Download Do distinct populations of dorsal root ganglion neurons account for

Am J Physiol Renal Physiol 297: F1427–F1434, 2009.
First published August 19, 2009; doi:10.1152/ajprenal.90599.2008.
Do distinct populations of dorsal root ganglion neurons account
for the sensory peptidergic innervation of the kidney?
Tilmann Ditting,1 Gisa Tiegs,2 Kristina Rodionova,1 Peter W. Reeh,3 Winfried Neuhuber,4
Wolfgang Freisinger,1 and Roland Veelken1
1
Department of Internal Medicine 4, Nephrology, and Hypertension, Departments of 3Physiology and 4Anatomy, University
of Erlangen, Nuremberg; and 2Division of Experimental Immunology and Hepatology, Center of Internal Medicine,
University Medical Center Hamburg-Eppendorf, Hamburg, Germany
Submitted 10 October 2008; accepted in final form 14 August 2009
classification of neurons; capacitance; TRPV1 channels; ASIC; renal
afferent nerve; rat
NERVES CONTAINING NEUROPEPTIDES such as calcitonin gene-related peptide (CGRP) and substance P (SP) are important
components of the sensory nervous system (13, 18). Although
these afferent nerves until recently have been thought to sense
stimuli in the periphery and transmit the information to the
central nervous system, they also have an “efferent” local
vasodilator function (10, 37). Acute administration of a CGRP
receptor antagonist increases blood pressure in various models
of hypertension, which indicates that this potent vasodilator
Address for reprint requests and other correspondence: R. Veelken, Dept. of
Medicine 4, Univ. of Erlangen-Nuremberg, Loschgestrasse 8, 91054 Erlangen,
Germany (e-mail: [email protected]).
http://www.ajprenal.org
plays a counterregulatory role to attenuate hypertension (12).
Furthermore, CGRP was seen as nephroprotective in hypertensive kidney damage in other publications (3, 31). The release of
neuronal peptides like CGRP is putatively dependent on the
stimulation of transient receptor potential vanilloid type 1
(TRPV1) channels (30). But also in other organs, e.g., the liver,
peptidergic afferent nerve fibers were able to release peptides
to influence inflammatory and sclerotic processes (33). Hence,
it is possible that CGRP release may be under the control of
TRPV1 receptors also with respect to peptidergic afferents in
the kidney, so that CGRP released from afferent renal nerve
fibers can exert its nephroprotective potential due to its vasodilatory properties during inflammatory kidney disease, e.g., in
hypertension (3, 31).
Concerning these important studies in animal models of
hypertension pointing to considerable role of afferent sensory
peptidergic nerve fibers and the TRPV1 channels, we are
unfortunately still lacking information about the afferent fibers
and respective neurons involved, their specific characteristics,
and in how far organ specificities with respect to sensitivity,
e.g., in the kidney occur. It is widely accepted that tissue
injury, inflammation, and ischemia are accompanied by local
tissue acidosis (16), and TRPV1 (29) as well as members of the
acid-sensing ion channels (ASIC) family (8, 42) have been
described as putative transducers of peptidergic afferent nerves
to interact with altered proton concentration. Are peptidergic
afferent renal nerves sensible or perhaps very sensible to
stimulation by an increased proton concentration possibly via
TRPV1 receptors or ASIC (30)? Such a finding could suggest
mechanisms to be studied further to putatively explain neurogenic CGRP release being so effective in ameliorating hypertensive kidney damage (3, 31).
One major problem in this respect lies in the fact that
afferent sensory nerve fibers, mainly C-fibers, are extremely
difficult to record (4). Furthermore, recording of C-fibers does
not often allow to investigate distinct receptor mechanisms,
since we are experimentally too much restricted to the thin
fibers and not able to investigate the respective neurons or
receptive endings (9). We recently presented results that
showed in how far cultured sensory neurons from the dorsal
root ganglia (DRG) with axonal projections from hindlimbs
and kidneys allowed for the investigation of changes in mechanosensitivity of these very neurons in various models of
secondary hypertension (19).
Hence, we tested the hypothesis that distinct, clearly characterizable neurons of the DRG with renal peptidergic afferents
are more sensitive to the stimulation by protons possibly via
TRPV1 receptors than neurons with nonrenal afferents from
0363-6127/09 $8.00 Copyright © 2009 the American Physiological Society
F1427
Downloaded from http://ajprenal.physiology.org/ by 10.220.33.1 on June 18, 2017
Ditting T, Tiegs G, Rodionova K, Reeh PW, Neuhuber W,
Freisinger W, Veelken R. Do distinct populations of dorsal root
ganglion neurons account for the sensory peptidergic innervation of
the kidney? Am J Physiol Renal Physiol 297: F1427–F1434, 2009.
First published August 19, 2009; doi:10.1152/ajprenal.90599.2008.—
Peptidergic afferent renal nerves (PARN) have been linked to kidney
damage in hypertension and nephritis. Neither the receptors nor the
signals controlling local release of neurokinines [calcitonin generelated peptide (CGRP) and substance P (SP)] and signal transmission
to the brain are well-understood. We tested the hypothesis that PARN,
compared with nonrenal afferents (Non-RN), are more sensitive to
acidic stimulation via transient receptor potential vanilloid type 1
(TRPV1) channels and exhibit a distinctive firing pattern. PARN were
distinguished from Non-RN by fluorescent labeling (DiI) and studied
by in vitro patch-clamp techniques in dorsal root ganglion neurons
(DRG; T11-L2). Acid-induced currents or firing due to current injection or acidic superfusion were studied in 252 neurons, harvested from
12 Sprague-Dawley rats. PARN showed higher acid-induced currents
than Non-RN (transient: 15.9 ⫾ 5.1 vs. 0.4 ⫾ 0.2* pA/pF at pH 6;
sustained: 20.0 ⫾ 4.5 vs. 6.2 ⫾ 1.2* pA/pF at pH 5; *P ⬍ 0.05). The
TRPV1 antagonist capsazepine inhibited sustained, amiloride-transient currents. Forty-eight percent of PARN were classified as tonic
neurons (TN ⫽ sustained firing during current injection), and 52%
were phasic (PN ⫽ transient firing). Non-RN were rarely tonic (15%),
but more frequently phasic (85%), than PARN (P ⬍ 0.001). TN were
more frequently acid-sensitive than PN (50 –70 vs. 2–20%, P ⬍ 0.01).
Furthermore, renal PN were more frequently acid-sensitive than
nonrenal PN (20 vs. 2%, P ⬍ 0.01). Confocal microscopy revealed
innervation of renal vessels, tubules, and glomeruli by CGRP- and
partly SP-positive fibers coexpressing TRPV1. Our data show that
PARN are represented by a very distinct population of small-tomedium sized DRG neurons exhibiting more frequently tonic firing
and TRPV1-mediated acid sensitivity. These very distinct DRG neurons might play a pivotal role in renal physiology and disease.
F1428
RENAL SENSORY AFFERENTS
METHODS
Male Sprague-Dawley rats (Ivanovas, Kisslegg, Germany) weighing 250 –300 g were maintained in cages at 24 ⫾ 2°C. They were fed
a standard rat diet (no. C-1000, Altromin, Lage, Germany) containing
0.2% sodium by weight and were allowed free access to tap water.
All procedures performed in animals were done in accordance with
the guidelines of the American Physiological Society and approved by
the local government agency (Regierung von Mittelfranken, Ansbach,
Germany).
Labeling of dorsal root ganglion neurons with renal afferents. To
identify neurons with putative renal afferents, these cells were labeled
using the dicarbocyanine dye DiI (1,1⬘ dioleyl-3,3,3⬘ tetramethylindocarbocyanine methansulfonate, ⌬9-DiI, DiI, 50 mg/ml in DMSO;
Molecular Probes, Eugene, OR) by subcapsular application of DiI (5
␮l; 10 mg/ml) in both kidneys 1 wk before harvesting neurons from
DRG T11-L2 (5, 20). Care was taken to prevent back leak or
spreading of DiI to surrounding tissue. To this end, rats were anesthetized with intraperitoneal bolus injection of methohexital (500 ␮l;
25 mg/ml); additional doses were given when appropriate. We allowed 1 wk for DiI to be transported retrogradely to the neuronal cell
bodies residing in the DRG. The subcapsular renal DiI application
could be brought about with minor surgery. Only small flank incisions
through skin and muscles were necessary to expose renal poles.
Analgesic drugs were not necessary and the animals recovered
readily.
Neuronal cell culture. Respective rats were anesthetized with
methohexital as described above, the animals were decapitated, and
DRG from T11-L2 was dissected. Primary cultures of neurons from
the DRG were obtained by mechanical and enzymatic dissociation by
adapting protocols previously described by others (6). The ganglia
were incubated with collagenase (1 mg/ml), trypsin (1 mg/ml), and
DNase (0.1 mg/ml) and resuspended in modified L-15 medium for 1 h
at 37°C that contained 5% rat serum and 2% chick embryo extract as
well as necessary inorganic salts, amino acids, and vitamins (SigmaAldrich, Munich, Germany). Enzymatic activity was terminated by
the addition of soybean trypsin inhibitor (2 mg/ml), bovine serum
albumin (1 mg/ml), and CaCl2 (3 mmol/l) in modified L-15 medium,
and the ganglia were triturated using sterile siliconized Pasteur pipettes to dissociate individual cells. After centrifugation, the cells
were again resuspended in a modified L-15 medium with 5% rat serum
and 2% chick embryo extract and plated on poly-L-lysine-coated glass
AJP-Renal Physiol • VOL
coverslips. 5-Fluorodeoxy-2-uridine (80 ␮mol/l) was added to suppress the proliferation of nonneuronal cells especially fibroblasts. The
cells were cultured on coverslips for 1 day for electrophysiological
experiments again in a modified L-15 solution (20). To demonstrate
that labeled cells were neurons, all cells used for experimental
procedures were tested for fast sodium currents during repolarization
that are characteristic for neuronal cells. A laser beam (532 nm) was
mounted to the patch-clamp recording set-up to allow for the detection
of DiI-stained DRG cells during the experiments using respective
optical filters. From previous experiments, we knew that the nonrenal
neurons in the DRG dissected have to the greatest extent axonal input
from the hindlimbs of the experimental animals (19).
Patch clamp. Patch recordings (19) were obtained from respective
neurons using a recording solution containing (in mM) 104 KCl, 16
KOH, 1 magnesium ATP, and 10 HEPES. Electrode resistance was
3– 6 M⍀, seal resistance was 2–10 G⍀, and series resistance was
⬎100 M⍀. Whole cell voltage-clamp recordings were obtained with
an Axopatch 200B amplifier (Axon Instruments, Foster City, CA).
Data were sampled at 5 kHz for voltage-clamp protocols and 20 kHz
for current-clamp protocols, filtered at 2 kHz, and stored on a
computer hard disk using a commercially available software package
(pClamp 8.2, Axon Instruments). We only included neurons in the
analysis if their resting membrane potentials were below ⫺40 mV. All
experiments described in our study were done in whole cell mode.
Only cells that stained brightly for DiI with laser excitation were
accepted as neurons with renal axons used for the experiments after
harvest and culturing. The experimental protocol was also performed
in neurons with nonrenal afferents likewise located in DRG of
vertebrae T11-L2 to use the results from a neuron population with
afferents of a different anatomical site as comparison, that according
to our previous work represented afferent innervation first and foremost from the animals’ hindlimb as already mentioned above (19). All
recordings were done at room temperature, i.e., 24°C, as described by
others (26, 27).
Voltage-clamp protocols. Acid-sensitive currents were induced by
superfusing the respective neuron under investigation for 10 s with
solutions of either pH 6 or pH 5 once every minute, respectively,
using a multibarrel perfusion pipette that was connected to computercontrolled magnetic valve system controlling the flow of gravity-fed
lines that were connected to reservoirs containing the respective
solutions. Every cell was exposed to both solutions that were administered to the bath in random order; pH was adjusted to 7.4 with NaOH
and to 6 and 5 with HCl. 2-N-morpholinoethansulfonic acid (MES)
was used instead of HEPES in the acidified solutions. The observed
current/time relationship during acidic stimulation of the neurons
showed a characteristic pattern consisting of a transient and a sustained component (see Fig. 1A). The TRPV1 channel inhibitor capsazepine (32) and amiloride (21), an unselective inhibitor of ASIC,
were used to clearly distinguish and characterize TRPV1 and ASICdependent currents, respectively. Currents were adjusted to cell capacitance to express current densities. Whole cell capacitance as an
indicator for cell surface was related to a classification of the cells by
their size as assessed through the microscope using an ocular-mounted
micrometer caliper.
Current-clamp protocols. To allow for a classification of DRG
neurons with renal vs. nonrenal afferent projections (i.e., DiI-positive
or DiI-negative, respectively) in terms of firing patterns, we used an
adaptation of a current-clamp approach recently described by others
(26, 27). Action potentials in response to current injection were
recorded by use of an Axopatch 200B amplifier (Axon Instruments).
The pClamp 8.2 software (Axon Instruments) was used to control
current-pulse generation, to record membrane potentials, as well as for
off-line data analysis. Action potentials were induced by rectangular
current-pulse injections as follows: a 5-ms prepulse, followed by a
600-ms pulse with an interpulse delay of 100 ms was delivered in two
consecutive trains of stepwise increased intensity; 40 – 400 and 400 –
4,000 pA, in 10 consecutive steps, with each step lasting 5.16 s. This
297 • NOVEMBER 2009 •
www.ajprenal.org
Downloaded from http://ajprenal.physiology.org/ by 10.220.33.1 on June 18, 2017
the very same DRG, that according to our previous work
mainly obtain afferent input from the hindlimb vasculature
(19). We used different proton concentrations in the culture
medium as stimulus for TRPV1 since they are easily applied in
cultured neurons and relevant with respect to inflammatory
processes (19, 41). We measured inward currents using a
standard patch-clamp equipment for electrophysiological investigations in the whole cell mode while voltage was clamped
(20). The investigated neurons were grouped according to their
capacitance. These groups corresponded quite well to the
classes of morphologically distinguishable dorsal root ganglion
neurons if sorted by size as described by others (15). To further
substantiate the putative existence of a distinct population of
renal afferent DRG neurons that are exceedingly sensitive to
acidic stimulation, in a separate set of experiments the electrical and acidic excitability was investigated to classify the DRG
neurons in terms of firing patterns as recently described by
others (26, 27). In the latter experiments, current was clamped.
Furthermore, we investigated kidney slices via confocal microscopy to demonstrate that all major parts of the kidney
(glomeruli, tubuli, vessels) were innervated by peptidergic
afferent nerve fibers (33).
RENAL SENSORY AFFERENTS
F1429
RESULTS
Fig. 1. A: amplitudes of acid-induced currents were determined at pH 6 and
pH 5. Peaks of transient currents were assessed at pH 6. The plateau of
sustained currents was calculated at pH 5. B: sustained currents could be
blocked with capsazepine in a dose-dependent manner (⌬), which points to
the expression of transient receptor potential vanilloid type 1 (TRPV1)
channels in renal dorsal root ganglia (DRG) neurons (n ⫽ 8). C: peaks of
transient currents could be blocked with amiloride in a dose-dependent
manner, which points to the expression of acid-sensing ion channels
(ASIC) in renal DRG neurons (n ⫽ 7).
protocol allowed to categorize each DRG neuron as “tonic” or
“phasic” (26, 27): tonic activation means generation of repetitive
action potentials that are sustained at a rather constant rate throughout
the period of depolarizing current injection. In contrast, phasic activation means that only a single or a few action potentials at the
beginning of the period of depolarizing current injection are observed,
even during very high levels of current injection.
After a recovery time of at least 5 min, when the resting membrane
potential returned to its individual baseline value, each cell was
superfused with solutions of either pH 6 or pH 5 once in a minute, for
10 s in randomized order, over a period of 8 min. When action
potentials were generated in response to acidic stimulation, either
sporadic or in regular trains, the respective DRG neuron was categorized as acid sensitive; neurons that remained silent were categorized
as acid resistant.
Immunohistochemistry. For detection of sympathetic and sensory
nerve fibers, immunocytochemistry for tyrosine hydroxylase (TH) and
SP or CGRP, respectively, as well as TRPV1 receptors was utilized
(2, 25, 40). Twelve-micrometer-thick sections from three paraformaldehyde perfusion-fixed rat kidneys (10 sections from each kidney,
spaced 60 ␮m apart) were incubated for single or double immunofluAJP-Renal Physiol • VOL
Labeled cells in situ and in culture. In slices of harvested
DRG, those cells that were positively stained with DiI could be
viewed under epifluorescence with modulation contrast optics
as red-orange cells. A clear difference in the pattern of distribution between neurons with renal or nonrenal afferents could
not be observed, although the neurons with renal afferents, i.e.,
DiI positive, appeared to be found more often closer to the
ventral portion of the ganglia. After 1 day in culture, dorsal
root ganglion cells could be distinguished in most cases from
fibroblast and other cells by their larger soma. A staining with
DiI could be observed in 40% of cultured neurons, suggesting
afferent axonal input from the kidneys. A fraction of these cells
(between 15–20% of all cultured cells) was brightly labeled
with DiI. These brightly labeled cells were defined as renal
afferent DRG neurons and were compared with clearly unlabeled neurons which received afferent input from nonrenal
sites. No significant differences in size could be detected
between labeled and unlabeled cells.
Voltage-clamp protocols. Renal (i.e., DiI labeled) as well as
nonrenal (i.e., unlabeled) DRG neurons mostly showed the
complex pattern of acid-induced transient as well as sustained
currents at pH 6 and pH 5 (Fig. 1A).
The sustained component of the acid-induced inward current
could be blocked by capsazepine in a dose-dependent manner
at pH 6 and pH 5 which points to the existence of TRPV1
receptor channels in these cells (Fig. 1B). No effects of amiloride could be observed on the sustained currents.
Additionally, observed transient currents could be blocked
by amiloride at pH 5 and pH 6 indicating the existence of
ASIC-type channels on the DRG neurons under investigation
(Fig. 1C). The transient currents were not different at either pH
(Fig. 1A). No effects of capsazepine could be observed on
transient currents.
297 • NOVEMBER 2009 •
www.ajprenal.org
Downloaded from http://ajprenal.physiology.org/ by 10.220.33.1 on June 18, 2017
orescence with primary antibodies (sheep anti-TH, Novus Biologicals,
Littleton, CO, 1:2,000; rabbit anti-SP, Peninsula, San Carlos, CA,
1:750; rabbit anti-CGRP, Peninsula, 1:1,000; goat anti-CGRP,
Biotrend, Köln, Germany, 1:250; rabbit anti-TRPV1, Calbiochem/
Merck, Darmstadt, Germany, 1:400) dissolved in Tris-buffered saline
(TBS; pH 7.33) overnight at room temperature. After three rinses in
TBS, appropriate fluorochrome-tagged secondary antibodies (donkey
anti-goat and anti-rabbit IgGs tagged with either Alexa 488 or Alexa
555, 1:1,000 in TBS, all from Molecular Probes) were applied for 1 h
at room temperature. Controls included omission of primary antibodies, replacement of primary antibodies with the respective normal
serum, and incubations with primary antibodies preadsorbed with the
respective antigen. Sections were investigated using a Bio-Rad MRC
1000 confocal system attached to a Nikon Diaphot 300 inverted
microscope. The blue 488-nm and yellow 568-nm lines of a kryptonargon laser were used for excitation of Alexa 488 and Alexa 555,
respectively. The blue 488-nm laser line was also used for eliciting
green autofluorescence of various tissue elements in the kidney and
thus used as counterstain. Merged two-channel confocal images were
adjusted for contrast and brightness using Adobe Photoshop.
Data analyses. The data were statistically analyzed with ANOVA
and Newman-Keuls post hoc test, or t-test (where appropriate) using
a CSS statistical software package (StatSoft, Tulsa, OK). Only a priori
fixed comparisons were tested. Statistical significance was defined as
P ⬍ 0.05. Data are given as means ⫾ SE (28). Furthermore, z-test was
used to test for significant differences in the frequency distribution of
characteristics of the DRG neurons (e.g., tonic, phasic, acid sensitive,
acid resistant).
F1430
RENAL SENSORY AFFERENTS
Fig. 2. A: renal (black bars) and nonrenal neurons (white bars) were divided in
3 groups of cell size based on their membrane capacitance. Medium (80 –160
pF)- and large-sized (⬎160 pF) renal cells exhibited significantly greater
sustained TRPV1 currents than nonrenal neurons (white). *P ⬍ 0.05, renal vs.
nonrenal. B: small neurons (⬍80 pF) showed no ASIC currents. Medium
(80 –160 pF)- and large-sized (⬎160 pF) renal neurons (gray bars) had
transient ASIC currents which were significantly greater than in nonrenal
neurons (white). *P ⬍ 0.05, renal vs. nonrenal.
AJP-Renal Physiol • VOL
innervation, current-clamp experiments were done to further
classify the T11-L2 DRG neurons. Based on DiI labeling, renal
DRG neurons (n ⫽ 96) were compared with nonrenal DRG
neurons (n ⫽ 84). Forty-eight percent of renal neurons could
be classified as “tonic neurons” (TN ⫽ sustained AP generation during depolarizing current injection); 52% were classified
as “phasic neurons” (PN ⫽ transient AP generation during
depolarizing current injection). Nonrenal DRG neurons were
far less tonic (15%), but more often phasic (85%, P ⬍ 0.001).
Furthermore, acid sensitivity was found much more frequently
in TN than in PN (50 –70 vs. 2–20%, P ⬍ 0.01). Moreover,
even renal PN were more frequently acid sensitive than nonrenal PN (20 vs. 2%, P ⬍ 0.01). These findings are illustrated
in Fig. 3. Renal TN tended to be larger than nonrenal TN in
terms of capacitance (92.9 ⫾ 7.5 vs. 56.8 ⫾ 7,2 pF, not
significant); however, this difference failed statistical significance. In acid-sensitive neurons of any group, membrane
depolarization due to the same acid load was much more
pronounced (⌬35.7 ⫾ 2.4 vs. ⌬13.9 ⫾ 1.2 mV, P ⬍ 0.0001).
Thus, the current-clamp approach strongly substantiates the
existence of a functionally distinct group of renal afferent DRG
neurons of medium-to-large size. Baseline membrane parameters from the cells investigated in the current-clamp experiment (renal vs. nonrenal) are given in Table 1. All cells, phasic
and tonic, exhibited phasic responses at low-current injection
levels. This current level needed to induce a phasic response
was the same in renal an nonrenal neurons; however, the
membrane threshold potential for phasic response was significantly more negative in nonrenal cells, and the change in
membrane potential was more pronounced in renal neurons.
Furthermore, the shift in membrane potential due to acidic
stimulation was more pronounced in renal cells (Table 1).
When the data obtained from current-clamp experiments
were grouped by size as above (⬍80, 80 –160, ⬎160 pF), the
frequency distribution of acid sensitivity was significantly
higher in small- and medium-sized renal neurons (Fig. 4). The
difference in membrane potential shift due to acidic stimulation
was similarly distributed between capacitance groups (Fig. 5).
Confocal microscopy. Renal nerves containing bundles of
sympathetic TH-positive and sensory primary afferent axons
specifically stained for CGRP and SP entered the kidney along
the arterial tree. As a rule, SP-positive axons were less frequent
than CGRP-positive fibers. Typically, sympathetic axons
(stained red in Fig. 6, left) and primary afferent CGRP-positive
axons (stained green in Fig. 6, left) were coursing side by side
within the same bundles sometimes resulting in yellow mixed
color due to overlay. The wall of the renal pelvis was densely
innervated by both sensory and sympathetic fibers (data not
shown). Both types of nerve fibers ramified in the renal cortex,
whereas the medulla was devoid of terminal axons. In every
section investigated, all components of the cortex, i.e., glomeruli,
tubules, and blood vessels, were contacted by these nerve fibers
(Fig. 6). Almost all of the CGRP-positive primary afferent
nerve fibers costained for TRPV1 receptor and vice versa
(Fig. 7). These nerve fibers were seen both along arteries as
well as in the interstitium between tubules where they ramified
(Fig. 7, middle). Controls did not show any specific immunostaining. Furthermore, we could show recenty that after renal
denervation all these nervous structures described above were
no longer detectable (35).
297 • NOVEMBER 2009 •
www.ajprenal.org
Downloaded from http://ajprenal.physiology.org/ by 10.220.33.1 on June 18, 2017
Neurons could be readily divided in three groups of capacitance (⬍80, 80 –160, ⬎160 pF) corresponding to small-,
medium-, and large-sized renal and nonrenal neurons (⬍30,
30 – 45, ⬎45 ␮m) as described by others (15).
In the sample of DRG neurons that underwent the voltageclamp protocol (n ⫽ 72), the largest group of cells is represented
in the medium-capacitance, medium-sized group (Fig. 2). The
TRPV1 occurence as assessed by the respective sustained currents induced by proton stimulation, which could readily be
blocked by the TRPV1 antagonist capsazepine, was significantly higher in renal than in nonrenal cells with respect to the
medium- and large-capacitance neurons, but no difference
occurred with respect to the small-capacitance ones (Fig. 2A).
Likewise in medium- and large-capacitance renal neurons,
ASIC seem to be significantly more expressed than in nonrenal
neurons. In small-sized cells with respective capacitance, no
ASIC currents could be observed regardless of the original site
of afferent innervation, i.e., renal or nonrenal, respectively
(Fig. 2B).
Current-clamp protocols. Since the voltage-clamp approach
strongly indicated the existence of a functionally distinct population of DRG neurons that account for the renal afferent
RENAL SENSORY AFFERENTS
F1431
Fig. 3. Relative frequency distribution of cell characteristics from currentclamp experiments is shown. A: firing pattern: tonic (striped bars) and phasic
neurons (blank bars) are just as frequent in the group of renal DRG neurons
(black bars). However, tonic neurons are far less frequent in the group of
nonrenal DRG neurons (white bars). The frequency distribution of both neuron
types is significantly different, when renal and nonrenal DRG neurons are
compared (*P ⬍ 0.001). B: acid excitability: in both groups tonic neurons
more frequently generated action potentials due to acidic stimulation (pH 6 and
pH 5) than phasic neurons (#P ⬍ 0.001). Furthermore, renal phasic neurons
were more frequently sensitive to acidic stimulation than nonrenal phasic
neurons (*P ⬍ 0.01).
kidney are distinctively more sensitive to stimulation with
protons compared with neurons with axons from other sites.
Furthermore, this greater susceptibility of renal afferent neurons to acidic stimulation correlated with a higher rate of tonic
firing response due to depolarizing current injection into respective neurons in culture. Vice versa, nonrenal neurons
DISCUSSION
We could demonstrate for the first time that subgroups of
sensory peptidergic neurons with renal afferents likely involved in renal function and inflammatory processes within the
Table 1. Renal and nonrenal parameters
Parameter
Units
Renal (n ⫽ 96)
Nonrenal (n ⫽ 84)
Revers-Mpot
Cm
Rm
Rest-Mpot
Phasic Thres-pot
Tonic Thres-pot
⌬Mpot phasic
⌬Mpot tonic
Phasic current
Tonic current
⌬Mpot pH 6
⌬Mpot pH 5
mV
pF
M⍀
mV
mV
mV
mV
mV
pA
pA
mV
mV
⫺49.2⫾0.7
94.1⫾5.38
164.7⫾11.7
⫺49.5⫾0.77
⫺28.3⫾1.32
⫺15.0⫾1.14
21.6⫾1.17
33.7⫾1.37
800.0⫾78.5
1,426.1⫾185.7
27.6⫾2.04
35.6⫾2.04
⫺51.2⫾0.6
80.9⫾4.23
163.7⫾18.0
⫺52.0⫾0.79
⫺35.4⫾0.94
⫺17.7⫾1.87
17.0⫾0.7
32.9⫾2.21
805.7⫾77.5
1,332.3⫾318.1
11.6⫾0.79
24.5⫾0.8
Statistics
P
P
P
P
P
P
P
P
P
P
P
P
⫽
⫽
⫽
⬍
⬍
⫽
⬍
⫽
⫽
⫽
⬍
⬍
ns
ns
ns
0.02
0.0001
ns
0.002
ns
ns
ns
0.0001
0.0001
Numbers are given as means ⫾ SE. Revers-Mpot, reversal potential; Cm,
membrane capacitance; Rm, membrane resistance (Cm and Rm measured at
⫺80-mV holding potential, ⫺5-V rectangular test pulse, 5 Hz); Rest-Mpot,
resting potential; Phasic Thres-pot, threshold potential for phasic response;
Tonic Thres-pot, threshold potential for tonic response; ⌬Mpot phasic/⌬Mpot
tonic, difference between resting potential and phasic or tonic threshold
potential, respectively; Phasic current /tonic current, current injected for phasic
or tonic threshold response, respectively; ⌬Mpot pH 6/⌬Mpot pH 5, membrane
potetial shift due to acidic stimulation (pH 5/pH 6); ns, not significant.
AJP-Renal Physiol • VOL
Fig. 5. Change of membrane potential due in renal (gray/black) and nonrenal
(white) neurons divided in the 3 cell size groups, based on their membrane
capacitance, due to acidic stimulation (pH 6 and pH 5). Again, small (⬍80 pF)and medium-sized (80 –160 pF) renal cells were significantly more sensitive to
acidic stimulation in terms of membrane potential change. *P ⬍ 0.01, renal vs.
nonrenal.
297 • NOVEMBER 2009 •
www.ajprenal.org
Downloaded from http://ajprenal.physiology.org/ by 10.220.33.1 on June 18, 2017
Fig. 4. Acid sensitivity obtained from current-clamp experiments in renal
(black bars) and nonrenal neurons (white bars) divided in the 3 groups
according to cell size, based on their membrane capacitance. Small (⬍80 pF)and medium-sized (80 –160 pF) renal cells were significantly more frequently
sensitive to acidic stimulation in terms of action potential generation. *P ⬍
0.01, renal vs. nonrenal.
F1432
RENAL SENSORY AFFERENTS
exhibited far more frequently a phasic firing pattern while
being far less frequently acid sensitive.
Acidic stimulation with protons stimulated predominantly
neuronal TRPV1 receptors on respective neurons but also
ASIC, distinguishable by the use of the respective channel
blockers capsazepine and amiloride.
The idea to categorize peptidergic afferent sensory neurons
in terms of proton susceptibility was based on the following
considerations: a major and unique task of the kidney is to
excrete an amount of acid equal to daily production to maintain
systemic acid-base balance (24). Thus, the existence of some
neuronal acid sensory system suggests itself. Furthermore, the
importance of an acidic milieu in the kidney is likely further
enhanced during inflammatory processes in nephritis and renal
hypertensive disease, when respective afferent nerve fibers
from the kidney might play an important role in the pathology
of these diseases: generally, tissue injury, inflammation, or
ischemia is accompanied by local tissue acidosis (16), and
TRPV1 (29) and ASIC (8, 42) have been described as putative
transducers on peptidergic afferent nerves. Furthermore, although being less specific proton stimulation does not show
tachyphylaxis and desensitization as the specific TRPV1 ligand
capsaicin.
The neurons investigated were not morphologically homogenous, but could be classified, as described before (15, 36),
according to cell size into three groups, that approximately
correlated with neuronal capacities measured in our experiments. The mentioned studies on which we based this “size
classification” could show direct correlation of neuronal soma
size with peripheral nerve conduction velocity in L4 DRG:
summarized, small- and medium-sized (mainly B type or small
dark) neurons mostly give rise to unmyelinated C and thin
myelinated A␦-fibers, whereas large (mainly A type or large
light) neurons give rise to A␣ and A␤ fibers (15). There is,
however, some overlap in cell size, mostly in A␦ fibers. In
general, sensory neurons innervating inner organs are composed mainly of the peptidergic subpopulation of B-type primary afferent neurons (22, 35).
Fig. 7. Three examples of CGRP-containing (red) primary
afferent nerve fiber bundles costaining for the TRPV1 receptor
(green) within the rat renal cortex in the adventitia of small
arteries (A; right and left) and between tubules (T; middle).
The axons in the middle are varicose and branching, indicative
of terminals. Almost all CGRP-positive axons also contain
TRPV1 immunoreactivity and vice versa as shown by orangeyellow mixed color. Bars are 25 ␮m.
AJP-Renal Physiol • VOL
297 • NOVEMBER 2009 •
www.ajprenal.org
Downloaded from http://ajprenal.physiology.org/ by 10.220.33.1 on June 18, 2017
Fig. 6. Left: nerve fiber bundle in the rat renal
cortex containing both calcitonin gene-related
peptide (CGRP)-positive afferent (green) and
tyrosine hydroxylase (TH)-positive sympathetic efferent (red) nerve fibers. Close proximity of both fiber types results in yellow
mixed color. Note fine varicosities adjacent to
both artery (A) and renal tubules (T). All-infocus projection of a stack of 6 1-␮m-thick
confocal optical sections. Right: substance P
(SP)-positive afferent nerve fibers (red) accompany the afferent arteriole (aa) with one
terminal branch entering the glomerulus (G).
SP-containing afferent nerve fibers were less
frequent than CGRP-containing nerve fibers.
Green autofluorescence of parenchyma, in
particular proximal tubules, is used as a counterstain. Single 1-␮m-thick confocal optical
section. Bars are 50 ␮m in both panels.
RENAL SENSORY AFFERENTS
AJP-Renal Physiol • VOL
functional investigation of these intrarenal fibers is as yet not
possible.
We used different proton concentrations as a general stimulus to induce neuronal inward currents dependent on TRPV1
receptors (29, 41). However, the fact that not only the occurrence of TRPV1 but also of ASIC was specifically increased in
medium-capacitance neurons with renal afferents points to the
possibility that protons may be of greater importance for the
stimulation of peptidergic renal afferent nerve fibers.
Our morphological studies revealed that afferent peptidergic
nerve fibers were mostly in close vicinity to sympathetic
efferent nerve fibers. It is known that biogenic amines required
for neurotransmission are transported into respective storage
vesicles utilizing an electrochemical gradient across the vesicular membrane established by proton pumping into the vesicle
via a vacuolar ATPase (11). Hence, the vesicles having a
membrane impermeable to protons must contain a considerable
amount of protons to be able to accumulate singly positively
charged amines into the relatively proton-impermeable acidic
secretory vesicles at the expense of proton antiport through a
transporter protein (23). This means that release of biogenic
amines from the vesicles will increase locally the proton
concentration that could stimulate TRPV1 and ASIC receptors
on afferent nerve fibers in the close vicinity. The putative
relevance of such a mechanism in health and disease needs
further evaluation.
In a recent study, renal afferent innervation was linked to the
amelioration of the consequences of oxidative stress (38). In
rats with destruction of afferent sensory nerves due to early
postnatal capsaicin treatment, enhanced O⫺
2 could be measured
in the cortex and medulla of the kidney if these animals were
on a high-salt diet. Increased superoxide levels in these animals
were attributed to increased expression and activity of
NAD(P)H oxidase, which was thought to contribute to decreased renal function and increased blood pressure. Since
oxidative stress is connected with decreased vasodilatation and
reduced tissue perfusion (7), altered regional proton concentration in the kidney under these circumstances could stimulate
TRPV1 receptors to promote vasodilatory CGRP release.
It is known that protease-activated receptor 2 (PAR2) sensitizes TRPV1 receptors (1). PAR2 expression was observed in
renal diseases like IgA nephropathy potentially aggravating the
pathogenesis of interstitial fibrosis (14, 39). Furthermore, since
TRPV1 receptors are eventually stimulated not only by protons
or sensitized by PAR2 but affected by a large variety of
substances and metabolic products potentially involved in
inflammatory responses (17), it is possible that also a wide
variety of different pathophysiological circumstances with inflammation and progressive fibrosis will stimulate these receptors. Since we are now able to classify and investigate specific
afferent sensory peptidergic neurons with projections to important cardiovascular organs like the kidney and have enough
evidence that afferent peptidergic nerve fibers play a role in the
pathophysiology of inflammation, it is now important to clarify
the specific mechanisms in target organ damage and inflammation that lead to altered expression and release of neuronal
peptides like CGRP. Specifically, it will be challenging to
figure out which of the various possibilities of TRPV1 receptor
stimulation is of actual importance in this context to eventually
develop respective therapeutic strategies beneficially influencing these neurogenic systems involved in hypertension.
297 • NOVEMBER 2009 •
www.ajprenal.org
Downloaded from http://ajprenal.physiology.org/ by 10.220.33.1 on June 18, 2017
The group of neurons with medium size and capacity likely
to be polymodal and hence responsible for mechano- and
chemoception (15, 34) is thus probably linked mainly to C
fibers. In this group, renal neurons were not only more sensitive to acidic stimulation in terms of action potential generation
than neurons with projections from other, nonrenal sites but
showed also higher membrane currents after acidic receptor
stimulations with protons, which was the case for TRPV1- as
well as for ASIC-mediated currents, as shown by use of the
respective channel blockers capsazepine and amiloride. Higher
membrane currents upon stimulation with protons might suggest a generally higher responsiveness of cells also with respect
to other cellular processes than action potential generation.
Such processes may also encompass the release of CGRP and
SP from axonal nerve endings within the kidney already
described often for these neurons.
However, this question must be addressed in further specific
experiments for the kidney. Such studies will then also have to
take into account the recent finding in DRG neurons (from
segments L4-S3) that neurokinines, like SP (26, 27), are able to
enhance the excitability of capsaicin-responsive (i.e., TRPV1
expressing) neurons via neurokinin receptor 2 (NK2) when
cells were stimulated by current injection or capsaicin. It is
likely that such a positive feedback system plays a significant
role in the kidney in health and disease. It remains, however, to
be determined whether such feedback mechanisms are detectable in proton stimulation.
The group of neurons with small size and low capacity is
often seen as involved into noci- and thermoception (15).
Renal cells in that group were also more sensitive to acidic
stimulation with respect to action potential generation. Since
these modalities are poorly defined for the parenchymal part of
the kidney, the putative role of these afferent renal neurons is
somewhat elusive. They may serve comparable regulatory
purposes as renal medium-sized neurons, but the fact that
small-sized neurons with renal axons were not unique when it
comes to membrane currents as compared with small-sized
neurons with nonrenal afferents suggests that these neurons
cannot be easily compared with neuronal cells of medium size.
Of note is that these cells did not exhibit ASIC-mediated
currents.
The group of renal neurons with large size, high capacity,
and putatively high conduction velocity is rather likely to
innervate the renal capsule than subserving any role in renal
cortex or medulla. Conceivable is also nociception from the
renal pelvis, if we take into account what is known from
neurons with axons exhibiting the modalities of A␣ and A␤
fibers (36).
Our studies using confocal microscopy revealed that all
major functional elements of the kidney like glomeruli, tubuli,
and vessels were innervated with afferent nerve fibers containing CGRP and to a lesser extent SP, which is in accordance
with our previous findings (35). One could argue that the nerve
fibers within the kidney will not always respond like the
respective neuronal bodies in the DRG to any kind of stimulation. However, we detected TRPV1 receptors colocalized on
CGRP-containing nerve fibers within the kidney. Hence, the
functionally relevant receptors are expressed not only on the
neurons as suggested by our patch-clamp experiments but can
also be found on the respective intrarenal nerve fibers. A direct
F1433
F1434
RENAL SENSORY AFFERENTS
ACKNOWLEDGMENTS
The technical assistance of S. Heinlein, A. Hecht, K. Löschner, and H.
Symowski is gratefully acknowledged.
GRANTS
This work was supported by a grant-in-aid from the Deutsche Forschungsgemeinschaft (SFB 423).
REFERENCES
AJP-Renal Physiol • VOL
297 • NOVEMBER 2009 •
www.ajprenal.org
Downloaded from http://ajprenal.physiology.org/ by 10.220.33.1 on June 18, 2017
1. Amadesi S, Nie J, Vergnolle N, Cottrell GS, Grady EF, Trevisani M,
Manni C, Geppetti P, McRoberts JA, Ennes H, Davis JB, Mayer EA,
Bunnett NW. Protease-activated receptor 2 sensitizes the capsaicin receptor transient receptor potential vanilloid receptor 1 to induce hyperalgesia. J Neurosci 24: 4300 – 4312, 2004.
2. Bergua A, Schrodl F, Neuhuber WL. Vasoactive intestinal and calcitonin gene-related peptides, tyrosine hydroxylase and nitrergic markers in
the innervation of the rat central retinal artery. Exp Eye Res 77: 367–374,
2003.
3. Bowers MC, Katki KA, Rao A, Koehler M, Patel P, Spiekerman A,
DiPette DJ, Supowit SC. Role of calcitonin gene-related peptide in
hypertension-induced renal damage. Hypertension 46: 51–57, 2005.
4. Coleridge HM, Coleridge JC, Kidd C. Cardiac receptors in the dog with
particular reference to two types of afferent ending in the ventricular wall.
J Physiol 174: 323–339, 1964.
5. Cunningham JT, Wachtel RE, Abboud FM. Mechanical stimulation of
neurites generates an inward current in putative aortic baroreceptor neurons in vitro. Brain Res 757: 149 –154, 1997.
6. Cunningham JT, Wachtel RE, Abboud FM. Mechanosensitive currents
in putative aortic baroreceptor neurons in vitro. J Neurophysiol 73:
2094 –2098, 1995.
7. Davignon J, Ganz P. Role of endothelial dysfunction in atherosclerosis.
Circulation 109: III27–III32, 2004.
8. Deval E, Noel J, Lay N, Alloui A, Diochot S, Friend V, Jodar M,
Lazdunski M, Lingueglia E. ASIC3, a sensor of acidic and primary
inflammatory pain. EMBO J 27: 3047–3055, 2008.
9. Ditting T, Hilgers KF, Scrogin KE, Stetter A, Linz P, Veelken R.
Mechanosensitive cardiac C-fiber response to changes in left ventricular
filling, coronary perfusion pressure, hemorrhage, and volume expansion in
rats. Am J Physiol Heart Circ Physiol 288: H541–H552, 2005.
10. Ditting T, Veelken R, Hilgers KF. Transient receptor potential vanilloid
type 1 receptors in hypertensive renal damage: a promising therapeutic
target? Hypertension 52: 213–214, 2008.
11. Eiden LE, Schafer MK, Weihe E, Schutz B. The vesicular amine
transporter family (SLC18): amine/proton antiporters required for vesicular accumulation and regulated exocytotic secretion of monoamines and
acetylcholine. Pflügers Arch 447: 636 – 640, 2004.
12. Gangula PR, Lanlua P, Wimalawansa S, Supowit S, DiPette D,
Yallampalli C. Regulation of calcitonin gene-related peptide expression
in dorsal root ganglia of rats by female sex steroid hormones. Biol Reprod
62: 1033–1039, 2000.
13. Gontijo JR, Smith LA, Kopp UC. CGRP activates renal pelvic substance
P receptors by retarding substance P metabolism. Hypertension 33: 493–
498, 1999.
14. Grandaliano G, Pontrelli P, Cerullo G, Monno R, Ranieri E, Ursi M,
Loverre A, Gesualdo L, Schena FP. Protease-activated receptor-2 expression in IgA nephropathy: a potential role in the pathogenesis of
interstitial fibrosis. J Am Soc Nephrol 14: 2072–2083, 2003.
15. Harper AA, Lawson SN. Conduction velocity is related to morphological
cell type in rat dorsal root ganglion neurones. J Physiol 359: 31– 46, 1985.
16. Huang CW, Tzeng JN, Chen YJ, Tsai WF, Chen CC, Sun WH.
Nociceptors of dorsal root ganglion express proton-sensing G proteincoupled receptors. Mol Cell Neurosci 36: 195–210, 2007.
17. Inoue R, Jensen LJ, Shi J, Morita H, Nishida M, Honda A, Ito Y.
Transient receptor potential channels in cardiovascular function and disease. Circ Res 99: 119 –131, 2006.
18. Kopp UC, Cicha MZ, Smith LA. Endogenous angiotensin modulates
PGE2-mediated release of substance P from renal mechanosensory nerve
fibers. Am J Physiol Regul Integr Comp Physiol 282: R19 –R30, 2002.
19. Linz P, Amann K, Freisinger W, Ditting T, Hilgers KF, Veelken R.
Sensory neurons with afferents from hind limbs: enhanced sensitivity in
secondary hypertension. Hypertension 47: 527–531, 2006.
20. Linz P, Veelken R. Serotonin 5-HT3 receptors on mechanosensitive
neurons with cardiac afferents. Am J Physiol Heart Circ Physiol 282:
H1828 –H1835, 2002.
21. Mercado F, Lopez IA, Acuna D, Vega R, Soto E. Acid-sensing ionic
channels in the rat vestibular endorgans and ganglia. J Neurophysiol 96:
1615–1624, 2006.
22. Millan MJ. The induction of pain: an integrative review. Prog Neurobiol
57: 1–164, 1999.
23. Parsons SM. Transport mechanisms in acetylcholine and monoamine
storage. FASEB J 14: 2423–2434, 2000.
24. Preisig PA. The acid-activated signaling pathway: starting with Pyk2 and
ending with increased NHE3 activity. Kidney Int 72: 1324 –1329, 2007.
25. Schrodl F, Brehmer A, Neuhuber WL, Nickla D. The autonomic facial
nerve pathway in birds: a tracing study in chickens. Invest Ophthalmol Vis
Sci 47: 3225–3233, 2006.
26. Sculptoreanu A, Aura Kullmann F, de Groat WC. Neurokinin 2
receptor-mediated activation of protein kinase C modulates capsaicin
responses in DRG neurons from adult rats. Eur J Neurosci 27: 3171–3181,
2008.
27. Sculptoreanu A, de Groat WC. Neurokinins enhance excitability in
capsaicin-responsive DRG neurons. Exp Neurol 205: 92–100, 2007.
28. Snedecor G, Cochran W. Statistical Methods. Ames: Iowa Univ. Press,
1972.
29. St Pierre M, Reeh PW, Zimmermann K. Differential effects of TRPV
channel block on polymodal activation of rat cutaneous nociceptors
in vitro. Exp Brain Res 196: 31– 44, 2009.
30. Strecker T, Messlinger K, Weyand M, Reeh PW. Role of different
proton-sensitive channels in releasing calcitonin gene-related peptide from
isolated hearts of mutant mice. Cardiovasc Res 65: 405– 410, 2005.
31. Supowit SC, Rao A, Bowers MC, Zhao H, Fink G, Steficek B, Patel P,
Katki KA, Dipette DJ. Calcitonin gene-related peptide protects against
hypertension-induced heart and kidney damage. Hypertension 45: 109 –
114, 2005.
32. Szallasi A. The vanilloid (capsaicin) receptor: receptor types and species
differences. Gen Pharmacol 25: 223–243, 1994.
33. Tiegs G, Bang R, Neuhuber WL. Requirement of peptidergic sensory
innervation for disease activity in murine models of immune hepatitis and
protection by beta-adrenergic stimulation. J Neuroimmunol 96: 131–143,
1999.
34. Veelken R, Stetter A, Dickel T, Hilgers KF. Bimodality of cardiac vagal
afferent C-fibres in the rat. Pflügers Arch 446: 516 –522, 2003.
35. Veelken R, Vogel EM, Hilgers K, Amann K, Hartner A, Sass G,
Neuhuber W, Tiegs G. Autonomic renal denervation ameliorates experimental glomerulonephritis. J Am Soc Nephrol 19: 1371–1378, 2008.
36. Villiere V, McLachlan EM. Electrophysiological properties of neurons in
intact rat dorsal root ganglia classified by conduction velocity and action
potential duration. J Neurophysiol 76: 1924 –1941, 1996.
37. Wang DH. The vanilloid receptor and hypertension. Acta Pharmacol Sin
26: 286 –294, 2005.
38. Wang Y, Chen AF, Wang DH. Enhanced oxidative stress in kidneys of
salt-sensitive hypertension: role of sensory nerves. Am J Physiol Heart
Circ Physiol 291: H3136 –H3143, 2006.
39. Wang Y, Pratt JR, Hartley B, Evans B, Zhang L, Sacks SH. Expression
of tissue type plasminogen activator and type 1 plasminogen activator
inhibitor, and persistent fibrin deposition in chronic renal allograft failure.
Kidney Int 52: 371–377, 1997.
40. Worl J, Mayer B, Neuhuber WL. Spatial relationships of enteric nerve
fibers to vagal motor terminals and the sarcolemma in motor endplates of
the rat esophagus: a confocal laser scanning and electron microscopic
study. Cell Tissue Res 287: 113–118, 1997.
41. Yao X, Garland CJ. Recent developments in vascular endothelial cell
transient receptor potential channels. Circ Res 97: 853– 863, 2005.
42. Yen YT, Tu PH, Chen CJ, Lin YW, Hsieh ST, Chen CC. Role of
acid-sensing ion channel 3 in sub acute phase inflammation. Mol Pain 5:
1, 2009.