Advertisement
Research ArticleArticle

Interaction of MDIMP with the Voltage-Gated Calcium Channels

Juan A.M. De La Rosa, Maricela García-Castañeda, Takuya Nishigaki, Juan Carlos Gómora, Teresa Mancilla-Percino and Guillermo Ávila
Molecular Pharmacology September 2020, 98 (3) 211-221; DOI: https://doi.org/10.1124/mol.120.119982

Abstract

Amino acid–derived isoindolines are synthetic compounds that were created with the idea of investigating their biological actions. The amino acid moiety was included on the grounds that it may help to avoid toxic effects. Recently, the isoindoline MDIMP was shown to inhibit both cardiac excitation-contraction coupling and voltage-dependent calcium channels. Here, we revealed that MDIMP binds preferentially to low-voltage–activated (LVA) channels. Using a holding potential of −90 mV, the following IC50 values were found (in micromolars): >1000 (CaV2.3), 957 (CaV1.3), 656 (CaV1.2), 219 (CaV3.2), and 132 (CaV3.1). Moreover, the isoindoline also promoted both accelerated inactivation kinetics of high-voltage–activated Ca2+ channels and a modest upregulation of CaV1.3 and CaV2.3. Additional data indicate that although MDIMP binds to the closed state of the channels, it has more preference for the inactivated one. Concerning CaV3.1, the compound did not alter the shape of the instantaneous current-voltage curve, and substituting one or two residues in the selectivity filter drastically increased the IC50 value, suggesting that MDIMP binds to the extracellular side of the pore. However, an outward current failed in removing the inhibition, which implies an alternative mechanism may be involved. The enantiomer (R)-MDIMP [methyl (R)-2-(1,3-dihydroisoindol-2-yl)-4-methylpentanoate], on the other hand, was synthesized and evaluated, but it did not improve the affinity to LVA channels. Implications of these findings are discussed in terms of the possible underlying mechanisms and pharmacological relevance.

SIGNIFICANCE STATEMENT We have studied the regulation of voltage-gated calcium channels by MDIMP, which disrupts excitation-contraction coupling in cardiac myocytes. The latter effect is more potent in atrial than ventricular myocytes, and this could be explained by our results showing that MDIMP preferentially blocks low-voltage–activated channels. Our data also provide mechanistic insights about the blockade and suggest that MDIMP is a promising member of the family of Ca2+ channel blockers, with possible application to the inhibition of subthreshold membrane depolarizations.

Introduction

Voltage-gated ion channels are proteins found in most cells of an organism and are responsible for regulating rapid ion transport across the plasma membrane (Terlau and Stühmer, 1998). In particular, voltage-gated Ca2+ channels (VGCCs) allow the entry of this second messenger and thereby influence several cellular processes. The VGCCs can be classified as low- and high-voltage–activated (LVA and HVA), as well as of type L, N, P/Q, R, and T. Interestingly, LVA and HVA channels differ in the residues providing electronegativity to the four homologous domains of their selectivity filter. The former contains two glutamates and two aspartates (EEDD), whereas the latter has four glutamates (EEEE; Catterall, 2011). The L-type channels are sensitive to low concentrations (<10 µM) of dihydropyridines (DHPs; Hess et al., 1984), and each cell type possesses its kit of channels. The case of VGCCs of atrial and ventricular myocytes is typical. In adult rats, for example, the former expresses both HVA (CaV1.2, CaV1.3, and CaV2.3) and LVA (CaV3.1 and CaV3.2) channels, whereas the latter primordially expresses CaV1.2 (Piedras-Rentería et al., 1997; Larsen et al., 2002; Perez-Reyes, 2003; Zhang et al., 2005; Zamponi et al., 2015; Ríos-Pérez et al., 2016).

The isoindolines are heterocyclic compounds consisting of a six-membered benzene ring fused to a five-membered ring containing a nitrogen atom at position two (Speck and Magauer, 2013). Molecular docking studies suggest that some of these compounds (isoindolines 2-substituted derivatives from α-amino acids) could bind to an L-type Ca2+ channel (CaV1.2). In particular, both the isoindoline MDIMP (also known as 1b) and a DHP antagonist showed similar binding profiles for CaV1.2. Moreover, the in silico model also predicts that MDIMP binds to segments S5, S6, and loop P (of domain II) by interacting with residue F763 (S6) and the glutamate of the selectivity filter (E736; Mancilla-Percino et al., 2010).

A previous work challenged the possibility that MDIMP regulates VGCCs in cardiac myocytes. Remarkably, the compound reduced the magnitude of both LVA and HVA calcium currents, acting more potently on the former. Additionally, MDIMP inhibited excitation-contraction (EC) coupling in both atrial and ventricular myocytes (Santamaria-Herrera et al., 2016). This demonstrates that the isoindolines effectively target VGCCs and thereby also impair EC coupling.

Here, we have investigated whether MDIMP selectively blocks a specific type of Ca2+ channel by using heterologous expression systems. Indeed, our results show that MDIMP acts preferentially on CaV3.1 and CaV3.2 (LVA channels). Additional findings suggest that the occupancy of the binding site is state-dependent: the inactivated state enhances the affinity, and the site is also partially available in the closed configuration. We also describe experimental conditions that allow MDIMP to partially discriminate between two L-type calcium channels, CaV1.2 and CaV1.3. The effect on CaV3.1 channels was further characterized, and results from two mutations in the selectivity filter of this channel (DEDD and DDDD) suggest that this region might be important for the interaction (the potency of the blockade is reduced by 3- and 5-fold). Interestingly, however, we observed a similar percentage of inhibition on inward and outward currents, indicating that imposing a driving force for the efflux of ions does not unplug the ion conducting pathway. Finally, MDIMP did not alter the shape of the instantaneous I-V curve, suggesting that the effect on a particular channel is all-or-none.

A preliminary report was presented recently (De La Rosa et al., 2019).

Materials and Methods

Cell Culture and Transfection.

Human embryonic kidney (HEK) 293T/17 cells, a subclone of the HEK293 cell line, were acquired from the American Type Cell Culture Collection (ATCC CRL-11268) and were grown and transfected as described previously (Santamaria-Herrera et al., 2016). The following cDNAs (GenBank accession number included) were used: CaV1.2 (AY728090), CaV1.3 (AF370009), CaV2.3 (X67856), β2a (M80545), and α2δ-1 (M21948). The transfection mix consisted of equimolar amounts of cDNA encoding for one principal channel subunit, both accessory subunits, and Enhanced Green Fluorescent Protein, as a reporter gene. In another set of experiments, 293T/17 cells were transfected with cDNAs encoding Enhanced Green Fluorescent Protein and mutant proteins of CaV3.1 (either DEDD or DDDD; Garza-López et al., 2016). FuGENE HD (Promega, Madison, WI) was used as a transfection reagent according to the provider’s instructions. All recordings were performed 48–72 hours post-transfection in green fluorescent cells. The activity of wild-type LVA channels (CaV3.1-AF190860 and CaV3.2-AF051946) was also investigated in HEK293 cells stably transfected with the corresponding cDNAs (Díaz et al., 2005; Lopez-Charcas et al., 2012).

Voltage-Clamp Experiments.

Calcium currents (ICa) were recorded using the whole-cell patch-clamp technique, as described elsewhere (Santamaria-Herrera et al., 2016). Unless otherwise specified, the holding potential (HP) was −90 mV. The whole-cell series resistance and cell capacitance were estimated using the capacitance cancellation feature of the amplifier (Axopatch 200B; Molecular Devices, LLC, San Jose, CA), and other linear components of the current signal were digitally eliminated (with a leak subtraction protocol, termed ‘P/N’). Series resistance was electronically compensated (by 60%–70%), and its final values were typically 1 to 2 MΩ. ICa was analogically filtered (at either 2 or 10 kHz) using the four-pole low-pass Bessel filter of the amplifier, digitalized by an analogical-to-digital converter (Digidata 1322; Molecular Devices, LLC), and stored for off-line analysis. ICa was assessed before, during, and after bathing the cells with MDIMP, its enantiomer ((R)-MDIMP ), or the vehicle (dimethyl sulfoxide [DMSO]).

In general, ICa was elicited by voltage steps delivered every 15 seconds, from the HP to either −20 mV (LVA channels) or +20 mV (HVA channels). For each particular cell, ICa values were normalized to those observed at ∼105 seconds after break-in into the whole-cell mode, and at that time, MDIMP was applied, typically in two consecutive concentrations, followed by washout. In some cells, the behavior of ICa was assessed in the presence of only the control external solution (here termed “basal” ICa). To estimate the percentage of MDIMP effect, the ICa values observed in the presence of the compound were normalized to those of time-matched basal. Unless specified otherwise, the concentration-response curves were built from peak values of current traces and fitted according to the following Hill equation:Embedded Image(1)where ICa represents the percentage of inhibition, [MDIMP] is the MDIMP concentration, IC50 is the [MDIMP] required to 50% of the effect, and nH is the Hill coefficient.

To investigate the steady-state voltage dependence of inactivation, 5-second conditioning pulses to different potentials preceded a test pulse to −20 or +20 mV (LVA and HVA channels, respectively). Subsequently, the relative values of ICa were plotted as a function of membrane potential during prepulses, and the resulting inactivation curves were fitted according to the following Boltzmann equation:Embedded Image(2)where Imax represents the maximum current recruited at very negative potentials, V1/2 is the potential required to reach 50% of inactivation, and k is the slope factor.

Finally, the voltage dependence of activation was assessed by fitting current-voltage curves (I-V curves) to the following Boltzmann equation:Embedded Image(3)where Gmax is the maximal conductance, Vm is the membrane potential, Vrev represents the apparent reversal potential, V1/2 is the membrane potential required to activate 50% of Gmax, and k is the slope factor.

Recording Solutions.

Unless otherwise specified, cells were bathed in an external solution containing (in millimolars) 10 CaCl2, 150 tetraethylammonium chloride, 10 HEPES, and 10 glucose (pH 7.3). Besides, the internal (pipette) solution contained (in millimolars) 140 Cs-Asp, 10 Cs-EGTA, 10 CsCl, 2 CaCl2, 5 Mg-ATP, 0.05 Tris-GTP, 5 glucose, and 10 HEPES (pH 7.3).

Stock solutions (100 mM) of MDIMP and its enantiomer ((R)-MDIMP ) were prepared using DMSO as the solvent and then were diluted in the external recording solution for obtaining final working concentrations. When high molarities of the compounds (i.e., 0.5–1 mM) were used, the external solution was warmed to avoid precipitation (10 minutes, ∼50°C; in a water bath). Indeed, concentrations higher than 1 mM were not investigated because MDIMP tends to precipitate beyond this concentration. New working solutions were prepared daily, and direct exposure to light was avoided (although we do not know whether MDIMP is light-sensitive). The exchange of external solutions was performed by a fast perfusion system (SF-77B; Warner Instruments, LLC, Hamden, CT) that was digitally controlled. For all channel types, the vehicle alone (DMSO) was tested at concentrations in which its mixture with MDIMP elicited a drastic reduction on ICa, and the solvent per se failed in producing significant changes.

Synthesis of (R)-MDIMP .

The strategy for synthesizing the new enantiomer (R)-MDIMP is illustrated as auxiliary material (Supplemental Fig. 1). Briefly, 1) the d-leucine methyl ester hydrochloride was prepared from d-leucine and trimethylsilyl chloride in methanol according to the method described elsewhere (Li and Sha, 2008). 2) Then, the new (R)-MDIMP was obtained from the reaction of d-leucine methyl ester hydrochloride, α,α’-dibromo-o-xylene, and KHCO3 in acetonitrile following the same strategy as for MDIMP (Mancilla-Percino et al., 2001). The structure of (R)-MDIMP was characterized by 1H and 13C NMR and high-resolution mass spectrum (see Supplemental Fig. 1 for further details).

Statistical Analysis.

All data are presented as means ± S.D. and were analyzed and plotted with the combined use of software suites: p-CLAMP (version 10.2; Molecular Devices) and SigmaPlot (version 12.3; Systat Software Inc; San Jose, CA). Statistical significance was evaluated using two-tailed paired Student’s t tests. All experiments were performed at room temperature (22–24°C).

Results

Our main aim was to investigate the effects of MDIMP on cardiac VGCCs expressed in heterologous systems. This experimental approach guarantees the identity of each particular channel and thus contributes to a better understanding of the underlying mechanisms. We began using an HP of −90 mV, which keeps most of the channels in the closed state.

In addition, we initially focused on the L-type Ca2+ channels CaV1.3 (Fig. 1, A–D) and CaV1.2 (Fig. 1, E and F). For each channel type, the basal levels of ICa were estimated under control conditions; that is, the behavior of ICa was studied as a function of time and in the absence of MDIMP. Figure 1A (open circles) provides an example of such basal behavior. It can be seen that the basal ICa of CaV1.3 decreases steadily within the next 10 minutes after having initiated the recordings (to an ∼85% of its initial magnitude). Another pool of cells was consecutively exposed to 750 and 1000 µM of MDIMP, and this led to a higher decrease in the magnitude of ICa compared with the basal behaviour (Fig. 1A, green triangles). MDIMP was then washed out, and the ICa slightly recovered from inhibition (Fig. 1A, open triangles). Results obtained from cells exposed to the compound were then divided by their time-matched basal to only focus on MDIMP effects (which applies to all subsequent data). This analysis revealed that 750 and 1000 µM of MDIMP reduce the magnitude of ICa by ∼20% and ∼40%, respectively (Fig. 1B).

Fig. 1.
Fig. 1.

MDIMP effects on L-type Ca2+ channels. (A) The open circles represent ICa of CaV1.3 that was recorded as a function of time under control or basal conditions (n = 6). Another cell group (triangles, n = 3) was consecutively exposed to 750 and 1000 µM of MDIMP (green symbols), and then the compound was washed out (open triangles). All results obtained from each particular cell were normalized by the absolute ICa value that was observed at minute 1.75 (just before MDIMP application). (B) Same data set shown as triangles in (A) but divided by the corresponding time-matched basal [(A) circles]. (C) Concentration-response curve built from data obtained as in (B). MDIMP (250 μM) increased the peak magnitude of ICa (data point with negative values), and thus the curve was fitted according to a slightly modified version of eq. 1. Specifically, a constant was added to account for the negative deviation (−20.2%). (D) Representative current traces obtained from CaV1.3-expressing cells in the presence of distinct MDIMP concentrations (numbers, in micromolars). The currents were normalized to the peak value of the control (black trace). (E and F) Concentration-response curve (E) and representative traces (F) obtained from CaV1.2-expressing cells (as in A–D). In both cases (CaV1.3 and CaV1.2), MDIMP accelerated the rate of ICa inactivation [(D and F), Supplemental Fig. 3].

With additional data, a concentration-response curve was built (Fig. 1C), and the following values of IC50 and Hill coefficient were obtained: 957 μM and 3.5 (these parameters are also shown in Supplemental Table 1, together with those obtained for all channel types and experimental conditions). Of note, a low MDIMP concentration (250 µM) promoted a slight increase in the activity of the CaV1.3 channel (see representative traces of Fig. 1D). As a result of this effect, the concentration-response curve showed a small but significant deviation toward negative values (of ∼20%, Fig. 1C). Thus, in the presence of MDIMP, the ICa of CaV1.3 likely reflects a balance between degrees of stimulation and inhibition.

Figure 1 also shows a summary of MDIMP effects on CaV1.2 (Fig. 1, E and F). The corresponding IC50 was 656 µM, which is higher than a value previously reported (452 µM; Santamaria-Herrera et al., 2016). This apparent inconsistency is likely due to an essential methodological difference: in the previous study, a depolarized HP of −50 mV was used. Below, we will come back to this point.

Next, we focused on CaV2.3, an HVA R-type Ca2+ channel (Fig. 2). In comparison with results from L-type Ca2+ channels (Fig. 1), MDIMP did not reduce the peak ICa of CaV2.3, even at the highest concentration tested (i.e., 1 mM). Nevertheless, the kinetics of ICa revealed a fascinating aspect: the rate of inactivation was drastically accelerated in a concentration-dependent manner (Fig. 2, inset). Accordingly, the compound exerted a marked inhibition at the end of the test pulse (i.e., 200 milliseconds), with an IC50 of 717 µM (Supplemental Fig. 2).

Fig. 2.
Fig. 2.

Effects on CaV2.3 channels. Average values of peak ICa expressed as a percentage of inhibition by different concentrations of MDIMP. The inset shows examples of current traces carried by CaV2.3, which were used to obtain the concentration-response curve. Rather than being reduced, the peak ICa showed a tendency to be increased in response to MDIMP. Additionally, the compound promoted an acceleration on the kinetics of current decay (see also Supplemental Fig. 3). The results were obtained and analyzed as in Fig. 1.

Indeed, a similar effect on the inactivation rate of L-type Ca2+ channels was also observed [see, for example, the representative traces of Fig. 1 and those of Santamaria-Herrera et al. (2016)]. Moreover, by analyzing the ICa remaining at the end of the test pulse, the IC50 values obtained from the data in Fig. 1 were smaller than those estimated from the peak ICa. Specifically, they range from 957 to 290 µM (CaV1.3) and from 656 to <250 µM (CaV1.2; Supplemental Fig. 2). The kinetics of inactivation were also studied by fitting exponential equations to the decaying phase of the current. In the case of CaV1.3, a second-order exponential was necessary, whereas a first-order exponential sufficed in both CaV1.2 and CaV2.3. In general, MDIMP reduced the value of time constants by ∼50%–60% (except for a fast component of ∼10 milliseconds in CaV1.3, which was not altered; Supplemental Fig. 3). Thus, MDIMP accelerates the rate of inactivation in all HVA Ca2+ channels, in spite of this process arising from distinct levels of complexity among channel subtypes.

Previous work in atrial myocytes showed that MDIMP preferentially inhibits LVA Ca2+ currents (ICaLVA), suggesting a selective effect on CaV3 channels (Santamaria-Herrera et al., 2016). In that work, however, this effect was not pursued further because the amplitude of cardiac ICaLVA is minimal (∼30 pA). Thus, we aimed to investigate whether MDIMP effectively inhibits LVA Ca2+ channels with a higher potency than that observed for HVA channels. Figure 3A shows examples of ICa that were obtained for CaV3.2 in the absence and presence of MDIMP (250 and 500 µM). The time course of average ICa values is also shown in Fig. 3B. It can be observed that MDIMP rapidly downregulates CaV3.2 in approximately 30 seconds. Notably, 250 and 500 µM of MDIMP exert an inhibition of ∼45% and ∼80%, respectively. Thus, after analyzing the entire concentration-response curve, it was possible to conclude that MDIMP inhibits CaV3.2 with an IC50 of ∼220 µM, corroborating the idea that MDIMP preferentially acts on CaV3 channels. This conclusion was further supported by results obtained for CaV3.1, as this channel exhibited the highest sensibility to MDIMP (Fig. 4). For example, the percentages of inhibition were, approximately, 75% and 95% for 250 and 500 µM. Accordingly, the IC50 was ∼130 µM (the lowest among all Ca2+ channel types that we have investigated).

Fig. 3.
Fig. 3.

MDIMP effects on CaV3.2 channels. (A) Representative traces of ICa that were recorded in a CaV3.2-expressing cell in the absence (control), presence, and washout of MDIMP (250 and 500 µM). (B) Time course of average values of ICa [recorded as in (A), n = 3]. (C) Concentration-response curve obtained from experiments as in (A and B). The green and open symbols indicate the presence and absence of MDIMP, respectively.

Fig. 4.
Fig. 4.

MDIMP effects on CaV3.1 channels. (A–C) Recordings of ICa carried through CaV3.1 in the absence and presence of distinct molarities (micromolars) of MDIMP (A) and corresponding time course of inhibition (B) and concentration-response curve (C). Data were obtained as described in Fig. 3. The orange and open symbols indicate the presence and absence of MDIMP, respectively.

It is also worth noting that MDIMP did not promote changes in the kinetic properties of CaV3 channels (Fig. 3A; Fig. 4A, see also Supplemental Fig. 4), which is in striking contrast with results obtained from HVA Ca2+ channels. In the latter, the time course of inactivation was drastically accelerated (as described above and also shown in representative traces of Figs 1,2). Thus, MDIMP promotes “inactivation” but only in those channels exhibiting slow inactivation (i.e., HVA channels).

It is well known that many drugs exert their effects depending on conformational states of the channels, and thus their modulatory actions depend indirectly on Vm. In the next section, therefore, we investigated the voltage dependence of inactivation of Ca2+ channels in the presence and absence of MDIMP. The corresponding data are shown and summarized in Fig. 5 and Table 1. In all channel types, the isoindoline shifted the steady-state inactivation curve toward more negative membrane potentials, suggesting that the inactivated state facilitates MDIMP binding or vice versa. The extent of this effect, however, is somewhat variable. More specifically, the shift in V1/2 values is close to −20 mV for both T- (CaV3.1, CaV3.2) and L-type Ca channels (CaV1.2 and CaV1.3), whereas CaV2.3 exhibits the most considerable shift (of approximately −40 mV; Fig. 5; Table 1). Indeed, the shape of inactivation curves was also significantly altered, but not in all types of channels. In particular, the k value was decreased for CaV1.2 and increased for both CaV3.1 and CaV3.2 (in all cases, the change in k was close to 5 mV; Fig. 5; Table 1).

Fig. 5.
Fig. 5.

The MDIMP shifts the voltage dependence of inactivation toward more negative potentials. (A and B) Representative current traces that were used to investigate the steady-state voltage dependence of inactivation in the absence (Control) and presence of MDIMP. A 5-second prepulse to distinct membrane potentials (Vm) preceded test pulses to either +20 mV [(A) HVA channels] or −20 mV [(B) LVA channels]. (C and D) Average inactivation curves that were obtained from recordings as in (A and B). The circles represent Control conditions, whereas colored symbols indicate the presence of MDIMP. For each data set, the peak ICa was normalized to its maximum value and plotted as a function of prepulse potential, and the resulting inactivation curves were fitted according to a Boltzmann equation (eq. 2). The corresponding parameters are given in Table 1, and their mean values were used to calculate the smooth lines. Results are from a total of four to five investigated cells (per channel type). The MDIMP concentration was (in micromolars) 130 (CaV3.1), 220 (CaV3.2), 500 (CaV1.2 and CaV2.3), and 1000 (CaV1.3). Open circles − Control, upward triangles − CaV1.2, downward triangles − CaV1.3, squares − CaV2.3, diamonds − CaV3.1, hexagonal symbols − CaV3.2.

View this table:
TABLE 1

Parameters of fitted steady-state inactivation and I-V curves

In Fig. 6, the view that MDIMP preferentially binds to the inactivated state was further challenged for HVA channels. In particular, the MDIMP effects were now assessed from an HP of −50 mV (a potential that allows transitions from closed to inactivated states; see Fig. 5). The corresponding IC50 values were (in micromolars) 445 (CaV1.2), 231 (CaV1.3), and 557 (CaV2.3); that is, much smaller than those observed with an HP of −90 mV (compare data of Fig. 6A with those of Figs. 1 and 2). These data reinforce our conclusion that MDIMP preferentially binds to VGCCs in the inactivated state. Besides, similar to the observed with an HP of −90 mV, the IC50 values are also smaller if the ICa is assessed at the end of 200-millisecond pulses (Fig. 6B and inset).

Fig. 6.
Fig. 6.

A depolarized HP increases the sensitivity of HVA Ca2+ channels to MDIMP. (A) Percentage of peak ICa inhibition as a function of MDIMP concentration. ICa was elicited as described in Figs. 1 and 2 but using an HP of −50 mV. (B) Same data as in (A), except that ICa was measured 200 milliseconds after test pulse onset. (C) Examples of current traces that were recorded in presence of distinct MDIMP concentrations (numbers, in micromolars) and then used to analyze the degrees of inhibition showed in (A and B). The inset illustrates IC50 values that were obtained using either the peak magnitude of ICa or the current remaining at 200 milliseconds. The purple, green, and yellow squares indicate data obtained for the CaV1.2, CaV1.3, and CaV2.3 channels, respectively.

Data showed in Fig. 6 suggest that an HP of −50 mV may help MDIMP to discriminate between CaV1.2 and CaV1.3 channels; the IC50 of the former nearly doubles that of the latter (Fig. 6). Thus, we decided to re-examine the effects of 250 µM MDIMP on these two types of calcium channels over an extended time scale (∼10 minutes, as opposed to the typical 90-second applications). The data obtained from this experimental series are summarized in Fig. 7. In keeping with the results of Fig. 6A, 90 seconds of MDIMP application caused a notable inhibition of CaV1.3 (by approximately 60%), whereas CaV1.2 was barely affected (∼7%). Moreover, although in both channel types the extent of inhibition continued growing, CaV1.3 showed the higher rate (τ = 1.55 minutes, vs. τ = 4.94 minutes), and by 10 minutes, the fraction of inhibition was 0.93 (CaV1.3) and 0.43 (CaV1.2). Thus, 250 µM MDIMP indeed helps to distinguish between CaV1.2 and CaV1.3.

Fig. 7.
Fig. 7.

MDIMP discriminates between CaV1.2 and CaV1.3. (A) Normalized ICa values obtained from CaV1.2- and CaV1.3-expressing cells in either absence (open symbols) or presence of 250 µM MDIMP (colored symbols). Test pulses (to +20 mV) were applied from an HP of −50 mV, and data represent the means ± S.D. from three cells per experimental condition. (B) Examples of ICa traces that were analyzed to generate the results shown in (A). The numbers indicate the time in presence of MDIMP (minutes). Blue symbols/traces − CaV1.2, green symbols/traces − CaV1.3

We next wondered whether MDIMP can also bind to the closed state of VGCCs, and this point was investigated for CaV2.3 (Fig. 8) and the two types of LVA channels (i.e., CaV3.1 and CaV3.2; Supplemental Fig. 5). The experiment consisted in keeping the channels closed; that is, at an HP of −90 mV but exposed to MDIMP for a couple of minutes. Then, the percentage of effect on ICa was assessed at the first depolarizing pulse. In the case of CaV2.3, the action of the isoindoline was somewhat complex. A small increase in peak ICa (of 13%) preceded a slight reduction in the magnitude of current remaining at 200 milliseconds (of 21%), and a striking change in current kinetics reflected these two effects (Fig. 8). Accordingly, the compound promoted a 2-fold increase in the fraction of inactivation (i.e., from 0.25 ± 0.09 to 0.50 ± 0.10; P = 0.005). These data suggest that the closed state of the CaV2.3 HVA channel binds MDIMP, and this results in marked alterations in gating. Moreover, MDIMP also acts on the closed state of CaV3.1 and CaV3.2 channels, as evidenced by results from a similar protocol as in Fig. 8, showing a significant reduction in the peak ICa (of 39% and 26%, respectively; Supplemental Fig. 5). Thus, we conclude that the MDIMP binds to VGCCs not only in the inactivated (Fig. 5) state but also in the closed one.

Fig. 8.
Fig. 8.

MDIMP binds to the closed state of CaV2.3. (A) Relative magnitude of ICa that was estimated both at its maximum level (peak, squares) and the end (200 milliseconds, triangles) of test pulses before (1.75 minutes) and 3.25 minutes after MDIMP exposure. During the MDIMP application, the membrane potential was kept at −90 mV (i.e., in absence of depolarizing test pulses). Then, a train of pulses was delivered (15-second cycle length), followed by the washout of MDIMP (ICa measurements are shown only for the first pulse of the train, minute 5). aP = 0.011, bP = 0.034; compared with the corresponding control (1.75 minutes, n = 5). (B) Representative current traces that were used to generate data of (A).

We next focused on activation curves. In these experiments, the HVA Ca2+ channels were examined from an HP of −50 mV to facilitate inhibition. As can be seen in Fig. 9, MDIMP promoted significant reductions on the maximal conductance (Gmax) of all channel types. In contrast, the slope factor (k) and midpoint (V1/2) were in general unaltered: the only exception was for CaV1.3 and CaV2.3, whose V1/2 values were shifted by nearly −4 mV (Fig. 9; Table 1).

Fig. 9.
Fig. 9.

MDIMP does not promote major changes in the voltage dependence of Ca2+ channel activation. (A) I-V curves obtained in the absence (gray symbols) and presence of MDIMP (closed symbols). The results were normalized to the maximum ICa value observed in each particular cell in the absence of MDIMP. Subsequently, the I-V curves were fitted according to a Boltzmann equation (eq. 3). The resulting parameters and number of experiments are given in Table 1. (B) Activation curves that were estimated from the ICa values shown in (A) and their corresponding Boltzmann parameters. The values of Ca2+ conductance (GCa) were normalized to Gmax. (C) Examples of ICa families that were used to generate the data of (A and B). The concentration of MDIMP was (in micromolars) 500 (CaV1.2 and CaV2.3), 230 (CaV1.3), 220 (CaV3.2), and 130 (CaV3.1).

Thus far, CaV3.1 is the channel with the highest affinity for MDIMP. This finding, combined with a previous suggestion that MDIMP binds to the selectivity filter of VGCCs (Mancilla-Percino et al., 2010), motivated us to investigate whether substituting amino acids of the selectivity filter of CaV3.1 modifies the affinity for MDIMP. In particular, we studied the following point mutations: DEDD and DDDD (Garza-López et al., 2016). The corresponding results are shown in Fig. 10A (representative traces) and Fig. 10B (concentration-response curves). It can be observed that MDIMP inhibited the activity of both mutant proteins, but the inhibition occurred at concentrations higher than those observed for the wild-type channel. For example, the mutant proteins were barely affected by 100 and 250 µM (Fig. 10), whereas these concentrations inhibited the wild-type channel by approximately 40% and 80% (Fig. 10, dashed line; see also Fig. 4). These results are in line with the view that the selectivity filter of VGCCs represents an important domain for MDIMP binding.

Fig. 10.
Fig. 10.

Mutations in the selectivity filter of CaV3.1 interfere with the affinity for MDIMP. (A) Calcium currents carried through CaV3.1 mutant channels (DEDD and DDDD) in the presence of distinct MDIMP concentrations (micromolars). (B) Concentration-response curves that were obtained from recordings as in (A); the estimated IC50 values are 387 µM (DEDD) and 543 µM (DDDD). A dashed line representing data gathered for the wild-type CaV3.1 is shown for comparison (see Fig. 4). The orange and yellow squares indicate data obtained for the DEDD and DDDD mutations, respectively.

To obtain more insight about a possible pore-blocking mechanism, we decided to determine whether the ions flowing in the outward direction may unplug the CaV3.1 channel (possibly by removing MDIMP from its putative binding site in the selectivity filter). However, under our standard ICa recording conditions, depolarizations to 80 and 100 mV were unable to elicit a possible efflux of Cs+ through CaV3.1. Therefore, we decided to study an outward Na+ current instead, using a protocol described elsewhere (Fig. 11; Lopin et al., 2012). The corresponding results indicate that this current is reduced to a ∼50% in the presence of 130 µM MDIMP (Fig. 11, A and C). This effect is in agreement with the data of Figs. 4 and 9, in which the IC50 for ICa reduction was about 130 µM.

Fig. 11.
Fig. 11.

An outward current cannot release the inhibition by MDIMP. (A) Experimental protocol and representative traces of ionic current (Im) used to compare MDIMP (130 µM) effects on inward and outward currents. (B) Average values of Im that were obtained from tail currents as those shown in (A) as a function of Vm (instantaneous I-V curves). The inset shows the same I-V curves but after dividing every data set by its corresponding value observed at −120 mV. The decaying phase of tail currents was fitted to a single exponential equation, and Im represents an extrapolation to the time of repolarization. No significant differences were found among experimental groups regarding time constants for current decay (i.e., closing kinetics; data not shown). (C) Percentages of Im inhibition by MDIMP that were estimated at the indicated membrane potential. The results were obtained from five CaV3.1-expressing cells, using the following recording solutions (in millimolars): 2 CaCl2, 128 NaCl, 5 ascorbic acid, 10 glucose, and 20 HEPES (external); and 2 CaCl2, 1 MgCl2, 120 NaCl, 10 HEPES, 4 MgATP, and 11 EGTA. In both cases, the pH was adjusted to 7.3 with NaOH. The absence and presence of MDIMP is indicated by closed and green symbols, respectively.

Moreover, MDIMP reduced the magnitude of both outward and inward (tail) currents to a similar extent in the same pool of cells (Fig. 11C). The similarity of effects on inward and outward currents suggests that a driving force for the efflux of ions fails in releasing the MDIMP-CaV3.1 interaction. Unfortunately, this result does not support the pore-blocking hypothesis.

Figure 11 also shows instantaneous I-V curves that were obtained by analyzing the initial amplitude of the tail current (Fig. 11, A and B). It can be observed that MDIMP did not alter the shape of the normalized curves (Fig. 11B, inset), indicating that the effect on a single channel is all-or-none (according to the terminology of Armstrong, 1971).

Discussion

In this work, we studied the regulation of cardiac VGCCs by MDIMP, a novel inhibitor of both calcium channels and EC coupling, with atrial selectivity. Indeed, a previous study showed that MDIMP preferentially acts on LVA calcium currents of atrial myocytes, suggesting that CaV3 channels exhibit the highest affinity for this compound (Santamaria-Herrera et al., 2016). In these cells, the LVA calcium current is thought to reflect the activity not only of CaV3 channels (i.e., CaV3.1 and CaV3.2) but also CaV1.3 and CaV2.3 (Piedras-Rentería et al., 1997; Larsen et al., 2002; Zhang et al., 2005; Ríos-Pérez et al., 2016). Thus, the present study represents the first unequivocal demonstration that MDIMP preferentially binds to CaV3 channels. In addition, our data suggest that the compound regulates these channels by acting in all-or-none fashion (for the instantaneous I-V curves have the same shape in the absence and presence of MDIMP).

Moreover, apart from describing the molecular and biophysical properties of the regulation, our study investigated the hypothesis that the selectivity filter represents the high-affinity receptor site for MDIMP. Indeed, evidence in favor of this view was obtained as the mutations of the selectivity filter decreased the affinity for the compound. Nevertheless, the pore-blocking model could not be supported by analyzing effects on outward currents. Thus, further investigation is needed to more firmly establish the precise mechanism.

Below, we discuss tentative mechanisms for the inhibition as well as the possible pharmacological and clinical relevance of MDIMP. From now on, the term “blockade” is used to indicate that MDIMP might bind within the selectivity filter of the channels. On the other hand, “modulation” (i.e., inhibition or stimulation), is reserved to imply that the MDIMP can bind to either the pore or an alternative site, resulting in allosteric consequences (i.e., alterations in gating).

Mechanisms of MDIMP Effects on VGCCs.

The functional properties of a VGCC emerge from the movement of many segments in at least one of the proteins that form the channel (i.e., main subunit). Thus, when a compound binds to an ion channel, it seems reasonable to expect that the function of the latter might be modified. The influence of the pore region on functional properties is not exempt from this principle. On the contrary, there are examples in which pharmacological or molecular modifications of the pore result in significant functional effects. The pore blockade of Na+ channels by Ca2+, which also shifts the open probability toward more positive potentials, is an emblematic example (Armstrong and Cota, 1991). Besides, it was recently shown that substituting a single amino acid, located next to the glutamate that belongs to the selectivity filter of human CaV1.2 (located in domain II), eliminates Ca2+-dependent inactivation, a process that begins intracellularly with allosteric modulation (by Ca2+ and calmodulin) of the COOH-terminal domain (Abderemane-Ali et al., 2019). Moreover, it has been reported that the point mutations of the selectivity filter significantly alter the gating of CaV3.1 (Talavera et al., 2003). Thus, although the pore-blocking mechanism would seem to represent the most straightforward explanation to the finding that the potency of MDIMP varies with the selectivity filter configuration, it is also possible that, instead, the latter exerts a remote control on another binding site.

Currently, it is difficult to explain the apparent paradox that, under certain circumstances, a Ca2+ channel antagonist also elicits a significant increase in the activity of the channels. Something roughly similar has been reported for DHPs (e.g., Bean, 1985), and the explanation relies on using racemic mixtures and the well known agonist and antagonist actions of DHP stereoisomers. In our study, however, this explanation can be ruled out because the MDIMP was obtained from pure l-leucine. Indeed, we have also synthesized the D enantiomer ((R)-MDIMP, Supplemental Fig. 1) to assess its potential to inhibit CaV3 channels. The results show that CaV3.2 exhibits roughly the same affinity for both MDIMP and (R)-MDIMP (Supplemental Fig. 6). CaV3.1, on the other hand, showed a slightly lower sensitivity to (R)-MDIMP (i.e., the IC50 was ∼260 µM, which is nearly 2-fold higher than that of MDIMP; Supplemental Fig. 6). Accordingly, results from docking experiments indicate that the (R)-MDIMP has a lower affinity for the pore region of VGCCs compared with MDIMP (data not shown).

Shifting the steady-state inactivation curve toward more negative potentials was a common feature of MDIMP effects. This can be interpreted to suggest that in all channels, the inactivated state facilitates binding of MDIMP, either to the pore or to an alternative site, and thereby the inactivation also promotes the corresponding blockade or inhibition. Preferential binding of drugs to the inactivated state is frequently observed. The example of DHPs acting on CaV1 (L-type) channels is emblematic. For example, when the HP changes from −80 to −15 mV (i.e., from privileging the resting state to promoting inactivation), then the IC50 for the effect of nitrendipine dramatically decreases (up to ∼2000-fold, from 730 to 0.36 nM; Bean, 1985). Another similitude between DHP antagonists and MDIMP is their ability to accelerate the inactivation rate of L-type Ca2+ channels (e.g., Bean, 1985; this study). Interestingly, it has been proposed that the DHPs bind to the lateral, lipid-facing side of the pore module, and this promotes an allosteric modification (asymmetry) of ion binding sites, which in turn results in high-affinity Ca2+ binding and blockade of the selectivity filter (Tang et al., 2016). Thus, it will be important for future studies to determine whether the MDIMP and DHPs share a common binding site in VGCCs (as suggested by a theoretical model; Mancilla-Percino et al., 2010).

In contrast with the observation that in all types of VGCCs, the inactivated state facilitates the interaction with the isoindoline, some of our findings were somewhat channel-type specific. For example, the apparent inactivation rate was accelerated by MDIMP, but only in HVA channels. In theory, this effect could be explained by the following two possible mechanisms:

  1. MDIMP binds to an alternative site, which results in a configuration that is primed to accelerate inactivation. This mechanism implies that MDIMP somehow modulates the particles (or segments) that provoke inactivation. According to our results, this site is partially available in the closed configuration (Fig. 8; Supplemental Fig. 5), and the inactivated state promotes its occupancy (Fig. 5), thereby creating a positive feedback loop.

  2. The open state facilitates the blockade. This possibility arises from assuming that the movement of activation gates to the open position liberates access to a binding site located in the selectivity filter. A precedent exists for this mechanism regarding LVA channels and antiepileptic drugs (succinimides). More precisely, it has been reported that these drugs not only show a high affinity for the inactivated state but also accelerate the apparent inactivation rate of ICaLVA. To investigate whether a possible pore-blocking mechanism explains these findings, Gomora et al. (2001) compared the percentage of effect on inward and outward currents and found that these compounds significantly alter the influx of Ca2+ but have little impact on outward currents (carried through CaV3.3). This unidirectional action was interpreted to suggest that succinimides indeed clog the pore. Moreover, it also indicates that the ions flowing in the outward direction unplug the channel (Gomora et al., 2001).

In the present study, we did not observe an unplug of CaV3.1 channels by the outward current, despite the view that MDIMP might bind to the selectivity filter. Thus, the identification of the actual binding site(s) warrants further investigation.

Currently, it is hard to understand why an outward current failed to release the supposed interaction of MDIMP with an outer pore domain. This apparent inconsistency could be explained by proposing that the interaction promotes a pore collapse. Under these circumstances, the intracellular ions would be trapped; that is, unable to pass through a collapsed pore, which represents a physical barrier for reaching—and expelling—MDIMP. Apart from serving as a conciliatory mechanism, this idea of “pore collapse” is consistent with our results suggesting that MDIMP acts in an all-or-none fashion (Fig. 11B).

Potential Pharmacological and Clinical Significance of MDIMP.

The relevance of CaV3 channels for the cardiac function is related to pacemaker depolarization, and thereby their blockade is thought to be cardioprotective in the context of atrial fibrillation. Besides, they can be re-expressed under certain pathological conditions as part of the neonatal program of gene expression. More generally, the CaV3 channels participate in regulating the function of a number of cell types, including not only excitable but also nonexcitable cells (Perez-Reyes, 2003; Zamponi et al., 2015). Thus, a selective blockade of CaV3 channels by MDIMP could be of therapeutic benefit. The finding that this compound preferentially inhibits EC coupling in atrial versus ventricular myocytes, combined with the fact that CaV3 channels are only expressed in the former cells, indirectly supports this interpretation.

On the other hand, the affinity of CaV3 channels for MDIMP is not higher than that found for other compounds. Indeed, certain compounds exert their effects with IC50 values in the low micromolar range (Perez-Reyes, 2003; Hashimoto and Kawazu, 2015). However, apart from selectivity, the tolerability is also essential for a compound to be clinically relevant. Therefore, it is crucial to keep in mind that an amino acid was included in MDIMP, with the idea of making a more tolerable compound on the grounds that adducts of amino acids are highly stable (Flückiger et al., 1984; Mancilla-Percino et al., 2010; Santamaria-Herrera et al., 2016). Indeed, they give rise to less variety of species in solution (e.g., Hofer et al., 2013). Thus, future research may lead to concluding that MDIMP is advantageous for in vivo conditions, possibly based on less adverse effects.

The calcium channels of sperm cells present the same pore motif as that of the CaV3.1 DDDD mutant protein (Navarro et al., 2008). This similitude, combined with the finding that the DDDD mutation decreases the affinity of CaV3.1 for MDIMP, strongly suggests that calcium channels of sperm cells should also exhibit relatively low MDIMP affinity. Likewise, other HVA channels not investigated here (but expressing four glutamates, EEEE) are also expected to be MDIMP-resistant. CaV3.3, on the other hand, is expected to share the same high affinity as CaV3.1 and CaV3.2. Further work is needed to test all these predictions.

In addition to VGCCs, other molecular targets likely mediate the biological activities of MDIMP. For example, it has been shown that the isoindolines derived from α-amino acids exert antitumoral effects and also act as enzyme inhibitors (of COX-1, COX-2, and histone deacetylase 8). The corresponding IC50 values for MDIMP were, approximately (in millimolars), 3.2 (cell viability), 0.7 (COX-1), 1.4, (COX-2), and 0.2 (histone deacetylase 8) (Trejo-Muñoz et al., 2013, 2014; Mancilla-Percino et al., 2016). On the other hand, we found that depolarized membrane potentials enhance the affinity of MDIMP for VGCCs. Indeed, at −50 mV 130 µM of the compound inhibits CaV3.1 by ∼90% ( instead of the 50% that is observed at an HP of −90 mV). This indicates the presence of a use-dependent block that could be relevant in electrically active cells. Therefore, we speculate that blood levels as low as 30–40 µM (which, in theory, do not compromise cell viability) should be sufficient for reverting an exacerbated excitability in vivo. It is important to emphasize, however, that the therapeutic window of MDIMP has yet to be experimentally determined.

Independently of its potential to be used clinically, MDIMP may also be helpful to dissect the activity of non–T-type calcium channels in native expression systems, specifically in conditions in which HVA channels exhibit characteristics typically associated with LVA channels, for example, in rat atrial myocytes in which the LVA calcium current is thought to reflect the activity of not only CaV3.1 and CaV3.2 but also CaV1.3 and CaV2.3 (Piedras-Rentería et al., 1997; Larsen et al., 2002; Zhang et al., 2005; Ríos-Pérez et al., 2016).

To date, a pharmacological tool that selectively targets CaV1.3 does not exist (Zamponi et al., 2015). A pyrimidine-2,4,6-trione derivate (Cp8) was thought to represent a selective blocker because it seems to discriminate between CaV1.2 and CaV1.3 (Kang et al., 2012). However, this observation was unconfirmed in more recent studies (Huang et al., 2014; Ortner et al., 2014). Moreover, one study found that Cp8 instead stimulates the activity of both channel types (Ortner et al., 2014). Our results suggesting that MDIMP could help to distinguish the activity of L-type Ca2+ channel subtypes in native cells (Fig. 7) are relevant in this context. Nevertheless, we believe that additional efforts are needed to find the most selective compound. MDIMP forms part of an extensive list of molecules termed “amino acid–derived isoindolines” (Mancilla-Percino et al., 2001), whose biological activities remain practically unexplored. Thus, our ongoing work is focused on characterizing a possible regulation of VGCCs by other members of this family.

In conclusion, we have investigated the mechanisms and specificity of MDIMP effects on VGCCs. Our results indicate that this isoindoline preferentially binds to CaV3 channels, possibly in the selectivity filter, which in turn may elicit a collapse of the pore. In addition, our results suggest that MDIMP is useful to discriminate between CaV1.2 and CaV1.3 (L-type) channels. The corresponding pharmacological relevance for the in vivo condition and the possibility that other amino acid–derived isoindolines also regulate VGCCs have yet to be investigated.

Acknowledgments

We thank Robert T. Dirksen and Ulises Meza for proving cDNA clones used in this study. We also thank Salvador Carrasco for technical assistance.

Authorship Contributions

Participated in research design: De La Rosa, Ávila.

Conducted experiments: De La Rosa, García-Castañeda, Ávila.

Contributed new reagents or analytic tools: Nishigaki, Gómora, Mancilla-Percino.

Performed data analysis: De La Rosa, Ávila.

Wrote or contributed to the writing of the manuscript: De La Rosa, Ávila.

Note Added in Proof—The steroisomer nomenclature for the enantiomer of MDIMP was not included in the Fast Forward version published June 25, 2020. The manuscript and supplemental material have now been corrected.

Footnotes

    • Received March 10, 2020.
    • Accepted June 11, 2020.

  • This work was supported by Consejo Nacional de Ciencia y Tecnologia [Grant 253009] to G.Á.

  • Part of this work was presented at the Biophysical Society Annual Meeting. March 02–06, 2019; Baltimore, MD.

  • https://doi.org/10.1124/mol.120.119982.

  • ↵Embedded ImageThis article has supplemental material available at molpharm.aspetjournals.org.

Abbreviations

CaV
voltage-gated calcium channel
COX
cyclooxygenase
DHP
dihydropyridine
EC
excitation-contraction
Gmax
maximal conductance
HEK
human embryonic kidney
HP
holding potential
HVA
high-voltage–activated
ICa
calcium current
ICaLVA
low-voltage–activated Ca2+ current
I-V curve
current-voltage curve
k
slope factor
LVA
low-voltage–activated
MDIMP
methyl (S)-2-(1,3-dihydroisoindol-2-yl)-4-methylpentanoate
(R)-MDIMP
methyl (R)-2-(1,3-dihydroisoindol-2-yl)-4-methylpentanoate
VGCC
voltage-gated calcium channel
Vm
membrane potential

References

  1. ↵
    1. Abderemane-Ali F,
    2. Findeisen F,
    3. Rossen ND, and
    4. Minor DL Jr..
    (2019) A selectivity filter gate controls voltage-gated calcium channel calcium-dependent inactivation. Neuron 101:1134–1149.e3.
    OpenUrl
  2. ↵
    1. Armstrong CM
    (1971) Interaction of tetraethylammonium ion derivatives with the potassium channels of giant axons. J Gen Physiol 58:413–437.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    1. Armstrong CM and
    2. Cota G
    (1991) Calcium ion as a cofactor in Na channel gating. Proc Natl Acad Sci USA 88:6528–6531.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    1. Bean BP
    (1985) Two kinds of calcium channels in canine atrial cells. Differences in kinetics, selectivity, and pharmacology. J Gen Physiol 86:1–30.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Catterall WA
    (2011) Voltage-gated calcium channels. Cold Spring Harb Perspect Biol 3:a003947.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. De La Rosa JAM,
    2. García-Castañeda M,
    3. Nishigaki T,
    4. Gomora JC,
    5. Mancilla-Percino T, and
    6. Avila G
    (2019) Pore-blocking effect of isoindoline MDIMP on voltage-gated calcium channels. Biophys J 116:113a–114a.
    OpenUrl
  7. ↵
    1. Díaz D,
    2. Bartolo R,
    3. Delgadillo DM,
    4. Higueldo F, and
    5. Gomora JC
    (2005) Contrasting effects of Cd2+ and Co2+ on the blocking/unblocking of human Cav3 channels. J Membr Biol 207:91–105.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Flückiger R,
    2. Henson E,
    3. Hess GM, and
    4. Gallop PM
    (1984) Mass spectral and HPLC analysis of biological compounds with diphenylborinic acid. Biomed Mass Spectrom 11:611–615.
    OpenUrlPubMed
  9. ↵
    1. Garza-López E,
    2. Chávez JC,
    3. Santana-Calvo C,
    4. López-González I, and
    5. Nishigaki T
    (2016) Cd(2+) sensitivity and permeability of a low voltage-activated Ca(2+) channel with CatSper-like selectivity filter. Cell Calcium 60:41–50.
    OpenUrl
  10. ↵
    1. Gomora JC,
    2. Daud AN,
    3. Weiergräber M, and
    4. Perez-Reyes E
    (2001) Block of cloned human T-type calcium channels by succinimide antiepileptic drugs. Mol Pharmacol 60:1121–1132.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Schaffer SW and
    2. Li M
    1. Hashimoto N and
    2. Kawazu T
    (2015) Pharmacological rofiles of T-type calcium channel antagonists, in T-type Calcium Channels in Basic and Clinical Science (Schaffer SW and Li M 37–48, Springer, Vienna, Austria.
  12. ↵
    1. Hess P,
    2. Lansman JB, and
    3. Tsien RW
    (1984) Different modes of Ca channel gating behaviour favoured by dihydropyridine Ca agonists and antagonists. Nature 311:538–544.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Hofer A,
    2. Kovacs G,
    3. Zappatini A,
    4. Leuenberger M,
    5. Hediger MA, and
    6. Lochner M
    (2013) Design, synthesis and pharmacological characterization of analogs of 2-aminoethyl diphenylborinate (2-APB), a known store-operated calcium channel blocker, for inhibition of TRPV6-mediated calcium transport. Bioorg Med Chem 21:3202–3213.
    OpenUrlCrossRef
  14. ↵
    1. Huang H,
    2. Ng CY,
    3. Yu D,
    4. Zhai J,
    5. Lam Y, and
    6. Soong TW
    (2014) Modest CaV1.342-selective inhibition by compound 8 is β-subunit dependent. Nat Commun 5:4481.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Kang S,
    2. Cooper G,
    3. Dunne SF,
    4. Dusel B,
    5. Luan C-H,
    6. Surmeier DJ, and
    7. Silverman RB
    (2012) CaV1.3-selective L-type calcium channel antagonists as potential new therapeutics for Parkinson’s disease. Nat Commun 3:1146.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Larsen JK,
    2. Mitchell JW, and
    3. Best PM
    (2002) Quantitative analysis of the expression and distribution of calcium channel alpha 1 subunit mRNA in the atria and ventricles of the rat heart. J Mol Cell Cardiol 34:519–532.
    OpenUrlCrossRefPubMed
  17. ↵
    1. Li J and
    2. Sha Y
    (2008) A convenient synthesis of amino acid methyl esters. Molecules 13:1111–1119.
    OpenUrlCrossRefPubMed
  18. ↵
    1. Lopez-Charcas O,
    2. Rivera M, and
    3. Gomora JC
    (2012) Block of human CaV3 channels by the diuretic amiloride. Mol Pharmacol 82:658–667.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Lopin KV,
    2. Gray IP,
    3. Obejero-Paz CA,
    4. Thévenod F, and
    5. Jones SW
    (2012) Fe2+ block and permeation of CaV3.1 (α1G) T-type calcium channels: candidate mechanism for non-transferrin-mediated Fe2+ influx. Mol Pharmacol 82:1194–1204.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Mancilla-Percino T,
    2. Carrillo L,
    3. Zamudio-Rivera LS,
    4. Beltrán HI, and
    5. Farfán N
    (2001) Synthesis and characterization of new 2-substituted isoindoline derivatives of α-amino acids. Org Prep Proced Int 33:341–349.
    OpenUrl
  21. ↵
    1. Mancilla-Percino T,
    2. Correa-Basurto J,
    3. Trujillo-Ferrara J,
    4. Ramos-Morales FR,
    5. Acosta Hernández ME,
    6. Cruz-Sánchez JS, and
    7. Saavedra-Vélez M
    (2010) Molecular modeling study of isoindolines as L-type Ca(2+) channel blockers by docking calculations. J Mol Model 16:1377–1382.
    OpenUrl
  22. ↵
    1. Mancilla-Percino T,
    2. Trejo-Muñoz CR,
    3. Díaz-Gandarilla JA,
    4. Talamás-Rohana P,
    5. Guzmán Ramírez JE,
    6. Cervantes J, and
    7. Figueroa Ortíz A
    (2016) Isoindoline derivatives of α-amino acids as cyclooxygenase 1 and 2 inhibitors. Arch Pharm (Weinheim) 349:175–185.
    OpenUrl
  23. ↵
    1. Navarro B,
    2. Kirichok Y,
    3. Chung J-J, and
    4. Clapham DE
    (2008) Ion channels that control fertility in mammalian spermatozoa. Int J Dev Biol 52:607–613.
    OpenUrlCrossRefPubMed
  24. ↵
    1. Ortner NJ,
    2. Bock G,
    3. Vandael DHF,
    4. Mauersberger R,
    5. Draheim HJ,
    6. Gust R,
    7. Carbone E,
    8. Tuluc P, and
    9. Striessnig J
    (2014) Pyrimidine-2,4,6-triones are a new class of voltage-gated L-type Ca2+ channel activators. Nat Commun 5:3897.
    OpenUrlCrossRefPubMed
  25. ↵
    1. Perez-Reyes E
    (2003) Molecular physiology of low-voltage-activated t-type calcium channels. Physiol Rev 83:117–161.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Piedras-Rentería ES,
    2. Chen CC, and
    3. Best PM
    (1997) Antisense oligonucleotides against rat brain alpha1E DNA and its atrial homologue decrease T-type calcium current in atrial myocytes. Proc Natl Acad Sci USA 94:14936–14941.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Ríos-Pérez EB,
    2. García-Castañeda M,
    3. Monsalvo-Villegas A, and
    4. Avila G
    (2016) Chronic atrial ionic remodeling by aldosterone: potentiation of L-type Ca2+ channels and its arrhythmogenic significance. Pflugers Arch 468:1823–1835.
    OpenUrl
  28. ↵
    1. Santamaria-Herrera MA,
    2. Ríos-Pérez EB,
    3. de la Rosa JAM,
    4. García-Castañeda M,
    5. Osornio-Garduño DS,
    6. Ramos-Mondragón R,
    7. Mancilla-Percino T, and
    8. Avila G
    (2016) MDIMP, a novel cardiac Ca(2+) channel blocker with atrial selectivity. Eur J Pharmacol 781:218–228.
    OpenUrl
  29. ↵
    1. Speck K and
    2. Magauer T
    (2013) The chemistry of isoindole natural products. Beilstein J Org Chem 9:2048–2078.
    OpenUrl
  30. ↵
    1. Talavera K,
    2. Janssens A,
    3. Klugbauer N,
    4. Droogmans G, and
    5. Nilius B
    (2003) Pore structure influences gating properties of the T-type Ca2+ channel alpha1G. J Gen Physiol 121:529–540.
    OpenUrlAbstract/FREE Full Text
  31. ↵
    1. Tang L,
    2. Gamal El-Din TM,
    3. Swanson TM,
    4. Pryde DC,
    5. Scheuer T,
    6. Zheng N, and
    7. Catterall WA
    (2016) Structural basis for inhibition of a voltage-gated Ca2+ channel by Ca2+ antagonist drugs. Nature 537:117–121.
    OpenUrlCrossRef
  32. ↵
    1. Terlau H and
    2. Stühmer W
    (1998) Structure and function of voltage-gated ion channels. Naturwissenschaften 85:437–444.
    OpenUrlCrossRefPubMed
  33. ↵
    1. Trejo-Muñoz CR,
    2. Mancilla-Percino T,
    3. Mera-Jiménez E,
    4. Correa-Basurto J, and
    5. Trujillo-Ferrara JG
    (2013) Partition coefficient determination of a series of isoindolines-2-substituted and its correlation with their antiproliferative activity on HeLa cells. Med Chem Res 22:4039–4045.
    OpenUrl
  34. ↵
    1. Trejo-Muñoz CR,
    2. Mera-Jiménez E,
    3. Pinto-Almazán R,
    4. Díaz-Gandarilla JA,
    5. Correa-Basurto J,
    6. Trujillo-Ferrara JG,
    7. Guerra-Araiza C,
    8. Talamás-Rohana P, and
    9. Mancilla-Percino T
    (2014) Effect of two series of isoindolines over HDAC8 activity and expression. Med Chem Res 23:3227–3234.
    OpenUrl
  35. ↵
    1. Zamponi GW,
    2. Striessnig J,
    3. Koschak A, and
    4. Dolphin AC
    (2015) The physiology, pathology, and pharmacology of voltage-gated calcium channels and their future therapeutic potential. Pharmacol Rev 67:821–870.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Zhang Z,
    2. He Y,
    3. Tuteja D,
    4. Xu D,
    5. Timofeyev V,
    6. Zhang Q,
    7. Glatter KA,
    8. Xu Y,
    9. Shin H-S,
    10. Low R, et al.
    (2005) Functional roles of Cav1.3(alpha1D) calcium channels in atria: insights gained from gene-targeted null mutant mice. Circulation 112:1936–1944.
    OpenUrlAbstract/FREE Full Text
View Abstract
Back to top

In this issue

Download PDF
Article Alerts
Email Article
Citation Tools

Research ArticleArticle

MDIMP, a Novel Calcium Channel Antagonist

Juan A.M. De La Rosa, Maricela García-Castañeda, Takuya Nishigaki, Juan Carlos Gómora, Teresa Mancilla-Percino and Guillermo Ávila
Molecular Pharmacology September 1, 2020, 98 (3) 211-221; DOI: https://doi.org/10.1124/mol.120.119982
Reddit logo Twitter logo Facebook logo Mendeley logo

Jump to section

Advertisement