Q&A: Are plants the key to solving energy and food crises worldwide?

UNIVERSITY PARK, Pa. — Changing market conditions are increasing the need for cost-effective ways to produce biorenewable chemicals, biofuels and materials that can serve as alternatives to oil-based products. According to Costas Maranas, Robert V. & Gloria H. Waltemeyer Chair and Donald B. Broughton Professor of Chemical Engineering at Penn State, solutions to these problems could come from applying tools used in synthetic biology to plants and their microbial partners across the globe.

Maranas’s research team develops computer models and algorithms to help scientists better understand, analyze and redesign biological organisms, including plants. In this Q&A, Maranas, who holds an additional affiliation with the Huck Institutes of the Life Sciences, discussed how recent breakthroughs in modeling, artificial intelligence (AI) and systems biology are accelerating their work, helping scientists better understand plants and repurpose their feedstocks into fuel, plastics and more.

Q: What is the connection between bioengineering and biomaterials science?

Maranas: Bioengineering is a more broad term. A bioengineer’s ultimate goal could be to make a particular molecule, detoxify something or develop a new biological organism that can sense a specific stimulus. Biomaterials are a specialized topic within the broader field of bioengineering. When scientists refer to biomaterials, they are talking about using plants and microbes — microscopic organisms like fungi or bacteria — to synthesize plastics or the chemical building blocks they consist of, as well as various biomaterials.

Take just one of many examples: We can bioengineer organisms to overproduce lactic acid, a chemical building block of polylactic acid, which is a biodegradable plastic that can be used to make diverse products from disposable packaging to medical implants. Many plastics in use today have alternatives that can be manufactured, in part, using plants or microbes. Synthetic biology could theoretically be used to bioengineer substitutes for many plastics used today, but the key challenge is to do so in a cost-effective manner.

Q: How can computational tools make bioenergy production more efficient?

Maranas: My research specializes in metabolic modeling, a process that analyzes the genomes of different organisms and identifies which enzymes they code for in their DNA. Knowing the enzymes present, you can figure out the complete chemical repertoire of an organism, which helps us build a highly detailed blueprint of the organism’s structure.

Once we have this blueprint, we can reverse engineer the organism to make new chemicals or biomaterials. This can optimize its growth, fabricate more sustainable building materials, or even convert the sugars from plant feedstocks into biofuels such as ethanol. This process would be much more difficult without the help of these detailed computer models.

Q: Why is data science essential to modern agricultural and biological technology?

Maranas: Data science is fundamental to our work, and plays a multitude of different roles across agricultural, biological and engineering research. In agriculture, for example, multispectral imaging is one such application, where specialized cameras respond to different wavelengths of light emitted from fields of plants — with that information, you can deconstruct the plant species present in an area or even the health of a crop. This process generates gigantic sets of data, otherwise known as big data, just from photos taken with a drone or a satellite.

My research uses big data to identify a sequence of enzymes that can convert a starting chemical to a specific chemical of interest. This is typically referred to as chemical retrosynthesis. We’ve developed an online tool, known as CatPred, that uses big data and AI to predict how effectively different enzymes can react with different biochemicals.

Q: Has AI-driven modeling accelerated bioengineering breakthroughs?

Maranas: The emergence of AI has been a quantum leap for this field, and specifically large language models — AI systems designed to understand and produce text. The same way that tools such as ChatGPT have a vocabulary of words to generate text, teams have developed specialized protein language models that have a vocabulary of amino acids, meaning these tools can tell you which protein sequence will be best for a specific biochemistry. Additionally, many traditional AI tools have been repurposed for specialized tasks like designing enzymes that will perfectly bind to a particular molecule and react in a certain way.

A lot of these tools are being developed in many different labs right now, so it's a very exciting time in bioengineering, as things are developing incredibly quickly.

Q: What is systems biology? How might it shape a more sustainable bioeconomy?

Maranas: Systems biology looks at living organisms as integrated systems rather than isolated parts. Instead of studying a single gene or enzyme in isolation, it combines data, computational modeling and biology to understand how all the components of an organism work together, as well as how those interactions can be tailored to produce new molecules, materials, or possibly biofuels.

This approach has been responsible for the synthesis and commercialization of several biomaterials — polylactic acid is one — but each of these required years of effort and in many cases decades of incremental improvements to get them to a commercial stage. What systems biology could do is greatly shorten the time scale of arriving at commercially viable designs, learning from past failures and coming up with workflows that have a high promise of success. I believe this field will industrialize and systematize the process of discovering something in a lab and translating it into a product that people and businesses can use.