Extraction of Lipids from Human Blood Plasma Using Compressed Gases

Lipids were extracted from human blood plasma using suband supercritical fluids. An extraction plant was specifically designed to allow the treatment of plasma with compressed propane, dimethyl ether (DME) and nitrous oxide (N2O) as solvents. Experiments were performed using suband supercritical gases as mild solvents to remove the lipid fraction without denaturing valuable proteins or affecting their activity. Liquid propane, liquid dimethyl ether, and suband supercritical nitrous oxide were investigated with regard to their impact on the lipids and proteins in blood plasma at 30 MPa and at temperatures of 25°C to 40°C. Besides the extraction with pure gases the effect of modifiers and liquid solvents were studied. The treatment of plasma with small amounts of tributyl phosphate and Triton X-100 and subsequent extraction with supercritical nitrous oxide showed promising results with respect to a complete removal of plasma lipids. Cholesterol and triglycerides could be removed to an extent of 96 percent whereas the valuable proteins were retained unaffected. Since the removed lipids were shown to cumulate in a foamy layer that is formed during the extraction with supercritical nitrous oxide it is very likely that a selective precipitation of lipoproteins takes place caused by a deformation of their hydrophilic shell. The influence on the activities of the valuable proteins is still under investigation. INTRODUCTION Proteins from human blood plasma are widely used in the treatment of life-threatening diseases. Since the supply of specific plasma proteins is not possible by means of plasma transfusion, the fractionation of human blood plasma plays an important role in therapy of hereditary or infectious diseases. Certain steps in the fractionation of blood plasma are affected by the lipids occurring in plasma in the shape of micelle-like complexes called lipoproteins. The rapid inactivation of adsorbents used in fractionation processes is attributed to the lipoproteins in blood plasma. The removal of the lipids prior to the fractionation is assumed to overcome the problem of rapid adsorbent inactivation and hence to be favourable for fractionation processes. Lipids also cause problems in storage of plasma and lyophilisation of plasma products. Hence a removal will generally lead to a decrease in perishability of plasma and less denaturing of proteins during freeze drying. It is obvious that the transport of the per definition hydrophobic lipids in the aqueous milieu of blood plasma cannot be established in their original free form and is enabled by the formation of lipoproteins. The general assembly is the same for different classes of lipoproteins. The core of the globular lipoprotein complex consists of the more hydrophobic lipids, the triglycerides and the esterified cholesterol, and is encircled by proteins, amphiphilic phospholipids and free cholesterol pointing their charged groups out towards the aqueous environment. Due to this specific structure it is evident that lipoproteins are affected by solvent-detergent treatments which are used to inactivate lipid-enveloped viruses by disrupting their lipoprotein shell whereas the biologically active plasma proteins remain unaffected. The solvent-detergent method involving the incubation of plasma with 1% tri-n-butyl phosphate (TBP) and 1% Triton X-100 at 30°C for several hours was described by Horowitz et al. [1, 2] who also remarked affections to lipoproteins. Bouzidi et al. [3] used supercritical carbon dioxide and supercritical nitrous oxide to inactivate lipid-enveloped viruses in human blood plasma and also determined denaturation of lipidic components. In this study quarantine plasma without stabilizing agents was subjected to suband supercritical nitrous oxide, to the solvent-detergent treatment described by Horowitz et al. and to combinations of both treatments, in order to determine the effect on the concentration of lipids in the treated blood plasma. MATERIALS AND METHODS Pretreatment of plasma: Thawed quarantine plasma was portioned, refrozen in acetone + dry ice and stored at –25°C. Prior to each experiment the plasma portions were thawed in a water bath at 36°C. To study the effect of modifiers 1% detergent (Triton X-100 or Tween 20) and in some experiments additionally 1% tributyl phosphate (TBP) was added to the thawed plasma according to the solvent detergent method for virus inactivation reported by Horowitz et al. [1, 2]. Analytical methods: Due to severe scattering of the analytical results with test kits for triglycerides and cholesterol, it was decided to charge a medical laboratory with the analyses. Besides the analyses for triglycerides and cholesterol in later experiments also the content of total protein (electrophoresis) and IgG (nephelometry) was analyzed. Triton X-100 was determined in the extracts by HPLC qualitatively. TBP could be determined in raffinates and in extracts with a gas chromatographic method quantitatively. Extraction with compressed gases: For the extraction with liquid and supercritical gases (below termed as solvent) two different high pressure liquid gas contactors were used. Cell 1 is a horizontal stainless steel high pressure liner of 25mm ID with sapphire windows at both ends. Plasma was pumped into the pressurized cell until half of the volume of the cell was filled with plasma. Using a high pressure gear pump the solvent was conveyed across the phase boundary surface from one end of the liner to the other while the plasma was stirred by means of a magnetic stirrer. Figure 1: High pressure extraction cells for the extraction of blood plasma with compressed gases. Cell 1 (left) and cell 2 (right). magnetic stirrer plasma solvent Cell 2 consists of a vertically placed stainless steel liner of 25mm ID with two pairs of sapphire windows. Between the pairs of sapphire windows a stainless steel gauze packing (Sulzer CY) was placed to increase the mass transfer surface. Plasma was pumped into the pressurized cell until the phase boundary line was in the middle of the lower pair of sapphire windows. Plasma was withdrawn at the bottom and fed to the top of the cell with an HPLC pump and hence circulated through the gauze packing. The solvent enters the cell directly above the lower pair of sapphire windows and is withdrawn at the level of the upper pair of windows. By this technique a countercurrent flow of solvent and plasma was established. In Figure 1 a sketch of the cells is shown. Both cells were incorporated into a closed high pressure system consisting of a 300 ml buffer autoclave, a high pressure syringe pump, and a compressor generator. The cell, the buffer autoclave and the gear pump were immersed in a water bath that was temperature controlled by means of a thermostat. The high pressure syringe pump (Isco 260 D) was connected to a pressure transducer and programmed to maintain constant pressure during sampling. The controller of the syringe pump was simultaneously used to indicate the pressure of the system. After flushing the system two times with the solvent the system was pressurized up to 5 MPa using the compressor generator. Prior to the adjustment of the final extraction pressure the plasma was pumped into the cell by a HPLC pump. When the desired extraction pressure was reached the gear pump was started to circulate the solvent through the cell. In case of using cell 1 the magnetic stirrer was started at the same time. For experiments with cell 2 the HPLC pump was started to circulate the plasma through the gauze packing. Each experiment was prepared the preceding afternoon and the solvent was circulated overnight. 30 minutes before sampling the gear pump and the magnetic stirrer were switched of and the valve for the gas phase sampling was heated to 50°C to prevent freezing due to the Joule-Thomson effect. Figure 2: System for batch extraction of human blood plasma The plasma sample was directly expanded to plastic test tubes as delivered from the laboratory performing the analyses of triglycerides and cholesterol. In some experiments the amount of gas dissolved in the plasma was evaluated by passing the expanded gas to a burette system. The sampling unit for the gas phase consisted of a trap-like sample vial which was connected to a gas meter by a hose. The solvent phase was expanded to ambient pressure into the sample vial and the gas was passed through the subsequent gas-meter to evaluate the amount of expanded gas. In some experiments an additional cooling trap immersed in an aceton+dry ice bath was used to retain residual water from the expanding gas. Experiments with Triton, Tween and TBP exhibited formation of a foam layer at the phase boundary during extraction. In order to analyze the foam, Ringer’s solution was pumped into gear pump compressor generator HP syringe pump buffer