The reverse micelle (RM) extraction of human IgG4 monoclonal antibodies (MAbs),
humanized IgG4 MAb B72.3 Fab fragments and horse heart cytochrome-c was
investigated. The effects of system parameters on forward (FE) and backward (BE)
extraction was examined, and optimal extraction conditions were determined.
Sodium bis(2-ethylhexyl)sulfosuccinate (AOT), bis(2-ethylhexyl)phosphate (HDEHP),
isooctane and corn oil were the model surfactants and solvents. Precipitate formation
was investigated, and non-ionic (polyoxyethylene(4) lauryl ether i.e. Brij 30) and
counterionic (trioctylmethylammonium chloride i.e. TOMAC) surfactants were
assessed to reduce precipitate formation and increase extraction yields. Protein in
the precipitate was measured using acetone precipitation. Conventional RM
extraction of IgG4 for AOT- and HDEHP-isooctane resulted in yields of up to 99%
FE, 58% BE for AOT-isooctane, and 92% BE for HDEHP-isooctane. It was found for
AOT-isooctane with TOMAC for cytochrome-c 81% FE and 98% BE; and 41% FE
and 85% BE for AOT-isooctane with Brij 30 and TOMAC, 56% FE and 71% BE for
AOT-corn oil, 62% FE and 89% BE for HDEHP-isooctane, and 57% FE and 59% BE
for HDEHP-corn oil with TOMAC for Fab fragments. Hollow fibre membrane (HFM)
extraction of cytochrome-c resulted in complete FE and 40% BE with TOMAC; and
with Fab fragments resulted in 99% FE and 79% BE for AOT, and in 88% FE and
46% BE for HDEHP with Brij 30. HFM FE using two aqueous phases showed that the
quantity of Fab fragments adsorbed to the membrane surface and/or stuck in the
membrane pores were negligible. Structural analysis revealed that precipitate formation during HFM module extraction was not an issue compared to conventional
FE, where it structurally damaged and prevented the Fab fragments and MAbs from
successfully transferring into the RM phase. RM phase water content analysis from
both extraction methods confirmed that RM extraction was successful.