Oncogenic potential of TASK3 (Kcnk9) depends on
Lin Pei*†, Ofer Wiser‡, Anthony Slavin*, David Mu§, Scott Powers§, Lily Yeh Jan‡, and Timothy Hoey*
*Tularik Inc., 1120 Veterans Boulevard, South San Francisco, CA 94080;‡Department of Neuroscience, University of California, 533 Parnassus Avenue,
San Francisco, CA 94143; and§Tularik Inc., Genomics Division, 266 Pulaski Road, Greenlawn, NY 11740
Contributed by Lily Yeh Jan, April 24, 2003
TASK3 gene (Kcnk9) is amplified and overexpressed in several
types of human carcinomas. In this report, we demonstrate that a
point mutation (G95E) within the consensus K?filter of TASK3 not
only abolished TASK3 potassium channel activity but also abro-
resistance to apoptosis, and promotion of tumor growth. Further-
more, we provide evidence that TASK3G95Eis a dominant-negative
mutation, because coexpression of the wild-type and the mutant
TASK3 resulted in inhibition of K?current of wild-type TASK3 and
its tumorigenicity in nude mice. These results establish a direct link
between the potassium channel activity of TASK3 and its onco-
genic functions and imply that blockers for this potassium channel
may have therapeutic potential for the treatment of cancers.
channels, whose structure consists of two-pore forming regions
flanked by four transmembrane domains (1, 2). Like other
two-pore domain members, these channels show little time or
voltage dependence; thus, they have characteristics of leaky K?
channels, generating background currents that contribute to
membrane potential and the regulation of cell excitability. The
4), neurotransmitters (5, 6), as well as extracellular pH in the
among the normal tissues except in the brain, where high levels
expression of TASK3 were detected (7–9). The physiological
functions of TASK channels are largely unknown, though their
roles in the regulation of breathing (12, 13), aldosterone secre-
tion (5) and anesthetic-mediated neuronal activity (14) have
been proposed. Recently, we showed that TASK3 is amplified in
10% of breast cancers and is overexpressed at a higher frequency
of breast, lung, colon, and metastatic prostate cancers (15),
suggesting that TASK3 may play a role in pathogenesis of some
Is the dysregulated expression of TASK3 in tumor cells a
consequence of their abnormal growth or is this K?channel
involved in promoting tumor growth? To begin to answer this
question, we created an inactivating mutation of TASK3. We
report here that TASK3G95Eis a dominant-negative mutation
that abolishes not only TASK3 K?channel activity but also its
oncogenic functions. These results provide molecular basis for
developing specific blockers for this K?channel in the treatment
ASK (TWIK-related acid-sensitive K?channels) channels
are members of the two-pore domain family of potassium
Materials and Methods
Plasmids and Mutagenesis. The coding region of TASK3 was
cloned at BamHI and EcoRI sites of the pLPC retroviral
expression vector (15) to generate pLPC-TASK3. Site-directed
mutagenesis was performed to change Gly-95 to Glu to create
pLPC-TASK3G95E, by using QuickChange site-directed mu-
tagenesis kit (Stratagene) according to the manufacturer’s pro-
tocol. pTracer-TASK3G95Ewas generated by excising
TASK3G95Efrom pLPC vector and cloned at KpnI and XbaI sites
of pTracer (Invitrogen). To generate TASK3 expression vector
for expression in Xenopus oocytes, the wild-type and mutant
forms of TASK3 were excised from pLPC vector and cloned at
BamHI and EcoRI sites of the pGEM3ZHEM vector.
Electrophysiological Recordings from Xenopus Oocytes. Capped
cRNAs were synthesized in vitro from linearized plasmids by
using T7 RNA polymerase (Epicenter, Madison, WI). Stage V
and VI oocytes were defolliculated manually and kept in stan-
dard ND 96 solution (96 mM NaCl?2 mM KCl?1.8 mM CaCl2?2
mM MgCl2?5 mM Hepes, pH 7.4) supplemented with penicillin
(100 units?ml), and streptomycin (100 ?g?ml), at 16°C. Oocytes
were injected with 5 ng per oocyte of the appropriate RNA
solution, and electrophysiology studies were carried out 24–48 h
after injection. Membrane currents were recorded by two-
electrode voltage clamp. Currents were filtered at 1 kHz, digi-
tally sampled at 5 kHz with a Digidata Interface (Axon Instru-
ments, Foster City, CA). Recording and data analysis were
performed by using pCLAMP software (Axon Instruments).
Experiments were carried out at room temperature, and solu-
tions were applied by a gravity-driven perfusion system. High
[K?] solution contained 40 mM KCl (38 mM NaCl of ND96
solution was replaced by KCl).
Cell Culture and Transfection. Partially transformed mouse embry-
onic fibroblast cell line C8 was grown in DMEM?F12 (50:50)
supplemented with 10% FBS. Retroviral packaging cell line
Phoenix cells and lung carcinoma cell line Ben were cultured in
DMEM with 10% FBS. Transfections were performed by using
Lipofectamine 2000 reagent (Invitrogen) following the manu-
facturer’s instructions. Retroviral infection was carried out by
using described protocols (16). Transfectants were selected in
puromycin (2 ?g?ml) or Zeocin (500 ?g?ml) for cells transfected
with pLPC or pTracer expression vector, respectively. After 2–3
weeks in selective media, clones were pooled and analyzed for
Immunofluorescence. Cells grown in chamber slides were fix in
methonal?aceton (1:1) at ?20°C for 10 min, and permeablized
in 0.1% saponin and 0.1% BSA at room temperature for 10 min.
Cells were then incubated with affinity purified anti-peptide
antibody to TASK3 (REEEKLKAEEIRIKGKYNISSEDYRQ)
(10 ?g?ml) at room temperature for 1 h, washed three times with
PBS, and then incubated with anti-rabbit FITC (Pierce) diluted
1:100 for 1 h at room temperature. After three washes in PBS,
Cells. Cells were placed on the stage of an inverted microscope
and continuously superfused with a solution containing 130 mM
NaCl, 5 mM KCl, 10 mM Hepes, 10 mM glucose, 2 mM CaCl2,
and 1 mM MgCl2(pH 7.4) with NaOH. Patch electrodes were
pulled from borosilicate glass to resistances of 2–6 megaohms.
Internal solution contained 140 mM KCl, 4 mM NaCl, 1 mM
MgCl2. 0.5 mM CaCl2, 10 mM Hepes, 10 mM EGTA, 3 mM
Abbreviations: TNF, tumor necrosis factor; FLIPR, fluorescence imaging plate reader.
†To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
www.pnas.org?cgi?doi?10.1073?pnas.1232448100 PNAS ?
June 24, 2003 ?
vol. 100 ?
no. 13 ?
ATP, and 0.3 mM GTP (pH 7.4) with NaOH. Recordings were
obtained by using an Axopatch 200A amplifier and digitized with
a Digidata 1200 A?D converter (Axon Instruments). Cells were
held at a membrane potential of ?60 mV and current evoked by
voltage steps from ?80 mV to ?60 mV in 100-ms duration.
Current responses filtered at 2 kHz and digitized at 10 kHz were
stored for later analysis using PCLAMP software.
Membrane Potential Assay. The functional expression of TASK3
as a potassium channel in mammalian cells was determined by
using a membrane potential sensitive dye on fluorescence im-
aging plate reader (FLIPR) (Molecular Devices). Cells were
plated at 2 ? 104per well on black 384-well clear-bottom plates
and incubated at 37°C overnight. Medium was then removed and
the cells incubated with 50 ?l of the membrane potential dye in
assay buffer (Hanks’ balanced salt solution plus 20 mM Hepes,
pH 7.4) at 37°C for 30 min. Assays were carried out in the FLIPR
at an excitation wavelength of 488 nm and by using an emission
filter provided by Molecular Devices specifically for the fluo-
for 10 s, 25 ?l of a 3? concentration of KCl (270 mM) in assay
for 90 s. Assays were carried out at room temperature.
Apoptosis and Cell Proliferation Assays. To induce apoptosis, cells
were treated with 1 ng?ml tumor necrosis factor (TNF) for 16 h.
Apoptosis was measured by DNA fragmentation assay using Cell
Death Detection ELISA (Roche Diagnostics, Mannheim, Ger-
many) following the manufacturer’s instructions. For cell pro-
liferation assay, cells were grown in media containing 1% FBS
for 48 or 72 h, and the rate of cell proliferation was measured by
using CellTiter Aqueous One Solution Cell Proliferation Assay
Tumorigenicity Assays in Nude Mice. Cells from various stable cell
lines were harvested and resuspended in PBS. The cell suspen-
sion (106cells per injection) was injected s.c. into athymic nude
mice. Mice were observed weekly for the visual appearance of
tumors at injection sites, and tumor sizes were measured each
week. At the end of the third week, mice were killed, and the
tumors were excised and cultured.
Point Mutation Within the GYG K?Selective Filter Results in an
Inactive TASK3 Channel. We used site-directed mutagenesis to
change glycine at amino acid 95 within the conserved GYG K?
filter (17) to glutamate and tested the potassium channel activity
of the mutant TASK3 in Xenopus oocyte and in mammalian
cells. When wild-type TASK3 cRNA was injected in oocyte, it
induced a K?current of 17 ?A at ?120 mV (in 40 mM KCl, n ?
5), whereas injection of TASK3G95Edid not induce any measur-
able K?current (n ? 3) (Fig. 1A). When wild-type TASK3 was
coinjected with mutant TASK3G95E, macroscopic current am-
plitudes were reduced from 5.2 ?A (wild type alone, at ? 30 mV,
in 2 mM KCl, n ? 10) to 2.05 ?A (Fig. 1B). On the other hand,
coexpression of TASK3G95Ewith the closely related channel
TASK-1 did not affect TASK-1 current (data not shown). These
results suggest that TASK3G95Eis an inactive K?channel when
expressed in Xenopus oocytes and that the expression of the
mutant TASK3 inhibits the channel activity of the wild-type
To determine the effect of this point mutation on TASK3
channel activity in mammalian cells, we generated stable cell
any measurable K?current. (B) Expression of TASK3G95Einhibits K?current of the wild-type TASK3. Shown are whole cell currents in response to voltage ramp
(Left) Representative traces of whole cell currents recorded in response to 175-ms voltage ramp between ?120 and ?40 mV. (Right) Average currents (?SE) of
all of the oocytes tested in each group. P value was determined by a Student’s t test. The extracellular [K?] concentration was 40 mM (A) and 2 mM (B).
www.pnas.org?cgi?doi?10.1073?pnas.1232448100Pei et al.
lines in a partially transformed mouse embryonic fibroblast C8
cells that overexpress either the wild-type or mutant form of
TASK3. We verified cell surface expression of the channel
protein by indirect immunofluorescence assay (Fig. 2A). To test
whether these cell lines express a functional channel, we used a
membrane potential sensitive dye (Figs. 4 and 5, which are
published as supporting information on the PNAS web site,
www.pnas.org) to measure the change in fluorescence signals on
depolarization of cells by KCl. Expression of wild-type TASK3
resulted in increase in fluorescence signal (?13,000 counts) on
addition of 90 mM KCl (Fig. 2B). In contrast, cells overexpress-
ing TASK3G95Eshowed no change in fluorescence signal com-
the wild-type and the mutant TASK3 were cotransfected into
cells, the fluorescence signals were reduced 75% compare with
wild-type TASK3 alone (Fig. 2B). We also determined the
functional channel expression in stable cell lines by whole cell
current recording. Cell membrane potential was held at ?60 mV
and stepped to various potentials for 100-ms duration. In cells
stably transfected with the mutant form TASK3, only very small
currents of ?50 pA were recorded (Fig. 2C). In cells stably
expressing wild-type TASK3, the same voltage step produced
large instantaneous and noninactivating currents (Fig. 2C).
These results agree with the results in Xenopus oocytes, suggest-
ing that TASK3G95Eis a dominant negative mutant.
TASK3G95ELoses Oncogenic Potential and Inhibits Oncogenic Function
of the Wild-Type TASK3. To determine whether the oncogenic
activity of TASK3 is correlated with its function as a potassium
channel, we tested the effects of overexpression of wild-type
TASK3 and TASK3G95Emutant on cell proliferation in low
serum, resistance to TNF-induced apoptosis as well as tumori-
genicity in nude mice.
To elucidate the role of TASK3 in cell proliferation, mouse
embryonic fibroblast cells expressing either the wild-type or
mutant TASK3 were cultured in 1% serum for 72 h, and the
growth rate was quantified by using 3-(4,5-dimethylthiazol-2-yl)-
(MTS) assay. Cells expressing either vector alone or the
TASK3G95Emutant proliferate in a much slower rate comparing
to cells expressing wild-type TASK3, resulting in a 2-fold in-
crease in cell number in TASK3 overexpressing cells after 72 h
in culture (Fig. 3A). These results suggest that overexpression of
TASK3 confers growth advantage in low serum and that this
function requires a functional potassium channel.
To test whether overexpression of TASK3 would result in
resistance to apoptosis, we treated embryonic fibroblast cells
fragmentation ELISA assay. Overexpression of wild-type
TASK3 led to a 50% reduction of TNF-induced apoptosis,
whereas expression of the mutant TASK3 had no effects (Fig.
expression of the wild-type and mutant TASK3 by indirect immunofluorescence. (B) K?channel activity of the stable cell lines expressing wild-type or mutant
TASK3 individually or together measured by FLIPR assay. Twenty thousand cells were plated in a 384-well plate and grown for 24 h. Cells were loaded with the
voltage-sensitive fluorescence dye for 30 min and assayed on FLIPR in assay buffer containing 90 mM [K?]. (C) TASK3 whole cell currents were recorded from
the stable cell line expressing either wild-type or mutant TASK3. Pipette and bath solutions contained 140 and 5 mM K?, respectively. Membrane potential was
held at ?60 mV and stepped from ?80 to ?60 mV in 10-mV increments.
Effect of the point mutation within the consensus K?filter sequence on TASK3 currents expressed in mouse embryonic fibroblast cells. (A) Cell surface
Pei et al.
June 24, 2003 ?
vol. 100 ?
no. 13 ?
3B). This result suggests that TASK3 confers partial resistance
to TNF-induced apoptosis and that the K?channel activity of
TASK3 is required for this function.
To determine the effect of TASK3G95Emutation on TASK3
tumorigenicity, we injected athymic nude mice with C8 cells
expressing vector alone, wild-type or mutant form of TASK3, as
well as C8 cells expressing both the wild-type and the mutant
TASK3 together. All animals injected with C8 cells expressing
wild-type TASK3 developed large tumors within 2 weeks. Al-
though mice injected with cells expressing vector alone or cells
expressing mutant TASK3 also developed tumor after 3 weeks,
the size of the tumors are much smaller in comparison to tumors
expressing wild-type TASK3 (Fig. 3C). Furthermore, cell coex-
pressing both the wild-type and mutant form of TASK3 only
developed small tumors as seen in the vector control (Fig. 3C).
To determine whether the tumor cells continue to express a
function K?channel, the tumors were excised, cultured for 1
week and then assayed for K?channel activity by FLIPR. The
tumor cells overexpressing the wild-type TASK3 maintain a
functional K?channel, whereas no channel activity was detected
for tumor cells overexpressing the mutant TASK3 (Fig. 3D).
These results suggest that K?channel activity is required for
TASK3 to promote tumor formation in nude mice and that
TASK3G95Eacts as a dominant-negative mutant for expression
of the mutant protein inhibited tumorigenicity of the wild-type
To test whether expression of TASK3G95Ewould affect func-
tions of endogenous TASK3, we stably transfected TASK3G95E
into human lung carcinoma cell line Ben that overexpresses
TASK3 (Fig. 6, which is published as supporting information on
the PNAS web site). As shown in Fig. 3E, expression of
TASK3G95Esignificantly reduced Ben-cell proliferation within
48 h, suggesting that TASK3G95Einterferes with endogenous
TASK3 function to inhibit proliferation of these lung carcinomas
K?channel activity plays important roles in the signaling path-
ways that regulate cell proliferation and apoptosis. Inhibition of
cell proliferation in normal human lymphocytes (18–21), human
melanoma cells (22, 23), small cell lung cancer (24), breast (25),
induced by treatment of cells with 1 ng/ml TNF for 16 h. Cell apoptosis is expressed as enhancement factor (fold over untreated cells in each group). Data are
representative of three independent experiments (n ? 3 for each experiment). (C) Tumorigenicity in nude mice. Mouse embryonic fibroblast cells stably
transfected with vector alone, wild-type, or mutant TASK3 individually or together were injected (106cells per injection) into nude mice (two experiments, n ?
10 for each cell line). Tumor formation was monitored. The tumors were measured each week at their longest points by using a caliper. The bar in the graph
plate, and the cell proliferation rated was determined as in A.
The inactivating mutation of TASK3 abolished its oncogenic functions. (A) Cell proliferation in low serum. Cell growth rate is expressed as absorbance
www.pnas.org?cgi?doi?10.1073?pnas.1232448100Pei et al.
andprostate(26)cancercells.TheroleofK?channelsincellular Download full-text
proliferation has been thought to be indirect, by either the
possible influence of K?channels on the intracellular Ca2?
concentration (27) or, alternatively, the control of cell volume
via K?channels (28). Recent studies have shown that enhance-
ment of K?current is directly involved in apoptosis (29, 30) and
oncogenesis (31). In mouse neocortical neurons, treatment of
cell with the K?ionophore valinomycin or KATPchannel opener
cromakalin induced apoptosis (29), whereas inhibition of out-
ward K?currents with tetraethylammonium, but not with block-
ers of Ca2?, or Cl?reduced apoptosis (30). The human ether-
of tumor cell lines (32), and HERG conductance markedly
promotes H2O2-induced apoptosis of various tumor cells (33).
The expression of rEAG favors tumor progression when the
transfected cells are injected in nude mice, and overexpression
of rEAG K?in NIH 3T3 cells induces significant features
characteristic of malignant transformation (31). Taken together,
these studies suggest that K?channels play a crucial role in
Recently, we showed that TASK3 (Kcnk9) is amplified in 10%
of breast cancers and is overexpressed at a higher frequency of
breast, lung, colon, and metastatic prostate cancers (15). This
was the first time that a genetic modification in K?channels has
here expended the results in this previous study and established
a direct link between K?channel activity of TASK3 and its
oncogenic function. We have characterized a dominant-negative
mutation of TASK3 that not only abolishes its K?channel
activity but also abrogates its oncogenic functions. Our data
showed that wild-type TASK3 confers growth advantage to cells
overexpress this K?channel, whereas the inactivating mutant
had no effect on cell growth, suggesting that TASK3 K?channel
activity is directly involved in cell proliferation. We provided
evidence that TASK3 overexpression resulted in partial resis-
tance to TNF-induced apoptosis, a mechanism that causes
cancer cells to escape host immune defenses. However, cells
expressing the mutant TASK3 are sensitive to TNF-induced
apoptosis, suggesting the requirement for a functional K?chan-
nel. Most importantly, we demonstrated the TASK3 mutant not
only was incapable of promoting tumor growth in nude mice, but
also inhibited the tumor promotion function of the wild-type
TASK3. We cannot at the present completely rule out the
possibility that this dominant-negative mutant of TASK3 could
interfere with the activity of other TASK channels through
heterodimerization (34). A previous study showed the same
TASK3 mutant did not have effect on TASK1 current when
coexpressed in Xenopus oocytes (35), and we confirmed this
observation in this study (data not shown). In addition, TASK1
gene amplification or overexpression was not detected in tumor
samples where TASK3 was amplified (15).
In summary, the experimental results presented in this study
provide the biological basis for the development of TASK3
antagonists that should potentially inhibit growth of certain
human carcinomas and therefore would represent a new family
of anticancer therapeutics.
We thank Chris Mathes (Axon Instruments) for help with whole cell
recording in mammalian cells.
1. Goldstein, S. A., Bockenhauer, D., O’Kelly, I. & Zilbergerg, N. (2001) Nat. Rev.
Neurosci. 2, 175–184.
2. Lesage, F. & Lazdunski, M. (2000) Am. J. Physiol. 279, F973–F801.
3. Meadows, H. J. & Randall, A. D. (2001) Neuropharmacology 40, 551–559.
4. Patel, A. J., Honore, E., Lesage F., Fink, M., Romey, G. & Lazdunski, M.
(1999) Nat. Neurosci. 2, 422–426.
5. Czirjak, G., Fischer, T., Spat, A. & Enyedi, P. (2000) Mol. Endocrinol. 14,
6. Talley, E. M. & Bayliss, D. A. (2002) J. Biol. Chem. 277, 17733–17742.
7. Duprat F., Lesage, F., Fink, M., Reyes, R., Heurteaux, C. & Lazdunski, M.
(1997) EMBO J. 16, 5464–5471.
8. Chapman, C. G., Meadows, H. J., Godden, R. J., Campbell, D. A., Duckworth,
M., Kelsell, R. E., Murdock, P. R., Randall, A. D., Rennie, G. I. & Gloger, I. S.
(2000) Mol. Brain Res. 82, 74–83.
9. Kim, G., Bang, H. & Kim, D. (2000) J. Biol. Chem. 275, 9340–9347.
10. Rajan, S., Wischmeyer, E., Liu, G. X., Preisig-Muller, R., Daut, J., Karschin,
A. & Derst, C. (2000) J. Biol. Chem. 275, 16650–16657.
11. Lopes, C. M., Zilbergerg, N. & Goldstein, S. A. (2002) J. Biol. Chem. 276,
12. Hartness, M. E., Lewis, A., Searle, G. J., O’Kelley, I., Peers, C. & Kemp, P. J.
(2002) J. Biol. Chem. 276, 26499–26508.
13. Buckler K. J., Williams, B. A. & Honore, E. (2000) J. Physiol. 15, 135–142.
14. Sirois, J. E., Lei, Q., Talley, E. M., Lynch, C., III, & Bayliss, D. A. (2000)
J. Neurosci. 20, 6347–6354.
15. Mu, D., Chen, L., Zhang, X., See, L. H., Koch, C. M., Yen, C., Tong,
J. J., Spiegel, L., Nguyen, K. C., Servoss, A., et al. (2003) Cancer Cell 3,
16. Pei, L. Peng, Y., Yang, Y., Ling, X. B., Van Eyndhoven, W. G., Nguyen,
K. C., Rubin, M., Hoey, T., Powers, S. & Li, J. (2002) Cancer Res. 62,
17. Heginbotham, L., Lu, Z., Abramson, T. & MacKinnon, R. (1994) Biophys. J.
Mol. Immunol. 27, 1259–1268.
19. Lin, C. S., Boltz, R. C., Blake, J. T., Nguyen, M., Talento, A., Fischer, P. A.,
Springer, M. S., Sigal, N. H., Slaughter, R. S., Garcia, M. L., et al. (1993) J. Exp.
Med. 177, 637–645.
20. Rader, R. K., Kahn, L. E., Anderson, G. D., Martin, C. L., Chinn, K. S. &
Gregory, S. A. (1996) J. Immunol. 156, 1425–1430.
21. Jensen, B. S., Odum, N., Jorgensen, N. K., Christophersen, P. & Olesen, S. P.
(1999) Proc. Natl. Acad. Sci. USA 96, 10917–10921.
22. Nilius, B. & Wohlrab, W. (1992) J. Physiol. 445, 537–548.
23. Lepple-Wienhues, A., Berweck, S., Bohmig, M., Leo, C. P., Meyling, B., Garbe,
C. & Wiederholt, M. (1996) J. Membr. Biol. 151, 149–157.
24. Pancrazio, J. J., Tabbara, I. A. & Kim, Y. I. (1993) Anticancer Res. 13,
25. Woodfork K. A., Wonderlin, W. F., Peterson, V. A. & Strobl, J. S. (1995)J. Cell.
Physiol. 162, 163–171.
26. Skryma, R. N., Prevarskaya, N. B., Dufy-Barbe, L., Odessa, M. F., Audin, J. &
Dufy, B. (1997) Prostate 33, 112–122.
27. Santella, L. (1998) Biochem. Biophys. Res. Commun. 244, 317–324.
28. Rouzaire-Dubois, B. & Bubois, J. M. (1998) J. Physiol. 510, 93–102.
29. Yu, S. P., Yeh, C. H., Sensi, S. L., Gwag, B. J., Canzoniero, L. M., Farhangrazi,
Z. S., Ying, H. S., Tian, M., Dugan, L. L. & Choi, D. W. (1997) Science 278,
J. Neurochem. 73, 933–941.
31. Pardo, L. A., del Camino, D., Sanchez, A., Alves, F., Bruggemann, A., Beckh,
S. & Stuhmer, W. (1999) EMBO J. 18, 5540–5547.
32. Bianchi, L., Wible, B., Arcangeli, A., Taglialatela, M., Morra, F., Castaldo, P.,
Crociani, O., Rosati, B., Faravelli, L., Olivotto, M. & Wanke, E. (1998) Cancer
Res. 58, 815–822.
33. Wang, H., Zhang, Y., Cao, L., Han, H., Wang, J., Yang, B., Nattel, S. & Wang,
Z. (2002) Cancer Res. 62, 4843–4848.
34. Czirjak, G. & Enyedi, P. (2002) J. Biol. Chem. 277, 5426–5432.
35. Karschin, C., Wischmeyer, E., Preisig-Muller, R., Rajan, S., Derst, C.,
Grzeschik, K-H., Daut, J. & Karschin, A. (2001) Mol. Cell. Neurosci. 18,
Pei et al.
June 24, 2003 ?
vol. 100 ?
no. 13 ?