Meet the founder: Brightseed’s Dr. Jim Flatt illuminates the ‘dark matter of nutrition’

[Disclosure: AgFunderNews’ parent company AgFunder is an investor in Brightseed.]

Launched by Sofia Elizondo, Dr Lee Chae, and Dr Jim Flatt in 2017, Brightseed​​ harnesses AI and machine learning to illuminate the ‘dark matter’ of nutrition lying untapped in plant and fungal kingdoms.

The San Francisco-based startup—which is working with partners ranging from major CPG companies and ingredients suppliers to pharmaceutical companies—has built a vast database of plant-based compounds associated with biological targets. Its AI-powered Forager platform then predicts how previously untapped plant components might confer specific health benefits and helps identify the optimal way to source or produce them at commercial scale.

Once Brightseed has identified fruitful targets, more traditional analytical methods and clinical testing can then be used to validate its computational work.

Its first product—sourced from upcycled hemp hulls—contains N-trans-caffeoyl tyramine (NCT) and N-trans-feruloyl tyramine (NFT), two bioactives shown in pre-clinical data to maintain gut barrier integrity, clear fat from the liver, and support weight management.

AgFunderNews (AFN) caught up with cofounder and CEO Dr. Flatt (JF) to discuss the genesis of Brightseed and lessons learned after 30 years of working in biomanufacturing (“the problem with a lot of companies in the algae space was that they went after low value, high-volume, capital-intensive applications and the technology wasn’t ready”).

AFN: You spent several years at Martek Biosciences producing omega-3s from microalgae. What did that teach you about the economics of biomanufacturing?

JF: Martek Biosciences (acquired by DSM for $1.1bn in 2011) was ahead of its time, producing a high-value ingredient [DHA omega-3] from a renewable source with clinically proven health benefits for brain and eye development. So I really gained an appreciation for how important it is to create a simple but powerful story for consumers, in this case, ‘Give your baby the best start in life’ [Martek’s DHA was targeted at infant formula].

But from a commercial perspective, the big challenge in algae, and in biotech more generally, is selecting the right targets. We were making a product with a strong health benefit and we could price it accordingly. The problem with a lot of companies in the [microalgae] space was that they went after low value, high-volume, capital-intensive applications [such as biofuels and proteins] and the technology was not ready.

If you look at what [microalgae specialist] Solazyme [later TerraVia, which was acquired by Corbion in 2017] did, it pivoted from biofuels to higher-value food products [proteins and high-oleic oils], but they weren’t really that high-value or particularly distinguished. There were some modest advantages to the products but not enough to really change consumer behavior.

AFN: After Martek, you led a team at Mascoma engineering yeast and bacteria to convert cellulose to ethanol. What did you take away from that experience?

JF: In this case, we were trying to break down natural plant material for sugars to make ethanol and other biofuels, but the lesson there was we had developed a good, cost-effective approach if you didn’t consider the investment you needed to make in the production plant itself!

And that was a huge wake up call for everyone. We all saw the promise for renewable sustainable fuels, but there’s literally trillions of dollars of infrastructure that has to be built out and that was a bridge too far for Mascoma.

But we did develop a contingency plan and maybe that’s my second lesson! We had designed a catalyst that could be useful for corn ethanol production, which actually went into the market with pretty good success and probably saved the company.

The lesson for me was in essence that you can’t swing for the fences too early if the technology is not ready, so at Brightseed from the outset we’ve tried to approach developing our products and processes with a clear-eyed view around the economics.

AFN: You headed up R&D at Hampton Creek (now Eat Just) between 2015-2017. How did that help you forge the idea for Brightseed?

JF: During that time I had the incredible luck to meet [Dr.] Lee [Chae] who became our CSO/CTO, and Sofia [Elizondo], who became our COO, and we worked quite closely on building a protein discovery platform, which planted the seed for Brightseed.

While plants can be great sources for proteins, they have so much more to offer. For a long period of our history, they have been the only way to promote health. Look at Traditional Chinese Medicine, Ayurvedic medicine in India, or medicines used by indigenous people all over the world; they are all based on the healing properties of plants.

AFN: So what was the basic premise behind Brightseed?

JF: Lee led a major program at Stanford’s Carnegie Institute for plant biology to develop omics data around the metabolites and the genomes of many plants but also pioneered some of the first artificial intelligence and machine learning methods that we have built on today.

So I’d say Lee really had the germ of the idea for Brightseed. Could we build a very powerful AI system to much more rapidly discern health beneficial compounds in plants?

Martek was a great success, but it took a very long time to get there, and it was very costly. The promise of [Brightseed’s core technology platform] Forager, was that it could greatly accelerate product discovery and translation to a commercial product and reduce those costs.

AFN: Unlike many startups seven years into operations, all three cofounders are still running the business…

JF: I think we have very complementary skills, and we’ve often been told by investors that it’s unusual to have three cofounders that are still enjoying working together after all this time! It’s not all that common.

 AFN: In simple terms, how does your tech work?

JF:  We’re focused on natural, new-to-the-world compounds that can help restore gut health, metabolic health, and cognitive health and function.

Through Forager, once we have identified these compounds of interest, we can then identify the best ways to produce them [on a commercial scale], whether it’s from a natural plant source, a microbial source [eg. algae, for example] or by understanding the metabolic pathways that plants use to produce these products and then using those pathways as a blueprint for producing that compound via precision fermentation [engineering a microbe to express the target compound].

So through Forager, not only are we providing a new to the world high value product, we also have the information and knowledge to identify the best way to produce it.

Forager becomes not only the way to make amazing discoveries but also to translate those discoveries to cost-effective, scalable ways to produce them. And that’s how I think we have learned the lessons from algal biotechnology and biotechnology more generally.

AFN: Could any of these compounds of interest be produced via plant cell culture?

JF:  We’re agnostic about how to produce these products. We’ve had discussions with a couple of companies in the plant cell culture space because there are some specific molecules that are very complex where it would not be possible to produce them via microbial fermentation for example.

So I think plant cell culture has its place but one challenge is that plant cells grow very slowly, and as such, they have some of the same challenges and limitations that we see in mammalian cell culture. So these cells might double once a day, whereas a fast-growing yeast might double 10, 12 times or more a day.

 AFN: Explain how your AI works…

JF: Forager generates new data around the natural product world; helps us understand the value of those compounds in restoring and promoting health; and then finally helps us identify the best way to produce these compounds at the target cost.

We have three learning objectives for the artificial intelligence system.

First, what’s in the plant or microbe? So for that we use AI-enabled untargeted, deep metabolomics methodology. So as an example, we did some early work in soy [via a partnership with Danone], where about 150 compounds had been reported on. When we applied Forager’s methodologies, we found over 1,400 compounds. We also did a case study with Ocean Spray where we quadrupled the knowledge of cranberries and their health potential.

The second question we ask is what role do these compounds play in health? So we look at biological targets such as the enzyme predominantly responsible for producing cholesterol in the body, which is the target for statin drugs, and use artificial intelligence to train on very large data sets that could give us insights to the keys that unlock these biological locks that are the determinants of health.

The third question is what are the best sources [of the compounds of interest] in nature?

AFN: What datasets are you interrogating?

JF: We generate new to the world data with Forager, so we’re not just using AI to mine data about compounds that are known, and that’s something that I think often gets lost.

Science knows a little more than 100,000 compounds from plants, which have yielded things like aspirin and metformin [a drug for type two diabetes]. We’ve now identified over 7 million distinct compounds from just under 4,000 different plant species that make the bulk of our food chain, as well as several thousand medicinal plants.

AFN: What health areas are you focusing on?

JF: We are very commercially focused, so it starts with deep insights from consumers, who all want to live a longer, healthier life without therapeutic drugs. We’re looking at their unmet health needs in areas such as gut health and immunity; metabolic health including weight management and glucose management; and then finally, cognitive health and function, including sleep, stress and anxiety management.

We also get insights from our partners from Danone, ADM and Pharmavite to Kallyope, a promising pharmaceutical company developing solutions for weight management and diabetes management, which is a hot area with all the interest in drugs such as Ozempic and Wegovy.

 AFN: Tell me about your first product…

JF: When we started Brightseed, we knew we needed to focus on a major unmet health need and frankly, something that pharma hadn’t been able to do. So we looked at a biological target called HNF4α, which is like a symphonic conductor in the body that’s there to help you stay in metabolic homeostasis.

Our primary researcher had found a few synthetic compounds that had modest effects on HNF4α, so we used Forager to see if we could generate new leads and within three months we discovered two potent compounds NCT and NFT [N-trans-caffeoyl tyramine and N-trans-feruloyl tyramine], which Forager then further ascertained were produced by 80 different plants, including black pepper and hemp hulls.

Our first product is a minimally processed whole food ingredient that’s derived from hemp hulls, a waste product, so here’s a product that can promote gut strength and gut health by improving gut barrier function.

[Editor’s note: The intestinal barrier is a semi-permeable layer of cells that allows the uptake of essential nutrients and immune sensing but keeps out pathogens and unwanted bacteria.]

 AFN: What clinical data do you have to show that your ingredient can improve gut barrier integrity?

JF: We have several published pre-clinical studies, and we have three clinical studies underway to strengthen the product-specific evidence around the gut fiber product itself, and these will all be reporting out this year.

But there’s already a body of literature around the effects of these compounds, so that gave us a good degree of confidence in terms of translating these discoveries to the clinical setting.

The exciting thing is that these compounds address symptoms that aren’t solved by gut health products currently on the market.

We’ve just finished one of those studies in the microbiome space that we think is going to support new claims around this product as a prebiotic.

AFN: Given your initial interest in these compounds was around metabolic health more generally, aren’t there other claims you could potentially make beyond gut health?

JF: Absolutely. And that is a source of clinical studies we’re about to get underway this year. This regulator [HNF4α] is present in the gut, but it’s also present in the liver, pancreas and kidneys.

So what we’ve found is that NCT and NFT are also able to clear harmful fat from the liver, and as you know, there is currently no treatment for fatty liver disease, so that’s very exciting.

We’ve also found a significant weight management benefit, so we’re specifically looking at clinical studies focused on how these compounds can promote fat loss but also help to keep weight off.

And then finally, we’re also looking at its effects on glucose management.

AFN: What’s the manufacturing setup for this first ingredient?

JF: Sourcing from plants [hemp hulls] is one method. But Forager also gives us the blueprint of how to produce NCT and NFT by precision fermentation. I can’t say any more now but suffice to say we have a very low-cost way to produce this product.

 AFN: Is it on the market yet? And what’s the regulatory pathway?

JF: We have [self-affirmed] GRAS status in the US and we will be notifying the FDA this year. We also have the ability to market the product in Europe, although we do not currently. We have sales activity in Canada, but it’s not yet on the market.

Our business model is to sell our functional ingredients to food and beverage companies and medical foods, especially nutrition companies.

 AFN: What about your partnership model?

JF: One way we work with our 20+ partners is through funded programs where we create a custom product for our partners and they have exclusive rights to sell and market it. In those cases, mostly there’s the option for the partner to have someone else produce the ingredient or for Brightseed to produce it.

We’ve already earned several milestone payments based on success of one of our earliest partners, so we’re generating revenue already.

With our partners, plus our in-house work, within a couple of years, you’re going to see one or two truly new to the world bioactive introductions that deal with major health needs and conditions coming out from Brightseed each year. And that’s a level of productivity that’s unheard of in this industry.