The Three Catecholics Benserazide, Catechol and Pyrogallol are GPR35 Agonists

Nearly 1% of all clinically used drugs are catecholics, a family of catechol-containing compounds. Using label-free dynamic mass redistribution and Tango β-arrestin translocation assays, we show that several catecholics, including benserazide, catechol, 3-methoxycatechol, pyrogallol, (+)-taxifolin and fenoldopam, display agonistic activity against GPR35.


Introduction
Identification of the hidden mechanisms of action of drugs and natural products is essential to elucidate their in vivo effects and for drug repurposing. Catecholics, a family of catechol-containing compounds, is rich in natural products as well as in clinically used drugs. Catecholics are well known for their ability to protect against lipid peroxidation by non-enzymatic scavenging of free radicals with their catechol moiety. Furthermore, a recent study showed that out of the 8,659 drugs recorded in the Comprehensive Medicinal Chemistry database 78 are catecholics, and 17 are prescribed by the US Food & Drug Administration [1].
DMR assays were performed using an Epic ® system (Corning Inc., Corning, NY, USA). This system records a ligand-induced dynamic mass redistribution (DMR) signal of live cells as a shift in resonant wavelength (in picometer, pm) of a resonant waveguide grating biosensor. The DMR signal obtained is a real-time kinetic response that provides a holistic view of ligand-receptor interaction(s) and its functional consequence in the cells [16][17][18]. For HT-29 cells 30,000 cells per well were seeded into 384well Epic ® biosensor plates (Corning) and cultured overnight. For U2OS and U2OS-GPR35-bla cells 18,000 cells per well were seeded and cultured overnight. These cells were then washed twice using a plate washer, maintained with Hank's balanced salt solution (1 × HBSS) and further incubated inside the Epic ® system for 1 h. DMR agonist profiling proceeds with a 2-min baseline, followed by compound addition and monitoring the compound-induced responses for about 1 h. DMR antagonist assay proceeds with a 10min preincubation with a GPR35 antagonist (i.e., ML145 [19]), followed by recording a 2min baseline, adding an agonist at a fixed dose (typically at its EC 80 ), and recording the cellular response. All DMR signals were background corrected.
Tango assays were performed in engineered U2OS-GPR35-bla cell line. This cell line permits an endpoint measurement of the activity of agonists specific to the GPR35 activation-induced β-arrestin translocation [9][10][11]. Specifically, 10000 U2OS-GPR35-bla cells per well were seeded in 384-well, black-wall, clear bottom assay plates with low fluorescence background (Corning). After overnight culture, the cells were stimulated with ligands for 5 h at 37 °C under 5% CO 2 , and then loaded with the cell permeable LiveBLAzer™ fluorescence resonance energy transfer (FRET) B/G substrate. After 2 h incubation the coumarin to fluorescein ratio was measured using Victor 4 plate reader (PerkinElmer, Waltham, MA, USA). Results obtained were normalized to the maximal response of zaprinast obtained within the same plate. The maximal response of zaprinast was set to be 100%.

Results and Discussion
We performed literature mining to identify important catecholics, particularly drugs used clinically and food additives. This led to identification of benserazide and fenoldopam, the two catecholic drugs from DrugBank [20], and propyl gallate the food additive antioxidant [1], beside several catecholics that are known to be GPR35 agonists. Benserazide is a peripherally-acting aromatic L-amino acid decarboxylase or DOPA decarboxylase inhibitor which is used in combination with levodopa for the management of Parkinson's disease [21]. Fenoldopam is a selective dopamine D 1 receptor partial agonist used as an antihypertensive agent [22]. Propyl gallate is a food additive (E310, European Food Safety Authority) used to protect oils and fats in products from oxidation. Propyl gallate was recently identified to be an estrogen antagonist [23]. For comparative studies we also tested syringic acid, 3,4-dimethoxycinnamic acid, catechol, 3-methocatechol, pyrogallol, (−)-catechin, and (+)-taxifolin ( Figure 1). Gallic acid, zaprinast and YE210 were used as positive controls.  We combined label-free DMR assays in both native HT-29 and engineered U2OS-GPR35-bla cells with Tango β-arrestin translocation gene reporter assay in U2OS-GPR35-bla cells to ascertain the agonist activity and specificity at the GPR35 of these catecholics. Previously we found that gallic acid is a moderately potent GPR35 agonist with an EC 50 of 1.16 μM to trigger DMR in HT-29 cells, but caffeic acid is a weak partial agonist [13]. Thus, we first profiled gallic acid and caffeic acid analogues. Results showed that the ester analogue of gallic acid, propyl gallate, still triggered a dose-dependent DMR in HT-29 cells (Figure 2a), yielding a single EC 50 of 20.2 ± 1.9 μM (n = 4) (Figure 2b). The propyl gallate-induced DMR was dose-dependently and completely blocked by ML-145 (Figure 2c). Similarly, syringic acid, a 3,5-dimethoxy analogue of gallic acid, also triggered a dose-dependent DMR (Figure 2d), but with relatively low potency (EC 50 of 147 ± 23 μM, n = 4) and efficacy (Figure 2e). ML-145 also dose-dependently and completely blocked the syringic acid DMR (Figure 2f). Furthermore, the 3,4-dimethoxyl analogue of caffeic acid, 3,4-dimethoxycinnamic acid, was found to be inactive. Together, these results suggest that beside the carboxylic acid group the pyrogallol or catechol group is also important to the agonistic activity at the GPR35 of these catecholics, and catechol may be the minimal fragment of these catecholics to activate the GPR35.  We then examined such a possibility by profiling catechol and its analogues using the DMR assay. Results showed that catechol, 3-methoxycatechol and pyrogallol all triggered a dose-dependent DMR in HT-29 (Figure 3a-c), whose characteristics were similar to those induced by other known GPR35 agonists including zaprinast, tyrphostin-51 and YE210 [9,10]. The three catecholics displayed distinct potency, with an EC 50 rank order of pyrogallol (1.3 ± 0.1 μM) > 3-methoxycatechol (147 ± 15 μM) > catechol (319 ± 27 μM) (n = 4). However, the three catecholics displayed similar efficacy, as evidenced by similar maximal DMR signals (Figure 3d). The antagonist blockage experiments showed that ML-145 completely blocked the catechol DMR, but neither the β-adrenergic receptor antagonist betaxolol nor α 2 -adrenergic receptor antagonist yohimbine had any obvious effects (Figure 3e). HT-29 is known to endogenously express both α 2A -and β 2 -adrenergic receptors [24,25]; catechol is a partial agonist of β 2 -adrenergic receptor [26]. ML-145 gave rise to a monophasic inhibition of the pyrogallol-induced DMR (IC 50 of 6.0 ± 0.5 μM, n = 4), but a biphasic inhibition of the 3-methoxycatechol-induced DMR (IC 50 of 1.9 ± 0.2 and 84.9 ± 7.7 μM, respectively; n = 4) (Figure 3f). One possibility accounted for such distinct inhibition profiles of ML-145 is that the binding site between catecholics and ML-145 is not entirely overlapped [26]; such a possibility is currently under investigation. Alternatively, such distinct inhibitor profiles may be originated from the varying modes of action of ML-145 to block the responses arising from the activation of GPR35 by different agonists [27]. Nonetheless, these results suggest that all three catecholics are GPR35 agonists.  We further examined four other catecholics including benserazide, fenoldopam, (−)-catechin and (+)-taxifolin. Results showed that (−)-catechin was inactive, while all other catecholics are active in HT-29 ( Figure 4). The potency rank order was (+)-taxifolin (6.6 ± 0.5 μM) > benserazide (11.0 ± 0.9 μM) > fenoldopam (206 ± 19 μM) (n = 4). Interestingly, only benserazide triggered a DMR whose characteristics are distinct from other catecholics (Figure 4a). Blockage experiments showed that ML-145 only partially blocked the DMR of 16 μM benserazide (Figure 4b), but completely blocked the DMR of (+)-taxifolin or fenoldopam (Figure 4d). Given that the DMR of a ligand is a holistic view of its action in the cells and many ligands display polypharmacology [17,18], these results suggest that benserazide also activates another target in HT-29, beside GPR35. Next, we compared the DMR of benserazide in U2OS-GPR35-bla cells with that in the parental U2OS cells. Results showed that both benserazide and zaprinast were inactive in the parental cells (Figure 5a). In contrast, both ligands triggered a similar negative DMR in the GPR35 expressing cells (Figure 5b,c). Interestingly, compared to zaprinast, benserazide led to a smaller DMR with a lower potency (Figure 5d). The EC 50 was found to be 80.8 ± 7.1 nM and 1.00 ± 0.13 μM (n = 4) for zaprinast and benserazide, respectively. These results further confirmed that benserazide is a GPR35 agonist.

Conclusions
In summary, we hypothesized and confirmed that the catechol group alone can interact with and activate the GPR35 receptor. This is significant from three different perspectives. First, comparing to gallic acid, pyrogallol displayed much higher potency and efficacy to trigger β-arrestin translocation in U2OS-GPR35-bla, but almost identical potency and efficacy to trigger DMR in native HT-29. This suggests that the pyrogallol group may interact with GPR35 via a site that is distinct from the site of interaction of the negatively charged carboxylic group; such a possibility is currently under investigation. Second, the agonistic activity at the GPR35 is a commonly observed property of drugs used for treating Parkinson disease; these drugs include entacapone [9], tolcapone and nitecapone [15], L-DOPA [4], and benserazide (this study). A recent study suggests that entacapone, tolcapone, pyrogallol, gallic acid, caffeic acid, and quercetin all can block fibril formation of α-synuclein and β-amyloid and thus protect against amyloid-induced toxicity [28]. Although the mechanism behind is largely unknown, all these compounds are GPR35 agonists, suggesting a possible role of GPR35 in Parkinson's disease and other neuronal disorders, which is currently under investigation. Third, we had showed previously that the DMR in HT-29 is mostly sensitive to receptor activity associated with low receptor occupancy state, while the β-arrestin translocation signal in U2OS-GPR35-bla cells appears to be biased toward receptor activity associated with high receptor occupancy state [15]. As a matter of fact, the majority of GPR35 agonists examined gave rise to a potency to mediate the DMR in HT-29 that is much higher than its respective potency to trigger β-arrestin translocation in U2OS-GPR35-bla cells [4,[9][10][11][12][13][14][15]. However, our present results showed that catechol, pyrogallol, 3-methoxycatechol and benserazide, the four analogues without any carboxylic acid moiety, all displayed comparable potency in both types of assays, suggesting that the mechanism for these catechols to active GPR35 may be quite unique.