 |
INTRODUCTION |
It is widely accepted that Ca2+ serves as an essential
signaling molecule, and for that reason, the level of intracellular
calcium ([Ca2+]i) is strictly controlled. This
control extends beyond regulating static levels of Ca2+ to
include regulation of Ca2+ levels in subregions of the cell
and regulation of the frequency and amplitude of Ca2+
oscillations within the cell. This multifaceted regulation is required
for countless cellular functions including muscle contraction, protein
secretion, cell proliferation, and apoptosis (1). Mutations inducing
drastic alterations in intracellular Ca2+ homeostasis are
most likely not compatible with life (2).
In most cells, there are two sources of Ca2+ that can be
tapped in order to modify resting Ca2+ levels, the
intracellular Ca2+ stores and the extracellular space.
Intracellularly, Ca2+ is released from the sarco- and
endoplasmic reticulum stores via two types of Ca2+
channels: inositol 1,4,5-trisphosphate receptors and ryanodine receptors. Cytosolic Ca2+ is taken up into the sarco- and
endoplasmic reticulum by one of the three sarco- or endoplasmic
reticulum Ca2+-ATPase Ca2+-pumps. Extracellular
Ca2+ can enter the cell via cation channels in the plasma
membrane (1), and in nonexcitable cells it is well documented that
depletion of internal Ca2+ stores leads to the activation
of store-operated channels
(SOCs)1 in the plasma
membrane. This allows Ca2+ to enter the cell on a continual
basis during times of elevated inositol 1,4,5-trisphosphate levels,
resulting in sustained elevations of cytosolic Ca2+ levels.
When inositol 1,4,5-trisphosphate levels decline, the Ca2+
entry persists only until the internal Ca2+ stores are
refilled, at which time the SOCs turn off (3). The net result is that,
at high doses of agonist, one normally sees a biphasic Ca2+
response made up of a high transient peak representing release of
internal Ca2+ stores followed by a long sustained plateau
phase that results from Ca2+ entry.
Sometimes changes in [Ca2+]i in response to
receptor agonists can be far more complex. The signal patterns can
include repetitive oscillations or waves of elevated
[Ca2+]i. The functional importance of
Ca2+ oscillations may relate to regulation of kinases and
phosphatases, energy metabolism, secretion, and certain transcriptional
events (1, 4). Oscillation of Ca2+ often occurs at low
concentrations of agonist that generates levels of InsP3,
which release submaximal amounts of Ca2+ but sufficient
amounts to trigger the positive feedback effect of Ca2+ on
the release of store Ca2+ via inositol 1,4,5-trisphosphate
receptors. In many instances Ca2+ influx is required in
order to sustain [Ca2+]i oscillations (4-7).
In the Drosophila eye, photoactivation of rhodopsin leads to
the opening of the light-sensitive cation influx channels Trp (transient receptor potential) and
Trpl (Trp-like) (8). When expressed in a
heterologous cell system, Trp channels can be activated by depletion of
Ca2+ from the internal stores (9). Therefore, SOCs may be
functionally related to Trp proteins. To date, seven mammalian Trp
homologs (short isoforms) have been cloned, and they vary significantly in their channel behavior and mode of activation (10). Human Trps are
found at highest levels in brain, heart, testis, and ovary; they are
~450 amino acids shorter than Drosophila Trp; they have
six-transmembrane domain regions; and regions in the N-terminal domain
have similarity to domains found in ankyrin and in the C terminus
similar to dystrophin (11, 12). Although, when expressed in a
heterologous cell system, the Drosophila Trp channel appears
to be store-operated, there is considerable evidence to suggest that
Trp is regulated by a different mechanism during the phototransduction
process (13). A similar controversy has arisen from numerous studies
where various mammalian Trp homologs have been overexpressed in various
cell systems; results suggest that mammalian Trps either may (14-18)
or may not (19-21) respond to store depletion. For example, if we
focus on the overexpression studies for the Trp4 isoform, there are
significant disagreements concerning whether the channel is regulated
by store depletion. The initial reports on bovine Trp4 described it as
a capacitative calcium channel, since its overexpression in HEK-293
cells (16) or Chinese hamster ovary cells (22) led to channel activity that could be stimulated by depletion of Ca2+ stores with
thapsigargin. A role for rat Trp4 in capacitative calcium entry was
suggested by the demonstration of a potentiated capacitative
calcium entry-mediated chloride current in oocytes expressing rat Trp4
(23) and was further supported by the observation that expression of a
Trp4 antisense significantly reduced calcium release activated
channel-like currents in adrenal cells (24). However, two recent
studies argue that Trp4 does not participate in forming store-operated
calcium channels. The expression of murine Trp4 in HEK-293 cells was
reported to have no effect on channel activity following depletion of
intracellular calcium stores but was reported to enhance the channel
activity in response to activation of Gq-coupled receptors
or receptor tyrosine kinases (25). The expression of HTRP4 in HEK-293
cells was reported to have no effect on barium entry stimulated by
calcium store depletion, supporting the findings for murine Trp4.
However, in contrast to results for murine Trp4, the HTRP4 produced a
constitutively active channel which was not stimulated by either
phospholipase C-linked receptor activation or by OAG (26). Given these
dramatically different results from previous Trp4 overexpression
studies, there is a compelling need to utilize alternative approaches
to help resolve the role of Trp4 in the regulation of capacitative
calcium entry.
Similar discrepancies in findings between laboratories exist for Trp1
(14, 15, 27-29), and Trp3 (14, 20, 30-35). One can think of a number
of theoretical reasons for overexpression studies to produce results
that may not reflect the role of endogenous Trps in native
Ca2+ channel activity. The endogenous channels may be
heterotetramers, and the channels formed in the overexpression studies
may be homotetramers due to their high expression level, or the
expressed Trp may be unable to form the appropriate heterotetramer
channels due to the lack of the appropriate Trp isoforms in the cell
being tested. We have also published data on clone-to-clone variation
within the HEK-293 cell population that points out the potential
pitfalls of interpreting data based on studies with a small number of
stable clones expressing Trps (36). To circumvent many of the problems of overexpression studies, we have used an antisense approach to
investigate these questions.
In a previous study, we reported that HEK-293 cells express mRNA
for HTRP1, HTRP3, HTRP4, and HTRP6, and evidence was presented that
HTRP1 and HTRP3 are involved in mediating store-operated Ca2+ entry (32). In the present study, we attempt to
determine whether multiple Ca2+ entry pathways are mediated
by endogenous HTRPs in HEK-293 cells. We have used the inhibitor,
2-aminoethyoxydiphenyl borane (2-APB), to analyze and compare
Ca2+ entry induced by thapsigargin, carbachol, and OAG in
HEK-293 cells. We also have generated antisense constructs for HTRP1, HTRP3, and HTRP4 and stably expressed them in HEK-293 cells. We provide
evidence that in HEK-293 cells endogenous HTRP4 proteins do not
participate in the formation of cation channels regulated by store
depletion with thapsigargin but do participate in channel activity
activated by CCh or OAG. More importantly, we provide evidence that
HTRP4 proteins are involved in the formation of channels responsible
for both the arachidonic acid-induced Ca2+ entry and the
Ca2+ entry needed to sustain long term Ca2+
oscillations in HEK-293 cells in response to low doses of CCh.
| |
EXPERIMENTAL PROCEDURES |
Materials--
Fura-2-free acid, fura-2/AM, and pluronic F-127
were purchased from Molecular Probes, Inc. (Eugene, OR); thapsigargin
was from LC Laboratory; G418 was from Mediatech; Nusieve GTG-agarose was from FMC BioProducts; Hanks' balanced salt solution (HBSS), Ca2+-free, Mg2+-free,
HCO
-free HBSS and Dulbecco's modified Eagle's medium, penicillin/streptomycin,
L-glutamine, and trypsin-EDTA were from Invitrogen;
Chelex-100 was from Bio-Rad; arachidonic acid was from Biomol; and
carbachol, OAG, and 2-APB, along with other chemicals, were purchased
from Sigma.
Isolation of cDNA-Encoding Fragments of Trps by
RT-PCR--
The poly(A)-RNA was isolated from HEK-293 cells using
guanidinium thiocyanate extraction followed by an oligo(dT) binding method (QuickPrep Micro mRNA Purification Kit; Amersham
Biosciences). The extracted mRNA was then treated with DNase
(Invitrogen). The first strand cDNA was reverse transcribed using
an oligo(dT) primer, and it was then amplified directly using PCR
(SuperScript Preamplification System; Invitrogen). Sequences of primers
to amplify specific regions of HTRP1, HTRP3, and HTRP4 as well as human
actin are described in Table I. Sequence similarity analysis was
performed using Seglab software by Genetics Computer Group. We used the inner primers to amplify fragments of endogenous HTRP cDNA in normal HEK-293 cells for the productions of the sense and antisense constructs and the outer primers to amplify HTRP cDNA fragments in
HEK-293 cells in which sense or antisense constructs were stably expressed. The hot start PCR was performed with Taq
polymerase; the initial denaturation was at 94 °C for 5 min,
followed by 30 cycles of 94 °C for 1 min, 60 °C for 1 min, and
72 °C for 30 s. A final extension was performed at 72 °C for
3 min. The PCR conditions were optimized to 30 cycles of annealing at
55 °C for endogenous HTRP1 and HTRP3 and to 35 cycles of annealing
at 53 °C for endogenous HTRP4, using the inner primers. For the
outer primers, the PCR was cycled 30 times, and the annealing was at
60 °C for all HTRP isoforms. For the experiments to test for
mRNA expression for the 
and isoforms of HTRP4, the PCR was
cycled 30 times, and the annealing was performed at 52.5 °C. The PCR
mixture (48 µl) consisted of ~22% of first stand cDNA as a
template, 1 µl of 100 pM solution of each primer, 5 µl
of 25 mM MgCl2, 1 µl of 10 mM dNTP mix, and 2.5 units of Taq DNA polymerase (Invitrogen).
For a positive control and to determine whether equal amounts of
mRNA were used for each condition, we used a human -actin
primer, and a negative control without RT was performed alongside all experimental samples.
The cDNA fragments of HTRP1 (369 base pairs), HTRP3 (323 base
pairs), and HTRP4 (412 base pairs) from RT-PCRs were separated by
electrophoresis in a 3% GTG-agarose gel. The corresponding bands were
cut out of gels, extracted (QIAEX II Gel Extraction 150), and subcloned
into an eukaryotic TA cloning vector pCR3.1 that accepts products in
both the forward and reverse directions (eukaryotic TA cloning kit,
bidirectional; Invitrogen). The clones were selected randomly, and the
cDNAs were purified (Qiagen Plasmid Maxi Kit 25) and sequenced. The
antisense constructs encoding HTRP1, HTRP3, and HTRP4 were expressed in
HEK-293 cells, whereas the corresponding short sense constructs were
expressed for the controls.
Cell Culture--
HEK-293 cells were cultured in Dulbecco's
modified Eagle's medium supplemented with 10% fetal bovine serum, 50 units/ml penicillin, 50 µg/ml streptomycin, and 2 mM
glutamine. Cells were grown in an incubator at 37 °C with humidified
5% CO2 and 95% air.
Stable Transfection of Antisense HTRP1, HTRP3, and HTRP4
Constructs--
HEK-293 cells were transfected with sense or antisense
constructs using the Ca2+ phosphate method. G418-resistant
transformants were selected using 400 µg/ml G418. In total, three
antisense cell lines (HTRP1AS, HTRP3AS, and HTRP4AS) and three sense
cell lines (HTRP1S, HTRP3S, and HTRP4S) were established. For each cell
line expressing sense or antisense constructs, all of the surviving
clones (~200) were pooled together to generate a cell line stably
expressing one of the Trp construct(s). This was done in order to avoid
the problems inherent with using small numbers of selected clones to
compare the effects of gene transfection. In a previous publication, we reported significant clone-to-clone variation of SOC activity within
the parent HEK-293 cell population, and, utilizing Monte Carlo
analysis, we estimated the probability that this clone-to-clone variation would lead to misinterpretations of the effect of Trp channel
expression (36). Cells up to passage 30 were used for Ca2+ imaging.
Immunoblotting--
Western blots of endogenous HTRPs were
successful only after extensive characterization of HTRP antibodies
(Alomone Laboratories, Jerusalem, Israel). We tested various
cell lysis buffers, sample heating temperatures, support membranes,
wash buffers, and antibody incubation times on samples of total cell
protein, cell membrane protein, and immunoprecipitated tagged Trp
protein prepared from HEK-293 cells overexpressing Trp proteins.
Initial Western blot experiments were performed with anti-tag
antibodies, whereas subsequent experiments compared results with
anti-Trp antibodies with those obtained with anti-tag antibodies.
Conditions were considered optimum when we could see clear distinct
bands of the appropriate molecular weight on Western blots of
Trp-expressing cells and the intensity of these bands was in great
excess of corresponding bands in the lane for control HEK-293 cells.
Since control samples of HEK-293 total cell protein run on minigels
often did not show endogenous Trp proteins, we scaled up the procedure
in order to routinely observe endogenous Trp proteins. We went to a
large gel format (16 × 16 cm) so that we could load more protein,
and we included a protein precipitation and wash step in order to reduce the amount of contaminants associated with loading a larger protein sample onto the gels. The precipitation and wash step was found
to greatly reduce the background of the Western blots. Optimum blocking
and wash conditions for Westerns with HTRP4 antibodies were slightly
different from those for HTRP1 and HTRP3 antibodies (see below).
Cells were lysed in modified radioimmune precipitation buffer (10 mM Tris-HCl, pH 7.5, 500 mM NaCl, 0.1% SDS,
1% Nonidet P-40, 1% sodium deoxycholate, 2 mM EDTA).
Protease Arrest (Geno Technology, Inc.) was added to prevent
proteolysis. A quantitative precipitation of all of the proteins in the
total cell extract was done with PAGE-Perfect (Geno Technology), and
the protein precipitate was then washed to remove agents that would
interfere with running high quality gels. Precipitated protein pellets
were mixed with 2× Laemmli buffer (plus 100 mM
dithiothreitol) and heated for 30 min at 70 °C, and sample pairs
(antisense cells versus control cells) containing equal
amounts of protein (200 µg for HTRP1 and HTRP3 and 400 µg for
HTRP4) were loaded on a regular size gel (16 × 16 cm, 7.5%
SDS-PAGE). Electrophoresis was performed overnight, and then proteins
were electrotransferred onto Immobilon membranes (Millipore Corp.). For
blots with anti-HTRP1 and anti-HTRP3, the membranes were blocked with
5% milk solution in TBS-T (0.1% Tween 20) (for HTRP4 Western blots,
the Tween 20 was omitted from the TBS solution) for 1 h and were
then incubated with primary antibodies raised against the individual
HTRPs (Alamone Laboratories) overnight at room temperature. The
antibodies were diluted 1:200 in the blocking solution. Membranes were
washed four times for 15 min each with TBS-T (for Trp4 Western blots,
three times for 15 min each with TBS and one time for 15 min with
TBS-T), incubated for 30 min at room temperature with secondary
anti-rabbit antibody (1:10000 in 5% milk/TBS-T (TBS for HTRP4 Western
blots)), washed under the same conditions, and developed with
SuperSignal Chemiluminescent Substrate (Pierce).
Ca2+ Imaging--
[Ca2+]i
concentration was measured in cells loaded with the fluorescent
indicator fura-2. Transfected cells were plated onto 25-mm coverslips 1 day before the experiment. On the next morning, cells were washed twice
with a HEPES-buffered HBSS (HHBSS); loaded for 30 min with 5 µM fura-2/AM that was dissolved in HHBSS supplemented
with 1 mg/ml bovine serum albumin, 0.025% pluronic F127; and then
unloaded in HHBSS for another 30 min. The coverslips were mounted
as the bottom of a chamber that was placed on the stage of a
Nikon Diaphot inverted epifluorescence microscope equipped with a Zeiss
Fluor ×10 (or ×20 for single cell analysis) objectives. Cells in the
chamber were perfused via an eight-channel syringe system (Anspec). A
suction pipette maintained a constant volume of solution (~0.5 ml) in
the chamber.
An InCyt IM2TM fluorescence imaging system (dual wavelength
fluorescence imaging system; Intracellular Imaging Inc., Cincinnati, OH) was used to measure [Ca2+]i during the
experiment. Excitation light from a xenon light source was
alternately passed through 340 and 380 nm narrow pass filters mounted
in a Sutter filter wheel (Lambda 10-C). The 510 nm emissions
were captured by a cooled CCD camera (Cohu 4915). The images were
transmitted to a computer (Intel Pentium Pro 500-MHz based PC) and
processed with the imaging software InCyt IM2TM 4.6. [Ca2+]i was calculated by measuring the ratio of
the two emission intensities for excitation at 340 and 380 nm. Calcium
standard solutions, which were prepared with fura-2 potassium salt,
were used to create a graph of fluorescence ratio
(F340/F380) as a function
of Ca2+ concentration (in nM). This graph was
then used to convert fluorescence ratios in an experiment to calcium concentrations.
In experiments in which Ba2+ and Sr2+ influx
were measured, the data are reported as the 340/380 ratio (R340/380),
since the Ba2+ and Sr2+ calibration curves for
fura-2 differ from the calibration curve for Ca2+.
During most experiments, an average response of ~800 cells from a
single field on each coverslip was represented as one trace. The level
of store-operated Ca2+ entry in the cells expressing Trp
sense or antisense constructs was obtained by subtracting the slope of
the Ba2+ leak (before stimulation) from the slope of
Ba2+ influx (after stimulation) for each coverslip. In
Ca2+ oscillation experiments, responses of single cells
were reported.
Nominally Ca2+-free HBSS was prepared by stirring
Ca2+-free, Mg2+-free, and
HCO-free HBSS with Chelex-100 beads.
After filtering out the Chelex-100 beads, MgCl2 was added to a final concentration of 1 mM.
| |
RESULTS |
Effects of 2-APB on Ca2+ Signaling Induced by
Thapsigargin, Carbachol, and OAG--
The membrane-permeant inhibitor
of the InsP3 receptor (37), 2-APB, has been utilized in
several recent studies to address the question of whether a direct
interaction occurs between the inositol 1,4,5-trisphosphate receptor
and overexpressed HTRP channels. However, recent reports suggest that
2-APB may be directly inhibiting the store-operated channels (38, 39).
Given the lack of a well defined mechanism of action, we set out simply
to determine whether 2-APB could help distinguish between the various
endogenous Ca2+ entry pathways in our HEK-293 cells.
To investigate thapsigargin-stimulated Ca2+ entry, cells in
HBSS were switched to a Ca2+-free HBSS medium and allowed
to establish a new base line, and then the cells were treated with
thapsigargin (1 µM). As expected, in the absence of
Ca2+, thapsigargin induced a initial transient
Ca2+ peak that reflects the depletion of intracellular
stores. The level of [Ca2+]i subsequently
declined, suggesting the removal of [Ca2+]i from
the cell by the plasma membrane Ca2+-ATPase. After
[Ca2+]i returned to basal levels,
Ba2+ (5 mM) was added into
Ca2+-free medium, and the initial slope of Ba2+
entry was taken to indicate the level of store-operated
Ca2+ entry (SOCE) (Fig.
1A). In unstimulated cells in
Ca2+-free medium, 2-APB (100 µM) slightly
elevated Ca2+ above basal levels (Fig. 1B),
possibly through its inhibitory effect on Ca2+-ATPase in
the endoplasmic reticulum (37). As reported for cells overexpressing
HTRP3 (40), we observed that 2-APB produced a concentration-dependent inhibition of
thapsigargin-stimulated Ba2+ entry (Fig. 1B). At
75 µM 2-APB, there was an ~50% reduction (n = 3) of Ba2+ entry, whereas at 100 µM 2-APB, a higher level of inhibition (75%) was
observed (n = 3).
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 1.
Inhibitory effect of 2-APB on
Ca2+ entry induced by thapsigargin, CCh, or OAG. In
the absence of Ca2+, fura-2-loaded HEK-293 cells were
treated with 1 µM thapsigargin (A and
B) or 100 µM CCh (C and
D). The SOCE or CCh-stimulated entry was assessed by the
readdition of BaCl2 (Ba2+; 5 mM) in
Ca2+-free medium following store depletion with either
thapsigargin or CCh. A, the SOCE induced by thapsigargin in
control HEK-293 cells; B, the inhibitory effect of 2-APB
(100 µM) on SOCE. C, the CCh-stimulated
Ba2+ entry in control HEK-293 cells; D, the
CCh-stimulated Ba2+ entry in 2-APB-treated cells. 2-APB
(100 µM) was added at 3 min prior to carbachol addition.
To monitor OAG effects (E and F),
SrCl2 (Sr2+; 5 mM) was added in
Ca2+-free medium. Stimulated Sr2+ entry was
initiated by the addition of OAG (100 µM) into the
chamber immediately after solution flow was stopped. 2-APB was added 3 min before OAG stimulation. Shown is the OAG-stimulated
Sr2+ entry in control cells (E) and in cells
treated with 100 µM 2-APB (F). Each
trace represents one coverslip (~800 cells). Three
coverslips (n = 3) were done for each condition, and
similar results were observed. The basal [Ca2+]i
was ~50 nM, and the peak of Ca2+ following
thapsigargin stimulation was ~350 nM. The peak of
Ca2+ following CCh stimulation was ~850
nM.
|
|
We used a similar protocol to investigate the effect of 2-APB on
carbachol-induced Ca2+ entry. In the absence of
Ca2+, carbachol (100 µM) induced a large,
transient increase in [Ca2+]i, indicating
Ca2+ release from intracellular stores following
InsP3 receptor activation. As the cytosolic
Ca2+ returned to basal levels, Ba2+ (5 mM) was added into the bath, and receptor-stimulated
Ba2+ entry was observed (Fig. 1C). In a series
of subsequent experiments, we applied 2-APB (100 µM) to
the bath for 3 min prior to carbachol stimulation. Following a 3-min
preincubation with 2-APB (100 µM), carbachol-induced
Ba2+ entry was dramatically reduced (Fig. 1D), a
finding consistent with the inhibitory effect of 2-APB reported
originally in cells overexpressing HTRP3 (40). Interestingly, under
these conditions, the Ca2+ release from internal stores was
not significantly altered (Fig. 1D). With longer periods of
preincubation with 2-APB (100 µM), the transient
Ca2+ release induced by carbachol was significantly or
completely inhibited (data not shown).
We next sought to assess the effect of 2-APB on OAG-induced
Ca2+ entry. Since previous studies investigating OAG
activation of overexpressed, human Trp proteins monitored
Sr2+ entry, we measured Sr2+ entry in the
presence of 100 µM OAG and the presence or absence of
2-APB (100 µM). As noted for the Ca2+ data in
Fig. 1B, there was a dose-dependent effect of
2-APB on basal Sr2+ levels (Fig. 1F). However,
2-APB, regardless of dose used, had no effect on the large slope of
OAG-induced Sr2+ entry (Fig. 1, E and
F). The data suggest that the cation entry pathway activated
by OAG is distinct from cation entry pathways activated either by store
depletion with thapsigargin or by CCh stimulation of HEK-293 cells.
Level of Expression of mRNA for Trp Homologs in HTRP1AS,
HTRP3AS, and HTRP4AS Cells--
Our previous studies indicated that
HEK-293 cells express mRNA for HTRP1, HTRP3, HTRP4, and HTRP6 (32).
To evaluate the role of HTRP1, HTRP3, and HTRP4 in mediating
thapsigargin-, CCh-, and OAG-stimulated Ca2+ entry, we made
short antisense cDNA constructs specific for HTRP1, HTRP3, and
HTRP4 that we stably transfected into HEK-293 cells. This allowed us to
generate three cell lines, i.e. HTRP1AS, HTRP3AS, and
HTRP4AS. Stable transfection of short sense cDNA constructs (300-400 nucleotides) for each of these Trp homologs was used as a
control (i.e. HTRP1S, HTRP3S, and HTRP4S). As we described previously, we mixed all of the surviving clones (~200) to generate heterogeneous populations of transfected cells for each cell line; therefore, the large cell-to-cell variations of SOCE we previously reported in the parent HEK-293 population should have no impact on our
interpretations (36).
We extracted poly(A) RNAs from all cell lines mentioned above, reverse
transcribed first strand cDNAs, and then specifically amplified the
appropriate cDNAs via PCR using primers specific for regions just
outside the sequence of the sense and antisense constructs (outer
primers, Table I). This was done so as
not to amplify the expressed sense and antisense constructs along with
the regions from the endogenous mRNA. As shown in Fig.
2, expression of HTRP1, HTRP3, and HTRP4
mRNA was detected in HTRP1S, HTRP3S, and HTRP4S cells,
respectively. Since a previous report has shown alternative splice
variants of HTRP4 (41), we used a different set of primers to determine
whether HTRP4 mRNA expressed was for the or the isoform. We
detected both isoforms, with the isoform being the more abundant
species (the isoform was about 20% of the level of the isoform; data not shown). In addition, the data in Fig. 2 show that the
expression of the mRNA for the specific Trp homolog was
significantly reduced in HTRP1AS, HTRP3AS, and HTRP4AS cells in which
antisense cDNA constructs were stably expressed.
View this table:
[in this window]
[in a new window]
|
Table I
Sequences of inner primers and outer primers for detection of HTRPs
The sense and antisense constructs for human Trps spanned
~100-130-amino acid-long segments of Trps, starting from the
transmembrane domain 5 and ending 12-20 amino acids downstream of
transmembrane domain 6.
|
|
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 2.
RT-PCR Detection of expression of mRNA in
control and antisense-expressing cells. Antisense cDNA
constructs for HTRP1 (HTRP1AS), HTRP3 (HTRP3AS), and HTRP4 (HTRP4AS)
were stably transfected into HEK-293 cells, whereas short sense
cDNA constructs (~300 nucleotides) for HTRP1 (HTRP1S), HTRP3
(HTRP3S), and HTRP4 (HTRP4S) were used as controls. For each cell line,
poly(A) RNA was extracted, treated with DNase, and reverse transcribed
with oligo(dT)12-18 primers and SuperscriptTM
II reverse transcriptase (Invitrogen). The first strand cDNA was
then amplified with PCR using the forward and outer reverse primers
that are specific for each Trp isoform (Table I) to amplify the various
Trp isoform cDNA. -Actin was amplified as a positive control to
assure that changes observed were not due to different levels of
starting material (first two lanes).
For a negative control, reactions were performed without reverse
transcriptase (No RT; last
two lanes). The results shown are representative
of at least three independent experiments.
|
|
Level of Expression of Trp Proteins in HTRP1AS, HTRP3AS, and
HTRP4AS Cells--
To confirm that antisense expression was reducing
the protein levels as well as the mRNA levels of the various
endogenous Trps, we performed Western blot analysis on all of the cells
utilized. Although we were ultimately successful in observing
endogenous Trp proteins on Western blots of HEK-293 cells utilizing the
Alomone anti-Trp antibodies, this occurred only after an extensive
characterization procedure for the anti-Trp antibodies (see
"Experimental Procedures"). Initial experiments gave high
background, multiband blots where the differences in the appropriate
molecular weight band was barely discernible between lanes for control
and Trp-overexpressing cells. However, after optimization of the
techniques, we were able to see single, clear, strong bands at the
appropriate molecular weight for endogenous HTRP1, HTRP3, and HTRP4.
The Western blot data in Fig. 3
demonstrate that HTRP1, HTRP3, and HTRP4 are endogenously expressed in
HEK-293 cells and that the expression of antisense constructs
specifically reduces individual HTRP protein levels. As seen in Fig.
3C, the introduction of HTRP1 antisense results in a
reduction of HTRP1 protein to a value that is 36 ± 2.8% of the
control protein level (significantly different from control value,
p < 0.001, n = 3), while having no
effect on either HTRP3 or HTRP4 protein levels. The expression of
antisense to HTRP3 results in a reduction of HTRP3 protein to a value
that is 43.0 ± 6.0% of the control protein level (significantly
different from control value, p < 0.005, n = 4), while having no effect on HTRP1 and HTRP4
protein levels (Fig. 3B). Finally, the expression of HTRP4
antisense results in a reduction of HTRP4 protein to a value that is
31 ± 14% of the control protein level (significantly different from control value, p < 0.02, n = 3),
while having no effect on HTRP1 and HTRP3 proteins levels (Fig.
3A). Therefore, the antisense constructs demonstrate the
specificity of action required to analyze the involvement of individual
Trp isoforms in various endogenous Ca2+ entry pathways.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 3.
Western blot analysis of HTRP1, HTRP3, and
HTRP4 protein levels in control and antisense-expressing cells.
Total cell lysates from HTRP4S, HTRP4AS, HTRP3S, HTRP3AS, HTRP1S, and
HTRP1AS cells were obtained using procedures described under
"Experimental Procedures." Total protein (200 or 400 µg) was
separated by 7.5% SDS-PAGE and transferred to Immobilon membrane.
Polyclonal anti-human Trp4, Trp3, or Trp1 antibody and horseradish
peroxidase-labeled goat anti-rabbit immunoglobulin were used as primary
and secondary antibodies, respectively. The signals were detected by
ECL using standard protocols. The results shown are representative of
at least three independent experiments.
|
|
Role of HTRP4 in Thapsigargin-stimulated, CCh-stimulated, and
OAG-stimulated Ca2+ Entry--
To investigate the
functional importance of HTRP4 in regulating Ca2+ influx,
we examined entry in HTRP4AS cells, compared with HTRP4S cells, in
response to thapsigargin, carbachol, and OAG. We monitored either
Ba2+ or Sr2+ entry prior to stimulation and
then again after stimulation. We summarized these results in Fig.
4. In the left
panels, we show traces from two representative coverslips
(HTRP4AS versus HTRP4S), each with the response averaged
over ~800 cells in the microscope field. In the study of
thapsigargin- or carbachol-induced Ca2+ entry pathways, we
measured levels of Ba2+ (or Sr2+) before (basal
leak) and after (total influx) stimulation. We then determined the
stimulated influx by subtracting the basal leak from the total influx.
In the study of the OAG-induced entry pathway, we simply measured the
slope of Sr2+ entry in the presence of OAG in HTRP4AS
versus HTRP4S cells, since no basal Sr2+ entry
was observed. In the right panel, we show the
mean values (with error bars showing the S.E.) of
the stimulated Ba2+ or Sr2+ uptakes plotted as
percentages of its control. In this case, we determined the mean value
of final Ba2+ (or Sr2+) influx for a series of
coverslips for each cell type, and we determine whether the difference
between HTRP4AS and HTRP4S cells was statistically significant by
Student's t test.
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 4.
Role of HTRP4 in thapsigargin-, CCh-, and
OAG-stimulated Ba2+ or Sr2+ entry. HTRP4
antisense cDNA constructs were stably transfected into HEK-293
cells, generating a HTRP4AS cell line. Stable transfection of HTRP4
sense cDNA constructs (~300 nucleotides) into HEK-293 cells
(i.e. HTRP4S cells) was used as a control. The
left panel shows the time course of
representative experiments during which levels of Ba2+ (or
Sr2+) before (basal leak) and after (total influx) agonist
stimulation were measured. SOCE was determined by subtracting the basal
leak from the total influx following store depletion with 1 µM thapsigargin (A and C).
CCh-stimulated entry was determined by subtracting basal
Ba2+ leak from total Ba2+ influx following
stimulation with 100 µM carbachol (E).
OAG-stimulated entry was monitored by measuring Sr2+ entry
following stimulation with 100 µM OAG (G).
Each trace represents a single coverslip with the response
averaged over ~800 HTRP4S cells (thick line) or
HTRP4AS cells (thin line). The right
panel shows the statistical analysis of the data, with the
mean value of the leak-subtracted Ba2+ (or
Sr2+) entry (for n coverslips) plotted as a
percentage of control. Statistical differences between values in
HTRP4AS cells and HTRP4S cells were analyzed by Student's t
test.
|
|
Fig. 4A shows both the basal Ba2+ leak (before
store depletion) and the total Ba2+ influx following store
depletion with thapsigargin (1 µM). When one compares the
responses in HTRP4AS and HTRP4S cell lines, there is little difference
in the initial transient Ca2+ peak (Fig. 4, A
and C), suggesting that the suppression of HTRP4 expression
does not significantly alter the size of the internal Ca2+
stores. Furthermore, there is no difference in thapsigargin-stimulated Ba2+ entry (slope: HTRP4S, 0.69 ± 0.07, n = 15; HTRP4AS, 0.67 ± 0.07, n = 15; p > 0.8), suggesting that HTRP4 plays no major
role in regulation of SOCE (Fig. 4B).
We also used Sr2+ to monitor SOCE. Interestingly,
Sr2+ leak was undetectable in both HTRP4AS and HTRP4S cells
(Fig. 4C). Furthermore, the Sr2+ influx
following store depletion was dramatically enhanced, independent of
cell lines (Fig. 4C). As shown in Fig. 4D, there
was no difference in Sr2+ entry in HTRP4AS compared with
HTRP4S cells (slope: HTRP4S, 0.62 ± 0.02, n = 12;
HTRP4AS, 0.64 ± 0.02, n = 12; p > 0.50). The Ba2+ and Sr2+ data both agree
that HTRP4 proteins are not involved in the Ca2+ entry
pathway activated by calcium store depletion with thapsigargin.
To evaluate whether HTRP4 plays a role in receptor-stimulated
Ca2+ entry, we measured Ba2+ influx before and
after carbachol (100 µM) stimulation of HTRP4AS cells,
using HTRP4S cells as a control. As shown in Fig. 4E, in the
absence of Ca2+, carbachol induced a large, transient
increase in [Ca2+]i followed by a decline to
basal levels. Stable transfection of antisense constructs for HTRP4 had
no measurable effect on the size of the CCh-stimulated Ca2+
release; however, it did cause a dramatic decrease in final
Ba2+ influx (Fig. 4E). As shown in Fig.
4F, the expression of HTRP4 antisense resulted in a 35%
inhibition of Ba2+ entry, which was a statistically
significant reduction (slope: HTRP4S, 1.19 ± 0.15, n = 36; HTRP4AS, 0.77 ± 0.10, n = 36; p = 0.03).
To evaluate whether HTRP4 plays a role in OAG-stimulated
Ca2+ entry, we measured Sr2+ influx before and
during stimulation with OAG (100 µM) in HTRP4AS cells,
using HTRP4S cells as a control. As observed in Fig. 4G, Sr2+ leak was undetectable in both HTRP4AS and HTRP4S
cells. However, OAG-induced Sr2+ entry was dramatically
reduced in HTRP4AS compared with HTRP4S. The expression of HTRP4
antisense resulted in a 46% inhibition of OAG-stimulated
Sr2+ entry, a statistically significant reduction (Fig.
4H) (slope: HTRP4S, 2.24 ± 0.23, n = 28; HTRP4AS, 1.20 ± 0.19, n = 28;
p = 0.001).
In summary, HTRP4 does not seem to play a role in regulation of SOCE.
However, it does play a role in carbachol-, and OAG-induced Ca2+ entry.
OAG-induced Calcium Signaling in HEK-293 Cells Stably Transfected
with HTRP1AS, HTRP3AS, and HTRP4AS--
In a previous study, we showed
that stable expression of HTRP3 antisense in HEK-293 cells resulted in
a 32% inhibition of SOCE. Transient transfection of an HTRP1 antisense
construct in HTRP3AS cells led to a higher level of inhibition
(55%) of store-operated Ca2+ entry (32). To evaluate
whether HTRP1 or HTRP3 also play a role in mediating OAG-induced
Ca2+ signaling, we measured Sr2+ influx induced
by OAG (100 µM) in HEK-293 cells expressing antisense cDNA constructs for HTRP1 (HTRP1AS) or HTRP3 (HTRP3AS). HTRP1S or
HTRP3S cells were used as the corresponding controls. Sr2+
leak was undetectable in any of these experiments. Sr2+
influx induced by OAG was dramatically reduced in HTRP3AS compared with
HTRP3S. The magnitude of the effect was a 52% reduction, which is
statistically significant (Fig. 5)
(slope: HTRP3S, 2.46 ± 0.18, n = 14; HTRP3AS,
1.17 ± 0.22, n = 11; p = 0.0002).
In contrast, there was no difference in OAG-induced Sr2+
influx between HTRP1AS and HTRP1S cells (Fig. 5) (slope: HTRP1S, 2.12 ± 0.36, n = 10; HTRP1AS, 2.20 ± 0.31, n = 11, p > 0.8). The data suggest
that HTRP3 and HTRP4, but not HTRP1, play an important role in
OAG-induced Sr2+ entry (Fig. 5).
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 5.
OAG-induced Sr2+ entry in HEK-293
cells stably transfected with HTRP1AS, HTRP3AS, or HTRP4AS.
Sr2+ influx induced by OAG (100 µM) was
measured in HTRP1AS, HTRP3AS, and HTRP4AS cells in comparison with
their corresponding controls HTRP1S, HTRP3S, and HTRP4S.
Sr2+ leak was undetectable in any of these experiments. The
mean value of the Sr2+ entry (for n coverslips)
is plotted as a percentage of control. Statistical differences between
values in control and antisense-expressing cells were analyzed by
Student's t test.
|
|
The Role of HTRP3 and HTRP4 in Ca2+
Entry-dependent, CCh-induced Ca2+
Oscillations--
Two recent studies (42, 43) indicate that CCh, at
low doses, can initiate long lasting Ca2+ oscillations in
HEK-293 cells. Sustained oscillations of Ca2+ in response
to CCh (sustained for at least 1 h) are strictly dependent on the
presence of extracellular Ca2+. While initial theories
suggested that the Ca2+ entry required for sustained
Ca2+ oscillations was mediated via store-operated channels,
very recent data suggest that the Ca2+ entry is via
non-store-operated channels (42, 44). Since we could reduce
Ca2+ entry via either store-operated or non-store-operated
pathways by expressing either HTRP3 antisense or HTRP4 antisense, we
investigated what effect the expressions of these antisense constructs
have on cell Ca2+ oscillations.
Since we observed above that HTRP4 does not play a role in SOCE but
does play a role in CCh-stimulated Ca2+ entry, it seemed
possible that HTRP4 is involved in mediating the Ca2+ entry
required to support the sustained Ca2+ oscillations in CCh.
To test this hypothesis, we monitored CCh-induced Ca2+
oscillations in individual cells for both the HTRP4S and HTRP4AS populations of cells. The data in Fig. 6
(left panel), show 10 representative single cell
responses recorded from one HTRP4S coverslip. Cells were stimulated
with 15 µM CCh in a medium containing 1.8 mM
Ca2+. These cells exhibited repetitive peaks over the
10-min time period shown. In Fig. 6 (right
panel), we show 10 representative individual cell responses
from one HTRP4AS coverslip. Note that there are dramatically fewer
repetitive Ca2+ spikes in HTRP4AS cells in comparison with
HTRP4S cells. In Fig. 7, we show the
statistical analysis of the responses recorded from over 300 individual
cells of each type. It is clear that many more HTRP4S cells than
HTRP4AS cells undergo in excess of three Ca2+ oscillations,
the number of damped oscillations seen in normal HEK-293 cells in a
Ca2+-free medium (data not shown). However, it is also
clear that the amplitude of the first spike is similar in the two cell
lines, indicating that the difference is not in the initial
intracellular Ca2+ pool content but rather in the inability
of the HTRP4AS cells to continue to refill the internal
Ca2+ stores after the initial depletion.
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 6.
The role of HTRP4 in the sustained
Ca2+ oscillations induced by low doses of CCh.
Ca2+ responses to CCh (15 µM) were recorded
in individual HTRP4AS or HTRP4S cells over a period of 10 min in
Ca2+-containing HBSS. Traces for 10 representative cells
from a single HTRP4S coverslip (left panel) and
10 representative cells from a single HTRP4AS coverslip (right panel)
are shown. Sampling rate was 90 image pairs/min.
|
|
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 7.
Frequency plot for CCh-induced
Ca2+ oscillations in HTRP4AS versus HTRP4S
cells. The number of HTRP4S cells (upper
panel) or HTRP4AS cells (lower panel)
responding with a given number of Ca2+ spikes within a
10-min period is plotted versus that number of
Ca2+ spikes. The data summarizes the results in 311 HTRP4S
cells and 309 HTRP4AS cells.
|
|
To determine whether the effects observed with HTRP4
antisense were specific and therefore supported a role for HTRP4 in
maintaining the ongoing Ca2+ oscillations, we determined
whether the expression of an HTRP3 antisense construct, which reduces
store-operated Ca2+ entry in HEK-293 cells (32), would
affect the Ca2+ oscillations induced by low doses of CCh.
The data in Fig. 8 (left
panel), show 10 representative single cell responses
recorded from one HTRP3S coverslip. Cells were stimulated with 15 µM CCh in a medium containing 1.8 mM
Ca2+. These cells exhibited repetitive peaks over the
10-min time period shown. In Fig. 8 (right
panel), we show 10 representative individual cell responses
from one HTRP3AS coverslip. Note that there are no obvious differences
in the number of Ca2+ spikes in HTRP3AS cells compared with
HTRP3S cells. In Fig. 9, we show the
statistical analysis of the responses recorded from over 300 individual
cells of each type. It is clear that HTRP3S and HTRP3AS cells undergo a
similar number of Ca2+ oscillations, arguing that reduction
of store-operated Ca2+ entry has no effect on the sustained
Ca2+ oscillations.
View larger version (32K):
[in this window]
[in a new window]
|
Fig. 8.
The role of HTRP3 in the sustained
Ca2+ oscillations induced by low doses of CCh.
Ca2+ responses to CCh (15 µM) were recorded
in individual HTRP3AS or HTRP3S cells over a period of 10 min in
Ca2+-containing HBSS. Traces for 10 representative cells
from a single HTRP3S coverslip (left panel) and
10 representative cells from a single HTRP3AS coverslip
(right panel) are shown. The sampling rate was 90 image pairs/min.
|
|
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 9.
Frequency plot for CCh-induced
Ca2+ oscillations in HTRP3AS versus HTRP3S
cells. The number of HTRP3S cells (upper
panel) or HTRP3AS cells (lower panel)
responding with a given number of Ca2+ spikes within a
10-min period is plotted versus that number of
Ca2+ spikes. The data summarize the results in 301 HTRP4S
cells and 303 HTRP4AS cells.
|
|
The Role of HTRP3 and HTRP4 in Arachidonic
Acid-stimulated Ca2+ Entry--
Since a recent report
suggested that the Ca2+ entry pathway required for
sustained Ca2+ oscillations is an arachidonic
acid-activated channel (43), we sought to determine whether expression
of HTRP4 antisense had any impact on the arachidonic acid-stimulated
Ca2+ entry. The data in Fig.
10A are representative
traces that show that the addition of arachidonic acid to previously
unstimulated HTRP4S cells produces a dramatic rise in Ca2+
levels, while a similar addition to HTRP4AS cells produced much less of
a response. The data in Fig. 10B summarize our results and
demonstrate that expression of HTRP4 antisense results in a 75%
reduction in the level of arachidonic acid-stimulated Ca2+
entry (slope: HTRP4S, 0.21 ± 0.03, n = 19;
HTRP4AS, 0.05 ± 0.01, n = 20; p < 0.0005). To determine whether this was a specific effect of HTRP4
antisense, we expressed HTRP3 antisense and determined the effect on
arachidonic acid stimulation of Ca2+ entry. As seen in Fig.
10, C and D, there is no significant effect of
expressing HTRP3 antisense on the level of arachidonic acid-stimulated Ca2+ entry (slope: HTRP3S, 0.17 ± 0.02, n = 12; HTRP3AS 0.14 ± 0.02, n = 11; p > 0.35).
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 10.
The role of HTRP4 in arachidonic
acid-stimulated Ca2+ entry. In the presence of HBSS
(Ca2+; 1.8 mM), arachidonic acid (40 µM) was added into the chamber as soon as the solution
flow was stopped, and the consequent Ca2+ responses were
measured in cells expressing antisense constructs in comparison with
their corresponding controls. The left panel
shows the time course of Ca2+ entry for representative
traces; each trace represents a single coverslip with the
response averaged over ~800 cells. A, HTRP4AS cells
(thin line) versus HTRP4S cells
(thick line). C, HTRP3AS cells
(thin line) versus HTRP3S cells
(thick line). The right
panel shows the statistical analysis in which the mean
values of arachidonic acid-induced Ca2+ entry (for
n coverslips) were plotted as a percentage of control.
Statistical differences between values in HTRP4AS cells and HTRP4S
cells (B) and between HTRP3AS and HTRP3S cells
(D) were analyzed by Student's t test.
|
|
| |
DISCUSSION |
Ever since the Trp mutation in the Drosophila
phototransduction cascade was identified (8), an investigation into the
mechanisms by which this light-sensitive ion channel is activated has
ensued. There is good agreement that Trps are activated by
G-protein-coupled receptors, which induce phospholipase C-mediated
phosphoinositide breakdown. However, the phototransduction pathways
that finally activate Drosophila Trp, as well as Trpl,
remain highly controversial (13). The same is true for the mammalian
Trp isoforms. For nearly all of the functionally expressed mammalian
Trps (short isoforms), there is at least one report concerning a
store-operated mechanism of activation (14-16, 24, 28-30, 32-35). On
the other hand, there is evidence for the involvement of
store-independent pathways in the regulation of Trp4 (26), Trp3 (20,
31, 45), and Trp1 (27). The reasons for differences in the functional
properties of Trps reported by individual laboratories remain unclear.
While most of the early studies on Trp function were done by
overexpressing exogenous Trps, our approach has been to investigate the
role of the endogenous Trp isoforms by using antisense techniques to
suppress the expression of various Trp proteins. In a previous publication (32), we showed RT-PCR evidence for the endogenous expression of HTRP1, HTRP3, HTRP4, and HTRP6 in our population of
HEK-293 cells, a finding that agreed qualitatively with a previous publication showing RT-PCR evidence for the expression of these four
Trp isoforms in HEK-293 cells (46). Although we did provide Northern
blot evidence for endogenous expression of HTRP3 (32) and another
publication provided Western blot evidence for endogenous expression of
HTRP1 protein (47) in HEK-293 cells, there is not extensive data in the
literature to confirm the RT-PCR evidence of endogenous expression of
Trp isoforms in HEK-293 cells. Thus, our Western blot data in Fig. 3
represents important evidence for the endogenous expression of HTRP1,
HTRP3, and HTRP4 proteins in HEK-293 cells. However, based on our
RT-PCR and Western blot data, it would appear that HTRP1 and HTRP3
proteins are more abundant than HTRP4 proteins. In the PCR analysis,
cDNA fragments for HTRP1 and HTRP3 were visible after 30 PCR
cycles, whereas the cDNA fragment for HTRP4 was visible only after
35 cycles, suggesting that less mRNA for HTRP4 is present.
Likewise, for the Western blot experiments, we loaded the gels with
twice as much protein and exposed the films for much longer time
periods for the HTRP4 blots than for the HTRP1 and HTRP3 blots.
Contrary to a number of earlier published results on exogenously
expressed Trp4 (16, 22-24), we were not able to detect a role for
endogenous HTRP4 in channel activation by store depletion with
thapsigargin (Fig. 4, A-D). Also, our results run counter to those described in a paper that was published during the preparation of this manuscript, describing studies on vascular endothelial cells
from Trp4 knockout mice (48). In that report, a reduction of
store-operated Ca2+ entry was observed in the absence of
murine Trp4. On the other hand, our results, using the antisense
approach, agree with at least part of the conclusions of two recent
overexpression studies, namely that murine Trp4 (25) and HTRP4 (26) do
not contribute to store-operated channel activity. In addition, a paper
that became available in electronic form during the revision our
manuscript argues that neither the nor splice variant of human
Trp4 is regulated by depletion of internal calcium stores (49).
Although we could see no evidence for HTRP4 involvement in capacitative
calcium entry, we were able to detect a significant effect of HTRP4
antisense expression on channel activation by the muscarinic receptor.
It is likely that endogenous HTRP4 functions as a component of
receptor-mediated calcium entry channels (Fig. 4, E and
F). Thus, our results agree with overexpression studies showing that murine Trp4 is regulated by agonist stimulation (25), but
they disagree with the report that overexpressed HTRP4 is constitutively activated and not regulated by agonists (26). In terms
of cation selectivity, we see a nonselective permeability for
Ba2+ and Sr2+ (Fig. 4, A-D). This
finding agrees with results from human Trp4 (26) and mouse Trp4 (25)
but not bovine Trp4 (22). Subtle differences between HTRP4 and Trp4
from other species may affect, at least partially, its contribution to
channel properties.
We also observed that HTRP4 was involved in the pathway responsible for
OAG stimulation of Sr2+ entry (Fig. 4, G and
H). Since phospholipase C activation by carbachol should
produce an elevation of diacylglycerol, one might expect the
Ca2+ entry pathway activated by OAG to be the same pathway
that is activated following CCh stimulation. Consistent with this
hypothesis are the findings that both the OAG-induced and the
CCh-stimulated Ca2+ entry pathways are at least partially
dependent on the presence of HTRP4. We observed that expression of
HTRP4 antisense can reduce the CCh-stimulated Ca2+ entry by
35% and can reduce the OAG-stimulated Ca2+ entry pathway
by 46%. However, based on the effects of 2-APB on these two pathways,
it seems unlikely that CCh stimulates the OAG-inducible pathway in
HEK-293 cells. We see that most of the CCh-stimulated entry pathway(s)
is inhibited by 100 µM 2-APB, while none of the
OAG-induced entry pathway is inhibited by this compound. Clearly, 2-APB
can block the capacitative portion of the CCh-stimulated entry, based
on the ability of 2-APB to block the thapsigargin-stimulated
Ba2+ entry (Fig. 1). However, it is also clear that the
noncapacitative component of Ca2+ entry stimulated by CCh
is sensitive both to HTRP4 expression levels and to the presence of
2-APB. These data strongly suggest that the physiological levels of
diacylglycerol produced by CCh stimulation of cells is not sufficient
to activate the Ca2+ entry pathway activated by
pharmacological levels of OAG. While one could certainly argue that
high concentrations of diacylglycerol near the membrane could be
produced following the activation of phospholipase C, the ability of
2-APB to block most of the CCh-stimulated Ca2+ entry, while
having no effect on the OAG-stimulated entry pathway, indicates that
this is not the case.
Our data demonstrating that Ba2+ entry due to store
depletion with thapsigargin is inhibited by 2-APB (Fig. 1) is, on the
surface, consistent with the notion that 2-APB blocks SOCE via its
interaction with the InsP3 receptor (40). Original reports
showed that 2-APB inhibits agonist-induced Ca2+ release in
platelets, neutrophils, and aorta, probably via inhibition of
InsP3 receptors, type 1 or type 3. While it does not affect InsP3 binding to its receptor, it does partly inhibit
Ca2+ uptake into the store, like thapsigargin, at least at
high doses (100 µM) (37). In recent studies, it was
reported that 2-APB completely blocks Ca2+ entry induced by
different store-depleting agents, and the results were observed in
several cell lines, including HEK-293 cells (40, 42). 2-APB was found
to block Ca2+ entry mediated by overexpressed HTRP3
channels if the channels are activated through phospholipase C-coupled
receptors but not by diacylglycerol. The results were interpreted
to argue for a direct interaction between the InsP3
receptor and HTRP3 (40). In our 2-APB studies, we found that there is a
striking difference between the short and long term effects of 2-APB on
InsP3 signaling. While a 3-min incubation with 2-APB had
essentially no effect on the CCh-induced Ca2+ release (Fig.
1D), longer incubation times dramatically reduced the
CCh-stimulated release of store Ca2+ (data not shown).
Surprisingly, the 3-min exposure to 2-APB, which had no measurable
effect on CCh-induced Ca2+ release, dramatically inhibited
CCh-induced Ba2+ entry (Fig. 1D). Therefore, the
inhibitory effect of 2-APB on SOCE is not correlated with the ability
to reduce the initial depletion of the Ca2+ stores. This
suggests that 2-APB may be directly blocking plasma membrane
Ca2+ channels, a finding that is consistent with recent
observations that 2-APB blocks thapsigargin-induced Ca2+
entry, even in cells lacking functional InsP3 receptors
(38, 39).
Cyclical processes are part of the essence of life, and oscillations in
cell calcium are one example of such cyclical processes, one that seems
to be fundamental to the growth and differentiation of many cell types.
Cytosolic calcium oscillations can be induced by a variety of agonists.
Although the underlying mechanisms remain unclear, several models have
been proposed, with InsP3 being the center of attention (4,
50-52). In the current study, we set up to record oscillations induced
by carbachol (15 µM) in fura-2-loaded HEK-293 cells. We
observed that the addition of CCh in the continued presence of external
Ca2+ led to a sustained oscillatory pattern that is
truncated shortly after the removal of extracellular Ca2+
(data not shown). We set out to investigate the role of HTRP4 in
mediating the Ca2+ entry that is necessary for this
sustained oscillatory pattern. We found that in cells expressing an
antisense construct for HTRP4 (Fig. 6, right
panel), the oscillations induced by CCh were more similar to
the damped oscillations seen in nominally Ca2+-free medium
than those seen in the presence of Ca2+ (Fig. 6,
left panel). In Fig. 7, we plot the number of
cells versus the number of oscillations that occur in a cell
over a 10-min period. It is clear that most of the cells expressing
HTRP4 antisense are clustered in the leftmost
portion of the bar graph, indicating
that they are unable to sustain oscillations over a long period of
time. This appears to be a specific effect of the HTRP4 antisense,
since the expression of the HTRP3 antisense construct had no effect on
Ca2+ oscillations (Figs. 8 and 9). The determination that
HTRP4 is involved in regulating agonist-induced Ca2+
oscillations is a very important one. Although much of the work in the
Ca2+ signaling field focuses on the plateau phase of the
Ca2+ response following high dose agonist addition, it is
likely that under physiological conditions, cells see agonist at much
lower doses. Thus, the Ca2+ oscillations seen in response
to 15 µM CCh are likely to be physiologically more
important than the plateau response seen at 100 µM CCh.
Thus, the determination that HTRP4 is involved in supporting
Ca2+ oscillations is an extremely important one.
While recent publications from the Shuttleworth laboratory have
suggested that the Ca2+ entry pathway required for
sustained Ca2+ oscillations is regulated by arachidonic
acid, this hypothesis is based on the use of pharmacological inhibitors
(43), which can always raise questions of drug specificity. When we
observed that expression of HTRP4 antisense produced a block of the
CCh-induced Ca2+ oscillations, we recognized that this
provided an opportunity to test the arachidonic acid hypothesis
utilizing a molecular approach. We first determined that the addition
of arachidonic acid to control cells (HTRP4S cells) produced a
significant level of Ca2+ entry (Fig. 10A).
While the dose of arachidonic acid used in Fig. 10 is somewhat higher
than those used to define the arachidonic acid-regulated channel I(ARC)
(53), we observed that the exact dose for activating Ca2+
entry depended on the batch of arachidonic acid received from the
supplier. With some batches, we did see activation with doses in the
range of 5-10 µM, while with other batches, it took
30-40 µM to stimulate robust levels of Ca2+
entry. The variability is probably due to the known instability of
arachidonic acid in solution. However, no matter which batch was used,
we would first define a window of doses to use by titrating the amount
that would stimulate Ca2+ entry without giving a rise in
Ca2+ concentration when added to cells in a
Ca2+-free environment.
We next determined whether expression of HTRP4 antisense would block
the arachidonic acid-stimulated Ca2+ entry pathway. The
data in Fig. 10, A and B, illustrates that HTRP4AS expression blocks ~75% of the arachidonic acid-stimulated Ca2+ entry. This appears to be a specific effect of HTRP4
antisense, since the expression of HTRP3 antisense has no significant
effect on the arachidonic acid-stimulated Ca2+ entry (Fig.
10, C and D). Therefore, this finding provides
the first evidence for the involvement of one of the Trp proteins in
the arachidonic acid-regulated Ca2+ channel. It also
supports the hypothesis that the arachidonic acid-regulated pathway is
the one that provides the Ca2+ influx necessary for the
maintenance of sustained Ca2+ oscillations.
Among all mammalian Trp homologs, HTRP3 has been the most extensively
studied. Early studies showed that expression of full-length cDNA
encoding HTRP3 increased SOCE (14), and expression of a partial
cDNA fragment of HTRP3 in the antisense orientation significantly reduced SOCE (32), following store depletion with thapsigargin. Additional studies suggested that HTRP3 might also be activated by a
conformational coupling mechanism by interaction with InsP3 receptor (30, 40, 54). Other recent studies show that overexpressed HTRP3 can be activated by application of OAG (26, 45, 55). For
comparison, human Trp1 has been shown to be a nonselective cation
channel (15), which, when expressed in mammalian cells, is activated by
store depletion by thapsigargin as well as inositol 1,4,5-trisphosphate
(15). Our recent studies indicated that expression of HTRP1 antisense
in combination with HTRP3 antisense produces a further reduction
in thapsigargin-stimulated Ca2+ entry in addition to that
seen with expression of HTRP3 alone (32). When HTRP1 is expressed in
Sf9 insect cells, it may (15) or may not (27) be sensitive to
depletion of internal Ca2+ stores. In addition,
overexpression of HTRP1 leads to channels that are sensitive to
stimulation by carbachol (14) but insensitive to OAG (45, 56). The
conclusion from the present study is that HTRP3 proteins as well as
HTRP4 proteins play a role in forming the endogenous channels, which
are activated by OAG. In contrast, endogenous HTRP1 proteins appear to
play no role in the activation of channels by OAG. As shown in Fig. 5,
in HEK-293 cells expressing antisense cDNA constructs for HTRP3 or
HTRP4, Sr2+ entry induced by OAG was dramatically reduced.
In contrast, Sr2+ entry induced by OAG was unaffected in
cells expressing antisense cDNA for HTRP1. This provides useful
evidence that OAG is indeed acting on the HTRP3 and HTRP4 but not HTRP1 channels.
We should point out that these results differ from earlier results, in
that OAG stimulates Sr2+ entry via endogenous channels in
our HEK-293 cells, while in other studies OAG had no effect on parental
HEK-293 cells but did stimulate Sr2+ entry in cells
overexpressing HTRP3 (26, 40). This difference in results is probably
due to the variation of characteristics of cultured lines grown
in different laboratories. It is widely known for cell lines such as
PC12 cells that there can be dramatic differences in cultures carried
for a number of years in different laboratories. Different laboratories
utilize serum from different sources, and other more subtle differences
in tissue culture procedure also exist. Since our previous work
demonstrates dramatic clone-to-clone variation in levels of SOCE within
our HEK-293 cell population, it is not difficult to imagine that
culture conditions in one laboratory might offer a competitive
advantage for growth of some clones over others, thereby leading to
high endogenous levels of SOCE in HEK-293 populations in one laboratory
versus low endogenous levels of SOCE in HEK-293 populations
in another laboratory. Thus, while we see substantial SOCE via
endogenous channels, some laboratories see very little endogenous SOCE.
The cell lines with the low endogenous SOCE have been popular for
electrophysiological studies of the overexpression of Trp proteins,
since there is a low basal current, and an increment in current due to
expression of Trp can be easily observed. One possible drawback to the
use of cell lines with low endogenous currents might exist if the
transfected Trp needs to form heterotetramers with another Trp isoform
to show the correct channel characteristics. Cell lines having
vanishingly small endogenous currents may not express the full
complement of endogenous Trps needed to form the true physiological
configuration of the channel. Thus, the characteristics of the channel
formed by overexpressed Trp may be misleading in terms of the true role
of that Trp protein. On the other hand, the HEK-293 cell line that we
use is perfect for our approach. It has substantial levels of both SOCE
and receptor-activated Ca2+ entry so that we can study the
role of endogenous Trps in either Ca2+ entry pathway.
One extremely important question in the Trp field is why there are so
many conflicting reports about the role of various Trp isoforms in
mediating either store-operated or agonist-stimulated Ca2+
entry pathways. One possible reason is the use of different cellular expression systems with the possibility of differing genetic
backgrounds in terms of the types of endogenous Trp subunits available
for combination with the overexpressed Trp isoform. It is also possible that overexpression of Trps leads to the study of homotetramers that
may have very different properties from endogenous channels that may
normally be assembled from heterogeneous subunits. In addition to these
problems, we published a recent report that cautions about the study of
low numbers of stable clones expressing Trps (36). That paper estimates
the probability that clone-to-clone variation in the population of
HEK-293 cells contributes to the misinterpretation of data in clones
overexpressing Trp isoforms. In light of these potential problems, we
decided that a strategy that reduces the level of endogenous Trps would
be a more desirable way to study the functional role of individual Trp isoforms.
In summary, our results define some interesting similarities and
differences in endogenous Ca2+ entry pathways activated by
thapsigargin, CCh, or OAG in HEK-293 cells. The endogenous
Ca2+ entry pathways activated by store depletion or by
carbachol can be inhibited by 2-APB, while the endogenous
Ca2+ entry pathway stimulated by OAG is not inhibited by
2-APB. The Ca2+ entry pathway activated by store depletion
does not require HTRP4 as a channel subunit. On the other hand, the
Ca2+ entry pathways activated by either carbachol or OAG do
require HTRP4 expression. The CCh- and OAG-stimulated,
HTRP4-dependent Ca2+ entry pathways appear not
to be one and the same, since 2-APB inhibits most of the CCh-stimulated
pathway but none of the OAG-stimulated pathway. HTRP3 proteins appear
to be more widely used than HTRP4 proteins, since they are required in
both the store-operated channels and the OAG-stimulated channels. In
contrast, HTRP1 appears to be a functional subunit of endogenous
store-operated channels but not OAG-stimulated channels. And of greater
physiological importance, HTRP4 plays a role in both the
Ca2+ entr
,
TOP
ABSTRACT
INTRODUCTION
