On-DNA Synthesis of Multisubstituted Indoles

The increasing role of the DNA-encoded library technology in early phase drug discovery represents a significant demand for DNA-compatible synthetic methods for therapeutically relevant heterocycles. Herein, we report the first on-DNA synthesis of multisubstituted indoles via a cascade reaction of Sonogashira coupling and intramolecular ring closure. Further functionalization by Suzuki coupling at the third position exploits a diverse chemical space. The high fidelity of the method also enabled the construction of an indole-based mock library.

D NA-encoded library (DEL) technology is now an important hit discovery tool that covers an unprecedentedly large part of the bioactive chemical space through the effective screening of billions of molecules.−9 On the basis of the substitutional pattern of U.S. Food and Drug Administration (FDA)-approved drugs, synthetic options for functionalization at the second and third positions are preferred. 7−20 On the basis of the handbook of DEL, traditional methods, such as Fischer indole synthesis, applying high temperatures and strongly acidic or inert conditions, are either not feasible on-DNA or propose detrimental reaction conditions to DNA. 21,22Recently, Kazmierski et al. reported potent IDO1 inhibitors identified by indole-based DEL screening (Scheme 1a). 23Although the strategy of applying Larock indole synthesis has been revealed, no experimental details were disclosed.In contrast to the reported method, we considered an alternative synthetic approach that would not hinder the pharmaceutically essential third position of indole but would exploit the less important sixth position for DNA attachment (Scheme 1b).Inspired by the work of Sakamoto et al. on the off-DNA cyclization of 2-ethynyl sulfonamides 24 and the reports of Satz et al. and Neri et al. on the on-DNA Sonogashira couplings, 25,26 we envisioned a cascade reaction starting from 2-iodo sulfonamides to indoles in one step.Subsequent iodination at the third position of indole would enable further functionalization by Suzuki couplings.This approach would also enable the introduction of amino acids at the fifth or sixth position of indole, offering another position for variability.
Initially, we investigated the cascade Sonogashira coupling and intramolecular ring closure using DNA-conjugated orthoiodo N-methanesulfonamide (1a) and phenylacetylene (2).Starting from recently published DNA-compatible Sonogashira couplings, we first applied a Pd(OAc) 2 /TPPTS precatalyst system in the presence of CuSO 4 , sodium ascorbate, and K 2 CO 3 using DMAc as the organic co-solvent at 75 °C for 3 h (entry 1 in Table 1).These conditions enabled full conversion respective to the starting material, resulting in a 9% yield of the desired indole 3a and 78% yield of the intermediates 4 and 5 based on high-performance liquid chromatography−mass spectrometry (HPLC−MS) (see the Supporting Information).Further experiments revealed that only dimethyl sulfoxide (DMSO) was a suitable co-solvent promoting not only the Sonogashira coupling but also the sequential ring closure, resulting in a 73% yield of product 3a (entry 2 in Table 1).Other co-solvents, such as MeCN and N,N-dimethylformamide (DMF), facilitated the Sonogashira coupling only (entries 3 and 4 in Table 1).Investigating the effect of the base, we found that both Cs 2 CO 3 and K 3 PO 4 provided the desired indole in good yields (entries 5 and 6 in Table 1).N,N-Diisopropylethylamine (DIPEA), however, inhibited the formation of the indole core (entry 7 in Table 1).Switching from the precatalyst system to the simple addition of sSPhos Pd G2 and CuSO 4 at 85 °C for 2 h enhanced the yield to 78% (entry 8 in Table 1).Finally, we probed various sulfonamides, such as NTs (1b), NNs (1c), and NTf (1d), in the reaction (entries 9−11 in Table 1), with compounds 1b and 1d leading to the corresponding products 4b or 4d and 5b or 5d as major products and compound 1c yielding product 3a in a moderate 52% yield.
With the optimized conditions in hand, we evaluated the scope of the method on a diverse set of aromatic, heteroaromatic, and aliphatic acetylenes (6) starting from the N-Ms compound 1a (Scheme 2).Mostly, we observed high to quantitative conversions leading to moderate to good yields.HPLC−MS results indicated only minor deiodination, and no Scheme 1. Strategy for the On-DNA Synthesis of Multisubstituted Indoles spontaneous desulfonylation of the starting material was observed.
The developed method tolerated the presence of alkyl, alkoxy, hydroxymethyl, COOH, COMe, CONH 2 , and NO 2 groups and halogen atoms (3b−3m).Amine-containing derivatives 3n and 3o were obtained in 25 and 29% yields, respectively, as a result of the formation of unidentified side products.In the case of the 2-formyl derivative (3p), we observed the full conversion of the starting material 1a in the Sonogashira coupling; however, the reaction conditions did not promote the ring closure to the desired indole.CNcontaining 3q underwent partial hydrolysis, presumably resulting in terminal amide 3r, a phenomenon that was almost totally suppressed in the case of benzylic CN (3s and 3t).The applied reaction conditions enabled the incorporation of thiophene, pyrazoles, aminopyridine, quinoline, and indazole rings in the second position in 35−67% yields (3u−3z).
Notably, steric hindrance around the reaction center did not compromise the formation of indole, demonstrated by multisubstituted oxazole 3aa obtained in 58% yield.In the case of thiazole derivative 3ab, the Sonogashira coupling was not initiated.We observed 10% deiodination in the case of pyridine derivative 3ac, eventually leading to a 44% yield.Aliphatic acetylenes required 85 °C for 8 h or 75 °C for overnight reaction times to effectively promote the ring closure.We obtained cyclopropyl 3ad, pent-2-ol 3ae, ethylene benzene 3af, and piperidine-4-one 3ag derivatives in 49, 56, 56, and 59% yields, respectively, at 85 °C after overnight reaction.Propargyl amine provided the expected product 3ah in 21% yield as a result of low conversion in the Sonogashira coupling.In the case of octanoic acid, we observed full conversion; however, inefficient ring closure provided product 3ai in 20% yield.Benzylic derivative 3aj and N-Boc-protected tetrahydropyrrole 3ak were synthesized at 75 °C overnight, leading to yields of 45 and 71%, respectively.Under the applied reaction conditions, benzylic alcohol derivative 3al was completely degraded, while in the case of the urea and phthalimide derivatives (3am and 3an), we only obtained unidentified side products.Changing the position of iodine and the NHM group on the starting material provided expected indole 3ao in 87% yield.When the halogen atom was changed to bromine (3ap), however, the starting material remained intact.
Applying the on-DNA iodination conditions of Lu et al., 27 we briefly optimized the reaction conditions for the Suzuki coupling starting from iodoindole (7) and phenylboronic acid (8) 26 (see the Supporting Information for the complete optimization).The nature of the organic co-solvent and the base greatly influences the efficiency of the Suzuki coupling, with the main side reaction being the deiodination of the starting material to product 3a.In the presence of DMSO as a co-solvent, we obtained the desired product 9a in 17% yield, experiencing substantial deiodination of the starting material (entry 1 in Table 2).Switching to DMAc or DMF suppressed the side reaction and also enhanced the yield of product 9a to 55 and 53%, respectively.The best results were obtained using  dioxane, leading to a 60% yield (entries 2−4 in Table 2).With regard to the base, Et 3 N favored the deiodination and the application of K 3 PO 4 led to only moderate yields (entries 5 and 6 in Table 2).Further experiments revealed that KOAc and KF are both interchangeable with K 2 CO 3 , giving flexibility in synthetic design (entries 7 and 8 in Table 2).Eventually, the best results were obtained when we adjusted the reaction temperature and time to 65 °C and 1 h, leading to a 63% yield of product 9a (entry 9 in Table 2).Aside from phenylboronic acid, the optimized method tolerated a wide range of functional groups, such as alkyl, OCF 3 , CF 3 , halogen atoms, hydroxymethyl, nitrile, and formyl groups, leading to yields between 54 and 77% (9b−9j; Scheme 3).We observed excessive deiodination of the starting material 7 in the case of the 3-acrylic amide, 4-OH, and 4-NH 2 derivatives, leading to low yields (9k−9m).In the presence of ortho-CF 3 , 2,4-OMe, and bis-ortho-substituted boronic acids, the starting materials were mostly deiodinated (9n−9q).The trimethoxy and pentafluoro phenyl boronic acids inhibited the reaction, only enabling moderate conversions and no generation of the expected products 9r and 9s.
Next, we checked the HPLC−MS spectra for typical damage types and validated the feasibility of enzymatic ligation subsequent to the indole synthesis. 28Additional quantitative polymerase chain reaction (qPCR) experiments indicated insignificant damage to the DNA barcode compared to the control compound.Finally, we designed a 1 × 2 × 3 mock library to demonstrate the application of the developed method in DEL synthesis starting from compound 1a using selected building blocks of the validation phase (see the Supporting Information).
In conclusion, we disclosed the first on-DNA indole synthesis using a Sonogashira coupling and intramolecular ring-closure cascade.Iodination at the third position, followed by Suzuki coupling, resulted in the derivatization of the heterocycle.This process enables the generation of indolebased DELs functionalized at the pharmacologically most relevant positions.We believe that bridging this method with known indole modifications could expose the chemical space of therapeutically relevant indoles toward DEL applications.

■ ASSOCIATED CONTENT Data Availability Statement
The data underlying this study are available in the published article and its Supporting Information.
General methods, synthetic procedures, copies of HPLC−MS and deconvoluted MS spectra, DNA damage and qPCR evaluation, cell-based affinity screening and product characterization (PDF) ■