An Israeli startup has emerged from stealth with patent-pending technology it claims could slash production costs for allulose, a sweetener many formulators regard as the best all-round sugar replacer, but which currently comes with a hefty price tag.
Founded in 2020 by biochemist Dr. Ziv Zwighaft, Tel Aviv-based Ambrosia Bio has raised seed funding from investors including an undisclosed beverage company. This enabled it to build IP, produce samples, and sign its first commercial agreements.
It is now looking to raise $8-10 million in a series A round to scale up the technology at a sugar production facility in Central Europe, and is now working with “one of the largest international food and beverage producers as well as a global sugar manufacturer.”
According to Zwighaft: “We are orders of magnitude more efficient than the industry standard, so we strongly believe we can reach cost parity with conventional sugars. Allulose currently sells at a few dollars per kilo, sometimes 10s of dollars. Once commoditized, my belief is that the end price [for allulose produced using Ambrosia’s approach] will be anywhere from $1.5-2.5/kg.”
The appeal of allulose
A low-glycemic, low-calorie, non-cariogenic sweetener, allulose is particularly attractive to formulators seeking to replicate the sensory and functional properties of sugar without the calories.
The rare sugar—which has 70% of the sweetness of sucrose but only 0.4 calories per gram (vs 4cals/g for sucrose)—has a negligible effect on blood sugar and insulin, and has garnered significant interest following the FDA’s 2019 decision to exclude it from the total and added sugars declarations on the Nutrition Facts panel.
As it has the texture and bulk of regular sugar, allulose can be used to reduce or replace it in everything from beverages, yogurt and ice cream to baked products and candies, and now features in brands from Chobani Zero Sugar and Magic Spoon to Nick’s ice cream.
It also browns during baking (unlike erythritol), depresses the freezing point when making frozen products, and disperses well in batters and dough without the need for additional water.
The production challenge
The problem, says Zwighaft, is that it’s still too expensive for mass market adoption because it’s so costly to manufacture.
While allulose is found naturally in plants such as figs and raisins, it is produced on a commercial scale by firms such as Tate & Lyle, Ingredion/Matsutani, and CJ CheilJedang via an enzymatic conversion process typically starting with sugar or corn starch.
Fructose from these starting materials is then converted into allulose with a naturally occurring enzyme called epimerase, which is both unstable and inefficient, claims Zwighaft. “It’s highly unstable and its shelf-life can be measured in days. By contrast, our proprietary version is extremely stable and can perform for months without a drop in activity in high temperatures.”
Given the instability of natural epimerase enzymes, he claimed, many companies add them to their fructose solution, wait for it to reach a state of equilibrium (a syrup with a 70:30 ratio of fructose to allulose) over about 20-24 hours, and then inactivate them.
‘We can use the same enzymes for many months’
Ambrosia, by contrast, packs its enzymes into a bed of resin to immobilize them. The fructose solution passes through the bed of enzymes, which convert fructose into allulose “in minutes.” Given their superior stability, Ambrosia can “use the same enzymes for many months.”
While the syrup emerging from Ambrosia’s process has the same 70:30 ratio of fructose to allulose, the quantity of enzymes required to produce a fixed amount of allulose is significantly lower than the industry standard, slashing production costs, claims Zwighaft.
The syrup then goes through additional steps including chromatography to separate allulose from fructose and other sugars, steps Zwighaft claims Ambrosia can also perform more efficiently than key players in the industry.
He is now working with synthetic biology company Ginkgo Bioworks to develop a microbial strain that can produce his proprietary enzymes at scale.
The business model
According to Zwighaft: “We’re basically offering [sugar producers] an end-to-end solution. So we would provide the technology that would enable a producer of a low-margin commodity ingredient such as sugar to make a higher-margin ingredient such as allulose.”
Separately, Ambrosia is also exploring adding its acid-stable enzymes to fruit juice to convert some of the fructose in the juice to allulose, reducing the drinks’ glycemic index and calorie content.
Bonumose: Alternative approach
Bonumose, a Virginia-based startup working on more efficient production processes for rare sugars such as allulose and tagatose, told AFN that it has enzymes that can convert maltodextrin to allulose with a high yield.
CEO Ed Rogers explained: “Bonumose has focused on tagatose to date, but our patented process can make allulose, too. Our allulose process is not the fructose-to-allulose process used by other companies. Instead, Bonumose converts starch (not fructose) to >80% yields of allulose using a continuous production process and a combination of immobilized enzymes that catalyze an irreversible enzymatic conversion.
“So our conversion yield is ~3X higher than fructose-to-allulose conversions and our higher purity allulose can more easily be purified to 100% pure allulose.”
He added: “Furthermore, starch is cheaper than fructose. Fructose typically is derived from (a) converting starch to glucose; (b) enzymatically converting the glucose to 42% fructose; and (c) purifying the fructose. We’ve been doing this for tagatose at commercial scale since December 2022 and will roll out allulose when we think the timing is right.”
Zwighaft at Ambrosia acknowledged that Bonumose’s conversion yield is higher than the standard approach, which yields 30:70 allulose to fructose, but he claimed that Ambrosia outperformed in other respects including enzyme stability, salt consumption, and ion exchange capacity and noted that Bonumose’s process requires multiple enzymes working in parallel.