Bicistronic GFP Expression Vectors as a Tool to Study Ion Channels in Transiently Transfected Cultured Cells

  • Jan Eggermont
  • Dominique Trouet
  • Gunnar Buyse
  • Rudi Vennekens
  • Guy Droogmans
  • Bernd Nilius
Part of the Methods in Pharmacology and Toxicology book series (MIPT)


The availability of ion-channel cDNAs has greatly increased our insight in the structure, function, pharmacology, and regulation of ion channels at the molecular level. Much of this knowledge has been obtained by expressing wild-type or mutant ion channels in a heterologous host system, thereby facilitating functional approaches and analyses, which are not possible, when the native channel is studied in its in situ context (1).


Green Fluorescent Protein Internal Ribosomal Entry Site Green Fluorescent Protein Expression Transcription Unit Green Fluorescent Protein Fluorescence 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


The availability of ion-channel cDNAs has greatly increased our insight in the structure, function, pharmacology, and regulation of ion channels at the molecular level. Much of this knowledge has been obtained by expressing wild-type or mutant ion channels in a heterologous host system, thereby facilitating functional approaches and analyses, which are not possible, when the native channel is studied in its in situ context (1).

Xenopus oocytes have proven to be a reliable and efficient expression system, especially for voltage-gated cation channels and channels gated by extracellular ligands. However, Xenopus oocytes have several drawbacks with respect to the study of ion channels: (1) Xenopus oocytes contain several endogenous currents, most notably anion currents, which can interfere with the characterization of expressed channels (for a discussion of endogenous ion currents in Xenopus oocytes, see refs. 2 and refs. 3). (2) Because Xenopus oocytes are highly specialized nonmammalian cells surrounded by a layer of follicular cells, the processing of the expressed channel protein (4,5), its pharmacological properties (6,7) and/or the signal transduction pathways may diverge from those in mammalian cells. (3) Single-channel analysis as well as control of the internal medium are more cumbersome than in mam malian cells and require specialized techniques (8). Most of these disadvantages can be overcome by expressing ion channels in mammalian cells, although, evidently, mammalian cells are not entirely devoid of endogenous background currents. A specific problem with transfection in mammalian cells is, however, that the transfection efficiency (i.e., the proportion of transfected cells that express the ion channel of interest) varies greatly depending on the cell type, the transfection method, purity of the DNA (contamination with bacterial endotoxins), quality of the cell line, and the biological activity (problem of toxicity). In principle, there are two experimental strategies to overcome the problem of low efficiency. (1) Stable selection of positive cells by drug resistance (e.g., resistance to G418 [an aminoglycoside], puromycin, hygromycin). However, selection and characterization of a monoclonal cell line stably expressing the ion channel under study is a time-consuming process that can take up to several weeks or months. The use of stable cell lines is therefore generally restricted to pharmacological screening experiments or applications that require large amounts of protein (e.g., biochemical characterization of the expressed channel). (2) Identification of transiently transfected cells using a marker that can be detected under patch-clamp conditions without compromising the cellular integrity and function. Ideally, such a marker should not interfere with microfluorescencebased techniques that allow measurement of intracellular Ca2+, intracellular pH, or cell volume. In this chapter, we describe a detection method for transiently transfected cells that uses a variant of green fluorescent protein (GFP) as an in vivo cell marker and an internal ribosomal entry site (IRES)-dependent bicistronic transcription unit to ensure strict coupling between expression of the channel and the GFP marker.


2.1 Green Fluorescent Protein

Wild-type GFP is a 26.8 kDa protein (238 amino acids) from the jelly fish Aequorea victoria that contains a fluorochrome with two excitation peaks (major at 395 nm; minor at 475 nm) and a single-emission peak at 510 nm (9). Formation of the fluorochrome occurs post-translationally by spontaneous cyclization of a Ser-Tyr-Gly tripeptide (amino acid 65 to 67) followed by oxidation (10). The spectral properties of GFP (excitation and emission wave lengths, fluorescence intensity) can be changed by specific mutations that modify the chromophore structure or promote correct folding of the protein (for a review, see refs. 11). Several properties promoted the use of GFP as an identification tool for cells transfected with ion channels. (1) GFP fluorescence does not require additional cofactors or substrates (9) thereby obviating the need for cell manipulation prior to detection. (2) GFP seems to be non-toxic for mammalian cells as evidenced by the observation that transgenic mice expressing GFP develop normally (12). However, it has been suggested that fluorochrome formation is accompanied by a stoichiometric release of H2O2 which could be harmful in case of high level GFP expression (13). (3) Because GFP is a cytosolic protein, it is less likely to interfere with the function of membrane proteins such as ion channels, thereby reducing the possibility of artefacts owing to the expression system. Yet, the application of wild-type GFP as a transfection marker is at least in mammalian cells seriously limited by the weakness of the fluorescent signal and by the background autofluorescence generated by cellular excitation in the UV range (13). Fortunately, these problems have been (partially) resolved by generating mutant GFPs with altered spectral properties (13). One of the first GFP mutants available was enhanced GFP (eGFP), which bears the double mutation F64L and S65T (14,15). When compared to wildtype GFP, eGFP has a red-shifted excitation spectrum with a single excitation peak at 488 nm and a similar emission peak at 510 nm, but the emission light is approx 30-fold more intense. We therefore selected eGFP as an identification tool for transfected cells.

2.1.1 Internal Ribosomal Entry Site

The next question is how to couple the expression of the ion channel to that of eGFP. One possibility is to construct a fusion protein between the ion channel and GFP leading to an ion channel with a fluorescent tag (16, 17, 18). Alternatively, mammalian cells can be co-transfected with two vectors, one expressing the ion channel and another GFP (16). We opted for a third possibility in which the ion channel and the eGFP marker are expressed from an IRES-containing, bicistronic mRNA.

IRES is an RNA sequence that allows cap-independent translation of mRNAs in mammalian cells. Translation initiation of the vast majority of cellular mRNAs requires the presence of a methylated cap at the 5− end of the mRNA ( Fig. 1 ). The eukaryotic initiation factor 4F (eIF-4F) binds to the methylated cap and initiates the recruitment of the ribosomal apparatus to the mRNA (19). However, translation initiation of some viral RNAs (e.g., picorna viruses, see refs. 20) and of a small subset of cellular mRNAs (19) does not require a 5− cap structure in the mRNA. Instead, cap-independent translation is initiated by binding of the ribosomal apparatus to an internal RNA sequence in the 5− untranslated region (UTR), which is aptly named the internal ribosomal entry site or IRES. The presence of an IRES in the 5− UTR allows cap-independent translation of viral RNA during picornaviral infection when cap-dependent translation of host-cell mRNAs is shut down owing to the virus-mediated cleavage of eIF-4G, a subunit of the eIF-4F complex (21). Intact eIF-4G is required for cap-dependent translation, but cleaved eIF-4G is still capable of sustaining IRES-mediated translation initiation (22,23). Although the naturally occurring viral and cellular IRES sequences are found in the 5− UTR of monocistronic RNAs (i.e., RNAs with a single open reading frame), IRES sequences can be put to work in the context of bicistronic mRNAs (24, 25, 26). If positioned between the first and second open reading frame of a bicistronic mRNA, IRES significantly increases the translation of the downstream open reading frame. Structural and functional criteria allow the picornaviral IRESes to be classified in 3 groups with the IRES of cardioviruses (e.g., the Encephalomyocarditis virus or EMCV) and aphtoviruses being the most efficient element for IRESFig. dependent translation initiation (27). Thus, owing to their unique property of directing internal translation initiation, IRES elements can be used to couple the expression of two proteins such as GFP and an ion channel on the condition that both are expressed from an IRES-containing bicistronic mRNA.

Cap-dependent vs IRES-dependent translation initiation. Recruitment of the eIF-4F factor is the first step in the translation of cellular mRNAs, picornaviral RNAs and IRES-containing bicistronic RNAs. Both the methylated 5− cap (cellular mRNAs) and the IRES sequence (picornaviral RNAs) function as binding sites for eIF-4F and they can therefore direct ribosomal assembly. The stop codon signals the end of the open reading frame and causes the ribosome to dissociate from the RNA. Thus, translation of the second open reading frame in a bicistronic mRNA requires an eIF-4F binding site upstream of the start codon of the second open reading frame to promote ribosomal assembly.

2.2 pCINeo/IRES-GFP Vector

The pCINeo/IRES-GFP is schematically shown in Fig. 2 . The vector was constructed starting from a pCI-neo backbone (Promega), which was modified by inserting an IRES-eGFP cassette in the multiple cloning site (28). The bicistronic transcription unit consists of the following elements: (1) the Cytomegalovirus immediate-early enhancer and promoter for efficient transcription in a broad range of mammalian cells; (2) an artificial intron; (3) a cloning site that consists of unique NheI, XhoI, and EcoRI sites and that allows insertion of cDNAs encoding the protein of interest; (4) an IRES sequence corresponding to nucleotides 243–833 of the EMCV RNA (GenBank/EMBL accession number M81861) and containing the essential sequence elements for cap-independent translation; (5) eGFP cDNA encoding the enhanced GFP (F64L/S65T) as described earlier; (6) an SV40 polyA signal for efficient 3− end processing. After transfection in a mammalian cell, a bicistronic mRNA is transcribed: the first open reading frame corresponds to the cDNA inserted in the cloning site upstream of IRES and it is translated in the classical cap-dependent way; the second open reading frame encodes eGFP and is translated via the IRES.
Fig. 2.

pCINeo/IRES-GFP vector. The Cytomegalovirus immediate/early enhancer/promoter drives the expression of the bicistronic transcription unit. cDNAs of interest can be inserted in the NheI, XhoI, or EcoRI sites. Transcription results in a bicistronic mRNA. The first open reading frame is derived from the inserted cDNA (encoding the protein of interest) and it is translated in a cap-dependent way. The second open reading frame encodes eGFP, which requires the IRES sequence for translation.

The pCINeo/IRES-GFP vector contains a second transcription unit that allows expression of the neomycin phosphotransferase gene. This transcription unit confers two additional properties: (1) it allows stable selection of transfected cells owing to resistance to neomycin and analogs such as G418; (2) the SV40 enhancer/promoter contains the SV40 origin of replication thereby allowing episomal replication in Large T-antigen expressing hosts such as COS cells.

In addition to the pCINeo/IRES-GFP vector other bicistronic IRES-GFP vectors have been reported, for example pTF-G6/IRES/GFP (29), pCAGGSM2/IRES-GFP (30) and commercially available vectors such as pIRES-EGFP and pIRES-EYFP (Clontech, Palo Alto, CA).

2.3 Evaluation of GFP Fluorescence in pCINeo/IRES-GFP Transfected Cells

2.3.1 Patch Clamp Set Up

For current recordings in pCINeo/IRES-GFP transfected cells, a patch clamp set up containing a Zeiss Axiovert 100 microscope, a Xenon light source, and epifluorescence optics (Zeiss, XBO 75 and Zeiss-EPI unit, fluorescence condenser) was used to visualize green fluorescent cells. The excitation light reached the cells via a band-pass filter (Zeiss filter set 9, 487909; BP 450–490) and subsequently a dichroic mirror (Zeiss FT 510). The emitted light passed a 520 nm long-pass filter (Zeiss, LP 520) and was visually detected (28).

2.3.2 FACScan Analysis

Cells transfected with pCINeo/IRES-GFP are trypsinized 24–48 h after transfection and GFP fluorescence is measured with a FACScan (Fluorescence Activated Cell Scan; Becton Dickinson) with FITC settings (excitation at 488 nm; emission detected in FL1-H channel). FACScan of cells transiently transfected with the pCINeo/IRES-GFP vector typically shows an elongated tail which corresponds to the cell population expressing GFP and the protein encoded by the first open reading frame ( Fig. 3 ). In cells stably transfected with pCINeo/IRES-GFP, the entire distribution plot is shifted to higher fluorescence levels ( Fig. 4 ).
Fig. 3.

FACScan of transiently transfected COS cells. (A) COS cells were transiently transfected with pCINeo/IRES-GFP/RCK1 and analyzed by FACScan. Nontransfected control cells display background fluorescence as indicated by the black curve (cut off for background fluorescence indicated by the dotted line). The transfected cell population (gray-filled curve) contains a subpopulation (7% of the cells) with increased fluorescence (extended gray-filled tail to the right). Electrophysiological analysis of the transfected population revealed that 100% of the green cells (n = 32) expressed a Kv1.1 delayed rectifier K+ current typical of RCK1, which was not detected in nonfluorescent COS cells (28). (B) COS cells were transfected with pCINeo/IRES-GFP/CD2 and subjected to dual FACScan analysis to simultaneously measure GFP expression (x-axis) and CD2 (a lymphocyte surface antigen) expression (y-axis). The graph shows that virtually every green fluorescent cell also expresses the CD2 antigen thereby validating the use of the GFP signal as a marker for co-expression.

Fig. 4.

FACScan of stably transfected CHO cells. CHO cells were transfected with pCINeo/IRES-GFP/CFTR and selected with G418 during 6 wk. The polyclonal G418-resistant cell population was then subjected to FACScan analysis (black curve) and compared to control cells (gray-filled curve). The fluorescence profile of the G418-resistant population is clearly shifted to the right with 76% of the cells being above background fluorescence (cut off for background fluorescence indicated by the dotted line).

2.3.3 Fluorescence Microscopy

GFP fluorescence is easily assessed on a classical fluorescence microscope (Nikon Epi-fl microscope) with standard epifluorescence optics (excitation filter: 450 to 490 nm; dichroic mirror at 505 nm and a barrier filter at 520 nm). Transfected cells are seeded on cover slips and turned upside down on a microscope glass.

Imaging of pCINeo/IRES-GFP transfected cells was performed with a Biorad 1024 confocal microscope with a Argon/Krypton laser (488 nm) for excitation and an emission filter (FITC settings) of 522 ± 16 nm. For this purpose, cells were grown in LabTek culture chambers.


3.1 Indications for Bicistronic IRES-GFP Vectors

In general, the need for a bicistronic IRES-GFP vector to study ion channels after transfection in mammalian cells is inversely correlated with the efficiency of transfection. In some cell lines and under optimal conditions, it is possible to obtain transfection efficiencies of 50% or more. For example, transient transfection of the wild-type α1G subunit (cloned in a pcDNA3 vector) in HEK cells generated typical T-type Ca2++ currents in approx 30% of the patched cells without GFP selection (J. Prenen and B. Nilius, unpublished observations). Efficient “blind patching” also requires that the transfected channel has a characteristic fingerprint that is easily and rapidly recognized after seal formation. However, “blind patching” may cause problems when testing mutant channels of which the expression level and/or the phenotype may drastically differ from those of the wild-type channel. Thus, even if control experiments indicate a high-transfection efficiency for the wild-type channel, a bicistronic IRES-GFP expression system may still facilitate the functional analysis of less-efficiently expressing mutant channels.

A bicistronic IRES-GFP vector is virtually indispensable when working with cell lines with low transfection efficiency irrespective of the transfected ion channel. For example, the transfection efficiency in calf pulmonary artery endothelial cells (CPAE) is generally less than 5%. However, selection of positive cells by means of GFP fluorescence allowed us to efficiently study membrane currents (and cytosolic Ca2+ signals) in CPAE cells transfected with the α-subunit of the human large conductance Ca2++ activated K+ channel (hslo), the human trp3 channel (31,32), bovine trp1 and trp4 (33), and human CFTR Cl– channel (34).

3.2 Alternatives to Bicistronic IRES-GFP Vectors

An interesting question is whether there are alternative strategies for coupled expression of ion channel and GFP and if so, what the respective advantages and disadvantages are. A first alternative is to construct an expression vector encoding a fusion protein between the ion channel and GFP, which results in the expression of an ion channel with a fluorescent tag at its amino or carboxyl terminus (16). The construction of a fusion protein vector is technically more demanding than the insertion of the channel cDNA in the pCINeo/IRES-GFP vector, because synthesis of a full-length fusion protein requires that the ion channel cDNA and GFP cDNA are ligated in frame to preserve the open reading frame. Although the fusion protein approach has been successfully used for several ion channels (17,18), it is impossible to predict a priori whether or not the attachment of a 238 amino acid peptide to the NH2- or COOH-terminus of an ion channel will alter the biosynthesis, localization, stability, or functional properties of the channel. A second alternative is cotransfection of mammalian cells with two vectors, one expressing the ion channel and another GFP (16). This approach has the advantage that it works with any mammalian expression vector because there is no specific need to clone the ion channel cDNA in a bicistronic IRESGFP vector. However, the fraction of fluorescent cells that will also express the ion channel depends on the efficiency of cotransfection and may therefore vary. Whereas this may not pose problems for cells with a high transfection efficiency, it can seriously slow down experiments on cells with a low transfection efficiency. A third possibility is to couple the expression of fluorescent protein and protein of interest via a bi-directional expression system. Fang et al. (35) described a mammalian expression vector in which the eGFP gene and a second gene encoding the protein of interest are transcribed from oppositely oriented CMV promoters controlled by a single tetracycline-regulated element (TRE). Binding of the tetracyclinecontrolled transactivator (tTA) to TRE simultaneously switches on both CMV promoters resulting in co-expression of the two transcription units (36,37). Moreover, transcription can be shut down by the addition of tetracycline or doxycycline because these antibiotics bind to tTA thereby dissociating it from TRE. Although this method ensures strict coupling, it suffers from the disadvantage that the tTA has to be provided in trans, thus limiting the application of this system to cell lines that have previously been stably transfected with a tTa expression plasmid (a series of tTA positive cell lines is available from Clontech). Finally, one could resort to an expression vector with two independent transcription units (cf pBudCE4; Invitrogen, Carlsbad, CA) with one promoter driving the expression of GFP and the other promoter controlling the expression of a second protein.
Fig. 5.

GFP expression does not interfere with ICl,swell in CPAE cells. (A,B) Time course of ICl,swell in control (A) and pCINeo/IRES-GFP transfected (B) CPAE cells at +100 mV (upper) and –100 mV (lower). ICl,swell was triggered by perfusing the cells with a hypotonic solution (HTS; 27% reduction) as indicated. (C,D) Currentvoltage relationships for control (C) and pCINeo/IRES-GFP transfected (D) CPAE cells. IV plots were reconstructed from ramps at the times (a,b) indicated in the time course. The inset shows the mean ICl,swell (error bar = SEM) at +50 mV for control (n = 10) and transfected (n = 9) cells. ICl,swell characteristics are not different between control and GFP-expressing cells.

In conclusion, in comparison with the aforementioned strategies, the bicistronic IRES-GFP vector offers the following advantages: (1) obligate co-expression of fluorescent marker and ion channel; (2) no structural modification of the ion channel; (3) no requirement for additional trans-activating factors. A technical problem with the pCINeo/IRES-GFP may be the small number of unique restriction sites, which may occasionally complicate subcloning or mutagenesis procedures.

3.3 Effects of GFP Expression on Membrane Currents

Because the purpose of the GFP marker is to identify cells for electrophysiological analysis, one critical question is whether GFP expression affects the functional characteristics of the host cell. The dominant membrane current in nonstimulated CPAE cells is an inwardly rectifying K+ current (38), which is not affected by the expression of eGFP (e.g., see Fig. 2 in ref. 32). Furthermore, we systematically studied the potential effect of eGFP on I Cl,swell, the chloride current through the volume-regulated anion channel (VRAC), which is typically activated by exposing CPAE cells to an extracellular hypotonic solution (for a review of VRAC, see ref. 2). ICl,swell current densities were not significantly different between control CPAE cells and green fluorescent CPAE cells that had been transfected with an “empty” pCINeo/IRESGFP vector (see Fig. 5 and 34). Similarly, expression of eGFP alone had no effect on ICl,swell in Caco-2 cells (39). Smith et al. (bi40) com pared the electrophysiological properties of dorsal root ganglion cells before and after infection with a GFP-expressing adenovirus and found no difference between the two conditions. Finally, agonist-induced Ca2+ release as measured by the peak cytosolic Ca2++ concentration after application of a Ca2++ mobilizing agonist was not significantly different between control cells and GFP-expressing cells, either CPAE (32) or COS-7 cells (Fig. 6). To conclude, the available data indicate that expression of GFP does not disturb the electrophysiological properties of the host cell as can be expected from a nontoxic, cytosolic protein.
Fig. 6.

GFP expression does not interfere with ATP-induced Ca2++ transient in COS cells. COS cells were transiently transfected with pCINeo/IRES-GFP and stimulated with 0.5 μM ATP. Free cytosolic Ca2++ concentration was ratiometrically measured with indo-1 as a fluorescent indicator. The time course as well as the amplitude of the ATP-triggered [Ca2++]cyt transients were identical in control and GFP-expressing cells. The inset shows no difference between the basal and peak [Ca2+]cyt (mean ± SEM) in control (n = 25) and GFP-expressing (n = 8) COS cells.


4.1 Functional Characterization of Ion Channels in GFP Transfected Cells

Using the bicistronic IRES-GFP vector we have been able to express a wide variety of ion channels in different mammalian cell lines: voltagedependent K+ channels RCK1 (28), voltage-dependent ClC-1 Cl– channels (41), and CFTR Cl– channels in COS cells (34,42); large conductance Ca2+ activated K+ channels hslo (32), Ca2+ permeable trp3 channels (31), bovine trp1 and trp4 (33), and CFTR Cl– channels (34) in CPAE cells; the regulatory β-subunit of the large conductance Ca2+ activated K+ channels hslo in endothelial EA.hy926 cells (Papassotiriou, Nilius; unpublished observations); modulators of VRAC such as caveolin-1 in Caco-2, T47D, and MCF-7 cells (39); ECaC (epithelial Ca2+ channel) (Vennekens, Nilius, Bindels; unpublished observations) in HEK cells.

4.2 Ca2+ Measurements in GFP Expressing Cells

One interesting application of the bicistronic technology is to measure cytosolic Ca2+ in transfected mammalian cells. For example, CPAE cells do not express a large-conductance Ca2+-activated K+ channel (BKCa), which provided us with a tool to study the effect of hslo (the human large-conductance Ca2+-activated K+ channel) on the agonist-induced [Ca2+]cyt in an endothelial cell line. In this study cytosolic Ca2+ was measured by using fura-2 as a Ca2+ indicator (excitation: 360/380nm; emission > 510 nm) signals with a photomultiplier tube. Stimulation of hslo-transfected CPAE cells with Ca2+ mobilizing agonists such as ATP resulted in a significant hyperpolarization and concomitantly an increased Ca2+ plateau that depended on Ca2+ influx (32). Thus, these experiments allowed us to correlate the effects of BKCa on membrane potential, membrane currents and [Ca2+]cyt during agonist stimulation. In a similar approach, we were able to show that CPAE cells transfected with htrp3 acquired an agonist-sensitive nonspecific cation current that could be blocked by PLC-β inhibitors and also an increase in the agonist-induced [Ca2+]cyt plateau (31).

4.3 Cell Volume Measurements in GFP Expressing Cells

We have recently developed a method to record simultaneously cell thickness (representing cell volume) and membrane currents (43). Cell thickness is measured using fluorescent beads (Red Neutravidin: excitation peak 580 nm; emission peak 605 nm; Molecular Probes), which are seeded above and below cells plated on coverslips, and which allow to monitor changes in cell height (see ref. 44). Because there is sufficient difference between the fluorescent properties of eGFP and Red Neutravidin-labeled beads, we have been able to measure cell thickness in pCINeo/IRES-GFP-transfected endothelial cell lines (34). Using this technique, we were able to show that the inhibitory effects of transiently expressed CFTR on the swelling-induced activation of VRAC in CPAE cells cannot be accounted for by altered cell volume responses in CFTR-expressing cells.

4.4 Immunofluorescence of pCINeo/IRES-GFP-Transfected Cells

Usage of the pCINeo/IRES-GFP vector is also compatible with immunofluorescent analysis of transfected cells. In a recent study, we showed that caveolin-1 is required for efficient activation of VRAC and that this effect is isoform-specific: caveolin-1?? is completely inactive, whereas caveolin-1β is equipotent with respect to caveolin-1 (39). To find out whether differences in expression or intracellular distribution might account for the isoform-specific effects, we looked at caveolin-1 distribution in transfected Caco-2 cells by immunofluorescence on a BioRad 1024 confocal fluorescence microscope with the following settings: excitation with Krypton/Argon laser at 488 nm; emission filter for FITC (or GFP): 522 ± 16 nm; emission filter for R-Phycoerythrin: 580 ± 16 nm. Because the cells were transfected with the pCINeo/IRES-GFP vector, we could first identify the transfected cells by their green fluorescent signal and subsequently assess the caveolin-1 expression using a monoclonal antibody (MAb) and R-Phycoerythrin-conjugated secondary antibodies. This revealed that caveolin-1β is expressed in the cell periphery and plasma membrane, whereas expression of caveolin-1α is restricted to the cell interior (39).

The pCINeo/IRES-GFP vector was also used to study the subcellular distribution of transiently transfected ClC-6 isoforms in COS and CHO cells (41). This indicated that transiently expressed ClC-6 co-localized with the endoplasmic reticulum Ca2+ pump SERCA2b, which is compatible with an intracellular location of ClC-6.

4.5 Creation of Stable Cell Lines

The pCINeo/IRES-GFP vector allows selection of stably transfected cells with geneticin (G418) owing to the presence of the phosphotransferase gene. We have successfully created CHO cell lines stably transfected with pCINeo/IRES-GFP containing the cDNA for CFTR (data not shown). FACS analysis of the G418-resistant polyclonal cell pool indicated that 70% of the G418-resistant cells showed green fluorescence above background after 6 wk of drug selection ( Fig. 4 ). However, we also observed that the number of GFP-positive cells in the polyclonal population gradually declined during cell culturing. This is probably owing to the fact that the phosphotransferase gene and the bicistronic IRES-GFP unit constitute two different transcription units that can be separated during chromosomal integration of the plasmid DNA. Consequently, only a fraction of the drug-resistant cells will be green fluorescent and express the cDNA that is cloned in the bicistronic IRES-GFP transcription unit.

This problem can be circumvented by using bicistronic vectors in which the resistance gene is placed behind the IRES sequence ensuring that every drug-resistant cell also expresses the cDNA of interest (45, 46, 47). Because these vectors do not contain a cDNA for a fluorescent marker, they do not allow visual selection of transfected cells. GFP selection would require the modification of existing vectors. One possibility is to create a tricistronic vector in which the second transcription unit encodes the fluorescent marker and the third transcription unit (also preceded by an IRES) contains the resistance gene. It has indeed been shown that IRES mediates translation of tricistronic RNAs (24,48). An alternative possibility is to construct a bicistronic vector with the second open reading frame encoding a fusion protein between a GFP variant and a resistance marker. Functional GFPdrug resistance fusion proteins have recently been described (49) and monocistronic vectors with such fusion proteins are commercially available (Clontech).

4.6 Expression of Heteromeric Channels

An important theme that has emerged from biochemical and molecular biological studies on ion channels is that a functional channel is often composed of a heteromeric complex. Heteromeric channel complexes are found among voltage-dependent cation channels (e.g., Ca2+ channels, see ref. 50), ligand-gated channels (51), degenerin/ENaC channels (52), and possibly voltage-dependent Cl– channels (53). Functional analysis of heteromeric channels requires the simultaneous co-expression of all subunits in a single transfected cell. This has been achieved by co-transfecting the different subunits, but as already mentioned this method suffers from the fact that cotransfection efficiency is never 100% thereby introducing a degree of uncertainty. Kawashima et al. (54) have recently described an alternative strategy to study heteromeric channels after transfection in mammalian cells. A tricistronic expression vector encoding the P2X2 subunit, the P2X3 subunits and neomycin phosphotransferase was used to establish stably transfected cells that expressed functional heteromeric ATP-activated channels. It remains to be seen whether such tricistronic constructs will gain more widespread acceptance to study heteromeric channels. Alternatively, one could use two bicistronic vectors containing different versions of green fluorescent protein (e.g., eGFP and BFP or blue fluorescent protein) in a cotransfection procedure followed by identification of cells containing both green and blue fluorescence (55,56). Finally, a growing theme in channel research is the importance of direct protein-protein interactions for channel function and regulation as illustrated by the interaction of PDZ proteins with Drosophila TRP (57) or human CFTR (58). Efficient co-expression vectors and/or systems will be extremely useful to study the functional implications of such interactions.


The bicistronic GFP-expressing vector provides an elegant and efficient tool to identify transiently transfected cells. The main advantages of this technology are: (1) there is a strict coupling between GFP expression and expression of the protein of interest; (2) it does not require covalent modifi cation of the protein of interest; (3) GFP is a nontoxic marker that does not interfere with membrane currents; (4) GFP expression does not interfere with other fluorescent signals such as fluorescent Ca2+ indicators or fluorescent beads provided that the appropriate filters are used. As such the pCINeo/ IRES-GFP vector is particularly well-suited to study ion channels in cell lines that are transfected with low efficiency.


  1. 1.
    Whiting, P. J., Wafford, K. A., Pribilla, I., and Petri, T. (1995) Channel cloning, mutagenesis and expression, in Ion channels. A Practical Approach (Ashley, R. H., ed.), IRL Press, Oxford, pp. 133–169.Google Scholar
  2. 2.
    Nilius, B., Eggermont, J., Voets, T., Buyse, G., Manolopoulos, V., and Droogmans, G. (1997) Properties of volume-regulated anion channels in mammalian cells. Prog. Biophys. Molec. Biol. 68, 69–119.CrossRefGoogle Scholar
  3. 3.
    Weber, W. M. (1999) endogenous ion channels in oocytes of Xenopus laevis: recent developments. J. Membrane Biol. 170, 1–12.CrossRefGoogle Scholar
  4. 4.
    Pasyk, E. A., Morin, X. K., Zeman, P., Garami, E., Galley, K., Huan, L. J., Wang, Y., and Bear, C. E. (1998) A conserved region of the R domain of cystic fibrosis transmembrane conductance regulator is important in processing and function. J. Biol. Chem. 273, 31759–31764.PubMedCrossRefGoogle Scholar
  5. 5.
    Smit, L. S., Strong, T. V., Wilkinson, D. J., Macek, M. J., Mansoura, M. K., Wood, D. L., et al. (1995) Missense mutation (G480C) in the CFTR gene associated with protein mislocalization but normal chloride channel activity. Hum. Molec. Genet. 4, 269–273.PubMedCrossRefGoogle Scholar
  6. 6.
    Madeja, M., Musshoff, U., and Speckmann, E. J. (1997) Follicular tissues reduce drug effects on ion channels in oocytes of Xenopus laevis. Eur. J. Neurosci. 9, 599–604.PubMedCrossRefGoogle Scholar
  7. 7.
    Krafte, D. S., Davison, K., Dugrenier, N., Estep, K., Josef, K., Barchi, R. L., et al. (1994) Pharmacological modulation of human cardiac Na+ channels. Eur. J. Pharmacol. 266, 245–254.PubMedCrossRefGoogle Scholar
  8. 8.
    Stefani, E. and Bezanilla, F. (1998) Cut-open oocyte voltage-clamp technique. Methods Enzymol. 293, 300–318.PubMedCrossRefGoogle Scholar
  9. 9.
    Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C.(1994) Green fluorescent protein as a marker for gene expression. Science 263, 802–805.PubMedCrossRefGoogle Scholar
  10. 10.
    Heim, R., Prasher, D. C., and Tsien, R. Y. (1994) Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proc. Natl. Acad. Sci. USA 91, 12,501–12,504.PubMedCrossRefGoogle Scholar
  11. 11.
    Tsien, R. W. (1998) Key clockwork component cloned. Nature 391, 839,840.PubMedCrossRefGoogle Scholar
  12. 12.
    Okabe, M., Ikawa, M., Kominami, K., Nakanishi, T., and Nishimune, Y. (1997) Green mice as a source of ubiquitous green cells. FEBS Lett. 407, 313–319.PubMedCrossRefGoogle Scholar
  13. 13.
    Tsien, R. Y. (1998) The green fluorescent protein. Annu. Rev. Biochem. 67, 509–544.PubMedCrossRefGoogle Scholar
  14. 14.
    Heim, R., Cubitt, A. B., and Tsien, R. Y. (1995) Improved green fluorescence. Nature 373, 664,665.CrossRefGoogle Scholar
  15. 15.
    Cormack, B. P., Valvidia, R. H., and Falkow, S. (1996) FACS-optimized mutants of the green fluorescent protein (GFP). Gene 173, 33–38.PubMedCrossRefGoogle Scholar
  16. 16.
    Marshall, J., Molloy, R., Moss, G. W. J., Howe, J. R., and Hughes, T. E. (1995) The jellyfish green fluorescent protein: a new tool for studying ion channel expression and function. Neuron 14, 211–215.PubMedCrossRefGoogle Scholar
  17. 17.
    Levitan, E. S. (1999) Tagging potassium ion channels with green fluorescent protein to study mobility and interactions with other proteins. Methods Enzymol. 294, 47–58.PubMedCrossRefGoogle Scholar
  18. 18.
    Lopatin, A. N., Makhina, E. N., and Nichols, C. G. (1998) Novel tools for localizing ion channels in living cells. Trends Pharmacol. Sci. 19, 395–398.PubMedCrossRefGoogle Scholar
  19. 19.
    Gray, N. K. and Wickens, M. (1998) Control of translation initiation in animals. Annu. Rev. Cell Dev. Biol. 14, 399–458.PubMedCrossRefGoogle Scholar
  20. 20.
    Jackson, R. J. and Kaminski, A.(1995) Internal initiation of translation in eukaryotes: the picornavirus paradigm and beyond. RNA 1, 985–1000.PubMedGoogle Scholar
  21. 21.
    Stewart, S. R. and Semler, B. L.(1997) RNA determinants of picornavirus cap-independent translation initiation. Sem. Virol. 8, 242–255.CrossRefGoogle Scholar
  22. 22.
    Pestova, T. V., Shatsky, I. N., and Hellen, C. U. (1996) Functional dissection of eukaryotic initiation factor 4F: the 4A subunit and the central domain of the 4G subunit are sufficient to mediate internal entry of 43S preinitiation complexes. Mol. Cell. Biol. 16, 6870–6878.PubMedGoogle Scholar
  23. 23.
    Roberts, L. O., Seamons, R. A., and Belsham, G. J. (1998) Recognition of picornavirus internal ribosome entry sites within cells; influence of cellular and viral proteins. RNA 4, 520–529.PubMedCrossRefGoogle Scholar
  24. 24.
    Jang, S. K., Davies, M. V., Kaufman, R. J., and Wimmer, E. (1989) Initiation of protein synthesis by internal entry of ribosomes into the 5− nontranslated region of encephalomyocarditis virus RNA in vivo. J. Virol. 63, 1651–1660.PubMedGoogle Scholar
  25. 25.
    Ghattas, I. R., Sanes, J. R., and Majors, J. E. (1991) The encephalomyocarditis virus internal ribosome entry site allows efficient coexpression of two genes from a recombinant provirus in cultured cells and in embryos. Mol. Cell. Biol. 11, 5848–5859.PubMedGoogle Scholar
  26. 26.
    Kim, D. G., Kang, H. M., Jang, S. K., and Shin, H. S. (1992) Construction of a bifunctional mRNA in the mouse by using the internal ribosomal entry site of the encephalomyocarditis virus. Mol. Cell. Biol. 12, 3636–3643.PubMedGoogle Scholar
  27. 27.
    Borman, A. M., Bailly, J. L., Girard, M., and Kean, K. M. (1995) Picornavirus internal ribosome entry segments: comparison of translation efficiency and the requirements for optimal internal initiation of translation in vitro. Nucleic Acids Res. 23, 3656–3663.PubMedCrossRefGoogle Scholar
  28. 28.
    Trouet, D., Nilius, B., Voets, T., Droogmans, G., and Eggermont, J.(1997) Use of a bicistronic GFP-expression vector to characterise ion channels after transfection in mammalian cells. Pflügers Arch. Eur. J. Physiol. 434, 632–638.CrossRefGoogle Scholar
  29. 29.
    Blair, L. A. C., Bence, K. K., and Marshall, J. (1999) Jellyfish green fluorescent protein: a tool for studying ion channels and second messenger sgnaling in neurons. Methods Enzymol. 302, 213–225.PubMedCrossRefGoogle Scholar
  30. 30.
    Warnat, J., Philipp, S., Zimmer, S., Flockerzi, V., and Cavalie, A. (1999) Phenotype of a recombinant store-operated channel: highly selective permeation of Ca2+. J. Physiol. (Lond.) 518, 631–638.CrossRefGoogle Scholar
  31. 31.
    Kamouchi, M., Philipp, S., Flockerzi, V., Wissenbach, U., Mamin, A., Raeymaekers, L., et al. (1999) Properties of heterologously expressed hTRP3 channels in bovine pulmonary artery endothelial cells. J. Physiol. (Lond.) 518, 345–358.CrossRefGoogle Scholar
  32. 32.
    Kamouchi, M., Trouet, D., DeGreef, C., Droogmans, G., Eggermont, J., and Nilius, B. (1997) Functional effects of expression of hslo Ca2+ activated K+ channels in cultured macrovascular endothelial cells. Cell Calcium 22(6), 497–506.PubMedCrossRefGoogle Scholar
  33. 33.
    Vennekens, R., Kamouchi, M., Wissenbach, U., Phillip, S., Eggermont, J., Droogmans, G., et al. (1999) Functional expression of Trp1 and Trp4 in vascular endothelium. Pflügers Arch. Eur. J. Physiol. 437, R42.Google Scholar
  34. 34.
    Vennekens, R., Trouet, D., Vankeerberghen, A., Voets, T., Cuppens, H., Eggermont, J., et al. (1999) Inhibition of volume-regulated anion channels by expression of the cystic fibrosis transmembrane conductance regulator. J. Physiol. (Lond.) 515, 75–85.CrossRefGoogle Scholar
  35. 35.
    Fang, Y., Huang, C.-C., Kain, S. R., and Li, X. (1999) Use of coexpressed enhanced green fluorescent protein as a marker for identifying transfected cells. Methods Enzymol. 302, 207–212.PubMedCrossRefGoogle Scholar
  36. 36.
    Gossen, M., Freundlieb, S., Bender, G., Müller, G., Hillen, W., and Bujard, H. (1995) Transcriptional activation by tetracyclines in mammalian cells. Science 268, 1766–1769.PubMedCrossRefGoogle Scholar
  37. 37.
    Baron, U., Freundlieb, S., Gossen, M., and Bujard, H. (1995) Co-regulation of two gene activities by tetracycline via a bidirectional promoter. Nucleic Acids Res. 23, 3605–3606.PubMedCrossRefGoogle Scholar
  38. 38.
    Voets, T., Droogmans, G., and Nilius, B. (1996) Membrane currents and the resting membrane potential in cultured bovine pulmonary artery endothelial cells. J. Physiol. (Lond.) 497, 95–107.Google Scholar
  39. 39.
    Trouet, D., Nilius, B., Jacobs, A., Remacle, C., Droogmans, G., and Eggermont, J. (1999) Caveolin-1 modulates the activity of the volume-regulated chloride channel. J. Physiol. (Lond.) 520, 113–119.CrossRefGoogle Scholar
  40. 40.
    Smith, G. M., Berry, R. L., Yang, J., and Tanelian, D. (1997) Electrophysiological analysis of dorsal root ganglion neurons pre-and post-coexpression of green fluorescent protein and functional 5-HT3 receptor. J. Neurophysiol. 77, 3115–3121.PubMedGoogle Scholar
  41. 41.
    Buyse, G., Trouet, D., Voets, T., Missiaen, L., Droogmans, G., Nilius, B., and Eggermont, J. (1998) Evidence for the intracellular location of chloride chan nel (ClC)-type proteins: co-localization of CIC-6a and CIC-6c with the sarco/ endoplasmic-reticulum Ca2+ pump SERCA2b. Biochem. J. 330, 1015–1021.PubMedGoogle Scholar
  42. 42.
    Wei, L., Vankeerberghen, A., Cuppens, H., Droogmans, G., Cassiman, J.-J., and Nilius, B. (1999) Phosphorylation site independent single R-domain mutations affect CFTR channel activity. FEBS Lett. 439, 121–126.CrossRefGoogle Scholar
  43. 43.
    Voets, T., Droogmans, G., Raskin, G., Eggermont, J., and Nilius, B. (1999) Reduced intracellular ionic strength as the initial trigger for activation of endothelial volume-regulated anion channels. Proc. Natl. Acad. Sci. USA 96, 5298–5303.PubMedCrossRefGoogle Scholar
  44. 44.
    Van Driessche, W., De Smet, P., and Raskin, G. (1993) An automatic monitoring system for epithelial cell height. Pflügers Arch. Eur. J. Physiol. 425, 164–171.CrossRefGoogle Scholar
  45. 45.
    Gurtu, V., Yan, G., and Zhang, G. (1996) IRES bicistronic expression vectors for efficient creation of stable mammalian cell lines. Biochem. Biophys. Res. Commun. 229, 295–298.PubMedCrossRefGoogle Scholar
  46. 46.
    Rees, S., Coote, J., Stables, J., Goodson, S., Harris, S., and Lee, M. G. (1996) Bicistronic vector for the creation of stable mammalian cell lines that predisposes all antibiotic-resistant cells to express recombinant protein. BioTechniques 20, 102–110.PubMedGoogle Scholar
  47. 47.
    Hobbs, S., Jitrapakdee, S., and Wallace, J. C. (1998) Development of a bicistronic vector driven by the human polypeptide chain elongation factor 1alpha promoter for creation of stable mammalian cell lines that express very high levels of recombinant proteins. Biochem. Biophys. Res. Commun. 252, 368–372.PubMedCrossRefGoogle Scholar
  48. 48.
    Metz, M. Z., Pichler, A., Kuchler, K., and Kane, S. E. (1998) Construction and characterization of single-transcript tricistronic retroviral vectors using two internal ribosome entry sites. Somatic Cell Mol. Genet. 24, 53–69.CrossRefGoogle Scholar
  49. 49.
    Lybarger, L., Dempsey, D., Franek, K. J., and Chervenak, R. (1996) Rapid generation and flow cytometric analysis of stable GFP-expressing cells. Cytometry 25, 211–220.PubMedCrossRefGoogle Scholar
  50. 50.
    Walker, D. and De Waard, M. (1998) Subunit interaction sites in voltage-dependent Ca2+ channels: role in channel function. Trends Neurosci. 21, 148–154.PubMedCrossRefGoogle Scholar
  51. 51.
    Unwin, N. (1993) Neurotransmitter action: opening of ligand-gated ion channels. Cell 72, 31–41.PubMedCrossRefGoogle Scholar
  52. 52.
    Fyfe, G. K., Quinn, A., and Canessa, C. M. (1998) Structure and function of the Mec-ENaC family of ion channels. Sem. Nephrol. 18, 138–151.Google Scholar
  53. 53.
    Lorenz, L., Pusch, M., and Jentsch, T. J. (1996) Heteromultimeric CLC chloride channels with novel properties. Proc. Natl. Acad. Sci. USA 93, 13,362–13,366.PubMedCrossRefGoogle Scholar
  54. 54.
    Kawashima, E., Estoppey, D., Virginio, C., Fahmi, D., Rees, S., Surprenant, A., and North, R. A. (1998) A novel and efficient method for the stable expression of heteromeric ion channels in mammalian cells. Recept. Channels 5, 53–60.PubMedGoogle Scholar
  55. 55.
    Stauber, R. H., Horie, K., Carney, P., Hudson, E. A., Tarasova, N. I., Gaitanaris, G. A., and Pavlakis, G. N. (1998) Development and applications of enhanced green fluorescent protein mutants. BioTechniques 24, 462–466, 468-471.PubMedGoogle Scholar
  56. 56.
    Yang, T. T., Sinai, P., Green, G., Kitts, P. A., Chen, Y. T., Lybarger, L., et al. (1998) Improved fluorescence and dual color detection with enhanced blue and green variants of the green fluorescent protein. J. Biol. Chem. 273, 8212–8216.PubMedCrossRefGoogle Scholar
  57. 57.
    Shieh, B. H. and Zhu, M. Y. (1996) Regulation of the TRP Ca2+ channel by INAD in Drosophila photoreceptors. Neuron 16, 991–998.PubMedCrossRefGoogle Scholar
  58. 58.
    Short, D. B., Trotter, K. W., Reczek, D., Kreda, S. M., Bretscher, A., Boucher, R. C., et al. (1998) An apical PDZ protein anchors the cystic fibrosis transmembrane conductance regulator to the cytoskeleton. J. Biol. Chem. 273, 19,797–19,801.PubMedCrossRefGoogle Scholar

Copyright information

© Humana Press Inc., Totowa, NJ 2001

Authors and Affiliations

  • Jan Eggermont
    • 1
  • Dominique Trouet
    • 1
  • Gunnar Buyse
    • 1
  • Rudi Vennekens
    • 1
  • Guy Droogmans
    • 1
  • Bernd Nilius
    • 1
  1. 1.Laboratory of PhysiologyCatholic University of LeuvenLeuvenBelgium

Personalised recommendations