotina,
V.
M.,
Najibi,
S.,
Palacino,
J.
J.,
Pagano,
P.
J.
&
Cohen,
R.
A.
(1994)
Nature
(London)
368,
850-853.
43.
Carl,
A.,
Frey,
B.,
Bayguinov,
0.
&
Sanders,
K.
M.
(1995)
Neurogastroen-
terol.
Motil.
7,
250
(abstr.).
44.
Hoyle,
C.
H.
V.
&
Burnstock,
G.
(1989)
in
Handbook
of
Physiology:
The
Gastrointestinal
System
I.
eds.
Schultz,
S.
G.
&
Wood,
J.
D.
(Am.
Physiol.
Soc.,
Bethesda),
pp
435-464.
45.
Gabella,
G.
(1995)
I.
Neurocytol.
24,
159-187.