Potential Drug Candidates to Treat TRPC6 Channel Deficiencies in the Pathophysiology of Alzheimer’s Disease and Brain Ischemia

Alzheimer’s disease and cerebral ischemia are among the many causative neurodegenerative diseases that lead to disabilities in the middle-aged and elderly population. There are no effective disease-preventing therapies for these pathologies. Recent in vitro and in vivo studies have revealed the TRPC6 channel to be a promising molecular target for the development of neuroprotective agents. TRPC6 channel is a non-selective cation plasma membrane channel that is permeable to Ca2+. Its Ca2+-dependent pharmacological effect is associated with the stabilization and protection of excitatory synapses. Downregulation as well as upregulation of TRPC6 channel functions have been observed in Alzheimer’s disease and brain ischemia models. Thus, in order to protect neurons from Alzheimer’s disease and cerebral ischemia, proper TRPC6 channels modulators have to be used. TRPC6 channels modulators are an emerging research field. New chemical structures modulating the activity of TRPC6 channels are being currently discovered. The recent publication of the cryo-EM structure of TRPC6 channels should speed up the discovery process even more. This review summarizes the currently available information about potential drug candidates that may be used as basic structures to develop selective, highly potent TRPC6 channel modulators to treat neurodegenerative disorders, such as Alzheimer’s disease and cerebral ischemia.


Introduction
Due to increased life expectancies, neurodegenerative diseases (NDD), such as Alzheimer's disease (AD), dementia, and cerebrovascular diseases, are considered by WHO the main cause of disability in the coming decades. Currently, there are no effective disease-modifying or preventing therapies for those NDDs.
AD is caused by the progressive loss of neurons in brain structures that are responsible for memory acquisition and preservation, such as the hippocampus and cortical areas. The exact mechanism that causes neurons to die is not known. Among the most studied toxic changes of proteins that cause neuronal degeneration in AD are extracellular amyloid-beta (Aβ) aggregates and intracellular hyperphosphorylated tau (p-tau) that forms neurofibrillary tangles. Recently, disruption of immune

TRPC Channels and Their Regulation in Cells
Transient receptor potential (TRP) cation channels form a large family of multifunctional cell sensors. There are 29 TRP channels that can be described and divided into six subfamilies based on sequence homology: seven canonical channels (TRPC), six vanilloid channels (TRPV), eight melastatin-related channels (TRPM), three polycystic channels (TRPP), three mucolipins (TPRML), and one ankyrin channel (TRPA) [28].
Analysis of the expression of different types of TRPC channels in situ using the Allen Brain Atlas and of gene expression in different regions of the mouse brain has demonstrated that TRPC6 is highly expressed in the hippocampus [14]. Western blot analysis of the rat hippocampus has shown that TRPC6 levels are enhanced in postsynaptic structures compared with synaptosomes. Moreover, electron microscopy has shown that TRPC6 is mostly located in the postsynaptic sites [13].
TRPs are cloned and identified, assuming that they are calcium-selective and activated by the emptying of internal Ca 2+ stores (store-operated channels (SOC)) [31]. After their initial functional characterization, it has turned out that both assumptions do not hold true, especially for TRPCs [31]. These ion channels only show a moderate Ca 2+ selectivity (PCa/PNa from~0.5 to 9), and the TRPC3/6/7 subfamily of TRPC channels can be activated by the second messenger diacylglycerol (DAG) produced by receptor-activated phospholipase C (PLC) without any involvement of internal stores (receptor-operated channels (ROC)) [32].
Store-operated Ca 2+ entry (SOCE) via TRPCs happens when 1,4,5-trisphosphate (IP3) or some other intracellular mechanism empties Ca 2+ stored in the endoplasmic reticulum (ER). The fall in the ER Ca 2+ concentration signals to the plasma membrane to open store-operated channels [33]. The major breakthrough in the understanding of SOCE physiology happened when Ca 2+ -sensing stromal interaction molecule (STIM) ER proteins and plasma membrane (PM) Orai 1-3 channels have been described [34][35][36][37]. STIM1 and 2 proteins reside in the ER and monitor ER Ca 2+ concentration with their EF-hand motif. When the ER Ca 2+ concentration drops, Ca 2+ dissociates from EF-hand, thus allowing STIM proteins to oligomerize and move to ER-PM tight junctions where they interact with PM Orai and TRP channels in order to facilitate Ca 2+ flow.

Role of TRPC6 in the Formation of Excitatory Synapses
TRPC6 plays a certain physiological role in the formation of excitatory synapses. Particularly, TRPC6 overexpression increases dendritic spine density [13] and attenuates mushroom spine loss in presenilin 1 knock-in (PS1-KI) and amyloid precursor protein knock in (APP-KI) hippocampal neurons to the level of wild type neurons [14]. At the same time, overexpression of TRPC1, TRPC3-5, and TRPC7 does not affect spine morphology [14]. Meanwhile, the downregulation of TRPC6 expression leads to spine density reduction, lowers the frequency of spontaneous miniature excitatory postsynaptic currents (mEPSCs) [13], decreases the expression of postsynaptic density protein 95 (PSD95), and inhibits phosphorylation of calcium-calmodulin-dependent protein kinase II (pCaMKII) [14].
Activation of TRPC6 can promote spine formation via the STIM2-neuronal SOCE-CaMKII pathway. The expression level of STIM2 is downregulated in hippocampal neurons from PS1-M146V-KI [38] and APP-KI [14] mice as well as in a cellular model of amyloid synaptotoxicity [39]. The STIM2 reduction may be a compensatory response to ER Ca 2+ overload in these models and may lead to the loss of mushroom spines in PS1-M146V-KI and APP-KI hippocampal neurons. TRPC6 and Orai2 channels are suggested to be the key components of neuronal SOCE in hippocampal cells [14]. It is hypothesized that the downstream molecule for TRPC6-mediated SOCE in hippocampal neurons is CaMKII since upregulation of neuronal SOCE activity recovers phosphorylation of CaMKII, restores the number of mushroom spines in hippocampal neurons in mouse models of AD [38,39], and induces long-term potentiation in PS1-KI, APP-KI, and 5xFAD hippocampal slices [14,21].
Alternative TRPC6-downstream signaling pathways are the Ca 2+ /calmodulin-dependent kinase IV (CaMKIV) pathway and the cAMP-response element-binding protein (CREB) pathway, which is important for dendritic growth in hippocampal neurons [17]; another study has shown that the CaMKIV-CREB pathway is important in promoting synapse and dendritic spine formation, spatial memory, and learning [13]. TRPC6 also acts as a negative regulator that suppresses NMDA-induced Ca 2+ influx in hippocampal neurons [15]. In addition, NMDAR has been shown to regulate transcription and degradation of TRPC6 in neurons in a bidirectional manner through NMDAR subunit 2A (NR2A) or NR2B activation [40].

Hypo-and Hyperactivation of TRPC6 Channels in Different Pathogenetic Forms of AD
There is evidence that different genetically inherited familial forms of AD (fAD) can cause TRPC6 dysfunction [14,21,24]. Both hypo- [14,21,22] and hyperactivation [24] of TRPC6 channels have been reported for different fAD-associated mutations in APP and PS genes.
fAD-associated PS2-N141I, M239V mutations cause downregulation of TRPC6-mediated Ca 2+ entry in transiently transfected HEK cells. Lessard et al. suggested that TRPC6 downregulation by PS2-N141I, M239V does not depend on ER Ca 2+ content but rather involves the interaction of PS2 with an intermediate protein of unknown origin [22]. Later on, contradictory results were obtained on ER Ca 2+ content in PS2-N141I expressing cells [41,42]. In double knockout fibroblasts, PS2-N141I increases ER Ca 2+ content [41] but lowers it in Hela cells [42]. According to the classical understanding of the SOCE physiology, overloaded ER Ca 2+ stores downregulate SOCE [33]; thus, the data provided by Tu et al. [41] seem to be more relevant.
PS1-M146V is an fAD-associated mutation that has been reported to downregulate TRPC6-dependent Ca 2+ entry in hippocampal neurons in store-operated mode [14]. PS1-M146V has been shown to increase ER Ca 2+ content in mouse embryonic fibroblasts (MEFs) [41] and in neurons [43]. It has been demonstrated in previous studies [38,44] that ER Ca 2+ stores are overloaded in neurons from AD mouse models. Furthermore, it has been discovered that overloaded ER Ca 2+ stores cause compensatory downregulation of TRPC6 channels in PS1-M146V neurons [14]. A similar impact on TRPC6 function has been reported for fAD-associated mutations in APP (KM670/671NL and I716F) [14] as well as for Aβ toxicity in a cell culture model of AD [21]. There are fAD mutations in PSEN1 (PSEN1M146L, PSEN1S170F, PSEN1I213F, PSEN1E318G, PSEN1P117R, PSEN1L226F, PSENA246E) [45,46] and PSEN2 (PSEN2M239I, PSEN2T122R) [42], which have been reported to downregulate SOCE, although their role in the regulation of TRPC6 function has not been investigated yet. TRPC6 activators have been shown to recover the percentage of mushroom spines in cell culture models of fAD and induce long-term potentiation in hippocampal brain slices taken from AD mouse models [14,21]. Based on these results, it is suggested that activators of TRPC6 may have a therapeutic value for the treatment of fAD with TRPC6 hypofunction [14,[47][48][49].
PS1-∆E9 mutation has been reported to empty ER Ca 2+ stores [41] and enhance SOCE [24,50]. There are other fAD mutations in PSEN1 (PSEN1D257A, PSEN1D385A), which have been reported to enhance SOCE [51]; however, their role in the regulation of TRPC6 function has not been investigated yet. Today, there is only one fAD-associated PS1-∆E9 mutation that has been shown to upregulate the TRPC6 function in store-operated mode [24]. A TRPC6 inhibitor has been shown to recover mushroom spine percentage in a cell culture model of fAD [24]. Inhibitors of TRPC6 have been proposed to have therapeutic effects in fAD with TRPC6 hyperfunction [24,49].
To conclude the section, in order to normalize TRPC6 function in neurons and preserve the stability of excitatory synaptic contacts, suitable pharmacological agents have to be used for distinct genetic forms of AD.

Cerebral Ischemia as a Risk Factor for AD Development
Recent experimental and clinical findings have demonstrated a high degree of correlation between cerebral ischemia and AD [4,52,53]. While some studies have indicated that ischemic stroke significantly increases the risk of AD [4], others, in turn, have associated AD with a higher risk of stroke [5,6]. Previous studies have suggested that almost 30% of AD subjects bear evidence of cerebral infarction at autopsy [54,55]. A meta-analysis that comprises seven cohort studies and two nested case-control studies has found that a history of stroke is associated with the development of AD [4]. Notably, several lines of evidence suggest that AD patients have a high risk of cerebral ischemia [5,6]. For example, among AD patients with no history of previous stroke, vascular dementia, or other cerebral degenerative diseases, the incidence of ischemic stroke amounts to 37.8 per 1000 persons (versus 23.2 in non-AD controls) [5]. In another register-based matched cohort study [6], patients with Alzheimer's dementia have a higher risk of hemorrhagic stroke, while there is no difference in ischemic stroke incidence. When the results are analyzed within different age groups, the risk of ischemic stroke is found to be increased among AD patients younger than 80 years [6].

Role of TRPC6 in the Development of Ischemia
Ca 2+ overload is one of the main molecular mechanisms involved in ischemic cell damage and death [56]. TRPC6, along with a few other prominent members of the family, has recently gained considerable attention as a promising target for the prevention of Ca 2+ overload [23]. Dysregulation of TRPC6 activity has been implicated in ischemic stroke [23,57], as well as retinal ischemia [58], and renal hypoxia following cerebral ischemia [59].
On the one hand, upregulation and maintenance of TRPC6 activity prevent NMDAR hyperactivation and the subsequent Ca 2+ influx, development of excitotoxicity, and neuronal death. In supporting this notion, both direct activation by 1-oleoyl-2-acetyl-sn-glycerol (a synthetic analog of diacylglycerol, the main endogenous TRPC6 agonist) and overexpression of TRPC6 inhibit NMDA-induced currents in cultured hippocampal neurons [15], and TRPC overexpression attenuates excitotoxic damage in hippocampal and cortical neurons [16]. Trpc6-transgenic mice with an elevated basal level of TRPC6 expression are less susceptible to cerebral ischemia than their wild-type littermates and have lower mortality rates, reduced infarct volumes, and better neurological outcomes after middle cerebral artery occlusion (MCAO) [16]. Recent studies have shown that TRPC6-mediated signaling promotes neuronal survival [60], the brain-derived neurotrophic factor-mediated axonal growth cone guidance [61], dendritic outgrowth and branching [17], and excitatory synapse formation [13]. In addition, positive modulation of TRPC6 activity allows for the sustained activation of the CREB/CaMK-IV and Ras/MEK/ERK pathways, which is vital for neuronal development, survival, and proper functioning [23]. Blocking CREB signaling hinders post-stroke recovery [62], and CaMK-IV inhibition impairs blood-brain barrier integrity and exacerbates ischemic injury [63]. In turn, elevated CREB and CaMK-IV activity is associated with improved post-stroke outcomes in a number of animal studies [62][63][64][65][66].
On the other hand, some experimental data suggest that the upregulation of TRPC6 activity increases intracellular Ca 2+ concentrations concomitantly with NMDAR activation, further exacerbating excitotoxic damage to neurons [16]. Oxygen-glucose deprivation in cultured cortical neurons and MCAO in wild-type mice are associated with elevated TRPC6 expression and activity, while TRPC6 deletion attenuates glutamate-and NMDA-induced cytotoxicity and reduced infarct volumes [25]. Knockdown of TRPC3, 6, and 7 prevents apoptosis in cultured astrocytes and ameliorates ischemic brain injury in mice [26]. Moreover, prevention of TRPC6 hyperactivation results in increased neuronal viability, reduced infarct volumes and brain edema, and improved functional recovery following acute ischemic stroke in rats [67][68][69] and crab-eating macaques [67]. Other TRPC6 inhibitors have also been reported to exert beneficial effects in experimental models to some extent relevant to ischemic brain injury (e.g., acute renal ischemia/reperfusion injury) [70][71][72][73][74].
Existing controversies regarding the TRPC6 function in the development of brain ischemia might be due to different experimental settings (i.e., rodent model, sex, age, the method used to model ischemia). However, similarly to AD, brain ischemia seems to be heterogenic, meaning that one group of patients has TRPC6 hypofunction, and another one has TRPC6 hyperfunction. This indicates the need to develop pathology-dependent strategies to treat different NDD patients.

Available Drug Candidates to Modulate TRPC6 Activity
To date, two strategies have been proposed for the pharmacological modulation of TRPC6 activity for the treatment of Alzheimer's disease and cerebral ischemia: (1) TRPC6 activation to allow Ca 2+ influx via neuronal SOCE and sustain the stability of postsynaptic contacts (for AD) and to attenuate NMDAR activity and prevent calcium-dependent excitotoxicity [15,16] (for ischemia); (2) inhibition of TRPC6 in order to prevent calcium overload and the subsequent cell damage [23,64] (for AD and ischemia). Although apparently mutually exclusive, both of these strategies are aimed to keep the intracellular Ca 2+ concentration within the normal range, which requires limiting its entry via transmembrane channels and/or release from intracellular stores [75]. Given the pivotal role of NMDAR in excitotoxic neuronal damage, NMDAR blockers have been proposed as potential neuroprotective agents, although most of them have failed to show substantial effectiveness in human patients so far [76]. Due to that fact, TRPC6 has emerged as an alternative therapeutic target for AD and ischemic stroke [23].

Endogenous Ligands and Analogs
DAG A compound containing both DAG and arachidonic acid fragments, 1-stearoyl-2-arachidonylsn-glycerol (SAG), elicits a rapid Ca 2+ flux into HEK293 cells, while 1-stearoyl-2-docosahexaenoylsn-glycerol (SDG), which does not contain an arachidonic acid moiety, has significantly lower potency [78]. 1-oleoyl-2-acetyl-sn-glycerol (OAG), a diacylglycerol analog and a TRPC3/6/7 channel modulator, is found to cross the plasma membrane and intracellularly activate the channels [32,90]. OAG has been shown to activate Ca 2+ -permeable channels, displaying TRPC6-like properties in cultured cortical neurons [91]. In addition, OAG has been demonstrated to increase field excitatory postsynaptic potential (fEPSP) levels in a TRPC-dependent manner in hippocampal slices from wild-type mice [92], indicating that it might have neuroprotective effects. However, as far as we are aware, those compounds have not yet been evaluated in in vivo models of AD and cerebral ischemia.

Lysophosphatidylcholine
Lysophosphatidylcholine (LPC) is produced from phosphatidylcholines via partial hydrolysis generally catalyzed by phospholipase A2. Increased LPC production has been observed in various disorders of the central nervous system, including stroke and AD [93], and is associated with acute and chronic brain ischemia [94,95]. Results of a cohort study have suggested that LPC levels could be used as a tool for ischemic stroke risk stratification in patients who have suffered a transitory ischemic attack before [96]. LPC is shown to activate TRPC6 channels and promote Ca 2+ flux into endothelial cells, hampering their migration and preventing endothelial healing, thus contributing to atherogenesis [97,98]. 20-Hydroxy-5Z,8Z,11Z,14Z-Eicosatetraenoic Acid 20-hydroxy-5Z,8Z,11Z,14Z-eicosatetraenoic acid (20-HETE) is the main eicosanoid metabolite of arachidonic acid and a potent inflammatory vasoconstrictor. In HEK293 cells, 20-HETE (half maximal effective concentration, EC 50 of 0.8 µM) elicits a three-fold increase in TRPC6 activity (as indicated by an increased inward, the non-selective current observed in whole-cell patch-clamp recordings) but does not affect intracellular Ca 2+ concentrations [77]. In isolated guinea pig airway smooth muscle cells, it induces a dose-dependent inotropic effect via TRPC6 activation and the subsequent promotion of Ca 2+ entry [99]. Nevertheless, 20-HETE has been shown to have detrimental effects in ischemic and traumatic brain injury, which might be explained by its vasoconstrictor properties, and has even been proposed as a predictor of poor prognosis in stroke patients [100]. To our knowledge, 20-HETE has not yet been tested in AD models, but it has been shown to activate TRPV1 channels in dorsal root ganglia cultures [101].
Lipopolysaccharide stimulation in 28 d post-MCAO isolated mouse astrocytes Co-incubation for 16 h increased viability [103] Lipopolysaccharide stimulation in 28 d post-MCAO isolated mouse cortical neurons increased viability [103] NMDA toxicity in rat hippocampal slices Co-incubation for 30 min reduced edema [104] Male Sprague-Dawley rats injected with fibrillary Aβ Intrahippocampal co-injection of Aβ with the drug for 14 days amyloid deposits disaggregation reduced spatial memory deficit [105] Rat hippocampal slice cultures Co-incubation for 24 h increased proportion of mature stubby spines [106] Hippocampal cultures from PS1-M146VKI and APPKI transgenic mice Incubation for 16 h increased percentage of mushroom spines in TRPC6-dependent manner increased neuronal SOCE in postsynaptic spines [14] Hippocampal cultures treated with synthetic Aβ42 peptides Co-incubation for 16 h increased percentage of mushroom spines increased neuronal SOCE in postsynaptic spines [21] Rat transient MCAO Icv injection at 2 h after reperfusion reduced infarct volume reduced sensory and motor deficits   increased viability [103] NMDA toxicity in rat hippocampal slices Co-incubation for 30 min reduced edema [104] Male Sprague-Dawley rats injected with fibrillary Aβ Intrahippocampal co-injection of Aβ with the drug for 14 days amyloid deposits disaggregation reduced spatial memory deficit [105] Rat hippocampal slice cultures Co-incubation for 24 h increased proportion of mature stubby spines [106] Hippocampal cultures from PS1-M146VKI and APPKI transgenic mice Incubation for 16 h increased percentage of mushroom spines in TRPC6-dependent manner increased neuronal SOCE in postsynaptic spines [14] Hippocampal cultures treated with synthetic Aβ42 peptides Co-incubation for 16 h increased percentage of mushroom spines increased neuronal SOCE in postsynaptic spines [21] Lipopolysaccharide stimulation in 28 d post-MCAO isolated mouse astrocytes Co-incubation for 16 h increased viability [103] Lipopolysaccharide stimulation in 28 d post-MCAO isolated mouse cortical neurons increased viability [103] NMDA toxicity in rat hippocampal slices Co-incubation for 30 min reduced edema [104] Male Sprague-Dawley rats injected with fibrillary Aβ Intrahippocampal co-injection of Aβ with the drug for 14 days amyloid deposits disaggregation reduced spatial memory deficit [105] Rat hippocampal slice cultures Co-incubation for 24 h increased proportion of mature stubby spines [106] Hippocampal cultures from PS1-M146VKI and APPKI transgenic mice Incubation for 16 h increased percentage of mushroom spines in TRPC6-dependent manner increased neuronal SOCE in postsynaptic spines [14] Hippocampal cultures treated with synthetic Aβ42 peptides Co-incubation for 16 h increased percentage of mushroom spines increased neuronal SOCE in postsynaptic spines [21] Mouse transient MCAO Intranasal administration q.d. for 7 d starting at day 7 post-MCAO increased hippocampal neurogenesis improved post-stroke depression and anxiety reduced memory deficit [107] Icv injections at 1, 24, and 48 h after MCAO reduced microglial activation reduced infarct volume reduced neurological deficit [108] Icv injections q.d. for 14 d starting at day 14 post-MCAO increased angiogenesis reduced motor deficit [109] increased angiogenesis increased subventricular neurogenesis reduced motor deficit [103] Mouse permanent MCAO Ip injection before ischemia onset no effect on infarct volume or brain edema [ Icv injections at 1, 24, and 48 h after MCAO reduced microglial activation reduced infarct volume reduced neurological deficit [108] Icv injections q.d. for 14 d starting at day 14 post-MCAO increased angiogenesis reduced motor deficit [109] increased angiogenesis increased subventricular neurogenesis reduced motor deficit [103] Mouse permanent MCAO Ip injection before ischemia onset no effect on infarct volume or brain edema [ reduced infarct volume following an initial and recurrent stroke reduced blood-brain barrier disruption following a recurrent stroke reduced brain edema following a recurrent stroke reduced astrogliosis following a recurrent stroke no effect on cerebral blood flow during or after recurrent stroke [112] Rat increased viability of male neurons no effect on female neurons [118] Glutamate toxicity in isolated rat embryonic hippocampal neurons Co-incubation for 10 min increased viability [119] Oxygen/glucose deprivation in isolated rat brain endothelial cells Preincubation for 3 d increased viability [112] Mouse transient MCAO Ip injection at 48 h before MCAO reduced infarct volume no effect on cerebral blood flow [113] Oral gavage q. reduced infarct volume following an initial and recurrent stroke reduced blood-brain barrier disruption following a recurrent stroke reduced brain edema following a recurrent stroke reduced astrogliosis following a recurrent stroke no effect on cerebral blood flow during or after recurrent stroke [112] Rat asphyxial cardiac arrest Ip injection at 48 h before cardiac arrest enhanced ATP synthesis efficiency in hippocampal mitochondria prevented hippocampal neuronal apoptosis [115] Gerbil transient bilateral common carotid artery occlusion Ip injections during occlusion or at reperfusion + at 24 h after reperfusion reduced hippocampal microglial activation prevented hippocampal-delayed neuronal death [116] Clinical trials in patients diverse affected neuroinflammation, Aβ deposition, and adaptive immunity in patients with mild to moderate Alzheimer's disease for a review, see [117]  increased viability of male neurons no effect on female neurons [118] Glutamate toxicity in isolated rat embryonic hippocampal neurons Co-incubation for 10 min increased viability [119] Oxygen/glucose deprivation in isolated mouse embryonic cortical neurons Incubation from 15 min before oxygen/glucose deprivation until 24 h of reoxygenation increased viability of male neurons no effect on female neurons [118] Glutamate toxicity in isolated rat embryonic hippocampal neurons Co-incubation for 10 min increased viability       [120] Aβ42-induced toxicity in primary hippocampal neurons Co-incubation for 16 h restored mushroom spines percentage induced neuronal SOCE in postsynaptic spines  increased CBF reduced infarct volume [121] Piglet neonatal hypoxia/ ischemia 5 min-infusion at 5 min after reperfusion + hypothermia at 3 h after reperfusion increased neuronal viability in the putamen, cortex, and thalamus prevention of seizures [122] Iv injection at 5 min after reperfusion increased neuronal viability in the putamen reduced neurological deficit no effect on cerebral blood flow (CBF) [123] 12

TS-011 N-(3-Chloro-4-morpholin-4-yl)phenyl-N′-hydroxyimidoformamide
Rat transient MCAO Iv injection at 30 min before MCAO + 1 or 2 h-infusion during MCAO reduced cortical, subcortical, and total infarct volumes reduced the delayed drop in CBF no effect on volume at risk [68] Iv injection at 20 min after MCAO + 2 h-infusion at reperfusion reduced cortical and total infarct volumes no effect on volume at risk no effect on CBF 1 h-infusion at reperfusion, 1, 2, or 4 h after reperfusion reduced infarct volumes [67] Primary hippocampal cell culture models from PS1-M146V and APPKI mice Incubation for 16 h increased percentage of mushroom spines in TRPC6-dependent manner increased neuronal SOCE in postsynaptic spines [14] Hippocampal brain slices from PS1-M146V and APPKI mice Pretreatment for 30 min recovered LTP induction increased CBF reduced infarct volume [121] Piglet neonatal hypoxia/ ischemia 5 min-infusion at 5 min after reperfusion + hypothermia at 3 h after reperfusion increased neuronal viability in the putamen, cortex, and thalamus prevention of seizures [122] Iv injection at 5 min after reperfusion increased neuronal viability in the putamen reduced neurological deficit no effect on cerebral blood flow (CBF) [123] 12

TS-011 N-(3-Chloro-4-morpholin-4-yl)phenyl-N′-hydroxyimidoformamide
Rat transient MCAO Iv injection at 30 min before MCAO + 1 or 2 h-infusion during MCAO reduced cortical, subcortical, and total infarct volumes reduced the delayed drop in CBF no effect on volume at risk [68] Iv injection at 20 min after MCAO + 2 h-infusion at reperfusion reduced cortical and total infarct volumes no effect on volume at risk no effect on CBF 1 h-infusion at reperfusion, 1, 2, or 4 h after reperfusion reduced infarct volumes [67] Rat pediatric asphyxial cardiac arrest increased CBF reduced infarct volume [121] Piglet neonatal hypoxia/ ischemia 5 min-infusion at 5 min after reperfusion + hypothermia at 3 h after reperfusion increased neuronal viability in the putamen, cortex, and thalamus prevention of seizures [122] Iv injection at 5 min after reperfusion increased neuronal viability in the putamen reduced neurological deficit no effect on cerebral blood flow (CBF) [123] 12 increased CBF reduced infarct volume [121] Piglet neonatal hypoxia/ ischemia 5 min-infusion at 5 min after reperfusion + hypothermia at 3 h after reperfusion increased neuronal viability in the putamen, cortex, and thalamus prevention of seizures [122] Iv injection at 5 min after reperfusion increased neuronal viability in the putamen reduced neurological deficit no effect on cerebral blood flow (CBF) [123] 12

TS-011 N-(3-Chloro-4-morpholin-4-yl)phenyl-N′-hydroxyimidoformamide
Rat transient MCAO Iv injection at 30 min before MCAO + 1 or 2 h-infusion during MCAO reduced cortical, subcortical, and total infarct volumes reduced the delayed drop in CBF no effect on volume at risk [68] Iv injection at 20 min after MCAO + 2 h-infusion at reperfusion reduced cortical and total infarct volumes no effect on volume at risk no effect on CBF 1 h-infusion at reperfusion, 1, 2, or 4 h after reperfusion reduced infarct volumes [67] Rat transient MCAO Iv injection at 30 min before MCAO + 1 or 2 h-infusion during MCAO reduced cortical, subcortical, and total infarct volumes reduced the delayed drop in CBF no effect on volume at risk [68] Iv injection at 20 min after MCAO + 2 h-infusion at reperfusion reduced cortical and total infarct volumes no effect on volume at risk no effect on CBF Co-incubation for 10 min increased cell viability [124] 3xTg mice AD model Administration by osmotic minipump over 28 days reduced cognitive deficit [125] Rat transient MCAO Iv injection before MCAO no effect on infarct and penumbra volumes and brain edema [124] Iv injections at 1 h before + at 1, 2, and 3 h after MCAO reduced infarct volume reduced brain edema Icv 24 h-infusion starting at 1 h before MCAO [119] 14 Meclofenamic acid 2-(2,6-Dichloro-3-methylanilino)benzoic acid Glutamate toxicity in isolated rat embryonic hippocampal neurons Co-incubation for 10 min increased viability [119] 15
In an ex vivo experiment, hyperforin (0.3 µM) promotes mature stubby spine formation and decreases the proportion of immature thin spine formation in rat hippocampal pyramidal neurons but does not affect mushroom spine density and morphology. Proper TRPC6 expression level and the presence of a fully functional TRPC6 channel are required for hyperforin to exert its effects, which suggests the key role of TRPC6 activation in its mechanism of action [106]. Using different rodent models of ischemic stroke, hyperforin has been shown to promote post-stroke neuro-and angiogenesis [103,109], inhibit microglial activation [108], attenuate brain edema [104], stimulate hippocampal neurogenesis, ameliorate post-stroke depression and anxiety, and restore memory function [107]. Chronic hyperforin treatment stimulates the expression of the tropomyosine receptor kinase B (TrkB) BDNF receptors as well as of TRPC6 in murine cortical neurons but has no effect on hippocampal neurogenesis [131]. When applied intracerebroventricularly to rats immediately after MCAO, hyperforin preserves TRPC6 activity, reduces infarct volumes, promotes functional recovery, and increases neurologic scores at 24 h after reperfusion [110].
There is a lot of evidence that hyperforin and its derivatives are highly selective towards the TRPC6 channel and do not exert similar effects on its closest relative, the TRPC3 channel [106,127,128,132]. Several studies have shown that hyperforin activates TRPC6 and increases its expression [106,133], leading to a decrease in the Aβ level and an improvement in cognitive performance in AD models [92,105,111].
The neuroprotective effect of hyperforin has been demonstrated in several rodent models of AD. In rats co-injected with amyloid fibrils and hyperforin in the hippocampus, hyperforin reduces amyloid deposit formation and, therefore, decreases the Aβ-induced neurotoxicity, reactive oxidative species formation, and attenuated behavioral impairments [105]. A more stable hyperforin derivative, tetrahydrohyperforin, also prevents the cognitive decline and synaptic impairment in double transgenic APPswe/PSEN1∆E9 mice in a dose-dependent manner. It has been shown that the neuroprotective mechanism of tetrahydrohyperforin is associated with a reduced rate of proteolytic processing of APP, decreased the total amount of fibrillar and oligomeric forms of Aβ, reduced level of tau hyperphosphorylation, and attenuated astrogliosis [132]. Tetrahydrohyperforin has been shown to specifically target TRPC6 [92].
Hyperforin is also thought to be responsible for the induction of the cytochrome P450 enzyme CYP3A4 by binding to and activating the pregnane X receptor [134], indicating that it might have side effects and undesirable drug-to-drug interactions. Moreover, hyperforin is difficult to synthetize [135], unstable when exposed to light, and irritant to the gastrointestinal tract [136]. Such side effects might limit the use of hyperforin as a TRPC6 activator.

Resveratrol
Resveratrol (5-[(E)-2-(4-hydroxyphenyl)ethenyl]benzene-1,3-diol) ( Table 1), which is found in grapes, berries, peanuts, and other plants, is among the best-known natural compounds with antioxidant and antihypoxic activity [66,137,138]. In the pioneering study of Wang et al. [116], resveratrol is shown to be able to cross the blood-brain barrier (BBB) and exert neuroprotective activity. Injected intraperitoneally either during or shortly after the induction of cerebral ischemia, it largely prevents delayed neuronal cell death and glial cell activation in a gerbil model of transient bilateral common carotid artery occlusion [116]. Based on the observed increase in TRPC6 and CREB activity, which has been prevented by PD98059 or KN62, inhibitors of MEK and CAMKIV/ CaMKII, respectively, it is suggested that resveratrol has exerted its effects via TRPC6 activation. Since its administration has been accompanied by a marked decrease in calpain activity, resveratrol is classified as an indirect positive modulator of TRPC6 activity [66]. When given to rats for 7 days before MCAO, resveratrol significantly reduces infarct volumes and enhances neurological scores at 24 h after reperfusion [66]. Low-dose oral resveratrol treatment for three consecutive days before and after an ischemic stroke induced in rats by middle cerebral artery clipping alleviates brain damage caused by the following recurrent stroke. Resveratrol normalizes BBB function and reduces cerebral edema without affecting regional cerebral blood flow and systemic blood pressure [112]. Resveratrol preconditioning (48 h before the induction of ischemia) effectively prevents neuronal cell loss in a mouse MCAO model of stroke [113]. In a rat model of global cerebral ischemia induced by asphyxial cardiac arrest, resveratrol (48 h before the induction of ischemia) is shown to protect the CA1 region of the hippocampus similarly to ischemic preconditioning [115].
Resveratrol is actively used in AD research [139][140][141]. Neuroprotective effects of resveratrol have been shown to be associated with the activation of silent mating type information regulation 2 homolog 1 (SIRT1) and vitagenes production [142]. In a phase II trial in AD patients, resveratrol is safe and well-tolerated, but its effectiveness is contradictory [143,144].
Although generally considered as a non-toxic therapeutic agent, high doses of resveratrol inhibit CYP3A4 activity in vitro [145] and in healthy volunteers [146], thus potentially inhibiting drug metabolic clearance, increasing bioavailability and toxicity of drugs taken concomitantly [147].

Synthetic Compounds
Flufenamic Acid Flufenamic (2-[[3-(trifluoromethyl)phenyl]amino]benzoic) ( Table 1) acid (FFA) is a member of the fenamate class of nonsteroidal anti-inflammatory drugs that have limited clinical applications due to their toxicity. FFA has been found to selectively activate TRPC6, at the same time, inhibiting TRPC3, 4, 5, and 7 [81]. Direct activation of TRPC6 by FFA has been confirmed in a number of in vitro studies [148], including those in glomerular podocytes [149] and ventricular cardiomyocytes [150]. Male (but not female) embryonic mice cortical neurons, which have been pre-treated with FFA 15 min before oxygen-glucose deprivation, have shown significantly higher viability, although this effect is not linked by the authors to TRPC6 activation [118]. Similar results are obtained in a recent in vitro glutamate toxicity assay using isolated rat embryonic hippocampal neurons [119].
Multiple experimental evidences suggest that FFA is a broad spectrum ion channel modulator, with a preference for non-selective cation channels and chloride channels (reviewed here [81]). Its activity seems to be dose-dependent since FFA inhibits TRPC6 with a half maximal inhibitory concentration (IC 50 ) of 17.1 µM [151] but activates the same channel at 100 µM [148]. TRPM8 is inhibited by 100 µM FFA but is slightly activated at higher concentrations [152]. A worse situation is reported for big calcium-activated potassium channels (BK Ca) modulation since FFA activates the channel below 10 µM, inhibits the channel between 10 to 50 µM, and then activates the channel above 50 µM [153]. Such opposing effects on the same channels and a huge number of other ion channels that are modulated by FFA makes it an inappropriate drug for usage in humans.
Piperazine derivatives as the majority of TRPC6 agonists cross-react with TRPC3 and TRPC7 [83], limiting their use as specific TRPC6 modulators. Among other side effects of piperazines is their hepatotoxicity [137]; however, hepatotoxicity has not been predicted by bioinformatical analyses for a 51164 compound [21]. Gastrointestinal hemorrhage and multiple organ failure have been predicted by bioinformatical analyses; thus, there is a need to search for the lowest therapeutic dose for the 51164 compound, and most likely, there is a need to modify its structure in order to minimize the mentioned side effects [21].  (Table 1), abbreviated as FDES, is an ethanolamine derivative known to possess antihypoxic, anti-ischemic, and neuroprotective properties [88,89,120,156]. Chronic FDES administration decreases mortality and improves motor function and coordination following permanent bilateral common carotid artery ligation [120] and reduces spatial memory deficit following middle cerebral artery (MCA) occlusion/reperfusion [89] in rats. Later, FDES was demonstrated to ameliorate fore-and hindlimb motor disturbances and increase overall locomotor activity in rats with unilateral traumatic brain injury [88].
FDES is shown to potentiate neuronal SOCE into postsynaptic spines in mouse hippocampal neurons [89]. Since TRPC6 knockdown abolishes the effects of FDES on neuronal SOCE (similarly to hyperforin), TRPC6 activation is suggested to be the primary mechanism of FDES neuroprotective action. Nanomolar concentrations of FDES effectively protect mushroom dendritic spines from amyloid synaptotoxicity, stabilizing and enhancing synaptic transmission, and preserving synaptic contact density [89]. Similarly to hyperforin, FDES decreases the proportion of immature thin and stubby spines in hippocampal neurons. When administered intraperitoneally to rats subjected to MCAO for 7 consecutive days starting from 1 h after reperfusion, FDES improves short-term spatial memory retention, as observed in the Barnes maze [89].
FDES is a precursor of choline and has been shown to have nootropic and actoprotective properties [89]. We assume that FDES would cross-react with muscarinic acetylcholine receptors, causing phospholipase C activation and production of IP3 and DAG. This cross-reactivity of FDES could further enhance its TRPC6-agonistic properties, although this hypothesis remains to be experimentally proven.  Table 1) was proposed by Zhang et al. as a novel positive modulator of the TRPC6/neuronal SOCE pathway acting in a manner similar to that of 51164. Despite its effectiveness in terms of improving mushroom spine morphology and TRPC6-mediated SOCE in PS1-KI and APP-KI hippocampal neurons and rescuing long-term potentiation in the APP-KI mouse model of AD, NSN21778 is found to have a rather poor pharmacokinetic profile and a low penetration of the blood-brain barrier [14].

Other TRPC6 Agonists
Several compounds described below have been confirmed to activate the TRPC6 channel using in vitro assays. However, to the best of our knowledge, their specific neuroprotective properties remain unexplored. Given their ability to interact with TRPC6, these compounds can be considered as potential neuroprotective agents. (EC 50 = 0.08 and 0.44 µM, respectively), acting directly and independent of protein lipase C signaling from the extracellular side [157]. OptoBI-1, an azobenzene moiety-containing photochromic derivative of GSK1702934A, is found to activate TRPC6 as well as TRPC3 and 7, although having a slightly higher affinity for TRPC3. Light treatment of cultured murine hippocampal neurons with OptoBI-1 suppresses action potential firing elicited by repetitive depolarizing current injections [84].
It can be assumed that the mechanism of action of C20 involves TRPC6 sensitization and not with its activation per se; that is, C20 allows TRPC6 to be activated at a low basal concentration of DAG [85].

TRPC6 Inhibitors
AD seems to be a multifactorial disease with different pathogenic cascades occurring in different patients. In terms of TRPC6 channel dysfunction, there are forms of fAD, which demonstrate TRPC6 hyperfunction [24]. For those patients, TRPC6 inhibitors might be used in order to normalize intracellular Ca 2+ homeostasis.
The pathogenesis of cerebral ischemia is not fully understood. Similarly to AD, cerebral ischemia involves several different pathological cascades. There are studies reporting that TRPC6-mediated Ca 2+ and Na+ influx facilitates NMDAR activation and exacerbates excitotoxicity, while TRPC6 deletion attenuates neuronal damage and death following focal cerebral ischemia [25]. In such cases, where excessive TRPC6 activity seems to be present, the use of TRPC6 inhibitors might be beneficial.
In mice with closed-head traumatic brain injury, larixyl acetate, a naturally occurring diterpene, ameliorates endothelial dysfunction [74], which is closely associated with cerebral ischemia as well [161]. In isolated mouse lungs, larixyl acetate prevents the development of the acute hypoxic ventilatory response [72]. TRPC6 inhibition is now considered to be the primary mechanism of larixyl's neuroprotective action [72,74].

Conclusions
The present review summarizes current knowledge on the role of TRPC6 channels in the development of two neurological disorders: Alzheimer's disease and cerebral ischemia. Cerebral ischemia serves as a risk factor for AD, and vice versa. It is becoming evident that both diseases can be caused by either upregulation or downregulation of TRPC6 channels. Thus, understanding the nature of the disruption of this molecular pathway in each particular patient is extremely important for appropriate drug prescription.
TRPC6 is structurally similar to TRPC3 and 7, and therefore the majority of compounds do not act selectively on either of these three isoforms, making it difficult to develop selective TRPC6 pharmacological modulators. Moreover, a number of TRPC6 channel modulators are cross-reactive to other cellular targets, thus limiting their pharmacological potential. Toxicity profiles of some known TRPC6 modulators (such as FFA) require further structural optimization. The presence of a wide range of different chemical substances that are known to interact with TRPC6 channels as well as the availability of cryo-electron microscopy structures of TRPC6 and 3 [169] may allow determining the pharmacophore in order to design selective TRPC6 activators and inhibitors in the future. In turn, the development of selective TRPC6 channel modulators could help slow down the progression of AD, cerebral ischemia, and, most likely, other TRPC6-dependent diseases.