Every US route-to-market analysis this publication has published for a precision-fermented ingredient has turned on the GRAS pathway: a notice, a no-questions letter, or a self-affirmed determination. For a colour, none of those exist. FDA states it in a single sentence, and it appears verbatim in its own recent letters to protein notifiers:
There is no GRAS provision for color additives.
That sentence sits in FDA’s response to All G’s recombinant bovine lactoferrin notice, signed 25 March 2026, and in its response to TurtleTree’s, signed 7 May 2025. In both cases FDA raised it because the ingredient is pink, and a coloured protein may itself trip the colour additive definition. But the underlying rule is categorical. A substance used to impart colour is a colour additive under section 201(t)(1) of the FD&C Act, and a colour additive can only be marketed if it is listed by regulation — a petition, a safety review, and a final order published in the Federal Register amending 21 CFR part 73 or part 74. There is no self- affirmation. There is no notification-and-no-objection. There is rulemaking, or there is nothing.
This is now the binding constraint on the most commercially urgent reformulation project in US food, and almost nobody is costing it correctly.
The record: what has actually happened to fermentation-derived colours
On 22 April 2025, HHS and FDA announced a series of measures to phase out petroleum-based synthetic dyes from the US food supply. Action three on the agency’s own list was “Working with industry to eliminate six remaining synthetic dyes — FD&C Green No. 3, FD&C Red No. 40, FD&C Yellow No. 5, FD&C Yellow No. 6, FD&C Blue No. 1, and FD&C Blue No. 2 — from the food supply by the end of next year.” The release is dated April 2025, so “next year” is 2026.
This is a request, not a ban. FDA Commissioner Marty Makary put it in those terms in the same release: “the FDA is asking food companies to substitute petrochemical dyes with natural ingredients.” No rule was issued and no compliance date attaches. It is nonetheless the deadline the entire natural-colour supply chain is now priced against.
Set against that deadline, here is the actual record for the two novel non-synthetic colours most relevant to fermentation.
| Galdieria extract blue | Beetroot red | |
|---|---|---|
| Petitioner | Fermentalg, Libourne, France | Phytolon Ltd |
| Petition | CAP 1C0320 | CAP 4C0326 |
| Docket | FDA-2021-C-0925 | FDA-2024-C-1085 |
| Filing announced | 9 September 2021 (86 FR 50495) | 12 March 2024 |
| Final order | 12 May 2025 | 6 February 2026 (91 FR 5295) |
| Regulation created | 21 CFR 73.167 | 21 CFR 73.39 |
| Filing to final order | 44 months | 23 months |
| Objection deadline | 11 June 2025 | 9 March 2026 |
| Effective date | 26 June 2025 | 23 March 2026 — never took effect |
| Status today | Effective; confirmed 21 August 2025 | Delayed indefinitely as of 20 March 2026 |
Phytolon makes its betalain pigments by precision fermentation, using engineered baker’s yeast — work developed in partnership with Ginkgo Bioworks. Beetroot red is, so far as the public record shows, the first colour produced by an engineered microorganism to reach a final FDA listing order.
It has never been lawful to use.
What stopped it, and why the mechanism matters more than the objection
The final order gave interested persons until 9 March 2026 to file objections and requests for a hearing. On 5 March, GMO/Toxin Free USA filed a submission. FDA’s notice of 24 March records what followed, in its own words:
We received a submission from GMO/Toxin Free USA containing objections that meet the conditions set forth in 21 CFR 12.22 to initiate a stay of the effective date of the final order.
And the operative date line:
As of March 20, 2026, the effective date of the final order published February 6, 2026 (91 FR 5295) is delayed indefinitely.
Three days before it would have taken effect. FDA was careful to say the delay “does not reflect a change in our determination that there is a reasonable certainty of no harm”, and does not constitute a finding that a hearing is justified. The agency has not conceded anything on the science. It had no choice about the delay.
That is the part operators need to understand. Under sections 701(e)(2) and 721(d) of the FD&C Act, any person adversely affected by a colour additive order may file objections within 30 days requesting a hearing, and the filing itself operates to delay the effective date until FDA takes final action on the objections. FDA’s own footnote to the notice is blunt: although the statute says “stay”, “this delay effectuates the same result.”
One organisation, one submission inside a 30-day window, and a listing that took 23 months to obtain is suspended for an indefinite period. FDA received other comments opposing the order, but noted none of them appeared to qualify as objections under 21 CFR 12.22. A single qualifying objection was sufficient.
As of the date of this article the docket carries three documents — the petition filing notice, the final order, and the delay — and no new effective date has been announced. That is more than five months of suspension.
Practical consequence: a colour additive listing is not final when the order publishes. It is final 30 days after the order publishes, and only if nobody objects. If you are building a reformulation timeline around a pending or recent colour listing, add the 30-day objection window as a hard gate, and treat the order date as provisional until the effective date is confirmed in a separate Federal Register document — as happened for galdieria extract blue on 21 August 2025, and has not happened for beetroot red.
The arithmetic that does not work
On 6 May 2026, Oterra — the world’s largest natural colour supplier — and San Diego biotech Debut announced a multi-million-dollar collaboration to scale Debut’s precision fermentation process into a natural alternative to Red 40, spanning orange, red and violet shades. The release states the companies will work with food and beverage manufacturers “while working toward an FDA approval filing, with the goal of bringing a commercially scaled product to market in approximately three years.”
Take those words precisely. As of May 2026 the companies were working toward a filing. They had not filed.
Now apply the only two data points that exist.
| Step | Galdieria benchmark | Beetroot red benchmark |
|---|---|---|
| Filing to final order | 44 months | 23 months |
| Order to usable | 45 days, uncontested | Indefinite, contested |
If Debut and Oterra file during 2027 — generous, given they had not filed by May 2026 — the beetroot red pace of 23 months puts a final order somewhere in 2029, and the galdieria pace of 44 months puts it in 2031. Either way the product is not lawfully usable in US food before 2029, and possibly not before 2031. The companies’ own “approximately three years” to a commercially scaled product lands around May 2029, which is consistent with the faster of the two benchmarks and only if the filing happens promptly and draws no objection.
The phase-out target for Red 40 is the end of 2026.
The gap is not a matter of effort or capital. It is structural, and it comes from the pathway. A precision-fermented protein can reach the US market through a GRAS notice in roughly a year — we counted twelve such clearances in the FDA’s inventory — or immediately, on a self-affirmed determination, subject to the mandatory-notification rule FDA has now proposed. A precision-fermented colour has neither option. It must be listed by rulemaking, and it can be stayed by a single letter.
Practical consequence: if you are a brand reformulating away from Red 40 for the end of 2026, biotech reds are not in the answer set for that deadline. They are a 2029-plus supply option. Your 2026 options are the currently listed colours — which is exactly why Oterra markets its existing Red 40 alternatives alongside the Debut partnership, and why its own sourcing team publicly describes a three-year cycle for agricultural pigments. Plan the agricultural route for the deadline and the fermentation route for the decade.
Why the objection route is a bigger risk for fermentation-derived colours than for others
There is an asymmetry here that the beetroot red docket makes visible and that deserves stating plainly.
The objector was GMO/Toxin Free USA. Beetroot red is produced by genetically engineered yeast. Galdieria extract blue, which cleared without a qualifying objection, is derived from a unicellular red alga, Galdieria sulphuraria, as FDA’s order describes it. Butterfly pea flower extract and calcium phosphate — both listed on 12 May 2025 — are a flower extract and a mineral.
A colour additive listing is a public rulemaking with a statutory objection right attached. That gives organised opposition a procedural entry point that the GRAS pathway simply does not provide: a GRAS notice is reviewed and closed by correspondence, with no comparable third-party veto window. So the very ingredients most likely to attract engineered-organism opposition are the ones routed through the process most exposed to it, while proteins made by the same organisms in the same fermenters go through a channel with no such window at all.
This is a coherent objection to make about the sector’s regulatory strategy, and it is one the Debut–Oterra partnership will have to plan around rather than hope past. The relevant preparation is not only a strong safety dossier. It is anticipating who will file in the 30-day window and what the dossier needs to contain to survive it.
The counter-argument
Two respectable ones.
First, the phase-out is voluntary and the deadline is soft. FDA asked; it did not ban. There is no enforcement date, no compliance obligation, and no penalty attached to end-2026. If the deadline slips — and voluntary industry deadlines usually do — then a 2029 biotech red is not late, it is simply the next generation of reformulation rather than the current one. State-level rules in California, Texas and Delaware are on their own timetables and may matter more than the federal request in practice.
That is fair, and it is the strongest defence of the Debut–Oterra timeline. But it concedes the point rather than rebutting it: the biotech red is not a solution to the phase-out as announced. It is a solution to whatever comes after.
Second, FDA said it would accelerate. The April 2025 announcement committed to “Authorizing four new natural color additives in the coming weeks, while also accelerating the review and approval of others”, and named the queue: “calcium phosphate, Galdieria extract blue, gardenia blue, butterfly pea flower extract, and other natural alternatives”. The beetroot red timeline of 23 months is roughly half the galdieria timeline of 44 months, consistent with that commitment working. If FDA sustains a 23-month pace, the arithmetic above is the pessimistic case.
Also fair, with one qualification worth noting: galdieria extract blue was itself on the fast-track list, so its 44 months already includes an accelerated finish. The unaccelerated baseline for a novel colour petition is therefore longer than 44 months, not shorter.
And acceleration operates on FDA’s review clock, not on the statutory objection window. It was the objection window — not the review — that stopped beetroot red. Speeding up the part that worked does not address the part that failed.
What we could not establish
- The current status of the beetroot red objections. FDA wrote: “We plan to address the objections in a future document.” We found no such document published as of 2 September 2026, and no new effective date. We cannot say whether a hearing will be granted, and neither, on the record, can FDA.
- The substance of the objections. We did not retrieve the GMO/Toxin Free USA submission itself from the docket, and characterise it only as FDA characterises it: a submission containing objections meeting the conditions of 21 CFR 12.22. We take no view on its merits.
- Spirulina. FDA’s expanded spirulina extract approval was reportedly placed on hold at the same time and for the same reason. We have relied on trade reporting for that and did not retrieve the corresponding Federal Register document, so we have kept it out of the table.
- Petition filing dates for butterfly pea flower extract and calcium phosphate. Both were finalised on 12 May 2025 and both are relevant context, but we could not confirm their petition-filing announcement dates from a primary Federal Register citation, so they are excluded from the duration table rather than estimated.
- Debut’s production organism and molecule. The companies describe “precision fermentation” and shades spanning orange, red and violet. Neither the organism, the pigment class, nor the intended part 73 or part 74 listing is stated publicly, and the difference matters: a colour requiring batch certification under part 74 is a different regulatory object again.
- Whether Fermentalg cultivates Galdieria by fermentation. FDA’s order describes the source as unicellular red algae. We did not verify the cultivation method, and have therefore not described galdieria extract blue as a fermentation-derived colour — only as the most recent novel non-synthetic colour to complete the pathway, which is what makes it a useful duration benchmark.
- Any deadline extension. We found no FDA document altering the end-of-2026 phase-out target.
What to watch
Whether FDA publishes a document on the beetroot red objections. This is the single most informative event available. If FDA denies the objections and sets a new effective date, the pathway works and the delay cost Phytolon roughly a year. If FDA grants a hearing, the pathway has a multi-year procedural tail, and every engineered-microbe colour behind it in the queue inherits that risk. Docket FDA-2024-C-1085 is where it will appear.
Whether Debut and Oterra announce a filing date. A petition filing is announced in the Federal Register — it is public, dated and checkable. Until that notice appears, the three-year commercial target has no clock running against it. If a filing notice publishes in 2027, apply the 23-to-44-month band and judge the target accordingly.
Whether anyone tries to route a colour through GRAS anyway. FDA has now twice, in letters that had nothing to do with colour petitions, warned protein notifiers that a pigmented ingredient may constitute a colour additive use and that its no-questions letter “is not an approval for use as a color additive”. That is the agency pre-empting a specific argument: that a fermentation-derived pigment sold as a functional ingredient might avoid the petition requirement. It will not, and FDA has now put that on the record twice. Anyone advised otherwise should read GRN 1284 and GRN 1219 before acting on the advice.