The N Terminus of a Schistosome β Subunit Regulates Inactivation and Current Density of a Cav2 Channel*

The β subunit of high voltage-activated Ca2+ (Cav) channels targets the pore-forming α1 subunit to the plasma membrane and tunes the biophysical phenotype of the Cav channel complex. We used a combination of molecular biology and whole-cell patch clamp to investigate the functional role of a long N-terminal polyacidic motif (NPAM) in a Cavβ subunit of the human parasite Schistosoma mansoni (βSm), a motif that does not occur in other known Cavβ subunits. When expressed in human embryonic kidney cells stably expressing Cav2.3, βSm accelerates Ca2+/calmodulin-independent inactivation of Cav2.3. Deleting the first 44 amino acids of βSm, a region that includes NPAM, significantly slows the predominant time constant of inactivation (τfast) under conditions that prevent Ca2+/CaM-dependent inactivation (βSm: τfast = 66 ms; βSmΔ2–44: τfast = 111 ms, p < 0.01). Interestingly, deleting the amino acids that are N-terminal to NPAM (2–24 or 2–17) results in faster inactivation than with an intact N terminus (τfast = 42 ms with βSmΔ2–17; τfast = 40 ms with βSmΔ2–24, p < 0.01). This suggests that NPAM is the structural determinant for accelerating Ca2+/calmodulin-independent inactivation. We also created three chimeric subunits that contain the first 44 amino acids of βSm attached to mammalian β1b, β2a, and β3 subunits. For any given mammalian β subunit, inactivation was faster if it contained the N terminus of βSm than if it did not. Co-expression of the mammalian α2δ-1 subunit resulted in doubling of the inactivation rate, but the effects of NPAM persisted. Thus, it appears that the schistosome Cav channel complex has acquired a new function that likely contributes to reducing the amount of Ca2+ that enters the cells in vivo. This feature is of potential interest as a target for new antihelminthics.

The cytoplasmic ␤ subunit of high voltage-activated Ca 2ϩ (Ca v ) channels targets the pore-forming ␣ 1 subunit to the plasma membrane and affects its biophysical phenotype (1,2). An important function of Ca v ␤ subunits is to modulate the inactivation rate of Ca v channels. Inactivation of Ca v channels is a life-sustaining property that maintains tight control of intracellular levels of Ca 2ϩ (3). For instance, in neurons, the spatialtemporal dynamics of Ca 2ϩ microdomains, which determine qualitative and quantitative aspects of neurotransmission, depend on the biophysical properties of the Ca v channels, such as conductance and inactivation rate as well as the nature of the intracellular buffers (for review, see Ref. 4).
Invertebrate Ca v channels are less well characterized than mammalian Ca v channels. However, Ca v channels are validated targets widely exploited in pharmacotherapy as well as by naturally occurring toxins (5). Understanding the idiosyncrasies of invertebrate Ca v channels could be useful for design of appropriate therapeutic strategies against invertebrate pathogens such as schistosomes. Schistosomes are parasitic flatworms that are the causative agents of schistosomiasis, a tropical disease that affects an estimated 200 million people worldwide. The current drug of choice against schistosomiasis is praziquantel, which affects Ca 2ϩ homeostasis in adult worms (6). Adults of Schistosoma mansoni express Ca v 2 and Ca v 1 orthologues as well as two ␤ subunit subtypes (7,8). The modus operandi of the schistosome Ca v channels remains largely unknown primarily due to technical challenges in expressing ␣ 1 subunits and to fast rundown of Ca 2ϩ currents in native preparations (9,10). It is expected that these channels would express better in helminthic clonal cell lines as these may have some specific factors that are needed for functional expression of these ␣1 subunits; to our knowledge these cells lines are not yet available. In contrast, the schistosome cytoplasmic ␤ subunits express robustly in mammalian cells (11) and in Xenopus oocytes (8). The less conventional of the two ␤ subunits has unusual structural and functional properties and appears to be involved in the action of praziquantel (8). In contrast, less data are available for the more conventional schistosome ␤ subunit, which does not appear to play a role in praziquantel action. This subunit, referred to as ␤ Sm heretofore, resembles mammalian ␤ subunits in that it significantly increases current density, modulates steady-state inactivation, and displaces the peak of the voltage-current relationship to the hyperpolarized direction. However, unlike other ␤ subunits, ␤ Sm has a long N-terminal poly acidic motif (abbreviated as NPAM herein) of 15 aspartate and glutamate residues that mediates rapid rundown of Ca v 2.3 currents (11). Although NPAM appears to represent the main difference between schistosome and mammalian ␤ subunits, one has to consider that even very minor differences between ion channel subunits can have major physiological and pharmacological consequences. Indeed, subtle changes are currently exploited in a variety of selective drugs and toxins. For example, pyrethroids, which are commonly used as insecticides, target the voltage-gated Na channels (Na v ) of insects but not those of mammals despite the extensive structural and functional simi-larities between the Na v channels of insects and mammals. Indeed, a single amino acid is responsible for the sensitivity of insect Na v channels to pyrethroid insecticides (12,13). Similarly, the capability of the other schistosome (variant) ␤ subunit to confer praziquantel sensitivity to an otherwise non-susceptible ␣1 subunit can be abolished by mutating a single amino acid residue (14). In another case the sensitivity of the GABA receptor to ethanol requires a string of 8 amino acids in one of the ␥ subunits (15). By analogy with this latter study, it is reasonable to suppose that the 15-amino acid-long polyacidic motif of the schistosome ␤ Sm subunit could serve as a drug target itself or might influence the action of drugs on schistosome Ca v channels. Nevertheless, the rationale for our study goes beyond identifying a pharmaceutical target against schistosomes and is geared to obtaining a better understanding of an important player in the physiology of the neuromuscular system of these invertebrates, i.e. Ca v channels. These insights will almost certainly be an advantage in the design and/or identification of effective pharmacophores.
Because acidic motifs that occur in channel and non-channel proteins bind Ca 2ϩ ions to modulate protein function (16,17), we hypothesized that the polyacidic motif in ␤ Sm would bind Ca 2ϩ to interfere with Ca 2ϩ /calmodulin-dependent inactivation. To this end we recorded Ca v 2.3 currents in conditions that allow Ca 2ϩ /calmodulin-mediated inactivation and in conditions that prevent it. Surprisingly, we found that the effects of NPAM 2 on Ca v inactivation appear to be Ca 2ϩ /calmodulinindependent. Here we show a detailed characterization of the role of NPAM in modulating the inactivation properties of Ca v 2.3 currents and provide a discussion of how the biophysical modulation of Ca v channels by NPAM-containing ␤ subunits, which are found so far only in parasitic trematodes, may contribute to the parasitic mode of life of these species.

EXPERIMENTAL PROCEDURES
Materials-Tissue culture dishes were purchased from Corning, Dulbecco's modified Eagle's media (DMEM) was purchased from Invitrogen, and poly-L-lysine, ATP, and the calcium phosphate transfection kit were purchased from Sigma. Restriction enzymes were purchased from New England Biolabs and oligonucleotide primers were from MWG biotech.
Cell Culture and Transfection-Human embryonic kidney (HEK) cells stably transfected with human Ca v 2.3d were cultured in DMEM supplemented with L-glutamine, glucose, and 10% fetal bovine serum in a humidified atmosphere (95%) with 5% CO 2 at 37°C. Cells were used for up to 25 passages and were split every 2-4 days. For electrophysiological recordings, cells were seeded in Petri dishes coated with poly-L-lysine, and transfection of ␤ subunits and, in some cases the ␣ 2 ␦-1 subunit (accession number AF286488.1), was performed with calcium phosphate on cells seeded on 60-mm Petri dishes at a confluence of 50 -60% using 10 g of the construct per dish. Using standard methods, we cloned all ␤ subunits into the pXOOM vector (18), which contains the gene for green fluorescent protein (GFP) as a marker for transfection.
Construction of ␤ Chimeras-Chimeric ␤ subunits were constructed by splicing the N-terminal region of the schistosome ␤ Sm (corresponding to amino acids  to the nearly fulllength coding regions of the three mammalian ␤ subunits. The respective portions were amplified from plasmid templates by PCR using GoTaq DNA Polymerase Green Master Mix (Promega) and primers containing appropriate restriction sites. For ␤ 1b and ␤ 3 chimeras, the forward primer for amplification of ␤ Sm (5Ј-GGGGATCCATGCAATGTTGTCGAGGATAT-TCA-3Ј) was designed to correspond to the first 8 amino acids of the sequence and contained a BamHI site at its 5Ј end. The ␤ Sm reverse primer was designed against amino acids 55-61 and contained a BglII site at its 5Ј end (5Ј-GGAGATCTTTTA-TAATCATCTTCATCATCT-3Ј). For the ␤ 2a chimeras, the ␤ Sm forward primer (5Ј-GGGGATCCATGCAATGTTGTCG-AGGATATTCA-3Ј) was designed to maintain the palmitoylated cysteine residues found in ␤ 2a at positions 3 and 4, thought to be important for the effects of this subunit on inactivation (19). The portions of the different mammalian ␤ subunits were amplified using subunit-specific primers with a BglII restriction site at the 5Ј end of the forward primers and a NotI restriction site at the 5Ј end of the reverse primers. For ␤ 1b (accession number X61394; amino acids 7-597), the forward primer was 5Ј-GGAGATCTATGTCCCGGGGCCCTTACCCA-3Ј; for ␤ 2a (accession number M80545; amino acids 9 -604), the forward primer was 5Ј-GGAGATCTTTTATAATCATCTTCATCA-TCT-3Ј; and for ␤ 3 (accession number M88751; amino acids 5-484), the forward primer was 5Ј-GGAGATCTTACG-TGCCCGGGTTTGAGGAC-3Ј. The reverse primers were: 5Ј-GGGCGGCCGCTCAGCGGATGTAGACGCCTTG-3Ј (␤ 1b ); 5Ј-GGGCGGCCGCTCATCATTGGCGGATGTATACAT-3Ј (␤ 2a ); 5Ј-GGGCGGCCGCTCAGTAGCTGTCCTTAG-GCC-3Ј (␤ 3 ). These reverse primers were all designed against the final amino acids and stop codon of the open reading frames. All fragments were amplified using standard PCR conditions (annealing temperature ϭ 50°C). After cleanup of the reactions over Qiaquik columns (Qiagen), the eluted products were digested with BglII, re-purified, and ligated. A small portion of this ligation mix was used as the template for another amplification of the full-length, chimeric construct using ␤ Sm forward and mammalian ␤ subunit reverse primers. This PCR product was gel-purified, digested with BamHI and NotI, and ligated into the BamHI/NotI-digested pXOOM, and this ligation was used to chemically transform Top 10 (Invitrogen)competent cells. All clones were sequenced, and only those without PCR (or other) errors were used for subsequent experiments. The strategy we used changes amino acid 47 of ␤ Sm from Glu to Arg.
Confocal Microscopy of Enhanced Green Fluorescent Protein (EGFP)-tagged ␤ Sm and ␤ Sm⌬2-44 Subunits-␤ Sm and ␤ Sm⌬2-44 were amplified by PCR using a reverse primer that removed the stop codon at the end of the open reading frame. Amplification with this primer, therefore, allowed for insertion of the ␤ subunit coding region into the pcDNA3-EGFP plasmid (Addgene #13031) such that the ␤ subunits would be tagged with EGFP at the C terminus. HEK cells stably expressing Ca v 2.3d were transfected with these EGFP-tagged ␤ subunit constructs as previously described. For confocal microscopy experiments, cells were plated onto poly-L-lysine-coated glass-bottomed 35-mm tissue culture dishes (MatTek). Confocal images of the cellular distributions of the EGFP-tagged ␤ Sm and its N-deletion mutant ␤ Sm⌬2-44 were acquired 24 h after transfection with a spinning disk confocal microscope (Leica) using a 100ϫ oilimmersion objective. Images were acquired using an argon laser (excitation, 488 nm; emission, BP emission filter).
Protein Extraction and Western Blots-Membrane proteins from EGFP-␤ Sm -or EGFP-␤ Sm⌬2-44 -transfected HEK cells were extracted using the ProteoJET Membrane Protein Extraction kit (Fermentas) according to the manufacturer's instructions. Briefly, cells were washed and permeabilized and then transferred to microcentrifuge tubes. After centrifugation at 10,000 ϫ g max , membrane protein extraction buffer was added to the supernatant and incubated for 30 min at 4°C with shaking, then the membrane protein extract was centrifuged at 10,000 ϫ g max for 15 min at 4°C to remove debris, and the supernatant was saved at Ϫ70°C for further analysis. Protein concentrations were determined using the Bradford Assay (Fermentas). Two micrograms of protein were separated on a 4 -12% Bis-Tris Nupage gel (Invitrogen) and transferred to a nitrocellulose membrane (Invitrogen). After blocking with TBS, Tween (TBST) plus 5% nonfat dry milk, the blot was incubated in a 1:1000 dilution of anti-GFP rabbit polyclonal antibody (ab-290, Abcam) in TBST plus 5% nonfat dry milk for 1 h at room temperature. After washing in TBST, blots were incubated in goat anti-rabbit IgG (1:15,000; Jackson ImmunoResearch), washed, and visualized with Supersignal West Pico Chemiluminescent Substrate (Thermo Scientific Pierce) according to the manufacturer's instructions.
Electrophysiology-Whole-cell patch clamp recordings were obtained 24 -48 h after transfection of the ␤ subunits using an Axopatch 200B (Molecular Devices). Cell capacitances were 10 -15 picofarads. Series resistance was compensated by 70%. Voltage pulses from Ϫ20 mV to ϩ70 mV were delivered in 5-mV intervals every 5 s from a holding potential of Ϫ80 mV. Data were acquired at sampling intervals of 50 s and filtered at 5 kHz during acquisition. The bath solution contained 10 mM CaCl 2 or BaCl 2 , 160 mM triethylammonium (TEA) chloride, 10 mM HEPES, and 0.1 mM EGTA, pH (TEA-OH) 7.4. Patch pipettes were pulled from borosilicate glass and fire-polished before each experiment and had resistances between 1 and 1.5 megaohms. Membrane seals were obtained by applying negative pressure. All experiments were performed at room temperature (22°C). The pipette solution contained 110 mM cesium methane sulfonate, 10 mM HEPES, 0.5 mM EGTA, and 5 mM Mg 2ϩ ATP, pH (CsOH) 7.3. In experiments designed to prevent Ca 2ϩ /calmodulin-dependent inactivation, EGTA was replaced with BAPTA (5 mM). In some control experiments, no chelators were added to the pipette solution. Currentvoltage relationships were obtained as a first step to ensure that the cell produced viable currents. Maximum amplitude currents were used to estimate the time constants of the fast and slow components of macroscopic inactivation. To estimate the kinetics of macroscopic inactivation, the decaying phases of maximal Ca 2ϩ currents evoked with depolarizing pulses from a holding potential of Ϫ80 mV were fitted to a double exponential equation of the form y ϭ y 0 ϩ A 1 exp(Ϫx/ 1 ) ϩ A 2 exp (Ϫx/ 2 ), where 1 and 2 are the time constants, A 1 and A 2 are the amplitudes, and y 0 is the offset. Some ␤ subunits induce faster inactivation than others. When inactivation was particularly slow, such as when ␤ 2a is coexpressed, longer pulses of at least 1 s were used. For consistency, the same portion of the pulse length is shown for all cases. To estimate the kinetics of macroscopic activation, the activating phases of maximal Ca 2ϩ currents evoked with depolarizing pulses from a holding potential of Ϫ80 mV were fitted to a single exponential equation of the form y ϭ y 0 ϩ A exp(Ϫx/). The voltage dependence of steady-state inactivation was determined by plotting the normalized peak current evoked by a depolarizing pulse to elicit the maximum current as a function of the voltage of a preceding 2-s pre-pulse test (between Ϫ120 and ϩ20 mV). Currents were normalized with respect to the current elicited from Ϫ120 mV. Steady-state inactivation curves were fitted by a Boltzmann function of the form where I max is the maximal current, V is the pre-pulse voltage, K is the slope factor, and V 0.5 is the voltage at which inactivation is half-maximal. Data were acquired with an Axopatch 200 B amplifier and Clampex 2.0 software and analyzed with Clampfit 2.0 software (Molecular Devices). Inactivation kinetics analyses were performed with IGOR PRO (WaveMetrics, Lake Oswego, OR). To determine the significance of differences in data, Student's t tests were performed. In the cases where the variances between the two groups being compared were significantly different, the Welch correction was applied.

RESULTS
We previously described an unusual, highly acidic domain (NPAM) near the N terminus of the "conventional" Ca v ␤ Sm subunit from S. mansoni (11). To test whether the NPAM region in ␤ Sm modulates Ca v channels in a Ca 2ϩ /calmodulindependent manner, we transfected HEK cells stably expressing Ca v 2.3 channels with wild type, deleted, or modified ␤ subunits that contain the schistosome NPAM domain and studied how they modulate inactivation of Ca v 2.3 in conditions that allow or suppress the Ca 2ϩ /calmodulin-dependent inactivation process. Fig. 1 shows a diagram of the experimental setup.
The N Terminus of ␤ Sm Accelerates Inactivation of Ca 2ϩ Currents through Ca v 2.3 when Ca 2ϩ Is Chelated with 5 mM Internal BAPTA-In this set of experiments 5 mM BAPTA was included in the pipette solution to prevent Ca 2ϩ /calmodulin-dependent inactivation in Ca v 2.3 channels, so that inactivation occurs via voltage-dependent inactivation (VDI) only (20,21). The subunits tested included native ␤ Sm ; ␤ Sm with the N-terminal 44 amino acids deleted (␤ Sm ⌬2-44), and chimeric mammalian ␤ subunits with the N-terminal region of ␤ Sm attached to the N terminus. The inactivating portion of the Ca 2ϩ currents was optimally fitted to a bi-exponential function. In these experiments, co-expression of ␤ subunits containing the N terminus of ␤ Sm robustly accelerated inactivation kinetics of Ca 2ϩ currents through Ca v 2.3 channels ( Fig. 2A). Both fast and slow components of inactivation were accelerated for all subunits, except for ␤ Sm -␤ 1b , which accelerated only the fast component of inactivation with respect to ␤ 1b (Fig. 2, B and C).

N Terminus of ␤ Accelerates Non-L-type Channel Inactivation
The N Terminus of ␤ Sm Does Not Significantly Modulate Ca 2ϩ Currents through Ca v 2.3 When Ca 2ϩ Is Chelated with 0.5 mM Internal EGTA-In these experiments 0.5 mM EGTA was included in the pipette solution, a condition that is expected to allow for accumulation of the intracellular Ca 2ϩ necessary for Ca 2ϩ /calmodulin-dependent inactivation of Ca v 2.3 (20,21). In most of these experiments, the addition or deletion of the N terminus of ␤ Sm to a ␤ subunit did not significantly alter the rate of inactivation of Ca v 2.3 currents (Fig. 3A). An exception is ␤ Sm -␤ 2a , which accelerated both fast and slow kinetic components of inactivation with respect to ␤ 2a (Fig. 3, B and C). Data obtained from whole-cell patch clamp experiments in which no chelators had been added to the patch solution (supplemental Fig. S1) were similar to those obtained with 0.5 mM EGTA in the patch pipette (Fig. 3).
The N Terminus of ␤ Sm Accelerates Inactivation of Ba 2ϩ Currents through Ca v 2.3-Taken together, the data shown in the previous section do not support our initial hypothesis that the acidic motif in the N terminus of ␤ Sm modulates the inactiva-tion process mediated by Ca 2ϩ /calmodulin. Instead, our data strongly point to the likelihood that the ␤ Sm N terminus accelerates the VDI process. To further confirm this finding, we tested the action of the N terminus of ␤ Sm on Ba 2ϩ currents, as Ba 2ϩ does not bind calmodulin (22), and therefore, using Ba 2ϩ as the charge carrier is an alternative approach to suppress Ca 2ϩ /calmodulin-dependent inactivation. As shown in Fig. 4A, Ba 2ϩ currents also inactivated significantly faster with coexpression of NPAM-containing ␤ subunits than with coexpression of ␤ subunits that did not contain NPAM. When NPAM was attached to ␤ Sm , ␤ 1b , or ␤ 3 , it accelerated only the predominant fast component of inactivation (Fig. 4, B and C). When attached to ␤ 2a , it modulated both fast and slow components of inactivation as when the combination Ca 2ϩ out /BAPTA in was used (compare with Fig. 3). To further assess the role of NPAM on inactivation kinetics of Ca v 2.3, we plotted the fast and slow time constants of Ba 2ϩ current inactivation as a function of voltage. Fig. 5, left and center panels, shows that inactivation kinetics were weakly dependent on voltage. The addition or deletion of NPAM did not alter the voltage sensitivity of inactivation kinetics. However, the relative weights of the two components of inactivation were modulated by NPAM (Fig. 5, right  panels). Specifically, the relative weight of the fast component was significantly augmented (concomitantly with a decrease of the relative weight of the slow inactivating component) in all cases when the coexpressed ␤ subunit contained NPAM. In the  To allow for accumulation of internal Ca 2ϩ , necessary to stimulate the Ca 2ϩ /calmodulin-dependent mechanism of inactivation of Ca v 2.3, 0.5 mM EGTA was included in the pipette solution. To prevent internal Ca 2ϩ accumulation and, thus, suppress the Ca 2ϩ /calmodulin-dependent mechanism of inactivation, 5 mM BAPTA was added to the pipette solution. Two versions for each tested ␤ subunit that differed in their N termini were compared; the schistosome ␤ Sm and the N-terminal deletion mutant ␤ Sm⌬2-44 , ␤ 1b and a chimera that contains the N terminus of ␤ Sm (␤ Sm -␤ 1b ), ␤ 2a and a chimera that contains the N terminus of ␤ Sm (␤ Sm -␤ 2a ), and ␤ 3 and a chimera that contains the N terminus of ␤ Sm (␤ Sm -␤ 3 ). The N terminus of ␤ Sm that was attached to the N termini of the mammalian ␤ subunits is magnified to show its structure; that is, a long NPAM is preceded by a shorter acidic motif and by an initial region of mostly uncharged amino acids. NOVEMBER 12, 2010 • VOLUME 285 • NUMBER 46 case of ␤ 3 , which increases the fast component of inactivation more than the other ␤ subunits, the addition of NPAM shifted all the weight of the inactivation process toward the fast component (Fig. 5D, right panel).

N Terminus of ␤ Accelerates Non-L-type Channel Inactivation
Acceleration of Ca v 2.3 Inactivation by the N Terminus of ␤ Sm Is Maintained in the Presence of ␣ 2 ␦-1-Because the ␣ 2 ␦-1 subunit is known to accelerate inactivation of Ca v 2.3 currents and ␣ 2 ␦-1 subunits are commonly found in high voltage-activated Ca v channel complexes in native cells (23), we set out to test the effect of the N terminus of ␤ Sm in the presence of an ␣ 2 ␦ subunit. Coexpression of rat ␣ 2 ␦-1 doubled the rate of both fast and slow components of inactivation. This effect was more pronounced when co-expressed with ␤ Sm , ␤ 1b , and ␤ 3 than with ␤ 2a . However, in the presence of ␣ 2 ␦-1, NPAM-containing ␤ subunits significantly further accelerated inactivation in all cases, except for ␤ 2a , where only a minor acceleration was observed (Fig. 6).
Deleting the Region N-terminal to NPAM in ␤ Sm Further Accelerates Inactivation-We next wanted to determine whether the region of ␤ Sm N-terminal to NPAM has a role in inactivation. In addition, there remained the possibility that the structural determinants for accelerating inactivation resided in this region rather than in NPAM itself. To this end we generated two additional N-terminal deletion mutants that left NPAM intact: ␤ Sm⌬2-17 , which lacks the first 17 amino acids, and ␤ Sm⌬2-24 , which lacks those same amino acids and additionally lacks a short acidic motif at position 18 -24 (Fig. 7A).
Whole-cell Ba 2ϩ currents were then recorded in the presence of either of these ␤ Sm versions. To our surprise, Ba 2ϩ currents inactivated faster with coexpression of either ␤ Sm⌬2-17 or ␤ Sm⌬2-24 than with coexpression of the full ␤ Sm (Fig. 7B). Both fast and slow components of inactivation were significantly accelerated by deleting the amino acids that precede NPAM (Fig. 7, C and D). This suggests that the region that precedes NPAM somehow suppresses NPAM-induced acceleration of Ca v 2.3 inactivation.
The N Terminus of ␤ Sm Accelerates Activation of Ca v 2.3-Although we focused on the inactivation kinetics, we could not fail to observe that currents that inactivated faster due to coexpression of NPAM-containing ␤ subunits also activated faster. Fitting the activating portion of the inward currents to single exponential functions yielded time constants of activation ( act ) for all channel combinations. In all cases, act had lower values (faster activation) in the presence of NPAM-bearing ␤ subunits than in the presence of the corresponding ␤ subunits without NPAM. In the case of ␤ 2a and ␤ 3 , this difference was statistically significant ( Table 1).
The N Terminus of ␤ Sm Decreases the Amplitude of Wholecell Currents through Ca v 2.3-Whole-cell Ba 2ϩ currents from cells co-expressing N-terminal-truncated ␤ Sm were consistently larger than currents produced in the presence of wild type ␤ Sm . Averaged current density-voltage plots of ␤ Sm as well as chimeric mammalian ␤ subunits show that this difference was significant in all cases except for ␤ 1b (Fig. 8, A-D). The

. The N terminus of a schistosome ␤ subunit (␤ Sm ) plays a significant role in modulating inactivation kinetics of Ca v 2.3 Ba 2؉ currents.
A, shown are whole-cell Ba 2ϩ currents through Ca v 2.3 elicited by a depolarizing pulse to maximum peak current from a holding potential of Ϫ80 mV with the following coexpressed ␤ subunits: ␤ Sm ; ␤ Sm⌬2-44 ; ␤ 1b ; the chimera ␤ Sm -␤ 1b ; ␤ 2a ; the chimera ␤ Sm -␤ 2a ; ␤ 3 ; or the chimera ␤ Sm -␤ 3 . Current amplitudes are normalized to highlight the differences in inactivation kinetics for each pair of ␤ subunits (with and without NPAM). Currents produced when NPAMcontaining subunits were coexpressed are shown in red. B, average values are shown of the fast time constant ( fast ) of macroscopic inactivation for Ca v 2.3 channels coexpressed with wild type and modified ␤ subunits. C, shown are average values of the slow time constant ( slow ) of macroscopic inactivation for Ca v 2.3 channels coexpressed with wild type and modified ␤ subunits. Fast and slow values were derived by two-exponential fits of the inactivating portions of Ca 2ϩ currents. Bars represent the mean Ϯ S.E. Statistical significance of differences between the means was obtained from unpaired Student's t tests. n ϭ 3-7.
activation portion of each individual I-V relationship was well fitted by a Boltzmann function. The presence of any ␤ subunit displaces the peak of the I-V to the left, which is well represented by the midpoints of activation given by this Boltzmann fit. However, midpoints of activation are not significantly altered by NPAM (Fig. 8E). I-V relationships obtained from patch clamp experiments where the cell interior was dialyzed with a chelator-free solution show that there are no appreciable differences between cells expressing NPAM-containing ␤ subunits and cells expressing NPAM-less ␤ subunits, except for ␤ 2a , which decreases average current amplitude if it contains the schistosome N terminus (supplemental Fig. S2).
To test whether the decrease of whole-cell current density by NPAMbearing ␤ subunits was due to reduced expression, we generated EGFP-tagged ␤ Sm and EGFP-tagged ␤ Sm⌬2-44 subunits and compared their subcellular distribution in HEK cells. Fig. 8F shows that there are no detectable differences between the cellular distributions of these two ␤ subunits. Both localize to the perinuclear membrane, to a localized region of the cytoplasm, presumably the endoplasmic reticulum and/or the Golgi apparatus, and to the plasma membrane. Supplemental Fig. S3 shows a Western blot of both subunits using a primary antibody against EGFP.
The N Terminus of ␤ Sm Does Not Modulate Steady-state Inactivation of Ca v 2.3-Because some structures of Ca v channels appear to have dual roles in determining inactivation kinetics and inactivation gating (24), we wanted to know whether the N terminus of ␤ Sm would also affect steady-state inactivation. However, we found no significant differences between midpoints of steady-state inactivation between the NPAM and non-NPAM versions for any given ␤ subunit (Fig. 9).

DISCUSSION
Here we report the role of a novel polyacidic motif of 15 aspartate and glutamate residues in a Ca v channel ␤ subunit from the human parasite S. mansoni (␤ Sm ). This motif is located at position 29 -44 in the N terminus and is preceded by a smaller acidic motif, at position 18 -24. A mutant  NOVEMBER 12, 2010 • VOLUME 285 • NUMBER 46 subunit that lacks the first 44 amino acids of ␤ Sm accelerates the decay of Ca 2ϩ currents through Ca v 2.3. We also created chimeric mammalian ␤ subunits that contained the N terminus of ␤ Sm . These chimeric ␤ Sm -␤ x subunits also accelerated inactivation of Ca v 2.3, compared with their wild type counterparts. Surprisingly, when the amino acids that precede the long polyacidic motif were deleted from ␤ Sm , inactivation of Ca v 2.3 currents occurred even faster than with an intact N terminus. The simplest explanation is that the amino acids that precede the polyacidic motif (NPAM) antagonize the NPAM function, perhaps by physically preventing NPAM from reaching its target. One can also hypothesize that the schistosome ␣1 subunit has an interaction site for these initial amino acids, which is not present in the mammalian ␣1 subunit. Taken together, the data presented here show that the N terminus of ␤ Sm contains structural determinant(s) with a role in accelerating inactivation, with the long polyacidic motif as a likely candidate.

N Terminus of ␤ Accelerates Non-L-type Channel Inactivation
Our data showing the accelerating effect of ␣ 2 ␦-1 on Ca v 2.3⅐␤ complexes is consistent overall with previous work. Interestingly, whereas Yasuda et al. (25), working with the mammalian cell line HEK ts-201, found that ␣ 2 ␦-1 does not accelerate the currents produced by Ca v 2.3⅐␤ 2a complexes, Qin et al. (23), using oocytes as the expression system, found that ␣ 2 ␦ accelerates inactivation of Ca v 2.3⅐␤ 2a complexes. We found that all Ca v 2.3⅐␤ complexes were accelerated by ␣ 2 ␦-1, but this effect was significantly less pronounced in the case of Ca v 2.3⅐␤ 2a . More specifically, the time constant of the fast inactivating component in the presence of ␤ 1b , ␤ Sm , and ␤ 3 was reduced by 50%, essentially doubling the rate of inactivation. When ␣ 2 ␦-1 was coexpressed with Ca v 2.3⅐␤ 2a , the time constant of the fast inactivating component was reduced by about 30% (compare Figs. 3B and 5B). Thus, because we observe a lesser effect of ␣ 2 ␦-1 on inactivation rate in the case of ␤ 2a , our data are consistent with the findings of Qin et al. (23) but do not completely agree with those of Yasuda et al. (25). Perhaps the causes for this difference lie in the slightly different materials and conditions used, such as the different cell lines, HEK ts-201 (25) and HEK AD 298 here. Given that ␤ 2a interacts with the plasma membrane, one might expect that ␤ 2a behaves differently in different expression systems. It is also possible that ␤ 2a modulation is affected by whether Ca v 2.3 is stably expressed (here) or whether it is transiently expressed (25).
Whereas recent work (26,27) has shown that the longer the N terminus of ␤ 1a , the greater its accelerating effect on a Ca v 1 channel, our data (Fig. 6) show that the N terminus of ␤ Sm does not follow this rule, as shortening the N terminus by deleting the first 27 amino acids (which precede the polyacidic motif) A, shown are whole-cell Ba 2ϩ currents through Ca v 2.3 elicited by a depolarizing pulse to maximum peak current from a holding potential of Ϫ80 mV with coexpressed ␣ 2 ␦-1 and the following ␤ subunits: ␤ Sm ; ␤ Sm⌬2-44 ; ␤ 1b ; the chimera ␤ Sm -␤ 1b ; ␤ 2a ; the chimera ␤ Sm -␤ 2a ; ␤ 3 ; or the chimera ␤ Sm -␤ 3 . Current amplitudes are normalized to highlight the differences in inactivation kinetics for each pair of ␤ subunits (with and without NPAM). Currents produced when NPAM-containing subunits were coexpressed are shown in red. B, average values of the fast time constant ( fast ) of macroscopic inactivation for Ca v 2.3 coexpressed with ␣ 2 ␦-1 and wild type or modified ␤ subunits. C, shown are average values of the slow time constant ( slow ) of macroscopic inactivation for Ca v 2.3 coexpressed with ␣ 2 ␦-1 and wild type or modified ␤ subunits. Fast and slow values were derived by two-exponential fits of the inactivating portions of Ca 2ϩ currents. Bars represent the mean Ϯ S.E. Statistical significance of differences between the means was obtained from unpaired Student's t tests. n ϭ 3-6.

TABLE 1 Activation time constants ( activation ) of Ba 2؉ currents produced by Ca v 2.3 channels in combination with wild type and modified ␤ subunits from S. mansoni (␤ Sm ) and mammals
activation was derived from single exponential fits to the activating portion of the inward currents. To generate chimeric ␤ subunits, the N terminus of ␤ Sm , which contains a long polyacidic motif of 15 glutamate and aspartate residues (NPAM), was attached to the N termini of mammalian chimeric ␤ subunits. Data represent the mean Ϯ S.E. n is shown in parentheses.

N Terminus of ␤ Accelerates Non-L-type Channel Inactivation
accelerates inactivation of Ca v 2.3. If the length of the N terminus were the only factor that determined the degree of the accelerating effect of ␤ Sm on Ca v 2.3, then deleting 27 amino acids, a significant length by comparison with the experiments performed by Herzig and collaborators (26,27), should have resulted in slowing of inactivation compared with that seen with wild type ␤ Sm . Instead, we observed increased acceleration of inactivation. These authors also found that the N terminus of ␤ 1a does not play a role in determining current density of an L-type channel, a result in stark contrast to our data showing that presence of the ␤ Sm N terminus results in reduced current density (Fig. 8). Clearly, the function of the N terminus of ␤ Sm , and NPAM in particular, is quite different from that of the mammalian ␤ subunits. A Possible Mechanism(s) Used by the N Terminus of ␤ Sm to Accelerate Inactivation of Ca v 2.3-The fact that acceleration of Ca v 2.3 inactivation mediated by the N terminus of ␤ Sm occurred only when Ca 2ϩ out /BAPTA in was used or when currents were carried by Ba 2ϩ indicates that this motif operates independently of the classic Ca 2ϩ /calmodulin-dependent mechanism. Although these results strongly point to VDI as the feature that is modulated by ␤ Sm to accelerate inactivation, we cannot rigorously rule out additional or alternative mechanisms of action. These are discussed in the next paragraphs.
VDI-The fact that currents consistently activated faster with coexpression of NPAM-containing ␤ subunits than with coexpression of NPAM-lacking ␤ subunits, would favor a model in which this motif modulates the voltage-dependent mechanism of inactivation. It is generally accepted that voltage dependence of inactivation occurs because the regions of the channel involved in this mechanism are coupled with the parts of the ion channel that sense a change in transmembrane voltage and move to produce VDI (28). This view of VDI largely derives from a model proposed by Aldrich et al. (29,30) in which a slowing in macroscopic inactivation is the result of a delay in channel opening at the single channel level. The ration- FIGURE 8. The N terminus of a schistosome ␤ subunit (␤ Sm ) reduces current density of Ca v 2.3 Ba 2؉ currents. A-D, shown are one-to-one comparisons of whole-cell current density-voltage relationships between cells coexpressing wild type ␤ subunits and ␤ subunits with modified N termini: ␤ Sm or the N-terminally deleted ␤ Sm⌬2-44 (A), ␤ 1b or the chimera ␤ Sm -␤ 1b (B), ␤ 2a or the chimera ␤ Sm -␤ 2a (C), ␤ 3 or the chimera ␤ Sm -␤ 3 (D). Open circles represent data obtained with coexpression of ␤ subunits that naturally or artificially lack NPAM; filled circles represent data obtained with coexpression of ␤ subunits that naturally or artificially contain NPAM. Current density-voltage relationships for Ca v 2.3 channels expressed alone are also shown (squares) for comparison. Depolarizing voltage-steps were delivered from a holding potential of Ϫ80 mV in 5-mV steps. E, shown is the midpoint of voltage dependence of activation, derived from Boltzmann fits to the activating portion of I-V curves. F, shown are differential interference contrast and fluorescence confocal images of HEK cells transfected with C terminus EGFP-tagged ␤ Sm and EGFP-tagged ␤ Sm⌬2-44 subunits. Data points represent the mean Ϯ S.E. n ϭ 3-5.  NOVEMBER 12, 2010 • VOLUME 285 • NUMBER 46 ale is that if channel opening is delayed, many channels would open while the channel inactivates, with an overall slowing of macroscopic inactivation. In this context, a possible explanation for our data would be that the acidic motif induces channel conformations that open more readily in response to depolarization, with the ensuing faster overall inactivation kinetics that fit the currently accepted model for activation-inactivation coupling of voltage-dependent ion channels. If, however, NPAM accelerates inactivation of Ca v 2.3 by interacting with structures that mediate VDI, one wonders what these might be, because in Ca v channels VDI appears to be a complex mechanism that depends on global conformational changes of the ␣1 subunit. The S6 transmembrane segments of the four domains, especially domains II and III, have an important role in VDI (31). However, the S6 segment of domain II appears to be exclusively involved in inactivation kinetics but plays no role in inactivation gating (24). Because the N terminus of ␤ Sm did not affect steady-state inactivation (Fig. 7), it is possible that NPAM interacts with a region exclusively involved in VDI such as S6 of domain II.

N Terminus of ␤ Accelerates Non-L-type Channel Inactivation
G-protein-mediated Inactivation-It is well known that Ca v 2 channel function is directly inhibited by G-protein ␤␥ subunits (for review, see Ref. 32). Furthermore, the binding site for G protein ␤␥ is in close proximity with the binding site for Ca v ␤ in the I-II linker (33,34). This supports the notion that the ␤ subunits regulate association between ␣1 subunits and membrane-bound G proteins. It is thought that G-protein-mediated inhibition is caused by stabilizing the closed state, making the channel reluctant to open (35,36). Therefore, an alternative hypothesis to that suggested in the previous section is that the acceleration of activation and inactivation induced by NPAM is not mediated by modulating the voltage sensors of Ca v 2.3 but by challenging association between G-proteins and Ca v 2.3. Less association with G-proteins would favor the process of activation, which would, therefore, result in accelerated inactivation as these two processes are coupled.
Ca 2ϩ Binding-Because polyacidic clusters in other voltagegated ion channels are involved in Ca 2ϩ binding (16,17), a model in which a putative Ca 2ϩ -sensing function of NPAM leads to fast inactivation remains a possibility. The finding that acceleration was also induced when Ba 2ϩ was used as the charge carrier cannot be regarded as evidence against this hypothesis, because the negatively charged residues of NPAM could interact electrostatically with Ba 2ϩ as well as with Ca 2ϩ ions.
NPAM in the Context of ␤ Subunit Structure-In the structural model proposed by Van Petegem et al. (37), the N terminus of the ␤ subunit is positioned away from the inner mouth of the channel and toward the plasma membrane that surrounds the channel. This model is based on ␤ 2a subunits, whose palmitoylation motif in the N terminus interacts with the plasma membrane. However, it is not known how the N terminus of other ␤ subunits is positioned (38). The current models of the three-dimensional structure of the ␤ subunits (for review, see Ref. 39) focus on core regions of ␤ subunits and their interaction with the I-II loop of the ␣1 subunit. Nevertheless, drawing from the model for ␤ 2a (37), we hypothesize that the addition of the polyacidic motif disrupts the interaction of the palmitoylation motif with the plasma membrane, thus preventing interaction of the N terminus with the plasma membrane.
Physiological Significance-Although it is fair to presuppose that the conditions in supplemental Fig. S2 (no internal chelators) or the addition of mild chelators (0.5 mM EGTA) are reasonable approximations of normal physiological conditions, by analogy with the mammalian system, it is not clear that they are physiological in the case of schistosome cells. Therefore, the fact that NPAM has little role in modulating Ca v activity under these low buffering conditions in vitro does not necessarily indicate that NPAM does not modulate Ca v channels in the native conditions. In fact, there are data indirectly suggesting that schistosomes have strong intracellular Ca 2ϩ buffering mechanisms. For instance, adult schistosomes reside in blood vessels and feed on blood, which contains significant amounts of Ca 2ϩ ; indeed, Shaw and Erasmus (40) propose the existence of special or additional cellular mechanisms to prevent excess Ca 2ϩ influx. Perhaps ␤ Sm represents one of these mechanisms, working to reduce Ca 2ϩ influx by increasing the rate of Ca v channel inactivation (and decreasing current density). The second, "variant" schistosome ␤ subunit also appears to dampen Ca 2ϩ currents (8). Additionally, it is known that schistosomes express a variety of Ca 2ϩ -buffering proteins. In parasites, these proteins appear to have an important role in adapting the parasite to the various environments they encounter in their complex life cycles and are also likely involved in secretion of proteins with roles in neutralizing host attack (41). The function of these buffering agents may not be restricted to host defense, but they may also serve to counteract deleterious increases in intracellular Ca 2ϩ . According to our data, the polyacidic motif of ␤ Sm would be functionally relevant in this buffered environment, further contributing to minimize an intracellular increase in Ca 2ϩ .
Modulation of channel kinetics without a concomitant modulation of state-state inactivation, although uncommon, may be relevant physiologically. For instance, consider the relatively slow inactivating rate of the cardiac L-type Ca v channels that leads to accumulation of Ca 2ϩ , which in turn determines the plateau of the cardiac action potential, making possible the con- traction of the heart. Mutations that minimally affect the inactivation rate of these channels are likely to have a significant impact on cardiac physiology even if they do not affect the voltage dependence of inactivation. In fact, data in support of this idea have been published and show that a splice variant of ␤ 2a , termed ␤ 2c , with a relatively short N terminus, is expressed in the heart and that these two variants differ in their inactivation kinetics but not in steady-state inactivation (42). The magnitude of the difference in inactivation kinetics is comparable with that observed in our study and is likely to have significant effects on cardiac physiology. By analogy, it seems likely that changes in inactivation kinetics of schistosome Ca v channels will have an impact in important physiological processes of these organisms. Non L-type Ca v 2 channels play important roles in synaptic transmission (43)(44)(45)(46), secretion (47,48), and long term potentiation (49). One would expect that both Ca v 2 channels of S. mansoni (6) are also relevant to the correct functioning of the nervous system of the parasite. However, it is possible that the environment of adult S. mansoni in the human circulatory system does not demand as many Ca 2ϩ -dependent responses as the lifestyle of free-living flatworms. Interestingly, none of the Ca v ␤ subunits of the free-living flatworms whose genome is available contain NPAM-type motifs, whereas all Ca v ␤ subunits from parasitic trematodes examined to date contain polyacidic motifs of 15 amino acids or longer (Table 2). Thus, the polyacidic motif in the N terminus of the only conventional ␤ subunit in S. mansoni (7) could represent a new gene function that appeared after the evolution of the parasitic lifestyle. In their native context, ␤ Sm subunits may decrease the total amount of Ca 2ϩ that enters through Ca v channels not only by accelerating inactivation and by decreasing current density (this work) but also by promoting Ca 2ϩ current run down (11). The putative Ca 2ϩ binding properties of NPAM suggest possible additional roles in the schistosome neuromuscular system, including perhaps as a transcription factor.
Conclusion-In summary, we have identified a structural determinant of Ca v channel inactivation, a polyacidic motif in the N terminus of a Ca v ␤ subunit, that differs greatly from those previously known. We have discussed several mechanisms by which this long polyacidic motif may induce fast kinetics of inactivation; (i) by modifying voltage-dependent inactivation, (ii) by preventing interaction of ␣1 and G-proteins, and (iii) by means of its putative Ca 2ϩ binding function. This study highlights a unique feature of the schistosome Ca v channel complex that is potentially amenable to pharmacological manipulation.