Nature-inspired remodeling of (aza)indoles to meta-aminoaryl nicotinates for late-stage conjugation of vitamin B3 to (hetero)arylamines

Despite the availability of numerous routes to substituted nicotinates based on the Bohlmann–Rahtz pyridine synthesis, the existing methods have several limitations, such as the inevitable ortho-substitutions and the inability to conjugate vitamin B3 to other pharmaceutical agents. Inspired by the biosynthesis of nicotinic acid (a form of vitamin B3) from tryptophan, we herein report the development of a strategy for the synthesis of meta-aminoaryl nicotinates from 3-formyl(aza)indoles. Our strategy is mechanistically different from the reported routes and involves the transformation of (aza)indole scaffolds into substituted meta-aminobiaryl scaffolds via Aldol-type addition and intramolecular cyclization followed by C–N bond cleavage and re-aromatization. Unlike previous synthetic routes, this biomimetic method utilizes propiolates as enamine precursors and thus allows access to ortho-unsubstituted nicotinates. In addition, the synthetic feasibility toward the halo-/boronic ester-substituted aminobiaryls clearly differentiates the present strategy from other cross-coupling strategies. Most importantly, our method enables the late-stage conjugation of bioactive (hetero)arylamines with nicotinates and nicotinamides and allows access to the previously unexplored chemical space for biomedical research.


Supplementary Figures and Tables (1) Investigation of hypotheses and optimizing the reaction conditions
Initial reactions were performed on the basis of the hypothesis to simulate the biosynthetic formation of kynurenine/N′-formylkynurenine via cleavage of indole rings. Note that all screening reactions were performed in a 4.0 mL sealed vial equipped with a magnetic bar and a Teflon-lined screwed cap. As shown in Supplementary Figure 1, however, there is no exact method to cleave C2=C3 bond of the indole ring for the formation of N′-formylkynurenine. Therefore, we modified the biosynthetic route of vitamin B3 and hypothesized that 3-formyl(aza)indoles can be transformed to meta-aminoaryl nicotinates by Aldol-type addition and dehydration followed by intramolecular cyclization, C-N bond cleavage and subsequent re-aromatization.

Reaction details and inference
We observed that N-acetyl-protected staring materials itself were not stable and susceptible to undergo deacetylation (based on TLC pattern). Reaction using commercially available N-acetyl- 3-formylindole and N-acetyl-3-formylazaindole (prepared using the procedure reported in the literature [1] ) with our optimized conditions showed messy TLC and 1 H NMR pattern. LC/MS also showed the presence of deacetylated product. However, we observed the formation of desired products using above reaction conditions, which supports our hypothesis of synthesis of vitamin B3 scaffold via the cleavage of (aza)indole rings.
Substrate for optimization: Arylsulfonyl scaffold is a well-known profound moiety in a number of pharmaceutically active compounds. We observed that 7-azaindole scaffold is more reactive (with minimal byproduct formation) than the corresponding indole substrates. Hence, we choose Nphenylsulfonyl-7-azaindole-3-carboxaldehyde (1a) as a suitable substrate for the optimization study and further substrate scope study.

(3) Reaction optimization
All screening reactions were performed in a 4.0 mL sealed vial equipped with a magnetic bar and a Notes: a Isolated yield given in the parenthesis is the average yield from at least 3 independent runs. b The reaction was not complete. c Byproduct B was not detected.
Under the standard reaction condition developed in Supplementary Table 1 (Entry 11), we observed the formation of undesired byproduct B in relatively high amount (~10-15%) and inseparable due to their similar polarity. Therefore, we performed solvent screening study for CH3CN vs. EtOH.
We generally observed that EtOH was generally the best solvent for complete conversion of starting material and the minimal formation of undesired byproducts, compared to CH3CN, except the case of N-sulfonyl protected indole substrates. CH3CN was selected as a good solvent for all N-sulfonylprotected indole-3-carboxaldehydes. To avoid transesterification, MeOH was used as a solvent in the case of methyl propiolate and EtOH was used as a solvent in the case of ethyl propiolate. CH3CN is used as solvent for all other propiolates.

b
Zn (OTf)2 (10 mol%), NH4OAc (5.0 equiv.), 6 h 76% (90%) Notes: a 1 H NMR yields obtained without catalyst were not consistent. b Isolated yield given in the parenthesis is the average yield from at least 3 independent runs.  Notes: a Generally all the reactions using ethanol produced undesired byproducts. b 1 H NMR yields obtained without catalyst were not consistent and messy. c Isolated products were not clean and contains inseparable byproducts. d Isolated yield given in the parenthesis is the average yield from at least 3 independent runs.

(4) Investigation of hypothesis to generate β-aminoacrylate equivalents from propiolates
As the reaction using β-aminoacrylate was not successful, we hypothesized the in situ generation of βaminoacrylate from propiolate (Supplementary Figure 4) based on the literature evidence where NH4OAc is known to form cation complexes by activating the alkyne triple bond via a cation-π interaction [2] and on the other hand, NH4OAc is also used as an ammonium source for enamine synthesis [3] .This hypothesis was encouraged by 1  we confirmed that this biomimetic re-modeling of 3-formyl(aza)indole is proceeded via Plausible Pathway B via Aldol-type condensation followed by intramolecular cyclization and C-N bond cleavage.
In addition, as described in Supplementary Figure 3, we observed very low reactivity when βaminoacrylates were used as a reactant and the addition of acetic acid or Zn(OTf)2 did not improve the reaction yields. Interestingly, we observed a slight improvement in yields ( 1 H NMR yield; ~35%) when reaction was performed in the presence of NH4OAc (5.0 equiv.), as shown in Supplementary Figure 6.
Therefore, NH4OAc plays an important role in driving the reaction equilibrium as a catalyst for Aldoltype addition. However, the reaction with β-aminoacrylates is still not efficient, compared to our optimized protocol using propiolates and NH4OAc via in situ generation of structural equivalents βaminoacrylate. Figure 6. Reaction using E/Z mixture of β-aminoacrylate in the presence of

NH4OAc
The above observation in Supplementary Figure 6 provides some clues for the overall mechanism of this transformation. As shown in Supplementary Figure 7, this transformation was initiated via two possible routes of Aldol-type addition, one by the addition of activated propiolates to 3- in parallel under identical reaction conditions featuring a short reaction time of 30 min to examine the initial rate of product formation, which may disclose the electronic effects of substituents on the reactivity of substrates. As shown in Supplementary Table 5, azaindole-3-carboxaldehyde 5a is more reactive than the corresponding indole-3-carboxaldehyde 5f (Entry 1), and C5-nitro-substituted indole-3-carboxaldehyde is more reactive than the corresponding methoxy-substituted one (Entry 2). These results indicate that the presence of EWG on the indole ring increases the reactivity of the substrates which is possibly due to the increased rate of C2-N bond cleavage. Similarly, the low reactivity of 1a might be due to the decreased rate of intramolecular cyclization which can be influenced by the density of lone pair electrons on indole nitrogen. Surprisingly, the C-4 substituted indole-3-caroboxaldehydes showed exceptionally high reactivity irrespective of electronic natures of substituent (Entry 4, Supplementary Table 5). This exceptional variation might be due to the influence of their torsional strains on the reaction rate as shown in the schematic representation. with an YMC-Pack silica column (SL12S05-2520WTX, 250 mm × 20 mm, 5 μm).

Materials
Supplementary Figure 8. Synthesis of starting materials azaindole-3-carboxaldehydes General Procedure: Substituted 3-formylazaindoles were synthesized using the general procedure reported in the literature [4] . The mixture of azaindole derivative, hexamethylenetetramine (HMTA, 1.1 equiv.), AcOH and water was heated at 120 °C for 12 h in a round bottom flask equipped with a magnetic bar and a reflux condenser. X mL of water and X/2 mL of AcOH used for X mmol of azaindole.
Precipitation was observed upon cooling the reaction mixture to room temperature (r.t.) and diluted with cold water. The precipitate was filtered, washed with cold water and hexane, and dried under the reduced pressure to afford the expected product. When precipitation was not observed especially with the small-scale reaction, the crude compound was extracted into ethyl acetate (EtOAc) or 10-20% DCM/MeOH and purified by silica-gel flash column chromatography. Acetal formation and product degradation were observed when MeOH/DCM mixture was used as an eluent for silica-gel flash column chromatography.

Supplementary Figure 9. Formylation of indole substrates
General Procedure: Substituted 3-formylindoles were synthesized using the general procedure reported in the literature [5] . To a stirred solution of an appropriate indole derivative in anhydrous dimethylformamide (DMF, 3.0-4.0 mL/g of indole) under argon atmosphere was slowly added phosphorous(V) oxychloride at 0 o C. The reaction mixture was brought to r.t. and stirring was continued for 45 min to 2 h. After completion of the reaction monitored by TLC and LC/MS, the reaction mixture was slowly poured into the cold aqueous saturated NaHCO3 solution and stirred for 15-30 min. The solid precipitated upon neutralization was filtered and the residue was washed with water followed by hot hexane to afford the desired product. When precipitation was not observed especially with the small scale reaction, the crude compound was extracted into EtOAc and purified by silica-gel flash column chromatography. Acetal formation and product degradation was observed when MeOH/DCM mixture was used as an eluent for silica-gel flash column chromatography.

Compound 5ag′: Butyl 1H-indole-5-carboxylate
Butyl 1H-indole-5-carboxylate 5ag′ (1g, 74% yield, off-white solid) was prepared by the following procedure reported in the literature [6] , from indole-5-carboxylicacid (1g, 6.21 mmol). 1   suspension was slowly added appropriate arylsulfonyl chloride (1.1 equiv.), and the reaction mixture was stirred at r.t. for 3-4 h. The progress of reaction was monitored by TLC and LC/MS. After completion of the reaction, the solvent was evaporated to provide a light brown solid residue which was washed with hexane and filtered through a sintered funnel. The solid residue was then given warm water wash followed by hexane, and dried under the reduced pressure to obtain the desired N-sulfonyl protected product. About 10 mL of CH2Cl2 was used for 1.0 g of (aza)indole-3-carboxaldehydes and about 100 mL of hexane and water used for washing the residue. When a gummy solid was obtained after evaporation of CH2Cl2 in vacuo, the crude compound was purified by silica-gel flash column chromatography using EtOAc/hexane mixture as an eluent. Compound 1aj: N- pyridin-1-yl)sulfonyl)phenyl)acetamide Following the general procedure described above with 1H-pyrrolo [2,3b]

Supplementary Figure 11. Synthesis of benzyl-protected (aza)indole-3-carboxaldehydes
General procedure: A round bottom flask placed under the ice bath, equipped with a magnetic bar, charged with (aza)indole-3-carboxaldehyde in anhydrous DMF was added NaH (2 equiv., 60% suspension in paraffin oil) in a portion-wise under argon atmosphere. After complete addition of NaH, the reaction mixture was warmed to r.t. and stirred for 30-45 min. The reaction mixture was then cooled to 0 °C, an appropriate benzyl bromide derivative (1.1 equiv.) was added dropwise (in case of liquid), the reaction mixture was allowed to attain r.t. and stirred for 3-4 h. The progress of the reaction was monitored by TLC and LC/MS. After completion of the reaction, it was quenched by ice cold water (20 mL), the precipitate obtained was filtered, washed with warm water followed by hot hexane, and dried under the reduced pressure to afford the desired product. About 5.0 mL of anhydrous DMF was used for a 1 g of (aza)indole-3-carboxaldehyde. When benzyl bromide derivative is solid, it was added by dissolving in a minimum amount of dry DMF and the crude product obtained after filtration was purified by silica-gel flash column chromatography using EtOAc/hexane mixture as an eluent.

Supplementary Figure 12. Synthesis of N-phenyl-protected azaindole-3-carboxaldehyde (1i)
Experimental procedure: The desired product was prepared by following the procedure reported in the literature [7] . In a 10 mL microwave vial equipped with magnetic stir bar and a Teflon-lined screw cap was added azaindole-3-carboxaldehyde (5a, 300 mg), KCl (1 equiv.), K2CO3 (3 equiv.), bromobenzene (1.1 equiv.), CuI (0.1 equiv.) and DMF (6 mL). The reaction mixture was continuously stirred under microwave irradiation for 4 h while temperature was maintaining at 100 o C. After completion of the reaction monitored by TLC and LC/MS, the reaction mixture was diluted with saturated aq. NH4Cl solution (10 mL) and the crude compound was extracted using EtOAc (2 × 10 mL). The combined organic extracts were washed with brine (10 mL), dried over anhydrous Na2SO4(s), filtered, and concentrated under the reduced pressure. The crude product obtained was then purified by silica-gel flash column chromatography using EtOAc/hexane (1:3) as an eluent to afford the desired starting material 1i in 65% yield (300 mg) as off-white solid.
After complete addition of NaH, the reaction mixture was warmed to r.t. and stirred for 30 min. The reaction mixture was cooled to 0 °C, 2-chloropyrimidine (1.1 equiv.) was added dropwise, the reaction mixture was allowed to attain r.t. and then heated at 100 °C for 12 h. The progress of the reaction was monitored TLC and LC/MS. After completion of the reaction, it was quenched by ice cold water (20 mL), the precipitate obtained was filtered, washed with warm water (50 mL) followed by hot hexane (30 mL), and dried under the reduced pressure to afforded the desired product (1k) in 350 mg with 57% yield as off-white solid.
Step1: To a stirred solution of 2-thienyl methanol (500 mg) and Et3N (1.2 equiv.) in DCM (6 mL) at 0 o C, was slowly added thionyl chloride and the reaction mixture was warmed to 30 °C. After being stirred for 2 h, the reaction mixture was neutralised with saturated aqueous NaHCO3 solution (10 mL) and crude compound was extracted with DCM (2 × 10 mL). The combined organic extracts were washed with brine (10 mL), dried over anhydrous Na2SO4(s), and filtered and concentrated in vacuo to afford crude (2-chloromethyl)thiophene. The product was a lachrymator agent in nature and used in the next step without further purification and characterization [8] .
Step 2: To a well stirred, ice-cold solution of (aza)indole (400 mg) in dry DMF (5 mL) was added of NaH (2 equiv., 60% suspension in paraffin oil) under argon atmosphere. The reaction mixture was then allowed to attain r.t. and stirred for 30 min. The reaction mixture was then cooled to 0 °C, added dropwise with the solution of crude (2-chloromethyl)thiophene (1.1 equiv.) in 1 mL of DMF obtained from the step 1, and allowed to attain r.t. After being stirred for 2 h, the reaction mixture was quenched with ice-cold water (50 mL) and the crude compound was extracted into EtOAc (2 × 20 mL). The combined organic extract was washed with brine (20 mL), dried over anhydrous Na2SO4(s), concentrated in vacuo, and purified by silica-gel flash column chromatography using EtOAc:hexane mixture (3:7) as an eluent to furnish the desired product.
After completion of the reaction, the reaction mixture was cooled to r.t., diluted with saturated aqueous NaHCO3, filtered through a short pad of celite, and washed with EtOAc (10 mL). The crude compound was then extracted from filtrate using EtOAc (3 × 20 mL), the combined organic extract was given a brine wash (3 × 10 mL), dried over anhydrous Na2SO4(s), and filtered and concentrated under the reduced pressure. The crude compound was purified by silica-gel flash column chromatography (EtOAc:hexane = 3:7) to provide the desired product (5z-5ac) as an off-white solid.

Method B:
In a dried round bottom flask equipped with a magnetic bar, the equimolar mixture of prop-2-ynoic acid, appropriate compound bearing hydroxyl group, and DMAP were dissolved in dry DCM.
To this reaction mixture was added DCC (1.1 equiv.) under argon atmosphere at r.t., and the reaction mixture was stirred for 2 h. The progress of the reaction was monitored by TLC via observing the complete consumption of alcohol. After completion of the reaction, the reaction mixture was passed through a short pad of celite, washed with DCM, and subjected to silica-gel flash column chromatography to obtain the desired propiolate.

Supplementary Figure 28 Ester hydrolysis of nicotinates to nicotinic acids
Experimental procedure: To a mixture of appropriate ethyl nicotinate derivative and K2CO3 (2.0 equiv.) was added a 1:1 mixture of EtOH/water and then the reaction mixture was heated at 100 °C for 2 h. The progress of the reaction was monitored by TLC and LC/MS. Upon complete consumption of starting material, the solvent was evaporated under the reduced pressure, the crude compound was diluted with water (10.0 mL) and washed with diethyl ether (2 × 10 mL). The desired product was precipitated out upon neutralizing the aqueous layer with saturated citric acid solution (pH 5-7). The precipitate was filtered, washed with water followed by hexane, and dried in vacuo to obtain the desired product.