Israeli startup Remilk makes dairy products without the need for a single cow in the process

Israeli startup Remilk hopes to be the first company in the world to bring real dairy milk products to market without needing a single cow in the process.

“Relying on animals to make our food is no longer sustainable,” said Aviv Wolff, who co-founded the startup up last year. “The implications of animal farming are devastating for our planet.”

The time has come, he said, for “a new revolutionary invention that will enable the transition to a food system that takes no more than what our planet can give.”

The technology Wolff and his scientist partner Ori Cohavi have developed produces milk proteins that are “chemically identical” to those present in cow-produced milk and dairy products, explained Wolff in a phone interview. That essentially means they are dairy products with dairy proteins, minus the cows.

The two entrepreneurs mapped out the chemical composition of milk, assessed the fat, lactose and sugar in the liquid, and determined that the key ingredient to making milk is the proteins.

They thus set out to recreate the proteins by taking the genes that encode them and inserting them into a single-cell microbe, which they manipulated genetically to express the protein “in an efficient and scalable way,” said Wolff.

Using a microbial fermentation process, they increased the number of proteins, which they then dried into a powder.

“We’re making dairy products that are identical to cow-milk products, with the same taste, texture, stretchiness, meltiness, with no cholesterol and no lactose,” he said.

“We’ve basically ported the whole mechanism of producing milk into a single-cell microbe. We don’t need the ‘rest of the cow,’ and we surely don’t need to spend resources in the process of creating a 900-kilogram animal.”

This model of food production will be up to 100 times more land efficient than the existing dairy system, 25 times more feedstock efficient, 20 times more time-efficient, and 10 times more water-efficient, he said.

When mixed with water, plant-based oils like coconut oil or sunflower oil, and plant-based sugar, the milk liquid and its derivatives can be produced with exactly the same properties, taste and structure, he said.

The dried protein will be sold to dairy companies, and manufacturers can add water and fat to create a range of cheeses, yogurts and ice creams, Wolff said.

Will the dairy products be considered milk-based, for Jewish dietary considerations?

“I am not a rabbi,” said Wolf. “But not a single cell is taken from a cow, as even the gene is a fully synthesized gene. There is no animal in any part of the process.”

So technically, the product is a non-dairy product. On the other hand, he said, the protein is actually a milk protein. “In the lab it is exactly the same” as a protein that came from a cow, he said.

Wolff said the company is not yet allowing observers to taste the milk. But the firm has held double-blind tastings for research and development purposes with independent audiences that were not able to distinguish the alt-cheese products from traditional cheeses, he said.

The product is ready but not yet ripe for commercialization, as the partners are now trying to lower the production price of the protein, he said. Because of the generally low price of milk globally, the company will not initially be able to compete with liquid milk and will focus initially on cheese products, Wolff said.

Wolff is a former combat commander in a special forces unit of the IDF. After eight years of military service, he worked in several startups until he set up Remilk.

His partner Cohavi has a PhD in protein biochemistry from the Weizmann Institute of Science and formerly led research and development activities in biotech firms.

The firm employs 10 workers and is seeking to add an additional 15 within the next six months, said Wolff.

The market for dairy alternative, or plant milk beverages made from soy, almond, coconut, oats and hemp, is projected to grow from $21.4 billion in 2020 to $36.7 billion by 2025, according to research firm MarketsandMarkets.

These alternatives offer nutritional perks, such as lower cholesterol levels, lower sugar levels and improved cardiovascular health, all of which have led to an increase in consumption, the report said.

According to a World Wildlife Fund report, demand for dairy products continues to rise due to the global population growth, rising incomes, urbanization and the westernization of diets in Asian giants such as China and India. This increases the pressure on natural resources, including soil and water.

There are some 270 million cows that produce milk — along with greenhouse gas emissions that add to climate change. Furthermore, dairy farming and feed production can lead to the loss of ecologically important areas such as wetlands, forests and prairies. Global estimates say that to produce one liter of milk, a massive 1,020 liters of water are needed.

Remilk is not the only tech company working on alternative milk products. Australian-US firm Change Foods uses bioengineering technology to create animal-free cheese and dairy products, its website says. San Francisco-based New Culture says it makes “cow cheese without the cow.”

Germany’s Legendairy Foods also uses microbial fermentation to produce the same proteins that are found in cow’s milk, and Belgium’s Those Vegan Cowboys says it makes products with the same of cows’ milk but without the use of animals.

“They use similar processes,” said Wolff. “But we have very strategic commercial partners working with us, and we will be quick to go to market.”

In a different vein, Israel’s Bio Milk, which is planning to go public in Tel Aviv by merging with a shell company, isolates the milk-producing cells in cows’ udders and transfers them to a bioreactor, where they are exposed to materials patented by the firm to produce milk.

The dairy industry has been criticized by animal rights organizations, such as People for the Ethical Treatment of Animals (PETA), which say that dairy cows are treated like milk-producing machines. Most cows raised in the dairy industry are intensively confined, repeatedly inseminated to give birth to calves and then are separated from their newborns at birth, so the milk produced to nourish their offspring can be sold to humans. Genetic manipulation, and, in some cases, antibiotics or hormones, are also used to cause cows to produce additional quantities of milk.

Wolff targeted early 2022 for Remilk to offer its first commercial, affordable product. “Remilk will be the first to market, we believe.”

---

The global population is estimated to reach 9.1–9.7 billion by 2050, with approximately an 1.8-fold increase in income per capita1,2. Food consumption is predicted to become increasingly based on animal derived products – mainly due to increased wealth in developing countries2.

The production of valued protein rich foods, such as milk is estimated to increase by 62% (from production levels of 2005–2007 to 2050, i.e. from 664 to 1 077 million tons)2. Animal derived products are less sustainable to produce compared to their vegetable counterparts.

This is why it is important to either produce products that can be used as an alternative or develop better ways of producing animal derived products. Certain subcomponents derived from milk, especially proteins, are important due to their functionality and nutritional value, which are why technologies that can produce these more sustainably would be beneficial.

Recombinant protein production could provide an alternative source of milk proteins, as engineered microorganisms have the advantage of producing specific compounds in elevated amounts3.

This approach has revolutionized the biotechnology and pharma industry4,5 though proteins for human consumption have not yet been subject to such trends, as competition from traditional industries and GMO controversies have been of concern. Proteins used as ingredients in the food industry have traditionally been based on isolates from natural sources – with milk proteins as a major driver6.

The nutritional and functional capacities of certain milk proteins make them an obvious choice when fortifying or formulating products and the production of milk proteins by recombinant means could offer a sustainable additional source in the future. It is however nontrivial to evaluate the economic feasibility of such an endeavor, as cost of substrate, price of milk proteins and technical performance becomes decisive.

The Genome-Scale Metabolic Network (GSMN) is the collective knowledge of metabolic routes within a given organism, while Flux Balance Analysis (FBA) is the simulation within these networks that allows metabolites to change accordingly7. The sandbox like nature of the GSMN makes it possible to use it as a tool for simulating metabolic conversions and growth.

Here we made a model where the upstream feasibility of recombinant protein production was explored within the GSMN of the E. coli (iAF1260)8 for the milk protein, Bovine Alpha Lactalbumin (α-La) (Uniprot: P00711). Briefly, this was achieved by associating the required metabolites for α-La production with costs and dynamically simulating the optimal production with respect to different objective functions, i.e. maximizing productivity using sugar and soybean meal as substrates.

Simulations within the model were assessed and evaluated in terms of economic potential. The results indicated a potential for recombinant food protein production in the future.

The dairy industry has within the last decades been revolutionized with respect to processing and purifying proteins from milk and whey. Proteins can be isolated from whey, a side product generated during cheese manufacturing, which historically has been a waste product.

Revenue generated from protein extraction has been increasing in recent decades and has comprised a growing part of the total revenue of the dairy industry. The market for whey proteins has seen a predominant growth within whey protein isolates (WPI – above 90% protein purity) compared to lesser whey protein concentrates (WPC34–34% protein purity) within the last decade, implying that pure proteins with functional properties are in high demand.

The total production of WPI was 36.800 mt in 2014 for the US, which is an increase of 2.5 fold compared to the total production in 2006–14.400 mt (http://usda.mannlib.cornell.edu/MannUsda/viewDocumentInfo.do?documentID=1052). WPI’s or fractions of these are produced by a variety of filtration systems and chromatographies6.

Among the products formulated with whey proteins are infant formulas, where especially the Bovine α-La is used to mimic the content of Humane α-La found in human milk due to its beneficial properties69. α-La is sold at different purities in powder form – ranging from 35% to 65% α-La purity in 80–90% total protein content. The price range lies approximately between 15 €/kg to 25 €/kg according to purity as of 2014 (Inquiry at international α-La vendor).

Processing

E. coli is a prokaryote industrial platform which has been widely used for recombinant protein production in biotech and pharmaceutical applications and could be a candidate for recombinant food protein production. While there are a variety of different technical obstacles that makes recombinant protein production challenging, it is important to recognize that the underlying biological system is capable of converting biomass into specific products at a rapid pace, which could be utilized to secure adequate amounts of quality food proteins.

Different strategies employed in recombinant protein production enable streamlined downstream processes and recombinant products can in general be rendered available for processing by multiple initial strategies, including intracellular recovery, extracellular secretion or intracellular recovery from inclusion bodies11.

The cost of purification is a significant factor that affects the overall feasibility, though the current market of GMO food graded vitamins, preservatives and enzymes suggests that cost effective processes are available.

The cost of handling and terminating GMO material is an additional expenditure for any recombinant process, though the cost can be mitigated by recycling the microbial biomass produced. One approach is to seek authorization to use specific terminated GMO as feed products. T

erminated GMOs used as feed material for animals can create value associated with sales and reduced cost in waste management, providing a sustainable alternative. Insulin producer Novo Nordisk and amino acid producer Ajinomoto Eurolysine SAS are examples of companies with different types of end-products which have sought such an authorization for some of their terminated GMO by-products (http://ec.europa.eu/food/dyna/gm_register/index_en.cfm).

Legislative concerns

The use of GMO derived ingredients intended for human consumption is under legislative control. Approval from governmental authorities is granted if sufficient documentation is presented in regards to production methodology and consumer safety. The probability of achieving usage authorization for α-La isolates from E. coli can be evaluated based on prior cases.

The US FDA GRAS Notice Inventory, lists food ingredients that are regarded as safe to use and includes protein isolates from milk and whey (See GRN No. 37 and GRN No. 444 (http://www.accessdata.fda.gov/scripts/fdcc/). E. coli strain K-12 is among other used in the production of the recombinant enzyme, Chymosin, applied in cheese manufacturing.

Chymosin is regarded as safe to use, when produced in certain non-pathogen microorganisms, including E. coli strain K-12 – according to § 184.1685 (http://www.gpo.gov/fdsys/pkg/CFR-2012-title21-vol3/pdf/CFR-2012-title21-vol3-sec184-1685.pdf). Chymosin is an example of a food ingredient that is dominating the market, as the natural source is not available in sufficient amounts nor at attractive prices.

For theoretical α-La protein isolates from E. coli K-12, both compound and organism have been approved separately in prior cases for food formulation, which indicate that authorization could be obtained.

Production

The formation of amino acids followed by their subsequent polymerization into proteins is a notable expenditure in metabolites, as it requires both substrate derived building blocks and energy in the form of ATP12,13. The increased cost of amino acid compared to sugar does not decrease the feasibility of using amino acids in the model, as higher amino acid fluxes results in higher productivity of the recombinant protein.

The capacity of the TCA cycle has been reported as a limiting factor during recombinant protein production in E. coli on minimal medium as the formation of intermediates for amino acids and intermediates designated for oxidative phosphorylation competes for substrates14.

Media composition is an important parameter for industrial processes. The shift between protein production and growth is only limited in the model by the metabolites required for mRNA and subsequently protein synthesis, which implies two things: (1) the translation initiation of recombinant protein is not limited by physical constraints and (2) the utilized strain is capable of mimicking the required production profile through a controllable system.

Elaborate algorithms and models exist that can overcome limitation at translation initiation by rational design15,16, while several controllable systems for protein production are available in bacteria at industrial scale17.

Protein translation involves four phases: initiation, elongation, termination and ribosome turnover, where initiation usually is the limiting factor in recombinant production11,16 implying that the applied genetic material is limiting its own protein synthesis (inadequate transcription levels, mRNA instability, unfavorable ribosomal-binding-site complex).

In the case where initiation limitations are overcome, the native genetic blueprint of the organism becomes the limiting factor for recombinant protein synthesis. In such cases the protein elongation rate in conjunction with the amount of ribosomes becomes the limiting rate – as detailed in the model. In vivo studies additionally suggest that the maximum elongation rate per ribosome is higher than the experimental observed18.

eferences

1. United Nations, Department of Economic and Social Affairs, Population Division (2015). World Population Prospects: The 2015 Revision, Key Findings and Advance Tables. Working Paper No. http://esa.un.org/unpd/wpp/publications/Files/Key_Findings_WPP_2015.pdf – Accessed May 2016.
2. Alexandratos, N. & Bruinsma, J. World agriculture towards 2030/2050: the 2012 revision. ESA Working paper No. 12–03. Rome, FAO, doi: 10.1016/S0264-8377(03)00047-4 (2012).
3. Adrio, J.-L. & Demain, A. L. Recombinant organisms for production of industrial products. Bioeng. Bugs 1, 116–131 (2010).Article Google Scholar 
4. Du, J., Shao, Z. & Zhao, H. Engineering microbial factories for synthesis of value-added products. J. Ind. Microbiol. Biotechnol. 38, 873–890 (2011).CAS Article Google Scholar 
5. Olempska-Beer, Z. S., Merker, R. I., Ditto, M. D. & DiNovi, M. J. Food-processing enzymes from recombinant microorganisms-a review. Regul. Toxicol. Pharmacol. 45, 144–158 (2006).CAS Article Google Scholar 
6. Tunick, M. H. Whey Protein Production and Utilization: A Brief History. Whey Process. Funct. Heal. benefits 1–13 (2008).
7. Orth, J. D., Thiele, I. & Palsson, B. Ø. What is flux balance analysis? Nat. Biotechnol. 28, 245–248 (2010).CAS Article Google Scholar 
8. Feist, A. M. et al. A genome-scale metabolic reconstruction for Escherichia coli K-12 MG1655 that accounts for 1260 ORFs and thermodynamic information. Mol. Syst. Biol. 3, 121 (2007).Article Google Scholar 
9. Kamau, S. M., Cheison, S. C., Chen, W. & Liu, X.-M. Alpha-Lactalbumin: Its Production Technologies and Bioactive Peptides. Compr. Rev. Food Sci. Food Saf. 9, 197–212 (2010).CAS Article Google Scholar 
10. Boland, M. J. et al. The future supply of animal-derived protein for human consumption. Trends Food Sci. Technol. 29, 62–73 (2013).CAS Article Google Scholar 
11. Huang, C.-J., Lin, H. & Yang, X. Industrial production of recombinant therapeutics in Escherichia coli and its recent advancements. J. Ind. Microbiol. Biotechnol. 39, 383–399 (2012).CAS Article Google Scholar 
12. Stouthamer, a. H. A theoretical study on the amount of ATP required for synthesis of microbial cell material. Antonie Van Leeuwenhoek 39, 545–565 (1973).CAS Article Google Scholar 
13. Russell, J. B. & Cook, G. M. Energetics of bacterial growth: balance of anabolic and catabolic reactions. Microbiol. Rev. 59, 48–62 (1995).CAS PubMed PubMed Central Google Scholar 
14. Heyland, J., Blank, L. M. & Schmid, A. Quantification of metabolic limitations during recombinant protein production in Escherichia coli. J. Biotechnol. 155, 178–184 (2011).CAS Article Google Scholar 
15. Villalobos, A., Ness, J. E., Gustafsson, C., Minshull, J. & Govindarajan, S. Gene Designer: a synthetic biology tool for constructing artificial DNA segments. BMC Bioinformatics 7, 285 (2006).Article Google Scholar 
16. Salis, H. M., Mirsky, E. a. & Voigt, C. a. Automated design of synthetic ribosome binding sites to control protein expression. Nat. Biotechnol. 27, 946–950 (2009).CAS Article Google Scholar 
17. Terpe, K. Overview of bacterial expression systems for heterologous protein production: from molecular and biochemical fundamentals to commercial systems. Appl. Microbiol. Biotechnol. 72, 211–222 (2006).CAS Article Google Scholar 
18. Lovmar, M. & Ehrenberg, M. Rate, accuracy and cost of ribosomes in bacterial cells. Biochimie 88, 951–961 (2006).CAS Article Google Scholar 
19. Goldberg, A. L. Protein degradation and protection against misfolded or damaged proteins. Nature 426, 895–899 (2003).CAS ADS Article Google Scholar 
20. Zhuang, K., Yang, L., Cluett, W. R. & Mahadevan, R. Dynamic strain scanning optimization: an efficient strain design strategy for balanced yield, titer, and productivity. DySScO strategy for strain design. BMC Biotechnol. 13, 8 (2013).CAS Article Google Scholar 
21. Nocon, J. et al. Model based engineering of Pichia pastoris central metabolism enhances recombinant protein production. Metab. Eng. 24, 129–138 (2014).CAS Article Google Scholar 
22. Sohn, S. B. et al. Genome-scale metabolic model of methylotrophic yeast Pichia pastoris and its use for in silico analysis of heterologous protein production. Biotechnol. J. 5, 705–715 (2010).CAS Article Google Scholar 
23. WorldBank. World Bank Commodities Price Data (The Pink Sheet)http://www.worldbank.org/en/research/commodity-markets.
24. Selvarasu, S. et al. Characterizing escherichia coli DH5? growth and metabolism in a complex medium using genome-scale flux analysis. Biotechnol. Bioeng. 102, 923–934 (2009).CAS Article Google Scholar 
25. Amthor, J. S. The McCree-de Wit-Penning de Vries-Thornley Respiration Paradigms: 30 Years Later. Ann Bot 86, 1–20 (2000).CAS Article Google Scholar 
26. Bremer, H. & Dennis, P. Modulation of chemical composition and other parameters of the cell by growth rate. In Escherichia coli and Salmonella typhimurium. Neidhardt, F. (ed.). Washington, DC Am. Soc. Microbiol. Press 1553 (1996).
27. Milo, R., Jorgensen, P., Moran, U., Weber, G. & Springer, M. BioNumbers The database of key numbers in molecular and cell biology. Nucleic Acids Res. 38, 750–753 (2009).Article Google Scholar 
28. Mahadevan, R., Edwards, J. S. & Doyle, F. J. Dynamic flux balance analysis of diauxic growth in Escherichia coli. Biophys. J. 83, 1331–1340 (2002).CAS ADS Article Google Scholar 
29. Varma, A., Palsson, B. O., Varma, A. & Palsson, B. O. Stoichiometric Flux Balance Models Quantitatively Predict Growth and Metabolic By-Product Secretion in Wild-Type Escherichia coli W3110. Appl. Environ. Microbiol. 60, 3724–3731 (1994).CAS PubMed PubMed Central Google Scholar