Week 1 HW: Final Project Ideas

(1) What you would like to biomanufcture and why?

I would like to biomanufacture an evolved enzyme that has better fluorinase activity for the production of lenacapavir. Lenacapavir is an effective drug used to treat HIV infection. One of its characteristics is that it has a lot of fluorine atoms. Incorporating them into the molecule is difficult, but having enzymes that can do that would make the process of producing lenacapavir much more efficient.

(2) What living system would you use?

I would use, at first, an initial panel of wild-type fluorinases, taking each gene from the bacteria that naturally produce them and later modifying bacteria to express them. Later I would use tools like machine learning, directed evolution and AlphaFold to evolve the enzymes.

(3) How would you use automation?

I would use liquid-handling robots to isolate the genes encoding fluorinases, AI to make predictions about their structures, and, once I have the correct genes for the evolved enzyme, I would use automation to culture bacteria that express the enzyme.

(4) How much would you need to make, and how long would it take?

a. How much?

Lenacapavir maintenance requires about 1.854 g per patient per year, so 1,000 patients would require approximately 1.85 kg of lenacapavir, equivalent to about 1.91 mol. Assuming one fluorinated precursor molecule is needed per lenacapavir molecule and an overall biotransformation yield of 80%, approximately 2.4 mol of precursor would need to be processed.

If the evolved fluorinase reaches a target total turnover number (TTN) of 1,000, then about 1,9x10-3 mol of enzyme would be required. With a fluorinase molecular mass of approximately 32.3 kDa, this corresponds to about 62 g of enzyme, or roughly 75 g including a 20% operational margin. Published recombinant fluorinase expression in E. coli is around 50 mg/L, which would require roughly 2,100 L of culture after allowing for purification losses; if expression were optimized to an assumed 0.5 g/L, the required culture volume would fall to approximately 210–250 L.

b. How long?

For large-scale manufacture, assuming an optimized recombinant E. coli process producing about 0.5 g/L soluble fluorinase, a 250-L fed-batch fermentation could generate enough enzyme in a single batch. A reasonable estimate would be about 36–48 hours for cell growth, protein expression, harvest, lysis, and enzyme recovery.

For the fluorination reaction itself, if the evolved enzyme achieves a Kcat close to 1 min⁻¹ and a TTN of 1,000, the theoretical minimum reaction time would be around 17 hours under ideal saturated conditions. Allowing for substrate limitations, enzyme deactivation, mixing, and scale-up effects, I would estimate approximately 24 hours for the biotransformation, giving a total manufacturing time of about 3–4 days per production campaign once the strain and process are established.

The prototyping phase would take much longer because the enzyme first has to be evolved. Using automated liquid handling, microplate expression, LC-MS screening, and ML-guided library design, one directed-evolution round could take roughly 7–10 days. Three iterative rounds plus confirmation and kinetic characterization would reasonably require about 6–8 weeks before moving to the manufacturing stage.

(5) What I don’t know yet:

I’m not sure about the mechanism that the substrate undergoes, I’m not entirely sure how to evolve an enzyme or what kinds of biological constraints this project has.