(Received for publication, January 29, 1997, and in revised form, April 2, 1997)
From the Center for Biological Research, Neurobiology Unit, Roche Bioscience, Palo Alto, California 94304
Dorsal root ganglion neurons express a wide
repertoire of sodium channels with different properties. Here, we
report the cloning from rat, dorsal root ganglia (DRG), cellular
expression, and functional analysis of a novel tetrodotoxin-sensitive
peripheral sodium channel (PN), PN1. PN1 mRNA is expressed in many
different tissues. Within the rat DRG, both the mRNA and PN1-like
immunoreactivity are present in small and large neurons. The abundance
of sodium channel mRNAs in rat DRG is rBI > PN1 PN3
>>> rBIII by quantitative reverse
transcription-polymerase chain reaction analysis. Data from reverse
transcription-polymerase chain reaction and sequence analyses of human
DRG and other human tissues suggest that rat PN1 is an ortholog of the
human neuroendocrine channel. In Xenopus oocytes, PN1
exhibits kinetics that are similar to rBIIa sodium currents and is
inhibited by tetrodotoxin with an IC50 of 4.3 ± 0.92 nM. Unlike rBIIa, the inactivation kinetics of PN1 are not
accelerated by the coexpression of the
-subunits.
Voltage-gated sodium channels play a critical role in the rising
phase of action potential and are, thus, important for impulse generation and conduction in most excitable cells. Sodium channels are
integral membrane proteins that are usually comprised of one large
-subunit (>200 kDa) and one or more smaller
-subunits (1, 2).
Several
-subunit sodium channel genes have been isolated from
different tissues and functionally analyzed in heterologous expression
systems (3). We and others have previously isolated clones from a
dorsal root ganglia (DRG)1 cDNA library
for a novel tetrodotoxin-resistant sodium channel, PN3/SNS, which is
expressed only in sensory neurons (4, 5). In addition to the PN3
cDNA clones, we have isolated cDNAs for other known and novel
sodium channels (4). One of them, PN1 (peripheral sodium channel 1)
(PN1), is a sodium channel that is regulated by nerve growth factor in
PC12 cells (6, 7). Nerve growth factor increases the mRNA
expression levels of brain type II/IIa and PN1 by two distinct
signaling mechanisms (7). To further understand the molecular basis of
neuronal excitability, we now describe the isolation of full-length
cDNA clones for PN1, quantitation of PN1 mRNA levels as
compared with other sodium channels in DRG, tissue distribution and its
cellular localization in DRG, and functional characterization of this
TTX-sensitive sodium channel in Xenopus oocytes. In
addition, we discuss the isolation of partial clones of the human
ortholog of PN1 (human neuroendocrine channel (hNE)) and the
distribution of hPN1/NE sodium channel by RT-PCR analysis. Our studies
indicate that PN1 is a novel TTX-sensitive sodium channel that is
highly homologous to the rabbit sodium channel Nas (8) and expressed in
a wide variety of tissues, including all cell types within the DRG, and that the human PN1 ortholog is, indeed, the neuroendocrine channel, hNE
(9).
Two rat DRG cDNA libraries, one oligo(dT) and the other random hexamer-primed, were constructed and screened as described (4). The cDNA probe for screening the random-primed cDNA library was a cDNA fragment from domains I and II of an rBIIa clone (10). Overlapping cDNA clones 27.6, 62.5, and 69.1 were assembled in the Xenopus oocyte expression vector, pBSTA (16), at the AvrII/BspEI sites. The overlapping cDNA clones 27.6, 62.5, 63.1, and 69.1, the resulting clones at various stages of assembly, and the final pBSTA-PN1 clone were sequenced on both strands. Sequence analyses were performed as described previously (4).
RT-PCR AnalysisHuman tissues were provided by the
Department of Pathology, Stanford University, Palo Alto, CA. Total RNA
from human adrenal, heart, and brain tissues, and
poly(A)+ RNA from thyroid tissue were purchased from
CLONTECH (Palo Alto, CA). Extraction of total RNA
and synthesis of first strand cDNA from rat and human tissues were
as described earlier (4). Tissue distribution analysis and quantitative
RT-PCR (QRT-PCR) were performed with primer sets within the
3-noncoding region of the gene. The forward primers targeted the same
sequence, whereas the reverse primer for QRT-PCR was more upstream in
the noncoding region. The amplicon sizes for tissue distribution
analysis and the QRT-PCR experiment were 646 and 180 bp, respectively.
Thermal cycler parameters were 30 s at 94 °C, 30 s at 61 °C,
1 min at 72 °C (34 cycles) and 30 s at 94 °C, 30 s at
61 °C, 5 min at 72 °C (1 cycle). For QRT-PCR, 100 ng total RNA
were reverse transcribed to single-stranded cDNA, and the PCR was
performed in the presence of 1 µCi [32P]dCTP and the following
parameters: 30 s at 94 °C, 30 s at 55 °C and 2 min at
72 °C (35 cycles). Specific primers targeting the 3
-noncoding region for PN3, rBI, and rBIII were similarly employed for
QRT-PCR. An aliquot of the reactions was run on a polyacrylamide gel
that was dried prior to quantitation on a PhosphoImager (Bio-Rad). An
external control cRNA comprising the same fragment was constructed for
QRT-PCR as described by Ramakrishnan et al. (11). A standard curve was generated with the external controls for every experiment. 1 ng of pBK-PN1 plasmid DNA was included as positive control for the
tissue distribution analysis by RT-PCR. Primers for
glyceraldehyde-3-phosphate dehydrogenase were employed to demonstrate
tissue viability (4) and to normalize the RNA content in QRT-PCR
experiments.
Nested, degenerate primers from cloned human sodium channel sequences (12-15) were used to amplify a 687-bp fragment spanning domain IVS1 to IVS6. Thermal cycler parameters were: first set of primers, 1 min at 94 °C, 2 min at 50 °C, 4 min at 72 °C (34 cycles) and 1 min at 94 °C, 2 min at 50 °C, 8 min at 72 °C (1 cycle); and for the second set of primers, 1 min at 94 °C, 2 min at 55 °C, 4 min at 72 °C (34 cycles) and 1 min at 94 °C, 2 min at 55 °C, 8 min at 72 °C (1 cycle). Specific primers for interdomain I/II and interdomain II/III based hNE channel sequence (9) defined amplicon sizes of 813 and 352 bp, respectively. Thermal cycler parameters were: 30 s at 94 °C, 30 s at 60 °C, 2 min at 72 °C (35 cycles) and 30 s at 94 °C, 30 s at 60 °C, 7 min at 72 °C (1 cycle). Blots were prepared from the agarose gels that were used to fractionate the PCR fragments and hybridized with a 32P-labeled DNA probe representing the entire coding sequence of the hNE channel (9).
In Situ HybridizationRats (125-150 g, Fisher and Charles
River) were anesthetized with 10% chloral hydrate and perfused with
phosphate-buffered saline (PBS) followed by 4% paraformaldehyde (pH
7.6). DRG (L4-5) were dissected out, cryoprotected by incubation in
PBS-sucrose (10-20%), then frozen and sectioned (10 µm) with a
cryostat. Sections were thawed, digested in proteinase K (1 µg/ml)
for 1 h at 37 °C, dehydrated in ethanol (50-100%), and air
dried. Specific activity of the oligonucleotide probes (targeted for
the 3-untranslated region of PN1 mRNA) was 5 × 107 cpm/µg. Hybridization, washing conditions, and
development of sections were identical to those used for PN3 in
situ hybridization (4). Selected sections were counterstained with
propidium iodide. To perform in situ hybridization using
non-radioactive oligonucleotide probes, the following protocol was
used. After dewaxing/deparaffinization series in xylene and ethanol,
4-µm sections were permeabilized with Triton X-100 (0.3%, 15 min,
RT) and subsequently with HCl (0.2 M), pepsin (0.1%, 10 min, 37 °C), and proteinase K (10 mg/ml, 30 min, 37 °C).
Following the permeabilization, sections were postfixed in 4%
formaldehyde, washed in PBS, and acetylated on a shaking
platform with acetic anhydride (0.25%, in triethanolamine buffer, pH
8). Sections were pre-hybridized for 2 h at 37 °C in a
hybridization solution containing 2 × SSC buffer, 1 × Denhardt's solution, 10% Dextran sulfate, 250 mg/ml yeast tRNA, 0.05 pM/ml of Randomer Oligoprobe (DuPont), 0.1 mg/ml poly(A)
(Boehringer Mannheim), 500 mg/ml of denatured salmon testis DNA, and
50% deionized formamide. Hybridization was done in the same solution,
containing specific antisense or sense oligoprobe, overnight at
37 °C. Following hybridization, samples were washed twice with 2, 1, and 0.25 × SSC, for 15 min at 37 °C each, followed by a rinse in
buffer A (100 mM Tris-HCl (pH 7.5) plus 150 mM
NaCl). Next, samples were incubated in blocking solution (buffer A plus
0.1% Triton X-100 plus 150 mM NaCl, 1 h, RT) and then
in anti-digoxigenin antibodies conjugated with alkaline phospatase (2 h, RT). Samples were subsequently washed in buffer A, incubated in
detection buffer (100 mM Tris-HCl plus 100 mM
NaCl plus 50 mM MgCl2), and then in 200 mM nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl
phosphate plus 10 mM of levamisol in 10 ml of detection
buffer A (Boehringer Mannheim detection kit) overnight in the dark.
Next day, sections were washed with water and mounted with
Krystalon.
Three peptides, two 16-mers and one 14-mer (located in interdomain I/II, II/III and domain IVS1/2), were synthesized according to the multiple antigen peptide technology. Polyclonal antisera were raised in rabbits (Research Genetics, AL) and affinity purified. The antipeptide antibody raised against the 16-amino acid peptide (1035-1050 amino acid in the protein sequence) located in the extracellular loop between domain IVS1 and S2 was employed in the immunocytochemical localization of PN1 (see below).
ImmunocytochemistryMale Sprague-Dawley rats (200-250 g; Harlan, Indianapolis, IN) were anesthetized with 10% chloral hydrate and perfused with PBS followed by 10% formalin. Brain, DRG (L4-5), and lumbar spinal cord were removed and postfixed overnight in 10% formalin. Following fixation, tissue was embedded in paraffin and sectioned at 4 µm. After deparaffinization, sections were preincubated in PBS containing 20% normal goat serum and 0.2% Triton X-100 and incubated with PN1 antibodies (1/1000, overnight, at 4 °C). Tissue sections were washed, incubated in biotinylated anti-rabbit IgG (1/200, RT; Vectastain Elite) followed by peroxidase avidin-biotin complex (1/50, 90 min, RT; ABC, Vectastain Elite), and visualized with a diaminobenzidene reaction. Finally, the tissue sections were washed again, dehydrated, and coverslipped for light microscopy observations. For the control experiment, antibodies were preabsorbed with the corresponding peptide. A Nikon Microphot SA microscope was used for sample observation following both in situ hybridization and immunocytochemistry.
Oocyte ExpressioncRNA was synthesized from PN1 subcloned
into pBSTA (16) using an Ambion mMessage mMachine kit. Defolliculated
Xenopus oocytes (stage V and VI) were prepared (17) and
injected with PN1 cRNA using a Nanojector (Drummond). For two-electrode
voltage clamp (TEVC) recordings, oocytes that had been injected with
30-60 pg of cRNA for 1-2 days previously were perfused with a
solution containing (in mM) 81 NaCl, 2 KCl, 1 MgCl2, 0.3 CaCl2, and 20 HEPES-NaOH (pH 7.5),
impaled with agarose-cushion electodes (0.3-0.8 megohm) (18), and
voltage clamped with a GeneClamp 500 amplifier (Axon Instruments) in
TEVC mode. Oocytes were held at 100 mV and pulsed to different
voltages for 15 ms. For macropatch recordings, oocytes that had been
injected with 10 ng of PN1 cRNA several days previously were
devitellinized (20) after osmotic shrinkage in ND-96 (17, see below)
plus 100 mM NaCl. The bath solution contained (in
mM) 9.6 NaCl, 88 KCl, 1 CaCl2, 1 MgCl2, 11 EGTA, and 5 HEPES-KOH (pH 7.5). The bath
electrode was electrically coupled to the bath via an agarose bridge
(1% agarose in bath solution). 1.2-2.6 megohm patch elecrodes were
pulled from either borosilicate (Garner Glass) or aluminosilicate
(Sutter Instruments) glass on a horizontal puller (Sutter P-87) and
filled with ND-96 (17), which contains (in mM) 96 NaCl, 2 KCl, 1.8 CaCl2, 1 MgCl2, and 5 Hepes-NaOH (pH 7.5). Borosilicate electrodes were coated with dental wax and firepolished before use. Recordings were made from attached
macropatches using a GeneClamp (Axon Instruments) in patch clamp mode
(CV5-1GU headstage). To the extent possible, pipette capacitance was
compensated using the amplifier capacitance compensation. Voltages were
not corrected for the liquid junction potential. Macropatches were held
at
110 mV and were pulsed to different voltages for 9 ms. For both
TEVC and macropatch recordings, stimulation and recording were
controlled by a computer running pClamp 6.0.3 (Axon Instruments). Data
was sampled at 50 kHz after filtering at 5 kHz (TEVC) or 10 kHz
(macropatch) with a 4-pole Bessel filter. For activation data, P/
4
(TEVC) or P/
6 (macropatch) leak subtraction (19) was used. For
macropatch recording, each pulse was repeated 5 times with digital
on-line averaging. Steady-state inactivation curves used test pulses to
+10 mV and were leak subtracted by calculating the passive leak, which
was determined from the holding current at
100 mV and
120 mV after
each test pulse.
The sequence of a clone from the
RT-PCR analysis of rat DRG using degenerate primers, corresponding to
rat sodium channel domain IV, indicated a novel voltage-gated sodium
channel. Subsequently, this was found to be identical to part of the
sequence of clone 27.6, isolated from the oligo(dT)-primed cDNA
library. Nucleotide sequence analysis of the full-length cDNA
assembled from clones 27.6, 62.5, and 69.1 revealed a 5952-base pair
open reading frame coding for a 1984-amino acid protein (Fig.
1). In addition, clone 27.6 had a 3.3-kb 3-noncoding
sequence, and clone 63.1 had a 320-base pair 5
-noncoding upstream
sequence. The assembled clone thus comprised 9.5 kb of the 11-kb band
detected in Northern blots (see below; 6, 7). There was an out-of-frame
ATG upstream of the genuine ATG, akin to other cloned sodium channels
(10). All the hallmark features of the voltage-gated sodium channels
were preserved in the deduced amino acid sequence of PN1, including consensus motifs for protein kinase A. Between segments 5 and 6 in
domain I, there is the aromatic amino acid, Tyr, that confers sensitivity to low concentrations of TTX (see below) (20). PN1 shares
highest homology with the human NE channel and rabbit NAS sodium
channel (93% identity), rBII (79%), rBIII (78%), and rBI (78%).
Indeed, PN1 mapped very close to the brain sodium channels, rBI, rBII,
and rBIII in mouse chromosome 2 (21). A survey of the sodium channel
-subunits in the human DRG was performed by RT-PCR with nested,
degenerate primers from domain IV. The nucleotide sequence of one of
the clones from this region was identical to the hNE channel (data not
shown; 8).
Expression of PN1
Northern analyses with a cDNA probe from rBII revealed an 11-kB band that was intense in DRG and to a lesser extent in brain (data not shown; 6, 7). RT-PCR analyses (Table I) indicated that PN1 mRNA was expressed in a wide variety of rat tissues, except skeletal muscle. The mRNA levels of PN1, PN3, rBI, and rBIII in the adult rat DRG by QRT-PCR are presented in Table II, showing that PN1 and PN3 mRNAs were expressed at comparable levels. rBI mRNA was found to be more abundant than PN1 and PN3, and rBIII was the least abundant of all.
|
|
RNAs from a variety of human tissues were used to perform RT-PCR (Table
I) with specific primers for hNE interdomains I/II and II/III, and the
PCR products were characterized by sequence analysis and Southern blot
hybridization under high stringency conditions. The results in Table I
show that the hNE channel mRNA was expressed in the same tissues as
rat PN1, including adrenal and thyroid tissues as was described by
Klugbauer et al. (9). Neither hNE nor rat PN1 was expressed
in skeletal muscle. The sequences of the PCR fragments spanning
interdomains I/II and II/III amplified from DRG were identical to the
hNE channel. Thus, cloning/detection of sequences spanning many regions
of the coding sequence of hNE in human DRG demonstrated that the NE
channel was expressed in this tissue. However, the disparity in
mRNA sizes of rat PN1 (11 kB) and hNE/PN1 (9.5 kB) (9) will have to
be resolved. It is likely that rat PN1 has a much longer
3-untranslated region (3.3 kb) than hNE. The high sequence homology
shared between rat PN1 and hNE, together with the above observation,
suggests that hNE is the human ortholog of the rat PN1 and rabbit NAS
sodium channels (8).
PN1 mRNA is expressed in small and large DRG neurons, as shown by
radioactive in situ hybridization in Fig.
2A. Fig. 2B illustrates the same
section labeled with propidium iodide to show the general morphological
characteristics and size of neurons. Similar results were obtained with
nonradioactive in situ hybridization (Fig. 2C).
In agreement with the in situ hybridization data,
immunocytochemistry with PN1 antipeptide antibody demonstrated that all
DRG neurons labeled with varying intensity, but no particular pattern
could be detected (Fig. 3A). In control
experiments where the antibody preabsorbed with the peptide was used,
there was hardly any immunoreactivity (Fig. 3B), thereby
demonstrating the specificity of the antibody. In addition, a
widespread distribution of PN1 immunoreactivity was observed in rat
spinal cord and brain (data not shown).
Functional Analysis of PN1
TEVC recordings from
Xenopus oocytes injected with PN1 cRNA (Fig.
4A) indicated that expression of PN1 produced
an inward current that inactivated somewhat slowly compared with
TTX-sensitive sodium currents found in brain, skeletal muscle, and DRG
(22-24). The V1/2 for activation using TEVC was
20 ± 2.0 mV, k =
6.5 ± 0.61 mV
(n = 5). Coexpression of the rat
1-subunit is known to accelerate the inactivation kinetics of the rBIIA, rBIII, and rat
skeletal muscle (rSkM1) sodium channels recorded using TEVC, perhaps by
stabilizing a fast gating mode (data not shown; 25, 26). Surprisingly,
coinjection of PN1 cRNA with 5 ng of r
1 (25) cRNA, 1 ng of r
2
(27) cRNA, or a combination of r
1 and r
2 cRNAs did not appear to
accelerate the inactivation kinetics (data not shown).
Due to the large capacitance of Xenopus ooctyes, the time
resolution of TEVC is limited. To study PN1 currents with more temporal fidelity, we recorded from cell-attached macropatches of oocytes expressing high levels of PN1 (Fig. 4B). In these recordings, the
V1/2 for activation (Fig. 4C) was
found to be 31 ± 3.8 mV, k =
8.9 ± 0.47 mV (n = 4). This is similar to the voltage dependence of activation of rBIIA (28) and hH1 (29) using this recording technique. As expected, with the increased temporal resolution, the
currents activated and inactivated much more rapidly (Fig. 4B). The inactivation was biphasic, with 55-60% of the
current inactivating very rapidly while the rest inactivated with
somewhat slower kinetics. Both phases of inactivation were
voltage-dependent, with half times of 0.46 and 20 ms at
30 mV and 0.1 and 1.8 ms at +10 mV. These results are consistent with the idea that
this channel exhibits more than one gating mode, as has been found for
other sodium channels (26, 28-31).
Analysis of steady-state inactivation using TEVC and 10 s
prepulses (Fig. 4C) yielded a V1/2 of
78 ± 1.1 mV, k = 5.8 ± 0.41 mV
(n = 4). This voltage dependence is very close to that
of hH1 using this protocol (V1/2 =
78 ± 1.6 mV, n = 11) but more negative than that of rBIIA
(V1/2 =
65 ± 1.2 mV,
n = 15).
The PN1 current was blocked by TTX with an IC50 of 4.3 ± 0.92 nM (n = 4; Fig. 3D). This indicates that PN1 is a TTX-sensitive sodium channel, with TTX sensitivity similar to that of the rBIIA (32), rSkM1 (33), rBIII (34, 35), and hNE (8) channels but with much higher sensitivity than the cardiac (36, 37) and PN3/SNS (4, 5) sodium channels.
In summary, PN1 is a novel voltage-gated sodium channel that is
similar, but not identical, to previously characterized sodium channels, such as rBIIa. It is sensitive to low concentrations of TTX.
However, in contrast to rBIIa, when recorded using TEVC, PN1 exhibits
inactivation kinetics that are not accelerated by coexpression of the
-subunits. In DRG, PN1 appears to be expressed at significant levels
that are comparable to those of PN3. Unlike PN3, PN1 is expressed in
all types of neurons in DRG. In view of their high sequence homology,
we believe that rat PN1, hNE, and rabbit NAS are orthologs of the same
sodium channel gene. The wide distribution of PN1/hNE suggests that
they contribute to excitability in several types of neurons and
excitable cells.
We thank C. Yee, P. Zuppan, and C. Bach for sequence analysis and oligonucleotide synthesis and E. Shelton for input on PN1 antisera. We are grateful to A. Tischler for oocyte preparations and recording from rBIIa and hH1. We are also thankful to R. L. Whiting for keen interest and encouragement during the course of this work.
While this publication was in review, the authors became aware of a paper (Toledo-Aral, J. T., Moss, B. L., He, Z., Koszowski, A. G., Whisenand, T., Levinson, S. R., Wolf, J. J., Siols-Santiago, I., Halegoua, S., and Mandel, G. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, 1527-1532) that describes the cloning of an identical gene, PN1 from PC12 cells and its expression in peripheral neurons.