Extinct Notes
Unearthing the Smell of Lost Flowers
Every perfumer knows that when you change one material in a formula, even by a fraction of a percent, the resulting smell can change completely. Fragrance is fragile that way. It’s a very delicate balance. There’s a fascinating project from the field of synthetic biology where the DNA from an extinct flower was taking in order to reconstruct what it might have smelled like. I’ve known about this project for a while, before I’d studied perfumery formally, but coming back to it now, with training under my belt, I can’t figure out how sure anyone could possibly be about this hypothetical smell. If a formula based on a flower sitting in front of me is this amoebous, what happens when you’re working from a two hundred year old scrap?
This project, Project Cretaceous, named after the period when flowers first evolved, started in 2014, in a conversation between Jason Kelly, CEO of Ginkgo Bioworks, and a consultant for Givaudan about the Givaudan Scent Trek program, which sends people into rainforests to capture the air around rare living flowers. They do this so the smells can then be reverse engineered. Kelly was really inspired by the idea and wondered about the flowers people could not travel to anymore, because they no longer exist.
Ginkgo’s creative director, Christina Agapakis, took the idea and ran with it. Working off the International Union for Conservation of Nature (IUCN) list of plant extinctions, her team found twenty extinct species in the collection, and fourteen with enough intact material to attempt to analyze it. She negotiated access to Harvard University’s Herbaria, an archive of more than five million preserved specimens, so she could use a very small amount of tissue from each of the candidates.
Note: Many people were involved in this exploration, so I have strove to name as many as I can. That might get tedious for you, reader, but we recognize the minds involved, don’t we?
What the Yeast Kept
Josh Kapp and Beth Shapiro , paleogenomicists, pulled what fragments of DNA they could out of the fourteen samples. From there, computational biologist, Jue Wang, had to reassemble genes from the scattered fragments of fifty or so letters apiece. He then used equivalent genes in living relative species as a template for whatever was missing. He built 2,738 different reconstructed versions across the fourteen plants.
Each version went into a yeast strain, before molecular microbiologist, Scott Marr, ran the fermented results through a mass spectrometer to see if any scented molecules emerged. Most of the fourteen species yielded genes that didn’t produce anything a mass spectrometer would register as a scent molecule. Only three did. Hibiscadelphus wilderianus, a hibiscus relative that grew on Maui, in lava fields, until cattle ranching wiped it out around 1912. Orbexilum stipulatum, a legume last seen on a Kentucky river island in 1881 before a dam drowned its only habitat. And Leucadendron grandiflorum, a South African shrub last recorded in a London garden in 1805.
Those three allowed the team to explore “the smell of an extinct flower,” but only because the hypothetical genetic codes of those three functioned in yeast in the right way. Agapakis has always been transparent about this project. “We will not bring back the plant,” she said, “but we hope that by experiencing the scent of the flower that went extinct, people will get a greater appreciation for plants that are lost.” Not cloning, not germination, not living tissue, nothing close to the de-extinction work being attempted on animals. This is just to produce a molecule list, assembled from fragments of a genome, with the reference codes of still-living plants, alluding to something that did used to exist, but never an exact reconstruction.
Art, Science, and Product
Alexandra Daisy Ginsberg, an artist who built an installation around these three molecule lists, has said the results “aren’t accurate because we don’t know which molecules were actually in the flower, in what quantities, whether the molecules’ function was smell-related, or even if the genes were switched on to produce those molecules.” She pointed out that hibiscus flowers barely smell at all, because they’re pollinated by birds, not insects, and birds don’t hunt by scent. Hibiscadelphus, the flower behind the whole project’s most celebrated result, likely then, would not have had much of a smell to reconstruct in the first place.
At this point, I wondered how such an elaborate exploration of curiosity was even possible and reminded myself that none of this started as science for its own sake. IFF backed the work early, and Ginkgo’s business is selling exactly this kind of synthetic biology capability to companies like it. The extinct flower idea was, first, a way to show potential clients what the platform could do. What was essentially a sales demonstration turned into something deeper with the inclusion of artists. Ginsberg and smell researcher Sissel Tolaas built the molecule lists into an installation called Resurrecting the Sublime. Tolaas split each reconstructed smell into four separate fragments, diffused from the ceiling so they mix differently with every visitor and with every breath. There’s no single fixed scent of the room, everyone experiences it differently in different areas of the space at different times, so the work really leans into the interpretation of the extinct material even at the level of the audience reception.
In 2023, Jasmina Aganovic, who studied chemical and biomedical engineering at MIT and later worked out of Ginkgo as an entrepreneur in residence, turned the same underlying research into Future Society, a commercial fragrance line built on those three flowers and a few more (six perfumes in total). One of the three perfumers who worked on the line, Olivia Jan, has described building “Grassland Opera” from the Orbexilum molecule with no real brief at all. One synthesized compound and, as she put it, total freedom for creativity. A rare gift in this industry for sure.
I haven’t smelled the entire extinct flower line from Future Society myself so I can’t tell you firsthand whether the bottled version of the extinct flowers reads as an answered question. I’d guess it feels settled given the format, and the brand’s now very strong reputation, but even this is a composition of interpretations. I love tracing how each mind and form held certain uncertainty in so many different ways. Ginsberg’s own phrase for the outcome was “a blurry picture of the past, a false yet powerful memory.” I find that very beautifully put and makes me consider the way that science can be held with softness rather than absolute rigor and truth through the arts. Don’t get me wrong. Both approaches to science have have value, but this side isn’t as easy for me to see. This side isn’t as peer reviewed. Perhaps this soft science is most often exploited commercially, but this project has shown me how even through commodity, it can be held with respect, and curiosity, and creativity. This is just further proof to me that art is a form of experimentation itself. It makes me wonder what other questions aren’t even being asked in the hard sciences because their answers are not attainable. I wholeheartedly believe that questions should always be asked even if never answered.
References
Jacobsen, Rowan. “Ghost Flowers.” Scientific American, February 2019.
Brouillette, Monique. “Would You Feel Sexy Wearing Eau de Extinction?” MIT Technology Review, December 5, 2016.
Melton, Lisa. “Wake Up and Smell the Extinct Leucadendrons.” Interview with Alexandra Daisy Ginsberg. Nature Biotechnology, September 2021.
Agapakis, Christina. “Reviving the Smell of Extinct Plants.” Ginkgo Bioworks blog, May 3, 2019.
“About.” Resurrecting the Sublime.
Parrillo, Angelina. “A Boston Beauty Brand Used ‘Resurrected’ Extinct Flowers to Create Perfume.” The Boston Globe, September 9, 2024.
Lewis, Ricki. “Perfume from Extinct Flowers, Thanks to Ancient DNA and Synthetic Biology.” PLOS DNA Science, December 14, 2023.