Modulation of Proton-Gated Channels by Antidepressants - ACS

Nov 26, 2018 - The chemical structures of some antidepressants are similar to those of recently described amine-containing ligands of acid-sensing ion...
0 downloads 0 Views 1MB Size
Subscriber access provided by Stockholm University Library

Article

Modulation of proton-gated channels by antidepressants Maxim Nikolaev, Margarita Komarova, Tatiana Tikhonova, Anastasia Korosteleva, Natalia Potapjeva, and Denis B Tikhonov ACS Chem. Neurosci., Just Accepted Manuscript • DOI: 10.1021/acschemneuro.8b00560 • Publication Date (Web): 26 Nov 2018 Downloaded from http://pubs.acs.org on December 3, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

Modulation of proton-gated channels by antidepressants Maxim V. Nikolaev*, Margarita S. Komarova, Tatiana B. Tikhonova, Anastasia S. Korosteleva, Natalia N. Potapjeva and Denis B. Tikhonov

I.M.Sechenov Institute of Evolutionary Physiology and Biochemistry Russian Academy of Sciences, St-Petersburg, Russia

ABSTRACT The chemical structures of some antidepressants are similar to those of recently described amine-containing ligands of acid-sensing ion channels (ASICs). ASICs are expressed in brain neurons and participate in numerous CNS functions. As such, they can be related to antidepressants action or side effects. We therefore studied the actions of a series of antidepressants on recombinant ASIC1a and ASIC2a, and on native ASICs in rat brain neurons. Most of the tested compounds prevented steady-state ASIC1a desensitization evoked by conditioning acidification to pH = 7.1. Amitriptyline also potentiated ASIC1a responses evoked by pH drops from 7.4 to 6.5. We conclude that amitriptyline has a twofold effect: it shifts activation to less acidic values while also shifting steady-state desensitization to more acidic values. Chlorpromazine, desipramine, amitriptyline, fluoxetine and atomoxetine potentiated ASIC2a response. Tianeptine caused strong inhibition of ASIC2a. Both potentiation and inhibition of ASIC2a were accompanied by the slowdown of desensitization suggesting distinct mechanisms of action on activation and desensitization. In experiments on native heteromeric ASICs, tianeptine and amitriptyline demonstrated the same modes of action as on ASIC2a although with reduced potency. KEYWORDS: antidepressants, ASICs, CHO cells, patch-clamp, inhibition, potentiation

1 ACS Paragon Plus Environment

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 38

INTRODUCTION Pharmacological treatment of depression mainly employs drugs that target monoaminergic transmission1. Important classes of antidepressants (ATDs) behave as serotonin reuptake inhibitors, serotonin-norepinephrine reuptake inhibitors and monoamine oxidase inhibitors. However, small molecules of ATDs have numerous targets in the central nervous system (CNS). A classic example is fluoxetine, which affects different potassium channels2-4,

calcium

channels5,

channels6,

sodium

GABAa

receptors7,

nicotinic

cholinoreceptor8, NMDA9 and AMPA10 types of ionotropic glutamate receptors. Other ATDs also demonstrate complex pharmacological profiles. At the beginning of the century, antidepressant action was described for the NMDAR channel blocker ketamine11,which had previously been considered to be just an anaesthetic12. The acid-sensing ion channels (ASICs) open for sodium (and sometimes for calcium) permeation in response to extracellular acidification. They are widely distributed in the CNS and PNS13. Recent data show that ASICs are activated during synaptic transmission because of the acidic content of synaptic vesicles14,15. Although the recorded ASIC-mediated synaptic currents are small, they can play a modulatory role and contribute to synaptic plasticity effects, learning and memory processes16. Moreover, ASICs are thought to have a part in the fear and anxiety behaviour17-20, mechanosensitivity21, chemosensing22 and vision23,24. ASICs are also involved in the pathologies of CNS and PNS, such as ischemia, angina, diabetes, stroke, epilepsy, multiple sclerosis, Alzheimer disease, Parkinson disease, psychiatric diseases and others16,25,26. Therefore, ligands of ASICs attract increasing attention. We have previously found that rather simple amine-containing compounds (Fig. 1a) affect ASICs in a subunit-specific manner, leading to a channel block and/or a shift of pHdependence of activation and steady-state desensitization (SSD)27-29. The NMDAR channel blocker memantine30 and the H4R antagonist A94393131 block the pore of ASIC1a homomers as well as previously described amiloride32,33, diminazene34,35 and 4-aminopyridine36. Another 2 ACS Paragon Plus Environment

Page 3 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

NMDAR channel blocker, 9-aminoacridine, causes ASIC1a inhibition27 by shifting the pH dependence of activation to more acidic values. Histamine shifts the pH dependence in the opposite direction, resulting in a potentiating effect29. Both histamine and memantine reduce SSD and allow ASIC1a to function under conditions of acidification30,37. Phenylcyclohexyl derivatives IEM-1921 and IEM-2044 potentiate ASIC2a; with the latter compound also having a potentiating effect on ASIC1a28. Despite the complexity of structure-function relationships, which require further analysis, a common feature of the chemical structures is the presence of hydrophobic/aromatic moiety connected to the amino group. Many well-known ATDs have a similar structural organization (Fig. 1a and b). In particular, IEM-1921 can be considered to be a ‘simplified’ ketamine. This obvious similarity inspired us to look for new ASIC ligands among ATDs. We selected several representative structures for this study (Fig. 1b) and examined the action of these drugs on recombinant ASIC1a and ASIC2a homomers. Representative experiments were also performed on native heteromeric ASICs in the rat hippocampal interneurons.

RESULTS AND DISCUSSION To determine the action of ATDs on ASIC1a and ASIC2a homomeric receptors, we employed CHO expression system. Perfusion of the non-transfected CHO cells by acidic solutions did not produce any noticeable effect on the recorded currents (Fig. 2a). Application of ATDs (300 µM) under neutral (pH = 7.4) conditions did not typically evoke a response; the exceptions were fluoxetine and chlorpromazine, which occasionally produced slowly increasing reversible inward currents of about 10–30 pA (Fig. 2a).

3 ACS Paragon Plus Environment

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Estimation of activities on ASIC1a homomers Perfusion of the ASIC1a expressing cell with ATDs (300 µM) at neutral pH = 7.4 did not produce ASIC1a-mediated responses (data not shown). Fluoxetine and chlorpromazine caused inward currents, which were similar to the currents in non-transfected cells. For analysis of drug action we used only records from the cells, in which these currents were absent or constituted less than 10% of the ASIC1a-mediated response. Based on the results of our previous studies30,37 we selected three different experimental protocols to reveal different types of drug activities (Fig. 2d). In the first protocol ASIC1a were activated with pH = 6.5, which caused 27 ± 5 % (n = 12) of maximal response. ATDs were applied simultaneously with the activating pH drop to reveal the action on the open channels. Under such conditions, most of the drugs did not have a noticeable effect on ASIC1a responses (Fig. 2d). Only amitriptyline produced statistically significant potentiation (30 ± 13%, n = 6, P < 0.05). The potentiating effect developed in the first activation in the presence of the drug and was fully reversible after 1–2 washout activations. Previous studies38,39 have demonstrated that ASIC1a ligands often affect SSD of the receptors. To reveal this mode of action for ATDs, we used conditioning pH = 7.1 that causes 77 ± 10 % (n = 12) desensitization. To discriminate drug action on both open and desensitized channels, two application protocols were used: continuous perfusion by the ATD-containing solution and ATD application simultaneously with activation. When applied simultaneously with pH = 6.5, the drugs did not have a significant effect on ASIC1a-mediated currents (Fig. 2b and d), except amitriptyline (53 ± 40% potentiation, n = 6, P < 0.05). This value does not differ significantly from the amitriptyline effect at conditioning pH = 7.4 (P > 0.05, one-way ANOVA followed by Tukey post-hoc test (ANOVA)). Drastically different results were obtained when the drugs were present in both conditioning (pH = 7.1) and activating (pH = 6.5) solutions (Fig. 2c and d). Citalopram, chlorpromazine, desipramine, amitriptyline, fluoxetine and atomoxetine produced a strong (3–6-fold) increase of amplitude. There were 4 ACS Paragon Plus Environment

Page 4 of 38

Page 5 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

no statistically significant differences in this group (P>0.05, ANOVA). The significantly smaller potentiating effect was found for tianeptine (85 ± 30%, n = 6, P < 0.05) whereas ketamine did not affect ASIC1a-mediated currents.

Mechanisms of ASIC1a modulation - amitriptyline Among the compounds tested, only amitriptyline had a significant potentiating activity in all three protocols used for the initial screening. Therefore, this drug was selected for detailed analysis. We have previously demonstrated that interaction of the amine-containing ligands with ASICs is complex and can be mediated by different molecular mechanisms31,37. Therefore, we compared the effects of amitriptyline in three drug application protocols: (i) application of amitriptyline for 30 seconds before receptor activation, when channels are mainly closed or desensitized, depending on conditioning pH; (ii) application of amitriptyline simultaneously with pH drop, affecting the open channels; (iii) continuous perfusion, when the drug is present in both conditioning and activating solutions. To minimise influence of cell-to-cell variations all three protocols were tested on the same cells in different order. In all protocols, amitriptyline demonstrated potentiating effects on ASIC1a current amplitudes elicited by a pH drop from 7.4 to 6.5 (Fig. 3a). Continuous perfusion and application of amitriptyline before activation had a similar (P > 0.1, repeated measurements one-way ANOVA followed by Tukey post-hoc test (RM-ANOVA)) potentiating effect on ASIC1a currents (80 ± 24%, and 76 ± 32%, correspondingly). These effects were significantly stronger (P0.05, n = 6) (Fig. 3b), but in all protocols with conditioning pH = 7.4, the decay time constant was 5 ACS Paragon Plus Environment

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

decreased in the presence of amitriptyline (Fig. 3a and c). For continuous drug presence the effect was 37 ± 12 % (n=6), with similar effect in the other protocols (P > 0.05, RMANOVA). At a glance, acceleration of response decay looks strange since response decay reflects development of desensitization. Yet, acceleration of the response decay may also be the result of pore block of ASIC1a31. Charged pore blockers usually demonstrate voltage dependence of action including previously described amiloride32,33, memantine30 and diminazene34,35. However, experiments at -20, -80 and -120 mV holding voltages did not reveal a significant difference of the amitriptyline effects on amplitude and kinetics (data not shown). Another factor affecting shape of the response is activating pH. Modest acidifications activate ASICs asynchronously, making the response shape smooth with rather slow decay while strong acidifications activate the channels almost simultaneously, resulting in sharper shape of the response and its fast decay. Thus, presence of amitriptyline changes the response shape and amplitude in the same way as activation by saturating proton concentrations does (data not shown). Analysis of pH-dependence of ASIC1a activation demonstrated parallel shift of the activation curve in the presence of amitriptyline (Fig. 3e). Fitting by Hill equation has shown that pH50 value increased from 6.07 ± 0.02 in control to 6.54 ± 0.03 in the continuous presence of 300 µM of the drug. The parallel shift of activation may be due to allosteric modulation of pH sensitivity or by direct binding to the one of the proton-sensing groups. We also compared the amitriptyline effects in three above-mentioned application protocols in the case of partially desensitized ASIC1a in paired experiments (Fig. 3b). The smallest potentiating effect (43 ± 22%, n = 6, P < 0.05) was observed for simultaneous application of the drug with activation by pH = 6.5. The potentiation was much higher if amitriptyline was applied before activation (302 ± 117%, n = 6, P < 0.05); and in the case of continuous presence of amitriptyline (450 ± 84%, n = 6, P < 0.05). There was no statistically significant difference between the latter values (P > 0.1, RM-ANOVA). It should be noted 6 ACS Paragon Plus Environment

Page 6 of 38

Page 7 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

that despite large relative potentiation of the response its amplitude remained smaller that the maximal ASIC1a response evoked by pH drop from 7.4 to 5.0 (Fig. 3d). Analysis of the SSD of ASIC1a showed the shift of the curve to the more acidic values; pH50 changes from 7.17 ± 0.04 in control to 7.08 ± 0.03 in the presence of amitriptyline (Fig. 3e). The potentiation of response at conditioning pH= 7.4 is due to the shift of the pH-dependence of activation. Our data obtained in different application protocols strongly suggest that amitriptyline binds to the closed and desensitized channels. However, relatively weak activity in the protocol of simultaneous application (Fig. 3b) does not necessarily mean state-dependence of binding. If binding kinetics is slow, the effect cannot completely develop in this protocol and long pre-application is required. Concentration dependence of amitriptyline action was estimated under conditions of its maximal activity - partial desensitization by conditioning pH = 7.1 and incomplete activation by pH = 6.5. Fig. 3f shows that the drug applied continuously or before activations by pH drop produces the same effect. The minimal concentration causing significant potentiation is 30 µM (44 ± 28%, n = 11, P < 0.05). The effect did not saturate at 300 µM, so it was not possible to estimate the EC50 value. Higher concentrations caused non-specific influence on the baseline current both in transfected and non-transfected CHO cells and affected stability of the recordings. The most common effect of the drugs on ASIC1a is the response potentiation under conditions of partial SSD. Only ketamine was completely inactive under such conditions. More detailed analysis of amitriptyline action demonstrated that the potentiation is due to acidic shift of pH-dependence of SSD. This mode of action was earlier demonstrated for GMQ40, histamine37, dynorphyn39 and memantine30. The opposite shift, which results in ASIC1a desensitization at neutral pH values, was revealed only for psalmotoxin38. Thus, the anti-desensitizing effect on ASIC1a does not appear to be highly specific.

7 ACS Paragon Plus Environment

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The second effect, potentiation of partially activated ASIC1a due to alkaline shift of activation, was found only for amitriptyline. The same mode of action was reported for histamine29, whereas GMQ40, mambalgins41 and 9-aminoacridine27 demonstrate the acidic shift of activation. Various GMQ derivatives studied in the recent work42 also cause acidic shift of activation but some compounds induce alkaline shift. Thus, there is no apparent correlation between drug actions on ASIC1a activation and desensitization. Although bold conclusion on the structural determinants of acidic and alkaline shifts of activation cannot be made at present, it seems important to note, that for 9-aminoacridine, GMQ and its derivatives the charged group is directly connected to flat aromatic moiety, whereas in the structures of compounds causing alkaline shift of activation the amino group is connected to hydrophobic moiety through flexible spacers. The third known mode of action, inhibition by channel block, which was reported for memantine30, diminazene34,35 and 4-aminopyridine36, was not found among the drugs tested in the present study.

Estimation of activities on ASIC2a homomers To activate ASIC2a, we used extracellular solutions with pH = 5.45 and pH = 5.0, which cause 26 ± 7% (n=10) and 66 ± 1% (n=8) of maximal response, correspondingly. Four protocols illustrated in Fig. 4a were used to reveal different types of activity. First, we tested the effects of ATDs on current amplitudes when the drugs were co-applied with pH drops. Amplitudes of responses evoked by pH drop to 5.45 were strongly potentiated by chlorpromazine (275 ± 98%, n = 5, P < 0.05) and desipramine (202 ± 58%, n = 5, P < 0.05). Potentiation by amitriptyline (57 ± 27%, n = 8, P < 0.05) and fluoxetine (44 ± 36%, n = 6, P < 0.05) was significantly weaker (P < 0.05, ANOVA). 8 ACS Paragon Plus Environment

Page 8 of 38

Page 9 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

Only desipramine significantly potentiated the ASIC2a responses (37 ± 24%, n = 8, P < 0.05) to stronger acidification (pH = 5.0) as shown in Fig. 4c. For all these drugs including desipramine the effects at pH = 5.45 and pH = 5.0 were significantly (P < 0.05, unpaired ttest) different. Tianeptine and atomoxetine caused inhibition at both values of activating pH, 80 ± 6% (n = 5) at pH = 5.45 and 70 ± 9% (n = 9) at pH = 5.0 for tianeptine, and 17 ± 6% (n = 5) at pH = 5.45 and 26 ± 9% (n = 8) at pH = 5.0 for atomoxetine (P < 0.05 in all cases). In contrast to potentiating action, for both these drugs the effects at different pH were not significantly different (P > 0.1, ANOVA). To reveal possible effects on ASIC2a desensitization we used pH drop from 7.4 to 5.45 during 30 s as conditioning acidification evoking desensitization, which was followed by testing acidification to 5.0. In the control experiments, this conditioning pH resulted in a 65– 75% decrease of response to pH = 5.0. The amplitudes of testing responses to pH = 5.0 were measured relative to the baseline current at pH = 7.4. In the case of drug application simultaneously with testing acidification their effects were not dependent on conditioning pH (P > 0.05, unpaired t-test). In contrast, presence of the drugs in both conditioning and testing solutions changed their effects. Chlorpromazine, amitriptyline and desipramine, which potentiated response to pH = 5.45, also potentiated responses to pH = 5.0 if they were preceded by pH = 5.45 conditioning acidification. Atomoxetine, which demonstrated modest inhibition in the protocols with conditioning pH=7.4, caused potentiation of pH = 5.0 response (95 ± 25%, n = 5, P < 0.05) if it was present during conditioning by pH = 5.45 (Fig. 4d). Inhibitory action of tianeptine (41 ± 9%, P