Hoxton Farms co-founder Max Jamilly: Computational modelling redefines cellular agriculture

The co-founder of the British cell-based fat producer explains why the combination of synthetic biology with computational modelling is its secret sauce for commercial success

By Murielle Gonzalez

In today’s emerging cellular agriculture ecosystem, fat is a sought-after ingredient. Why? Because it helps alternative protein products achieve the flavour and texture sensations of the animal counterpart – and this is the $20 billion market opportunity that UK start-up Hoxton Farms is tapping into with its proprietary biotechnology platform.

“We grow real cell-based animal fat without the animals,” says Hoxton Farms co-founder Max Jamilly. “We’re selling cultivated fat as a business-to-business ingredient to the meat alternative industry, starting with plant-based meat so that we can finally make meat alternatives that look, cook, and taste just like it.”

Based in London, Hoxton Farms is a newcomer in the market. Jamilly co-founded the company with long-time friend Ed Steele last year, and it has already attracted investors. San Francisco-based venture capital firm Founders Fund led a seed funding round for the company, which closed at £2.7 million in February with participation from other six investment firms.

The entrepreneurial duo is building an interdisciplinary team of stem cell biologists, engineers and food scientists. Jamilly is a synthetic biologist with two degrees from Cambridge and a PhD from Oxford, while Steele is a mathematician with degrees from Oxford and Imperial College London. They have known each other for 25 years as their friendship started at primary school in the early 90s.

Jamilly took centre stage on the opening day of the virtual summit Future Food-Tech last week and presented the business to Pae Wu, partner at venture capital fund SOSV, and Jim Thorne, president of Nourish Ventures, the venture arm of ingredients company Griffith Foods.

He explained that Hoxton Farms combines Steele’s background in machine learning and optimisation with his know-how in synthetic biology to build “incredible models” that redefine how to grow stem cells for cellular agriculture. But how do these models work, and what’s so incredible about them? NutritionInvestor caught up with Jamilly at the summit to find out more about the company and the challenges ahead.

Hoxton Farms: Cellular agriculture based on digital twins

Simply put, Hoxton Farms takes a handful of cells from an animal and grow them up in a bioreactor. The cells are differentiated into fat tissue, and the company harvest that fat.

Jamilly argues that two aspects unique to Foxton Farms’ process set the company apart from other cell-based companies – growing fat rather than muscle and the capacity to optimise the process thanks to computer modelling.

“Fat cells are much easier to grow,” he says. “They don’t need any of the stretching forces that align muscle cells as they grow. And fat cells don’t need any of the complicated cell fusion steps that go on into developing muscle fibres.” For Jamilly, these characteristics mean that the technology is a lot more scalable.

Jamilly is bullish about the scalability of the technology because that’s how Hoxton Farms saves time and resources while achieving a commercial product with added value to its customers.

“What makes our process unique is that we’re able to optimise it using the computational models that we build,” says Jamilly, and the conversation unfolds the beauty of working with digital twins – the virtual representation of a physical object or system across its life-cycle.

Typically, a multi-step bioprocess optimisation in cellular agriculture requires time and resources spent in experimentation in the lab before taking it to production scale – and half the time, the latter doesn’t work as expected, so work has to start from the beginning. Computational modelling gets rid of this approach.

“The models that we build, and you can think of them as a digital twin of the entire process, allow us to recapitulate the bioprocess on the computer without having to go anywhere near as much as in vitro experimental optimisation,” says Jamilly. He notes computational modelling makes the experimentation more reliable.

Biological data for customised products

Hoxton Farms is a cellular agriculture company like no other because most of the heavy lifting of the R&D work is done by its computational modelling. This approach allows the company to understand the process at the cellular level – in the lab and at scale.

Moreover, for Jamilly, the company’s digital twin of the biological process is the secret sauce for making fat with the optimised flavour, texture, and melting points that consumers love. Knowing this data allows for the production of fat that ticks all the boxes for its clients.

Jamilly argues that nobody understands entirely how it is that cells biosynthesise different kinds of fat. And nobody understands the exact relationship between the fat composition of the cell and the overall sensory performance of the resulting tissue. “With machine learning, even if we don’t understand the underlying biology, we can fill in the gap,” he says.

The models that Hoxton Farms builds describe how cells behave and interact with other cells and how they react with the hardware in the bioprocess. “We start with biological building blocks and use them to predict the way that a fat cell makes fat,” says Jamilly.

The resulting fat tissue produced by Hoxton Farms is exactly the same as its animal counterpart. “The challenge of doing cell culture is convincing a cell used to growing inside an organism to grow outside from it. But we do our very best to recreate the same conditions that the cell experiences inside the cow inside a bioreactor,” says Jamilly.

And because the fat is identical to its animal counterpart, it also has the same nutritional profile – and it’s cleaner and made sustainably.

Plant-based meat today contains various processed ingredients and high salt content. Many of these ingredients are there to make up for plant oil deficiencies, including flavour and melting temperature. Jamilly says optimised cultured fat gets rid of those, making the product healthier.

“A huge amount of animal-borne disease comes into humans through consumption of meat produced from intensive agriculture. And the interesting but gruesome statistic is that the maximum threshold for faecal contamination in the meat that comes out of traditional abattoirs and slaughterhouses is not zero,” says Jamilly. “So, the disease risks are much lower with cultured products, and nutritional profile is the same, if not better.”

Looking ahead

Work continues apace on the computational modelling front. Hoxton Farms is trying to create animal fat from stem cells and produce the ingredient optimised for taste, texture and melting point specific to the formulation required by its clients.

“We’re at the early stages of developing the scaledown models, but those combined with the optimisation hugely improve our predictive validity. And that’s what we’re really excited about,” says Jamilly.

Hoxton Farms is building the know-how to create the key ingredient for next-generation alternative meats, but the consumer perception of synthetic biology and cultured products is yet to be seen.

“We think that all of the signs points in the right direction and that consumer attitudes are changing,” says Jamilly. “People want something that’s really tasty, healthy, good for the environment, good for animal ethics and sustainability – and we think we tick all those boxes,” Jamilly concludes.