(302c) Protein Crystallization with Gas Bubble Templates and Scaling-up Studies

With growing interest in the protein drugs in the pharmaceutical company, crystallization of the protein becomes potential technology in downstream manufacturing purified products with advantages of stability, storage and delivery. Moreover, the crystallization process is easy to scale up with much lower manufacturing cost ¹.

One of the useful method to facilitate crystallization process is to use heteronucleants to promote nucleation ², such as nano-templates. However, concerns were raised due to generality of engineered liquid-solid interfaces in the larger industrial ³ by introducing potential impurities. Therefore, by using gas bubbles as heteronucleants is a potential method to accelerate crystallization process without introducing more impurities. Previous literatures have proven that the gas-liquid interface could promote the nucleation and achieve better control of crystal sizes for organic molecules ^4–6.

This study investigated the effect of gas bubbles on the protein crystallization in a scaling-up methodology: air bubbles were started adding into the protein crystallization system in μL-level hanging-drop experiment; then different gas bubbles were mixed with crystallization solution by a mL-level microfluidic system; lastly gas flow was injected into a L-level batch crystallizer. In the small-scale system, the experiment containing gas bubble was compared to the one without gas bubble under the same conditions to gain insights who the presence of the gas-liquid interface affect the protein crystallization behavior. Microfluidic studies focused on how different gas feed altered the crystallization process. By using Process Analytic Tools (PATs), the crystallizer experiment gained more insights in the gas bubbles influence on the industrial scale-up of the protein crystallization.

Experimental Methods

Lysozyme in 0.1 M sodium acetate buffer with pH 4.2 was selected as the model crystallization system and sodium chloride with a range of concentrations were used to facilitate the crystallization process. The hanging-drop experiments were performed in 24-well plates sealed with glue to allow controlled vapor diffusion in closed systems as shown in Figure 1 (a), with a crystallization droplet of ~3 μL. For the 1 mL-scale study, a microfluidic chip was used to produce stable gas bubbles as shown in Figure 1 (c). The samples from both set-ups were stored at 20°C environment inside the thermostatic incubator and were taken out to be observed under the optical microscope. In the batch crystallizer of 500 mL illustrated in Figure 1 (e), the temperature was controlled by a heat exchanger covered the crystallizer and CryPRINS information system at 20°C and an array of PATs including FBRM, PVM, and UV/Vis probes was deployed to monitor the process characteristics.

Results and Implications

The results showed that the crystals were attached to the bubble. The crystals attached to the surface showed a curved facet as shown in Figure 1 (b) instead of the tetragonal geometry as in the bulk solution surface, confirming that crystal was actually growing on the gas-liquid interface. This piece of evidence also agreed with the heterogenous nucleation mechanism, where the air bubble provides the heterogenous nucleation site for lysozyme crystallization. To further understand the static air-bubble effect on crystallization, the crystal density near the injected air-bubble was compared to the one outside the air-bubble. The crystal density was increased up to 4 times due to the air-bubble in the range of the tested condition. As the gas-liquid interface provides a heterogenous nucleation sites, the free energy required for the lysozyme crystallization is lowered giving higher probability for lysozyme to nucleate near the air bubble.

In the microfluidic studies, four different types of gas, helium, oxygen, carbon dioxide and nitrogen, were injected in the microfluidic device with 100 mg/mL lysozyme and 1.1 M NaCl mixed solution. Figure 1 (d) depicts the gas bubble and crystal formation situation. For the experimented gas, it was found that the population density of lysozyme crystal showed a monotonical dependance on the gas solubility. The crystal tended to form greatest population density induced by carbon dioxide. Regarding the crystal formation locations, lysozyme crystals had a greater tendance to form on the gas-liquid interface, supporting a heterogeneous nucleation mechanism induced by the gas bubble. The solution treated with carbon dioxide contained the highest proportion, more than 80% of lysozyme formed on the carbon dioxide bubble surface. When gas bubbles were mixed into the batch protein crystallizer, the onset of nucleation was found to start faster, in agreement to the findings in the μL-scale hanging-drop study. The addition of gas bubbles is able to promote protein nucleation by acting as a heterogenous nucleation surface, paving the way for using the gas bubble as a soft template for protein crystallization in downstream processing.

Conclusions

Lysozyme crystals were discovered to be attached on the bubble surfaces. Air bubbles in the hanging-drop experiment accelerated the crystallization process, by reducing the induction time and increasing the population density in most of pH and lysozyme concentration conditions. Different gas bubbles in the microfluidic study shows the tendency of population density formed in the aqueous solution follows same as the one of the gas solubility. Meanwhile, CO₂ has the greatest possibility of protein crystals attached to the bubble surface. The promotion effect of gas bubbles were successfully applied in larger scale crystallization process, reducing the induction time and increase the yield. All the results demonstrated that the protein crystallization with air bubble as templates can be potential to accelerate protein crystallization in the industrial scale.

1. Yang, H., Chen, W., Peczulis, P. & Heng, J. Y. Y. Development and Workflow of a Continuous Protein Crystallization Process: A Case of Lysozyme. Cryst. Growth Des. 19, 983–991 (2019).
2. Artusio, F. & Pisano, R. Surface-induced crystallization of pharmaceuticals and biopharmaceuticals: A review. Int. J. Pharm. 547, 190–208 (2018).
3. Zhou, R. Bin, Cao, H. L., Zhang, C. Y. & Yin, D. C. A review on recent advances for nucleants and nucleation in protein crystallization. CrystEngComm 19, (2017).
4. Wada, Y., Matsumoto, M. & Onoe, K. Antisolvent crystallization of NaCl using the minute-bubble technique – Effects of different antisolvent types. J. Cryst. Growth 448, 76–81 (2016).
5. Kleetz, T., Pätzold, G., Schembecker, G. & Wohlgemuth, K. Gassing Crystallization at Different Scales: Potential to Control Nucleation and Product Properties. Cryst. Growth Des. 17, 1028–1035 (2017).
6. Adhikari, B. M., Truong, T., Bansal, N. & Bhandari, B. Influence of gas addition on crystallisation behaviour of lactose from supersaturated solution. Food Bioprod. Process. 109, 86–97 (2018).

Figure 1. (a) Schematic diagram of the hanging-drop experiment with gas-bubbles; (b) microscopic image of the crystal on bubble surface in the droplet; (c) schematic diagram of the microfluidic experiment with gas-bubbles; (d) microscopic image of the crystal formation inside capillary after mixing gas bubbles by microfluidic device; (e) schematic diagram of batch crystallizer fitted with gas bubble inflow and integrated PAT array; (f) PVM image of bubble and crystal formation inside the crystallizer.