Silencing the Activity and Proliferative Properties of the Human EagI Potassium Channel by RNA Interference*

EagI potassium channels are natively expressed in the mammalian brain as well as in many cancer cell lines and tumor tissues. The role of EagI in malignant transformation has been suggested by several experiments, but the lack of specific EagI inhibitors has made it difficult to examine the influence of the channel on oncogenesis and its potential as a therapeutic target. We have used short interfering RNA to test the effects of EagI reduction on the behavior of tumor cells in vitro. By generating and optimizing an EagI-specific short interfering RNA system, we were able to study the effects of EagI depletion on several cancer cell lines that endogenously express this protein. We show here that our short interfering RNA sequences act specifically on EagI, reproducibly induce a significant decrease in the proliferation of tumor cell lines, and do not trigger any observable nonspecific responses.

Ether à go-go (Eag) potassium channels were first detected in Drosophila melanogaster (1)(2)(3). In vertebrates, the Eag family is formed by eight members distributed in three subfamilies (e.g. Refs. 4 and 5). The prototypic channel (EagI) is preferentially expressed in the brain where two alternatively spliced variants (Eag1a and Eag1b) have been detected (6). It encodes a voltage-gated potassium channel expressed in most brain regions, except for the thalamus (7). Despite the efforts of several laboratories, the physiological role of EagI in the brain remains elusive. Those efforts led nevertheless to the characterization of some interesting properties of the channel, like its interaction with Ca 2ϩ -calmodulin (8 -10), the strong influence of the external magnesium concentration on the channel kinetics (11), or the steep dependence of the time constant of activation on the prepulse potential (12). Thus, the activation of the channel is slower the more hyperpolarized the starting membrane potential is. This particular property has proven very useful when attempting to identify the channel in native tissues and has been postulated to be important for the physiological function of EagI, because the properties of the current in a particular cell will not only depend on the actual stimulus, but also be determined by the average membrane potential previous to the stimulus. hEag1 mRNA has also been detected in many cancer cell lines, although the corresponding normal tissue exhibits no such expression. Furthermore, hEag1 protein expression is detected in up to 75% of tumor tissues derived from different organs. 2,3 This is especially significant if one takes into account that the corresponding normal tissues were in all cases negative for EagI expression, except in some restricted populations.
EagI overexpression in heterologous systems leads to a phenotype compatible with malignant transformation of the cells, like higher proliferation rate, loss of contact inhibition, and generate larger and more aggressive tumors than control cells when implanted subcutaneously into severe combined immunodeficient mice (6). These data suggest that EagI may not only represent a potential marker for the diagnosis and prognosis of cancer but also that it may contribute to the formation and progression of cancers. Indeed, it has been shown for breast cancer cells (13) and for melanoma cells (14) that the use of EagI blockers leads to a reduction in proliferation. In those reports, the antihistamine astemizole and the tricyclic antidepressant imipramine were used, respectively. Although both are efficient EagI blockers, they are rather nonspecific (15). Not only the study of the physiological functions of EagI, but also the characterization of the potential usefulness of EagI as a cancer target still require the use of potent and specific inhibitors of the channel.
RNA interference is a relatively new technology with the potential to investigate molecular physiology by means of specific and effective knock down of a particular gene both in vitro and in vivo (16). Short interfering RNAs are 21-23 base pairs double stranded RNA molecules with 2-nucleotide 3Ј-overhangs (in practice modified to increase stability). When entering the cell, they bind to the RISC complex (RNAinitiated silencing complex), which promotes the cleavage of the RNA containing that particular sequence. The mechanism is very effective, leading to virtual abolition of the expression of the target gene, but its most interesting feature is the virtually absolute specificity that can be reached. The major drawback of the technique is the possible occurrence of off-target effects, either mediated by the siRNA itself (for example acting as a micro-RNA) or by the nonspecific stress response of the cell upon uptake of double stranded RNA (see Refs. 17 and 18). A number of independent control experiments need to be performed to minimize the risk of such off-target effects, a summary of which has been recently agreed on by an expert panel (19).
In this study, we examined the effectiveness of siRNA 4 -mediated EagI silencing as a tool to reduce its tumorigenic properties in vitro. We initially used an EagI-transfected cell line (HEK293) to carefully characterize the action (with highest attention to specificity) of RNA interference in a heterologous system. Importantly, a panel of control experiments (some of them specifically designed for this study) suggest that the siRNA acts very specifically, i.e. only on EagI, without inducing spurious responses. We show not only a reduction of EagI at the mRNA, protein, and functional levels using electrophysiology but also the effects EagI silencing has on cell proliferation. Furthermore, we observe a similar behavior in several cancer cell lines that endogenously express EagI, derived both from the brain and peripheral tissues. Although it has generally been described that the silencing of most genes requires 2 days or more (20), our results indicate that EagI mRNA degradation starts within 4 h of the siRNA transfection and maximal inhibition occurs at 8 h in all three cell lines studied. In agreement with this data, we observed maximal protein and functional knock down at later time points. As the first report of siRNA used as a tool to reduce EagI expression, our data reinforce the concept of an implication of EagI in tumor cell proliferation and indicate a causal link between eag1 expression and increased proliferation. It could therefore serve as a basis for the development of anti-EagI siRNAs as a therapeutic strategy for tumors, while representing a useful tool for the exploration of EagI functional role in normal tissue.
siRNA Transfections-Four different siRNAs were designed against both EagI and EagII with accession numbers NM_172362 and NM_ 172376, respectively, using the HiPerformance siRNA Design Algorithm. The algorithm uses a neuronal network approach to design the siRNAs (21) and discards those sequences with high similarity to regions of other genes. Areas of the target sequence with known single nucleotide polymorphisms were avoided. The generation of multiple functional siRNAs against EagI allowed independent confirmation of the phenotypic effects (18). There are no reports in which different phenotypes induced by the two known splice variants of EagI have been described. Only Eag1a has been studied in terms of cell proliferation, and we therefore did not attempt to design siRNAs that could discriminate between the two EagI splice variants.
The following siRNAs were used: siRNAs directed against hEAG1 with NM_172362 as target sequence, Kv10.1 nt 1509 -1529, Kv10.  [1][2][3]. These molecules have the same base composition and closest structure to the active siRNA. The sense strand of the annealed siRNA duplexes was modified with two dT at the 3Ј-ends, the antisense strand with the sequencecorresponding desoxynucleotides to stabilize the siRNAs against degradation (22).
All siRNAs were synthesized by Qiagen and annealed prior to use, except the commercial Negative Control #1 and the human GAPDH siRNA (Ambion), which we used as negative and positive controls, respectively. The cells were incubated with the siRNA and the transfection reagent for between 4 and 24 h. Cells were harvested for the experiments at different time points after the start of transfection. Additionally cells treated only with OptiMEM and Oligofectamine, Lipofectamine, or Dharmacon Transfection Reagent 2 were included as controls.
Real Time PCR-Total RNA was extracted with the RNeasy mini kit (Qiagen) following the manufacturer's recommendations and stored at Ϫ80°C until used. 5 g of total RNA were used for cDNA synthesis using SuperScript (Invitrogen) or M-MLV reverse transcriptase (Promega). Real time PCR was performed in triplicates on 100 ng of cDNA using the two-step TaqMan Core Kit (Applied Biosystems) in an ABI 7700 (PerkinElmer Life Sciences) sequence analyzer.
Western Blot-To obtain cell lysates, cultures were washed with phosphate-buffered saline and lysed in phosphate-buffered saline containing 1% SDS and protease inhibitor tablets (Roche Diagnostics). The crude cell lysate was homogenized 10 times through a 23-gauge injection needle, sonicated to disrupt DNA, and subsequently centrifuged for 10 min at 14,000 rpm in a benchtop centrifuge. The protein concentration was determined with the Bio-Rad protein assay (Bio-Rad). The proteins (30 g) were denatured and separated on 10% SDS-PAGE and transferred to nitrocellulose filters at pH 10. The membranes were blocked (2% casein, Roche Diagnostics) for 1 h, incubated with polyclonal anti-EagI antibody in 1% casein (23) for 2 h, washed (3 ϫ 10 min Tris-buffered saline ϩ 0.05% Tween 20), and incubated with peroxidase-coupled anti-rabbit antibody (Amersham Biosciences) for 1 h. The membranes were developed using chemoluminescence (ECL kit, PerkinElmer Life Sciences).
Flow Cytometry-After treatment with siRNA, cells were plated in 6-well plates and were incubated for 24 -144 h before measurements. Cells were trypsinized, and the suspension was incubated in phosphatebuffered saline containing 0.1% saponin and a mixture of two anti-EagI monoclonal antibodies directly labeled with quantum dots. After 2 h on ice, cells were washed twice with phosphate-buffered saline, 0.1% saponin and analyzed on a BD FACSAria TM flow cytometer (Becton Dickinson, Heidelberg, Germany). Quantum dots were excited at 488 nm, and fluorescence was collected using a LP556 dichroic and BP660/20 filter. Gates for viable cells (using propidium iodide) were established, and the mean of fluorescence intensity was determined after the exclusion of dead cells.
Electrophysiology-hEag1a-transfected HEK293 cells were plated on multiple poly-L-lysine-coated glass coverslips in Petri dishes and allowed to attach overnight. Cells were treated with siRNAs for 4 h. 24 or 72 h after treatment, the coverslips were removed from the dish and used for electrophysiological measurements.
Electrophysiological recordings were performed in the whole-cell configuration of the patch clamp (24) using an EPC9 amplifier and Pulse software (HEKA, Germany). Currents were digitized at 50 KHz (with 5-fold over-sampling). Patch pipettes were pulled from Corning #0010 glass (World Precision Instruments) to resistances of 1-2 megohms when filled with the internal solution 100 mM KCl, 45 mM NMDG, 5 mM 1,2-bis(O-aminophenoxy)ethane-N,N,NЈ,NЈ-tetraacetic acid, 10 mM HEPES, 5 mM EGTA, 1 mM MgCl 2 , pH 7.4 (KOH). The bathing solution contained 160 mM NaCl, 2 mM CaCl 2 , 1 mM, 2 mM MgCl 2, 2.5 mM KCl, 8 mM glucose, 10 mM HEPES, pH 7.4 (NaOH). We used the automated capacity compensation of the amplifier to estimate series resistance, which was compensated to 85%. Under such conditions, the heterologous EagI currents are robust enough to mask endogenous currents almost completely (see for example Ref. 15). To measure EagI currents, we applied a conditioning pulse to 100 mV for 1500 ms to slow down the activation of EagI (11,12), and the outward currents were then elicited by a square depolarization to ϩ40 mV. We measured the mean steady state current between 350 and 450 ms after the start of depolarization and subtracted from this value the current amplitude between 7 and 13 ms after the start of depolarization. Using this protocol we assume that any fast activating currents are unrelated to EagI (endogenous). The final current values were divided by the measured capacitance of the cell to obtain current densities.
Proliferation Assay-Proliferation was estimated based on the ability of metabolically active cells to either reduce resazurin to fluorescent resorufin (Alamar Blue, BIOSOURCE) or tetrazolium salts to colored formazan (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, Roche Applied Science). siRNA-and control-treated cells were trypsinized and plated in 96-well plates in a volume of 200 l at densities ranging from 1000 to 2500 cells/well depending on the cell line exam-ined. To determine the metabolic activity, 20 l of Alamar Blue or 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reagent was added to each well and returned to the incubator for 2-4 h. The absorbance was determined in a 1420 Victor 2 Multilabel Counter (Wallac) using a 570 nm filter and the fluorescence using 544 nm emission and 590 nm excitation filters.

RESULTS
Specific siRNAs Reduce EagI mRNA Levels-Three cell lines were selected to observe the effects of silencing human EagI with RNA interference. EagI-transfected HEK293 cells were chosen as a heterologous system, whereas two additional cell lines that endogenously express hEag1 were picked as models for either "normal" (Daoy, derived from brain) or ectopic expression of the channel (MDA-MB435S, derived from breast carcinoma).
As a positive control for transfection efficiency, we used siRNA directed against hGAPDH. Knock down of the protein concentration was confirmed by Western blot (not shown). As GAPDH is a key enzyme of glycolysis, reduction of the protein should be reflected on the metabolic activity of transfected cells. As expected, GAPDH-siRNA treatment reduced levels of GAPDH RNA in all three cell lines (Fig. 1,  A-C). The metabolic activity of the cells was immediately reduced and remained clearly reduced for several days after transfection, which we interpret as a combination between slower cell proliferation and metabolic inhibition (Fig. 1D).
The use of multiple functional siRNAs allows independent confirmation of the phenotypic effects (18,19), as different siRNAs targeting the same gene are often differentially effective in silencing. This behavior is because of their different thermodynamic properties, stabilities, and positioning of either the siRNAs themselves or of the target region of the gene (17). We therefore tested four siRNAs directed against EagI, which have been designated Kv10.1-1-Kv10.1-4 to distinguish them from the channel. These siRNAs recognize distinct target sequences in EagI and differ in their silencing potency in both EagI-transfected HEK293 and endogenously expressing cells as shown in Fig. 2. Kv10.1-1 did not reduce the amount of hEag1a RNA significantly in hEag1a-transfected HEK293 cells (Fig. 2A). In contrast, all four sequences were effective at the RNA level in endogenously expressing cell lines (Fig. 2, B and C). In all cell lines, Kv10.1-3 was the most potent siRNA reducing the hEag1a RNA content by 63.2 Ϯ 9.7% (HEK-hEAG1a, Fig. 2A Fig. 2C).
Off-target effects and nonspecific cellular responses are major concerns when interpreting the phenotypes induced by siRNA. To address this issue, we used several independent approaches. First, we determined and subsequently used the minimal effective concentration of each EagI siRNA. We also utilized both commercial nonsense siRNAs   (Negative Control #1), which have been established not to induce a noticeable phenotype in several cell lines, as well as two negative controls specially designed for this work. These negative controls each consist of siRNA with the reversed (but not complementary) sequence of the corresponding targeting siRNAs that are directed against two different regions in the EagI sequence. This means that these controls not only comprise the same nucleotides but also the same sequence of nucleotides as the targeting EagI siRNAs. In all experiments described in this study, none of these control siRNAs produced significant changes in any of the parameters measured (Figs. 2, 5, 8, and 9). Finally, siRNA directed against the close relative EagII failed to reduce the mRNA content of EagI in cells and conversely, EagI siRNA did not alter the levels of EagII mRNA, indicating specific effects on the particular target sequence (Fig. 3).
Interferon response mediates the defense against viral RNA during infection and causes a global nonspecific suppression of gene expression in immune (25) and non-immune cells (26,27). Although synthetic siRNA duplexes are not able to induce an interferon response (22,28), some genes unrelated to interferon can be induced by siRNA treatment, as is the case of pyruvate dehydrogenase kinase isoenzyme 1 (27). None of the specific siRNAs reported here induced changes in the expression levels of PDK1, indicating that the phenotypic effects observed are not related to nonspecific double stranded RNA-induced responses (Figs. 4 and 5).
We also determined the effects of varying the duration that the cells are in contact with the siRNA-transfection reagent mix. For transfection, a minimum time of 4 h is generally recommended, but the optimal time depends on the system used. Cells were incubated with equilibrated siRNA-transfection reagent mix for 4, 8, 12, and 24 h, and the relative reduction in EagI RNA levels elicited by Kv10.1-3 was monitored by real time PCR. Additionally, we measured the hPDK1 mRNA levels to assess the nonspecific response. We detected optimal EagI reduction after an 8-h incubation with siRNA in hEag1a HEK293-transfected cells. Longer incubations resulted in non-significant changes in the levels of hPDK1, suggesting that incubation over longer time periods would induce nonspecific changes. We therefore established the 8-h incubation to be optimal and used this time for all experiments (Fig. 5).
Dose-response Relationship of EagI Inhibition-To reduce both off-target effects and any nonspecific responses to the siRNAs, the minimum concentration of siRNAs required to produce the maximum effect should be used (19). Therefore the dose-response relationship of siRNA-mediated EagI inhibition was determined in both transfected HEK293 and naturally expressing cells. IC 50 values were in the tens of nM range (19.56 Ϯ 36 nM (HEK-hEAG1a), 11.77 Ϯ 9 nM (MDA-MB435S), and 10.43 Ϯ 10 nM (Daoy)), and a nonspecific response was not observed. Based on this information, we selected 25 nM as the optimal siRNA concentration to perform all subsequent experiments (Fig. 6).
Time Course of siRNA Action-There is abundant literature addressing the time course of siRNA action on the RNA, protein, and functional  levels for many targets (20,29,30). Nevertheless, we decided to follow the time course for EagI targeting and discovered surprisingly rapid effects. A reduction in EagI RNA levels was already detectable at 4 h, and they reached their minimum level at 8 h. A recovery in EagI RNA was only measurable after 72 h (Fig. 7).
We therefore expected to also detect effects on the protein and functional levels earlier than that reported for other proteins. Western blots (Fig. 8A) and flow cytometry (Fig. 8B) show a significant reduction of hEag1 protein level, apparent 8 h after transfection with a minimum protein level observed at 48 h. Levels recovered to those of the controls at 144 h. In functional terms, the current was completely abolished 24 h after transfection and completely restored 72 h after transfection (Fig. 8C). This is in contrast to the total EagI protein levels, which are still significantly reduced at this latter time point (see Fig. 8, A and B).
EagI has also been proposed to be important for cell growth in several cell lines (6,13,14), and we therefore investigated the possible effect that EagI knock down could have on the proliferation of cells. Because maximal knock down of the protein levels occurs at 48 h, we expected to detect an impact on proliferation relatively late after transfection. Specific anti-EagI siRNA progressively reduced the proliferation of the three model cells over a tested period of 96 h. At this time point, proliferation was reduced by 57.2 Ϯ 4.8% (HEK-hEAG1), 44.9 Ϯ 3.7% (MDA-MB435S), and 31.9 Ϯ 1.6% (Daoy) (Fig. 9).  EagI Silencing Reduces Proliferation of Other Tumor-derived Cell Lines-After optimizing and characterizing the effects of EagI siRNA on the described model cell lines in some detail, we extended our investigations to several other cell lines derived from major classes of cancer. We therefore tested cell lines isolated from sarcomas (HT-1080, A-204, Hs633t, TE-671), breast carcinoma (MDA-MB435S, MCF-7), melanoma (IPC298, IGR-39), cervical carcinoma (HeLa), neuroblastoma (SHSY-5Y), and medulloblastoma (Daoy). We found that treatment with the different EagI siRNAs reduced the corresponding RNA levels in all cases and that the proliferation of some of these cell types was impaired, albeit to different degrees. Specifically, as indicated in Table 1, particular EagI siRNAs could reduce the proliferation of A204 (rhabdoyosarcoma) and HT1080 (fibrosarcoma).

DISCUSSION
In recent years evidence has accumulated that implicates the involvement of potassium channels in the generation and/or progression of cancer (31). In 1999, our group described the oncogenic potential of the EagI potassium channel and its wide distribution in cancer cell lines, whereas most healthy cells outside the brain showed virtually no expression (6). We proposed a role for EagI in proliferation and tumor forma-tion based on in vitro and in vivo experimental results (6). This together with its restricted expression outside the brain suggested that the EagI channel may represent a potential target for cancer therapy. Until now, no specific inhibitor for the channel has been found as neither the tricyclic antidepressant imipramine nor the antihistamine astemizole represent specific blockers of EagI. Imipramine blocks many cardiac and neuronal sodium, calcium, and potassium channels as well as EGL2channels, whereas astemizole is the paradigmatic inhibitor of HERG channels (15). Unfortunately, there is a broad co-expression of these other channels with EagI in both naturally expressing cell lines and cancer tissues (data not shown) and the expected, severe therapeutic side effects associated with use of these drugs have been reported (32). Moreover, the need for a specific blocker of EagI stems not only from a possible therapeutic use but also from the difficulties in clarifying both the physiological role of the channel in brain cells and its pathophysiological relevance in tumor cells.
siRNAs represent a suitable alternative to functional blockers of channel activity. This method has proven to be a useful and specific tool to inhibit the function of proteins (including ion channels, e.g. TASK-1 (33)) and represents a promising approach for cancer therapy (16). The most important factor for the successful application of siRNAs is the targeting of an appropriate segment of mRNA. Several factors (17) must be considered as highlighted by our approach. The algorithm used for the design of our siRNAs (HiPerformance siRNA Design algorithm, Qiagen) takes into account thermodynamic properties and predicted specificity of the siRNAs, but only experimental evidence can give an insight as to the actual efficacy and specificity of a particular siRNA. Control experiments should include the use of at least two independent specific siRNAs, as well as controls consisting of either modified (mismatch) siRNA or scrambled siRNAs (19). We have tested six independent EagI sequences (four of which are reported here) and observed varying levels of successful EagI knock down. Using one of the sequences, we show that low concentrations of EagI siRNA can reduce EagI expression in several cell lines at the mRNA, protein, and functional levels. This phenotype is not reproducible by commercial nonsense siRNAs, a mixture of four independent siRNAs against EagII, or reverse (not complementary) siRNAs. To our knowledge, reverse siRNAs have not been used before as a specific control, but we believe that a control possessing an identical nucleotide composition and the closest similar structure available to the specific targeting siRNA, should provide the best means to elicit most of the potential nonspecific effects of siRNA. Additional control experiments whereby RNA levels of the closely related EagII gene and a nonspecific gene (PDK1) were monitored indicate that the EagI siRNAs used in this study act very specifically on their target sequence.  High siRNA concentrations can also induce nonspecific effects. The amount of siRNA needed for maximal target RNA degradation might vary depending on the expression level of the target (34), and we therefore determined the dose-response relationship of EagI siRNA in transfected and naturally expressing cells. The IC 50 was in all cases in the range of tens of nM, implying that effects because of high concentration are unlikely.
A recent discovery is the potential of double-stranded siRNAs to induce an interferon response (25) thereby causing a global nonspecific suppression of gene expression (27,29) There is evidence that 21-mer siRNAs are able to not only induce an interferon response but also to affect the expression of completely unrelated genes. To avoid misinterpretation of these nonspecific responses, the recommended controls are to monitor target related and unrelated genes, to use the minimal effective concentrations, to optimize the times of exposure to siRNA, and to apply several scrambled and specific siRNAs alone or in combination (19). In our work, none of the control experiments indicated any nonspecific or off-target effects. We therefore conclude that the phenotypic changes observed are because of the silencing of EagI expression.
When interpreting siRNA results, it is important to consider not only the sequence and the turnover of siRNAs themselves but also many factors related to the model cell line (20). We used three cell lines of distinct origin in our experiments, which possess varying amounts of EagI RNA and probably differ with respect to EagI turnover and the presence or absence of interaction partners. HEK293 cells transfected with the Eag1a splice variant were used as a heterologous overexpression system, where currents could be easily measured, and large amounts of RNA (1-5 ϫ 10 6 copies/l) and EagI protein were available. However this cell line is not derived from a tumor but was immortalized by viral infection, and as such it represents an artificial system that might lack some interacting factors or the EagI channel might be aberrantly processed. We also used Daoy cells as a CNS-derived model. Medulloblastoma is the most frequent malignant brain tumor in children and is thought to derive from the undifferentiated granular layer cells of the cerebellum. The Daoy cell line possesses neuronal and glial characteristics (35) and expresses a moderate amount of endogenous EagI (10,000 -50,000 copies/l). Additionally, Daoy cells also express EagII naturally, which provided a good internal control. The third cell line, MDA-MB435S, is derived from a pleural effusion of a mammary carcinoma, although it seems more related to melanomas rather than carcinomas as revealed by gene array experiments (36,37) and expresses a lower amount of EagI (1000 -5000 copies/l). MDA-MB435S cells also express small amounts of EagII. These two cell lines are therefore good candidates to have the set of factors specific for interaction in cancer cells.
EagI siRNA treatment resulted in essentially identical phenotypic changes in all three cell lines that can be interpreted as a reduction in proliferation. We therefore went on to study the effects on cell proliferation induced by EagI silencing in a number of other cancer cell lines derived from very diverse tumors, such as rhabdomyosarcoma, fibrosarcoma, breast carcinoma, melanoma, cervix carcinoma, neuroblastoma, and medulloblastoma. In these studies, we found a consistent reduction of the speed of growth in most cell types. These findings highlight the participation of EagI in tumor cell proliferation and reinforce the hypothesis that EagI might be a suitable target for cancer therapy.
The time course of the reduction of RNA, protein, and functional activity of EagI upon exposure to siRNA was unusually fast, indicating a very rapid turnover of the molecule at all levels. Normally 2-4 days are required to detect a reduction of the protein after exposure to siRNAs (30,33). We report here not only a reduction of protein within 8 h of the transfection, but also a nearly complete abolition of the EagI conductance at 24 h. That RNA levels recovered after 72 h, at which point the restitution of the electrophysiological activity of EagI was complete, suggests the existence of an intracellular storage pool of EagI protein.
The reasons and implications of this fast turnover are currently under investigation in our laboratory.
In conclusion, we report here the design and characterization of siRNAs that specifically target the EagI message. These siRNAs will serve as tools in the future to facilitate the elucidation of both the physiological and pathophysiological functions of this intriguing protein.