In one breath. Every exosome product makes a claim about what is inside. Almost none can tell you how many are in the bottle, or whether they are still there at month six. The lab trying to answer that question is not in Seoul or Los Angeles. It is in Gifu, and it is working with food.
Every section carries an evidence label: 観測 (observation), 研究知見 (research), 本人の言葉 (his words), 私の読み (my read). This is piece 2 of a series, "What the US is about to learn". Piece 1 was The last good year for the word exosome.
Where I met him · 観測 observation
At COSME Week Osaka at the start of October, the academic zone was a short row of posters between the ingredient booths. Prof. Hiroshi Takemori of Gifu University walked me through his one by one. His background is not cosmetics. He spent two decades on signalling enzymes and melanin biology, at Osaka University and then at a national biomedical institute, and came to extracellular vesicles only in the last few years, through a pair of fluorescent probes his lab designed to stain them. That order matters. He arrived at vesicles as a biochemist with a measuring tool in his hand, not as a marketer with a product.
What the lab actually does · 研究知見 research
Three things, and the third is the one I keep thinking about.
First, it finds vesicles, and vesicle-like things, in food. Lactic acid bacteria (Lactobacillus delbrueckii, L. helveticus, Lacticaseibacillus paracasei). Euglena. Ginger. A fermented botanical drink. Bovine milk, where the particles are true exosomes and where the whole programme began. Koji is a funded project in his lab but, as far as I can find, not yet a paper, so I will leave it there.
A note on the words, because the words are the whole point. Read his papers side by side and you notice something I did not expect: the lab does not use one term. It chooses the word by what the particle is. The Euglena particles are called extracellular vesicle-like nanoparticles, and the "like" is doing real work: the paper says plainly that they are "artificially created products", made by autoclaving the whole culture at 120 degrees for twenty minutes, which reconstitutes cell membranes into particles. The discovery was an accident with a flask that was on its way to disposal, and the authors say so. The Lacticaseibacillus paracasei particles, by contrast, are secreted by living bacteria, and that paper calls them extracellular vesicles, no hedge. Nowhere does either paper say "exosome" for its own particles. When I asked him, his answer was precise: "exosome" is fine on a product, not in a paper, where the word is extracellular vesicle. His lab's abbreviation EV-LNP expands, in the paper, to EV-like nanoparticle; in conversation he also uses LNP in the drug-delivery sense, lipid nanoparticle, and the Euglena paper itself once calls the heat-made particles "reconstituted lipid nanoparticles". If you are reading this in English, be careful: to a US or EU reviewer, LNP means the vaccine carrier and nothing else. I asked him which expansion he wanted in English and he chose the long form, extracellular vesicle-like nanoparticle, so that is the term in this piece, and I would advise any Japanese company writing an English dossier to do the same. The particle does not change. The reader does.
Second, it tests what they do, at the level the data supports, and it reports the negatives. The 2026 paper on L. delbrueckii particles looks at keratinocyte barrier function in a human cell line: the lipid fraction, especially hydroxylated fatty acids, restored tight-junction proteins under cold stress through a TRPV4-linked calcium pathway. That is the one result in this body of work that points at skin, and his own development slide frames it as a candidate for skincare and quasi-drugs. The Euglena paper shows the heat-made particles cutting LPS-induced IL-6 in mouse macrophages to about a tenth of the induced level, with nitric oxide and TNF-alpha barely moved; the colitis experiment in that paper used a water-soluble extract of the heat-treated cells, not the particles themselves, because there were not enough particles to feed to mice, and the authors say that too. The L. paracasei paper is not about inflammation at all: those vesicles showed no anti-inflammatory activity in the same macrophage assay and the paper says so, then goes on to show they make a laboratory gene-delivery method work better through a surface protein and a kinase pathway. Ginger particles lower IL-6 in macrophages; a fermented botanical drink reduces fat absorption in mice. Cells and mice. I have not found a human skin trial from this group, and neither, as far as I can tell, has he claimed one. That is the right shape for this stage of a field, and reporting the negative result in your own paper is the right shape for a scientist.
Third, it is building the ruler: standardisation. The Gifu research-seed page and the company's own slides say it plainly: no standard method exists for quantifying these particles across species, and that is a barrier to using them in products. So the group designed fluorescent compounds that bind the particles and allow simple counting, validated first against the usual protein markers in milk, and it is working on extraction and storage methods per food type so that the count means something. The same work showed that high-temperature sterilised milk contains no intact exosomes while low-temperature pasteurised milk does, which is the kind of unglamorous fact a quality standard is built from. The Euglena particles turned out to fluoresce red on their own. The 2025 paper proposes a chlorophyll derivative formed on heating as the cause, and when I asked, he told me the lab now uses that fluorescence as its probe.
In his words · 本人の言葉
"A unique chlorophyll derivative obtained by heating Euglena exhibits high affinity for EV-like nanoparticles as a whole and emits red fluorescence, making it a useful probe for quantifying nanoparticles. What I would like to emphasize is that we routinely consume this fluorescent compound in our daily diet, and its safety is therefore supported by our long history of dietary exposure."
That is his position, stated as his; the probe is a counting tool, not a skincare claim. A university spin-out, GIFU EXOSOME, runs contract measurement for companies, about forty jobs in 2023 according to its own deck, and it says out loud that measurement does not make money. It does it anyway, because without a count there is no standard, and without a standard every claim is a story.
Why this is the better half of the exosome story · 私の読み my read
In my last piece I wrote about the word: Korea took "exosome" off cosmetic advertising for human-derived material in January 2025, the EU bans human-origin ingredients outright, and a US lawsuit is testing the claims. Non-human particles pass that first gate. A yogurt bacterium is not a human cell. But passing the origin gate only means you are allowed to start. The claims have to be rebuilt from zero, from cells to animals to people, and the product has to prove that what it says is inside is actually inside, now and at month six.
That second problem, the counting problem, is the one nobody in the market talks about, and it is the one this lab chose. If a Japanese ingredient company ever puts a lactic-acid-bacteria particle into a serum and wants to sell it in Los Angeles, the first question a reviewer will ask is not "does it work". It is "how many, by what method, measured when". There are very few places on earth that can answer that for a food-derived particle today. One of them is a university lab in Gifu whose company has trademarked a phrase that translates as "delicious exosome". I find that combination of rigour and whimsy completely Japanese, and I mean it as a compliment.
What I would ask before anyone bottles it · 私の読み my read
- Which particle, from which strain or plant, secreted or made, at what count per millilitre, by what method, and at what age of the product.
- Which effect, in which model. A keratinocyte barrier result is a keratinocyte barrier result; it becomes a skin claim only after a vehicle-controlled human study.
- What the word on the box will be in each market. "Vesicle" survives everywhere. "Exosome" does not. "LNP" means something else entirely once you cross the Pacific.
Prof. Takemori's lab and its spin-out are looking for consumer-brand partners for two of these particles. If you make skincare in Japan and want a vesicle ingredient that can survive an overseas review, this is where I would start the conversation, and I would start it with the count.
Keep reading
The last good year for the word exosome, piece 1 of this series. For brands and ingredient makers: what I check before a product page goes overseas.
Amisha, cell biologist, Osaka. MEng, University of Tokyo. Education, not medical advice. Prof. Takemori reviewed the two paragraphs about his work before publication; the sentences quoted as his are his own words, in English, from that exchange.
Sources. Yamamoto C et al., World J Microbiol Biotechnol 2026;42:563 (L. delbrueckii EV-like nanoparticles, keratinocyte barrier, TRPV4), doi 10.1007/s11274-026-05272-7. Yamamoto C et al., Curr Microbiol 2026;83:201 (L. helveticus surface-layer protein on nanoparticles), doi 10.1007/s00284-026-04783-8. Sholihah IA, Batubara I, Barlian A, Takemori H, BioNanoScience 2026;16:396 (L. paracasei extracellular vesicles enhance PEI transfection; no anti-inflammatory activity in RAW264.7), doi 10.1007/s12668-026-02631-4. Sakurai A et al., BioNanoScience 2025;15:271 (Euglena EV-like nanoparticles created by autoclaving; IL-6 in RAW264.7; autofluorescence attributed to a chlorophyll derivative; DSS colitis with a water-soluble extract of heat-treated cells), doi 10.1007/s12668-025-01887-6. Chikama K et al., J Food Sci 2025;90:e70518 (fermented botanical product nanoparticles, fat absorption, mice), doi 10.1111/1750-3841.70518. Muhammad DR et al., Adv Pharm Bull 2025;15:60 (ginger exosome-like nanoparticles, IL-6, RAW264.7), doi 10.34172/apb.42675. Furukawa S et al., Biochem Biophys Res Commun 2024;696:149505 (measuring milk exosomes with fluorescent probes; heat-sterilised milk contains none), doi 10.1016/j.bbrc.2024.149505. Gifu University research seeds on the measurement and standardisation of food-derived exosomes and extracellular vesicles; GIFU EXOSOME company materials, 2024. Prof. Takemori, emails to the author, October 2026 (terminology; review of the two paragraphs; the quoted sentences supplied by him in English). MISEV2023, J Extracell Vesicles 2024;13(2):e12404. I review the scientific basis of English skincare claims for a living; nothing here refers to any client. Education, not medical advice.
