Alexis St-Gelais, chimiste – Popularization & Plant profiles
The intensive breeding of cannabis produces strains that have different molecular signatures, and this is quite apparent within terpenes profile. These features can be useful to characterize a strain or extract, by putting forward what sets it apart from another variety. Based on our extensive experience with cannabis terpenes, let us have a look at some interesting molecular trends in the plant.
Chemotypes and Cannabis
The production of molecules within the plant obeys a metabolic logic. A crude parallel can be made with the color of eyes in humans: depending on the genetic characteristics of an individual, our body will have the ability to produce (or not) pigments that will in turn determine the iris’ shade. The latter (although with multiple subtleties) can then be classified into a limited number of categories, like blue eyes. When studying plants, the concept of chemotype can be used to designate this phenomenon where molecules are expressed in some individuals and not (or less) in others. Polatoglu suggested the following definition for this concept: organisms categorized under same species […] having differences in quantity and quality of their component(s) in their whole chemical fingerprint that is related to genetic or genetic expression differences [1].
Within cannabis, cannabinoids tend to follow a chemotypical pattern, where one or two dominant cannabinoids are found but can vary from one strain to another. In The Handbook of Cannabis, de Meijer proposes a model with three genetic and one morphological turning points that can lead a given strain to express one of nine possible chemotypes (or even more, considering that one can have a mixed chemotype where two molecules are co-dominant, most typically THCA and CBDA) [2]. The model is summarized in figure 1 below, where the term “locus” refers to a zone in a chromosome of the cannabis plant where the genes encountered will influence the metabolic expression of cannabinoids.
Figure 1. Genetic model proposed by de Meijer [2] (figure adapted by PhytoChemia) to explain cannabinoids chemotypes. A first locus controls the expression of enzymes which are necessary for the synthesis of metabolic precursors of the phenolic part of cannabinoids – if this locus is inactive, the plant will be short of raw metabolic material and no cannabinoids will be produced whatsoever. A second locus controls the length of the carbon chain attached to this phenolic backbone, defining a proportion between C3 and C5 phenolic acids (divarinolic acid vs olivetolic acid, the latter being the precursor of the familiar THCA and CBDA). A third locus works in a similar fashion to the A/B/AB/O blood types in humans (this parallel is ours, not de Meijer’s). If the locus is inactive (akin to type O), the metabolism stops at CBGA (or CBGVA); if the genetic information for type A is present, then THCA will be produced, whereas type B will lead to CBDA (and both types can be present, as with blood type AB, for a mixed chemotype). A fourth parameter can come into play and has to do with the morphology of the trichomes, which can in some case lead to the conversion of CBGA to CBCA and give more chemotypes.
As far as we are aware, a comprehensive model for cannabis terpenes has yet to be proposed (let us know if you know of one!). You can nevertheless see from the example of cannabinoids that genetic traits in a strain can define whether or not a given molecule will accumulate as an outcome of the plant’s metabolism – or in other terms, if the strain will pertain to a given chemotype.
Correlated Terpenes
Very often, a given metabolic transformation will be driven by an enzyme, i.e., a protein that is able to facilitate a specific chemical transformation. Enzymes can be more or less selective in what type of molecules they can transform. When several molecules are similar in structure, they can sometimes all be transformed by the same enzyme, although not necessarily all with the same efficiency or speed. This implies that if the plant expresses a given enzyme thanks to its genetic traits, not only one, but several molecules can sometimes arise. In other cases, if one transformation of a key molecule is permitted upstream, several other transformations downstream become possible since the “raw material” has been made available. In any case, even without knowledge of the exact genetic and metabolic mechanisms at play, one can therefore examine if there are correlations between molecules, which would imply that they have a common origin. If that is the case, they should be considered together, because it is unlikely that only one of them will be found.
There are several such cases in cannabis terpenes. Here are quantitatively important groups to consider:
- • Pinenes correlate, with varying proportions of α- and β- isomers;
- • When limonene is abundant, a series of oxygenated monoterpenes tends to also increase in content;
- • A large proportion of terpinolene will be accompanied by the presence of several other monoterpenes;
- • β-caryophyllene and α-humulene are strongly correlated;
- • A group of eudesmane-type (or selinane) sesquiterpenes are closely tied. Those include α- and β-selinenes, a few selinadiene isomers and juniper camphor. They also correlate with spirovetiva-1(10),7(11)-diene and eremophila-1(10),7(11)-diene;
- • α- and δ-guaienes co-occur;
- • Germacrene B is always associated with γ-elemene by GC, since the latter is a thermal degradation product of germacrene B and therefore partly generated during analysis. (E)-α-bisabolene and α-bisabolol tend to somewhat correlate with germacrene B;
- • Finally, a cluster of sesquiterpenols including guaiol, eudesmols, bulnesol and cryptomeridiol are clearly tied together in terms of abundance.
Behavior of Terpenes Groups Across Strains
As we have come to test thousands of samples, some trends have become apparent within or between the groups outlined above. Here are some recurring phenomenons we observe in our tests. Keep in mind that with intensive breeding, one can still stumble upon something unusual: these are trends, not absolute rules!
Monoterpenes
The profile of terpenes is most of the time dominated by one or several monoterpenes amongst the following: myrcene, α-pinene*, limonene*, terpinolene*, (E)-β-ocimene and linalool – the latter two very seldom being the dominant compound. Remember that those marked by an asterisk come with peers. From the perspective of chemical trends, the cases of terpinolene and limonene are particularly interesting.
In the case of terpinolene, its presence seems to be a metabolic key for the expression of several other molecules. Whenever terpinolene is a dominant terpene, a diverse group of molecules that are usually found at best as traces become more salient. Some of them are represented in figure 2.
Figure 2. Compounds associated with terpinolene in cannabis. These molecules tend to be more abundant whenever a strain is rich in terpinolene, while being trace constituents or even entirely missing otherwise. In addition to those molecules, an unknown oxygenated monoterpene is also closely tied to terpinolene. It is eluted near terpinen-4-ol on a DB-5 column.
As for limonene, its concentration is correlated with that of camphene and several oxygenated monoterpenes including α-terpineol, endo-fenchol, and borneol, as well as pinene hydrates (figure 3). The latter are in fact relatively rare in the field of essential oils, with cannabis being one of the rare botanicals where they have some abundance alongside the rather uncommon African wild sage (or leleshwa), Tarchonanthus camphoratus.
Figure 3. Structures of molecules closely correlated to limonene abundance in cannabis.
Sesquiterpenes
β-Caryophyllene and α-Humulene
These two sesquiterpenes (figure 4) can sometimes surpass monoterpenes in terms of concentrations in a strain. They are always expressed unless a strain or extract is almost devoid of terpenes. This is not so surprising, because these terpenes are amongst the most widely distributed in nature – relatively few essential oil bearing plants do not exhibit them. It will be extremely rare to see any cannabis where the sum of caryophyllene and humulene does not account for at least 1% (relative percentage) of total terpenes, and their distribution is relatively continuous across all their possible concentrations. As such, they do not really represent a chemotype, rather a continuum.
Figure 4. Structures of β-caryophyllene and α-humulene
Germacrene B
As mentioned earlier, germacrene B is not thermally stable (figure 5). Whenever it shows on a GC profile, it will inevitably be accompanied by γ-elemene, into which it partially rearranges within the heated injection port of the instrument [3]. These usually will not be featured in terpenes screens in most laboratories, because the germacrene B standard is hard to obtain – but it can nevertheless be a quantitatively important constituent of terpenes in some strains. We have seen the sum of germacrene B and γ-elemene reach well over 10 mg/g in some cases! And there are instances where germacrene B is almost missing entirely, with almost all possibilities in-between.
There is some degree of correlation between germacrene B and the pair of closely related compounds (E)-α-bisabolene and α-bisabolol, although in that case we sometimes observe examples where they are decoupled. The α-bisabolene/bisabolol pair exhibits some chemotypical behavior, where they are most of the time expressed in strains, but once in a while inhibited to very low levels.
Figure 5. Structures of germacrene B and fully or partially correlated compounds.
Guaienes
α-Guaiene and δ-guaiene (figure 6) are typically found in rose and patchouli essential oils, among others. In many cannabis strains, these sesquiterpenes will be rather faintly expressed, but in some cases, their expression is triggered to account for a few relative points of percentage of total terpenes. In all honesty, this is difficult to track from α-guaiene only, because it tends to coelute with another quantitatively abundant sesquiterpene of cannabis, trans-α-bergamotene, on many GC columns (DB-5 and DB-Wax included). δ-Guaiene is therefore the good cue to look at for this chemotype. There are exceptions, but together the guaienes will typically either account for under 0.5 mg/g of terpenes or be found in the 1-3 mg/g bracket, which would suggest that there is a genetic trait that either allows or inhibits their production. α-Guaiene can oxidize over time into a potent odorant compound, rotundone [4] – it is probably too faint to be monitored directly in cannabis but could contribute to the aroma of some strains.
Figure 6. Structures of guaienes.
Eudesmanes (Selinanes)
This is one group of sesquiterpenes (figure 7) that you do not want to miss if you want an accurate account of the terpenes content of a strain. Selinadienes are, in many strains, amongst the most abundant terpenes overall, sometimes contributing well over 10 mg/g in total. As far as we are aware, cannabis is also the botanical where these molecules are the most prominent. The fact that our screen takes them into account whereas many laboratories disregard them goes a long way to explain the difference in “total terpenes” reported – and keep in mind the concept of total terpenes requires precautions.
Figure 7. Structures of the main eudesmane-type sesquiterpenes found in cannabis, and the correlated spirovetiva-1(10),7(11)-diene and eremophila-1(10),7(11)-diene. The group also includes selina-4,7(11)-diene.
There appears to be chemotypes with regards to eudesmanes, too. In a few strains, they are almost absent, implying that the absence of a given gene inhibits their metabolism.
Bulnesol/guaiol/eudesmols Sesquiterpenols
One last interesting group comprises several molecules that are closely correlated (figure 8), including bulnesol, guaiol and several eudesmol isomers. Except for cryptomeridiol, they are all featured roughly in the same amounts, and they follow a presence/absence chemotypical pattern. Some strains clearly express the group, whereas the molecules are found in traces only in other cases. In our experience, this is one of the most variable metabolic traits between strains, along with the dominant monoterpenes.
Figure 8. Compounds correlated in a cluster of sesquiterpenols found in cannabis, with none of them clearly dominating the rest.
Bottom Line
The proportions between terpenes can be useful tools to describe strains. Our full terpenes service comes with a short conclusion that will highlight a few trends regarding the groups discussed above, and we keep thinking of good ways to capture the terpenes chemotypes in cannabis to better convey this information to our customers in the future.
References
[1] Polatoglu, K. “Chemotypes”– A Fact That Should Not Be Ignored in Natural Product Studies. Nat. Prod. J. 2013, 3 (1), 10–14. https://doi.org/10.2174/2210315511303010004.
[2] de Meijer, E. The Chemical Phenotypes (Chemotypes) of Cannabis. In The Handbook of Cannabis; Pertwee, R. G., Ed.; Oxford University Press: Oxford, 2014; pp 89–110.
[3] Venditti, A. What Is and What Should Never Be: Artifacts, Improbable Phytochemicals, Contaminants and Natural Products. Nat. Prod. Res. 2020, 34 (7), 1014–1031. https://doi.org/10.1080/14786419.2018.1543674.
[4] Huang, A.-C.; Burrett, S.; Sefton, M. A.; Taylor, D. K. Production of the Pepper Aroma Compound, (−)-Rotundone, by Aerial Oxidation of α-Guaiene. J. Agric. Food Chem. 2014, 62 (44), 10809–10815. https://doi.org/10.1021/jf504693e.