Technical improvement in the purification of enzymes from red algae using an aqueous two-phase partitioning system

The effective elimination of phycobiliproteins from crude enzyme preparation of the red alga Caloglossa continua (Okamura) King et Puttock (Ceramiales, Florideophyceae) was investigated in an aqueous two‐phase partitioning system (ATPS) by changing the concentrations of polyethylene glycol (PEG) and ammonium sulfate (AS). The phycobiliproteins shifted from the AS‐rich lower phase to the PEG‐rich upper phase in high PEG and AS concentrations. The best ATPS condition for the elimination of phycobiliproteins from the lower phase was obtained by the combination of 20% (weight/volume; w/v) PEG and 16% (w/v) AS. However, the recovery of aldolase and mannitol‐1‐phos‐phatase activities was significantly reduced. For purification of the enzymes, a combination of 15% (w/v) PEG and 16% (w/v) AS was the best ATPS condition, because a high specific activity and recovery of the enzymes were obtained. Under these conditions, 98% of the phycobiliproteins were removed from the lower phase. Therefore, the ATPS proved to be a very useful method as a first step in the purification of enzymes from red algae.     

Microbial production of optically active β-phenylalanine ethyl ester through stereoselective hydrolysis of racemic β-phenylalanine ethyl ester

The ability to produce (R)- or (S)-β-phenylalanine ethyl ester (3-amino-3-phenylpropionic acid ethyl ester, BPAE) from racemic BPAE through stereoselective hydrolysis was screened for in BPAE-assimilating microorganisms. Sphingobacterium sp. 238C5 and Arthrobacter sp. 219D2 were found to be potential catalysts for (R)- and (S)-BPAE production, respectively. On a 24-h reaction, with 2.5% (w/v) racemic BPAE (130 mM) as the substrate and wet cells of Sphingobacterium sp. 238C5 as the catalyst, 1.15% (w/v) (R)-BPAE (60 mM) with enantiomeric purity of 99% e.e. was obtained, the molar yield as to racemic BPAE being 46%. On a 48-h reaction, with 2.5% (w/v) racemic BPAE (130 mM) as the substrate and wet cells of Arthrobacter sp. 219D2 as the catalyst, 0.87% (w/v) (S)-BPAE (45 mM) with enantiomeric purity of 99% e.e. was obtained, the molar yield as to racemic BPAE being 35%. The enzyme stereoselectively hydrolyzing (S)-BPAE was purified to homogeneity from the cell-free extract of Sphingobacterium sp. 238C5. The enzyme was a monomeric protein with a molecular mass of about 42,000. The enzyme catalyzed hydrolysis of β-phenylalanine esters, while the common aliphatic and aromatic carboxylate esters were not catalyzed.     

Optimization of conditions for acid protease partitioning and purification in aqueous two-phase systems using response surface methodology

Aspergillopepsin I, an acid protease, was purified using an aqueous two-phase system that comprised various combinations of polyethylene glycol (PEG), NaH₂PO₄ and NaCl. Partition of the enzyme depended upon the molecular mass of the PEG and the presence of NaCl. With PEG 1500, 4000 and 6000, the partition coefficients were increased by 1,500-, 1,800- and 560-fold compared to values without NaCl. The presence of NaCl (8.75%, w/w) increased purification by 3.8, 9.5 and 2.8 times into these respective PEGs. The optimal aqueous two-phase system for acid protease purification was developed using response surface methodology. This system contained 17.3% of PEG 4000 (w/w), 15% NaH₂PO₄ (w/w) and 8.75% NaCl (w/w) and provided the best partition coefficient (Ke > 1,100) and yield over 99% in the same phase. The optimal ATPS purification factor of acid protease was over 5.     

Activation in the family of Candida rugosa isolipases by polyethylene glycol

We have investigated activation of two isoenzymes (lip1 and lip3) from Candida rugosa in polyethylene glycol (PEG) media. Aqueous solutions of PEG 8000 and 20,000 activate lip3 but not lip1 from C. rugosa. Maximum activation (260%) of lip3 requires 6 h of pre-incubation with PEG 8000 (4%, w/v). PEG seems to shift the equilibrium between the open and the closed forms of lip3 towards the active conformation. Inhibition experiments demonstrate that ligands have easier access to the lip3 active site than to the lip1 active site, both in the presence and the absence of PEG.

The presence of PEG in the crystallization medium is responsible for reported differences in the crystal structures of lip1 and lip3. A comparative analysis of crystallographic models of lip1 and lip3 suggests a role for PEG in activation of lip3 and further stabilization of the activated/open form via dimerization in aqueous media.     

Potential Aqueous Two‐Phase Processes for the Primary Recovery of Colored Protein from Microbial Origin

The primary recovery of c‐phycocyanin and b‐phycoerythrin from Spirulina maxima and Porphyridium cruentum, respectively, using an established extraction strategy was selected as a practical model system to study the generic application of polyethylene glycol (PEG)‐phosphate aqueous two‐phase systems (ATPS). The generic practical implementation of ATPS extraction was evaluated for the recovery of colored proteins from microbial origin. A comparison of the influence of system parameters, such as PEG molecular mass, concentration of PEG as well as salt, system pH and volume ratio, on the partition behavior of c‐phycocyanin and b‐phycoerythrin was carried out to determine under which conditions target colored protein and contaminants concentrate to opposite phases. One‐stage processes are proposed for the primary recovery of the colored proteins. PEG1450‐phosphate ATPS extraction (volume ratio (V_R) equal to 0.3, tie‐line length (TLL) of 34 % w/w and system pH 7.0) for the recovery of c‐phycocyanin from Spirulina maxima resulted in a primary recovery process that produced a protein purity of 2.1 ± 0.2 (defined as the relationship of 620 nm to 280 nm absorbance) and a product yield of 98 % [w/w]. PEG1000‐phosphate ATPS extraction (i.e., V_R = 1.0, PEG 1000, TLL 50 % w/w and system pH 7.0) was preferred for the recovery of b‐phycoerythrin from Porphyridium cruentum, which resulted in a protein purity of 2.8 ± 0.2 (defined as the relationship of 545 nm to 280 nm absorbance) and a product yield of 82 % [w/w]. The purity of c‐phycocyanin and b‐phycoerythrin from the crude extract increased 3‐ and 4‐fold, respectively, after ATPS. The results reported herein demonstrated the benefits of the practical generic application of ATPS for the primary recovery of colored proteins from microbial origin as a first step for the development of purification processes.     

Genetic transformation of Clostridium botulinum hall a by electroporation

Summary A method was developed for the introduction of plasmids into Clostridium botulinum by electroporation. A 4.4 kb plasmid vector, pGK12, which contains genes for resistance to erythromycin (Em^r) and chloramphenicol (Cm^r) was electroporated into C. botulinum type A (Hall A). The highest transformation efficiency was obtained using midlog phase cells, 10% PEG 8000 as the electroporation solution, and 2.5 kV field strength. The transformation efficiency was highest (10³ transformants/g of DNA) when 1 g of plasmid DNA and 4 × 10⁸ CFU/ml of recipient cells were used. Plasmid DNA recovered from the transformants was indistinguishable from that introduced on the basis of restriction enzyme digestion and agarose gel electrophoresis.     

Evaluation of recombinant phenylalanine dehydrogenase behavior in aqueous two-phase partitioning

Recombinant Bacillus sphaericus phenylalanine dehydrogenase (PheDH) partitioning was studied in polyethylene glycol (PEG) and ammonium sulfate aqueous two-phase systems (ATPS). The objectives of this work were to investigate influences; varying the molecular mass and concentration of PEG, pH, phase volume ratio (V_R), tie-line length (TLL) and concentration of (NH₄)₂SO₄ on the partition behavior of PheDH. It was revealed that the partitioning was not affected by V_R, while PEG molecular mass and concentration and (NH₄)₂SO₄ concentration had significant effects on enzyme partitioning. Longer TLL and higher pH resulted in better partitioning into the top phase. Under the most favorable partition conditions with 8.5% (w/w) PEG-6000, 17.5% (w/w) (NH₄)₂SO₄ and V_R = 0.25 at pH 8.0, partition coefficient (K_E), recovery (R%), yield (Y%) and TLL were achieved 58.7%, 135%, 94.42% and 39.89% (w/w), respectively. Overall, the promising results obtained in this research indicated that the ATPS partitioning can be provided an efficient and powerful tool for recovery and purification of recombinant PheDH.     

The association behaviour of beta-lactamases. Sedimentation equilibrium studies in ammonium sulphate solutions.

The beta-lactamases (EC 3.5.2.6) from TEM plasmid RP4, Bacillus licheniformis 749/C and Enterobacter cloacae P99 were studied in solution over a wide concentration range by equilibrium sedimentation. Though crystal symmetries indicate that all three enzymes are potentially dimeric in their crystal forms, in 50 mM-sodium cacodylate at pH 6.5 the enzymes show only a small tendency to associate, indicated by a weight-average Mr (Mw) at 3% (w/v) concentration about 9% greater than that of the monomer. Although the mode of association could not be determined, this extent of association corresponded to a dimerization constant of about 2 X 10(2) M-1. In 2.1 M-(NH4)2SO4 the B. licheniformis enzyme shows some association at concentrations over 1%, displaying an Mw value at 7% concentration about 60% more than the monomer. Under the same conditions Mw for the Entero. P99 enzyme is about 60% greater than the monomer near the solubility limit of about 2%. However, the Mw for the TEM enzyme is over twice that of the monomer at its solubility limit (3%) in 1.7 M-(NH4)2SO4. Fitting the sedimentation data of the TEM enzyme in 1.7 M-(NH4)2SO4 with a dimerization model and an indefinite-isodesmic-association model yielded equilibrium constants of 1.5 X 10(4) and 3.3 X 10(2) M-1 respectively, with the indefinite-isodesmic model giving the better fit. Fitting the data for the other two enzymes yielded values of 1.4 X 10(3) and 1.7 X 10(2) M-1 respectively for the Entero. P99 enzyme and 4.5 X 10(2) and 45 M-1 respectively for the B. licheniformis enzyme. It could not be determined which model was the better fit for these two enzymes. Since none of the beta-lactamases studied here showed strong evidence of the terminal aggregate being a dimer, we conclude that crystalline dimers, if they exist, will not be tightly associated or physiologically significant.     

α-Amylase production by Bacillus subtilis and B. amyloliquefaciens in different PEG solutions

-Amylase production by Bacillus subtilis and Bacillus amyloliquefaciens was investigated in polyethyleneglycol (PEG)-containing growth medium. Five different molecular weight PEGs (600, 3000, 4000, 8000 and 20,000) were used. Enzyme production with B. subtilis increased 21% in medium containing 5% PEG 3000, but enzyme production with B. amyloliquefaciens increased 31% in medium containing 5% PEG 600 and 21% in medium containing 2% PEG 8000.     

Effect of polyethyleneglycols (PEG 600 and PEG 6000) and bovine serum albumin on myeloperoxidase-mediated release of [3H]-serotonin from rat peritoneal mast cells in vitro

Rat peritoneal mast cells in vitro released [³H]-serotonin in the presence of halide (I-), myeloperoxidase (MPO), and a H₂O₂-generating system. The degranulation was partly inhibited with low concentrations of bovine serum albumin (BSA; 0.1–l% w/v), or polyethyleneglycols (PEG 600 or PEG 6000, 0.1% w/v). Furthermore, PEG 600 and PEG 6000 enhanced the stabilization achieved with 0.1% BSA. The results may provide a new aspect on antigenicity of PEG-modified allergens, and on mast-cell response at an inflammatory site.     

Investigation of chromatography and polymer/salt aqueous two-phase processes for downstream processing development of recombinant phenylalanine dehydrogenase

This work presents a comprehensive study between the polymer/salt aqueous two-phase systems (ATPS) and chromatography process for downstream processing of recombinant Bacillus badius phenylalanine dehydrogenase (PheDH). First, the partitioning behavior of recombinant PheDH in polyethylene glycol (PEG)/K₂HPO₄ ATPS was examined. For comparative purpose, a classical chromatographic protocol was performed as well. Investigation of chromatography and ATPS procedures revealed that the ATPS comprising of 9% (w/w) PEG-6000, 16% (w/w) K₂HPO₄ and 16% (w/w) KCl with pH of 8.0, volume ratio (V _R ) of 0.25, temperature of 25 °C and 40% (w/w) cell lysate ensured the most favorable approach for PheDH downstream process. A specific activity of 4,231.4 U/mg, a yield of 96.7% and a recovery of 162.0% were obtained. Furthermore, the shorter process time (4 vs. 48 h) and the lower total cost (4 vs. 20 €) were additionally features that confirmed the suitability of proposed technique.     

A self-dual network derived from a single neuronic equation

A model of a neuron with memory was viewed from the realibizability of a given periodic sequence of typeR _n R _n =u ₁ u ₂...u _{n 1 2}... _n(u _j = 0 or 1, _j = 1 –u _j). It is easily shown that the periodic sequenceR _n R _n is always obtained as an output of the model with memory of lengthn. However it is not so easy to decide the realizability of the given sequence in the case of the memory of length less thann. The necessary and sufficient condition for the realizability was given in a practical form. The condition is concerned with the determinant of a matrix associated with the given sequence.     

Fractionation of phosphate in sediments of four representative mangrove stages (French Guiana)

Phosphate was fractionated in Guianese mangrove sediments. Fe(OOH)P was extracted using a Ca-EDTA + Na-dithionite solution buffered at pH 8. CaCO₃P was extracted using Na₂-EDTA solution at pH 4.5. Next, Acid Soluble Organic Phosphate (ASOP) was extracted by H₂SO₄ 0.5 N. Finally, Residual Organic Phosphate (ROP) was digested with H₂SO₄ + H₂O₂. Four representative mangrove stages have been studied: sea edge pioneer mangroves, mature coastal mangroves, mixed riverine mangroves, and declining to dead mangroves. The sum of the P-fractions varied between 638 to 804 g g^-1 in pioneer and mixed mangroves respectively. In all the stages, the percentage of inorganic phosphate was larger than 50% of the total P. Fe(OOH)P varied between 221 (pioneer mangrove) to 426 g g^-1 (dead mangrove). CaCO₃P varied between 75 to 102 g g^-1 in mixed, dead or mature mangroves and attained 125 g g^-1 in pioneer mangrove. The sum of the concentrations of organic phosphate (ASOP + ROP) increased markedly from the dead mangrove (189 g g^-1) to the mixed mangrove (380 g g^-1). Guianese mangroves, are relatively rich in total phosphate, possibly because they are narrowly related to the 'Amazon dispersal system. Each mangrove stage can be characterised by a prevailing form of phosphate. The concentrations of these different forms were ascribed to the marked relations with the seawater which controls import or export of suspended matters and to the wave action which controls the resuspension of the sediments and subsequently exchange of phosphate between the suspended matter and the water column.     

Response surface methodology for the extraction of wedelolactone from Eclipta alba using aqueous two-phase extraction

Abstract

Aqueous two-phase extraction of wedelolactone from Eclipta alba was studied using the polymer-salt system. The system consisted of polyethylene glycol (PEG) as a top phase (polymer) and sodium citrate as a bottom phase (salt). Process parameters such as PEG concentration, PEG molecular weight, salt concentration, and pH have been optimized using response surface methodology (RSM) with the help of central composite design (CCD). The optimized conditions for aqueous two-phase system (ATPS), in the case of one factor at a time approach, were found as PEG 6000, PEG concentration 18% (w/v), salt concentration 16% (w/v), and pH 7; with maximum extraction yield of 6.52?mg/g. While, RSM studies showed maximum extraction yield of 6.73?mg/g with the optimized parameters as PEG 6000, PEG concentration 18% (w/v), salt concentration 17.96% (w/v), and pH 7. ATPS was found to give a 1.3 fold increase in the extraction yield of wedelolactone as compared to other conventional extraction methods.     

Application of aqueous biphasic systems as strategy to purify tannase from Aspergillus tamarii URM 7115

The aims of the current study are to assess the influence of polyethylene glycol (PEG) concentration, molar mass, pH, and citrate concentrations on aqueous biphasic systems based on 2⁴ factorial designs, as well as to check their capacity to purify tannase secreted by Aspergillus tamarii URM 7115. Tannase was produced through submerged fermentation at 26°C for 67?h in Czapeck-Dox modified broth and added with yeast extract and tannic acid. The factorial design was followed to assess the influence of PEG molar mass (M_PEG 600; 4,000 and 8,000?g/?mol), and PEG (C_PEG 20.0; 22.0 and 24.0% w/w) and citrate concentrations (C_CIT 15.0, 17.5, and 20.0%, w/w), as well as of pH (6.0, 7.0, and 8.0) on the response variables; moreover, partition coefficient (K), yield (Y), and purification factor (PF) were analyzed. The most suitable parameters to purify tannase secreted by A. tamarii URM 7115 through a biphasic system were 600 (g/mol) M_PEG, 24% (w/w) C_PEG, 15% (w/w) C_CIT at pH 6.0 and they resulted in 6.33 enzyme partition, 131.25% yield, 19.80 purification factor and 195.08 selectivity. Tannase secreted by A. tamarii URM 7115 purified through aqueous biphasic systems composed of PEG/citrate can be used for industrial purposes, since it presents suitable purification factor and yield.     

Edge preference of retinal and tectal neurons in common toads (Bufo bufo) in response to worm-like moving stripes: the question of behaviorally relevant ‘position indicators’

Summary Previous experiments have shown that during prey-catching behavior (orienting, snapping) in response to a worm-like moving stripe common toads.Bufo bufo (L.) exhibit a contrast-and direction-dependent edge preference. To a black (b) stripe moving against a white (w) background (b/w), they respond (R^*) preferably toward the leading (l) rather the trailing (t) edge (R _l ^* > R _t ^* ), thus displaying head preference. If the contrastdirection is reversed (w/b), the stripe's trailing edge is preferred (R _l ^* < R _t ^* ), hence showing tail preference. In the present study, neuronal activities of retinal classes R2 and R3 and tectal classes T5(2) and T7 have been extracellularly recorded in response to leading and trailing edges of a 3 ° × 30 ° stripe simulating a worm and traversing the centers of their excitatory receptive fields (ERF) horizontally at a constant angular velocity in variable movement direction (temporo-nasal or naso-temporal).The behavioral contrast-direction dependent edge preferences are best resembled by the responses (R) of prey-selective class T5(2) neurons (R_l R_t=101 for b/w, 0.31 for w/b) and T7 neurons (R_lR_t=61 for b/w, 0.41 for w/b); the T7 responses may be dendritic spikes. This property can be traced back to off-responses dominated retinal class R3 neurons (R_lR_t=61 for b/w, 0.51 for w/b), but not to class R2 (R_lR_t =1.21 for b/w and 0.91 for w/b). The respective edge preference phenomena are independent of the direction of movement.When stimuli were moved against a stationary black-white structured background, the head preference to the black stripe and the tail preference to the white stripe were maintained in class R3, T5(2), and T7 neurons. If the stripe traversed the ERF together with the structured background in the same direction at the same velocity, the responses of tectal class T5(2) and T7 neurons were strongly inhibited, particularly in the former. Responses of retinal R2 neurons in comparable situations could be reduced by about 50%, while class R3 neurons responded to both the stimulus and the moving background structure.The results support the concept that the prey feature analyzing system in toads applies principles of (i) parallel and (ii) hierarchial information processing. These are (i) divergence of retinal R3 neuronal output contributes to stimulus edge positioning and (in combination with R2 output) area evaluation intectal neurons and to stimulus area evaluation and (in combination with R4 output) sensitivity for moving background structures inpre tectal neurons; (ii) convergence of tectal excitatory and pretectal inhibitory inputs specify the property of prey-selective tectal T5(2) neurons which are known to project to bulbar/spinal motor systems.Abbreviations ERF excitatory receptive field - IRF inhibitory receptive field - N nasal - T temporal - R _w response to a worm-like stripe moving in the direction of its longer axis - R _A response to an antiworm-like stripe whose longer axis is oriented perpendicular to the direction of movement - R _l response to the leading edge of a worm-like moving stripe - R _t response to the trailing edge of a worm-like moving stripe - b/w black stimulus against a white background - w/b white stimulus against a black background - sm structured moving background - ss structured stationary background - u minimal structure width of a structured background consisting of rectangular black and white patches in random distribution - HRP horseradish peroxidase     

Influence of partition parameters on a recombinant antigen of Schistosoma mansoni expressed on E. coli using poly(ethylene glycol)–hydroxypropyl starch aqueous two-phase system

This work describes the partition of a Schistosoma mansoni tegumental antigen produced by a recombinant Escherichia coli strain using an aqueous two-phase system (ATPS) composed of polyethylene glycol (PEG) and purified hydroxypropyl-starch (Reppal PES 100). The effects of the polymer molecular weight, tie line length and pH on antigen partitioning were investigated. The detection of the antigen in both phases was determined by ELISA. The system composed of PEG 8000 (5.1% w/w) and Reppal PES 100 (13.0% w/w) led to a yield of 92% and a purification factor of 12 concerning the antigen in the PEG-rich phase. It was observed that antigen partition in ATPSs was strongly affected by the pH and tie line length. In addition, it was possible in a single step, to remove the cell debris, which precipitated at the interface of the system.     

Growth and nutrition of small Betula pendula plants at different relative addition rates of iron

Summary Small birch plants (Betula pendula Roth.) were cultivated in a hydroponic spray solution where the relative addition rate of iron (R_Fe; g g^–1 day^–1), was the growth-controlling variable. All other elements were added in free access. An additional treatment was performed where all nutrients, including iron, were in free access (FA). The plants showed deficiency symptoms at steady-state growth and severe limitation of iron, R_Fe 0.05 and 0.10 day^–1. There were few symptoms at R_Fe of 0.15 or above. Plant relative growth rate (R_G; g g^–1 day^–1), equalled the relative rate of increase in iron supply, R_Fe. Internal iron concentration of the plants ranged from 40 to 70 g g^–1 dry weight (DW) over the range for which iron supply was limiting growth. At FA, the internal concentration was approximately 200 g g^–1 DW without further increase in R_G, demonstrating that iron may be taken up in excess without affecting growth. Internal concentrations of macronutrients were stable at the different R_Fe, except for Ca and Mg in shoots which were higher at low iron supply. Uptake rates of iron, calculated per root growth rate (mol g^–1 root DW), were approximately twice as high at R_Fe 0.20 as at 0.05 day^–1. The effect of iron limitation on dry matter allocation to leaves was small, with increases in the root fraction being largely at the expense of the stem. Leaf area ratio was constant regardless of R_Fe and the specific leaf area tended to increase with increasing iron limitation. Net assimilation rate decreased by a factor of 6 from free access to severe iron limitation, largely accounting for the differences in plant R_G.     

Retained binding mode of various DNA-binding molecules under molecular crowding condition

Meso-tetrakis(N-methyl pyridinium-4-yl)porphyrin (TMPyP) intercalates between the base-pairs of DNA at a low [TMPyP]/[DNA base] ratio in aqueous solutions and molecular crowding conditions, which is induced by the addition of Poly(ethylene glycol) (PEG). Studied DNA-binding drugs, including TMPyP, 9-aminoacridine, ethidium bromide, and DAPI (4′,6-diamidino-2-phenylindole) showed similar binding properties in the presence or absence of PEG molecules which is examined by circular and linear dichroism. According to the LD^r (reduced linear dichroism) results of the binding drugs examined in this work, PEG molecules induced no significant change compared to their binding properties in aqueous buffering systems. These results suggest that the transition moments are not expected to be perturbed significantly by PEG molecules. In this study, the experimental conditions of PEG 8000 were maintained at 35% (v/v) of total reaction volume, which is equal to the optimal molar concentration (0.0536 M as final concentration for PEG 8000) to maintain suitable cell-like conditions. Therefore, there was no need to focus on the conformational changes of the DNA helical structure, such as forming irregular aggregate structures, induced by large quantities of molecular crowding media itself at this stage.     

Purification of CBS 819.72 α-amylase by aqueous two-phase systems: Modelling using Response Surface Methodology

The partition and purification of α-amylase from a culture supernatant of Aspergillus oryzae CBS 819.72 was made in aqueous two-phase system (ATPS). According to bibliography and preliminary studies, the factors polyethylene glycol (PEG) molecular weight (M_PEG) and concentration (C_PEG), buffer type (BU) and concentration (C_BU), temperature (T), salt nature (SALT) and concentration (C_SALT), bioligand (BL) and concentration (C_BL) and pH were investigated using a Plackett–Burman design to identify the factors affecting separation. Taking into consideration a simultaneous increase in enzyme recovery (R_Y) and purification factor (P_F), the best performance of the system was obtained at 4 °C and pH 6 using PEG 8000 g/mol, citrate buffer, KCl and sucrose. Experimental Box–Behnken design together with the Response Surface Methodology (RSM) have been used to find optimum C_PEG, C_Citrate and C_SALT. Quadratic models were predicted for P_F and R_Y in the top phase and a better compromise between these two parameters can be found by superimposing the contour plots of P_F and R_Y for 8% citrate. A region in the experimental space can be defined where the purification factor is always higher than 3 with yields exceeding 65%.