Cronos, through their partnership with Ginkgo Bioworks, has done the first of its kind; they created a new pathway to produce cannabinoids! By going through two recently published patent applications I was able to confirm they can theoretically make cannabinoids in yeast without infringing on intellectual property covering the biosynthetic pathway found in the cannabis plant. In this article I’m going to go over what they did, the system they’ve outlined and their next steps. The first patent can be found here, the second patent can be found here.

The first patent covers how they circumvented the patent I said would be the hardest to get around in the pathway to making cannabinoids. It will be useful to think of the proteins I’m talking about as machines in Ford’s assembly line, there are inputs and products to every step. In a similar fashion to how cars are made through a stepwise process of adding parts and making changes, proteins in cells build molecules like cannabinoids. Ginkgo’s solution to circumventing the most difficult step is simple: find a protein that does two steps in the pathway instead of finding a protein that does the identical chemical reaction to the patented protein. They screened thousands of proteins and did end up finding such a protein, instead of having two proteins doing two different steps, they have one protein doing two steps. However, unlike the patented protein from cannabis, the protein they found is not perfect, far from it. Instead of making only one product, the protein makes at least two products, olivetol and olivetolic acid. Olivetolic acid is made naturally in cannabis and is the native input to the protein in the next step of the pathway to make cannabinoids. Olivetol is an unnatural byproduct not found in the cannabis plant. As you’ll see later, this is problematic.

The second patent application filed by Ginkgo covers many potential proteins they discovered to circumvent another patent protecting the native pathway to producing cannabinoids in the cannabis plant. Similar to the protein I just outlined, many of the possible proteins outlined in this patent are also imperfect and create unnatural byproducts not found in the cannabis plant. I’ve included a diagram that outlines some of the potential products their system can generate, the pathway Cronos wants is highlighted in red while the undesirable byproducts are in black. 

The byproducts I’ve shown in the diagram are only a fraction of the potential byproducts outlined in the patent applications. It’s important to note that patents aren’t peer-reviewed papers, scientists can intentionally omit details crucial for the evaluation of information. I will highlight some of those missing pieces of information. Table 8 shows us the ‘best’ protein in the first patent produces two distinct products, olivetol and olivetolic acid in an approximate 1:3 ratio. We do not know if other products are also being made. It’s possible, considering that protein is capable of doing two very different reactions, that there are more byproducts in the mixture. The proteins outlined in the second patent are capable of producing five different products from a single input. They only quantify two of those products in the data supplied, those are the two products I’ve combined with the two products from the previous step to generate four potential products on the diagram above. As seen in the quote below from the second patent, the type of proteins conducting the second step accept a wide range of inputs, so it’s likely they’re converting both olivetol and olivetolic acid to a range of products. 

The patent applications don’t show them engineering the whole pathway in yeast to produce CBGA, but given they have all the steps they need, it’s not a big leap to assume their claim to have done so is accurate. The issue they’re working on now is trying to eliminate the byproducts. Cronos is licenced to sell cannabinoids found in the cannabis plant, any byproducts they’re making that look like cannabinoids but aren’t found in the plant are illegal to sell. Since the byproducts have similar physical and chemical properties to CBGA, they must go to great lengths to extract everything, then remove all the impurities from their extract to ensure they’re selling pure CBGA (or CBG when decarboxylated). Purification costs time, money and it sacrifices yield since no process is 100% efficient. This is what Cronos had to say in their annual report about going forward with the R&D produced through the Ginkgo deal.

The last sentence is particularly interesting, they believe there’s enough consumer demand to justify producing a non-psychoactive cannabinoid that isn’t well known, but not enough to justify producing THC or CBD. There are several potential medicinal applications for CBG, but Cronos can’t advertise any of those without going through the traditional pharmaceutical route and doing clinical trials.

In summary Ginkgo has created a system for Cronos capable of producing up to 9 molecules, based on the data provided in the patents. Only 1 of those molecules is a natural cannabinoid found in the cannabis plant (CBGA), the rest are byproducts that must be removed after extraction from yeast. Cronos currently has no plans to scale up production of any cannabinoids apart from CBGA, despite the only common demand for cannabinoids being THC and CBD. The next milestone to watch for is reducing the cost of producing CBGA to $1,000/KG, unlocking the first tranche of payments to Ginkgo (expected Q3 2021 according to Cronos).

The preceding is the opinion of the author, and is in no way intended to be a recommendation to buy or sell any security or derivative. The author holds no position in Cronos.