Meet the biotech startups brewing milk in bioreactors to improve on baby formula.
Lab-grown breast milk is gearing up to disrupt the infant formula market. New York–based startup Helaina raised $45 million in September 2024 to develop a breast-milk-equivalent protein, and California-based B corporation Checkerspot raised $55 million in 2023 to create human milk fat through microalgal fermentation. A number of startups are also embracing biotech tools to optimize milk for infant nutrition and achieve a product closer to the gold standard for babies: mother’s milk (Table 1).
When breastfeeding isn’t an option, human milk made in the lab could meet babies’ nutritional requirements. Credit: Kelvin Murray / Stone / Getty Images
Growing consumer enthusiasm for all kinds of dairy alternatives is helping lab-made breast milk gain traction. Lab-made alternatives — from cultured plants and animal cells — are also contributing to a shifting commercial landscape. The global cell-based milk market was valued at $450 million in 2023 and is expected to see a compound annual grow rate of 8.8% through 2029. The same biotech advances are turning lab-made human breast milk into a nascent industry.
The experts say that breast is best, but cultivating mammary cells could offer a product with nutritional characteristics closer to the unique composition of human breast milk for those who seek — or need — an alternative. Demand for infant formula is extremely high: only 25% of newborns in the United States receive are exclusively breastfed in their first six months.
Traditional infant formula is typically made of modified cow’s milk. Its aim is to mimic breast milk, but as human breast milk is a living fluid with over 2,000 ingredients, Rafael Perez-Escamilla, who is director of the Maternal and Child Health Promotion program at the Yale School of Public Health, a $70 million research program that contributes to improvements in breast feeding and infant nutrition, says this is an elusive goal. “There are no two women in the world that have the same composition of milk because when the baby latches to the mother's breast, the immune systems communicate with each other.” He adds that even during a single breastfeeding session, the composition of human milk changes.
Startups are rising to the challenge. Companies making cell-based milk are saying they can narrow the gap between human milk and infant formula by producing milk that features some of the sugars, proteins and fat found only in human breast milk.
Cell biologist Leila Strickland first looked at producing cell-based human milk in 2009, after she became a mother and found breast feeding challenging. She cofounded Biomilq, in Durham, North Carolina, to make cell-based milk and has since raised $25 million with backers that include Bill Gates’s investment firm Breakthrough Energy Ventures.
Biomilq uses mammary cells isolated from breast milk samples donated by lactating women. When placed in a nutrient-filled bioreactor, the cells fix themselves to a scaffold of perforated straws. While basal cells take in the nutrients from the culture medium, apical cells secrete the milk, which can be extracted from the bioreactor with a syringe. This process takes 30 days.
So far, Biomilq has successfully produced a complex mix containing human milk oligosaccharides (HMOs), which play an important part in the development of the infant’s microbiome, as well as the multifunctional protein osteopontin, which is abundant in human milk but not in bovine milk or infant formulas. Strickland says the HMOs and osteopontin components could be ready by 2028 and launched by 2035. Biomilq’s team has also been able to make the mammary cells express milk rather than individual compounds using this process, but it’s not yet developed to the point where they can scale it for commercial use.
Singaporean startup TurtleTree launched in 2019 with the goal of creating milk without cows. Former Google employee Fengru Lin first came up with the idea of cell-based milk production when faced with the difficulty of sourcing raw milk for her cheese hobby. With the rise of cell-based meat, Lin wondered if you could create cell-based milk too. The company set up to produce both dairy milk using bovine mammary cells and infant formula using human mammary cells.
After a $32.9 million Series A raise in May 2021, Lin and her cofounder, Silicon Valley entrepreneur Max Rye, were able to take TurtleTree to the United States. They now have 42 employees and offices in Singapore, at the University of California, Davis (UC Davis), and in Boston, Massachusetts.
They, too, have been able to produce milk by culturing mammary cells. At first, the cell-based milk was underwhelming. “We are able to detect some proteins, some sugars in the output, but the challenge is to get it to high enough volumes so we can purify the protein and make it into a product that is priced to scale,” says Lin.
To help bring a cultured milk to market faster, they turned to precision fermentation with engineered yeast. Using this method, they produced the immune-boosting protein bovine lactoferrin, in a form they call LF+. “We engineer the yeast to take in sugar and pump out a protein comparable to bovine lactoferrin, without the cow,” says Lin. It takes a week to produce the TurtleTree lactoferrin LF+, which Wisconsin’s Cadence Cold Brew coffee company plans to include in a sports drink. TurtleTree’s next goal is to add the bovine lactoferrin to infant formula. “We are aiming to get our infant approval about two years from now,” says Lin. Human lactoferrin will follow, but the process will be longer because of the need for regulatory approvals.
Helaina, founded in 2019, has produced human lactoferrin. The company uses gene-editing tools to modify yeast strains and have them produce a protein with a structure and function that are the same as those found in the human body.
The goal of Helaina’s food scientist founder Laura Katz was not to recreate whole milk for infants, but to focus on specific bioactive proteins with health-giving properties that could boost the health of all ages. Helaina uses a glycoengineered yeast, Komagataella phaffii, to produce human lactoferrin, which is found in colostrum, the first run of human milk. Helaina staff place the food and modified yeast in a bioreactor, and within a week the yeast spits out the protein. They filter the protein from the liquid and spin it into a powder, which is now sold to nutrition and food companies, such as Mitsubishi.
Massachusetts biotech company Conagen has also been working on lactoferrin and HMOs since 2022, but unlike others, its focus is solely on baby formula. Conagen is creating lactoferrin through precision fermentation and the HMO 2′-fucosyllactose through a bioconversion process using enzymes.
The Conagen team chose to optimize the sugars in baby milk. “Most of the sugars you’re going to get in infant formula [are] sucrose or glucose. There’s not a lot of that in breast milk,” says Casey Lippmeier, senior vice president for innovation at Conagen.
To make lactoferrin, Conagen grows human mammary cells in a bioreactor with glucose, macronutrients and trace minerals. The product is then separated from the saline solution in which the cells grow by chromatography and finally dried. The 2′-fucosyllactose, which helps protect against infectious disease such as influenza, is made from l-galactose through four enzymatically catalyzed reaction steps. Conagen is now seeking regulatory approval for it to be included in baby formula.
Human milk sugars and protein are not the only compounds in startups’ sights; the Rehovot-based Israeli company Wilk is focused on fats. Wilk was founded in 2020 by lactation scientist Nurit Argov-Argaman to create a product that would bridge the gap between baby formula and breastmilk. Taking inspiration from the techniques used by cell-based meat companies, Wilk uses donated cells from human milk to recreate HMOs and fats found in breast milk.
To make fat, the company feeds human mammary cells with plant-based fatty acids; for the sugars, it uses a patented technology that provides the building blocks for HMOs. Within 24 hours, the mammary cells make the fats or HMOs. The fats are removed from the nutrient-filled medium through centrifugation and the sugars through a process of sampling, drying and using special crystals to separate out the HMOs.
Zohar Barbash, Wilk’s CTO, says the biggest challenge is scaling up. As they move from 1-liter to 1,000-liter bioreactors, they need to adapt the growth medium to make sure the cells not only proliferate but continue to produce fats and sugars. They also want to bring down the cost. “Right now, the 1 gram we are producing costs us about €1,000,” says Barbash. Backed by the French dairy giant Danone, Wilk hopes to have a commercial product by 2027.
Others have palm oil, used in traditional baby formula, in their crosshairs. Californian B Corp Checkerspot specializes in creating materials such as lubricants from triglycerides. To make a more sustainable alternative to palm oil for use in formula, the company recreated the primary fat in human breast milk: OPO (1,3-dioleoyl-2-palmitoyl glycerol). The Checkerspot OPO was created by placing the microalga Prototheca moriformis in a bioreactor and feeding it sugar in the presence of nitrogen. When the nitrogen stream is shut off, the cells stop dividing and start converting sugar into triglyceride oil instead. The oil is then removed from the cells in the same way as in the oil seed industry.
Scott Franklin, CTO of Checkerspot, explains that they can control the degree of saturation of the fatty acid backbone and move the positions of hydroxylated fatty acids on that backbone to match the human equivalent. They are now seeking regulatory approvals for the release of the HMO and fat they’ve created.
Researchers at UC Berkeley and UC Davis have found that you can also recreate human HMOs in plants. In 2023, a team genetically engineered leaves from a Nicotiana benthamiana plant, which is a close relative of tobacco, to create HMOs.
Patrick Shih, a plant biologist and microbiologist at UC Berkeley, says that although it’s possible to genetically engineer microbes such as Escherichia coli to make sugars, it can be an Sisyphean task. But plants make sugar naturally: “Plants are sugar-producing factories, so we saw plants as a unique opportunity to make certain classic compounds.”
To emulate the path by which plants take CO2 from the air and turn it into sugar, the team introduces various enzymes into the plant’s leaves. This method makes it possible to create different glycosidic linkages and in turn build oligosaccharides that are close to the human equivalents. Within a week, the group can have their targeted HMO. They were able to make 11 HMOs from all three major groups of HMOs.
Shih says that by using plants, not only can they make a diverse group of HMOs, but the process will be cheaper to scale. They are now looking for partners to help them take this from the lab to the commercial scale.
Yale’s Perez-Escamilla says that although there could be a place for cell-based compounds, he has some concerns. And those are not only about the concentrations given to infants but also in regard to regulations. “I want regulatory agencies to understand these are bioactive substances. If it's not handled properly, they can also have negative side effects. It requires extensive ethical and regulatory conversations,” he says.
