Synthetic small molecule GLP-1 secretagogues prepared by means of a three-component indole annulation strategy

Abstract

Rational assembly of small molecule libraries for purposes of drug discovery requires an efficient approach in which the synthesis of bioactive compounds is enabled so that numerous structurally related compounds of a similar basic formulation can be derived. Here, we describe (4 + 3) and (3 + 2) indole annulation strategies that quickly generate complex indole heterocycle libraries that contain novel cyclohepta- and cyclopenta[b]indoles, respectively. Screening of one such library comprised of these indoles identifies JWU-A021 to be an especially potent stimulator of glucagon-like peptide-1 (GLP-1) secretion in vitro. Surprisingly, JWU-A021 is also a potent stimulator of Ca(2+) influx through TRPA1 cation channels (EC50 ca. 200 nM), thereby explaining its ability to stimulate GLP-1 release. Of additional importance, the available evidence indicates that JWU-A021 is one of the most potent non-electrophilic TRPA-1 channel agonists yet to be reported in the literature.

Figure 1. Cyclohepta[b]indole synthesis by (4 + 3) and (3 + 2) cycloaddition reactions.

(Left panel) “A-Series” heterocyclics generated by (4 + 3) cycloaddition reactions. ^aindole (1 equiv), carbonyl (2 equiv), diene (5 equiv), GaBr₃ (10 mol%), rt. ^bindole (1 equiv), carbonyl (2 equiv), diene (5 equiv), Ga(OTf)₃ (10 mol%), rt. ^cindole (1 equiv), carbonyl (1.1 equiv), diene (5 equiv), Ga(OTf)₃ (20 mol%), rt. ^dSingle-crystal X-ray analysis. ^e2 mmol scale. (Right panel) “B-Series” heterocyclics generated by (3 + 2) cycloaddition reactions. ^aindole (1 equiv), carbonyl (2 equiv), diene (1.5 equiv), TfOH (20 mol%), rt.

Figure 2. Cyclohepta[b]indole-stimulated GLP-1 release: effects of TRPA1 channel blockers.

(a–c) JWU-A019, JWU-A020, and JWU-A021 stimulated GLP-1 secretion from STC-1 cells. (d) Basal GLP-1 secretion was not significantly altered by A967079. (e–g) GLP-1 secretion stimulated by JWU-A021 was reduced by A967079, AP-18, and HC030031. Data are the mean + s.d. of N = 3 independent experiments (*p < 0.05; paired t test).

Figure 3. Actions of JWU-A021 and other A-series compounds to increase [Ca²⁺]_i.

(a) Fura-2 assays of STC-1 cell monolayers demonstrated the concentration-dependent action of JWU-A021 to increase [Ca²⁺]_i. (b,c) A967079 and HC030031 each exerted concentration-dependent actions to counteract the stimulatory effect of JWU-A021 (1 μM) in STC-1 cells. (d) The TRPA1 channel activator AITC (10 μM) increased [Ca²⁺]_i, and this action of AITC was also reduced by A967079 and HC030031 in STC-1 cells. For these panels and subsequent figures, JWU-A021 was administered by bolus injection (Inj.). (e,f) Ca²⁺-elevating actions of JWU-A016 (e) and the A029-A034 series of test agents (f). For all examples depicted here, the findings are representative of a single experiment repeated a minimum of five times on five different occasions with similar results.

Figure 4. Differential actions of JWU-A021 enantiomers to increase [Ca²⁺]_i.

(a,b) The dextrorotatory (Dextro-) enantiomer (+)-(6R,9S)-JWU-A021 exerted a more powerful stimulatory effect in comparison to the levorotatory (Levo-) enantiomer (+)-(6S,9R)-JWU-A021 when it was tested in the fura-2 assay using monolayers of STC-1 cells. (c,d) Differential stimulation of an increase of [Ca²⁺]_i by the dextrorotatory, levorotatory, and racemic forms of either 1 μM or 0.3 μM JWU-A021. For all examples depicted here, the findings are representative of a single experiment that was repeated a minimum of three times on three different occasions with similar results.

Figure 5. JWU-A021 promotes Ca²⁺ influx rather than Ca²⁺ mobilization.

(a) The Ca²⁺ channel blocker La³⁺ exerted a concentration-dependent action to abrogate the increase of [Ca²⁺]_i stimulated by JWU-A021 (1 μM). (b) The action of JWU-A021 (1 μM) to increase [Ca²⁺]_i was abrogated when the Ca²⁺ concentration of the SES was reduced to 100 nM. (c,d) ATP dose-dependently increased the [Ca²⁺]_i measured under conditions in which the SES contained either 2.6 mM CaCl₂ (c) or 100 nM CaCl₂ (d). (e) The L-type Ca²⁺ channel blocker nimodipine (5 μM) slowed the rate of onset but failed to reduce the end-point increase of [Ca²⁺]_i measured in response to JWU-A021 (1 μM). (f) The effectiveness of nimodipine (5 μM) as an inhibitor of Ca²⁺ influx was demonstrated by its ability to fully block the end-point increase of [Ca²⁺]_i measured in response to 28 mM KCl-induced depolarization. For all examples depicted here, the findings are representative of a single experiment repeated a minimum of three times on three different occasions with similar results.

Figure 6. Membrane currents and Ca²⁺ transients activated by JWU-A021.

(a) Whole-cell patch clamp analysis (V_h −60 mV) demonstrated inward membrane currents activated by repeated 5 sec focal applications of JWU-A021 (3 μM; red triangles) to a single STC-1 cell. The inset provides a current-voltage (I-V) relationship for the current activated by JWU-A021 (I_m, membrane current in pA normalized to membrane capacitance in pF). It is the difference current obtained by subtracting the IV relationships measured during (time point “i”) and after recovery (time point “ii”) of the response. Findings are representative of a single patch clamp experiment that was repeated with similar results using N = 10 cells. (b) RT-PCR validation that STC-1 cells express TRPA1 channel mRNA, as detected using two different primer pairs (RT, reverse transcriptase; MWM, molecular weight markers). Findings are representative of a single experiment repeated twice with similar results. (c) Averaged Ca²⁺ transients obtained from STC-1 cells stimulated by focal application (arrows) of JWU-A021 (3 μM) to N = 8 cells. (d) Ca²⁺ transients stimulated by focal application (arrows) of JWU-A021 (3 μM) to a single HEK-293 cell transfected with rat TRPA1 cDNA fused to EYFP cDNA (red trace), or a HEK-293 cell transfected with EYFP cDNA but not rat TRPA1 cDNA (black trace). EYFP fluorescence was used as a marker to positively identify cells that were transfected so that fura-2 based assays of [Ca²⁺]_i could be performed using these cells. Findings are representative of a single experiment repeated a minimum of three times on three different occasions with similar results.

Figure 7. Studies with HEK-293 cells transfected with recombinant TRPA1.

(a,b) HEK-293 cell monolayers transfected with wild-type (WT) rat TRPA1 cDNA, but not a negative control empty vector (EV), exhibited an increase of [Ca²⁺]_i in response to JWU-A021 (1 μM), and this action of JWU-A021 was abrogated by the TRPA1 channel blocker A967079. (c,d) HEK-293 cells transfected with WT rat TRPA1 cDNA, but not a negative control EV, exhibited an increase of [Ca²⁺]_i in response to AITC (10 μM), and this action of AITC was abrogated by the TRPA1 channel blocker A967079. This experiment confirmed the expected failure of HEK-293 cells to express endogenous TRPA1 channels. (e) A concentration-dependent action of JWU-A021 to increase [Ca²⁺]_i was measured in HEK-293 cell monolayers transfected with wild-type (WT) rat TRPA1 cDNA. (f) HEK-293 cell monolayers transfected with mutant C622S rat TRPA1 cDNA responded to JWU-A021 in a manner nearly identical to that of cells transfected with WT TRPA1 (compare panels e,f). Thus, the non-electrophile JWU-A021 acted independently of C622 covalent modification. (g,h) The TRPA1 activator AITC stimulated an increase of [Ca²⁺]_i in HEK-293 cell monolayers transfected with WT rat TRPA1, and this action of AITC to increase [Ca²⁺]_i was greatly diminished in HEK-293 cells transfected with mutant C622S TRPA1 cDNA (compare panels g,h). Thus, the electrophile AITC must covalently modify C622 in order to fully activate the channel. For all examples depicted here, the findings are representative of a single experiment repeated a minimum of three times on three different occasions with similar results.

Figure 8. JWU-A021 stimulates GLP-1 release from mouse intestinal cells.

(a) Primary cultures were stimulated for 30 min using serum-free DMEM assay buffer containing either 5.6 or 10 mM glucose so that glucose-stimulated GLP-1 secretion could be measured. Data are the mean + s.d. of 4 independent assays (*p < 0.05; paired t test) and are expressed as the fold-stimulation of GLP-1 release, so that a value of 1.0 corresponds to GLP-1 release measured for buffer containing 5.6 mM glucose. (b) HC030031 (10 μM) inhibited the actions of JWU-A021 (3 μM) and AITC (100 μM) to stimulate GLP-1 secretion from primary cultures. HC030031 was administered 15 minutes prior to addition of JWU-A021 or AITC, and it was also present during the 30 minutes test interval during which cells were exposed to JWU-A021 or AITC dissolved in serum free DMEM assay buffer containing 5.6 mM glucose. Data are the mean + s.d. of 4–6 independent assays (*p < 0.05; **p < 0.01; ANOVA with Bonferroni post test). (c) Immunocytochemical detection of GLP-1 in primary cell cultures. The top panel illustrates specific GLP-1 immunoreactivity (brown), as detected using the anti-GLP-1 monoclonal primary antibody in combination with an HRP conjugated secondary antiserum. The bottom panel illustrates negative control non-specific labeling obtained when using the secondary antiserum only. (d) qRT-PCR analysis demonstrated that JWU-A021 (3 μM) increased the relative abundance of TRPA1 channel mRNA in primary cell cultures, and that this effect was reduced by HC030031 (10 μM). For this analysis, cultures were maintained for 30 minutes in serum-free DMEM assay buffer containing 5.6 mM glucose and the test compounds. Data are the mean + s.d. of 6 independent assays (**p < 0.01; ANOVA with Bonferroni post test). The top inset illustrates qRT-PCR products detected by agarose gel electrophoresis.