Recovery of lactoferrin and lactoperoxidase from sweet whey using colloidal gas aphrons (CGAs) generated from an anionic surfactant, AOT.

The recovery of lactoferrin and lactoperoxidase from sweet whey was studied using colloidal gas aphrons (CGAs), which are surfactant-stabilized microbubbles (10-100 microm). CGAs are generated by intense stirring (8000 rpm for 10 min) of the anionic surfactant AOT (sodium bis-2-ethylhexyl sulfosuccinate). A volume of CGAs (10-30 mL) is mixed with a given volume of whey (1-10 mL), and the mixture is allowed to separate into two phases: the aphron (top) phase and the liquid (bottom) phase. Each of the phases is analyzed by SDS-PAGE and surfactant colorimetric assay. A statistical experimental design has been developed to assess the effect of different process parameters including pH, ionic strength, the concentration of surfactant in the CGAs generating solution, the volume of CGAs and the volume of whey on separation efficiency. As expected pH, ionic strength and the volume of whey (i.e. the amount of total protein in the starting material) are the main factors influencing the partitioning of the Lf.Lp fraction into the aphron phase. Moreover, it has been demonstrated that best separation performance was achieved at pH = 4 and ionic strength = 0.1 mol/L i.e., with conditions favoring electrostatic interactions between target proteins and CGAs (recovery was 90% and the concentration of lactoferrin and lactoperoxidase in the aphron phase was 25 times higher than that in the liquid phase), whereas conditions favoring hydrophobic interactions (pH close to pI and high ionic strength) led to lower performance. However, under these conditions, as confirmed by zeta potential measurements, the adsorption of both target proteins and contaminant proteins is favored. Thus, low selectivity is achieved at all of the studied conditions. These results confirm the initial hypothesis that CGAs act as ion exchangers and that the selectivity of the process can be manipulated by changing main operating parameters such as type of surfactant, pH and ionic strength.

Studies on the purification of peroxidase from horseradish roots using reverse micelles.

Horseradish peroxidase (HRP) was successfully purified from horseradish roots by a two-stage reverse-micellar extraction from the dialyzed aqueous extract. The anionic surfactant AOT dissolved in isooctane was used to produce the reverse-micellar phases. The narrow pH range at which HRP solubilization occurred was exploited to remove most of the contaminant proteins in the first forward extraction. In the second extraction stage, HRP was selectively solubilized and concentrated by using a volume ratio of 10 between the aqueous and organic phases. The HRP final specific activity was 86 guaiacol U mg-1, obtained with a purification factor of 80 and yield of 46%. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis showed two overlapping bands, with HRP corresponding to that at 43.8 kDa. Image analysis on isoelectric focusing (IEF) gels showed that the HRP was 80% pure. Ion exchange liquid chromatography showed that most of the specific activity was due to the basic isoenzyme with pI 8.5, which comprises 33.5% of the product. There were high HRP losses as a precipitate at the interface when direct reverse-micellar extraction was attempted from the crude extract. It is believed that the hydrophobic environment near the haem group of the HRP basic isoenzyme favors complex formation with the surfactant, and that this is promoted at higher protein concentrations.

Adsorption of Kluyveromyces marxianus pectinase on CM-Sephadex gels.

New results are presented on the equilibria and kinetics of the ion exchange onto CM-Sephadex of polygalacturonase (pectinase) produced by the fermentation of Kluyveromyces marxianus. It is found that the equilibrium behavior follows the form of the Langmuir isotherm; the equilibrium is strongly affected by pH. High partitioning onto the ion-exchange matrix, with good retention of enzyme activity, is achieved in the pH range 3.5-5.0, and this can be qualitatively explained in terms of simple models for protein adsorption by ion exchange. The kinetics of ion exchange is modeled by assuming that the transfer resistances can be lumped into a single coefficient, and the results show that this gives a reasonable description of the adsorption kinetics. Under optimum conditions protein adsorption is enhanced by electrostatic effects and is extremely fast, and it is suggested that in these circumstances external mass transfer resistance is significant. At pH values close to the isoelectric point, electrostatic interactions are weak and intraparticle diffusion is rate-limiting: pore-blocking by adsorbed proteins appears to be important under these conditions. The results also provide the basis for an efficient single-step purification scheme.

Selective separation of beta-lactoglobulin from sweet whey using CGAs generated from the cationic surfactant CTAB.

The selective separation of whey proteins was studied using colloidal gas aphrons generated from the cationic surfactant cetyl trimethyl ammonium bromide (CTAB). From the titration curves obtained by zeta potential measurements of individual whey proteins, it was expected to selectively adsorb the major whey proteins, i.e., bovine serum albumin, alpha-lactalbumin, and beta-lactoglobulin to the aphrons and elute the remaining proteins (lactoferrin and lactoperoxidase) in the liquid phase. A number of process parameters including pH, ionic strength, and mass ratio of surfactant to protein (M(CTAB)/M(TP)) were varied in order to evaluate their effect on protein separation. Under optimum conditions (2 mmol/l CTAB, M(CTAB)/M(TP) = 0.26-0.35, pH 8, and ionic strength = 0.018 mol/l), 80-90% beta-lactoglobulin was removed from the liquid phase as a precipitate, while about 75% lactoferrin and lactoperoxidase, 80% bovine serum albumin, 95% immunoglobulin, and 65% alpha-lactalbumin were recovered in the liquid fraction. Mechanistic studies using zeta potential measurements and fluorescence spectroscopy proved that electrostatic interactions modulate only partially the selectivity of protein separation, as proteins with similar surface charges do not separate to the same extent between the two phases. The selectivity of recovery of beta-lactoglobulin probably occurs in two steps: the first being the selective interaction of the protein with opposite-charged surfactant molecules by means of electrostatic interactions, which leads to denaturation of the protein and subsequent formation and precipitation of the CTAB-beta-lactoglobulin complex. This is followed by the separation of CTAB-beta-lactoglobulin aggregates from the bulk liquid by flotation in the aphron phase. In this way, CGAs act as carriers which facilitate the removal of protein precipitate.

The effects of pH and ionic strength on the partitioning of four proteins in reverse micelle systems.

Four proteins with different physicochemical properties have been partitioned in reversed micelle systems: thaumatin, ribonuclease A, soybean trypsin inhibitor, and alpha-lactalbumin. The organic phase was formed by sodium salt (AOT) in isooctane, and the aqueous phase contained KCl, KBr, MgCl(2), or NaCl. Aqueous phase pH was varied between 2 and 13 and ionic strength from 0.1 to 1.0 M. Small changes in pH [around the isoelecric point (pl)] were found to influence the solubilization of ribonuclease A and trypsin inhibitor, but for thaumatin the pH change necessary to affect partition was much greater as a consequence of the difference in net charge (titration curves) of these protein molecules as pH changes. The type of ions present in the system was also a determining factor for partition; the larger ions (K(+)) produced more electrostatic screening and hence less protein solubilization than the smaller ions (Na(+)). With changes in ionic strength surface hydrophobicity was a dominant factor affecting solubilization of thaumatin in NaCl-containing systems at high pH. Charge distribution and hydrophobicity are thought to be important parameters when partitioning the protein alpha-lactalbumin. (c) 1994 John Wiley & Sons, Inc.

Protein extraction by reverse micelles: studies on the recovery of horseradish peroxidase.

Phase transfer studies were carried out on the solubilization of horseradish peroxidase (HRP) (E.C. 1.11.1.7) in reverse micelles formed in isooctane using the anionic surfactant, aerosol OT, at concentrations between 50 and 110 mM. The selectivity of this methodology was tested, because the HRP used comprised a mixture of seven different isoenzymes with a wide range of isoelectric points. Forward and backward transfers were carried out in wellstirred vessels until equilibrium was reached. Significant protein partitioning could only be obtained by using NaCl to adjust ionic strength in pH range between 1.5 and 3.5, with a maximum at pH 3. The back transfer process was best at pH 8 with 80 mM phosphate buffer and 1 M KCI. A loss of 1% to 3% of the surfactant through precipitation at the interface at pH<4 was observed, which may be due to instability in this pH region, because, even without protein, a similar precipitate was noticed. Protein partitioning was approximately constant when the ionic strength was increased up to 1 MNaCl at pH 3, but protein recovery in back transfer decreased accordingly. Hydrophobic interactions together with association between the protein and surfactant might be responsible for that behavior. Protein partitioning remained the same when the surfactant concentration was decreased to 50 mM, at the expense of higher variability. HPLC chromatograms showed no apparent damage to the protein after reverse micellar extraction. Protein partitioning is best when the temperature is kept at 25 degrees C. The amount of protein and specific activity recovered strongly depends on the phase ratio used during forward transfer. Overall activity recovery varied from 87% to 136% when the phase ratio was increased from 1:1 to 30:1 in forward transfer. This behavior may be due to a change in the ratio of the three isoenzymes recovered after the backward transfer process, with the most active one being increasingly enriched at higher phase ratios.

Solid-state fermentation: a continuous process for fungal tannase production.

Truly continuous solid-state fermentations with operating times of 2-3 weeks were conducted in a prototype bioreactor for the production of fungal (Penicillium glabrum) tannase from a tannin-containing model substrate. Substantial quantities of the enzyme were synthesized throughout the operating periods and (imperfect) steady-state conditions seemed to be achieved soon after start-up of the fermentations. This demonstrated for the first time the possibility of conducting solid-state fermentations in the continuous mode and with a constant noninoculated feed. The operating variables and fermentation conditions in the bioreactor were sufficiently well predicted for the basic reinoculation concept to succeed. However, an incomplete understanding of the microbial mechanisms, the experimental system, and their interaction indicated the need for more research in this novel area of solid-state fermentation.

Emulsion liquid membrane extraction of lactic acid from aqueous solutions and fermentation broth.

Studies on the batch extraction of lactic acid using an emulsion liquid membrane system are reported. The membrane phase consists of the tertiary amine carrier Alamine 336 and the surfactant Span 80 dissolved in n-heptane/paraffin and aqueous solutions of sodium carbonate in the internal phase. The effects of internal phase reagent, extraction temperature, and initial external phase pH on the extraction efficiency and the emulsion swelling are examined. A statistical factorial experiment on extraction from clarified lactic acid fermentation broth was carried out to obtain knowledge of the performance of the extraction system from a broth. The extraction efficiency from the fermentation broth is found to be lower as compared to aqueous solutions of pure lactic acid. The effect of pH and the presence of other ionic species on selectivity are discussed.

Factors affecting the esterification of lauric acid using an immobilized biocatalyst: Enzyme characterization and studies in a well-mixed reactor.

The esterification of lauric acid with geraniol catalyzed by the commercially immobilized lipase preparation from Mucor miehei, Lipozyme(R), was studied in well-stirred flasks. The enzyme support was characterized in terms of its internal and external surface area, protein location, and protein content. It was found that the enzyme was mainly located on the external surface of the support, therefore, internal diffusional limitations were not important. It was also shown that the protein content of the support depends on the size of the particle, with smaller particles containing higher amounts of protein per unit weight. Under the conditions studied, the reaction was not under external mass transfer limitations, and the initial reaction rate depended on the size of the support particles. This was mainly due to the different protein contents on the support as a function of particle size and not to internal or external mass transfer limitations. Also, it was found that the inhibition exerted by water was predominantly a physical effect due to its accumulation around the enzyme. It was also found that the reaction was substrate inhibited by lauric acid, but not by geraniol. (c) 1995 John Wiley & Sons, Inc.

Extraction of lysozyme and ribonuclease-a using reverse micelles: Limits to protein solubilization.

Experiments are reported here on the equilibrium partitioning of lysozyme and ribonuclease-a between aqueous and reversed micellar phases comprised of an anionic surfactant, sodium di-2-ethylhexyl sulfosuccinate (AOT), in isooctane. A distinct maximum, [P](rm,max) was found for the quantity of a given protein that can be solubilized in the reverse micelle phase by the phase-transfer method. This upper limit depended upon the size of the protein, the surfactant concentration, and the aqueous phase ionic strength, and was determined by complex formation between protein and surfactant molecules to form an insoluble interfacial precipitate at high values of [P](rm). In this work, it was found to be possible to dissociate the protein-surfactant complex and recover the precipitated protein. The kinetics of protein-surfactant complex formation depended upon the nature and concentration of the solubilized protein and on the surfactant concentration. Calculations of micellar occupancy and the relative surface areas of protein molecules and surfactant head-groups suggested that it was the exposure of the solubilized protein to the bulk organic solvent which promoted protein-surfactant complex formation as [P](rm) --> [P](rm,max). In the light of the experimental results and calculations described above, a mechanistic model is proposed to account for the observed phenomena. This is based upon the competing effects of increasing the solubilized protein concentration and the corresponding increase in the rate of protein-surfactant complex formation. The dynamic nature of the reverse micelles is inherent in the model, explaining the formation of the interfacial precipitate with time and its dependence on the internal phase volume of the micellar phase. Experiments on the co-partitioning of water and measurement ofthe AOT concentration in both phases verified the loss of protein, water, and surfactant from the organic phase at high values of [P](rm). (c) 1995 John Wiley & Sons Inc.

The application of aqueous two-phase systems to the purification of pharmaceutical proteins from transgenic sheep milk.

Transgenic sheep milk containing the protein human alpha 1-Antitrypsin (AAT) was partitioned in Poly(ethylene glycol) (PEG)-Sulphate and PEG-Phosphate biphasic systems. Individual partition coefficients for AAT and some of the milk proteins were determined in these systems. The effects of PEG molecular weight, pH and the inclusion of NaCl on the partitioning of the proteins were also studied. It was found that increasing the concentration of NaCl and decreasing the molecular weight of the PEG resulted in an increase of the partition coefficients of the proteins to the upper (PEG) phase. This partitioning effect was greater for the more hydrophobic proteins and particularly in systems having a pH close to the isoelectric point of the protein. Solubilities of the proteins in increasing concentrations of ammonium sulphate were measured in order to investigate the effects of hydrophobic and electrostatic interactions on the partitioning of these proteins in aqueous two-phase systems. Those proteins that precipitated at low levels of ammonium sulphate showed an increase in partition coefficient at low concentrations of NaCl, or they were precipitated at the interface of the phase at low concentrations of NaCl. Proteins that had low salting out constants in ammonium sulphate solutions were relatively unaffected by NaCl in ATPS. It is probable however that conformational changes and the state of aggregation of proteins are also important and should be invoked in describing the partitioning behavior observed for beta-Lg for example. Comparison of theoretical and experimental values for AAT yield and purity showed clearly that partition coefficients are influenced by the degree of purity and values obtained with purified standards are not necessarily the same as for the same protein present in a complex mixture. Under the most favourable conditions using a 4% w/w loading of transgenic ovine milk, we obtained a 91% yield of AAT in the PEG phase with a purity of 73%.

Combined effect of operational variables and enzyme activity on aqueous enzymatic extraction of oil and protein from soybean.

The individual effect of two different enzymes-protease and cellulase-on oil and protein extraction yields combined with other process parameters-enzyme concentration, time of hydrolysis, particle size and solid-to-liquid ratio-was evaluated by Response Surface Methodology. The selection of the enzymes for the study was based on preliminary experiments that showed higher increments in the extraction yield with the use of the two enzymes when compared to hemicellulase and pectinase. The levels of the quantitative parameters studied were: i) enzyme concentration: 0.1, 0.45, 2 w/w %; ii) liquid-to-solid ratio: 0.05, 0.125, 0.2; iii) mean particle size: 212.5, 449.5, 855 µm; iv) time of hydrolysis: 30; 60; 120 min. Experimental data for both oil and protein extraction yields obtained with and without enzymes correlated very well with process parameters (P < 0.0001), resulting in models with high coefficient of determination for oil and protein extraction yields (r(2) = 0.9570 and r(2) = 0.9807, respectively). The use of protease resulted in significantly higher yields over the control (protein yield increased from 27.8 to 66.2%, oil yield increased from 41.8 to 58.7%) only when heat treated flour was used, or when non-heat treated flour with large particle sizes was used in the extraction. The yields of protein and oil from non-heat treated material in general decreased slightly with the use of enzymes.