Week 1 HW: Final Project Ideas

cover image cover image

(See references below)

Question 1:

Industrial natural rubber (cis-1,4-polyisoprene) production today depends almost entirely on the Hevea tree (1,3). I would like to biomanufacture rubber in engineered cells instead.

Natural rubber is a high-volume commodity whose performance petrochemical synthetic rubber still can’t fully match: its long chain length and co-produced proteins and lipids give it a resilience, tear/fatigue resistance, and strain-induced crystallization that keep it essential where failure isn’t an option. For example, commercial and military aircraft tires are made almost entirely from natural rubber because synthetic rubber cannot perform as well. Natural rubber appears in more than 40,000 products: tires, hoses, medical gloves, toys, clothing. The U.S. alone consumed roughly 1.78 billion pounds in 2020, and the EU has listed natural rubber as a critical raw material since 2017, one of its few biotic entries (2,3).

Yet the supply is narrow and fragile: over 90% of US natural rubber comes from Southeast Asia, from a single tree species with little genetic diversity that is highly susceptible to disease, with land-use, deforestation, and labor pressures compounding the risk and Europe fully dependent on imports (2,3).

Natural rubber’s value comes from its very high molecular weight, and that is exactly what past attempts have failed to reach: the best recent de novo result in yeast reached only ~10,000 Da, and its authors note that efficient synthesis at high molecular weight remains challenging — against natural rubber’s ~1,000 kDa or more (4). Getting there means pushing large carbon flux through the isoprenoid pathway and accumulating a long, hydrophobic polymer, both of which become a metabolic burden or outright toxic to the cell. In the plant, this is handled by building rubber on rubber particles: specialized organelles with a hydrophobic polyisoprene core enclosed by a lipid monolayer and membrane-bound proteins, where the key enzyme is a cis-prenyltransferase (1). So the question I want to test is whether we can mimic that. Produce high-molecular-weight rubber on protein-stabilized particles without poisoning the host.

Question 2:

I would use the yeast Pichia pastoris as my manufacturing chassis, fed on methanol as its sole carbon and energy source. I chose it for three reasons: it is a well-established, industrially proven host; methanol is a cheap, non-food carbon source; and it is the exact system in which Zhao et al. (2026) achieved the first de novo biosynthesis of cis-1,4-polyisoprene, with their plasmids and methods described in full. (4)

Zhao et al. (2026) the engineered yeast produced 87 mg/L of polymer at a molecular weight of about 10,000 Da — which the authors describe as among the highest achieved heterologously in a non-model microbe. (4) But ~10,000 Da is still roughly 100× smaller than natural rubber, which must reach 1,000,000 Da or more to have the same properties.

In the rubber tree, cells build this very-high-molecular-weight polymer on rubber particles: lipid-monolayer compartments studded with the proteins REF and SRPP. (1) The objective of my project is therefore to design a living system that forms similar particles inside the yeast — compartments that can safely store the growing polymer without making the cell toxic to itself.

Inputs: methanol (carbon + energy), a defined minimal medium (nitrogen, salts, trace metals), and oxygen, run as an aerobic fed-batch fermentation.

Biological process:

  1. The yeast takes up methanol through its native methylotrophic metabolism, feeding carbon into central metabolism and ultimately to acetyl-CoA.
  2. Acetyl-CoA is converted through the mevalonate pathway into the isoprenoid building blocks IPP and DMAPP. The rate-limiting step is HMG-CoA reductase, which Zhao et al. boosted using a rubber-tree enzyme (HbHMGR2).(4)
  3. Chain elongation. A cis-prenyltransferase — the Hevea rubber transferase (HRT1) — adds IPP units in the cis configuration onto the FPP primer, growing the cis-1,4-polyisoprene chain.
  4. Chain stabilization. The rubber-particle proteins REF1 and SRPP1 (from Hevea) are co-expressed to stabilize the growing polymer.

Output: cis-1,4-polyisoprene, accumulated inside the cells and recovered by harvesting, lysis, and organic-solvent extraction.

Assumptions: The plant rubber-particle machinery can be re-localized onto an engineered yeast compartment (a lipid droplet or peroxisome)

Question 3:

I would automate the bioreactor itself — the culturing and growth-control step. I’d use an automated parallel bioreactor system that continuously monitors biomass, dissolved oxygen, pH, and methanol, and uses software feedback to control the methanol feed and induction in real time. This step is worth automating because methanol is both the carbon source and toxic above a threshold: too little starves the isoprenoid pathway, too much kills the cells.

Question 4:

(a) How much. Target: enough natural rubber to cover U.S. demand, which is tire-dominated. The U.S. consumed 1.78 billion pounds of natural rubber in 2020 (3)

1.78 × 10⁹ lb × 0.454 kg/lb ≈ 8.1 × 10⁸ kg ≈ 810,000 tonnes/year

(b) How long. Using the demonstrated titer from Zhao et al., 0.087 g/L (4):

  • One large 500 m³ (500,000 L) reactor × ~50 fed-batch runs/year → 0.087 g/L × 500,000 L × 50 ≈ ~2 tonnes/year per reactor.
  • To make 810,000 t/year → ~370,000 such reactors. Not feasible.

So at current performance, this can’t be manufactured at national scale realistically. The gating problem isn’t reactor time; it’s raising titer (and molecular weight) by orders of magnitude first.

Question 5:

  • Whether the system can reach high molecular weight at all. The whole project depends on producing ~1,000,000 Da polymer, which no one has done. I’d need to build the compartmentalized (“artificial rubber particle”) strain and test it at small scale to see whether molecular weight rises meaningfully before anything else matters.
  • How to get the polymer out. It accumulates inside the cell, so recovery needs lysis and solvent extraction. That’s fine in a flask, but at industrial scale those extra downstream steps become a major cost and throughput bottleneck.
  • I don’t yet know if methanol metabolism can push enough carbon through the pathway to close the >1,000× gap to a useful titer.

References

[1] Yamashita S, Takahashi S. Molecular Mechanisms of Natural Rubber Biosynthesis. Annu Rev Biochem. 2020;89:821–851. https://www.annualreviews.org/content/journals/10.1146/annurev-biochem-013118-111107

[2] CBE JU / European Commission. Alternative Biomanufacturing Routes for Natural and Synthetic Rubber (HORIZON-JU-CBE-2025-RIA-03). https://cordis.europa.eu/programme/id/HORIZON_HORIZON-JU-CBE-2025-RIA-03

[3] U.S. Tire Manufacturers Association. Fast Facts: Sustainable Natural Rubber. https://www.ustires.org/resources/fast-facts-sustainable-natural-rubber

[4] Zhao X, et al. De Novo Biosynthesis of cis-1,4-Polyisoprene from Methanol in Pichia pastoris via Systematic Metabolic Engineering. ACS Sustainable Chem Eng. 2026. https://doi.org/10.1021/acssuschemeng.6c03644