Purification of recombinant and human apolipoprotein A-1 using surfactant micelles in aqueous two-phase systems: recycling of thermoseparating polymer and surfactant with temperature-induced phase separation.

An effective system has been developed for purification of apolipoprotein A-1 from Escherichia coli fermentation solution and human plasma using aqueous two-phase extraction and thermal-phase separation. The system included non-ionic surfactants (Triton or Tween) and as top phase-forming polymer a random copolymer of ethylene oxide (50%) and propylene oxide (50%), Breox PAG 50A 1000, was used. The bottom phase-forming polymer was either hydroxypropyl starch, Reppal PES 100 and PES 200, or hydroxyethyl starch, Solfarex A 85. The top-phase-forming polymer and the surfactants are thermoseparating in water solution, i.e., when heated a water phase and a polymer/surfactant phase are formed. Recombinant apolipoprotein A-1, the Milano variant, was extracted from E. coli fermentation solution in a primary Breox-starch phase system followed by thermal separation of the Breox phase where the target protein was recovered in the water phase. Both in the Breox-starch system and in the water-Breox system Triton X-100 was partitioned to the Breox phase. The addition of non-ionic surfactants to the Breox-starch system had strong effect on the purification and yield of the amphiphilic apolipoprotein A-1. In a system containing 17% Breox PAG 50A 1000, 12% Reppal PES 100 and addition of 1% Triton X-100 the purification factor was 7.2, and the yield 85% after thermal separation of the Breox phase. Recycling of copolymer and surfactant was possible after thermal separation of copolymer phase. Approximately 85% of the copolymer and surfactant could be recycled in each extraction cycle. DNA could be strongly partitioned to the starch phase in the primary-phase system. This resulted in a 1000-fold reduction of E. coli DNA in the apolipoprotein A-1 solution obtained after thermoseparation. In extraction from human plasma containing low concentrations of apolipoprotein A-1, it was possible to reach a purification factor of 420 with 98% yield. By reducing the volume ratio to 0.1 Apo A-1 could be concentrated in a small volume of top phase (concentration factor 10) with a yield of 85% and a purification factor of 110.

Purification of recombinant apolipoprotein A-1Milano expressed in Escherichia coli using aqueous two-phase extraction followed by temperature-induced phase separation.

A method for purification of recombinant apolipoprotein A1 in aqueous two-phase systems has been studied. A mutant of apolipoprotein A-1, the Milano variant, was expressed in E. coli. Phase systems containing ethylene oxide (EO)-propylene oxide (PO) random copolymers have been used. These polymers are thermoseparating and have the ability to separate into one water-rich and one polymer-rich phase when heated above a critical temperature i.e. the cloud point. The filtrate from an E. coli fermentation was added to a primary aqueous two-phase system composed of an EO-PO copolymer and Reppal, which is an inexpensive hydroxypropyl starch. Apolipoprotein A-1 was partitioned to the top EO-PO copolymer phase and contaminating proteins to the bottom starch phase. The phase diagrams for Reppal PES 100-EO50PO50 (Ucon) and Reppal PES 100-EO30PO70 were determined. The effect on partitioning, when changing parameters such as polymer concentration, type of polymer, protein concentration, pH, salt concentration and volume ratio, were studied. Studies on E. coli DNA partitioning showed that DNA could be partitioned strongly to the bottom phase. An optimal system was scaled up from 5 g to 5 kg with similar degrees of purification, i.e. 2.5 and 2.7 and yields of 79% and 82% respectively. Furthermore temperature-induced phase formation was used for separation of apolipoprotein A-1 from the copolymer by raising the temperature above the copolymer cloud point; thus, recovering protein in a 'clean' water phase.

Effects of recombinant apolipoprotein A-I(Milano) on aortic atherosclerosis in apolipoprotein E-deficient mice.

BACKGROUND: We previously reported marked inhibitory effects of recombinant apolipoprotein (apo) A-I(Milano)/phospholipid complex (A-I[Milano]/PC) on neointimal lesions in balloon-injured iliofemoral arteries of hypercholesterolemic rabbits. In this study, we tested the hypothesis that apo A-I(Milano)/PC would inhibit aortic atherosclerosis in apo E-deficient mice. METHODS AND RESULTS: Thirty-five apo E-deficient mice fed a high-cholesterol diet were included in the study. Control mice were killed at 20 (n=8) or 25 (n=7) weeks. Treated mice received 18 injections of either 40 mg/kg apo A-I(Milano)/PC (n=15) or PC only (n=5) intravenously every other day from 20 weeks until death at 25 weeks. Aortic atherosclerosis was identified with Sudan IV staining. Lipid and macrophage contents of the aortic sinus plaques were measured after oil-red O and Mac-1 antibody staining, respectively, and quantified with computed morphometry. In control mice, from 20 to 25 weeks, aortic atherosclerosis increased by 59% (11 +/- 1% versus 17 +/- 5% of the aortic surface, P=.002), and lipid content increased by 45% (22 +/- 8% versus 32 +/- 6% of plaque area, P=.02) without a significant change in macrophage content (10.8 +/- 2% versus 13.2 +/- 6%). Compared with 20-week-old untreated control mice, PC only-treated mice at 25 weeks demonstrated a 32% increase in aortic atherosclerosis (11 +/- 1% versus 15 +/- 4%, P=.01) and an increase in lipid content (22 +/- 8% versus 47 +/- 3%, P<.0001) without a change in macrophage content (10.8 +/- 2% versus 11 +/- 2%). In comparison with 20-week-old untreated control mice, 25-week-old apo A-I(Milano)/PC-treated mice demonstrated no increase in aortic atherosclerosis (11 +/- 1% versus 10 +/- 4%, P=NS), a 40% reduction in lipid content (22 +/- 8% versus 13 +/- 8%, P=.01), and a 46% reduction in macrophage content (10.8 +/- 2% versus 5.8 +/- 2.9%; P=.03). Serum cholesterol levels were markedly elevated in all groups and did not change significantly with apo A-I(Milano)/PC or PC only. In vitro, apo A-I(Milano)/PC stimulated cholesterol efflux from cholesterol-loaded FU5AH hepatoma cell lines in a dose-dependent manner, whereas PC only or PC-free apo A-I(Miano) had no effect. CONCLUSIONS: Recombinant A-I(Milano)/PC prevented progression of aortic atherosclerosis and reduced lipid and macrophage content of plaques in apo E-deficient mice despite severe hypercholesterolemia. Thus, A-I(Milano)/PC may have a role in inhibiting progression and promoting stabilization of atherosclerosis.

Recombinant apolipoprotein A-I Milano reduces intimal thickening after balloon injury in hypercholesterolemic rabbits.

BACKGROUND: Several epidemiological studies have shown an inverse relation between high-density lipoprotein (HDL) cholesterol levels and coronary heart disease. Recently, observational studies have suggested a similar inverse relation between HDL and restenosis after coronary balloon angioplasty. Despite these observations, it is unclear whether this inverse relation reflects a direct vascular protective effect of HDL or apolipoprotein (apo) A-I, the major apolipoprotein component of HDL. Therefore, to determine whether HDL directly influences neointima formation, we investigated the effect of recombinant apo A-I Milano (apo A-I M), a mutant of human apo A-I with Arg-173 to Cys substitution, on intimal thickening after balloon injury in cholesterol-fed rabbits. METHODS AND RESULTS: Cholesterol feeding was initiated 18 days before injury and continued until the time of death. Eight rabbits received intravenous injections of 40 mg of apo A-I M linked to a phospholipid carrier on alternate days, beginning 5 days before and continuing for 5 days after balloon injury of femoral and iliac arteries. Eight rabbits received the carrier alone, and four received neither apo A-I M nor the carrier. Three weeks after balloon injury, apo A-I M-treated rabbits had significantly reduced intimal thickness compared with the two control groups (mean +/- SD): 0.49 +/- 0.29 versus 1.14 +/- 0.38 mm2 and 1.69 +/- 0.43 mm2, P < .002 by ANOVA). The intima-to-media ratio was also significantly reduced by apo A-I M (0.7 +/- 0.2 versus 1.5 +/- 0.5 and 2.1 +/- 0.1, P < .002 by ANOVA) compared with the two controls. The fraction of intimal lesion covered by macrophages, as identified by immunohistochemistry using macrophage-specific monoclonal antibody, was significantly less in apo A-I M-treated rabbits compared with carrier-treated animals (25.3 +/- 17% versus 59.4 +/- 12.3%, P < .005). Aortic cholesterol content, measured in an additional 10 rabbits, did not differ significantly between apo A-I M-treated animals (n = 5) and carrier-treated controls (n = 5). CONCLUSIONS: Apo A-I M significantly reduced intimal thickening and macrophage content after balloon injury in cholesterol-fed rabbits without a change in arterial total cholesterol content. Although the precise mechanism of action remains to be defined, these findings are consistent with a direct vascular effect of apo A-I, which could have potential therapeutic implications.