Using deep learning methods to design thermostable rubisco activase

This project is pending review.
Pending Approval
  • $0
  • 0%
  • Private
    Not Launched

About This Project

Due to climate change, plants are subject to unprecedented levels of heat stress. Heat stress causes crucial thermolabile proteins, such as rubisco activase (RCA), to denature and aggregate, drastically reducing efficiency of carbon fixation. Using deep learning-based computational methods, however, we can design thermostable versions of RCA to confer thermotolerance, creating plants that can fix carbon and grow in increasingly extreme environments.

Ask the Scientists

Join The Discussion

What is the context of this research?

Plant thermotolerance is a complex phenotype with many contributing factors, but improving the thermostability of the protein rubisco activase (RCA) is a promising approach to increase thermotolerance.

At elevated temperatures in plants, RCA denatures and aggregates, losing its ability to reactivate inactive Rubisco—the central component of carbon fixation in plants. This causes a major reduction in photosynthesis and photorespiration. Thermophilic plants have evolved variants of RCA that are more thermostable, and some plants even conditionally express those thermostable isoforms at elevated temperatures. Indeed, thermostable RCAs have been shown to increase growth and thermotolerance in plants.

What is the significance of this project?

More frequent, severe, and unpredictable heat waves are threatening global agricultural yields. It is estimated that with every additional degree Celsius increase in global average temperature, wheat (Triticum aestivum) production is estimated to decrease by 6%, rice (Oryza sativa) production by 3.2%, and corn (Zea mays) production by 7.4%.

Recent advances in deep learning-based protein design methods can rapidly generate thermostable designs that preserve functional activity, such as RCA's activation of Rubisco. Such a functional thermostable design of RCA would create thermotolerant plants, combatting the predicted losses in yield described above.

What are the goals of the project?

Two criteria must be met for this stage of the project: (1) candidate RCAs must be more thermostable and (2) candidate RCAs must exhibit wild-type levels of functional activity at elevated temperatures.

Thermostability of de novo RCAs will be determined using established methods (e.g., circular dichroism or differential scanning calorimetry).

Functional activity of de novo RCAs will be measured at control temperatures (22°C) and at elevated temperatures (30°C, 35°C, or 40°C). RCA functional activity assays at these temperatures will characterize preservation of RCA's capacity for Rubisco activation as well as ATP hydrolysis.

Successful candidates will be selected for further testing.


Please wait...

We order custom protein sequences as encoded synthetic genes, in the form of eBlock Gene Fragments from Integrated DNA Technologies. Funding for this will enable me to experimentally characterize my designed protein sequence in vitro before I put them into plants. RCAs are all around 300 amino acids long , and at $.07/bp, this comes out to $12,000 for 192 unique encoded RCAs.

Consumables (e.g., reagents and substrates) will be used to run the actual assays to assess functional activity and melting temperature.

Tuition and salary would alleviate financial burden from my PIs, allowing me to work on this project more freely.

Project Timeline

This project will take about 9 months to complete. In general, this will be an iterative process going between generating designs, and ordering and testing said designs. As I learn more from experimental results, I can impose stricter in silico filters and order fewer designs to test each round.

Jan 01, 2024

Computationally generate first round of designs

Mar 01, 2024

Functionally assay designs

Apr 01, 2024

Computationally generate second round of designs

Jun 01, 2024

Functionally assay designs

Jul 01, 2024

Computationally generate final round of designs

Meet the Team

Joseph Min
Joseph Min
Graduate Student

Joseph Min

Hey! I'm Joe, a software-engineer-turned-grad-student. I'm broadly interested in applying computational methods of protein design to plants!

Project Backers

  • 0Backers
  • 0%Funded
  • $0Total Donations
  • $0Average Donation
Please wait...