3-Amino-chromanes and Tetrahydroquinolines as Selective 5-HT_2B, 5-HT₇, or σ₁ Receptor Ligands

Abstract

The phenethylamine backbone is a privileged substructure found in a wide variety of G protein-coupled receptor (GPCR) ligands. This includes both endogenous neurotransmitters and active pharmaceutical agents. More than 20 structurally unique heterocyclic phenethylamine derivatives were broadly evaluated for GPCR affinity. Selective ligands for the 5-HT_2B, 5-HT₇, and σ₁ receptors were identified, each with low nanomolar binding affinities. The σ₁ receptor affinity was supported in a cellular assay that provided evidence for increased cell survival under oxidative stress.

G protein-coupled receptors (GPCRs) are a prominent pharmacological target. More than 30% of FDA approved drugs target at least one GPCR.^1,2 Worldwide, more than 25% of drug sales come from GPCR modulating compounds.³ GPCR modulators are used to treat a wide variety of diseases and disorders including allergies,⁴ schizophrenia,⁵ depression,⁶ pain management,⁷ and asthma.⁸ Common GPCR targeting pharmaceuticals⁹ include the antihistamine loratadine,⁴ antidepressants fluoxetine and sertraline,⁶ antipsychotics aripiprazole, clozapine, and haloperidol,⁵ the opioids morphine, hydrocodone, and fentanyl,⁷ and bronchodilators salbutamol and tiotropium bromide.¹⁰ The phenethylamine core is a privileged substructure in GPCR ligands (Figure 1). This is partially because several endogenous neurotransmitters or neuromodulators are phenethylamines including dopamine¹¹ and epinephrine¹² (Figure 1a). The phenethylamine backbone is also found in a wide variety of GPCR targeting active pharmaceutical ingredients including morphine, salbutamol, and pseudoephedrine (Figure 1b).

Figure 1.

Phenethylamine containing GPCR ligands.

We recently developed a tandem Winstein rearrangement Friedel–Crafts alkylation that enabled the synthesis of differentially functionalized heterocyclic tertiary azides from an equilibrating mixture of allylic azides (Scheme 1a).¹³ This cascade was synthetically attractive because it constructed tetralins, chromanes, and tetrahydroquinolines featuring a tetra-substituted stereocenter while maintaining a diversifiable vinyl group. The heterocyclic products could be readily converted into substituted phenethylamine derivatives containing either a primary amine or pyrrolidine motif (Scheme 1b). It seemed prudent to investigate the potential GPCR affinity of these molecules because of their rigid structure, small molecular weight, and phenethylamine backbone. This report describes the results of an initial screen, which identified several individual molecules each with low nM binding affinity for 5-HT_2B, 5-HT₇, or σ₁ receptor. Furthermore, the small family of compounds screened demonstrates divergent selectivity across the GPCR family. A new lead compound was identified for the 5-HT₇ receptor, and a second lead compound was identified for the σ₁ receptor. The σ₁ receptor affinity was further supported by cellular assays.

Scheme 1. Source of Novel Phenethylamines.

The synthesis of these compounds began with the addition of vinyl Grignard to ketone 5, forming tertiary allylic alcohol 6 (Scheme 2). Subsequent cross metathesis with cis-1,4-but-2-enediol diacetate and Hoveyda–Grubbs second generation catalyst afforded allylic acetate 7. Azidation occurred upon exposure to TMSN₃ and catalytic Zn(OTf)₂ to generate allylic azide 8 as a mixture of equilibrating isomers. Methanolysis cleaved the acetate, and resulting alcohol 9 was activated as the trichloroacetimidate 1. The dynamic cyclization was performed with catalytic AgSbF₆, yielding tertiary azide 2, typically in >25:1 dr. Global reduction with H₂ and palladium on carbon afforded primary amine 3. Alternatively, exposure to HBCy₂ resulted in pyrrolidine 4. The pyrrolidine could be subjected to N-methylation (10) or acetylation (11).

Scheme 2. Synthesis of (±)-GPCR Ligands.

Reagents and conditions: (a) vinyl MgCl, THF, 0 °C, 30 min, 83–91%; (b) cis-1,4-diacetoxy-2-butene, 1–2 mol % Hoveyda–Grubbs second generation catalyst, 40 °C, 18 h, 68–86%; (c) TMSN₃, 10 mol % Zn(OTf)₂, rt, 90 min, 29–87%; (d) K₂CO₃, MeOH, rt, 30 min, 97%–quant.; (e) NCCCl₃, 20 mol % DBU, rt, 90 min, 73–97%; (f) 10 mol % AgSbF₆, CHCl₃, 40–60 °C, 24 h, 39–94%; (g) H₂, 10% w/w Pd/C, MeOH, 18 h, 52%-quant; (h) HBCy₂, DCM, 0 °C to rt, 18 h, 35–84%; (i) aq. CH₂O, NaBH₃CN, HOAc, NCCH₃, 0 °C to rt, 30 min, 70–89%; (j) Ac₂O, TEA, DMAP, 0 °C to rt, 18 h 73–96%.

The GPCR binding affinity of these compounds was assessed through the Psychoactive Drug Screening Program.¹⁴ Initially, three primary amines were screened (Table 1, see the Supporting Information for selected binding curves). Compounds 3a–3c represent one example each of the three synthetically accessible cores: tetralin (3a), chromane (3b), and tetrahydroquinoline (3c). These molecules were screened across a wide range of human GPCR including serotonin (5-HT), α- and β-adrenergic (Alpha and Beta), dopamine (D), histamine (H), muscarinic (M), opioid (OR), sigma (σ), and others (Table 1). Of the three compounds assayed, all exhibited at least moderate binding (K_i ≤ 200 nM) against one or more receptors. Tetralin 3a was the only compound to exhibit any affinity against the δ-, μ-, or κ-opioid receptor. This is consistent with the structures of other phenethyl opioid ligands that contain a cyclohexyl backbone (e.g., morphine, Figure 1b).^7,15 Compound 3a exhibited moderate affinity toward the 5-HT_2B and σ₁ receptors. Chromane 3b exhibited moderate affinity against the 5-HT_2B receptor but lacked any κ-opioid binding. Tetrahydroquinoline 3c exhibited the strongest initial hit, with a K_i of 74 nM against the 5-HT_1A receptor. Based on these initial results and lack of opioid affinity, a subsequent structure activity relationship (SAR) study was conducted on the chromane and tetrahydroquinoline scaffolds, screening against the 5-HT_1A, 5-HT_2B, 5-HT₇, σ₁, and σ₂ receptors.

Table 1. Initial GPCR Screen^a.

	pKia
receptor	compound 3a	compound 3b	compound 3c
5-HT1A	6.2 ± 0.1b	5.7 ± 0.1b	7.13 ± 0.06b
5-HT1B	<5	<5	<5
5-HT1D	<5	<5	<5
5-HT1E	<5	<5	<5
5-HT2A	<5	<5	<5
5-HT2B	6.84 ± 0.06b	6.75 ± 0.06b	6.1 ± 0.1b
5-HT2C	5.6 ± 1.0b	5.6 ± 3.5b	5.4 ± 1.3b
5-HT3	<5	<5	<5
5-HT5A	5.7 ± 1.5b	<5	<5
5-HT6	<5	<5	<5
5-HT7	<5	<5	6.0 ± 0.1b
Alpha1A	<5	<5	5.8 ± 0.1b
Alpha1B	<5	<5	<5
Alpha1D	<5	<5	<5
Alpha2A	5.2 ± 0.2b	<5	6.4 ± 0.1b
Alpha2B	5.3 ± 0.1b	<5	6.4 ± 0.1b
Alpha2C	<5	<5	<5
Beta1	<5	<5	<5
Beta2	5.3 ± 0.1b	<5	<5
Beta3	<5	<5	6.0 ± 0.1b
BZP	5.8 ± 0.1b	<5	5.1 ± 0.1b
D1	<5	<5	<5
D2	<5	<5	<5
D3	<5	<5	<5
D4	<5	<5	5.7 ± 0.1b
D5	<5	<5	<5
DAT	5.3 ± 0.2b	<5	5.5 ± 0.2b
GABAA	<5	<5	<5
H1	5.5 ± 0.5b	<5	<5
H2	<5	<5	<5
H3	<5	<5	5.4 ± 0.2b
H4	<5	<5	<5
M1	<5	<5	<5
M2	<5	<5	<5
M3	<5	<5	<5
M4	<5	<5	<5
M5	<5	<5	<5
NET	5.6 ± 0.1b	6.5 ± 0.1b	6.4 ± 0.1c
δ-OR	<5	<5	<5
κ-OR	6.5 ± 0.1b	<5	<5
μ-OR	5.7 ± 0.1b	<5	<5
PBR	<5	<5	<5
σ1	6.9 ± 0.1b	6.0 ± 0.1b	6.5 ± 0.1b
σ2	<5	<5	6.2 ± 0.1b

^aFor pK_i values reported as <5, the compound did not demonstrate ≥50% binding in an initial assay at 10 μM concentration.

^bThese data reflect triplicate measurements of a single experiment. The error reflects the error in the sigmoidal curve fit.

^cThese data reflect the average of triplicate experiments run in triplicate. The error reflects the standard deviation of the three calculated pK_i values. Values in italic type are noted for the reader’s convenience.

Assay results for five additional primary amine analogues are displayed (Table 2). Amines 3d and 3e did not demonstrate a K_i below 100 nM for any of the GPCRs assayed. Amine 3f bound the σ₁ receptor with a K_i of 16 nM and demonstrated reasonable selectivity relative to the 5-HT_2B receptor (10-fold), which was the second most sensitive receptor assayed. Amines 3g and 3h have higher affinities toward the 5-HT_2B receptor (K_i = 3.5 and 20 nM, respectively). Compound 3g demonstrated ∼150-fold selectivity relative to the σ₂ receptor, which was the second most sensitive GPCR assayed. The 5-HT_2B receptor is a member of the 5-HT family of receptors that is known to be an essential receptor during development.¹⁶ Long-term consumption of 5-HT_2B agonists can induce potentially fatal myofibroblast proliferation and valvular heart disease.^17,18 Thus, the 5-HT_2B receptor is considered an antitarget. Therefore, further optimization of this amine class was not pursued, opting instead to evaluate the σ₁ affinity demonstrated by amine 3f on more rigid pyrrolidine analogues.

Table 2. K_i Data for Primary Amino-chromanes^a.

^aFor pK_i values reported as <5, the compound did not demonstrate ≥50% binding in an initial assay at 10 μM concentration.

^bThese data reflect triplicate measurements of a single experiment. The error reflects the error in the sigmoidal curve fit.

^cThese data reflect the average of triplicate experiments run in triplicate. The error reflects the standard deviation of the three calculated pK_i values. Values highlighted in red are noted for the reader’s convenience. Values not determined are noted as n.d.

The σ receptors were initially thought to be members of the opioid receptor family.¹⁹ Since, it has been shown that σ₁ acts as a chaperone protein localized in the endoplasmic reticulum,²⁰ and it affects a wide variety of cellular functions including regulation of opioid receptors,²⁰ kinases,²¹ TRPV1,²² dopamine receptors,²³ apoptosis,²⁴ as well as cellular calcium and potassium levels.²⁵ Modulating intracellular calcium levels implicated σ₁ as a target for treating colon and breast cancer.^26,27 Modulating the σ₁ receptor also effects alcohol abuse,²⁸ pain management,²⁹ opioid analgesia,^30,31 and neuroprotection in models of retinal neural degradation.^32,33

Pyrrolidine containing chromanes were assessed for σ₁ binding (Table 3). A direct analogue of amine 3f was essentially equipotent against σ₁ (4f), while replacing the phenyl group with a methoxy group was disadvantageous (4e). Exploring substitution around the arene provided mixed results. Benzyloxy compound 4i was a high-affinity σ₁ ligand. Removing the benzyl group was detrimental to affinity (4d). Some affinity could be restored through N-methylation (10d). Unsurprisingly, masking the basic amine as an amide removed affinity in both cases examined (11d and 11l). Compound 4j could be a lead compound. While compound 4j demonstrated reduced affinity relative to amine 3f as a σ₁ ligand, it shows >200-fold selectivity versus the σ₂ receptor. Changing the methoxy groups for methyl groups enhanced 5-HT_2B binding (4k), and the other compounds assayed provided both reduced σ₁ affinity and selectivity (4l and 4m).

Table 3. K_i Data for Pyrrolidine Containing Chromanes^a.

^aFor pK_i values reported as <5, the compound did not demonstrate ≥50% binding in an initial assay at 10 μM concentration.

^bThese data reflect triplicate measurements of a single experiment. The error reflects the error in the sigmoidal curve fit.

^cThese data reflect the average duplicate experiments run in triplicate. The error reflects one-half the difference between the two calculated pK_i values.

^dThese data reflect the average of triplicate experiments run in triplicate. The error reflects the standard deviation of the three calculated pK_i values. Values highlighted in red are noted for the readers convenience. Values not determined are noted as n.d.

With an attempt to evaluate the 5-HT_1A affinity demonstrated by amine 3c (Table 1), other tetrahydroquinolines were investigated (Table 4). As a direct comparison to amine 3c (Table 1), pyrrolidine 4c (Table 4) was assayed. While primary amine 3c was ∼15-fold selective for 5-HT_1A over 5-HT₇, pyrrolidine 4c was not a suitable ligand for 5-HT_1A (K_i > 10 000 nM) and was instead a high-affinity 5-HT₇ ligand (K_i = 6.3 nM). Therefore, the relatively small structural change resulted in more than a 20 000-fold relative difference in the 5-HT_1A versus 5-HT₇ selectivity. The 5-HT₇ receptor has been implicated in the regulation of multiple biological functions including sleep, circadian rhythm, and mood.¹⁶ Various 5-HT₇ antagonists have been investigated for depression treatment along with other disorders.³⁴ The 5-HT₇ affinity could be further enhanced though N-methylation (10c), while acylation removed affinity (11c). Other substituted arenes displayed reduced 5-HT₇ affinity (4n and 4o). Even though compound 4c has a weaker affinity than compound 10c, compound 4c was selected for further assay because it exhibited enhanced selectivity, having a ∼75-fold lower affinity for the next most sensitive receptor, σ₁.

Table 4. K_i Data for Tetrahydroquinolines^a.

^aFor pK_i values reported as <5, the compound did not demonstrate ≥50% binding in an initial assay at 10 μM concentration.

^bThese data reflect triplicate measurements of a single experiment. The error reflects the error in the sigmoidal curve fit.

^cThese data reflect the average of triplicate experiments run in triplicate. The error reflects the standard deviation of the three calculated pK_i values. Values highlighted in red are noted for the reader’s convenience. Values not determined are noted as n.d.

Based on the results outlined in Tables 1–4, compounds 4c and 4j were screened more broadly against a wider array of GPCRs (Table 5). Gratifyingly, only minimal affinity was observed across the additional GPCRs that were investigated. This indicates that compound 4c could be considered a new lead targeting 5-HT₇ due to the low nM binding affinity (K_i = 6.3 nM) and selectivity (∼75-fold versus next most sensitive receptor). Compound 4j could be considered a new lead targeting σ₁ due to its respectable affinity (K_i = 44 nM), good σ₁ versus σ₂ selectivity (>200-fold), and good selectivity versus other GPCRs (>5-fold versus 5-HT_2B, and >20-fold versus 38 other GPCRs).

Table 5. GPCR Screen for Selectivity^a.

	pKia
receptor	compound 4c	compound 4j
5-HT1A	<5	6.0 ± 0.1
5-HT1B	<5	<5
5-HT1D	<5	<5
5-HT1E	<5	<5
5-HT2B	6.3 ± 0.1	6.6 ± 0.1
5-HT2C	<5	5.8 ± 0.1
5-HT3	<5	<5
5-HT6	<5	<5
5-HT7	8.2 ± 0.1	5.9 ± 0.1
Alpha1A	<5	<5
Alpha1B	<5	<5
Alpha1D	6.0 ± 0.1	<5
Alpha2B	<5	<5
Alpha2C	6.0 ± 0.1	<5
Beta1	5.5 ± 0.1	6.0 ± 0.1
Beta2	<5	6.24 ± 0.03
Beta3	<5	<5
BZP	<5	5.2 ± 0.1
D1	<5	<5
D2	<5	<5
D3	<5	<5
D4	<5	<5
D5	<5	<5
DAT	<5	<5
GABAA	<5	<5
H1	<5	<5
H2	5.7 ± 0.1	<5
H4	<5	<5
M1	<5	<5
M2	<5	<5
M3	<5	<5
M4	<5	<5
M5	<5	<5
NET	5.8 ± 0.1	<5
δ-OR	<5	<5
κ-OR	<5	<5
μ-OR	<5	<5
PBR	<5	<5
σ1	6.3 ± 0.1	7.4 ± 0.1
σ2	6.0 ± 0.1	<5

^aFor pK_i values reported as <5, the compound did not demonstrate ≥50% binding in an initial assay at 10 μM concentration. These data reflect triplicate measurements of a single experiment. The error reflects the error in the sigmoidal curve fit. Values in italic type are noted for the reader’s convenience.

The σ₁ receptor is a target for protecting retinal cells from neural degradation.^32,33,35 To date, the most studied agent is (+)-pentazocine (PTZ), which has shown promise in mice.^36,37 Unfortunately, PTZ is a potent opioid and is probably unsuitable for clinical use as a treatment for retinopathy.⁷ The protective effects of compound 4j were demonstrated in the well-characterized cone photoreceptor cell line 661W.³⁸ Treatment of 661W cells with up to 200 μM of compound 4j for 24 h did not alter cell viability as measured using the MTT assay (Figure 2). This indicated that compound 4j was not cytotoxic to 661W cells at the concentrations assayed.

Figure 2.

Cytotoxicity assay of compound 4j in 661W cells. 661W cells were treated with compound 4j (10, 30, 100, and 200 μM) for 24 h. Cell viability was assessed using the MTT assay. No significant difference in viability was observed at the 4j concentrations tested compared with nontreated cells. MRP = compound 4j.

Exposing 661W cells to tert-butyl hydroperoxide (tBHP, 55 μM) for 24 h induced oxidative stress, which decreased cell viability by more than 50% (Figure 3). Cotreatment with compound 4j resulted in a dose-dependent improvement in cell viability (see the Supporting Information for additional concentrations). At a 4j concentration of 30 μM, cell viability was significantly greater than tBHP-exposed cells and was >80% of nontreated cells. The beneficial effects of compound 4j (at 30 μM) were similar to those observed in cells treated with PTZ (50–100 μM, Figure 3). At the highest concentration tested (200 μM), the beneficial effects of compound 4j were no longer observed. The beneficial effects of compound 4j (30 μM) could be counteracted by the addition of BD1063 (10 μM), which is a known σ₁ receptor antagonist (see the Supporting Information).

Figure 3.

Enhanced 661W cell viability with compound 4j upon tBHP exposure. 661W cells were treated with tBHP in the presence/absence of increasing concentrations of compound 4j (10–200 μM) or the σ₁ receptor ligand PTZ (20–100 μM) for 24 h before cell viability assessment. Cell viability was assessed using the MTT assay. Data are presented as mean ± standard deviation (SD) of triplicate measurements; *p < 0.05; ** p < 0.01; *** p < 0.001; ns = not significant as compared to the tBHP treatment. (tBHP = tert-butyl hydroperoxide; MRP = compound 4j; PTZ = (+)-pentazocine.)

Earlier studies demonstrated the beneficial effects of σ₁ receptor modulation by PTZ in attenuating oxidative stress in 661W cells.³⁹ Compound 4j was evaluated for similar effects. Reactive oxygen species (ROS) were detected using the CellROX green reagent in an immunocytochemical assay upon exposure of 661W cells to tBHP (55 μM). CellROX is a cell-permeant dye that is weakly fluorescent in a reduced state but exhibits bright green photostable fluorescence upon oxidation by ROS. Cells that received no tBHP showed minimal fluorescence, while cells exposed to tBHP showed a marked increase in green fluorescence (Figure 4A). Treatment with compound 4j resulted in decreased green fluorescence, especially at the 30 μM concentration. These data were similar to the fluorescence suppression observed with PTZ treatment (Figure 4A). Quantification of fluorescence intensity indicated a marked increase in ROS in the tBHP-exposed cells, whereas the level of ROS was significantly reduced in tBHP-exposed cells treated with compound 4j or PTZ (Figure 4B).

Figure 4.

Attenuated oxidative stress induced by tBHP in 661W cells upon treatment with compound 4j or PTZ. 661W cells were seeded on coverslips for 18 h. Cells either were or were not (control) exposed for 2 h to tBHP (55 μM) in the presence/absence of compound 4j or PTZ. (A) Representative immunofluorescent images of cells incubated with CellROX green reagent to detect ROS; green fluorescent signals indicated ROS as visualized by epifluorescence. DAPI was used to label nuclei (blue). (B) Quantification of fluorescent intensity reflecting ROS levels of data shown in panel A. Data are presented as mean ± SD. Data represent three independent experiments performed in duplicate. Significant differences are indicated: *** p < 0.001; ****p < 0.0001 as compared to the tBHP treatment. (CT = control, tBHP = tert-butyl hydroperoxide; MRP = compound 4j; PTZ = (+)-pentazocine.)

In conclusion, the GPCR affinity was investigated for a variety of new phenethylamine-based primary amines and pyrrolidines. Several compounds demonstrated significant and selective binding against the 5-HT_2B, 5-HT₇, or σ₁ receptor. Compound 4j protected against oxidative stress when assayed in 661W cells. The potency of compound 4j was comparable to PTZ in these assays.

Supporting Information Available

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsmedchemlett.9b00225.

• Experimental procedures, characterization data, and additional information on assays (PDF)

Author Contributions

MP performed the synthesis of new compounds. HX and JW performed assays with 661W cells. The project was directed by JT and SS. The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

This research was supported by the National Institute of General Medical Sciences of the National Institutes of Health under Award Number R35GM124718. This work was supported by the National Institutes of Health under award number R01EY028103 and the Foundation Fighting Blindness award number TA-NMT-0617–021-AUG.

The authors declare the following competing financial interest(s): A provisional patent has been filed by the University of Minnesota on the substructure disclosed in this report; M.R.P. and J.J.T. filed US patent No. 16/428,343, submitted 5/31/2019, on the compounds in this article.