The affinity precipitation for the isolation of biomolecules

Product isolation increasingly becomes a major bottleneck in molecular biotechnology. This is especially the case for the sensitive proteineous substances (recombinant proteins, antibodies) and lately also DNA, usually produced in low concentration. The current challenge to preparative bioseparation can be summed up as the need to perform an economically sound, high-resolution separation at large scale, while maintaining "physiological" conditions throughout. Currently the only approach, which allows the direct, selective enrichment of the product from a highly diluted and complex feed, is to use biospecific interactions, i.e. "affinity" techniques. Affinity chromatography, the best known and understood of these methods may not be ideally suited early on during the downstream process because of danger of column fouling with most raw feeds, but also because of limited scale up potential and high costs (equipment, material). A promising technique combining the affinity interaction and the precipitation technique, called affinity precipitation, significantly increases the selectivity while retaining the main advantages of the precipitation method like the relatively simple equipment requirements and the applicability at large industrial scale. Affinity precipitation should also have advantages in terms of scalability, handling, and costs. In affinity precipitation, the ability of certain water-soluble polymers to form a separate phase under particular environmental conditions is used to isolate and purify biomolecules. The precipitation requires only small changes in the environment (pH, ionic strength, temperature, light, presence of specific substances). Such polymers are often called "stimuli-responsive" polymers. PolyNIPAAm is one of the most studied thermoresponsive polymers. Such molecules typically show a critical solution temperature (CST) in aqueous solution, i.e. they are soluble in cold water, but become insoluble and precipitate once the CST is surpassed. The phenomenon is usually fully reversible and the molecules redissolve readily once the temperature is lowered again. Thermoresponsive bioconjugates carrying an affinity ligand have been suggested for the specific purification of biologicals by affinity precipitation. The involved bioconjugate is called an affinity macroligand, AML. The polymer mediates the response to the stimulus (e.g. a change in temperature). The affinity tag enables the AML to attach itself to the target molecules. Affinity precipitation is to date not an established downstream processing technique in the biotech industry. The objective of this thesis was to find potential applications for the affinity precipitation by first choosing interesting target molecules and second by comparing the affinity precipitations with well established affinity bioseparation procedures based on chromatography or magnetic beads. The biomolecules chosen as target products were poly(A) mRNA, scFv antibody phages, hemoglobin and IgG 4E11. To my knowledge, it is the first time that these biomolecules have been isolated using the affinity precipitation technique. The first application, chapter 2, is about the purification of RT-PCR competent poly(A) mRNA from crude cell lysate by affinity precipitation. An AML-precursor consisting of avidin covalently linked to polyNIPAAm was used for the recovery of poly(A) mRNA hybridised to biotinylated poly(dT)-tags from crude cell lysates (Jurkat cells) by affinity precipitation. The results of the affinity precipitation were compared to those achieved with an accepted standard purification of poly(A) mRNA using avidin-activated magnetic beads. Both yield and quality / purity of the affinity precipitated poly(A) mRNA were found to be similar or better (especially removal of rRNA) than for poly(A) mRNA prepared by the magnetic particle-based protocol, while both mRNA-isolates performed equally well in standard reverse transcriptase amplification (RT-PCR) of a β actin transcript fragment. The second application, chapter 3, is about the screening of a synthetic human antibody phage display library against the MUC-1 surface antigen using smart bioconjugates. A synthetic antibody phage library (ETH-2) was screened against the MUC-1 peptide. Three methods were used, immuno tubes coated with the peptide antigen, streptavidin coated paramagnetic particles together with a biotinylated peptide, and a stimuli-responsive conjugate of avidin and polyNIPAAm. In combination with the biotinylated MUC-1, the avidin-polyNIPAAm conjugate was used to isolate specific phages from the library in several rounds of panning, consisting of repeated thermoprecipitation / redissolution cycles. Compared to the established techniques, affinity precipitation of the phages led to a greater variety of genetically different specific phage antibodies. The third application, chapter 4, is about the use of human haptoglobin-polyNIPAAm as AML for affinity precipitation of human hemoglobin. Affinity precipitation was included as the final purification step in a hemoglobin isolation protocol from blood. The first steps of the process were realized by traditional methods (lysis of red blood cells and protein precipitation with ammonium sulphate). Five different haptoglobin-polyNIPAAm AML were constructed changing the polymer:haptoglobin coupling ratio. Affinity precipitation was compared to affinity chromatography and batch adsorption methods using two binding / elution protocols. The harsh elution conditions (one at pH below 2, the other with 5 M urea) needed for dissociation of the hemoglobin-haptoglobin complex rendered the recycling of the haptoglobin AML difficult. Nevertheless, the affinity precipitation was a suitable method for the purification of hemoglobin. The results obtained with these AML confirm the values obtained by others using current methods in regard to binding ratio between haptoglobin and hemoglobin and the association constant (in the expected range). The fourth application, chapter 5, is about the use of affinity precipitation for antibody purification. To reach this goal, five AML were constructed and tested. The best results were obtained with the Protein A-AML, with nearly the expected ratio as found in the literature. Two approaches were used for the construction of the AML and it was observed that the direct coupling of polyNIPAAm reduced the binding ratio (blocking of the binding sites responsible for the affinity interaction) compared to the AML obtained by an avidin-biotin interaction, which apparently bypasses this problem. However this last AML is bigger and more costly. A Protein A mimic was also tested in affinity precipitation but the results weren't satisfactory although not worse than those for affinity chromatography using this ligand. FLAG-tag peptide was tested in affinity precipitation / chromatography for IgG 4E11 purification but again no satisfactory antibodies recovery was obtained. It seems that stressed (partly denatured) antibodies were used, affecting the interaction between capture molecules and IgG, but also giving an underestimation of the ELISA results. However, the first results with affinity precipitation of antibodies are promising but need to be further optimization in regard to the elution conditions and the quantification methods.

Related material


EPFL authors