Proteins are where the action is. The repository is where the problems are.

09/27/23   |   Written by Paul Reginato

If DNA is where the information is, proteins are where the action is. How can we get more action in climate biotech?

In this post in our series about Garden Grants, we discuss the role of proteins in nature’s economy, the rise of protein engineering in biotech, and some challenges to leveraging proteins in our own economy. We also introduce the new Homeworld Problem Statement Repository, where we will collect concise descriptions of high-impact problems in climate that we hope will catalyze community activity, starting with protein engineering. If there’s a problem you think the community should know about but you’re not submitting a full proposal to Garden Grants, submit it to the Repository! And if you’re a protein engineer looking for problems, keep an eye on the Repository as it grows!

If you have a project you want to get off the ground, we invite you to start the journey at the Homeworld Garden Grants website: our first call for proposals is scoped on protein engineering and sustainability, and the deadline is October 20, 2023. We are thrilled to help your ideas grow. 

Biosphere envy and nanomachines

I find myself wistfully looking to ecosystems, like a home I’ve left for the city but whose wisdom I feel I’ve lost and wish to reclaim. And I’m not alone: while few in our society want to fully “return to nature,” there is a collective feeling of biosphere envy. Year after year, the biosphere transforms over 100 Gt of air, water and rock into life and back. Can we revolutionize our unsustainable material economy to be circular and non-pollutive like the biological world?

At the molecular basis of biology’s material transformations are proteins, the nanomachines that by-and-large catalyze the work and organize the structure of the biosphere. A suite of proteins including RuBisCO catalyzes fixation of carbon into sugars that power all domains of life; nitrogenase catalyzes triple-bond breakage to convert nearly-inert nitrogen gas into ammonia to provide organisms with bioavailable nitrogen; ferric reductase catalyzes environmental iron into a form that can be used in metabolism across the tree of life; and PETase naturally evolved to digest polyethylene terephthalate plastic waste. It’s just a tad hyperbolic to claim that for every molecular structure there is a hypothetical protein that can selectively bind it, and for every chemical reaction there is a protein that can catalyze it. If DNA is where the information is, then proteins are where the action is.

Bacterial flagellar motor imaged using cryo-electron tomography, a protein nanomachine powered by catalysis of ATP hydrolysis. Panels A and B show cross-sections, while D and E show surface renderings. (Qin et al., BioRxiv, 2017, DOI:10.1101/303222).

Protein engineering – the modification or de novo creation of proteins imbued with desired properties – also happens to be the most rapidly-accelerating revolution in molecular biotechnology. We are constantly seeing new protein engineering capabilities emerge, enabled particularly by new AI tools. Most recently, diffusion models and large language models, most known for interpreting and generating images and language, have demonstrated astonishing success in de novo protein design. As biotechnology gets closer to the real action of biology and we look to the biosphere for inspiration, this begs the question: can we take a page from the book of life and shape a better future using proteins?

Structures of proteins designed and synthesized de novo using large language models (LLMs) in the Rives lab at Meta. (Hie et al., BioRxiv, 2022, DOI:10.1101/2022.12.21.521526).

Bridging nanoscale engineering to commodity-scale industry

Most innovation in protein engineering happens in the labs of bioengineering departments with focus and financing toward human health. We’ve seen frontier protein engineering revolutionize neuroscience through optogenetics, gene editing through CRISPR, and cancer therapeutics through monoclonal antibodies. In each of those cases, nanoscale proteins impact microscale cells: a tool too small to see impacts a system too small to see.

But for biotech to impact planetary health, we need tools too small to see to impact systems too big to see: gigaton-scale waste arises from the largest commodity industries. Can we use enzymes (which are a kind of protein) to produce nitrogen fertilizer, recycle plastics, decontaminate water, mine minerals, and circularly cultivate biomass from pure CO2 and electricity? For most of the protein engineering community, working on sustainability means learning not only a new set of applications and disciplinary adjacencies, but a whole new set of economic sensibilities.

One challenge in applying protein engineering to sustainability is predicting the properties required for a protein to be economically useful at scale: for example, if the protein is to catalyze commodity manufacture, the eventual product must be cheaper than the same product made by other methods. That is a very different economic situation than an engineered therapeutic antibody with no competitors. Techno-economic analysis (TEA) must therefore be used to predict what combination of activity and stability an enzyme must have under process conditions for a given enzyme unit cost in order to scale.

Another challenge for enzymes in commodity industries is engineering protein stability. To be cost-effective, an enzyme should be reusable for a long time under process conditions so that a large quantity of commodity product can be produced per unit protein.

A third challenge for engineered proteins in industrial processes is developing cost-effective protein screening assays that are relevant to industrial processes while using minute quantities of protein, to enable R&D. Engineered protein variants can cost >$100 per mg to test at lab scale and are typically used at > 1 mg/mL concentration, which means assays must use very small volumes while recapitulating industrially-relevant conditions in order to screen many variants in a single experiment. For commodities that are traditionally produced without enzymes, standard R&D experiments may use liters of material, so new assays may need to be developed.

As we at Homeworld Collective found in our recent roadmapping work for biotech in CDR, all three of those challenges are limiting efforts to engineer the enzyme carbonic anhydrase for application to direct air capture (DAC) of CO2, despite the possibility for an engineered enzyme to enable substantial reduction of DAC’s cost.

Nonetheless, protein engineering is a tool we frequently see in the hands of practitioners seeking climate impact, particularly in younger cohorts. Most are searching for impactful problems; others have new ideas but lack funding to work outside medicine; and of course many ideas in emerging application areas come from folks who have worked in sustainability for some time. As a practitioner-focused non-profit working to realize the potential for biotech in sustainability, we want to help these hungry protein engineers get to work. That’s why we scoped our first call for proposals to Garden Grants on protein engineering.

Growing ideas with the Homeworld Problem Statement Repository

Because many climate biotech practitioners are working to understand new application areas and find community, Garden Grants is designed to support the growth of knowledge and collaboration in addition to funding early-stage ideas. As discussed in our previous blog post, Garden Grants proposals have public problem statements and private solution statements. Our intention with that structure is to help applicants attract collaborators, thought partners, and funders while teaching the community about high-impact problems, all without divulging applicants’ secret sauce.

But what if you know of a great problem and aren’t yet working on a solution? What if you want to see more attention brought to a whole problem area? Shouldn’t problem statements be able to come first and stand alone, inviting brilliant minds to devise a solution statement? Now we introduce the Homeworld Problem Statement Repository, which hosts public problem statements that need solution statements. By hosting the Repository on PubPub, we enable problem statements to be honed via version updates in response to inline community discourse, and we enable problem statements to be cited via DOI. Anyone can share a problem statement in a Google Doc via this online form (also linked in the Repository page).

Screenshot from the Homeworld Problem Statement Repository on PubPub.

The repository currently contains three high-priority problem statements for enabling protein engineering of carbonic anhydrase to impact DAC, corresponding to the three common challenges mentioned above. They address needs for TEA to describe target protein properties, ultrastable enzymes that retain high activity, and development of miniaturized assays that test enzymes cost-effectively while being representative of scaled industrial processes.

Over the coming weeks of Garden Grants Call 1, we will add protein engineering problem statements to the Repository as we work to understand and support the space. We encourage you to participate in collective goal-setting by giving constructive criticism to problem statements inline or by submitting your own problem statements!

As biotechnologists, our work is to reason through dreams of a biological future, understand practical possibilities, clarify what is needed to get there, and build it. Let’s do it together.

By Paul Reginato