Ethyl (1S,3Ar,6As)-Octahydrocyclopenta[C]Pyrrole-1-Carboxylate Hydrochloride (1:1)

The hydrochloride salt is introduced directly into a paritaprevir (ABT-450) commercial manufacturing route at the P2 fragment elaboration stage. Before coupling, the ethyl ester is hydrolyzed to the corresponding (1S,3aR,6aS)-octahydrocyclopenta[c]pyrrole-1-carboxylic acid while liberating the secondary amine. In a standard production-scale protocol conducted in a 500 L glass-lined reactor under nitrogen, 1.0 eq of the hydrochloride ester is suspended in 2.5 volumes of deionized water and 2.5 volumes of tetrahydrofuran at 20 °C. An aqueous solution of 2.2 eq sodium hydroxide is added dropwise, maintaining the internal temperature below 25 °C to minimize base-catalyzed epimerization at the α-carbon of the ester. Hydrolysis is monitored by reverse-phase HPLC (C18 column, 210 nm) until residual ester content drops below 0.5 % area. The mixture is then acidified to pH 2.0–3.0 with concentrated hydrochloric acid at 0–5 °C, precipitating the free amino acid as a white solid. The precipitate is filtered, washed with chilled deionized water, and vacuum-dried at 40 °C for 12 h under ≤10 mbar to a moisture content below 0.3 % as determined by Karl Fischer titration. Pre-drying of the starting hydrochloride is mandatory when ambient relative humidity exceeds 60 % because the amino acid intermediate exhibits noticeable hygroscopicity after neutralization, leading to inaccurate mass balance and reduced coupling efficiency in the subsequent amidation step.

The dried amino acid is then coupled to the paritaprevir macrocyclic amine intermediate in dichloromethane (8 volumes) using 1.15 eq of N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride and 1.15 eq of 1-hydroxybenzotriazole hydrate, with 2.2 eq of N,N-diisopropylethylamine as base. Activation proceeds at 0–2 °C for 40 min, after which the amine component is added and the reaction mass is warmed to 25 °C over 3 h. A single-pass conversion of ≥92 % is typical; the crude product is isolated by solvent switch to ethyl acetate and sequential washes with 5 % aqueous citric acid, saturated sodium bicarbonate solution, and brine. Residual dichloromethane in the final paritaprevir drug substance must comply with ICH Q3C Option 2 limits (600 ppm). Enantiomeric purity of the coupled product is verified by chiral supercritical fluid chromatography (Chiralpak AD-H, 40 °C, 3 mL·min⁻¹, 220 nm) with an acceptance criterion of <0.15 % of the diastereomer originating from the (1R,3aS,6aR)-epimer. The bulk paritaprevir is further formulated into fixed-dose combination products designated under the Viekira Pak label; consequently, the intermediate manufacturing process must be operated under ICH Q7 GMP for active pharmaceutical ingredients starting materials from the stage at which the chiral octahydrocyclopenta[c]pyrrole fragment is introduced. Residual elemental impurities are monitored per ICH Q3D: palladium (≤10 ppm) and iron (≤100 ppm) are routinely tested when upstream hydrogenation steps that use transition-metal catalysts are part of the registered route. Avoid combination of the free amino acid with excess strong base or prolonged heating above 30 °C prior to coupling because racemization can occur at the C-1 carboxylate position, and the resulting diastereomeric content can exceed the 0.15 % threshold, rendering the batch unsuitable for subsequent crystallization-based purging.

Glecaprevir Assembly via Pivaloyl Mixed Anhydride

When the same bicyclic scaffold is required for glecaprevir (ABT-493), a mixed anhydride activation strategy is frequently selected to suppress racemization under strictly anhydrous conditions. The hydrochloride ethyl ester undergoes an identical hydrolysis–precipitation sequence to deliver the free amino acid, which is then subjected to azeotropic drying in tetrahydrofuran (6 volumes) at 35–40 °C under reduced pressure until the water content falls below 500 ppm. The dried acid is suspended in anhydrous tetrahydrofuran and treated sequentially with 1.05 eq of pivaloyl chloride and 1.10 eq of N-methylmorpholine at –18 to –15 °C in a 316L stainless-steel jacketed vessel. Mixed anhydride formation is confirmed by in-line FTIR monitoring of the carbonyl stretching band shift; the reaction is held for 25–30 min before the glecaprevir amine segment, dissolved in tetrahydrofuran, is metered in over 45 min while maintaining the internal temperature below –10 °C. Excess pivaloyl chloride must be avoided because residual pivalic acid esters can form persistent crystalline solvates with the API that require intensive reslurry purification. After coupling, the mixture is quenched with 5 % aqueous potassium bicarbonate and extracted into isopropyl acetate. The organic layer is concentrated and the glecaprevir free base is crystallized from n-heptane–isopropanol (4:1 v/v). The isolated free base is subsequently converted to the mesylate salt in acetone (3 volumes) with 1.02 eq of methanesulfonic acid at 20–25 °C. Chiral HPLC of the mesylate (Chiralcel OJ-RH, 35 °C, 1.0 mL·min⁻¹) limits the undesired diastereomer to ≤0.10 % area. Because the glecaprevir mesylate is a non-hygroscopic crystalline solid, the final drying step operates at 50 °C and 5 mbar for 8 h, achieving residual acetone below 5000 ppm as per ICH Q3C. The entire process, from free amino acid isolation to mesylate formation, must be executed under a nitrogen atmosphere because the mixed anhydride intermediate hydrolyzes rapidly at relative humidity above 30 %, causing a cascade of side reactions that lower the yield by 15–20 % and generate an N-pivaloyl amide impurity that is difficult to purge below the ICH Q3A reporting threshold of 0.05 %.

In the manufacture of danoprevir (RG7227), the bicyclic amino acid derived from the hydrochloride serves as the chiral P2 anchor before macrocyclization. After ester hydrolysis and neutralization, the free amino acid is protected as its N-allyloxycarbonyl derivative by treatment with allyl chloroformate (1.15 eq) in two-phase THF–aqueous sodium carbonate at 0–5 °C. The N-Alloc amino acid is activated as the pentafluorophenyl ester with 1.05 eq of pentafluorophenol and 1.2 eq of N,N′-diisopropylcarbodiimide in ethyl acetate, isolated, and coupled to the linear peptide segment carrying a terminal olefin. Ring-closing metathesis is conducted in toluene at 60 °C with Grubbs second-generation catalyst (1 mol%) to form the 15-membered macrocycle. Ruthenium removal to single-digit ppm levels is accomplished by treatment with activated carbon (10 % w/w) and triphenylphosphine oxide at 50 °C for 4 h, followed by filtration through a 0.2 µm PTFE membrane. The Alloc group is cleaved under palladium-catalyzed conditions using phenylsilane (2.5 eq) and tetrakis(triphenylphosphine)palladium(0) (0.02 eq) in dichloromethane. Palladium content in the final danoprevir free base is quantified by inductively coupled plasma mass spectrometry and must comply with the ICH Q3D oral concentration limit of 10 µg·day⁻¹, translating to ≤10 ppm at a 100 mg·day⁻¹ dose assumption. The hydrochloride salt of the starting material must be stored at 2–8 °C in sealed, polyethylene-lined aluminium pouches; exposure to ambient air for more than 2 h on the shop floor during dispensing leads to moisture uptake exceeding 0.5 % and detectable hydrate formation that alters the ester hydrolysis rate reproducibility. Operational personnel must pre-dry ancillary equipment because residual moisture in charging ports or transfer lines systematically depresses the pentafluorophenyl ester formation conversion below 95 % when humidity exceeds 55 %.

How Does Replacing a Proline Ring with an Octahydrocyclopenta[c]pyrrole Scaffold Impact Enzymatic Stability?

Medicinal chemists routinely substitute the pyrrolidine ring of L-proline with the (1S,3aR,6aS)-octahydrocyclopenta[c]pyrrole framework to restrict conformational flexibility and shield a peptide bond from enzymatic hydrolysis. The hydrochloride is converted into an Fmoc-protected amino acid building block suitable for solid-phase peptide synthesis (SPPS). In a typical derivatization, the ethyl ester is hydrolyzed as described previously, and the resulting amino acid is dissolved in a 1:1 (v/v) mixture of 0.5 M aqueous sodium carbonate and dioxane. Fmoc-N-hydroxysuccinimide ester (1.05 eq) is added in three portions at 0 °C over 1 h, and the mixture is stirred for another 3 h at 20 °C. After acidification and extraction into methyl tert-butyl ether, the Fmoc-protected monomer is purified by silica gel chromatography (eluent: hexane–ethyl acetate 3:2 containing 0.5 % acetic acid) and isolated as a white foam with a chemical purity of ≥98.5 % by HPLC. On a 2-chlorotrityl chloride resin preloaded at 0.3–0.8 mmol·g⁻¹, the Fmoc-amino acid is coupled using 3 eq of monomer, 3 eq of N,N′-diisopropylcarbodiimide, and 3 eq of ethyl (hydroxyimino)cyanoacetate in DMF, achieving a single coupling efficiency of >98 % as judged by the Kaiser test. Multiple insertions of this constrained residue produce peptide analogues where the amide bond adjacent to the bicyclic nitrogen adopts a preferred trans configuration, as confirmed by ROESY NMR. Literature pharmacokinetic studies indicate that replacement of a native proline with the octahydrocyclopenta[c]pyrrole residue extends the plasma elimination half-life of susceptible linear peptides from <0.5 h to >4 h in rodent models, primarily due to resistance toward prolyl oligopeptidase and dipeptidyl peptidase-IV. When the Fmoc derivative is supplied to peptide CDMOs, the excipient list must specify residual dioxane below 380 ppm and Fmoc-β-alanine-like impurities below 0.2 %; elevated Fmoc-β-alanine content causes premature chain termination in long sequences, lowering the crude peptide purity by 5–10 %. The Fmoc protection step is incompatible with the presence of primary amines or unprotected lysine side chains in the same vessel, and even trace ammonium ions from buffer carry-over provoke Fmoc-deprotection that yields a truncated sequence.

Direct liberation of the secondary amine from the hydrochloride under phase-transfer conditions yields a neutral catalyst investigated for stereoselective carbon-carbon bond formation. The hydrochloride is suspended in dichloromethane (10 volumes) and stirred with 1.5 eq of powdered potassium carbonate and a catalytic quantity of tetra-n-butylammonium bromide (0.05 eq) for 2 h at 20 °C. Filtration and evaporation provide the free amine as a pale-yellow oil that is stored over 4 Å molecular sieves. This free amine catalyzes the direct aldol reaction between cyclohexanone (10 eq) and 4-nitrobenzaldehyde (1.0 eq) in dimethyl sulfoxide at –18 °C with a catalyst loading of 10 mol%. After 72 h, the anti-aldol product is obtained in 88 % isolated yield with an enantiomeric excess of 94 % as determined by chiral HPLC (Chiralpak IA, hexane–isopropanol 90:10). A temperature rise to –10 °C drops the ee to 82 %, and operation above 0 °C renders the process non-selective (ee <50 %), placing the useful processing window at ≤–15 °C with a tolerance of ±3 °C. Michael addition to β-nitrostyrene using the same catalyst (15 mol%) in chloroform at 0 °C yields the 1,4-adduct with 91 % ee after 48 h. The free amine must be handled under argon because the unhindered secondary amine absorbs carbon dioxide from air forming a carbamate salt, which precipitates as a viscous mass and deactivates the catalyst within 10–15 min of atmospheric exposure. Scale-up experiments in a 100 mL jacketed glass reactor with a PTFE-coated temperature probe show that batch-to-batch variability in enantioselectivity can be minimized to ±2 % ee when the water content of the reaction medium is maintained below 200 ppm; above 500 ppm water, the ee drops to 75–80 % and the anti/syn ratio shifts from 12:1 to 6:1.