Chromatography-free route via an acylaminoisoxazole
8 stages · 31% overall · no column at any stageThe imidazole in sabizabulin is the hard part. Earlier routes built it directly and paid for it: one needed t-BuLi at −78 °C and finished at 0.4% overall, another ran ten steps with OsO4 and three protecting-group operations for 3.8%. This process never builds an imidazole at all. It builds an isoxazole instead, early and cheaply from a nitroketone, carries it through the whole sequence as a masked imidazole, and unmasks it at the end: hydrogenation snaps the weak N-O bond to give an enaminone, which closes onto the amide nitrogen to form the imidazole ring. Nothing is protected, nothing is chromatographed, and every intermediate is isolated by filtration or crystallisation. The first two steps here are Scheme 3 of the paper, the rest is Scheme 9.
Scale: 74.3 g of the starting acid; 51 g of nitroketone into the second half; 24.7 g of sabizabulin out
Step 1
98% isolated
SOCl2 (3 equiv), DMF (cat.), CH2Cl2, 40 °C, 2 h
then imidazole (1 equiv), Et3N (1 equiv), CH2Cl2, 0 to 20 °C
then imidazole (1 equiv), Et3N (1 equiv), CH2Cl2, 0 to 20 °C
the acid is activated as an acylimidazole, not as the acid chloride · imidazole was chosen over benzotriazole or a phenyl ester because imidazole is water soluble, so the excess simply washes out · no column, 90 g from one flask
no public CAS
no public CAS
Step 2
98% isolated
MeNO2 (1.2 equiv), t-BuOK (1.2 equiv), DMSO, 10-20 °C, 4 h
then quench into ice, water and HCl
then quench into ice, water and HCl
C-acylation of nitromethane · the potassium nitronate is made first at 16-18 °C, then the solid acylimidazole is added in one portion · quenching into ice and acid precipitates the product, so it is filtered rather than extracted · use the acylimidazole immediately, it hydrolyses slowly even at 4 °C
no public CAS
no public CAS
Step 3
77% by mass, used crude
DMF-DMA (1.5 equiv), CH2Cl2, 20 °C, 30 min
then crystallised by adding methanol
then crystallised by adding methanol
dimethylaminomethylenation, and the first place the process earns its keep · the side product is methyl 3,4,5-trimethoxybenzoate, which stays dissolved in methanol while the enaminone crystallises out · removing that ester matters: leave it in and the next step's yield drops, while the coloured impurities alongside it turn out to be harmless
no public CAS
no public CAS
Step 4
61% over two steps, 34.0 g
NH2OH·HCl (1.2 equiv) added over 20 min
NaOAc/AcOH buffer at pH 5.5, H2O-MeOH 3:1, 50 °C
NaOAc/AcOH buffer at pH 5.5, H2O-MeOH 3:1, 50 °C
the isoxazole ring, and the step that needed the most work · the starting material decomposes in acid and the product decomposes in base, so the window is narrow: yield peaks at pH 5.2-5.6 and falls away on either side · the hydroxylamine goes in as a solution, dropwise, into a pre-equilibrated buffered suspension, which avoids local pH excursions · homogeneous solvent only, because a two-phase mixture cannot be buffered properly
no public CAS
no public CAS
Step 5
84%, 25.6 g
Zn (4.5 equiv), NH4Cl (10.7 equiv), LiCl (1 equiv)
MeOH-H2O, 3-5 °C, then warm to room temperature
MeOH-H2O, 3-5 °C, then warm to room temperature
nitro to amine on an intact isoxazole · temperature is the whole story: at 20-30 °C the exotherm drives side products and the yield sits at 63%, at 0-5 °C it reaches 85% · lithium chloride adds about ten points on its own · the product is filtered off through Celite with the zinc salts and then washed off the cake with dichloromethane, which is what replaces the column
no public CAS
no public CAS
Step 6
89%, 35.7 g
indole-3-carboxylic acid (13) + SOCl2, DMF (cat.), CH2Cl2, 40 °C
then the amine with pyridine (1.15 equiv), CH2Cl2, 20 °C, 1 h
then the amine with pyridine (1.15 equiv), CH2Cl2, 20 °C, 1 h
the indole half is joined on as an acid chloride · pyridine rather than triethylamine or DIPEA, because a stronger base opens the isoxazole ring · dichloromethane is a poor solvent for hydrogen-bonded aggregates, so the amide precipitates during the aqueous workup and is simply filtered
no public CAS
no public CAS
Step 7
97% assay-corrected, used crude
H2 (1 atm), 10% Pd/C (4 mol%), EtOH
20 °C for 2 h, then 50 °C for 30 min
20 °C for 2 h, then 50 °C for 30 min
the mask comes off · hydrogenolysis breaks the weak N-O bond of the isoxazole and nothing else, revealing the enaminone that will become the imidazole · balloon pressure is enough · the crude product is carried straight into the last step, which is the point of doing it this way round
no public CAS
no public CAS
Step 8
72% over two steps, 24.7 g
K2CO3 (30 equiv), MeOH, reflux, 8 h, 0.025 M
then crystallised as the methanol solvate
then crystallised as the methanol solvate
the imidazole closes · two numbers decide this step and both are unusual: thirty equivalents of carbonate, and a dilution of 0.025 M · at 0.1 M the yield is 42%, at 0.025 M it is 72%, because the competing reaction is methoxide cleaving the enaminone apart into indole-3-carboxamide and trimethoxybenzoate · stronger bases work but destroy the substrate, organic bases do not work at all · the final purification is a crystalline methanol solvate that barely dissolves in methanol while the impurities do
31% overall yield vs 0.4 to 15.9%
none chromatography vs every earlier route
none protecting groups vs two of three routes
99.2% final purity vs by LC-UV
Against the three published routes the authors tabulate: Chen (5 steps, 0.4%), Wang (10 steps, 3.8%) and Peng (5 steps, 15.9%). All three require chromatography; the first two also require protecting groups. Yields here are isolated.
A Practical and Chromatography-Free Process for Sabizabulin (VERU-111) via an Acylaminoisoxazole Intermediate
Org. Process Res. Dev. 2026, accepted 21 October 2026 · Schemes 3 and 9, Tables 3-5
