(421h) Surrogate-Based Optimization for Biocatalytic Manufacturing of Diabetes Drug

Along with the rapid economic development and improved quality of life, the world diabetes prevalence has climbed to over 8% in recent years [1], and even more than 10% of population suffers from diabetes in countries such as the United States and so on. Type 2 diabetes accounts for the majority of patients [2]. Dipeptidyl peptidase IV (DPP-4) inhibitors have shown potent effect for treating type 2 diabetes, including sitagliptin. The mechanism of action of DPP-4 inhibitors is inhibiting the DPP-4 enzyme to reduce the rate of degradation of glucagon-like peptide-1, a hormone whose effect is increasing insulin and decreasing glucagon secretion [3]. Sitagliptin has advantages over other active pharmaceutical ingredients such as: instead of intravenous injection, it is orally active and patients do not suffer from hypoglycemia, etc. [4] In addition, sitagliptin has the highest market share of 14.8% in diabetes market in 2016, i.e., widely used by a majority of patients [5].

This paper demonstrates two case studies, maximizing product concentration and productivity, respectively, to show the application of surrogate-based optimization for sitagliptin manufacturing as this method has not been fully explored in the pharmaceutical industry. The advantages of this method include its open-source nature and being capable of approximating the global optimum of high-dimension objective function in a short time. A model of continuous manufacturing using plug flow reactors for the biocatalytic synthesis was proposed, based on the synthesis of sitagliptin using novel transaminase developed by Savile et al., that claims 10-13% higher yield, 53% higher productivity and 19% reduction in total waste [6]. The reaction of the biocatalytic transformation is described as below:

Prositagliptin ketone + i-PrNH_{2 ->}Sitagliptin + Acetone

The solutions of optimization were determined through pySOT package in Python 3.6 [7]. The sensitivity analysis was also conducted to identify critical process parameters (CPPs). The reaction kinetics constant was identified by assuming a first-order reaction due to the excess of i-PrNH₂. Also according to the study of Savile et al., we assumed feeding time to be proportional to the concentration of impurity formed, and the concentration of enzymes proportional to the overall catalyzing efficiency [6]. The results show that in case 1 of maximizing product concentration, the flow is extremely low and the influence of the concentration of enzymes is insignificant; while in case 2 of maximizing productivity, the flow rate is relatively high but still below the upper bound due to the formation of impurity. The results of sensitivity analysis show that in both case 1 and 2, the most influential CPP is the concentration of reactant. Most importantly, this method shows its ability to obtain the global optimum within a short computational time, and its application in pharmaceutical industry.

The present study also proposes several feasible downstream pathways including 1) the denaturation of enzymes, filtration, extraction and chromatography in series [6]; 2) differential extraction with membrane and chromatography in series [6]; and 3) ion-exchange adsorption of product, hydrophobic adsorption of enzyme, wash and chromatography in series [8][9]. However, it is similar to purify the transaminases as biocatalysts as to purify therapeutic proteins, which includes filtration, cell lysis, chromatography, etc. This may disobey the spirit of green chemistry. Therefore, we propose a manufacturing process in which the immobilized cell is adopted as biocatalyst instead of the enzyme, followed by differential extraction and chromatography steps. The techno-economic analysis is conducted to support such design decision.

In the future work, optimization could be conducted based on the whole process, from reaction to crystallization. Most importantly, not only the process optimization, but an economic optimization can also be performed with more information of the costs and revenues to obtain the maximized profits. Furthermore, life cycle assessment can be conducted to evaluate the environmental impact of each manufacturing manner. In the end, the optimal solution can be determined based on the trade-off between the economic and environmental performance.

Reference:

[1] World Health Organization, “Global Report on Diabetes,” p. 27, 2016. Retrieved from < http://apps.who.int/iris/bitstream/10665/204871/1/9789241565257_eng. pdf >

[2] World Bank database, Diabetes prevalence refers to the percentage of people ages 20-79 who have type 1 or type 2 diabetes. Retrieved from < https://data.worldbank.org /indicator/SH.STA.DIAB.ZS?view=chart >

[3] N. A. Thornberry and B. Gallwitz, “Mechanism of action of inhibitors of dipeptidyl-peptidase-4 (DPP-4),” Best Pract. Res. Clin. Endocrinol. Metab., vol. 23, no. 4, pp. 479–486, Aug. 2009.

[4] E. J. Verspohl, “Novel therapeutics for type 2 diabetes: Incretin hormone mimetics (glucagon-like peptide-1 receptor agonists) and dipeptidyl peptidase-4 inhibitors,” Pharmacol. Ther., vol. 124, no. 1, pp. 113–138, Oct. 2009.

[5] Evaluate Pharma, “World Preview 2017, Outlook to 2022,” Evaluate Pharma, no. 10th Edition. pp. 1–41, 2017.

[6] C. K. Savile et al., “Biocatalytic Asymmetric Synthesis of Chiral Amines from Ketones Applied to Sitagliptin Manufacture,” Science (80-. )., vol. 329, no. 5989, pp. 305–309, Jul. 2010.

[7] http://pysot.readthedocs.io/en/latest/index.html

[8] M. D. Truppo, J. D. Rozzell, and N. J. Turner, “Efficient Production of Enantiomerically Pure Chiral Amine at Conc 50 g/L Using Transaminase,” Org. Process Res. Dev., vol. 14, no. 1, pp. 234–237, 2010.

[9] M. D. Truppo, H. Strotman, and G. Hughes, “Development of an Immobilized Transaminase Capable of Operating in Organic Solvent,” ChemCatChem, vol. 4, no. 8, pp. 1071–1074, 2012.