Figure 3 . Modular development of a media for heterologous protein production in P. pastoris
A) Initial full-factorial screen of nitrogen source choice and buffer pH demonstrates that urea is preferred over ammonium sulfates and high buffer pH is preferred over lower values. B) A full-factorial concentration optimization identified methanol as the most concentration dependent variable. Other components in the base media were predicted to affect productivity with much lower levels of significance. C) Evaluation of the effect of methanol concentration on P[8] titer, using two different base media (urea, buffer, and YNB concentrations): the biomass accumulation base medium and the optimal base media composition predicted by our concentration DOE. D) Ranking of supplements according to their effect on P[8] titer. Supplements related to membrane fluidity or lipid metabolism ranked highly. E) Evaluation of combinations of lipid and surfactant supplements confirmed that cholesterol supplementation leads to the greatest improvement in P[8] titer. F) Concentration optimization of cholesterol demonstrated low concentration dependence, with similar performance observed over a 40-fold range (0.2-8x). G) Comparing cholesterol-free and cholesterol-supplemented cultures fed at various concentrations demonstrates that cholesterol supplementation results in a significant ~25% improvement in P[8] titers (p<0.001). H) No significantly beneficial supplements were observed when repeating the supplementation screen. I) Screening supplementation of 20 g/L of co-fed substrates individually or in 1:1 combinations by mass identified sorbitol supplementation as highly beneficial to P[8] titer. J) Examination of the effect of co-feed ratio and total carbon concentration on titer in DM2_dev1 supplemented media. K) Comparison of P[8] titer obtained with DM2 to previous iterations and other common P. pastoris media demonstrates a ~2x improvement in P[8] titer, relative to 1 v/v% methanol RDM and 1 v/v% methanol BMMY.