Extraction behaviors of aqueous PEG impregnated resin system in terms of impregnation stability and recovery via protein impregnated resin interactions on bovine serum albumin

Schematic overview of the experiment

The overall experimental process is presented in Fig. 1. The processes included the dry impregnation of PEG on the resins, extraction with aqueous salt phase (sodium citrate and 4% NaCl) and back-extraction prior to the analysis such as impregnation stability, adsorption kinetics, extraction behaviors and comparison with ATPS and resins adsorption.

Chemicals and reagents

PEG with different average molecular weights of 2,000 g/mol (25322-68-3), 4,000 g/mol (25322-68-3), 6,000 g/mol (25322-68-3) and 8,000 g/mol (25322-68-3) were purchased from Milipore, Merck (Darmstadt, Germany). Potassium citrate (≥99.0%, 6100-05-6), sodium chloride (≥99.0%, 7647-14-5), BSA (>96%, 9048-46-8) and Amberlite XAD4 (20–60 mesh, 37380-42-0) were sourced from Sigma Aldrich (St. Louis, MO). The porous VitraPor glass pellets with two different particle sizes of 4.0 mm 8.00 mm were supplied from ROBU Glassfilter-Gerate GmbH (Hattert, Germany). PageBlue protein staining solution and protein loading buffer were all from Fermentas (St. Leon-Rot, Germany).

Impregnation stability of PEG on porous solids

In order to determine the impregnation stability on the porous solids, an experiment on the impregnation stability was performed quantitatively. The experiments were started by preparing different concentrations of PEG (10% w/w, 20% w/w, 30% w/w and 40% w/w) with different molecular weights of 2,000 g/mol, 4,000 g/mol, 6,000 g/mol and 8,000 g/mol. In order to determine the percentage of leached polymer phase, the polymer phase was dyed with 1.0 mg/mL of Patent blue V. The effects of pore size, particle size and materials on the impregnation stability were evaluated. Three types of solid materials such as VitraPor glass beads with diameter size 4.0 mm and 8.0 mm and Amberlite XAD4 (diameter size distribution of 0.25–0.84 mm) were used to impregnate the PEGs via dry impregnation method (Van Nguyen et al., 2013). VitraPor glass beads are inert porous glass beads with defined pore sizes of approximately 60 µm while Amberlite XAD4 is polymeric adsorbent with a pore size of approximately 0.005 µm. Both VitraPor glass beads and Amberlite XAD4 have structural rigidity, good mechanical stability and zero or minimal expansion in organic solvents. These PEG impregnated resins were suspended into 20% (w/w) of sodium citrate solution and agitated in incubator shaker (Infors HT, Switzerland) at 200 rpm, 30 °C for 180 min. The contact time was fixed at 180 min based on our preliminary study, in which the leaching percentage was stable after 30 min. For the quantitative evaluation of leaching, the leaching percentage from each sample solution at 30 min interval for 180 min was determined by measuring the absorbance of Patent blue V-dyed PEG at 620 nm using microplate reader (Halo MPR-96 Visible Microplate Reader, Dynamica). The leaching percentage of PEG were calculated using Eq. (1): (1)${\displaystyle \text{Leaching percentage}\left(\text{\%}\right)=\frac{\text{Absorbance}620\text{nm of leached PEG}}{\text{Absorbance}620\text{nm of impregnated PEG}}\times 100.}$

Experimental design

Two different aqueous solutions, PEG and sodium citrate were prepared as described by Tan et al. (2018). Dry impregnation method was used to impregnate PEGs to resins. Briefly, the PEG solution was added drop by drop into the resins until all the resins were fully impregnated with PEG. A 2 g of impregnated resins was added into the aqueous salt solution containing 1% (w/w) of BSA and incubated at 30 °C for 30 min with agitation speed of 150 rpm to enhance the mass transfer rate. For back extraction, the impregnated porous solids containing the BSA were filtered out from the citrate solution containing 4% of NaCl using a sieve (100 mesh size). The excess salt solutions on the solid’s surface were dried and suspended into aqueous citrate solution in the absence of NaCl for desorption process. BSA is one of the major serum proteins; it plays an important role as a result of its functional and nutritional properties which have bioactive peptides. Its low cost and wide availability compared to other proteins, its structure and functional similarity to human serum albumin enable its various biotechnological applications. Based on these aspects, BSA has been extensively applied as the model protein in various extraction and purification studies. For instance, BSA has been used as a model in the separating process using hydroxyapatite and active babassu coal (Pereira et al., 2015; Ribeiro Alves et al., 2016).

Determination of bovine serum albumin separation performance using AIRS

To investigate the protein separation performance using AIRS, the effects of the influencing factors on partitioning behaviors were studied. The factors consisted of the PEG molecular weights (2,000 g/mol, 4,000 g/mol, 6,000 g/mol), concentration of PEG (5% w/w, 10% w/w, 15% w/w and 20% w/w), and 8,000 g/mol), pH of sodium citrate (5.0, 6.0, 7.0, and 8.0), method of extraction viz. sonication and agitation, type of solid materials (VitraPor glass beads 4.0 mm diameter, VitraPor glass beads 8.0 mm and Amberlite XAD4), and salt concentration (10% w/w, 20%w/w and 30% w/w). One-factor-at-a-time was used while the other factors were fixed. The recovery yield of the protein was calculated using Eq. (2) where C_o and C_t referred to the initial concentration of the protein in the aqueous citrate phase and concentration of the protein in the aqueous citrate phase after back-extraction. (2)${\displaystyle \text{Recovery yield}\left(\text{\%}\right)=\frac{Co-Ct}{Co}\times 100\text{\%}}$

Statistical analysis

All the data represent the mean of three independent experiments. The errors during the measurement were indicated by standard deviations. The statistical analysis and regression coefficient (R²) values of the linear form of Langmuir isotherm and Freundlich isotherm models were determined using statistical functions of Microsoft Excel 2010.

Impregnation stability

Results from the leaching percentage of different PEG molecular weights and concentration on different types of resins are presented in Table 1. The leaching factor for Amberlite XAD4 did not exceed 1%. The leaching factors for VitraPor glass beads with diameter sizes of 8.0 mm and 4.0 mm were higher as compared to XAD4. The pore size of Amberlite XAD4 (0.005 µm) is much smaller as compared to VitraPor glass beads (60 µm). The smaller pore size of the solid materials had higher impregnation stability. The PEG leaching profile depends on the PEG molecular weight and concentration, and a slower release rate is observed for the higher molecular weight and concentration. When the PEG 2000 was used, 23.74% to 20.19% of the PEG at concentrations of 10% (w/w) in the impregnation had been leached out from the VitraPor glass beads 4.0 mm and 8.0 mm, respectively. The leaching percentage decreased to 7.36% when the concentration of PEG 2000 increased to 40% (w/w). Similarly, PEG 4000 and PEG 6000 in both VitraPor glass beads had higher leaching percentages at lower concentrations (10% and 20%). The leaching percentage successfully reduced to less than 5% when the concentrations increased to 30% and 40%, respectively. Scanning electron micrographs (SEM) of Amberlite XAD4 before and after impregnation with PEG 4000 are presented in Fig. 2. Smooth surface of Amberlite XAD4 in Fig. 2B indicated that the PEG had been impregnated into the resins, by filling the porous of the resins.

Adsorption kinetics of BSA into resins

Comparison of different coefficients from Langmuir and Freundlich adsorption isotherm of BSA on PEG impregnated resins are presented in Table 2. Langmuir isotherm yielded a negative value of maximum BSA uptake by PEG impregnated resin, −3.87 mg/g while the Freundlich isotherm provided adsorption capacity, K_f of 6.28 mg/g. Freundlich adsorption isotherm has the highest R² value (0.95) which is near to 1. For Freundlich model, intensity of adsorption (n) gives an indication of favorable adsorption process. This indicates the adsorption of BSA on PEG impregnated resins followed the Freundlich isotherm.

Effect of molecular weight and concentration of PEG, pH and concentration of sodium citrate on the BSA recovery yield

Effects of the PEG molecular weight and concentration on the BSA recovery yield are presented in Fig. 3A. Four different PEG molecular weights (2,000 g/mol, 4,000 g/mol, 6,000 g/mol and 8,000 g/mol) were used, each at different concentrations (10% w/w, 20% w/w, 30% w/w and 40%w/w). The recovery of BSA increased when using lower molecular weights of PEG. PEG 2000 and PEG 4000 showed similar trend where the recovery of BSA increased as the concentration of PEG increased. In contrast, a decrease in depleted protein could be observed with higher molecular weights of PEG (6000 and 8000).

In order to determine the optimum pH for the maximum depletion of BSA, the pH of the sodium citrate solution was varied from pH 5, 6, 7 and 8 and the results are presented in Fig. 3B. According to our experiment, increasing the pH of sodium citrate solution from 5 to 7 increased the BSA yield. The yield of recovered BSA for pH 5, 6 and 7 are 30.98%, 34.97% and 51.08% respectively. However, further increase in the pH to pH 8 reduced the BSA yield (41.39%). From this experiment, pH 7.0 is selected to be used in subsequent experiment since it is the optimum pH for the BSA extraction.

The sodium citrate concentration up to 30% (w/w) was tested to determine the effect of salt concentration on the recovery of BSA in AIRS and the results are presented in Fig. 3C. The recovery of BSA was slightly different at different concentration of sodium citrate. From this experiment, 10% (w/w) of sodium citrate concentration shows highest depletion of BSA (50.55%).

Comparisons between AIRS, ATPS and adsorption method on BSA partitioning experiment

Comparison between BSA recovery yield using three different types of partitioning experiment (AIRS, ATPS and adsorption) are presented in Fig. 4. ATPS was carried out using 40% (w/w) of PEG 2000 and 10% (w/w) of sodium citrate like AIRS for comparison. AIRS yielded the highest recovery of BSA (53.72%), followed by BSA adsorbed by Amberlite XAD4 without impregnated PEG (46.38%). ATPS had the lowest BSA recovery yield (41.80%) among all the partitioning experiments. From this experiment, it was proven that by combining ATPS and adsorption method, the BSA recovery yield could be enhanced.