Fig. 1(a) depicts the binodal curves of 1-propanol /sucrose, 1-propanol/maltose, and 1-propanol/glucose ABS. Fig. 1(b) depicts the binodal curves of 2-propanol/sucrose, 2-propanol/maltose, and 2-propanol/glucose ABS. The biphasic region of 1-propanol/sugar ABS was larger than that of 2-propanol/sugar ABS, indicating that the 2-propanol/sugar ABS required a higher amount of phase-forming components to form ABS (cf. Supporting Information Data, Figs. A1–A3). According to the octanol/water partition coefficient of 1-propanol and 2-propanol, the hydrophilicity of secondary alcohol (i.e. 2-propanol) is higher than the hydrophilicity of primary alcohol (i.e. 1-propanol) (Ebrahimi and Sadeghi 2018). In other words, a higher concentration of sugar was required to sugar-out the 2-propanol to form ABS. Thus, the lower phase-forming ability of 2-propanol was observed. Ethanol was also investigated as a potential phase-forming component for the formation of alcohol/sugar ABS. Several systems constituted with various concentrations of ethanol and sugar were examined. However, no phase formation was observed. This could be due to the relatively close hydrophilicity degree between the ethanol and the examined sugars (Ebrahimi and Sadeghi 2018). Moreover, precipitation had occurred in the mixture which contained high concentration of ethanol and low concentration of sugar. Thus, ethanol and sugars are not suitable candidates to form stable sugaring-out assisted ABS.
As shown in Fig. 1(a), the binodal curve of 1-propanol/maltose ABS was closer to the axes compared to the binodal curve of 1-propanol/glucose ABS and 1-propanol/sucrose ABS. A similar trend was observed for 2-propanol/sugar ABS (Fig. 1(b)), indicating that the phase formation of alcohol/sugar ABS was also influenced by the sugaring-out ability of the sugars investigated. The sugaring-out ability and hydrophilicity of the sugars are affected by its stereochemistry and number hydroxyl groups. The molecular structures of sugars used to form the alcohol/sugar ABS are depicted in Fig. 2. Among the disaccharides (i.e. maltose and sucrose) and monosaccharides (i.e. glucose) applied in this study, maltose has the highest number of equatorial hydroxyl group and thus a higher aptitude to be hydrated (Freire et al. 2011). It has been reported that the ability of these sugars in sugaring-out ionic liquids (Ferreira et al. 2016), polymers (Sadeghi et al. 2016), and alcohols (Ebrahimi and Sadeghi 2018) to form ABS decreased in the order of maltose > sucrose > glucose. Fig. 1 showed the amount of maltose required to form 1-propanol/sugar and 2-propanol/sugar ABS was lesser compared to sucrose and glucose (Ebrahimi and Sadeghi 2016).
The partitioning behaviour of BSA in the 1-propanol/glucose, 1-propanol/maltose, 1-propanol/sucrose, 2-propanol/glucose, 2-propanol/sucrose, and 2-propanol/maltose ABS was investigated to evaluate the effect of various types and concentrations of phase-forming components on the partition efficiency of BSA. The systems investigated were chosen based on their relative position in the biphasic region which will give a VR of about 1.0 at equilibrium.
For 1-propanol/sugar ABS, BSA was preferentially partitioned to the alcohol-rich top phase (i.e. K > 1) at lower concentrations of alcohol and sugar (Table 1). The BSA was partitioned to the alcohol-rich top phase which is more hydrophobic because of the favourable hydrophobic interaction between the BSA molecules and the alcohol molecules (Ng et al. 2018). However, increasing the alcohol and sugar concentrations decrease the partition efficiency of BSA in both 1-propanol/sugar ABS and 2-propanol/sugar ABS. This decrease could be attributed to the gradual dehydration of both aqueous phases (Ebrahimi and Sadeghi 2018). When the concentrations of sugar and alcohol used to construct the 1-propanol/sugar ABS were increased, both aqueous phases with increasing phase-forming components’ concentration were gradually dehydrated, which to a stronger extent resulted in the insufficient free water molecules to solubilize the protein in the alcohol-rich top phase, and thereby leading to the decrease in the partition efficiency of the BSA (Ooi et al. 2009). For 2-propanol/sugar ABS, a high K of 2.41 ± 0.39 and Y of 70.51% ± 3.35 were recorded at 32% (w/w) 2-propanol/31% (w/w) maltose ABS. Similarly, increasing the concentrations of 2-propanol and sugars significantly reduced the partition efficiency of BSA in the 2-propanol/sugar ABSs. The K and Y of BSA in 2-propanol/sugar ABSs were mostly lower than unity and 50%, respectively. Moreover, the 1-propanol which is more hydrophobic compared to the 2-propanol has driven more of the protein to partition to the alcohol-rich top phase of the sugaring-out assisted ABS.
Table 1
The effect of types and concentration of phase-forming components on the partition efficiency of BSA with propanol/sugar ABS.
Types of ABS
|
Concentration of sugar,
% (w/w)
|
Concentration of alcohol,
% (w/w)
|
K
|
Recovery Yield, %
|
1-Propanol/Maltose
|
22
|
35
|
6.03 ± 0.17
|
87.84 ± 0.61
|
|
23
|
37
|
4.40 ± 0.43
|
81.70 ± 1.88
|
|
24
|
38
|
2.57 ± 0.46
|
72.58 ± 3.58
|
|
25
|
39
|
1.74 ± 0.14
|
64.39 ± 1.81
|
|
26
|
39
|
0.90 ± 0.25
|
49.40± 4.79
|
1-Propanol/Sucrose
|
26
|
33
|
1.17 ± 0.07
|
63.61 ± 1.48
|
|
27
|
34
|
1.26 ± 0.11
|
56.71 ± 2.11
|
|
28
|
35
|
1.36± 0.15
|
62.86 ± 2.64
|
|
29
|
36
|
0.93±0.10
|
48.21 ±2.72
|
|
30
|
37
|
0.72±0.08
|
42.33 ± 3.32
|
1-Propanol/Glucose
|
24
|
36
|
1.14 ± 0.08
|
50.42 ± 2.10
|
|
25
|
38
|
1.36 ± 0.05
|
58.11 ± 1.48
|
|
26
|
39
|
1.01 ± 0.08
|
50.76 ± 2.58
|
|
27
|
40
|
0.59 ± 0.02
|
38.04 ± 0.71
|
|
28
|
41
|
0.38 ± 0.09
|
30.75 ± 4.91
|
2-Propanol/Maltose
|
31
|
32
|
2.41 ± 0.39
|
70.51 ± 3.35
|
|
32
|
33
|
1.19 ± 0.19
|
54.19 ± 3.88
|
|
33
|
34
|
0.85 ± 0.03
|
42.10 ± 0.92
|
|
34
|
35
|
0.76 ± 0.14
|
43.06 ± 4.63
|
|
35
|
36
|
0.33 ± 0.03
|
23.38 ± 1.66
|
2-Propanol/Sucrose
|
33
|
38
|
1.01 ± 0.03
|
48.13 ± 0.84
|
|
34
|
39
|
0.75 ± 0.02
|
43.02 ± 0.60
|
|
35
|
40
|
0.74 ± 0.04
|
42.46 ± 1.19
|
|
36
|
41
|
0.22 ± 0.05
|
17.71 ± 3.15
|
|
37
|
41.5
|
0.08± 0.03
|
7.42 ± 2.90
|
2-Propanol/Glucose
|
31
|
39
|
0.82 ± 0.07
|
44.95 ± 2.00
|
|
32
|
40
|
0.59 ± 0.00
|
39.84 ± 0.10
|
|
33
|
41
|
0.35 ± 0.05
|
27.38 ± 2.70
|
|
34
|
42
|
0.32 ± 0.04
|
25.82 ± 2.51
|
|
35
|
43
|
0.48 ± 0.02
|
34.33 ± 1.10
|
The types of sugar applied also exerted a significant effect on protein partition efficiency in the alcohol/sugar ABS. The sugaring-out ability of sugars greatly depends on its hydration extensions (Ebrahimi and Sadeghi 2018). Apart from the number of hydroxyl group, the arrangement of the hydroxyl group on the pyranose ring also affects the intensity of the sugaring-out effect. The equatorial hydroxyl groups possess greater hydration potential than the axial ones as the former could interact with water molecules and form long-lived hydration structures (Ng et al. 2012; Sadeghi et al. 2016). For glucose, its C-2, C-3, and C-4 hydroxyl groups are all on the equatorial positions (Fig. 2). Both maltose and sucrose possess the same number of hydroxyl group. Maltose is composed of two glucose monomers whereas sucrose is composed of a glucose monomer and a fructose monomer. As such, maltose which consists of two six-membered pyranose rings is more easily hydrated than the sucrose (Freire et al. 2011). In other words, maltose has a higher number of equatorial hydroxyl group and could exert a stronger sugaring-out effect than sucrose and glucose, thereby facilitating the transfer of BSA to the alcohol-rich top phase. Results showed that the 1-propanol/maltose ABS gave an overall better protein partition efficiency compared to the 1-propanol/sucrose ABS and the 1-propanol/glucose ABS. A similar phenomenon was observed in the 2-propanol/sugar ABS. Therefore, 35% (w/w) 1-propanol/22% (w/w) maltose ABS which gave higher K of 6.03 ± 0.17 and Y of 87.84% ± 0.61 compared to other alcohol/sugar ABSs was selected to investigate the effect of amount of protein added on the partition efficiency of BSA in the 1-propanol/maltose ABS.
Effect of BSA amount added to the ABS on the partition efficiency of BSA
The amount of BSA added into the ABS was varied at a range of 5-25% (w/w) to evaluate the partition efficiency of BSA in the 35% (w/w) 1-propanol/22% (w/w) maltose ABS (Fig. 3). When the amount of protein added to the system increased from 5% (w/w) to 10% (w/w), the partition coefficient increased from 6.27 ± 0.16 to 10.21 ± 0.03 and the recovery yield increased from 87.12% ± 1.55 to 91.27% ± 0.02. It was observed that the increase in protein amount from 5% (w/w) to 25% (w/w) resulted in a decrease of VR by 8.7%. With the decrease in VR, the free volume available in the top phase to accommodate a higher amount of protein is greatly reduced (Ooi et al. 2009). Hence, the protein partition efficiency decreased to minimum K of 3.38 ± 0.28 and Y of 77.12% ± 1.49 when the protein amount was increased to 25% (w/w). The increase in the amount of BSA beyond 25% (w/w) is not feasible because of the increase in the tendency of protein precipitation at the interface as a result of phase saturation (Ng et al. 2014). Thus, 10% (w/w) of BSA solution which exhibited the highest protein partition efficiency was selected to further investigate the effect of pH on the partitioning behaviour of BSA in the 1-propanol/maltose ABS.
Effect of pH on the partition efficiency of BSA
The pH of the environment will affect the net surface charge and conformation of the protein. Hence, the pH of an ABS can be adjusted to steer and enhance the partition of protein to the targeted phase. In this study, the pH of the 35% (w/w) 1-propanol/22 % (w/w) maltose ABS, which contained 10% (w/w) BSA, was varied between pH 3.0 and pH 8.0 to investigate the effect of pH on the partition efficiency of BSA. The phase system was mixed well while adjusting the pH with the addition of 1.0 M of sulfuric acid or 1.0 M sodium hydroxide and followed by phase separation. The change in the volume of both phases was negligible with the increase in pH.
As shown in Fig. 4, the pH exerted a significant impact on the protein partition efficiency. When the pH was increased from pH 3.0 to pH 5.0, the partition efficiency of BSA increased remarkably from K of 5.71 ± 0.7 and Y of 86.94% ± 1.39 at pH 3.0 to maximum K of 20.01 ± 0.05 and Y of 95.42% ± 0.01 at pH 5.0. The isoelectric point (pI) of BSA is at pH 4.8 and its native form is relatively stable at pH 5.0-8.0 (Chow et al. 2015). When the pH was increased to 5.0, approaching the pI of BSA, the net surface charge of BSA was close to zero. Thus, the hydrophobic interaction between the BSA and 1-propanol molecules was intensified at this pH, facilitating the partition of more BSA to the alcohol-rich top phase (Ng et al. 2018). Thereafter, the protein partition efficiency decreased significantly when the pH was raised to pH 8.0. At pH 8.0, the K and Y of BSA dropped to a minimum of 2.52 ± 0.27 and 71.50% ± 2.17, respectively. As proteins prone to denature at high pH values, operating the ABS above pH 8.0 would not be favourable for effective protein recovery (Chow et al. 2015).
Effect of types of adjuvants on the partition efficiency of BSA
The effects of the addition of neutral salts (sodium chloride (NaCl) and potassium chloride (KCl)) or ILs, (1-butyl-3-methylimidazolium tetrafluoroborate, [Bmim]BF4; 1-ethyl-3-methylimidazolium tetrafluoroborate, [Emim]BF4; 1-butyl-3-methylimidazolium bromide, [Bmim]Br and 1-ethyl-3methylimidazolium bromide, [Emim]Br) on the partitioning behaviour of BSA in the 35% (w/w) 1-propanol/22% (w/w) maltose ABS at pH 5.0 were investigated. These adjuvants were added into the ABS at a fixed concentration of 1% (w/w) and their respective K and Y were shown in Fig. 5.
According to the Hofmeister series, the salting-out effect is more pronounced for KCl than NaCl (Wan et al. 2018). Thus, higher K (13.86 ± 1.03) and Y (93.26% ± 0.47) of BSA were obtained in the 35% (w/w) 1-propanol/22% (w/w) maltose ABS which was added with KCl compared to K (7.57 ± 0.78) and Y (88.48% ± 1.34) obtained with the addition of NaCl. Nevertheless, both K and Y values attained with the addition of these univalent neutral salts were lower than that without the addition of adjuvants, indicating that the presence of KCl and NaCl has a negative impact on the protein partition efficiency of BSA in the alcohol/sugar ABS.
Among the ILs investigated, only [Bmim]Br exhibited a positive effect in improving the partition efficiency of BSA in the 1-propanol/maltose ABS. When compared to the alcohol/sugar ABS without adjuvants, the addition of [Bmim]BF4 showed insignificant changes in the protein partition efficiency, while the addition of [Emim]BF4 and [Emim]Br reduced the K and Y significantly by more than 60% and 8%, respectively. When the effect of IL’s cation on the protein partition efficiency was compared, the K and Y values of BSA for ILs which share the same type of anion increase in the following orders: [Emim]BF4 < [Bmim]BF4 and [Emim]Br < [Bmim]Br. This phenomenon could be attributed to the alkyl chain length of IL’s cation. Longer alkyl chain length increases the hydrophobicity of ILs and enhances the molecular interaction between BSA and IL, thereby promoting the partition of the BSA in the top phase containing 1-propanol and IL (Ran et al. 2019).
For the same alkyl chain length, [Bmim]Br exhibited higher K and Y values compared to [Bmim]BF4. This variation in the protein partition efficiency is in accordance with the chaotropic order of the IL anions ( > ), whereby the chaotropic which has a higher tendency in unfolding the protein and destabilizing the hydrophobic aggregates will increase the dissolution of protein in the sugar-rich bottom phase (Ran et al. 2019). Thus, tetrafluoroborate-based IL with high salting-in ability could not act as a suitable adjuvant to enhance the partition efficiency of protein in the alcohol/sugar ABS. Comparing to the ABS without adjuvant (K of 20.01 ± 0.05 and Y of 95.42% ± 0.01), a slight increase in K (27.42 ± 0.02) and Y (96.27% ± 0.10) of BSA were observed with the addition of [Bmim]Br. Hence, [Bmim]Br was chosen as the suitable adjuvant for subsequent evaluation.
Effect of concentration of adjuvant [Bmim]Br on the partition efficiency of BSA
The effect of [Bmim]Br’s concentration on the partitioning behavior of BSA was evaluated within the range of 1.0-5.0% (w/w) and the results were presented in Fig. 6. When the concentration of [Bmim]Br was increased to 3.0% (w/w), the K (34.39 ± 2.26) was increased markedly by 70% compared to the system without adjuvant (K of 20.01 ± 0.05). This increase in protein partition efficiency indicates that the addition of IL could serve to enhance the affinity of the 1-propanol-rich phase to the protein (Ran et al. 2019). A maximum Y of 97.05% ± 0.35 was achieved at 3% (w/w) [Bmim]Br. The protein partition efficiency decreased thereafter to K of 10.08 ± 1.4 and Y of 90.88% ± 1.06 at 5% (w/w) [Bmim]Br. This decrease in protein partition efficiency was probably caused by the change in protein stability at a high concentration of IL (Hadzir et al. 2016).
Stability of BSA in 1-propanol/maltose ABS
The protein conformation is highly affected by the change in the pH of the environment and the presence of foreign solute. FTIR analysis was conducted to evaluate the structure of the BSA that partitioned in the top phase of the 35% (w/w) 1-propanol/22% (w/w) maltose ABS at pH 5 with and without the addition of 3% (w/w) [Bmim]Br IL. Proteins are made up of many amino acids which are joined to one another by peptide bonds. The basic unit of the peptide bond is amide. The amide I band, which falls between 1600-17000 cm-1, is associated with the C=O stretching vibration of the amide functional group [1]. As the most sensitive spectral region to the secondary structure of the protein, this absorption band is often used as a structural probe to determine the structural properties of protein. Fig. 7 shows the FTIR spectra of pure BSA and BSA partitioned in the alcohol-rich top phase of the 35% (w/w) 1-propanol/22% (w/w) maltose ABS at pH 5 with and without the addition of adjuvants. As shown in Fig. 7, the amide I absorption band of the pure BSA falls between 1600-17000 cm-1. For the spectra of BSA separated to the top phase of 1-propanol/maltose ABS with and without IL, the amide I absorption band was still identifiable and remained unchanged in the same region. Since there was no disappearance and significant shift of protein absorbance peak, the BSA conformation was unaltered and conserved in the phase solution of the 1-propanol/maltose ABS.