6-Furopyridine Hexamethylene Amiloride Is a Non-Selective P2X7 Receptor Antagonist

P2X7 is an extracellular adenosine 5′-triphopshate (ATP)-gated cation channel present on leukocytes, where its activation induces pro-inflammatory cytokine release and ectodomain shedding of cell surface molecules. Human P2X7 can be partially inhibited by amiloride and its derivatives at micromolar concentrations. This study aimed to screen a library of compounds derived from amiloride or its derivative 5-(N,N-hexamethylene) amiloride (HMA) to identify a potential P2X7 antagonist. 6-Furopyridine HMA (6-FPHMA) was identified as a novel P2X7 antagonist and was characterized further. 6-FPHMA impaired ATP-induced dye uptake into human RPMI8226 multiple myeloma cells and human P2X7-HEK293 cells, in a concentration-dependent, non-competitive manner. Likewise, 6-FPHMA blocked ATP-induced Ca2+ fluxes in human P2X7-HEK293 cells in a concentration-dependent, non-competitive manner. 6-FPHMA inhibited ATP-induced dye uptake into human T cells, and interleukin-1β release within human blood and CD23 shedding from RPMI8226 cells. 6-FPHMA also impaired ATP-induced dye uptake into murine P2X7- and canine P2X7-HEK293 cells. However, 6-FPHMA impaired ATP-induced Ca2+ fluxes in human P2X4-HEK293 cells and non-transfected HEK293 cells, which express native P2Y1, P2Y2 and P2Y4. In conclusion, 6-FPHMA inhibits P2X7 from multiple species. Its poor selectivity excludes its use as a specific P2X7 antagonist, but further study of amiloride derivatives as P2 receptor antagonists is warranted.


Introduction
Purinergic signalling has multiple physiological roles [1] and comprises a range of cell surface receptors including P2X and P2Y receptors, which are typically activated by extracellular adenosine 5 -triphosphate (ATP) [2]. P2X7 is found on multiple cell types, including leukocytes [3]. ATP-induced activation of P2X7 results in the flux of Ca 2+ , Na + and K + cations, as well as the transfer of organic cations, including fluorescent dyes, across the cell membrane [4]. Activation of P2X7 leads to multiple pro-inflammatory downstream events. This includes activation of the NOD-like receptor family pyrin domain containing 3 (NLRP3) inflammasome, and the subsequent release of the pro-inflammatory cytokine interleukin (IL)-1β [5], and the ectodomain shedding of cell surface molecules such as the low affinity IgE receptor CD23 [6]. As such, P2X7 is a potential therapeutic target for multiple inflammatory diseases [7]. Novel compounds that block P2X7 could be used as therapeutics in inflammatory diseases or as tools to better understand the roles of P2X7 at molecular and cellular levels. 2 of 15 Amiloride, a commonly used K + -sparing diuretic that inhibits the epithelial sodium channel [8,9], and its derivatives were first shown to inhibit human and murine P2X7 three decades ago. Amiloride inhibited ATP-mediated Na + fluxes in human leukaemic lymphocytes by 50% at 1 mM, while 5-(N,N-hexamethylene) amiloride (HMA) and 5-(N-ethyl-N-isopropyl) amiloride (EIPA) inhibited ATP-induced Na + fluxes by 95% and 72% at 40 µM, respectively [10]. Likewise, HMA also inhibited ATP-induced ethidium + uptake in human leukemic B cells by 50% at 20 µM, but concentrations > 100 µM were required for complete inhibition [11,12]. Another group demonstrated that HMA inhibited 2 (3 )-O-(4-benzoylbenzoyl) ATP (BzATP)-induced inward currents in human embryonic kidney (HEK)293 cells stably transfected with human (h) P2X7 (HEK-hP2X7 cells) by 30% at 10 µM [13]. Similarly, 40 µM HMA partially inhibited ATP-and BzATP-induced currents in Xenopus oocytes injected with mouse (m) P2X7 mRNA [14]. Species comparisons have revealed that HMA inhibits BzATP-induced uptake with reduced potency against mP2X7 compared to hP2X7 expressed in HEK293 cells [15]. This study also revealed that HMA potentiated mP2X7-mediated dye uptake at a relatively low concentration (16 µM) of BzATP, an effect not observed for hP2X7 [15]. Despite the inhibitory effects of amiloride and HMA against P2X7 reported above, to the best of our knowledge these compounds or other derivatives have not been pursued as P2X7 antagonists since these studies.
A library of amiloride and HMA derivatives has recently been described as potential inhibitors of human urokinase plasminogen activator (uPA) or as dual inhibitors of uPA and sodium-hydrogen exchanger isoform-1 (NHE1) [16][17][18]. It was hypothesized that this library may also contain potential P2X7 antagonists. Therefore, eighty of these compounds, including amiloride and HMA, were screened as potential novel human P2X7 antagonists with 6-furopyridine HMA (6-FPHMA) shown to inhibit hP2X7. This compound impaired hP2X7 in a non-competitive manner and blocked events downstream of P2X7 activation in human leukocytes. 6-FPHMA inhibited canine (c) P2X7 with similar potency to that for hP2X7 but was ten-fold less potent against mP2X7. However, this compound also inhibited hP2X4 and hP2Y receptors in a concentration-dependent manner with similar potency to hP2X7. As such, its poor selectivity excludes the use of 6-FPHMA as a specific P2X7 antagonist, but the future study of amiloride derivatives as single or dual P2 receptor antagonists is warranted.
Plasmid DNA used for all transfections was transformed into competent DH5α cells (Thermo Fisher Scientific), with positive colonies selected on LB agar plates containing 100 µg/mL ampicillin (mP2rx7) or 50 µg/mL kanamycin (cP2RX7 and hP2RX4) and single colonies were expanded in LB broth with the relevant antibiotic (37 • C, overnight). Plasmid DNA was isolated using the Wizard Plus SV Minipreps DNA Purification kit (Promega, Madison, WI, USA) according to the manufacturer's instructions.

Plate-Based Flow Cytometric Dye Uptake for Compound Screening
RPMI8226 cells were washed thrice with low-divalent medium (LDM) (145 mM NaCl, 2 mM KCl, 0.2 mM CaCl 2 , 13 mM D-glucose and 10 mM HEPES, pH 7.5) and 1 × 10 5 cells were added to each well of a 96-well flat-bottomed plate. Amiloride derivatives, at 10 µM (final concentration 5 µM) or vehicle (DMSO) were added, and plates pre-incubated (37 • C, 15 min). LDM, containing YO-PRO-1 iodide (basal) or both YO-PRO-1 iodide and ATP (250 µM, approximate EC 50 in this assay), was added and incubated (37 • C, 8 min) before the addition of ice-cold Mg 2+ stop solution (145 mM NaCl, 2 mM KCl, 0.2 mM CaCl 2 , 20 mM MgCl 2 and 10 mM HEPES, pH 7.5) to halt ATP-induced dye uptake. Wells were washed twice with LDM (300× g, 5 min) and cells were resuspended in LDM. Data was acquired with an Attune NxT Flow Cytometer an Autosampler (Thermo Fisher Scientific). Mean fluorescent intensity (MFI) of YO-PRO-1 + uptake was determined using FlowJo. The respective basal fluorescence was subtracted from each ATP sample to calculate ATPinduced uptake and samples were normalized to the maximal ATP-induced response in the presence of DMSO.

Flow Cytometric Uptake Assays for Characterization
ATP-mediated uptake assays were conducted as described with either ethidium bromide [22] or YO-PRO-1 iodide [23]. Briefly, RPMI8226, HEK-hP2X7, HEK-mP2X7, HEK-cP2X7 or cryopreserved human peripheral blood mononuclear cells (PBMCs) [24] were washed thrice with LDM (300× g, 5 min), resuspended in LDM and 1 × 10 6 cells were added to individual tubes. Cells were pre-incubated in the absence of compound (37 • C, 5 min) or in the presence of compound or vehicle (DMSO) (37 • C, 15 min) as indicated. Cells were then incubated with either ethidium bromide (25 µM) or YO-PRO-1 iodide (1 µM) in the presence of ATP or an equivalent volume of LDM (basal) (37 • C, 5 min). An equal volume of ice-cold Mg 2+ stop solution was added and cells centrifuged (300× g, 5 min). Cells were washed twice in LDM (300× g, 5 min), resuspended and data were collected with an LSR Fortessa X-20 flow cytometer. MFI of ethidium + or YO-PRO-1 2+ was determined using FlowJo. The respective basal fluorescence was subtracted from each ATP sample to calculate ATP-induced uptake and samples were normalized to the maximal ATP-induced response in the presence of DMSO.

Fura-2AM Ca 2+ Response Assay
Ca 2+ response assays were conducted as described [21]. Briefly, HEK293, HEK-hP2X7 or HEK-hP2X4 cells were harvested and plated in black, clear-bottom 96-well plates at 5 × 10 4 cells/well overnight (37 • C, 95% air/5% CO 2 ). Cells were washed with extracellular Ca 2+ solution (ECS) (145 mM NaCl, 2 mM CaCl 2 , 5 mM KCl, 13 mM glucose and 10 mM HEPES, pH 7.4) and incubated with Fura-2AM loading buffer (2.5 µM Fura-2AM and 0.2% pluronic acid in ECS) (37 • C, 95% air/5% CO 2 , 30 min, dark). Loading buffer was removed; cells were washed once with ECS and incubated in ECS (37 • C, 95% air/5% CO 2 , 20 min, dark) to allow Fura-2AM de-esterification. Where indicated, cells were pre-incubated with compound or vehicle (DMSO) (37 • C, 30 min, dark). Fura-2 fluorescence at 510 nm was measured every 5 s using a FlexStation 3 (Molecular Devices, San Jose, CA, USA). Baseline recordings were taken for 15 or 30 s and ATP or ECS (basal) were added using the built-in fluidics. Data was acquired using SoftMax Pro version 7.0 (Molecular Devices). Relative changes in intracellular Ca 2+ were calculated using the ratio of Fura-2AM fluorescence at 510 nm following excitation at 340 nm and 380 nm (F 340nm/380nm ). Ca 2+ responses were normalised to the baseline recordings using the formula ∆Ca 2+ = ∆F/F rest = (F − F rest )/F rest , where F is the F 340nm/380nm ratio in a well at a particular time point and F rest is the mean fluorescence of a well from 0-15 s or 0-30 s (prior to ATP or ECS addition) [25]. Ca 2+ responses over time were plotted in Prism software v8.4.2 (GraphPad Software, La Jolla, CA, USA). To determine relative P2X-mediated Ca 2+ responses, the area under the curve from 100 to 180 s was calculated with GraphPad Prism. To determine P2Y-mediated Ca 2+ responses the area under the curve from 30 to 100 s was calculated. In both cases, ATP-induced responses were calculated by subtracting the corresponding basal response. These responses were normalised to the maximal ATP-induced response within their respective experiment.

CD23 Shedding Assay
CD23 shedding from RPMI8226 cells was conducted as described [27], except using LDM. Briefly, 1 × 10 6 cells in LDM were added to individual tubes and pre-incubated in the absence or presence of compound (or vehicle) (37 • C, 15 min). ATP or LDM (basal) was added, cells were incubated (37 • C, 1 min), ice-cold Mg 2+ stop solution was added and tubes centrifuged (300× g, 5 min) to halt ATP-induced CD23 shedding. Cells were stained in the absence or presence of anti-CD23 mAb with 7AAD (on ice, 30 min, dark), washed once with D-PBS and data were collected using an LSR Fortessa X-20 flow cytometer. CD23 MFI on live cells was determined using FlowJo, with CD23 expression defined as the difference in MFI between cells stained in the presence or absence of anti-CD23 mAb. ATP-induced shedding was normalised to the maximal ATP-induced response in each experiment.

Statistical Analysis
Data are presented as mean ± standard error of the mean (SEM). All statistical analyses, graphing and curve fitting were conducted using GraphPad Prism. Where appropriate, data sets were tested for normality (Shapiro-Wilk normality test) and statistical differences were determined using a one-way ANOVA. Agonist responses were fitted to a 'log(agonist) vs. normalized response-variable slope' model and antagonist responses were fitted to a 'log(inhibitor) vs. normalized response-variable slope' model to determine the half-maximal effective concentration (EC 50 ) and half-maximal inhibitory concentration (IC 50 ), respectively.

6-FPHMA Is a Novel P2X7 Antagonist
Eighty compounds from a library of amiloride and HMA derivatives [16,17], including the parent compounds, were screened using a plate-based YO-PRO-1 2+ dye uptake assay. A total of seven compounds were found to impair ATP-induced dye uptake by >25% ( Figure S1, Supplementary Materials). Based on its chemical characteristics 6-FPHMA ( Figure 1A) was identified as a possible P2X7 antagonist and selected for further characterization. The corresponding amiloride analogue, 6-furopyridine-substituted amiloride (6-FPA) ( Figure 1B), was examined as a matched inactive control.

CD23 Shedding Assay
CD23 shedding from RPMI8226 cells was conducted as described [27], except LDM. Briefly, 1 × 10 6 cells in LDM were added to individual tubes and pre-incuba the absence or presence of compound (or vehicle) (37 °C, 15 min). ATP or LDM was added, cells were incubated (37 °C, 1 min), ice-cold Mg 2+ stop solution was and tubes centrifuged (300× g, 5 min) to halt ATP-induced CD23 shedding. Cell stained in the absence or presence of anti-CD23 mAb with 7AAD (on ice, 30 min, washed once with D-PBS and data were collected using an LSR Fortessa X-20 flo tometer. CD23 MFI on live cells was determined using FlowJo, with CD23 expr defined as the difference in MFI between cells stained in the presence or absence ti-CD23 mAb. ATP-induced shedding was normalised to the maximal ATP-induc sponse in each experiment.

Statistical Analysis
Data are presented as mean ± standard error of the mean (SEM). All statistica yses, graphing and curve fitting were conducted using GraphPad Prism. Where priate, data sets were tested for normality (Shapiro-Wilk normality test) and sta differences were determined using a one-way ANOVA. Agonist responses were fi a 'log(agonist) vs. normalized response-variable slope' model and antagonist resp were fitted to a 'log(inhibitor) vs. normalized response-variable slope' model to mine the half-maximal effective concentration (EC50) and half-maximal inhibitor centration (IC50), respectively.

6-FPHMA Is a Novel P2X7 Antagonist
Eighty compounds from a library of amiloride and HMA derivatives [16,1 cluding the parent compounds, were screened using a plate-based YO-PRO-1 2+ d take assay. A total of seven compounds were found to impair ATP-induced dye u by >25% ( Figure S1, Supplementary Materials). Based on its chemical characte 6-FPHMA ( Figure 1A) was identified as a possible P2X7 antagonist and selected f ther characterization. The corresponding amiloride analogue, 6-furopyridine-subs amiloride (6-FPA) ( Figure 1B), was examined as a matched inactive control.  3.2. 6-FPHMA, but Not 6-FPA, Inhibits hP2X7-Mediated Dye Uptake 6-FPHMA and 6-FPA were characterized using ATP-induced ethidium + uptake in human RPMI8226 multiple myeloma cells, which have repeatedly been shown to express endogenous P2X7 [27][28][29][30]. To determine the EC 50 for ATP against endogenous hP2X7, RPMI8226 cells were incubated with increasing concentrations of ATP and ethidium + uptake was determined by flow cytometry ( Figure S2A, Supplementary Materials). ATP induced ethidium + uptake in a concentration-dependent manner with a maximum response at 1 mM and an EC 50 of 311 ± 21 µM (Figure 2A). To determine the inhibitory effects of each compound, RPMI8226 cells were pre-incubated with DMSO (vehicle) or increasing concentrations of 6-FPHMA or 6-FPA followed by incubation with 300 µM ATP (approximate EC 50 ). 6-FPHMA inhibited hP2X7-mediated dye uptake in a concentrationdependent manner with 100% blockade at 30 µM and an IC 50 of 0.58 ± 0.05 µM ( Figure 2B).
In contrast, 6-FPA did not inhibit hP2X7-mediated dye uptake at any concentration up to 30 µM ( Figure 2B). uptake was determined by flow cytometry ( Figure S2A, Supplementary Materials). ATP induced ethidium + uptake in a concentration-dependent manner with a maximum response at 1 mM and an EC50 of 311 ± 21 μM (Figure 2A). To determine the inhibitory effects of each compound, RPMI8226 cells were pre-incubated with DMSO (vehicle) or increasing concentrations of 6-FPHMA or 6-FPA followed by incubation with 300 μM ATP (approximate EC50). 6-FPHMA inhibited hP2X7-mediated dye uptake in a concentration-dependent manner with 100% blockade at 30 μM and an IC50 of 0.58 ± 0.05 μM ( Figure 2B). In contrast, 6-FPA did not inhibit hP2X7-mediated dye uptake at any concentration up to 30 μM ( Figure 2B).  To provide additional evidence that 6-FPHMA inhibits hP2X7, ATP-mediated ethidium + uptake was also examined in stably transfected HEK-hP2X7 cells [19]. To determine the EC 50 for ATP against transfected hP2X7, HEK-hP2X7 cells were incubated with increasing concentrations of ATP and ethidium + uptake was determined by flow cytometry ( Figure S2B, Supplementary Materials). ATP induced ethidium + uptake in a concentrationdependent manner with a maximum response at 1 mM and an EC 50 of 466 ± 17 µM ( Figure 2C). To determine the inhibitory effects of each compound, HEK-hP2X7 cells were pre-incubated with DMSO (vehicle) or increasing concentrations of 6-FPHMA or 6-FPA followed by incubation with 450 µM ATP (approximate EC 50 ). 6-FPHMA inhibited hP2X7mediated dye uptake in a concentration-dependent manner with 95% blockade at 100 µM and an IC 50 of 1.63 ± 0.42 µM ( Figure 2D). Similar to RPMI8226 cells, 6-FPA did not inhibit hP2X7-mediated dye uptake at any concentration up to 30 µM ( Figure 2D).

6-FPHMA Inhibits hP2X7-Mediated Ca 2+ Responses
hP2X7 activation by ATP results in Ca 2+ influx, and this influx can be measured using the intracellular Ca 2+ indicator Fura-2AM. Thus, the effect of 6-FPHMA on ATP-induced Ca 2+ influx in HEK-hP2X7 cells was examined. To determine the EC 50 for ATP-mediated Ca 2+ responses in HEK-hP2X7 cells, HEK-hP2X7 and non-transfected HEK293 cells were incubated with ATP and Ca 2+ responses measured. HEK293 cells, which express endogenous hP2Y receptors [31], had an immediate Ca 2+ response following ATP stimulation that peaked within 5 s of ATP addition and decreased to baseline by 100 s regardless of ATP concentration ( Figure S3A, Supplementary Materials). In contrast, HEK-hP2X7 cells had a sustained Ca 2+ response lasting longer than 180 s ( Figure S3B, Supplementary Materials). As such, hP2X7-mediated Ca 2+ responses were determined by measuring the area under the curve from 100-180 s ( Figure S3B, Supplementary Materials). Using these criteria, ATP-induced Ca 2+ responses in HEK293 cells were negligible while ATP-induced Ca 2+ responses in HEK-hP2X7 were concentration-dependent with a maximal response at 2 mM and an EC 50 of 722 ± 76 µM ( Figure 2F).
P2X7 activation leads to the shedding of CD23 from human immune cells [6] and RPMI8226 cells have previously been used to examine such responses in this cell line model of human leukocytes [27,34]. To confirm that ATP could induce CD23 shedding, the EC 50 of ATP for CD23 shedding from RPMI8226 cells was first determined. RPMI8226 cells were incubated with increasing concentrations of ATP and CD23 expression on live cells was determined by flow cytometry (Figure S5, Supplementary Materials). ATP induced CD23 shedding in a concentration-dependent manner with a maximal response at 200 µM and an EC 50 of 112 ± 28 µM ( Figure 3E). To determine if 6-FPHMA could inhibit P2X7-mediated CD23 shedding, RPMI8226 cells were pre-incubated with DMSO (vehicle) or increasing concentrations of 6-FPHMA followed by 200 µM ATP. 6-FPHMA inhibited ATP-induced CD23 shedding in a concentration-dependent manner with 97% blockade at 30 µM and an IC 50 of 2.3 ± 0.5 µM ( Figure 3F).
As mentioned previously, HEK293 cells express endogenous hP2Y receptors, including hP2Y1, hP2Y2 and hP2Y4 [31] and activation of these receptors can induce Ca 2+ responses in these cells [21,31]. To confirm the presence of endogenous hP2Y receptors on HEK293 cells, these cells were incubated with increasing concentrations of ATP and Ca 2+
As mentioned previously, HEK293 cells express endogenous hP2Y receptors, including hP2Y 1 , hP2Y 2 and hP2Y 4 [31] and activation of these receptors can induce Ca 2+ responses in these cells [21,31]. To confirm the presence of endogenous hP2Y receptors on HEK293 cells, these cells were incubated with increasing concentrations of ATP and Ca 2+ responses measured. ATP-induced Ca 2+ responses were determined by measuring the area under the curve from 30 to 100 s ( Figure S7C, Supplementary Materials). ATP induced Ca 2+ in a concentration-dependent manner with a maximum response at 300 µM and an EC 50 of 2.50 ± 1.01 µM ( Figure S7D, Supplementary Materials). To determine if 6-FPHMA could block hP2Y receptors, HEK293 cells were pre-incubated with various concentrations of 6-FPHMA then incubated with 2.5 µM ATP (EC 50 ) and Ca 2+ responses were measured ( Figure S7E, Supplementary Material). 6-FPHMA inhibited endogenous hP2Y-mediated Ca 2+ responses in a concentration-dependent manner with 97% inhibition at 30 µM and an IC 50 of 0.33 ± 0.13 µM ( Figure 5B). responses measured. ATP-induced Ca 2+ responses were determined by measuring the area under the curve from 30 to 100 s ( Figure S7C, Supplementary Materials). ATP induced Ca 2+ in a concentration-dependent manner with a maximum response at 300 μM and an EC50 of 2.50 ± 1.01 μM ( Figure S7D, Supplementary Materials). To determine if 6-FPHMA could block hP2Y receptors, HEK293 cells were pre-incubated with various concentrations of 6-FPHMA then incubated with 2.5 μM ATP (EC50) and Ca 2+ responses were measured ( Figure S7E, Supplementary Material). 6-FPHMA inhibited endogenous hP2Y-mediated Ca 2+ responses in a concentration-dependent manner with 97% inhibition at 30 μM and an IC50 of 0.33 ± 0.13 μM ( Figure 5B).

Discussion
6-FPHMA was identified as a P2X7 antagonist following the screening of 80 amiloride or HMA derivatives, including the parent compounds. Characterization of this novel antagonist revealed that it inhibited hP2X7 in a non-competitive manner and that it could also block events downstream of P2X7 activation. 6-FPHMA displayed species-specific potency, with similar IC50 values against hP2X7 and cP2X7 but it was ten-fold less active against murine P2X7, similar to that observed for other P2X7 antagonists [19,20,[35][36][37][38]. The current study demonstrated that 6-FPHMA also inhibited hP2X4 and hP2Y receptors in a concentration-dependent manner with similar potency to hP2X7. Collectively, this study identified 6-FPHMA as a P2X7 antagonist, but this compound lacked selectivity, limiting its use as a specific P2X7 antagonist. Nevertheless, given this broad inhibitory action against P2 receptors, the possibility remains that other amiloride or HMA derivatives may be identified or developed as more selective P2X or P2Y antagonists.
6-FPHMA inhibited hP2X7 more potently than any other amiloride or HMA derivative described to date with IC50 values ranging from 0.46 μM to 2.3 μM depending on the assay and cell type used. Previous studies [10][11][12][13][14] and the screening assay in the current study indicated that amiloride blocks hP2X7 relatively poorly but derivatives with 5-alkylamino groups, like HMA and EIPA, have increased potency against this receptor. This hydrophobic 5-substitution may allow these compounds to better interact with the key residues in the previously identified allosteric P2X7 drug-binding pockets

Discussion
6-FPHMA was identified as a P2X7 antagonist following the screening of 80 amiloride or HMA derivatives, including the parent compounds. Characterization of this novel antagonist revealed that it inhibited hP2X7 in a non-competitive manner and that it could also block events downstream of P2X7 activation. 6-FPHMA displayed species-specific potency, with similar IC 50 values against hP2X7 and cP2X7 but it was ten-fold less active against murine P2X7, similar to that observed for other P2X7 antagonists [19,20,[35][36][37][38]. The current study demonstrated that 6-FPHMA also inhibited hP2X4 and hP2Y receptors in a concentration-dependent manner with similar potency to hP2X7. Collectively, this study identified 6-FPHMA as a P2X7 antagonist, but this compound lacked selectivity, limiting its use as a specific P2X7 antagonist. Nevertheless, given this broad inhibitory action against P2 receptors, the possibility remains that other amiloride or HMA derivatives may be identified or developed as more selective P2X or P2Y antagonists.
6-FPHMA inhibited hP2X7 more potently than any other amiloride or HMA derivative described to date with IC 50 values ranging from 0.46 µM to 2.3 µM depending on the assay and cell type used. Previous studies [10][11][12][13][14] and the screening assay in the current study indicated that amiloride blocks hP2X7 relatively poorly but derivatives with 5-alkylamino groups, like HMA and EIPA, have increased potency against this receptor. This hydrophobic 5-substitution may allow these compounds to better interact with the key residues in the previously identified allosteric P2X7 drug-binding pockets [39][40][41]. Whether amiloride, HMA or their derivatives, including 6-FPHMA, bind this pocket remains to be shown. Moreover, since 6-FPHMA was a more potent inhibitor of P2X7 than HMA, the addition of a second hydrophobic group, such as furopyridine, at the 6-position may increase the ability of these compounds to interact with the drug-binding pocket. It should be noted that the 5hexamethylene substitution is likely to be more important than the 6-furopyridine group for interactions in the drug-binding pocket, as 6-FPA, which contains a 6-furopyridine group but not a 5-hexamethylene substitution, did not inhibit P2X7. Studying other amiloride or HMA derivatives with similar functional groups to 6-FPHMA may help elucidate the binding mechanisms of these compounds.
6-FPHMA inhibited P2X7-mediated dye uptake into human T cells, including CD4 + and CD8 + T cell subsets, and human RPMI8226 multiple myeloma cells. This parallels earlier works that identified amiloride and its analogues as P2X7 antagonists using human leukemic lymphocytes [10,12]. Furthermore, the present study demonstrated for the first time that an HMA derivative can inhibit events downstream of hP2X7 activation, namely IL-1β release and CD23 shedding. Despite reduced inhibition of P2X7 on T cells compared to BBG and PPADS, 6-FPHMA inhibited IL-1β release in human blood more potently than either BBG or PPADS. This may indicate that 6-FPHMA would be a more potent P2X7 antagonist in vivo than BBG or PPADS, although it should be noted JNJ-47965567 inhibited P2X7 better than 6-FPHMA in either assay. It remains unclear as to why BBG or PPADS lacked inhibitory activity in whole blood. However, this finding has implications for the use of such antagonists in vivo and whether they also have reduced efficacy in rodent models of disease.
Contrary to the initial aim to identify a novel P2X7 antagonist, the current study provides the first evidence that an HMA derivative has the potential to inhibit multiple P2 receptors. 6-FPHMA inhibited hP2X7, hP2X4 and hP2Y receptors at similar potency. Further, 6-FPHMA can also inhibit both human uPA and NHE1 [17]. As mentioned above, five selective P2X7 antagonists bind non-competitively to the same allosteric hydrophobic drug-binding pocket on P2X7, but these compounds are unable to bind a similar pocket on P2X4 due to their relatively large size [39]. 6-FPHMA is larger than the smallest of these antagonists, so it is also unlikely that 6-FPHMA is binding to this hydrophobic pocket on P2X4. This suggests 6-FPHMA is binding to a different site on P2X4 to that of P2X7. Alternatively, 6-FPHMA may be binding both P2X4 and P2X7 at structurally similar allosteric drug-binding sites, such as the homologous amino acid residues I312 and I310 which are involved in the binding of the P2X4 antagonist BX430 [42]. If so, this would imply that 6-FPHMA does not bind the well-described allosteric binding pocket discussed above. However, the notion that 6-FPHMA binds structurally similar allosteric drug-binding sites on P2X4 and P2X7 does not explain the inhibition of hP2Y receptors by 6-FPHMA. This may suggest that 6-FPHMA is binding to a unique site on each P2 receptor studied, or it may indicate that 6-FPHMA is inserting directly into the cell membrane, as previously shown for the non-selective P2 receptor antagonists, PPADS and MRS2159 [43]. If true, this suggests that 6-FPHMA may have the ability to inhibit other P2 receptors. Future in vitro or in silico studies could examine where and how 6-FPHMA binds to P2 receptors. Moreover, 6-FPHMA may be suitable as a lead compound to develop a dual P2X7/P2X4 antagonist. Dual blockade of P2X7 and P2X4 may be beneficial as a therapy in a number of inflammatory diseases, such as cardiovascular disease [44,45] or neuroinflammatory disorders [46,47], where both receptors have important roles. However, concerning the development of such compounds as specific P2 receptor antagonists, the potential inhibition of uPA and/or NHE1 also needs to be considered.

Conclusions
In conclusion, this study identified and characterized a novel non-competitive P2X7 antagonist, 6-FPHMA. This compound had greater potency then either amiloride or HMA and impaired events downstream of P2X7 activation. Further, 6-FPHMA inhibited P2X7 from other species; however, it also inhibited human P2X4 and P2Y receptors, limiting its use as a P2X7 antagonist. Nevertheless, this suggests that 6-FPHMA may be a starting point for the development of more selective P2 receptor antagonists. Therefore, the further study of amiloride derivatives as P2 receptor antagonists is warranted.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/biom12091309/s1, Figure S1: Compound screening revealed 6-furopyridine substituted HMA as a candidate for further characterisation, Figure S2: Representative flow cytometric gating strategies to determine dye uptake into RPMI8226 and HEK-hP2X7 cells, Figure S3: Ca 2+ response traces in non-transfected and P2X7-transfected HEK293 cells, Figure S4: Representative flow cytometric gating strategy to determine dye uptake into human T cells, Figure S5: Representative flow cytometric gating strategy to determine CD23 shedding from RPMI8226 cells, Figure S6: Representative flow cytometric histograms of dye uptake into HEK-mP2X7 and HEK-cP2X7 cells, Figure S7: Transfection efficiency and Ca 2+ response traces in HEK-hP2X4 and HEK293 cells.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki and use of human tissues was approved by the Human Ethics Committee of the University of Wollongong (protocol code HE12/219 and approved 31 May 2012).

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.