3.3. Refolding of denatured proteins
To investigate the refolding of the model protein, a 0.25 mg/mL CA solution unfolded using 5 M GdnHCl was fed at a flow rate of 0.05 mL/minute, with a residence time, τ , of 20 minutes, from the feed-side inlet of the microchannel, and 0.05 M Tris-HCl buffer was fed from the permeation side inlet. Refolding was carried out by removing GdnHCl from the feed side to the permeation side through the dialysis membrane. The result of the microchannel refolding was compared with that of refolding using the conventional 10-fold dilution and batch dialysis methods (Fig. 6).
The CA concentrations shown in Fig. 6a represent the CA concentrations after refolding, and the protein recovery data shown in Fig. 6b is the ratio of the amount of soluble CA after folding (CA concentration [mg/mL] × recovered liquid amount [mL]) to the amount of soluble CA before refolding (CA concentration [mg/mL] × feed solution volume [mL]). The relative recovery shown in Fig. 6c represents the CA activity after refolding (activity per concentration) when the enzyme activity of native CA is considered to be 100%. The total recovery shown in Fig. 6d is the product of relative activity (Fig. 6c) and CA recovery amount (Fig. 6b), and represents the total recovery rate of active CA for each refolding method when that of native CA is taken to be 100%.
From the data shown in Fig. 6a, the CA concentration after refolding was 0.025 mg/mL using the conventional dilution method, which was almost the same as the dilution factor (10-fold), whereas in batch dialysis and microchannel dialysis, the concentration was about 0.20 mg/mL, about 20%–30% lower than the initial value. Table 2 shows the relationship between the amount of solution supplied from the microchannel inlet and that recovered at the outlet for microchannel dialysis. The solution amount increased by around 1.3 times during the dialysis. For batch dialysis, the solution amount before dialysis was 10.0 g and that after dialysis was 11.4 g, although not all of the solution was recovered. The results indicate that for both methods, the amount of solution increased during dialysis.
The water possibly moved from the permeate side to the feed side through the dialysis membrane due to the osmotic pressure of GdnHCl. When a permeation experiment was attempted with a 5 M GdnHCl feed solution, as described in Section 2.2.1, the amount of solution in the feed side increased, and that in the permeate side decreased, strongly suggesting the effect of osmotic pressure (data not shown). Nevertheless, the concentration of CA remained high in both dialysis processes, as shown in Fig. 6a.
The CA recovery rate for this process is shown in Fig. 6b. Almost 100% of CA was recovered using the 10–fold dilution and microchannel dialysis methods, and protein aggregation did not occur. Using the conventional dialysis method, the recovery rate is slightly lower, at about 80%.
A comparison of the enzymatic activity after refolding is shown in Fig. 6c. Since the enzymatic activity depends on the CA concentration, activity normalized by CA concentration was compared. As a reference point, the activity of native CA was taken to be 100%. The recovery rate of the activity using the 10-fold dilution method was about 40%, which is consistent with previous reports (Yamaguchi et al., 2010). CA is known to be a protein that is relatively difficult to refold. Activity reached 100% in batch dialysis refolding, and 10% higher activity in microchannel dialysis refolding on average. The raw data for enzymatic activity measurement are shown in the Supporting Information. Conventional batch dialysis was reported to cause collisions and agglomeration of intermediates during relatively long refolding times, (Yamaguchi et al., 2010) while in microchannel dialysis, the residence time during which these intermediates can be produced is short. It is possible that the activity normalized by enzyme concentration exceeds 100%, as has been reported in the literature (Batas & Chaudhuri, 1996). The state after refolding may be better than the native folding state, which was prepared by dissolving the purchased enzyme in buffer.
The total esterase activity of the CA as a product of the CA recovery (in Fig. 6b) and the normalized enzymatic activity (in Fig. 6c) was shown in Fig. 6d. Microchannel dialysis was shown to be superior among the three refolding methods.
As shown in Fig. 6a–d and Table 2, microchannel dialysis produced stable results with respect to the final concentration, activity, and flow rate, in 30–150 minutes. The CA concentration on the permeate outlet side was always less than 0.01 mg/mL, which was negligible sufficiently smaller than that on the feed side; the permeation of CA through the dialysis membrane was almost negligible (data not shown).
Table 3 summarizes the results of three protein refolding methods: conventional 10-fold dilution, conventional batch dialysis, and microchannel dialysis. These data clearly show that microchannel dialysis achieved high protein concentrations with sufficient recovery of active proteins, with a short residence time.
One problem with the use of microchannel dialysis is the limitation of the amount of solution which can be processed, due to the small size of the flow path. However, this problem can be addressed by increasing the number of the channels by using, for example, hollow fiber membrane. The concentration of 0.20 mg/mL produced by microchannel dialysis is about eight times higher than that of the dilution method. Nevertheless, even higher concentrations are expected industrially, and the issue of producing higher concentrations of active protein in a short time is currently under investigation.