Two splice variants derived from a Drosophila melanogaster candidate ClC gene generate ClC-2-type Cl⁻ channels

Abstract

Members of the ClC family of membrane proteins have been found in a variety of species and they can function as Cl⁻ channels or Cl⁻/H⁺ antiporters. Three potential ClC genes are present in the Drosophila melanogaster genome. Only one of them shows homology with a branch of the mammalian ClC genes that encode plasma membrane Cl⁻ channels. The remaining two are close to mammalian homologues coding for intracellular ClC proteins. Using RT-PCR we have identified two splice variants showing highest homology (41% residue identity) to the mammalian ClC-2 chloride channel. One splice variant (DmClC-2S) is expressed in the fly head and body and an additional, larger variant (DmClC-2L) is only present in the head. Both putative Drosophila channels conserve key features of the ClC channels cloned so far, including residues conforming the selectivity filter and C-terminus CBS domains. The splice variants differ in a stretch of 127 aa at the intracellular C-terminal portion separating cystathionate beta synthase (CBS) domains. Expression of either Drosophila ClC-2 variant in HEK-293 cells generated inwardly rectifying Cl⁻ currents with similar activation and deactivation characteristics. There was great similarity in functional characteristics between DmClC-2 variants and their mammalian counterpart, save for slower opening kinetics and faster closing rate. As CBS domains are believed to be sites of regulation of channel gating and trafficking, it is suggested that the extra amino acids present between CBS domains in DmClC-2L might endow the channel with a differential response to signals present in the fly cells where it is expressed.

• Drosophila melanogaster
• ClC-2 channel
• CBS domains

Introduction

Members of the ClC family of membrane proteins can function as Cl⁻channels or Cl⁻/H⁺ antiporters (Accardi & Miller [2004]), and have been found in a variety of species ranging from prokaryotes (Iyer et al. [2002]) to mammals (Jentsch et al. [2002]). In mammals, four plasma membrane ClC Cl⁻ channels have been identified and their function studied by heterologous expression (Jentsch et al. [2002]). Two of these are the closely related ClC-Ka and ClC-Kb, expressed mainly in kidney and concerned with Cl⁻ reabsorption. ClC-1 is the main resting conductance in muscle, where it plays a function in the stability of membrane potential. ClC-2 is widely distributed in human tissues and is thought to play a role in transepithelial transport and the control of intracellular Cl⁻ in neurons. Mutations in the genes for these channels have been implicated in human disease (Jentsch et al. [2005]). Mutations affecting ClC-Kb and barttin, the β-subunit accompanying ClC-K channels, are responsible for inherited salt-wasting renal disorders. Mutations in the ClC-1 gene lead to muscle hyperexcitability and myotonias. Certain mutations in the ClC-2 protein have been correlated with idiopathic epilepsies. Other members of the ClC family in humans are intracellular proteins and, at least some of them, function as Cl⁻/H⁺ antiporters rather than channels (Scheel et al. [2005], Picollo & Pusch [2005]).

The Drosophila melanogaster fruit fly is an important model organism for genetic and developmental studies. Its usefulness in physiological research stems from the availability of large panels of mutants and well characterized functional and behavioural assays, and the sequencing and annotation of its genome (Yoshihara et al. [2001], Dow & Davies [2003]).

Cl⁻ channels identified so far in Drosophila include histamine-gated Cl⁻ channels HisCl1 and HisCl2 (Gisselmann et al. [2002], Gisselmann et al. [2004]), GABA-gated channels (Henderson et al. [1994], Lee et al. [2003]), glutamate-activated Cl⁻ channels (Kane et al. [2000]), an inwardly-rectifying Cl⁻ channel (Asmild & Willumsen [2000]), the tweety maxi-Cl⁻ channels (Suzuki & Mizuno [2004]), and a novel family of ligand-gated Cl⁻ channels inhibited by protons (Schnizler et al. [2005]). The presence of three genes which should encode members of the ClC family has been reported in the genome of Drosophila, but there are no functional studies exploring the ability of their products to conduct chloride ions (Wang et al. [2004]). A ClC2-like conductance, nevertheless, has been observed in fly photoreceptors (Ugarte et al. [2005]), but there is no demonstration of an association with a ClC mRNA or protein. We have now cloned and expressed in a heterologous system, two functional splice variants of a Drosophila ClC Cl⁻ channel that we term DmClC-2. The mRNA for the two splice variants present a differential expression in the fly head and body. Differences in their C-terminus end, affecting the sequence between two cystathionate beta synthase (CBS) domains, suggest that they might be regulated differentially, interact differentially with other cellular proteins to have altered function or membrane trafficking.

Materials and methods

Total mRNA was isolated from the head or body of individual flies (Oregon R strain) using Trizol (Invitrogen) and RT-PCR was performed as follows. Three µg of total RNA were reverse transcribed with the SuperScript II system (Invitrogen), using the oligo(dT) primer. For PCR amplification flanking primers for the whole DmClC-2 open reading frame (ORF) obtained from the GeneBank (Accession N^o NM169432) were used. Primers chosen were 598–616 gcttcatgtttaacaacagc and 4111–4133 tcaacgatgatttccgttttct. The reaction mixture contained aliquots of cDNA, 0.2 µM of each primer, 2.5 units Taq DNA polymerase (Fermentas), 2 µM dNTPs, and 1.5 µM MgCl₂ in a total volume of 50 µl. Conditions were: initial denaturation at 95°C for 2 min, 30 cycles at 95°C for 30 s, annealing at 56°C for 45 s and extension at 72°C for 3 min, and final extension at 72°C during 5 min. The PCR products were subcloned in the pGEM-T vector (Promega).

For electrophysiology experiments, HEK-293 cells were grown and transiently transfected with expression plasmids for each splice variant as described previously (Cid et al. [2000]). The pipette solution (35 mM Cl⁻) contained (mM): 100 sodium gluconate, 33 CsCl, 1 MgCl2, 2 EGTA and 10 Hepes pH 7.4 adjusted with Tris. The bath solution contained (mM): 140 NaCl, 2 CaCl2, 1 MgCl2, 22 sucrose and 10 Hepes pH 7.4. No interference with cationic currents was ever detected in these experiments. Liquid junction potentials were calculated (Barry [1994]) and corrected for when appropriate. Standard whole-cell patch- clamp recordings were performed as described elsewhere (Cid et al. [2000]). To obtain an estimate of apparent open probabilities, steady-state relative conductance was plotted as a function of voltage and adjusted to a Boltzmann distribution of the form: G=G₀+G_max/(1 + exp{(V − V_0.5)/k}, where G, G₀ and G_max are conductance, voltage-independent residual conductance and maximal conductance (extrapolated), respectively. V_0.5 is the voltage for 50% activation and k is the slope factor. Relative permeabilities of foreign anions were calculated from the changes in reversal potential after partial extracellular replacement (Díaz & Sepúlveda [1995]). The expression used was: P_X/P_Cl={¹[Cl]_o exp(−F▵E_rev /RT) − ²[Cl]_o}/[X]_o, where ¹[Cl]_o is the original extracellular Cl⁻ concentration and ²[Cl]_o and [X]_o are the extracellular concentrations of Cl⁻ and the foreign anion, respectively, after the change of solution. ▵E_rev is the new reversal potential minus the original, and R, T and F have their usual meanings. Significance of differences between means was determined using Student's unpaired t-test.

Results

Two products could be amplified by PCR using primers generated as described in the Experimental Procedures section. As seen on lane 3 of the gel in Figure 2B, only one amplicon could be obtained when mRNA isolated from the body of flies was used. The same amplicon was generated using mRNA recovered form the head, but an additional, larger product was also present (lane 2). Sequencing of these PCR products demonstrated that both encode proteins of the ClC family and are present in the FlyBase. Sequence analysis revealed closest similarity with the ClC-1 and ClC-2 branch of the ClC family. The smaller product, that we term DmClC-2S (FlyBase CG31116-RC), has 41% identity (50% in an overlap of 573 aa) with human ClC-2 channel. Similar alignment with human ClC-1 gave 37% identity (44% in an overlap of 573 aa). Figure 1 shows alignments for the two DmClC-2 predicted proteins and the human ClC-2 protein. Both putative Drosophila channels conserve key features of the ClC channels cloned so far. Seventeen plasma membrane α-helices (B-R, see Figures 1 and 2A) were predicted by comparison with a bacterial ClC protein of known molecular structure (Dutzler et al. [2002]). Residues thought to form the selectivity filter (aa 272 to 276 and Y622) are identical in the mammalian sequence. Another part of the selectivity filter in the fruit fly sequences, residues 526 to 530, presents two changes compared to mammalian ClC-2: A527 (M in ClC-2) and M529 (I in ClC-2). The glutamate residue at position 274, responsible for fast gating in ClC-2, is also conserved in both Drosophila sequences. Two CBS domains are also present in the carboxy-terminus. The larger protein, termed here DmClC-2L (Fly Base CG31116-RA), presents an extra region coding for 127 amino acids at the intracellular C-terminal portion of the protein, between the two CBS domains (Figures 1 and 2A). This difference accounts for the larger PCR product observed in Figure 2B. The structure of the CG31116 gene predicts 17 exons. DmClC-2S originates in the splicing of exons 12–15. A third transcript present in the Fly Base (CG31116-RD) was not detected as it has an alternative first exon upstream from the 5′ primer utilized here, but is otherwise identical to DmClC-2L. This makes it unlikely that the presence of exons 12–15 is the consequence of an incomplete splicing process.

Figure 1.  Amino acid sequence alignments for DmClC-2S, DmClC-2L and hClC-2 proteins. Putative transmembrane segments are labelled B-R. C-terminus CBS domains are identified as CBS1 and CBS2. Notice a 127 amino acid insertion between CBS domains in DmClC-2L. Black boxes identify identical residues. Grey means conservative changes. Numbering is for DmClC-2L.

Figure 2.  Secondary structure and RT-PCR for DmClC-2. (A) Putative secondary structure predicted for the DmClC-2 channels obtained by sequence comparison with a bacterial ClC protein of known molecular structure (Dutzler et al. [2002]). The cylinders labeled A-R represent α-helices with the extracellular side above and intracellular region below the structure. Notice the site of insertion of 127 amino acids at the C-terminus, between the CBS domains in the DmClC-2L variant. (B) RT-PCR products amplified from fly mRNA. Lane 1, 1000bp ladder; Lane 2, head mRNA; Lane 3, body mRNA; Lane 4, water control for PCR.

When the fly cDNAs were transfected into HEK-293 cells, currents were generated which displayed a similar behaviour as that observed with their guinea-pig, rat and human ClC-2 orthologs (Cid et al. [2000], Varela et al. [2002], Niemeyer et al. [2004]). Figure 3A and B show currents recorded in cells transfected with DmClC-2S and DmClC-2L cDNA respectively. When voltages from −10 to −190 mV were applied, both splice variants developed similar inwardly rectifying currents, which were small at moderately negative potentials and became larger with stronger hyperpolarization. There was no significant outward current. Figure 3C and D plot the steady-state activation as a function of voltage for both channels. Boltzmann functions can be adjusted to the points, yielding values (means±SEM) for V_0.5 and k of −119±14 mV and −19.3±3 mV for DmClC2S (Figure 3C). Respective values for DmClC-2L (Figure 3D) were −110±6 mV and −23.4±3 mV. The differences between these parameters was not significant and they are close to those measured with ClC-2 channels of mammalian origin studied under similar conditions (Cid et al. [2000], Varela et al. [2002], Niemeyer et al. [2004]). The relative selectivity of DmClC-2 channels to anions was studied by partial Cl⁻ replacement. The chloride current was activated by hyperpolarization and then a fast voltage ramp allowed the measurement of reversal potentials. With high extracellular Cl⁻, the reversal potential was −36 mV, close to the predicted chloride reversal potential (−35 mV), and consistent with negligible cation permeation (Figure 3E). Cl⁻ was the most permeant anion, with replacement by other anions leading to a positive shift in reversal potential and increasing degree of blockade for Br⁻ NO₃⁻ and I⁻. The relative permeability sequence was Cl⁻>Br⁻>NO₃⁻> I⁻≫ gluconate for both variants (Figure 3F). There was no difference in selectivity between the Drosophila channel variants, or with mammalian ClC-2 channels, despite the changes affecting the selectivity filter at positions 527 (M for A) and 529 (I for M), which must therefore be considered conservative in terms of channel function.

Figure 3.  Functional assay of DmClC2 channel activity in HEK-293 cells. (A) and (B) show representative whole cell current traces elicited from a holding potential of −10 mV in response to pulses raging from −10 to −190 mV in 10 mV steps. These pulses were followed by a depolarization to 30 mV. The duration of the main pulses was increased at more positive voltages in order to approximate full activation of the conductance. For illustration proposes, the beginning of the tail currents at 30 mV were set at the same time. (C) and (D): Steady-state activation as a function of voltage for DmClC-2S and DmClC-2L respectively. Values are means ±SEM of n=7 for DmClC-2S and n=5 for DmClC-2L. Continuous lines are Boltzmann fits. (E) Anion selectivity experiment for DmClC-2S. The current-voltage relations were obtained by activating the channels by a pulse to −130 mV followed by a 100 ms voltage ramp to 30 mV. The extracellular solution contained 146 mM Cl⁻ or 16 mM Cl⁻ plus 130 mM of the indicated anion. (F) Summary of the selectivity experiments for DmClC-2S (filled bars) and DmClC-2L (open bars). Values are means ±SEM of five experiments for DmClC-2S. The results of a single experiment for DmClC-2L are shown.

To explore possible differences between DmClC-2 splice variants, we examined the rates of activation as a function of applied voltage. As for mammalian ClC-2, a double exponential plus an instantaneous component equation could be fit to the current relaxations. Figure 4A shows the time constants for the two resolved components. They were both voltage dependent, becoming faster with hyperpolarization, and there was no difference between the variants. The opening time constants for DmClC-2 were ∼3-fold slower than in mammalian ClC-2. The channels closed completely during a pulse to 30 mV (Figure 4A and B), and this relaxation followed a triple exponential time course. Figure 4B shows a summary of the results of such fit for the deactivation during a 30 mV pulse after activation to −130 mV. There were no differences in closing rates between the Drosophila splice variants. The deactivation rate was largely dominated, with a fractional weight of about 0.75, by a time constant of around 10 ms. This compares with a fast time constant of the order of 100 ms (weight ∼0.5) for the fastest closing rate measured in mammalian ClC-2 (Cid et al. [2000]).

Figure 4.  Time constants for activation and deactivation of DmClC-2 Cl⁻ currents. (A) and (B): Results of fitting a double exponential plus an instantaneous component equation to the activation process. Results for DmClC-2S are shown as open symbols and for DmClC-2L as solid symbols. Voltage-dependence of the slow (τ_s) and fast time constant (τ_f) are shown in (A). A_s, A_f and A₀ are fractional amplitudes for the slow, fast and instantaneous term respectively, and are shown in (B). Time constants (C) and fractional amplitudes (D) for the deactivation process during a pulse to 30 mV. A triple exponential equation was fitted and the time constants obtained are numbered 1-3. A₁ to A₃ are the fractional amplitudes. All values shown are means ±SEM of seven experiments for DmClC-2S (open symbols) and four experiments for DmClC-2L (solid symbols).

The DmClC-2L channel presents 13 consensus protein kinase C (PKC) phosphorylation sites. Two of those are in the N-terminal segment and, of the remaining eleven, six are in the extra 127 amino acid stretch between the CBS domains present in this variant. To test whether PKC could regulate channel activity, we recorded currents for both DmClC-2 channels, and added the PKC activator phorbol 12-myristate 13-acetate (PMA) in the bath solution. PMA (100–1000 nM) was without effect on the DmClC-2 variants-mediated currents (data not shown). There are 4 potential protein kinase A phosphorylation sites in DmClC2S. Assays with 100 µM dibutyryl cAMP did not show any activation of DmClC-2S currents (not shown).

Discussion

There are nine mammalian ClC genes, encoding Cl⁻ channels or H⁺/Cl⁻ antiporters. They fall into four branches: ClC-1/ClC-2; ClC-Ka/ClC-Kb; ClC-3/ClC-4/ClC-5; and ClC-6/ClC-7; only the first two groups are plasma membrane Cl⁻ channels (Jentsch et al. [2002]). The Drosophila genome has been completed and annotation is advanced (Celniker & Rubin [2003], Yandell et al. [2005]). There are three ClC genes in the fruit fly and they correspond most closely to the ClC-1/ClC-2 branch (FlyBase CG31116), the ClC-6/ClC-7 branch (FlyBase CG8594) and the ClC-3/ClC-4/ClC-5 branch (FlyBase 5284) (Dow & Davies [2003]). There is no predicted product to represent the kidney ClC-K channels in the fly. It would appear, therefore, that the only candidate gene to produce plasma membrane Cl⁻ channels in Drosophila is that related to the ClC-1/ClC-2 branch. We show that two splice variants of this putative ClC channel can be amplified from adult fly head, whilst only one was present in the body. Their sequences are most closely related to human ClC-2 and on that basis (but see also functional characteristics) we have termed them DmClC-2. Recently, a new variant for DmClC-2, with an alternative start site upstream of the primer that we used, was reported (FlyBase CG31116-RD).

Sequence comparison of the two DmClC-2 proteins with that of a bacterial homologue of known molecular structure (Dutzler et al. [2002]), suggests the presence of 17 membrane α-helices. The residues corresponding to the putative selectivity filter, and in particular a glutamate that is believed to form the gate of the channel (Dutzler et al. [2003], Niemeyer et al. [2003]), are all conserved in the fly sequences. Two highly conserved CBS domains are also present in DmClC-2. CBS domains are believed to form AMP or ATP binding sites and act as regulators of protein function (Scott et al. [2004]), and they have been proposed to mediate intracellular nucleotide effects in human ClC-2 and ClC-1 (Niemeyer et al. [2004], Bennetts et al. [2005]). CBS domains have also been shown to be involved in determining the subcellular distribution, trafficking and gating of ClC-1 and ClC-2 (Estévez et al. [2004], Hebeisen et al. [2004], Peña-Münzenmayer et al. [2005], Zúñiga et al. [2005]). The difference between the DmClC-2 variants is a 127 aa insertion between CBS-1 and -2. It has been postulated that the CBS domains within a single monomer interact with each other (Estévez et al. [2004]). The presence of a large intervening insertion might alter this interaction, with consequences for function, membrane trafficking or domain targeting.

DmClC-2 channels share the functional hallmarks of mammalian ClC-2 channels. They activate at hyperpolarized potentials and are inwardly rectifying. Their steady-state voltage-dependence parameters, V_0.5 and slope factor of the activation curve, are indistinguishable between the two fly variants and very similar to those of mammalian ClC-2 channels. Similarly, anion selectivity does not vary between fly channels and is similar to that of mammalian ClC-2. The only apparent functional differences between DmClC-2 variants and the mammalian ClC-2 are a slower opening at negative potentials and a markedly faster closing rate at positive potential.

Despite the fact that there are six putative PKC-phosphorylation sites in the extra amino acids inserted between CBS domains of DmClC-2L, we could not see any effect of a PKC activator on the currents. We cannot rule out, however, that PKC-regulation, or other effects of the insertion in DmClC-2L, might require the presence of an accessory protein. Recently an interaction between the heat shock protein Hsp90 and mammalian ClC-2 channel has been reported to affect function and plasma membrane expression of the channel (Hinzpeter et al. [2006]). As CBS domains can mediate effects on ClC-2 gating and membrane trafficking, it has been speculated that Hsp90 might be exerting its effects through interaction with the CBS domains (Cid et al. [2006]). We could, therefore, speculate that expression in cells with appropriate partner proteins might reveal functional differences between DmClC-2L and S that pass unnoticed in the heterologous expression system. Similarly, we did not see an effect of cAMP on DmClC-2 currents.

There are few clues about the localization of DmClC-2 in Drosophila. Studies of gene expression during development show the presence of DmClC-2 transcript in some cells of the midline primordium at s11, and in midline glial cells at s16 (Kearney et al. [2004]). The expression of the gene has also been reported in the ventral nerve cord, ventral midline and lateral cord glia, at stages s13–16 (Tomancak et al. [2002]). This might indicate a neuronal or glial expression in the adult fly, related to proposed roles of mammalian ClC-2 in the brain. In neurons expressing inhibitory GABA_A receptors, ClC-2 is proposed to limit intracellular Cl⁻ accumulation (Staley et al. [1996]). In astrocytes, ClC-2-like channels have been implicated in extracellular K⁺ buffering, and pH− and cell volume- regulation (Walz [2002]). ClC2-like currents have been recorded in fly photoreceptors (Ugarte et al. [2005]). Similar anion selectivity and rectification properties suggest that DmClC-2 could be the molecular counterpart of the current in photoreceptors and play some role in phototransduction. In mouse, genetic inactivation of ClC-2 leads to blindness through photoreceptor degeneration. It has been proposed that ClC2 activity in retinal pigmented epithelium and neuroretina, might be essential for photoreceptor development (Bösl et al. [2001]).

All three ClC Drosophila genes are expressed the Malpighian (renal) tubule, with DmClC-2 being the most abundant (Wang et al. [2004]). This might correspond to DmClC-2S, given the confinement of DmClC-2L to the head. It has been speculated that the channel might mediate the massive secretory flow of salt and water that occurs in the tubule (Wang et al. [2004]). The inwardly rectifying nature of DmClC-2, makes it suitable for Cl⁻ efflux from the cell, predicting its presence at an apical location in a secretory epithelium. The other mammalian channel related to DmClC-2 is the muscle Cl⁻ channel ClC-1 (Jentsch et al. [2002]), activated by depolarization and essential for the repolarization of muscle action potential. Mutations altering ClC-1 function lead to human myotonias (Jentsch et al. [2005]). The inward rectification of DmClC-2, makes it unlikely that they could fulfil a similar role in the fly muscle.

In summary, we demonstrate that one of the Drosophila genes predicted to encode for transcripts of ClC-type channels, does indeed generate products whose functional characteristics are closest to the mammalian ClC-2 Cl⁻ channel. Drawing a parallel with the function of ClC-2 in mice and humans, the Drosophila channels could be involved in epithelial transport. In the central nervous system, they could play roles in neurons and glia. Future mutagenesis work in the fly should throw light on the role of the DmClC-2 channels in the Drosophila.

This paper was first published online on prEview on 24 January 2006.

Research funded by Fondecyt Grant 1030627. CECS is funded by a Millennium Science Initiative Institute grant and by Fundación Andes, the Tinker Foundation and Empresas CMPC. C.A.F. was supported by CONICYT Chile.