9-Phenanthrol, also called 9-hydroxyphenanthrene or phenanthrene-9-ol in the IUPAC (International Union of Pure and Applied Chemistry) nomenclature, has been known for more than a century (Pschorr and Schroter, 1902; Moriconi et al., 1959). It is a benzo[c]quinolizinium derivative composed of three fused benzene rings (Figure 1). Despite primary results indicating fungitoxicity of the molecule (Rich and Horsfall, 1954), it was, until recently, only used as a biomarker of anthropogenic compound pollution. Indeed, 9-phenanthrol may originate from other hydroxylated polycyclic aromatic hydrocarbons (PAH). Those molecules come from incomplete combustion of organic matter, resulting, in part, from human activities. PAHs found in polluted soil or water can be ingested by organisms like fish. One of these, the phenanthrene, is partly metabolized to form molecules such as 9-phenanthrol, which can become concentrated in tissues because of its hydrophobicity (Escartin and Porte, 1999; Koenig et al., 2013). As PAHs such as phenanthrene are highly toxic, 9-phenanthrol may be applicable for depollution processes because it has a lower toxicity and can be produced after phenanthrene biodegradation by proteobacteria (Feng et al., 2012).
Far from these environmental concerns, we observed in 2008 that 9-phenanthrol inhibits the transient receptor potential melastatin (TRPM) 4 channel (Grand et al., 2008). TRPM4 forms a Ca2+-activated non-selective cation (NSCCa) channel widely expressed in tissues from several mammalian species including humans (Launay et al., 2002). The physiological roles for the TRPM4 channel were difficult to identify until the development of knockout mice and appropriate pharmacological tools (Guinamard et al., 2011). Regarding pharmacology, it has been confirmed that 9-phenanthrol specifically targets the TRPM4 channel (Table 1) and its use in a variety of biological preparations has unmasked the contributions of the TRPM4 channel in physiological processes.
|Current||Model||Effect||Concentration in mol·L−1||References|
|TRPM4||HEK-293||Inhibition||IC50 = 1.7 × 10−5||Grand et al., 2008|
|TRPM5||HEK-293||None||10−4||Grand et al., 2008|
|TRPC3||HEK-293||None||3 × 10−5||Gonzales et al., 2010b|
|TRPC6||HEK-293||None||3 × 10−5||Gonzales et al., 2010b|
|CFTR||CHO||None||2.5 × 10−4||Grand et al., 2008|
Here we review 9-phenanthrol as a TRPM4 channel inhibitor. We summarize the identification of 9-phenanthrol and document its specificity among ion channels. Further, we review physiological processes modulated by 9-phenanthrol and recommend appropriate applications of the drug, while acknowledging their caveats and limitations.
Looking for TRPM4 channel pharmacological inhibitors
The TRPM4 channel belongs to the transient receptor potential (TRP) protein family, whose members form non-selective cation channels (Alexander et al., 2013). It shares with its closest relative, the TRPM5 channel, a lack of selectivity for monovalent cations (i.e. equal permeability for Na+ and K+), and while these channels are Ca2+ impermeable their activation mechanism is simultaneously sensitive to internal Ca2+ concentration (Launay et al., 2002; Guinamard et al., 2011). In addition, both channels have a higher activity at depolarized voltages. Such NSCCa currents were recognized in native preparations from a variety of tissues (Teulon, 2000). Unmasking their physiological roles has depended on pharmacological tools. Several pharmacological agents inhibit TRPM4 channels and NSCCa currents more generally (Figure 1). The most commonly used NSCCa channel inhibitor is the non-steroidal anti-inflammatory drug flufenamic acid that inhibits both TRPM4 and TRPM5 channels (Ullrich et al., 2005). This molecule is advantageous because it rapidly affects channel activity and is reversible. However, it has a large spectrum of targets, particularly among ion channels (Guinamard et al., 2013). The bitter compound quinine and its stereoisomer quinidine similarly inhibit TRPM4 and TRPM5 channels (Talavera et al., 2008), but once again are not specific (White, 2007). Spermine is another NSCCa antagonist without specificity for underlying channel type (Nilius et al., 2004). The sulfonylurea glibenclamide was shown to inhibit native NSCCa currents and the TRPM4 channel (Demion et al., 2007), but it is also known to interact with other channels, mainly those belonging to the ATP binding cassette (ABC) protein family and the ATP-dependent K+ channel, which is a multimer of inwardly rectifying K+ channel subunits and the sulfonylurea receptor (SUR; Alexander et al., 2013). Note that the co-assembly of the TRPM4 channel with the SUR1, such as occurs in acute CNS injuries, potentiates its sensitivity to glibenclamide (Woo et al., 2013b). TRPM4-like currents are also inhibited by the closely related chloride channel blockers diphenylamine-2-carboxylic acid, 3′,5′-dichlorodiphenylamine-2-carboxylic acid and 5-nitro-2-(3-phenylpropyl-amino)-benzoic acid in a variety of tissues (Gögelein and Pfannmüller, 1989; Chraïbi et al., 1994; Teulon, 2000). Even though all of these molecules target the TRPM4 channel, none is a specific antagonist. Moreover, except for the presence of phenol rings, they share few (if any) chemical determinants (Figure 1).
Looking for new candidates to inhibit the TRPM4 channel, we were intrigued by several similarities between TRPM4 and ABC proteins. Both are sensitive to ATP and hold ATP binding sites within their amino acid sequence (Ullrich et al., 2005; Frelet and Klein, 2006). In addition, as referenced earlier, the antidiabetic sulfonylurea glibenclamide modulates the TRPM4 channel (Demion et al., 2007) as well as ABC proteins including SURs and the cystic fibrosis transmembrane conductance regulator (CFTR) chloride channel (Sheppard and Welsh, 1992; Frelet and Klein, 2006). We hypothesized that other modulators of ABC proteins might also affect the TRPM4 channel. Indeed this proved correct for the CFTR activator 5-butyl-7-chloro-6-hydroxybenzo[c]-quinolizinium chloride (MPB-104) (Figure 1), which inhibits the TRPM4 channel with an IC50 of ∼2 × 10−5 mol·L−1 (Grand et al., 2008). However, MPB-104 is not specific for the TRPM4 channel, as it also modulates the CFTR channel. Therefore, in searching for other related compounds, we tested 9-phenanthrol (Figure 1), which lacks several components necessary for CFTR channel activation. As expected, 9-phenanthrol did not modulate the CFTR channel, yet it inhibited TRPM4 channels heterologously expressed in HEK-293 cells (Grand et al., 2008).
Inhibition of recombinant TRPM4 channel by 9-phenanthrol
As mentioned earlier, and shown in concentration–response curves (Figure 2), 9-phenanthrol inhibited human TRPM4 channel activity in patch recordings from transfected HEK-293 cells (Grand et al., 2008). In the conventional whole-cell configuration, superfusion of 9-phenanthrol in the bath inhibited the TRPM4-mediated current with an IC50 of 1.7 × 10−5 mol·L−1 (Table 1). The inhibition was also observed in nystatin perforated whole-cell membrane currents from TRPM4-transfected COS-7 cells (Woo et al., 2013a). Similar results were observed in the inside-out configuration (Grand et al., 2008; Woo et al., 2013a). These data indicate that the molecule may interact with the channel on both sides, or more likely, that the molecule is able to cross the membrane because of its hydrophobicity. The interaction site of 9-phenanthrol within the channel structure is not known. However, the glycosylation state of the protein is not involved. Indeed, the N988Q substitution, at an extracellular location near the pore-forming loop between the transmembrane segments 5 and 6, which results in the disappearance of the N-glycosylated forms of the protein, did not modify its inhibition by 9-phenanthrol (Woo et al., 2013a). Because channel inhibition occurs within seconds in the inside-out configuration, but within a minute in the whole-cell configuration, a reasonable starting assumption is that 9-phenanthrol interacts with the TRPM4 channel intracellularly.
The Hill coefficient of the concentration–response curve is close to 1, indicating no cooperation in 9-phenanthrol interactions with the channel. The concentration–response curves performed in the inside-out configuration at positive and negative voltages showed no evidence of voltage-dependent inhibition (Grand et al., 2008).
Whether recorded in inside-out patches or the whole-cell configuration, TRPM4 channel inhibition by 9-phenanthrol is reversible, even though complete recovery can take several minutes after channel exposure to high concentrations (∼10−4 mol·L−1).
Inhibition of endogenous TRPM4 currents by 9-phenanthrol
After its identification as a recombinant TRPM4 channel inhibitor, the effects of 9-phenanthrol were tested on endogenous TRPM4-like currents from several native tissues (Table 2).
|Current||Model||Effect||Concentration in mol.L−1||References|
|TRPM4-like||HEK-293||Inhibition 80%||10−4||Amarouch et al., 2013|
|TRPM4-like transient inward||Rat cerebral arteries smooth muscle cells||Inhibition||IC50 = 10−5||Gonzales et al., 2010b|
|TRPM4-like transient inward||Monkey colonic smooth muscle cells||Inhibition 80%||10−4||Dwyer et al., 2011|
|TRPM4-like transient inward||Rat, guinea pig DSM cells||Inhibition 40%||3 × 10−5||Parajuli et al., 2013; Smith et al., 2013a,b|
|TRPM4-like||Mouse atrial cardiomyocytes||Inhibition 80%||10−5||Simard et al., 2013|
|TRPM4-like||Human adenocarcinoma AGS cells||Inhibition 90%||3 × 10−5||Kim et al., 2012|
|TRPM4-like||Human adenocarcinoma MKN-45 cells||Inhibition 90%||3 × 10−5||Kim et al., 2012|
|TRPM4-like depolarization-induced slow||Mouse Purkinje cells from cerebellar slices||Inhibition 100%||10−4||Kim et al., 2013|
|TRPM5-like||Mouse tongue taste cells||None||10−4||Liu et al., 2011|
|TRPM7-like||Human adenocarcinoma AGS cells||None||3 × 10−5||Kim et al., 2012|
|TRPM7-like||Human adenocarcinoma MKN-45 cells||None||3 × 10−5||Kim et al., 2012|
|TRPM7-like||Mouse intestinal cells of Cajal||None||3 × 10−5||Kim et al., 2011|
|Voltage-dependent K+||Rat cerebral arteries smooth muscle cells||None||3 × 10−5||Gonzales et al., 2010b|
|BKCa||Rat cerebral arteries smooth muscle cells||None||3 × 10−5||Gonzales et al., 2010b|
|KIR||Rat cerebral arteries smooth muscle cells||None||3 × 10−5||Gonzales et al., 2010b|
|Kv||Rat cerebral arteries smooth muscle cells||None||3 × 10−5||Gonzales et al., 2010b|
|ICa,L||Rat cerebral arteries smooth muscle cells||None||3 × 10−5||Gonzales et al., 2010b|
|ICa,L||Rat ventricular cardiomyocyte||None Inhibition 50%|| |
|Simard et al., 2012a|
|IK||Rat ventricular cardiomyocyte||None Inhibition 40%|| |
|Simard et al., 2012a|
The TRPM4 channel is highly expressed in epithelia such as kidney. HEK-293 cells obtained from this type of tissue weakly express an endogenous TRPM4-like current, which is inhibited by 80% by 10−4 mol·L−1 9-phenanthrol in whole-cell recordings (Amarouch et al., 2013).
The TRPM4 channel is also expressed in various smooth muscles cells (Earley, 2013). A TRPM4-like transient inward current is attenuated by 9-phenanthrol with an IC50 of 1 × 10−5 mol·L−1 in perforated patches, and totally inhibited by 3 × 10−5 mol·L−1 in whole-cell recordings from rat cerebral artery smooth muscle cells (Gonzales et al., 2010b; Gonzales and Earley, 2012). 9-Phenanthrol, at 3 × 10−5 mol·L−1, inhibits 40% of the same current during perforated-patch recordings from guinea pig and rat detrusor smooth muscle (DSM) cells (Parajuli et al., 2013; Smith et al., 2013a,b). In addition, 10−4 mol·L−1 9-phenanthrol attenuates TRPM4-like currents in inside-out patch recordings from monkey colonic smooth muscle cells (Dwyer et al., 2011).
We have reported the functional expression of TRPM4 channels in cardiomyocytes, including human cardiac tissue (Guinamard et al., 2004; Simard et al., 2013). Endogenous TRPM4 channel activity in inside-out patch recordings from mouse isolated atrial cardiomyocytes was inhibited by 80% in response to 10−5 mol·L−1 9-phenanthrol (Simard et al., 2013).
Purkinje cells from mouse cerebellum express a calcium-dependent depolarization-induced slow current whose properties match those of TRPM4 and TRPM5 channels (Kim et al., 2013). The current is smaller in Trpm4 null mice, but is not affected in Trpm5 null mice, which indicates that the calcium-dependent inward current probably corresponds to a TRPM4 current. This current is totally abolished by 10−5 mol·L−1 9-phenanthrol (Kim et al., 2013).
The effects of 9-phenanthrol were also tested on TRPM4-like endogenous currents in two human gastric adenocarcinoma cell lines, AGS and MKN-45. In both cell types, 3 × 10−5 mol·L−1 9-phenanthrol reduced the current by 90% (Kim et al., 2012).
9-phenanthrol effects on other ion channels
To evaluate specificity of 9-phenanthrol for the TRPM4 channel, its effects were also tested on other recombinant ion channels as well as native currents (Tables 1 and 2). Concentrations at or exceeding 3 × 10−5 mol·L−1 were used in all cases. Note that 9-phenanthrol concentrations greater than 10−4 mol·L−1 may be difficult to test precisely because the molecule partly precipitates at such levels even if dissolved in DMSO.
9-Phenanthrol was tested on several members of the TRP channel family responsible for a large variety of non-selective cationic currents. First of all, it was tested on the TRPM5 channel, the closest TRPM4 relative within the family. A recombinant TRPM5 current was not affected by 10−4 mol·L−1 9-phenanthrol in HEK-293 cells, which is an encouraging sign regarding its high selectivity for the TRPM4 channel (Grand et al., 2008). This lack of effect on the TRPM5 channel was confirmed in a preparation of isolated lingual taste cells from mouse, which exhibit native NSCCa activated by linoleic acid. This NSCCa current was abolished by Trpm5 gene disruption or application of the TRPM5 channel-specific inhibitor triphenylphosphine oxide, but not by 10−4 mol·L−1 9-phenanthrol (Liu et al., 2011). These data substantiate the effectiveness of 9-phenanthrol at discriminating TRPM4 from TRPM5 currents in native cells, which is challenging because TRPM4 and TRPM5 channels share biophysical and regulatory properties as well as sensitivity to intracellular Ca2+ yet lack of Ca2+ permeability, as mentioned earlier (Guinamard et al., 2011). It is interesting to note that the main difference between TRPM4 and TRPM5 channels is ATP sensitivity; TRPM4 is blocked by internal ATP, while TRPM5 is not (Ullrich et al., 2005). This discrepancy is probably due to the presence of only one ATP binding site within the TRPM5 channel that appears to be inaccessible, while TRPM4 channel possesses four ATP binding sites (Ullrich et al., 2005). This disparity between TRPM4 and TRPM5 might provide some insights into the mechanism(s) involved in selective TRPM4 inhibition by 9-phenanthrol.
Thirty micromolar 9-phenanthrol did not modulate transient receptor potential canonical (TRPC) 3 or TRPC6 currents after recombinant expression in HEK-293 cells (Gonzales et al., 2010b). This lack of effect is important because TRPC3 and TRPC6 channels often co-express with the TRPM4 channel, particularly in cardiac and smooth muscle preparations.
In addition to non-selective cation channels, 9-phenanthrol was tested on two recombinant chloride channels. Firstly, the effects of 9-phenanthrol were assessed via radioactive iodide efflux on recombinant CFTR Cl– channel in CHO cells. Even at 2.5 × 10−4 mol·L−1, 9-phenanthrol had no effect (Grand et al., 2008). Once again, TRPM4 and CFTR may be expressed in the same tissue. Secondly, an inhibitory effect of 9-phenanthrol on recombinant TMEM16A channels expressed in HEK-293 was reported in a recent abstract (Burris et al., 2013), but the results have not yet been scrutinized. However, pharmacological parity with TRPM4 is conceivable, because the TMEM16A channel also supports a Ca2+-activated chloride current (Terashima et al., 2013) and is widely co-expressed with the TRPM4 channel in smooth muscle cells (Bulley and Jaggar, 2013).
9-Phenanthrol was tested on several native ionic currents with properties different from TRPM4 current. A TRPM7-like endogenous current from human gastric adenocarcinoma cell lines AGS and MKN-45 was found to be insensitive to 3 × 10−5 mol·L−1 9-phenanthrol (Kim et al., 2012) as was a similar current from mouse interstitial cells of Cajal (Kim et al., 2011).
In cerebral artery smooth muscle cells from rats, neither voltage-gated K+ currents (KV), large conductance Ca2+-activated K+ currents (BKCa), inward rectifyer K+ currents (KIR), nor L-type Ca2+ currents (ICa,L) were modulated by 3 × 10−5 mol·L−1 9-phenanthrol (Gonzales et al., 2010b). However, it is more likely that higher concentrations of the drug are needed to have an effect on these channels. Indeed, as reported in mouse isolated ventricular cardiomyocytes, 10−5 mol·L−1 9-phenanthrol has no effect on any of these currents, but 10−4 mol·L−1 9-phenanthrol partially inhibits L-type Ca2+ currents and the delayed outward rectifier K+ current (IK) (Simard et al., 2012a).
In dopamine neurons of the substantia nigra from mice that expressed both TRPM4 and hyperpolarization and cyclic nucleotide-gated (HCN) channels (which gives rise to hyperpolarization-activated mixed cation current Ih in neurons), 10−4 mol·L−1 9-phenanthrol and the HCN channel blocker ZD7288 have distinct effects on rhythmic neuronal activity (Mrejeru et al., 2011). While indirect, these data suggest that 9-phenanthrol, even at high concentrations, does not affect HCN channels.
To be exhaustive, even if it is not in the field of ion channels, we have to mention the results from a study indicating a potent inhibitory effect of 9-phenanthrol, with other phenanthrene derivatives, on bovine heart cyclic-AMP-dependent PKA catalytic subunits (Wang et al., 1994). However, the experiments were only conducted in vitro by biochemical assays of the reaction medium, and this inhibitory effect on PKA catalytic subunits has not been demonstrated in model cells of native tissues. To the best of our knowledge, this biochemical experiment from the 1990s has never been repeated or confirmed. On the contrary, the effects of 9-phenanthrol in cardiac preparations were not precluded by the simultaneous application of the PKA inhibitor H-89, which argues against an effect of 9-phenanthrol via PKA inhibition (Simard et al., 2012a; 2013).
Altogether, these results indicate that 9-phenanthrol is a selective TRPM4 channel inhibitor. Nonetheless, as shown in cardiomyocytes, some side effects may be observed at concentrations at or exceeding 10−4 mol·L−1.
Smooth muscle cell contraction
Identifying 9-phenanthrol as a TRPM4 channel-specific inhibitor has provided the opportunity to distinguish TRPM4 currents from currents evoked through TRPM5 and it has provided a powerful tool to reveal the role(s) of TRPM4 channels in physiological processes. Through its modulation of TRPM4 channels, 9-phenanthrol affects smooth muscle behaviour, spontaneous firing in neurons and cardiomyocytes, as well as protecting against cardiac or vascular injuries (see later and Table 3).
|Physiological processes||Model||Effect||Concentration in mol·L−1||References|
|Artery constriction||Rat cerebral arteries||Inhibition||IC50 = 2.6 × 10−6||Gonzales et al., 2010b|
|Smooth muscle contraction||Rat, guinea pig isolated detrusor muscle||Inhibition||IC50 = 3 × 10−6||Parajuli et al., 2013; Smith et al., 2013a,b|
|Cardiac beating rate||Mouse, rat, rabbit isolated right atrium||Reduction||EC50 = 7.8 × 10−6||Hof et al., 2013|
|Cardiac atrial action potential duration||Mouse isolated atrium||Reduction||EC50 = 2.2 × 10−5||Simard et al., 2013|
|Hypoxia-reoxygenation induced cardiac arrhythmias||Mouse isolated ventricle|| |
|Simard et al., 2012a|
|Hypoxia-reoxygenation preconditioning||Rat isolated whole heart||Cardioprotection||3 × 10−5||Wang et al., 2013|
|LPS-induced cell death||Endothelial cells HUVECs||Reduction 50%||10−6||Becerra et al., 2011|
|Cell viability||Adrenocarcinoma cells AGS and MKN-45||None||3 × 10−5||Kim et al., 2012|
|Capillary structure formation||Endothelial cells HUVECs||Stimulation||5 × 10−6||Loh et al., 2013|
|Neuronal bursting activity||Mice coronal midbrain slice||Reduction||10−4||Mrejeru et al., 2011|
|Neuronal firing activity||Mice olfactory bulb slice||Reduction 50%||10−4||Shpak et al., 2012|
|Neuronal firing activity||Rat inhibitory prepositus hypoglossy nucleus neurons||Reduction 60%||10−4||Saito and Yanagawa, 2013|
A preliminary series of studies focused on arterial smooth muscles that are known to express TRPM4 (Earley, 2013). Arteries respond to pressure overload by a vasoconstriction in order to autoregulate blood flow. In an experimental model of isolated cerebral arteries from rat, 9-phenanthrol dilated arteries that were previously pressurized (Gonzales et al., 2010b). This occurred with an EC50 of 2.6 × 10−6 mol·L−1 indicating a high sensitivity to 9-phenanthrol. Interestingly, down-regulation of Trpm4 expression using antisense oligodeoxynucleoides also affected vasoconstriction in this preparation, which further implicates the TRPM4 channel in this tissue (Earley et al., 2004). Arterial contraction requires L-type Ca2+ current activation. Thus, control of membrane potential is a critical factor that influences the phenomenon. Using intracellular microelectrode recordings, 3 × 10−5 mol·L−1 9-phenanthrol was shown to hyperpolarize vascular smooth muscle cells in pressurized cerebral arteries. The mechanism of action probably involves the inhibition of a TRPM4-depolarizing current via a mechanism explained later (Gonzales et al., 2010b). 9-Phenanthrol was also shown to inhibit a depolarizing transient inward cationic current in the same preparation; this effect was prevented if the cells were treated with Trpm4 small interfering RNA (Gonzales et al., 2010a; Gonzales and Earley, 2012). The putative mechanism for vasoconstriction of cerebral arteries in response to pressure-dilatation involves a mechanosensitive membrane protein coupled to PLC, whose activation produces inositol 1,4,5-trisphosphate (IP3). IP3 binds to its receptor on the sarcoplasmic reticulum to induce calcium release and thus TRPM4 channel activation ensues. The TRPM4 channel is an inward charge carrier (with a reversal potential ∼0 mV) that induces cell depolarization in favour of L-type Ca2+ current activation, producing inward Ca2+ currents and cell contraction (Earley, 2013). The phenomenon also depends on PKC to promote translocation of TRPM4 channels to the plasma membrane. Small interfering Trpm4 RNA or treatment with 2 × 10−5 mol·L−1 9-phenanthrol similarly abolish the PKC-induced cerebral artery vasoconstriction (Crnich et al., 2010). Cerebral artery pressure-induced constriction involves sphingosine-kinase-1 translocation from the cytosol to the plasma membrane, which produces sphingosine-1-phosphate following membrane depolarization. In an experimental model of rabbit posterior cerebral arteries, 5 × 10−6 mol·L−1 9-phenanthrol attenuated this pressure-induced translocation (Lim et al., 2012).
Among smooth muscles that express the TRPM4 channel, an effect of 9-phenanthrol was also demonstrated in DSM that controls urinary bladder contraction. The expression level of Trpm4 mRNA is 2.6-fold greater in DSM cells than in cerebral artery myocytes (Parajuli et al., 2013). The intracellular signalling that leads to DSM contraction remains to be determined, but it may resemble the one described earlier for cerebral arteries. Variations in the intracellular Ca2+ concentrations activate a transient inward cationic current that is consistent with an NSCCa. This inward current causes cell membrane depolarization, inducing L-type Ca2+ current activation, which leads to significant accumulation of intracellular Ca2+ that then stimulates contraction (Smith et al., 2013b). 9-Phenanthrol affects several of these parameters in DSM. The transient inward cationic current with TRPM4-like properties was inhibited by 3 × 10−5 mol·L−1 9-phenanthrol in both rat and guinea pig freshly isolated DSM cells (Smith et al., 2013a,b). The same concentration of 9-phenanthrol also reduced the baseline intracellular Ca2+ level measured by Ca2+-sensitive fluorescent dyes (Smith et al., 2013a). Because this effect on baseline Ca2+ is reduced by nifedipine, but not by carbachol, it is attributed to a decrease in Ca2+ influx via L-type channels subsequent to a decrease in channel activation due to the lack of membrane depolarization (Smith et al., 2013a). Contractions of the DSM are also affected by 9-phenanthrol. Whereas pharmacological tools such as carbachol increase basal tension and spontaneous contractions, 9-phenanthrol reduces the force, amplitude and frequency of carbachol-induced contractions from rat isolated DSM strips with an IC50 ranging from 1.7 to 3.3 × 10−6 mol·L−1 (Smith et al., 2013b). DSMs are activated by parasympathetic inputs that induce bladder voiding behaviour. This neurogenic control can be mimicked by electrical field stimulation, which induces DSM strips to contract. During such recordings, 9-phenanthrol attenuates the amplitude of the contraction and muscle force with an IC50 of 2.3 and 2.8 × 10−6 mol·L−1 respectively (Smith et al., 2013b). These results obtained using 9-phenanthrol indicate that the TRPM4 channel is a major contributor to excitation–contraction coupling in the DSM.
Membrane current regulation has also been demonstrated in smooth muscle cells of the colon. The resting membrane potential is generally less negative than the K+ equilibrium potential as a result of a spontaneous inward current with non-selective cation permeability, which shares single-channel properties with TRPM4 (Dwyer et al., 2011). In freshly dispersed colonic smooth muscle cells from monkeys, this TRPM4-like current is inhibited by 10−4 mol·L−1 9-phenanthrol. No further experiments were conducted using 9-phenanthrol in this preparation, even though the authors demonstrated that La3+ and Gd3+ both modulated the baseline membrane potential, presumably by attenuating the same TRPM4-like current (Dwyer et al., 2011).
Modulation of heart function
Investigating the role of the TRPM4 channel in cardiac physiology and pathophysiology is of great importance, firstly, because the heart is among the foremost TRPM4-expressing tissues (Launay et al., 2002) and secondly, because the only known channelopathy related to TRPM4 dysfunction in humans perturbs cardiac electrical activity: for example, congenital conduction block and Brugada syndrome (Kruse et al., 2009; Liu et al., 2010; 2013). Endogenous TRPM4-like currents have been thoroughly characterized in cardiac preparations (see review, Guinamard et al., 2011), but demonstrating their specific roles in cardiac tissues has only become possible following the discovery of 9-phenanthrol as a TRPM4 channel inhibitor. 9-Phenanthrol revealed the roles of the TRPM4 channel in both basal cardiac activity and in pathological-like conditions.
Cardiac rhythm initiates in the specific pacemaker tissue dubbed the sinoatrial node at the flank of the right atrium. Intracellular microelectrodes can record spontaneous activity of isolated right atrium preparations suspended in a superfusion bath. The rate of spontaneous action potentials from mouse and rat preparations is reversibly reduced by 9-phenanthrol with an IC50 of 7.8 × 10−6 mol·L−1 (Figure 3; Hof et al., 2013). A particular characteristic of sinoatrial node cells is the presence of a slow diastolic depolarization (absent in other cardiac myocytes) that drives the membrane potential over the threshold for Ca2+ current activation, and is responsible for the action potential upstroke. While a major part of this slow diastolic depolarization is due to the hyperpolarization-activated funny current (If) (Monfredi et al., 2010), which is equivalent to Ih in neurons, an additional component of the slow diastolic depolarization is due to Ca2+-activated depolarizing currents triggered by cytoplasmic Ca2+ oscillations. In the rabbit isolated sinoatrial node, 10−5 mol·L−1 9-phenanthrol reduces both the beating rate and the slope of the slow diastolic depolarization, with no other effects on sinoatrial node action potentials. This 9-phenanthrol-induced reduction in sinoatrial node beating rate is absent in Trpm4−/− mice (Hof et al., 2013) and TRPM4 currents were characterized in isolated sinoatrial node cells from wild-type mice (Demion et al., 2007), which clearly demonstrates that the TRPM4 channel is involved in regulating the rate of beating. Interestingly, the ability of 9-phenanthrol to reduce the beating rate in mouse and rat isolated right atrium is more pronounced at low frequencies (Hof et al., 2013), which suggests that the TRPM4 channel could protect against bradycardia and, this might be modulated by 9-phenanthrol.
The results summarized earlier indicate that 9-phenanthrol acts in sinoatrial node cells to directly slow the diastolic depolarization through inhibiton of TRPM4 channels. However, a study conducted on the HL-1 cell line derived from mouse atrial cells revealed that 9-phenanthrol also modulates cytoplasmic calcium oscillations (Burt et al., 2013). HL-1 cells are derived from atrial cardiomyocytes, but not specifically sinoatrial node cells, and they acquire properties similar to those of pacemaker cells and develop free beating activity, with intracellular calcium oscillations being present in about 40% of the cells (Wondergem et al., 2010). Fluorescent Ca2+ recordings using Fura-2 revealed that 10−5 mol·L−1 9-phenanthrol abolishes these Ca2+ oscillations leading to a transient increase in cytoplasmic Ca2+, attributed to the release of Ca2+ from an intracellular pool different from the sarcoplasmic reticulum and thus, is most likely to be mitochondrial Ca2+ (Burt et al., 2013). These Ca2+ oscillations could participate in the spontaneous activity, since the application of 10−5 mol·L−1 9-phenanthrol in cell-attached recordings from HL-1 cells abolished the depolarizing inward ionic currents that contribute to the slow diastolic depolarization (Burt et al., 2013).
In addition to its effect on spontaneous beating, 9-phenanthrol also modulates action potentials evoked in mouse atrial cardiomyocytes; recorded using intracellular microelectrodes. 9-Phenanthrol reduced action potential duration, with an IC50 of 2.2 × 10−5 mol·L−1, without affecting other action potential parameters or the resting membrane potential (Simard et al., 2013). This effect of 9-phenanthrol on action potential duration is probably as a results of its inhibitory effect on TRPM4 channels because (i) a TRPM4 current can be recorded from mouse isolated atrial cells and this endogenous current is inhibited by 9-phenanthrol (Simard et al., 2013); (ii) the effect of 9-phenanthrol is strongly reduced in ventricular tissue, which is known to express lower levels of the TRPM4 channel compared with the atrium (Guinamard et al., 2006; Liu et al., 2010; Simard et al., 2013); and (iii) the 9-phenanthrol effect is completely absent in atria from Trpm4−/− mice (Simard et al., 2013). While these experiments were done in mice, their relevance probably extends to humans as human isolated atrial cardiomyocytes have been reported to express functional TRPM4 channels (Guinamard et al., 2004).
9-Phenanthrol does not only modulate basal cardiac activity, but has also been shown to prevent cardiac dysfunction induced by ischaemia-reperfusion episodes. Ischaemia, such as that observed in coronary occlusion, is a major source of cardiac damage. It produces electrical perturbations that lead to arrhythmias, it decreases contractility and it induces apoptosis. Although it corrects oxygen defects, reperfusion paradoxically exacerbates ischaemia-related perturbations, mainly through disturbance of internal Ca2+ homeostasis (Murphy and Steenbergen, 2008). As the TRPM4 channel is an ATP-sensitive Ca2+-activated channel and because it is highly expressed in heart, it was thought to be involved in ischaemia-related perturbations. Therefore, the TRPM4 channel inhibitor 9-phenanthrol was tested for possible cardioprotective effects. In an initial study, the effect of 9-phenanthrol against arrhythmias was investigated in a mouse isolated ventricle subjected to hypoxia-reoxygenation protocols, in order to mimic ischaemia-reperfusion conditions (Simard et al., 2012a). During hypoxia, as well as during reoxygenation, arrhythmias were observed in the form of early after depolarizations (EADs). Superfusion of 10−5 mol·L−1 9-phenanthrol during reoxygenation reduced the occurrence of EADs by 60% while 10−4 mol·L−1 totally abolished the EADs (Simard et al., 2012a). Figure 4 shows an example of such anti-arrhythmic effects in ventricular myocytes. As in the atrial tissue, 9-phenanthrol does not regulate the resting membrane potential at the ventricular level. In this model, it was hypothesized that EADs were formed in the conductive tissue such as Purkinje fibres, where TRPM4 channels are highly expressed (Liu et al., 2010), and were transmitted to the ventricular cells. In a second study, the cardioprotective effect of 9-phenanthrol was tested in a rat isolated whole heart perfused using a Langendorff apparatus (Wang et al., 2013). Ischaemia-reperfusion was mimicked by switching off and on the perfusion of physiological solution in the coronary vessels. The preconditioning effect of 2 × 10−5 mol·L−1 9-phenanthrol was tested by applying the drug 3 min before a 30 min ischaemia episode, which was then followed by reperfusion. 9-Phenanthrol prevented the decrease in contractile function, and the occurrence of ventricular fibrillation or tachycardia. In addition, as a proof of prophylaxis against cellular damage, 9-phenanthrol reduced lactate dehydrogenase activity and the size of the infarcted area (Wang et al., 2013). As this effect of 9-phenanthrol was not prevented by addition of the ATP-sensitive K+ channel inhibitor 5-hydroxydecanoate, its effects were attributed to the TRPM4 channel inhibition (Wang et al., 2013). These two complementary studies revealed 9-phenanthrol to be a new and potentially powerful cardioprotective agent against ischaemia-reperfusion injuries. The rapid kinetics of 9-phenanthrol's antiarrhythmic effects is consistent with the direct involvement of the TRPM4 channel in cardiac action potentials (Simard et al., 2012a). Inversely, the slow kinetics of 9-phenanthrol in the preconditioning model suggests an additional subcellular process subsequent to TRPM4 channel inhibition (Wang et al., 2013).
Prevention of cell death
The TRPM4 channel is involved to some extent in cell death (Schattling et al., 2012; Simard et al., 2012b). Oxidative stress induces HeLa cell death by necrosis, but not apoptosis, through TRPM4 channel activation (Simon et al., 2010). The effect of 9-phenanthrol on this process was investigated in a model of HUVECs, which endogenously express TRPM4 channels, but not TRPM5 channels (Becerra et al., 2011). In those cells, LPS induces cell death by promoting Na+ overload leading to cell depolarization and cell volume increase. LPS-induced cell death was estimated by lactate dehydrogenase release, thiazolyl blue tetrazolium bromide (MTT) salt colorimetric assay and trypan blue. Treatment with 10−6 mol·L−1 9-phenanthrol reduced cell death by half (Becerra et al., 2011). These protective effects of 9-phenanthrol were reproduced by glibenclamide (Becerra et al., 2011), which is also known to inhibit the TRPM4 channel (Demion et al., 2007). The LPS-induced Na+-current is inhibited by glibenclamide, as expected, but the effect of 9-phenanthrol on this current was not tested (Becerra et al., 2011). Note that in control conditions 9-phenanthrol at concentrations ranging from 10−7 to 5 × 10−6 mol·L−1 does not induce cell death in HUVECs, although higher concentrations may have deleterious effects (Loh et al., 2013).
The prophylactic effect of 9-phenanthrol observed in HUVECs is probably not a widespread phenomenon, as application of 9-phenanthrol, even at 4 × 10−5 mol·L−1 does not protect the H9c2 cardiomyocyte cell line against peroxide-induced damage (Wang et al., 2013).
The effects of 9-phenanthrol, at 1, 2 and 3 × 10−5 mol·L−1, on cell viability were investigated in human gastric adenocarcinoma cell lines AGS and MKN-45 by use of the MTT colorimetric assay. None of these concentrations affected cell viability (Kim et al., 2012).
Stimulation of angiogenesis after ischaemic stroke
Ischaemic stroke increases an ATP-sensitive NSCCa current in neurovascular cells and neuroendothelia (Simard et al., 2006). This current is supported by a combination of SUR-1 and the TRPM4 channel, whose activation leads to oncotic death of endothelial cells and consequently, capillary fragmentation (Simard et al., 2010; Woo et al., 2013b). Ischaemic stroke conditions were mimicked in cultures of HUVECs by incubating the cells in a hypoxic solution and depriving them of glucose and serum (Loh et al., 2013). Such deprivation increases cell death after 24 h. While 5 × 10−6 mol·L−1 9-phenanthrol did not reduce cell death, it did promote the formation of capillary structures on Matrigel, as detected by light microscopy (Loh et al., 2013). This effect was attributed to TRPM4 channel inhibition because 24 h of oxygen and glucose deprivation produces TRPM4 overexpression at the mRNA and protein level and activates a TRPM4-like current that is inhibited by 10−4 mol·L−1 9-phenanthrol (Loh et al., 2013). These results show that 9-phenanthrol, via TRPM4 channel inhibition, improves the functionality of endothelial cells under conditions that mimic ischaemic stroke.
Modulation of neuronal activity
9-Phenanthrol modulates electrical activity in neurons and neural systems. NSCCa current with the hallmarks of the TRPM4 channel are present in the CNS where they participate in generating bursting activity in networks such as the breathing-related rhythmogenic neurons from the pre-Bötzinger complex (Pace et al., 2007; Mironov, 2008; Mironov and Skorova, 2011) and in neurons of the substantia nigra pars compacta (Mrejeru et al., 2011). Glutamate receptor activation gives rise to cytoplasmic Ca2+ that activates the depolarizing NSCCa current, favouring rhythmic burst generation. The effect of 9-phenanthrol on this activity was tested in a coronal midbrain slice model from juvenile mice. Electrical bursting in nigrostriatal dopamine-releasing neurons was recorded under NMDA stimulation. Burst frequency was reversibly reduced by 10−4 mol·L−1 9-phenanthrol, which changed the burst-firing pattern into one characterized by tonic firing (Mrejeru et al., 2011).
Accessory olfactory bulb neurons involved in transmitting pheromonal stimuli also express Trpm4 mRNA and exhibit a TRPM4-like current activated by afferent sensory fibre stimulation. In mice brain slices, inhibition of this current by 10−4 mol·L−1 9-phenanthrol reduces their ability to discharge repetitively (i.e. persistent firing) (Shpak et al., 2012).
The prepositus hypoglossi nucleus of the brainstem is involved in maintaining horizontal gaze. Inhibitory neurons from this structure exhibit NSCCa channels (Saito and Yanagawa, 2010). In slice preparations from the rabbit brainstem, 10−4 mol·L−1 9-phenanthrol was shown to reduce firing rate by 60% in these neurons (Saito and Yanagawa, 2010; 2013).
The appropriate use of 9-phenanthrol in physiological studies
Our meta-analysis of the studies presented earlier suggests that 9-phenanthrol modulates physiological processes mainly through TRPM4 channel inhibition. (i) In most models, the drug reduces an NSCCa in Trpm4-expressing cells (Gonzales et al., 2010a; Simard et al., 2013; Smith et al., 2013a,b). (ii) When tested, the effect of 9-phenanthrol is not reproduced in Trpm4–/– mice (Hof et al., 2013; Simard et al., 2013) or its effects can be mimicked by small interfering Trpm4 RNA treatment (Crnich et al., 2010). (iii) The IC50 values for the effects of 9-phenanthrol on physiological processes are in the same range as those measured for TRPM4 channel inhibition in recombinant HEK cells (Tables 1-3). (iv) The effect of 9-phenanthrol can be reproduced by other TRPM4 inhibitors such as flufenamic acid (Gonzales et al., 2010a; Saito and Yanagawa, 2010; Mrejeru et al., 2011; Shpak et al., 2012; Simard et al., 2012a; 2013; Amarouch et al., 2013; Burt et al., 2013; Kim et al., 2013) and glibenclamide (Becerra et al., 2011; Kim et al., 2013).
We conclude that 9-phenanthrol can confidently be used as a specific TRPM4 channel inhibitor. Nevertheless, high concentrations (at or exceeding 10−4 mol·L−1) must be used with caution for several reasons.
Firstly, 9-phenanthrol may precipitate at high concentrations, thus the final diluted concentration may be overestimated. 9-Phenanthrol is usually diluted in DMSO in a stock solution at 10−1 mol·L−1 and then diluted in the physiological perfusion solution, such that the maximal DMSO ratio of 0.01% is not exceeded even when the final concentration of 9-phenanthrol reaches 10−4 mol·L−1. This concentration of DMSO in the bathing solution has no effect on the TRPM4 channel (Grand et al., 2008). Because 9-phenanthrol partly precipitates at higher concentrations, one can increase the final DMSO concentration if a relatively high concentration of 9-phenanthrol is called for. Nevertheless, in general, this use of DMSO should be avoided because a high concentration of DMSO may introduce new side effects. DMSO inhibits non-selective cation channels such as NMDA and AMPA glutamate receptors when used at or above 0.5% (Lu and Mattson, 2001), and it inhibits whole-cell non-selective cationic currents in erythrocytes at or above 1% (Nardid et al., 2013).
Secondly, in rat cardiomyocytes, 10−4 mol·L−1 9-phenanthrol affects both L-type Ca2+ currents and delayed outward rectifier K+ currents (Simard et al., 2012a). Even if these effects on active membrane properties do not explain the modulation of electrical activity by 9-phenanthrol in cardiomyocytes, the effects on Ca2+ and K+ channels must to be considered carefully because of the ubiquity and widespread influence of these channels on voltage trajectory and intracellular signalling in many types of excitable cells.
Thirdly, 9-phenanthrol autofluoresces under UV light (Moriconi et al., 1959) in particular at 340 nm, which can induce artefacts when measuring Ca2+ or voltage using fluorescent dyes (Burt et al., 2013). This can cause problems when using high concentrations.
Fourthly, while the effects of 9-phenanthrol are reversible in most studies, the recovery is more problematic at 10−4 mol·L−1, which is probably due to the accumulation of 9-phenanthrol in membranes, because of its high hydrophobicity.
The most appropriate concentrations of 9-phenanthrol for physiological studies range from 1 × 10−5 to 3 × 10−5 mol·L−1. The lower limit corresponds closely to the IC50 for TRPM4 channel inhibition, while the upper limit produces an 80% inhibition of the current. The 3 × 10−5 mol·L−1 upper limit is sufficient to strongly affect recombinant TRPM4 current (Grand et al., 2008) as well as physiological processes, and perhaps more importantly, is devoid of side effects. Although several studies have demonstrated the physiological effects of 9-phenanthrol, they are less apparent at lower concentrations.
Note that, while 9-phenanthrol dissolved in DMSO can be stored at −20°C, it degrades with time. We would thus recommend that once it is in solution it should be used within a month.
9-Phenanthrol, a TRPM4 channel inhibitor for what purpose?
Our reading and laboratory experience indicates that 9-phenanthrol is an appropriate tool to investigate the functional significance of the TRPM4 channel in physiological and pathological processes. The drug is suitable for a variety of physiological approaches such as: (i) electrophysiological recordings using patch-clamp pipettes in single-channel or whole-cell configurations; (ii) transmembrane potential recording in multicellular preparations using an intracellular microelectrode; (iii) calcium measurements by fluorescent dyes; (iv) measurements of smooth muscle contraction; and (v) cell culture (which is amenable to the same set of measuring tools listed earlier).
A particular strength of 9-phenanthrol is its ability to distinguish TRPM4 from TRPM5 channels, which are otherwise quite similar (Ullrich et al., 2005; Guinamard et al., 2011). Because these two ion channels hold ionic currents with similar properties, the discovery of 9-phenanthrol has provided a convenient tool to differentiate one from the other in native preparations. A variety of NSCCa reported in native cells remain to be identified at the molecular level. 9-Phenanthrol thus might help resolve whether the underlying channels are TRPM4, TRPM5, or another variety of TRP channel. This issue is especially relevant in kidney tissue, which is known to express high levels of Trpm4 mRNA (Launay et al., 2002) and a native NSCCa current, whose identity is still not known (Chraïbi et al., 1994; Teulon, 2000). This disparity also applies for neurons of the pre-Bötzinger complex that express both Trpm4 and Trpm5 mRNA, and a NSCCa current implicated in rhythmic cellular and network bursting that serves to generate inspiratory breathing movements (Crowder et al., 2007; Pace et al., 2007; Mironov, 2008; Mironov and Skorova, 2011).
The TRPM4 channel has already been shown to be involved in a variety of physiological processes, such as cerebral artery constriction (Earley, 2013), insulin secretion by pancreatic beta cells (Cheng et al., 2007), immune responses (Launay et al., 2004; Vennekens et al., 2007; Barbet et al., 2008), DSM contraction (Smith et al., 2013a,b) and cardiac action potentials (Hof et al., 2013; Simard et al., 2013). Moreover, the channel is implicated in pathologies such as autoimmune encephalomyelitis (Schattling et al., 2012), ischaemia and ischaemia-reperfusion injuries in the brain or the heart (Simard et al., 2012a; Loh et al., 2013), cardiac hypertrophy (Guinamard et al., 2006), as well as human cardiac-genetic diseases (Kruse et al., 2009; Liu et al., 2010; 2013). It is tempting to speculate that 9-phenanthrol, through its effects on TRPM4 current, may modulate such pathological processes and could potentially ameliorate or correct their perturbations. This has already been observed in several studies, as described earlier. By modulating cardiac rhythm, 9-phenanthrol might function as a bradycardic agent (Hof et al., 2013). Also, as it reduces cardiac injuries induced by hypoxia-reoxygenation 9-phenanthrol may have cardioprotective properties (Simard et al., 2012a; Wang et al., 2013). By promoting angiogenesis, the drug may inhibit cerebral oedema damage after ischaemic stroke (Loh et al., 2013). Its inhibitory effect on the detrusor muscle might mean it has potential as a drug for the treatment of overactive bladder (Smith et al., 2013a,b). By preventing LPS-induced endothelial cell death, 9-phenanthrol may benefit the treatment of endotoxaemia-derived sepsis (Becerra et al., 2011). However, while these possibilities are exciting and provocative, it is still too early to predict specific clinical applications because all of the studies reviewed earlier were performed in vitro or on isolated cells or tissues. Among the barriers that must be overcome before going further in that direction is the ability to reach the sufficiently high levels of circulating drug in vivo to inhibit the TRPM4 channel. In that regard, the low solubility of 9-phenanthrol might be an obstacle. In addition, the toxicity of 9-phenanthrol has to be carefully evaluated as PAH are known to have consistent toxic effects (Feng et al., 2012). This last point was highlighted by results from a very recent study using in vitro biochemical assays, which indicated that 9-phenanthrol inhibits the biosynthesis of androgen and oestrogen in subcellular fractions of carp gonads (Fernandes and Porte, 2013).