www.nytimes.com /2020/04/01/magazine/cannabis-science.html

Cannabis Scientists Are Chasing the Perfect High

Gary Greenberg
24-31 minutes

Credit...Illustration by Erik Carter

The Great ReadFeature

Chemists at some of the biggest legal-weed companies are after an elusive prize: a predictable, reliable product.

Credit...Illustration by Erik Carter

Listen to This Article

Audio Recording by Audm

To hear more audio stories from publishers, like The New York Times, download Audm for iPhone or Android.

The retail showroom of INSA, a farm-to-bong cannabis company in western Massachusetts, is a clean industrial space on the first floor of a four-story brick building in the old mill town Easthampton. When I visited recently, before the coronavirus shut down recreational sales and forbade crowds, the crew of eight behind the glass display cases looked a lot like the staff you’d see dispensing lattes at Starbucks or troubleshooting iPads at the Genius Bar: young, racially diverse, smiling. They were all wearing black T-shirts with the INSA motto, “Uncommon Cannabis.” Standing in line with me were a white-haired couple leaning on canes; a 40-something woman in a black pantsuit, who complained that the wait would be longer than her lunch break; a bald man in a tweed jacket; and a pair of women in perms and polyester discussing the virtues of a strain called Green Crack. We were all waiting at a discreet distance from the counter, as you would at the bank, for the next available “budtender.”

I got Ben, who described for me the wares that fill the cases like rings and watches in a jewelry store: waxes and dabs and oils and buds and edibles, most of them, he said, processed in a lab and kitchen on the other side of the wall behind him, using weed grown on the upper three floors. He sounded a little apologetic when he told me that while he knew why the bud I was pointing to was called Peyote Critical — “It speaks a little bit to its parentage, Peyote Purple and Critical Kush” — he hadn’t tried it, so he wasn’t entirely sure how it would affect me.

Ben took me around a corner to another glass case, this one displaying vaporizers in different shapes and sizes. He pulled a box off a shelf behind him. It was a $35, 350-milligram disposable vape pen loaded with Jack Herer, a strain named for a legendary grower. If I bought this, he said, I should “resist the temptation to take big rips — four seconds at the max, then pull that pen away and inhale to get a nice full set of lungs.” Ben felt more certain about the effects of Jack Herer than Peyote Critical, especially after he took a look at the label. “The primary terpene in here is limonene,” he said, which should make me “energetic and uplifted.” But there were more terpenes at work, Ben said. “You’ve got pinene coming in at 2.83 percent, good for memory retention and alertness, and then myrcene, which should help balance out some of the raciness from the limonene. Myrcene is good for your brain’s absorption of metabolizing THC but also has relaxing, sedating qualities.”

Terpenes are the compounds that give the different strains of cannabis their distinctive aromas. According to Ben, they are also what “modulates the high”; each variety of weed has its own terpene profile, which helps account for one of the riddles of cannabis: that even if two strains share the same psychoactive molecule — 9-delta tetrahydrocannabinol, or THC — one can make its users collapse in gales of laughter while the other produces paranoia and yet another seems (at least for a moment) to reveal the secrets of the universe. INSA is confident enough that it has figured out which kind of high each of its products will deliver that it offers a customer-satisfaction guarantee if the experience it advertises doesn’t match what the user gets.

The company, which was founded in 2013, was named for the two main varieties of cannabis: indica and sativa. Before budtenders and black-market dealers knew anything about terpenes, they told their customers that indica strains were sedating and sativa strains euphoric. But scientists have shown that while indica plants tend to be short and bushy and quicker to mature and sativas tall and spiky and slower, those differences do not correspond to any kind of consistent differences in the chemical profiles of the plants. “We’ve been trying to re-educate our consumers,” Peter Gallagher, chief executive of INSA, told me.

His firm isn’t the only one changing the way cannabis is marketed. Wherever weed is legal, companies are claiming that they have figured out how to produce a bespoke high. The promises are specific — one California company, MedMen, offers its customers “a surefire explosive orgasm” — and backed by scientific-sounding terminology like “terpene profile” and “cannabinoid breakdown.” Some of the research these companies cite to support what they are advertising has been published and peer-reviewed, but much of the recent work on the effects of cannabis has been conducted privately, and the companies are guarding their results as trade secrets. MedMen canceled my interview with its chief executive when it learned that I wanted to talk about the science behind its claims.

The cannabis business, then, has arrived at a critical moment. Now that pot has become something like a regular consumer product, customers are increasingly seeking the same “proven consistency” they expect from potato chips and soap. The financial stakes are clear: Despite lingering prohibitions in 17 states, legal cannabis is already an $8 billion industry in the United States. Domestic sales of alcohol, humankind’s other favorite intoxicant, topped $200 billion last year. But to make cannabis as popular as booze requires solving that original problem: It’s hard to imagine millions of people becoming new recreational users without being able to promise them that the product they’re spending money on — the average purchase at INSA is around $90 — will give them the effect they want.

Companies like MedMen and INSA may have decided that they’ve already cracked the code, but it remains to be seen whether that’s even possible with a plant as complex as cannabis. What those companies know for certain, however, is that the billion-dollar race to find out has already begun.


Credit...Illustration by Erik Carter

Cannabis has been consumed in one form or another for thousands of years, but it wasn’t until 1964 that a team led by the Israeli researcher Raphael Mechoulam identified THC as the molecule that got users high. By then, cannabis prohibition had been widespread for more than 25 years, creating formidable bureaucratic obstacles to researchers who wished to work with the plant. But Mechoulam kept at it, isolating another cannabinoid, cannabigerol (CBG), and mapping the structure of cannabidiol (CBD). All these chemicals, it turned out, had a role to play in the body’s response to cannabis. In 1998, Mechoulam coined the term the “entourage effect” to describe the complicated interplay between cannabinoids and the body’s own neurotransmitters in determining the drug’s effects.

While Mechoulam was still conducting his research, an American neuro­logist named Ethan Russo was zeroing in on terpenes as a major source of the variability in the effects of different strains of weed — or “chemovars,” as he prefers to call them. Not long ago, I went to visit him at his home on an island in Puget Sound, where he walked me through his past two decades of trying to conduct conventional research on this unconventional subject.

Russo told me that while he had been interested in botanical treatments since reading Euell Gibbons’s “Stalking the Healthful Herbs” while still in his teens, he knew little about cannabis as a medicine. That began to change when patients in his private practice in Montana began to report success with plant remedies, including cannabis, for chronic conditions like migraines. Intrigued by the results they reported, he began to study herbal medicine in earnest. Eventually, he wrote a textbook on the subject, which included a chapter on cannabis.

In 1996, while writing the book, Russo was introduced to aromatherapy with essential oils. “I realized how evocative they were,” he told me. He also knew that the same molecules that gave essential oils their punch — the terpenes — were present in cannabis, and that aficionados often said that “the nose knows,” meaning that a strain that smelled good to a user was likely to yield felicitous results. He began to suspect that the terpenes “were having a major modulatory effect on THC” and thus held a key to understanding the wildly variable effects of the drug.

He received Food and Drug Administration approval to run a clinical trial of cannabis as a treatment for migraines, but the National Institute on Drug Abuse, which must approve such research with illegal drugs, refused to sign off. The bureaucratic resistance “really stimulated my sense of adolescent rebellion,” he told me, and he decided to investigate terpenes’ role in the entourage effect on his own — legally, but using a method that, while once the backbone of medical research, had fallen into disrepute: self-experimentation.

He ordered terpenes from chemical-supply houses and, along with a few friends, began to blind-test them by transferring small amounts of them from coded bottles into a vaporizer designed to minimize odors and keeping track of their effects. In 2004, he went to Amsterdam, where he was able to obtain pure THC legally, pair it with different combinations of terpenes and record the effects they had on a group of volunteers.

Russo’s research was not without its problems. Scents could not be totally eliminated, the effects of THC couldn’t be successfully blinded and the prodigious daily cannabis intake of at least one participant made him a poor judge of the effects of individual THC/terpene combinations. Still, Russo found consistent correlations. THC alone, he found, lowered mood and distorted perception, and proved over all to be “really hard to function on.” He recalled one session in which, as it turned out, he had inhaled pure THC. “It was my turn to make dinner that night, and it was like: ‘Oh, God, I’m not sure I can do this. Where’s the knife? What do I need to do next?’ Everything was so hard.” But throw in pinene, the terpene that gives a pine woods its scent, and “all of a sudden that’s gone. You’re clear. You have no problem remembering anything.” Limonene, one source of citrus’s distinctive odor, also cured the THC blues, “making this unpleasant thing vibrant and alive and electric.” On the other hand, some terpenes just made things worse — like myrcene, an oil that smells a little like cloves and is present in high concentrations in hops, on which, Russo recalled, “I can’t function, I can’t think, I can’t move.”

In 2010, at a conference honoring Mechoulam, Russo presented a paper called “Taming THC,” which compiled more than 400 studies that strengthened the case for the role terpenes played in the variable effects of pot. It did not directly mention Russo’s D.I.Y. research, but a careful reader could find observations about the effects of specific combinations on memory, cognition and mood — that myrcene-heavy strains may produce “couchlock,” that pinene might be an “antidote” to the negative effects of THC — that were at least as indebted to Russo’s experiments in Amsterdam as to anything in the scientific literature. The paper was published the following year in the prestigious, widely read British Journal of Pharmacology.

Russo was not the only cannabis researcher studying terpenes, and “Taming THC” was not the first scientific article to speculate about their role in cannabis intoxication. It was also meant to be the starting point for more rigorous research into terpenes, not the final word on their effects. But the article, with its concise charts of correlations between terpenes and drug effects, came along at a crucial moment in the history of pot: By 2011, 15 states had approved medical marijuana, and Colorado and Washington were on the verge of making the drug legal for recreational use.

A new industry was ready to burst into being, and here, in the legitimate academic press, was a paper providing a map to what Russo called a “pharmacological treasure trove.” If the paper’s promises held up, a company could even take aim at the most tempting prize of all: the vast number of Americans who had never tried weed before, and others who had aged out of it but might be brought back on board. For that market, it wasn’t enough for cannabis to be legal; the drug had to be as predictable as a pre-dinner martini.

“Taming THC” laid out an ambitious scientific agenda for anyone seeking to further test the paper’s claims: “high throughput pharmacological screening,” animal experiments to specify mechanisms of action, molecular studies to establish just how terpenes and cannabinoids interact, animal-behavior studies, brain-imaging research and human clinical trials. Nearly a decade later, this agenda, which is modeled on pharmaceutical drug development, remains unfulfilled.

Recently, however, a few companies in the United States and Canada have begun an aggressive investigation into the entourage effect, though they are forgoing many protocols of the pharmaceutical industry. Last year, I met Jon Cooper, the founder of a company called Ebbu, at a co-working space in Denver. Cooper had been toying with the idea of a cannabis start-up ever since Colorado legalized the drug, but he was deterred by his own history with pot. “I’d had some awesome experiences that I wished I could have all the time,” he told me. But he’d also had “some completely horrific experiences that I never ever wanted again.” Cooper says he couldn’t sell something he didn’t believe in; but what if he could figure out how to “capture in a bottle the awesome experience, so every store I walk into, I could get that same experience. Wouldn’t that be amazing?”

A year after Cooper started Ebbu in 2013, he approached Brian Reid, who was running a lab at the University of Colorado’s school of pharmacy, hoping they could collaborate. Reid’s specialty was exactly the “high throughput” screening Russo had called for, in which algorithms are used to quickly determine which potential drugs would interact with which cellular targets. The university’s lawyers, worried about a possible loss of federal funding, nixed the deal, but Reid eventually decided to go to work for Ebbu directly; in 2016, he became its chief science officer.

Reid began to, in Cooper’s words, “crank data.” As long as he did not ask for government money, he could do high-grade pharmaceutical research using human subjects without the usual regulatory scrutiny. Ebbu went straight into human trials of the most likely drivers of the entourage effect. That’s not as reckless as it might appear. Reid points out that no one is known to have ever died from an overdose of cannabis.

Colorado law forbids cannabis companies to give away products, so Ebbu offered samples for $1 to people who agreed to fill out an online questionnaire about their experiences. Their responses were correlated with the chemical profiles of the extracts in order to gather evidence about which combinations produced which effects. In June 2016, the company announced that it had identified eight terpenes and three cannabinoids that modify the effect of THC in a predictable way. The company said this was a “significant milestone” in its quest to develop formulations that would “enable consumers to choose a desired experience.”

In late 2016, Ebbu started distributing a formulation called Genesis to dispensaries around Colorado. It yielded a “really happy, focused high,” Cooper told me, one “that was extremely blissful.” It did not cause anxiety or make its users stupid or sleepy or goofy: “They could always communicate, they could talk to their kids, they were able to do all these things and function while still having this amazing experience.” Cooper loved it. His wife and his brother-in-law and his friends loved it. The consumers who bought Genesis loved it so much, according to Cooper, that it gained a “cult following” — people who “stopped consuming all other products and started consuming only Genesis.”

Still, Cooper kept funding more research. “Every single dollar that was coming in the door,” he said, “I was investing in science and intellectual property.” His approach seemed vindicated in the fall of 2018, when Canopy Growth, the largest cannabis conglomerate in Canada, bought Ebbu for a reported $330 million in cash and stock. According to Bruce Linton, the founder and former chief executive of Canopy, it wasn’t Genesis or any other product that made Ebbu such an attractive buy. It was the data the company had cranked out about breeding, extracting, formulating and using cannabis, about what consumers wanted and about which combinations of chemicals could be counted upon to give it to them. The value, according to Linton, was obvious: Whether customers want to go to a comedy club, or listen to music, or just “diminish the anxiety of the workweek,” they’re all “buying outcomes,” which Canopy, using the scientific knowledge it bought from Ebbu, could “containerize.” (Linton left Canopy last year as part of a shake-up at Constellation Brands, the multinational corporation with a controlling stake in Canopy.)

Russo’s dream of open scientific exchange, however, has remained elusive. The biggest companies in the cannabis world are keeping their research under wraps, and without the necessity of answering to regulators, it’s likely to stay that way. Ebbu did release some findings from its consumer research at the 2018 meeting of the International Cannabis Research Society, where it presented a poster called “Cannabinoid and Terpene Formulations Elicit Distinct Mood Effects.” The poster shows correlations between different mixtures and consumer experience, using terms like “active” and “chill” or “conflicted” and “Zen.” But the poster, which like all posters was not peer reviewed, did not explain how those categories were defined or which terpenes or cannabinoids were in the products, nor did it disclose the data underlying the findings. Several participants thought the poster was contrary to the spirit of the conference, which was about open scientific inquiry. To them, it looked like nothing more than an advertisement.

As more of the compounds in cannabis are isolated, a few companies are looking at ways to eliminate one stubborn source of variability: the plants themselves. Ebbu’s intellectual property includes a patent for using an inkjet printer to spit out cannabinoids and terpenes in precisely measured ratios determined by the user. Brought in from the black-market wilderness by deep-pocketed, consumer-savvy companies, cannabis may become just another designer drug.

At INSA, the Jack Herer vape oil may be named after a known strain, but it is not made by extracting or distilling a Jack Herer plant. Rather, it’s formulated in INSA’s lab to emulate the chemical profile of that variety. The company can obtain its THC and other cannabinoids from any cannabis plant, and it buys its terpenes from outside suppliers. Peter Gallagher says INSA does not hide the fact that its vape oils are manufactured products that, like pharmaceutical drugs, are created by isolating and combining compounds. Indeed, he envisions an exciting future when “you could come into the store and build your own blend of certain proportions of cannabinoids and terpenes.”

Recent research has shown that it’s possible to grow cannabinoids from yeast, cutting out the need for any horticulture at all — a prospect that has already attracted industry attention. After all, greenhouses take up more space than laboratories, molecules are easier to patent than plants and once you figure out how to do it all in a petri dish, you don’t have to worry about weather or insects. Ethan Russo, however, thinks producers should be cautious in taking this approach. “The idea that you’re going to bottle this up and eliminate cannabis” is a bad one, he told me. He doesn’t doubt that a few of the more than 500 chemicals in the plant can be identified as critical to its effects, but, he says, “that doesn’t tell the whole story” any more than the flute and violin lines alone can convey the entire impact of a symphony. What’s missing is the way the entourage works together not only to create the effects of the plant but also to provide a counterpoint to its potential dangers. “It’s vastly preferable to take the effort, time and money to develop a specific chemovar of cannabis that’s going to do the same thing and do it better and demonstrably more safely,” he says.

At least one researcher is making that effort, with the help of some willing human volunteers. For the past three years, a neuroscientist named Adie Rae has designed the survey used to judge the winners of the Cultivation Classic, an annual competition sponsored by an alt-weekly in Portland, Ore. Cannabis competitions are common, but the Cultivation Classic may be the only one that requires its judges to spend an afternoon listening to a scientist talk about predictive algorithms and blinded studies.

Last year, I joined Rae as she addressed a crowd of 160 people in a conference room in downtown Portland. The participants had been handed black zipper bags that contained a dozen tiny glass jars, each labeled with a number and housing a single bud of locally sourced, organically grown weed. Their mission, Rae explained, was to take a 48-hour “tolerance break” and then, over the course of the next month, sample each flower, paying careful attention to their psychological state before, during and afterward. A four-digit PIN enclosed in their kits would give them access to a website on which they were to rate the extent to which the sample gave them the experience they wanted, whether it made them sleepy or stimulated, sociable or introspective, cognitively impaired or creative, if it gave them side effects like redeye or anxiety, and if its aroma and appearance and taste were to their liking.

Rae makes information on the chemistry of the winning plants and the effects that users reported available on the Cultivation Classic website. “We wouldn’t take these data and hand them over to Monsanto or some other corporate juggernaut,” she says. “We want the folks who have social-justice components in their workplaces, who are mindful of the resources they use.”

Her most recent findings might disappoint any cannabis company, large or small. After she crunched the Cultivation Classic numbers, Rae could not find strong correlations between any single terpene and the high that resulted. “If we just look at the individual terpenes and try to correlate them with any of our measures about the experience, we have a total scatterplot,” Rae told me. “It’s like a shotgun. The points are everywhere.” She could reach only one conclusion: “There is just no association with any singular terpene for any question we have.”

Rae was neither surprised nor disheartened. After all, she explained, the entourage effect relies on the interactions among hundreds of chemicals, and the attempt to parse it is still in its infancy. “This is a fishing expedition,” Rae says. “Are there any meaningful conclusions to be drawn at all? Is it all about what a person expects from their experience? Is it all about their own endogenous cannabinoid system? Is there any pattern whatsoever besides that THC is intoxicating? We don’t really know.”

Not that she is about to abandon her research: A deeper analysis of the interactions of terpenes and cannabinoids, which she plans to perform, may yet yield correlations. Rae likens it to music, suggesting that “terpenes might be the timbre of the experience, while THC is the volume.” She’s also aware that the biggest confounding factor in the attempt to parse the entourage is the person who is hosting it. Someone who smokes a bowl at the end of a long day of physical labor may well have a different experience if she smokes the same pot with friends at a party or before she does yoga. Someone who buys weed with a particular expectation may well have an outcome shaped by that expectation. And perhaps most important, the particulars of an individual’s neurochemistry can change the way a plant’s chemicals affect the brain.

The best hope for someone seeking a predictable high may come directly from users, who can inform one another about the virtues of particular varieties — just as they did in the old days, only this time with help from big data. Rae foresees an app that can tell a person what cultivars are liked by people who liked what she liked, and for which purposes. It wouldn’t promise quite the “surefire” experience marketed by some of the bigger cannabis companies, but it would be reasonably reliable, as far as psychoactive substances go.

Of course, that would complicate the prospects of an industry that is going all in on the idea that the entourage effect, whatever its constituents and dynamics, can be unpacked and put to use in the market. That doesn’t bother Rae. In the end, she says, there’s not much point to taking drugs if the outcome is written on the label and if the drug is not taken mindfully. “We have to understand that there is always going to be some level of exploration,” she told me. The ongoing elusiveness of the entourage effect “is what’s exciting, because it’s not necessarily telling us more about the plant; it’s telling us more about ourselves.”

Rae’s approach to cannabis is similar to that envisioned by Russo when he put down his copy of “Stalking the Healthful Herbs” and started his study of terpenes. But the future of the drug probably doesn’t belong to them — it belongs to companies like MedMen and Canopy and maybe, one day, even Monsanto. Consumer capitalism is endlessly resourceful at transforming almost any human desire into a standardized product on the shelf. The smart money has already placed its bets, and the humble cannabis plant can put up a fight for only so long.

Gary Greenberg is the author, most recently, of “The Book of Woe: The D.S.M. and the Unmaking of Psychiatry” and a contributing editor for Harper’s Magazine.

A version of this article appears in print on , Page 44 of the Sunday Magazine with the headline: Decoding Cannabis . Order Reprints | Today’s Paper | Subscribe