Shiru Vice President of Business Development Julian Lewis
Shiru Vice President of Business Development Julian Lewis

What makes a burger patty mouthwatering? While the simple answer is because they are made of meat (typically beef), there’s more to this question than meets the eye. The answer lies in meat’s unique mixture of fat and umami and the culinary magic that occurs due to the Maillard reaction between amino acids and reducing sugars that gives browned food its distinctive flavor. When trying to make a replacement for meat, these food features need to be considered.

The world eats a lot of meat, especially beef, with some estimates of global consumption weighing in at more than 27 billion pounds. The average American eats somewhere close to 70 pounds of beef per year. But because cattle are the least efficient livestock for converting feed into edible meat, environmentalists have long argued that reducing beef consumption is an important conservation act. The missions for many plant-based food companies like Beyond Meat and Impossible Foods, which are recreating some fast-food favorites like ground beef for burger patties, hinge on the notion that the greatest effect an individual can have on the environment lies in the decision of what we put on our plates.

Shiru is a biotech company pioneering next-generation ingredients in food manufacturing and production. Powered by a team of technology experts and food industry veterans, Shiru is on a mission to reduce the world’s reliance on animals for food by providing delicious, cost-effective, and sustainable plant-based alternatives.

Founded in 2019 as a Y Combinator-backed startup, Shiru has developed a precision biofermentation process combined with machine-learning algorithms to discover and create novel plant-based food ingredients. Founded by protein biochemist and entrepreneur Jasmin Hume, PhD, Shiru’s goal is to make a full suite of sustainable ingredients to help conserve water, decelerate global warming, and halt the extinction crisis. The biotech startup’s patent-pending discovery platform combines machine learning and biofermentation to produce plant-based alternatives to eggs, meat, dairy, and gelatin—using a fraction of their animal counterparts’ land, water, and energy.

The biotech startup closed a $17 million Series A round in October 2021, led by S2G Ventures, a multi-stage venture fund investing across the food, agriculture, oceans and seafood markets. Returning investors Lux Capital, CPT Capital, Y Combinator, and Emles Venture Partners as well as new investors The W Fund, SALT, and Veronorte also participated, bringing Shiru’s total funding to date to more than $20 million.

GEN Edge spoke with Julian Lewis, VP of Business Development at Shiru, about how the company develops functional ingredients for various products that currently require animal-derived proteins, from packaged baked goods and sauces to burgers and yogurt.

GEN Edge: What has Shiru set out to do? 

Lewis: Shiru’s mission is to become an ingredients business and provide a better toolkit for food developers. We’re not going to produce food as a finished product. Instead, we’re going to be in the B2B supply chain, providing better ingredients to enable the food industry to develop plant-based foods and better ingredients.

We believe that the diversity of natural plant-based proteins is so extensive and potent that we can achieve many of the things that food developers need. Depending on the source, there are anywhere between 30–50,000 known edible plants on earth, and every individual plant has got around tens of thousands of different proteins. Essentially, there are millions, if not billions, of potential plant-based proteins out there. Yet, when we look at what plants we consume, three types of plants—rice, wheat, and maize (corn)—occupy 60% of our caloric intake. Out of the tens of thousands of plants, the food industry primarily uses only three.

We’ve built a database of proteins to find new functional proteins, pulling together many different sources and putting them into one interrogatable well-organized database. We’ve built out lots of features alongside each protein that describe some of the performance characteristics of those proteins. If we can match a performance characteristic of a protein to the desired functionality in food, then we’ve found a new ingredient. We’ve built a sizable proprietary feature set against each of these hundreds of millions of proteins in the database, and we then run machine learning algorithms to interrogate the database. Some of those interrogations are pretty mainstream standard stuff that many people could do, but some of it is very unconventional, resulting in unexpected results.

We’re finding alternative proteins that can perform functions you’d never consider in a million years! Some of the results are pretty exotic. Once we’ve identified a shortlist of candidates that may deliver the functionality we’re looking for, we validate them in the lab. We express them through fermentation, validate them using automated high-throughput screening, and then filter more into specific assays that will validate the performance we’re looking for. We’ll then start to create samples that we can put into real food that you can eat just to check it does work.

Our first mission was to look for proteins that could deliver food with the quality of gelation—things that gel. Eggs often bind or gel foods together, but plant-based foods don’t use eggs. Instead, they will use ingredients like methylcelluloses or gums, which the food industry tells us they want to get rid of and don’t like for various reasons. So, we look for a natural-based protein that could replace egg or methyl celluloses, which we’ve discovered.

Gelation is just one type of function. There are many other functions of ingredients in food, like color, binding, softness, or resilience. There are a million different things that a food developer needs to do to affect the texture of foods, and they need a tool kit of ingredients to help achieve that.

GEN Edge: What does Shiru’s R&D pipeline look like? 

Lewis: We start with a dry lab. The origins of our search start digitally—in silico. We’ve got data scientists organizing the data, bioinformaticists unpacking what the data are telling us about protein functionality, and then machine learning engineers building the algorithms. We then need to validate them in the wet lab, where we ferment the protein, expressing them at a micro-titer level. Our method of production is precision fermentation. Once we’ve identified a protein—it might, for example, come from a very exotic mushroom grown on the steppe of Mongolia. Instead of going to Mongolia, we reproduce that protein using precision fermentation. We’ve got fermentation engineers to produce the protein at scale.

The proteins then go through an automated high-throughput screening and then into the characterization team, where they build assays that relate to the functionality we’re looking for. To find proteins with specific functionality, we need food scientists. For example, for gelation, we develop assays that will say whether a protein gels or not in a scientific way and whether it would work in food, which is a complex system. Checking protein functionality in a complicated recipe with many parameters—pH and interactions with other ingredients, such as sugars and salts in the food environment—is tested in the lab.

 GEN Edge: Does Shiru have all of these capabilities in-house?

Lewis: We’ve got the skeleton of it all, but we’re working with partners. In terms of the dry lab and high throughput, high-value characterization, we’ve built that. That’s the core of what we do. In terms of food science, we’re not experts in food science. We’re experts in protein science, and we’re working very closely with several global food manufacturers because they know their food science. So we’re working with them while building our own capability for strain engineering and fermentation.

We do have those engineers in-house, but we haven’t got all the different scales of fermentation tanks yet. We’re working with CROs and partnering with large current B2B ingredient manufacturers. CP Kelco is in the business of fermentation and helping us scale up. We’re both learning together in terms of how we manage that fermentation. They’ve got some expertise but don’t understand our materials. We understand our materials, but we don’t have all the kits. Our business model relies heavily on partnerships to fill in the gaps of what we don’t have. But ultimately, our ambition is to build that as a complete business.

 GEN Edge: How important to your operation are machine learning algorithms?

Lewis: The more data you’ve got, the better the machine learning. We have a lot of failures in our search, but failures are just as good as successes because it’s data. Our search starts [by examining and understanding the makeup of] an existing material. For example, when we’re looking for something to replace eggs, an excellent place to start what’s the makeup of an egg. Understanding the structures in them that deliver the gelatinous quality is where we start. One conventional thing that every kind of R&D business could do is look for structural homology.

We also look for unconventional connections, such as the feature sets of each protein, like physical parameters, surface hydrophobicity, and thermostability. We’ve got over 200 of these features, creating quite an extensive database. If we can understand the features of the starting material or other materials—what’s in our database that has similar features— [this highlights] features which we believe are important. Not all features are essential. When looking for feature similarity, you might find answers that you would never find with sequence homology. That gives us surprising results.

Another part of the algorithm borrows from neurolinguistic programming and the machine learning science of embeddings to look at connections between combinations of features, structure, and sequence. This takes us to very bizarre parts of our database, and the results are astonishing. Some of those results appear valid, which will feed into finding deeper connections.

 GEN Edge: Why was the company named Shiru? 

Lewis: Shiru is an ancient Chinese word that means “like meat.”  One of the things we’re enabling are products that are “like meat.” We discovered there’s a very ancient tradition in the old Chinese empires of creating huge banquets, which were planted, creating all kinds of foods that look like big joints and meat that were all plant-based. As with so much in life, it’s already been thought of before, and we’ve forgotten. Here we are, trying to reinvent stuff that’s already been thought of before.

Shiru also is found in other languages with different meanings; fortunately, they’re all quite pleasant. In Japanese, the word has to do with wisdom. Something in ancient Hebrew is along those lines. We’ve yet to act if they come across like one very unfortunate translation.

GEN Edge: Tell us about the company’s trajectory.

Lewis: Over the next year, we will focus on the proof of concept to lead us to commercialization. We’ve got commercial partnerships with businesses that are scaling up with us, and they will physically make food, which will work or won’t or will fall somewhere in between. We want proof of concept that will lead to commercialization. These partners will prove that what we’re doing works in their systems. Then we’ll scale up with them and get something on the market in the next year or two. We probably won’t be on the market within a year, but we’ll be on the pathway within five years. In the next five years, we’ll have brought to market more materials in different areas, and we’ll be looking to produce those ourselves. So we’re looking to grow the business with our manufacturing. In five years, there will be products on the market with Shiru-discovered proteins.

Our overall strategy has three steps: replace, unlock, and transform. At the moment, we’re looking for solutions that can replace undesirable ingredients and food. We recognize that the future of plant-based food is limited by how appealing and delicious it is. Some of it’s great and some of it sucks! So, how can we unlock the potential of plant-based foods and make something that doesn’t exist? How can we make plant-based cheese tasty? How can we unlock the potential of these foods by creating new texture solutions to unlock them?

The food industry hasn’t transformed for hundreds of years. They’re still making the same stuff and versions of the same stuff. But there’s new stuff to make, which might be new types of food that we can’t even imagine. We’re not just trying to replicate a meat burger or dairy cheese. We’re thinking about what else is new and how to facilitate that. We’ve got an eye on the future. That’s how we will make food in space when there’s no gravity, which might make food production tricky, or on Mars.

 GEN Edge: What are the biggest challenges you’re facing?

Lewis: The biggest one is just our ability to scale quickly. We want to partner with the large food businesses that see this challenge but do not know how to move forward equally. We’ve got some of the answers, and they’ve got some of the solutions, resources, and capabilities. It’s all about partnership and sharing because not one of us on our own will solve this. The challenge is finding partners willing to genuinely partner to blow away some of the barriers.

Previous articleTwo-Step QS-21 Purification Simplifies Process for Greater Use
Next articleNeuroString Is a Tissue-Mimicking Neurotransmitter Sensor