Probe-Dependent Negative Allosteric Modulators of the Long-Chain Free Fatty Acid Receptor FFA4

High-affinity and selective antagonists that are able to block the actions of both endogenous and synthetic agonists of G protein–coupled receptors are integral to analysis of receptor function and to support suggestions of therapeutic potential. Although there is great interest in the potential of free fatty acid receptor 4 (FFA4) as a novel therapeutic target for the treatment of type II diabetes, the broad distribution pattern of this receptor suggests it may play a range of roles beyond glucose homeostasis in different cells and tissues. To date, a single molecule, 4-methyl-N-9H-xanthen-9-yl-benzenesulfonamide (AH-7614), has been described as an FFA4 antagonist; however, its mechanism of antagonism remains unknown. We synthesized AH-7614 and a chemical derivative and demonstrated these to be negative allosteric modulators (NAMs) of FFA4. Although these NAMs did inhibit FFA4 signaling induced by a range of endogenous and synthetic agonists, clear agonist probe dependence in the nature of allosteric modulation was apparent. Although AH-7614 did not antagonize the second long-chain free fatty acid receptor, free fatty acid receptor 1, the simple chemical structure of AH-7614 containing features found in many anticancer drugs suggests that a novel close chemical analog of AH-7614 devoid of FFA4 activity, 4-methyl-N-(9H-xanthen-9-yl)benzamide (TUG-1387), will also provide a useful control compound for future studies assessing FFA4 function. Using TUG-1387 alongside AH-7614, we show that endogenous activation of FFA4 expressed by murine C3H10T1/2 mesenchymal stem cells is required for induced differentiation of these cells toward a more mature, adipocyte-like phenotype.

Herein, we examine the antagonistic properties of AH-7614, and a small number of chemical derivatives, at FFA4. This demonstrates that AH-7614 is able to inhibit FFA4-mediated signals induced by a range of distinct fatty acids and synthetic FFA4 agonist ligands across all endpoints tested. Analysis of the mechanism of antagonism indicates that these compounds are negative allosteric modulators (NAMs) of FFA4 function, which display "probe dependence" in the nature of their inhibitory properties, depending on the specific agonists they are inhibiting. This work also identifies a close chemical derivative of AH-7614 that lacks antagonistic activity at FFA4, which, given the current lack of knowledge about potential off-target effects of AH-7614, will be a useful tool in defining FFA4-specific biologic effects. We have demonstrated the utility of this approach in assessing the role of FFA4 in adipogenesis of murine C3H10T1/2 mesenchymal stem cells.
Plasmids. Plasmids encoding either the human (short isoform) or mouse FFA4 receptors with enhanced yellow fluorescent protein (eYFP) or the related fluorescent protein mVenus fused to their C terminus and incorporating an N-terminal FLAG epitope tag in the pcDNA5 FRT/TO expression vector were generated as previously described Butcher et al., 2014;Alvarez-Curto et al., 2016;Prihandoko et al., 2016).
b-Arrestin-2 Interaction Assay. b-Arrestin-2 recruitment to either human or mouse isoforms of FFA4 was measured using a bioluminescence resonance energy transfer (BRET)-based approach. HEK293T cells were cotransfected with eYFP-tagged forms of each receptor in a 4:1 ratio with a b-arrestin-2 Renilla luciferase plasmid using polyethylenimine. Cells were then transferred into white 96-well plates at 24 hours post-transfection. At 48 hours post-transfection, cells were washed with Hanks' balanced salt solution (HBSS) and then incubated in fresh HBSS prior to the assay. Cells were preincubated for 15 minutes with HBSS supplemented with vehicle [1% (v/v) DMSO], AH-7614, TUG-1506, or TUG-1387. Cells were incubated with 2.5 mM Renilla luciferase substrate coelenterazine h (Nanolight Tech, Pinetop, CA) at 37°C for 10 minutes and then stimulated with various FFA4 agonists for a further 5 minutes at 37°C. BRET resulting from receptor-b-arrestin-2 interaction was then determined by measuring the ratio of luminescence at 535 and 475 nm using a Pherastar FS fitted with the BRET1 optic module (BMG Labtech, Aylesbury, UK).
Ca 21 Mobilization. Calcium assays were carried out on either Flp-In T-REx 293 cell lines, generated to inducibly express hFFA4 upon treatment with doxycycline, or 1321N1 cells stably expressing the hFFA1 receptor. One day prior to the experiment, cells were seeded at 50,000 cells/well (Flp-In REx 293) or 25,000 cells/well (1321N1) in poly-D-lysine-coated, black, clear-bottom, 96-well microplates. Cells were allowed to adhere for 3-4 hours before the addition of 100 ng/ml doxycycline to induce receptor expression in the case of Flp-In T-REx 293 cell lines. The following day, cells were incubated in culture medium containing the calcium-sensitive dye Fura2-AM (3 mM) for 45 minutes. Cells were then washed three times and then allowed to equilibrate for 15 minutes in HBSS prior to conducting the assay. Fura2 fluorescent emission was measured at 510 nm following excitation at both 340 and 380 nm during the course of the experiment using a Flexstation plate reader (Molecular Devices, Brambleside, UK). Calcium responses were then measured as the difference between 340/380 ratios before and after the addition of the relevant compounds. For antagonism, the cells were preincubated for 15 minutes prior to agonist addition with HBSS supplemented with vehicle [1% (v/v) DMSO], AH-7614, or TUG-1387 (10 mM) prior to the addition of 50 mM a linolenic acid (aLA), 500 nM TUG-891, or 13 nM TUG-770.
High-Content Imaging Quantitative Internalization Assay. hFFA4-mVenus Flp-In T-REx 293 cells were plated 75,000 cells/well in black with clear-bottom 96-well plates. Cells were allowed to adhere for 3-6 hours before the addition of doxycycline (100 ng/ml) to induce receptor expression. After an overnight incubation, culture medium was replaced with serum-free DMEM containing the ligand to be assessed and incubated at 37°C for 30 minutes before fixation with 4% paraformaldehyde. After washing with phosphate-buffered saline, cell nuclei were stained for 30 minutes with Hoechst33342.
Plates were subsequently imaged using a Cellomics ArrayScan II high-content plate imager (Thermo Fisher Scientific, Paisley, UK). Images were processed to identify internalized mVenus, which was then normalized to cell number based on nuclei identified by Hoechst33342 staining, to obtain a quantitative measure of hFFA4-mVenus internalization.
HTRF-Based Inositol Monophosphate Assay. Inositol monophosphate assays (Cisbio Bioassays, Codolet, France) were performed according to the manufacturer's instructions. In brief, a suspension of 7500 cells/well was incubated with the stated concentrations of agonist for 1 hour in the presence of 10 mM lithium chloride (LiCl). Inositol monophosphate accumulation was subsequently measured using a Pherastar FS plate reader.
RNA Isolation and Reverse-Transcription Quantitative Polymerase Chain Reaction (RT-qPCR). Total RNA was isolated from C3H10T1/2 cells using an RNEasy mini kit (Qiagen, Manchester, UK). Following RNA isolation, reverse-transcription polymerase chain reaction was performed using Superscript III (Life Technologies), and the resultant cDNA was used as a template for quantitative polymerase chain reaction analysis using an ABI Prism 7300 sequence detector (Applied Biosystems, Paisley, UK). Cycling conditions were as follows: 50°C for 2 minutes, 95°C for 10 minutes, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. FFA4, PPARg, and Runx2 expressions were then defined relative to cyclophilin using the 2 2DDCt method.
Western Blotting. Analysis of receptor phosphorylation was performed on mFFA4-eYFP cells that were pretreated with either 10 mM AH-7614 or TUG-1387 for 30 minutes, after which time they were treated with 10 mM TUG-891 for 5 minutes. Cell lysates were prepared and size fractionated on 4-12% SDS-PAGE gels and transferred to nitrocellulose membranes. Nitrocellulose membranes containing resolved receptor proteins were incubated in Tris-buffered saline LI-COR blocking buffer (LI-COR Biosciences, Cambridge, UK) and incubated subsequently with a phospho-specific antiserum produced in house that is able to recognize phosphorylated Thr 347 and Ser 350 residues of the receptor . Membranes were air-dried and scanned using an LI-COR Odyssey CLx Imager.
Oil Red O Staining. Untreated or differentiated C3H10T1/2 cells were fixed in 10% (v/v) formalin for 60 minutes at room temperature. The cells were washed in 60% (v/v) isopropanol and then incubated for 10 minutes with isopropanol supplemented with 0.6% (w/v) Oil Red O. The Oil Red O solution (Sigma-Aldrich) was then removed, and cells were washed Â4 with dH 2 O and then stored in phosphate-buffered saline at 4°C until visualization using an Invitrogen EVOS FL Auto Imaging System (Life Technologies). After removal of all liquid, the Oil Red O stain was dissolved in isopropanol and the absorbance measured at 405 nm using a Pherastar FS reader.

632
Statistical Analysis. Data are presented as the mean 6 S.E.M. of a minimum of three independent experiments performed in triplicate, unless otherwise stated in the respective figure legends. Data analysis was carried out using GraphPad Prism software (package v5.02; GraphPad Software, La Jolla, CA), with concentration-response data fitted to three-parameter sigmoidal curves. For allosteric modulation experiments, data sets were globally fit to an operational model of allosterism described previously . To fit these data, the t value for all allosteric modulators was constrained to reflect the lack of detectable agonism seen within these compounds, which allowed for curve fits yielding estimates of loga, logb, logK A , and logK B . Statistical analyses were carried out using one-way analysis of variance followed by Tukey's post-hoc test.

Results
We synthesized and assessed the functional activity of AH-7614 (Fig. 1), a xanthene derivative of a diarylsulfonamidebased FFA4 agonist. This compound was originally described as an antagonist of FFA4, able to block effects of both the polyunsaturated v-6 fatty acid linoleic acid and the synthetic FFA4 agonist GSK137647A (Fig. 1) (Sparks et al., 2014), and has since been used in a number of functional studies (Quesada-López et al., 2016;Houthuijzen et al., 2017;Villegas-Comonfort et al., 2017) but without analysis of its mode of action. Indeed, AH-7614 inhibited, in a potent and concentration-dependent manner (pIC 50 vs. EC 80 concentrations of the agonists), the ability of either the v-3 fatty acid aLA (pIC 50 7.51 6 0.08, n 5 3) or a synthetic orthobiphenyl-based agonist of FFA4, TUG-891 ( Fig. 1) (pIC 50 8.13 6 0.08, n 5 3) (Shimpukade et al., 2012;Hudson et al., 2013;Butcher et al., 2014), to promote Ca 21 mobilization in Flp-In T-REx 293 cells induced to express hFFA4-eYFP (Fig.  2, A and B). Similar results were obtained when measuring the ability of AH-7614 to inhibit agonist-induced interactions of hFFA4 with b-arrestin-2 following coexpression of these two proteins in HEK293T cells (Fig. 2, C and D). Importantly for subsequent studies, AH-7614 also functionally antagonized effects of both aLA and TUG-891 at the corresponding mouse ortholog of the receptor (mFFA4) (Fig. 2, E and F). AH-7614 showed selectivity for FFA4, as it was unable to antagonize activation of the other long-chain fatty acid receptor, FFA1 (Milligan et al., 2015, in studies in which the synthetic alkyne-based FFA1 agonist TUG-770  promoted elevation of intracellular Ca 21 in 1321N1 cells stably transfected to express a human FFA1 construct (Fig. 2G).
In similar EC 80 inhibition experiments, AH-7614 was also effective in blocking TUG-891-mediated internalization of FFA4 from the cell surface (Fig. 3A) (pIC 50 5 7.70 6 0.10, n 5 3). We therefore used this internalization assay, with an endpoint 30 minutes after addition of agonist to facilitate achievement of ligand equilibrium, to probe the mechanism of AH-7614mediated blockade of FFA4. Receptor internalization experiments were conducted in which the ability of defined concentrations of preadded AH-7614 to block internalization in response to various concentrations of TUG-891 was assessed (Fig. 3B). In these experiments, it was clear that the primary inhibitory effect of AH-7614 was to produce a decrease in the maximal response to  . Both aLA and TUG-891 induced calcium release that was potently inhibited in the presence of AH-7614 (aLA, pIC 50 = 7.51 6 0.08, n = 3; TUG-891, pIC 50 = 8.13 6 0.08, n = 3). HEK293T cells transiently expressing hFFA4-eYFP and b-arrestin-2-Renilla luciferase were preincubated with either vehicle [1% (v/v) DMSO] or increasing concentrations of AH-7614. The cells were then treated with either 50 mM aLA (C) or 500 nM TUG-891 (D), and b-arrestin-2 recruitment was subsequently determined using a BRETbased assay. Both aLA and TUG-891 promoted b-arrestin-2 recruitment that was potently inhibited (aLA, pIC 50 = 7.66 6 0.05, n = 5; TUG-891, pIC 50 = 7.55 6 0.07, n = 5) in the presence of increasing concentrations of AH-7614. mFFA4-dependent b-arrestin-2 recruitment in the presence of either 50 mM aLA (E) or 500 nM TUG-891 (F) was also potently inhibited (aLA, pIC 50 = 8.05 6 0.08 n = 3; TUG-891, pIC 50 = 7.93 6 0.06, n = 3) by AH-7614. However, AH-7614 had no effect on TUG-770-induced calcium release in 1321N1 cells stably transfected with hFFA1 receptor (G), indicating that AH-7614 is not an antagonist at this receptor. Data represent experiments performed in triplicate three times or more.

FFA4 Negative Allosteric Modulators
TUG-891. Although this is consistent with outcomes reported by Sparks et al. (2014) using intracellular Ca 21 assays, the longer time course of the internalization studies largely eliminates potential issues of ligand hemiequilibrium that can greatly influence outcomes and mechanistic interpretation in assays, such as those based on Ca 21 mobilization, that use much shorter endpoints (Charlton and Vauquelin, 2010). These findings indicated that AH-7614 is not a competitive antagonist of FFA4. Interestingly, even at the highest concentration of AH-7614 tested, a residual response to TUG-891 remained. This response to TUG-891 was similar in the presence of 1 and 10 mM AH-7614, and as such, the effect of AH-7614 appeared to saturate. Such an outcome is consistent with AH-7614 functioning as an NAM of FFA4.
As it is common for small chemical changes in allosteric GPCR ligands to significantly alter the details of their allosteric behavior (Wood et al., 2011; Hudson et al., 2014), we next synthesized derivatives of AH-7614 and assessed their activity at FFA4. Two key derivatives were considered. In the first, the sulfonamide of AH-7614 was replaced with an amide (TUG-1387), whereas in the second, the xanthine was replaced with a thioxanthene (TUG-1506) (Fig. 4A). These analogs were tested for their ability to antagonize EC 80 concentrations of TUG-891 (Fig. 4B) in a b-arrestin-2 recruitment assay. The thioxanthene-containing TUG-1506 retained the ability to inhibit TUG-891 (pIC 50 5 6.38 6 0.09, n 5 3) but with lower potency than the xanthene containing AH-7614. In contrast, the amide-containing TUG-1387 lacked all ability to inhibit TUG-891-mediated activation of FFA4.
As AH-7614, and likely TUG-1506, appeared to be NAMs of FFA4 function, we characterized and quantified the nature and extent of their NAM properties against synthetic FFA4 agonists from several distinct chemotypes. A common feature of GPCR allosteric modulators is that they may show "probe dependence," producing different allosteric effects depending on the orthosteric agonist being modulated (May et al., 2007). Therefore, we conducted experiments in which AH-7614 (Fig.  5, A-D) and TUG-1506 (Fig. 5, E-H) were used to antagonize b-arrestin-2 recruitment to FFA4 induced by synthetic agonists from each of four distinct chemotypes: TUG-891 (Shimpukade et al., 2012), TUG-1197 (Azevedo et al., 2016), GSK137647A (Sparks et al., 2014), and Cpd A (Oh et al., 2014) (Fig. 1). Data from these experiments were then fit to an operational model of allosterism to quantify the allosteric effect each NAM had on affinity of each orthosteric agonist (a), the allosteric effect on efficacy of each orthosteric agonist (b), the affinity of the orthosteric agonists (K A ), and the affinity of the modulator (K B ) ( Table 1). AH-7614 reduced the maximal response to each agonist tested, and produced a modest reduction in agonist potency (Fig. 5, A-D). This corresponded to loga ( Fig. 5I; Table 1) and logb ( Fig. 5J; Table 1) values that were both less than zero for each of the four agonists. However, although the loga values were broadly similar for each agonist, the logb for GSK137647A was significantly lower (P , 0.01) than the value obtained for either TUG-891 or Cpd A, suggesting a subtle level of probe dependence in AH-7614 modulation of FFA4 activity. This can be noted in the concentration-response curves by the observation that, although high concentrations of AH-7614 eliminated measureable response to GSK137647A (Fig. 5C), a residual response was apparent for all other agonists, even at 30 mM AH-7614 (Fig. 5, A, B, and D). Importantly, the agonist responses were similar in the presence of between 3 and 30 mM AH-7614, indicating that the inhibitory properties of AH-7614 were saturable, a hallmark of allosteric modulators.
In examining the allosterism between TUG-1506 and the various FFA4 agonists (Fig. 5, E-H), much more striking probe dependence was apparent. Most notably, whereas TUG-1506 primarily shifted the potency of TUG-891 (Fig. 5E) and Cpd A (Fig. 5H) with little effect on maximal response, the opposite was true for TUG-1197 and GSK137647A, where TUG-1506 depressed the maximal response without affecting agonist potency. This was reflected in the operational model curve fit parameters for these experiments, where a large negative loga was observed for TUG-891 and Cpd A ( Fig. 5K; Table 1) but a large negative logb was observed for TUG-1197 and GSK137647A (Fig. 5L; Table 1). This is a clear indication of the probe dependence in the NAM properties of TUG-1506. Interestingly, this may be related to the chemical nature of the Cells were pretreated with varying concentrations of AH-7614 and exposed to increasing concentrations of TUG-891 (B) for 30 minutes, after which they were fixed. Cells were imaged using a Cellomics ArrayScan II, and the extent of internalization of the receptor construct was quantified.
agonists, with TUG-891 and Cpd A both being carboxylatebased FFA4 agonists, whereas TUG-1197 and GSK137647A are each sulfonamide-based agonists. Importantly, once again, the inhibitory effects of TUG-1506 were saturable in all cases, entirely consistent with defining this compound as an NAM at FFA4. Similar experiments were also conducted with TUG-1387 against each FFA4 agonist, although in this case, given that TUG-1387 had already been shown to lack activity at FFA4, only a single high concentration of TUG-1387 was used (Fig. 6, A-D). TUG-1387 did not produce an NAM effect on TUG-891, TUG-1197, or Cpd A, and only produced a very modest effect on GSK137647A. Overall, these results are consistent with either TUG-1387 having no affinity for FFA4 or it being a silent allosteric modulator that, even if it does bind the receptor, does not affect orthosteric agonist function.
To further assess the potential for using AH-7614 in combination with TUG-1387 to define biologic functions of FFA4, we aimed to confirm the broader ability of AH-7614, but not TUG-1387, to block additional FFA4 signaling pathways. We   . Estimated parameters for loga and logb are shown for both AH-7614 (I and J) and TUG-1506 (K and L) when assessed at each of the four agonists. Data shown are from representative experiments (N numbers for each agonist-NAM combination are reported in Table 1) fit to an operational model of allosterism .

FFA4 Negative Allosteric Modulators
explored both FFA4-mediated accumulation of inositol monophosphates (Fig. 7A) and agonist-induced phosphorylation of Thr 347 and Ser 350 residues within the C terminus of FFA4 detected with previously described phospho-FFA4 antisera (Butcher et al., 2014;Prihandoko et al., 2016) (Fig. 7B). In each case, the agonist response was functionally inhibited by AH-7614, but was not affected by TUG-1387. Importantly for studies on more physiologically relevant systems, we also found that TUG-1387 lacked activity at the related fatty acid receptor FFA1 in experiments measuring calcium release in 1321N1-hFFA1 cells (not shown).
Recently, it has been suggested that FFA4 is able to act as a bipotential regulator of osteogenic and adipogenic differentiation of bone marrow-derived mesenchymal stem cells (Gao et al., 2015). Furthermore, FFA4 has been proposed to be a key player in the activation of brown adipose tissue (Quesada-Lopez et al., 2016), further reinforcing a role for FFA4 in adipose tissue physiology. To further assess the role of FFA4 in adipogenic differentiation, and to demonstrate the utility of the FFA4 NAM (AH-7614) when used in combination with TUG-1387 to interrogate FFA4 biology, we used the murine mesenchymal stem cell line C3H10T1/2 (Tang et al., 2004;   636 Hashimoto et al., 2016;Lee et al., 2016). Here, differentiation toward an adipocyte phenotype was produced by maintaining the cells in the presence of an induction mixture (IID) consisting of 100 nM insulin, 500 mM IBMX, and 10 nM dexamethasone for 5 days. Oil Red O staining effectively visualized the development of triglyceride deposits associated with adipogenesis (Fig. 8A). In concert, exposure to the IIDcontaining medium produced clear upregulation of mRNA for the adipogenic development marker, PPARg (Fig. 8B). This effect was not produced simply by cell confluence on the tissue culture plate (Fig. 8B). In parallel, levels of Runt-related transcription factor 2 (Runx2) mRNA, a key transcription factor associated with osteoblast differentiation Yuan et al., 2016), were greatly reduced (Fig. 8C). Over this time, frame levels of mFFA4 mRNA were also markedly increased by cell maintenance in the IID medium; interestingly, however, an equivalent upregulation of mFFA4 mRNA was also produced simply by maintaining C3H10T1/2 cells at confluence for a 5-day period (Fig. 8D). The ability of C3H10T1/2 cells to differentiate toward the adipocyte phenotype was clearly dependent upon signaling from FFA4, because treatment of cells with AH-7614 along with the IID medium greatly reduced visual observation of triglyceride stores as detected by Oil Red O staining and its quantification (Fig. 9, A and B). Treatment with AH-7614 also limited IIDinduced PPARg mRNA induction (Fig. 9C), without affecting the associated downregulation of Runx2 expression (Fig. 9D).
In contrast, treatment of C3H10T1/2 cells with TUG-1387 during the induction period with the IID-containing medium had no significant effect on the development of Oil Red O staining (Fig. 9, A and B). Taken together, these data indicate that TUG-1387, although structurally related, lacks activity at both FFA4 and the related FFA1, and therefore indicate this compound will be a useful chemical control to confirm that biologic effects produced by AH-7614 and related compounds, such as TUG-1506, are in fact FFA4-mediated.

Discussion
The availability of pharmacological tool compounds is central to efforts to define functions of many cellular proteins, including GPCRs, and to assess such proteins as potential therapeutic targets. However, for poorly and recently characterized receptors, such tool compounds are often lacking; described largely in the chemical/patent literature and, therefore, often of limited availability; and/or are poorly characterized in terms of selectivity and potential off-target effects. Within the group of GPCRs that respond to nonesterified, or "free," fatty acids, the most broad-reaching pharmacopoeia to date targets the long-chain fatty acid receptor FFA1 . This reflects that FFA1 has been targeted therapeutically and in human clinical trials to assess whether activation of this receptor might provide an approach to treat type II diabetes (Watterson et al., 2014;Li et al., 2016;Mohammad, 2016;Ghislain and Poitout, 2017). Although the second GPCR that is activated by longer-chain free fatty acids, FFA4, is also frequently suggested as a potential therapeutic target for diabetes and other metabolic disorders Moniri, 2016;Hansen and Ulven, 2017), this lags behind studies on FFA1, and no synthetic ligands with activity at this receptor have yet entered clinical studies . As a result, even tool compounds that are useful for assessing roles of this receptor are less widely available or understood . Moreover, as clear consensus is based on the therapeutic potential of agonism of FFA4 Moniri, 2016), little effort Fig. 7. AH-7614, but not TUG-1387, blocks agonist-induced elevation of intracellular inositol monophosphates and phosphorylation of FFA4. (A) FFA4-mediated inositol monophosphate (IP1) production was measured in cells expressing hFFA4-mVenus. TUG-891 (closed circles) promoted accumulation of IP1. Pretreatment with either 10 mM AH-7614 (inverted triangles) or increasing concentrations of this molecule (open squares) blocked IP1 accumulation. By contrast, pretreatment with TUG-1387 was without effect (open diamonds). (B) mFFA4-eYFP cells were pretreated with vehicle or various concentrations of AH-7614 or TUG-1387 for 30 minutes followed by treatment with 10 mM TUG-891 for 5 minutes. Cell lysates were prepared and resolved by SDS-PAGE and immunoblotted with an mFFA4specific phospho-antiserum that recognizes phosphorylation of residues Thr 347 and Ser 350 . Quantification of a series of such immunoblots is shown in the right-hand panel. ns, P . 0.05.

FFA4 Negative Allosteric Modulators
has been expended to identify, report, or characterize antagonists of this receptor, despite antagonist tool compounds being a traditional touchstone for detailed pharmacological analysis.
The only inhibitor of FFA4 function reported to date is AH-7614 (Sparks et al., 2014). Although Sparks et al. (2014) suggested that this compound would likely act as a competitive antagonist of FFA4, the limited data presented in their report did not actually suggest such a mode of action, which appeared to be noncompetitive, at least in assays measuring regulation of intracellular Ca 21 levels. The studies reported here expand on this, defining AH-7614 as an NAM of FFA4 function. This conclusion is supported by several hallmarks of allosterism (May et al., 2007), including that the ability of AH-7614 to inhibit FFA4 function is saturable, and that AH-7614 displays some level of probe dependence. More striking, however, in this regard is the chemical derivative of AH-7614, TUG-1506, described for the first time in these studies, which shows clear probe dependence depending on the chemical nature of the FFA4 agonist it is modulating. Specifically, this compound is an NAM primarily of ligand affinity for the two carboxylate-containing agonists tested, but primarily an NAM of efficacy for the noncarboxylate, sulphonamide-based agonists. Such changes in allosteric modulation are commonly observed with small chemical changes in allosteric ligands (Wood et al., 2011;Hudson et al., 2014), and indeed, the observed pattern of inhibition by both AH-7614 and TUG-1506 can only be explained by an allosteric mode of action. . RT-qPCR analysis of PPARg (n = 5) (C) and Runx2 (n = 4) (D) levels was then performed on differentiated (IID) and nondifferentiated (vehicle) C3H10T1/2 cells in the presence or absence of 10 mM AH-7614. Data are presented as the mean 6 S.E. (**P , 0.01, and ***P , 0.001). Scale bar in (A) = 400 mm.

FFA4 Negative Allosteric Modulators
Whereas Sparks et al. (2014) more broadly assessed the selectivity of the FFA4 agonist GSK137647A described in their study, they did not extend this to AH-7614. Therefore, having identified TUG-1387 in the current work as a close structural analog of AH-7614 that lacks activity at FFA4, we used each of these ligands to define specific roles for FFA4. Indeed, the FFA4 NAM AH-7614 prevents differentiation of a mouse mesenchymal stem cell line toward an adipocyte phenotype, and does so in a manner that is consistent with blockade of endogenous signaling via FFA4.
We describe AH-7614 and TUG-1506 as the first allosteric ligands for the FFA4 receptor. Despite only a minor structural difference between these two compounds, substantial differences in the nature of their allosterism were observed, perhaps suggesting that further examination of the structure activity relationship within the AH-7614 chemical series may yield novel allosteric FFA4 ligands with diverse pharmacological properties. The fact that the observed NAM properties within the AH-7614 series were clearly probe dependent indicates that care must be taken when designing future studies using these ligands. Overall, with appropriate caution, AH-7614 should prove to be a valuable tool in further unraveling functions of the FFA4 receptor.
Performed data analysis: Watterson, Hudson, Alvarez-Curto. Wrote or contributed to the writing of the manuscript: Watterson, Hudson, Alvarez-Curto, Milligan.