Indeed, many observations have accumulated, which are supportive of a struc-
tural multiple-loci model coupled with a single-gene inheritance mode; these
multiple loci should be tightly linked, possibly arranged in tandem, to account for
the virtual absence of segregation of the main chemotypes observed by all authors.
This new view suggests that duplication events followed by sequence divergence
might have generated the cannabinoid synthase gene family, as frequently observed
in other plants especially for secondary metabolism genes (Ober 2005 ).
Support for the multi-locus model has been recently provided by the construc-
tion of aC. sativalinkage map, where a strong statistical association between a
QTL for the chemotype (defined as CBDA/THCA ratio) and the location of the two
cannabinoid synthases was reported (Weiblen et al. 2015 ).
CBDAS sequences in drug-type plants were detected after completion of thefirst
genome and transcriptome sequencing of the THCA-predominantCannabisstrain
Purple Kush, which generated thefirstC. sativasequence database available online,
“The Cannabis genome browser” (http://genome.ccbr.utoronto.ca/). Van Bakel
et al. ( 2011 ) identified three different CBDAS pseudogenes; a premature-stop
CBDAS transcript was also detected in the same plant, most likely deriving from
one of the pseudogenes. Reads corresponding to THCAS were also discovered in
fiber-type cultivar Finola, though these were attributed to the presence of pseudo-
genic copies due to inability to assemble them into functional protein-coding genes.
Weiblen et al. ( 2015 ) reported nine unique THCAS and CBDAS sequences in
two highly inbredfiber- and drug-strains and in the F1 plants from their cross. All
F1 plants showed simultaneous presence of all four homologues, providing further
evidence of a heterozygosity state in separate loci. F2 plants, however, only
reproduced parental chemotypes (with one single exception), suggesting therefore
that these multiple loci may be tightly linked.
The same authors also proposed the use of the CBDAS gene as a“genotyping
tool”, predictor of the“drug”nature of a plant: homozygous plants for a functional
CBDAS, even when producing intermediate levels of the two main cannabinoids,
are not considered able to produce amounts of THCA that exceed the limit value of
0.2%. They suggested, therefore, a higher efficiency of CBGA conversion by
CBDAS compared to THCAS, as observed by hybrid plants characterized by a
CBDA/THCA ratio that was always shifted towards CBDA, in their experiments.
The observation of a perfect association between drug phenotype and the non-
functional CBDAS homologue, and the detection of a single recombinant indi-
vidual with a clear drug phenotype that was homozygous for the hemp-type
THCAS and homozygous for the marijuana-type nonfunctional CBDAS, suggested
that the absence of a functional CBDAS was essential for the complete drug
phenotype expression (Weiblen et al. 2015 ). However, fully inherited CBDA/
THCA ratio values skewed towards THCA production rather than CBDA, have
been reported by other authors in F1 and F2 individuals from different parentals (de
Meijer et al. 2003 ) (Fig.15.4); therefore, the deviation from the unity, caused by a
putative higher efficiency of CBDAS compared to THCAS, proposed by Weiblen
et al. ( 2015 ), does not seem to be a general feature of allCannabisgermplasm and
cannot be considered reliably diagnostic.
15 Genomics and Molecular Markers inCannabis sativaL. 335