Week 1 HW: Final Project Ideas

(1) What I want to biomanufacture and why

I want to biomanufacture thyroxine (T4) at scale.

It’s a very important hormonal drug. People take it every day for life, and there’s no substitute for it, but in Egypt the supply sometimes goes down, and when it does, patients are forced to look for it on the black market and get exploited there. This isn’t something I read about. I witnessed it first hand when my mom, god bless her, was diagnosed with Hashimoto’s disease and we had to go looking for her medication.

Also, while part of thyroxine production is domestic, the precursors are mostly imported, which is a huge bottleneck.

(2) The living system I would use

Mammalian cells look like the obvious option here, since there’s a native enzyme that synthesizes thyroxine. But I don’t think that route works. Thyroid peroxidase isn’t efficient at T4 synthesis on its own, because it needs a separate scaffold protein, thyroglobulin, to present the tyrosines to it. On top of that, both of these proteins are highly glycosylated, which means they aren’t easily expressed in the known synthetic biology workhorses.

That being said, the chemical synthesis of thyroxine is quite mature and efficient, but it’s capped at around 50% yield because of chirality problems. So I believe the answer to a democratized thyroxine synthesis is to use simpler, readily available precursors instead.

Looking at T4’s structure, it’s mainly a diiodophenol attached to a diiodotyrosine through an ether (C-O) bond. So I wonder if we can just supply phenol and tyrosine into a cell-free system together with an FAD-dependent iodinase and a laccase. The iodinase carries out the iodination with the right regioselectivity on both rings, and the laccase catalyzes the C-O coupling through radical chemistry. T4 is very poorly soluble, so it should precipitate out of the medium as it forms, which both pulls the reaction forward and hands me a first purification step for free.

So in terms of inputs and outputs: in goes tyrosine, phenol, iodide, oxygen, and a cofactor regeneration system for the flavin, and out comes T4 as a solid I can filter off.

(3) How I would use automation

Enzyme screening on a programmable microfluidics platform like Nuclera’s. The nice thing about that platform for this specific project is that it expresses proteins cell-free, and my production system is also cell-free, so a variant that works in the screen goes straight into the real reaction without a host-context gap, and I skip cloning, expression and purification for every candidate. That’s where I’d hunt for an iodinase variant with the regioselectivity I need on phenol.

(4) How much, and how long

(a) How much?

There is a study where a cohort of 430 people were tested at Zagazig University’s clinic, and 20.6% of them had hypothyroidism. Assuming the sample is representative of the Egyptian population, that’s about 20 million people suffering from hypothyroidism in Egypt. Assuming median age and weight, each person needs roughly 3500 mcg, or 3.5 mg, so we are sitting at a demand of 70 kg per month. Assuming 50% of production is domestic, we are looking at a monthly production of 35 kg per month.

(b) How long?

This is a question I don’t really have an answer to. In principle, radical chemistry is quite fast. The problem is that laccase’s kinetics and thermodynamics on iodinated phenols haven’t been measured anywhere.

(5) Things I don’t know yet

1. Can laccases accommodate large iodinated compounds?
2. Can laccase oxidize diiodophenol and diiodotyrosine, or does their oxidation potential exceed that of laccase?
3. Will T4 really precipitate out of the mixture, or would all the substrates stay insoluble since they aren’t particularly hydrophilic?
4. Can regioselectivity be achieved using laccases?

Final Project