SummaryCannabis sativa is an economically important source of durable fibers, nutritious seeds, and psychoactive drugs but few economic plants are so poorly understood genetically.Marijuana and hemp were crossed to evaluate competing models of cannabinoid inheritance and to explain the predominance of tetrahydrocannabinolic acid (THCA) in marijuana compared with cannabidiolic acid (CBDA) in hemp. Individuals in the resulting F 2 population were assessed for differential expression of cannabinoid synthase genes and were used in linkage mapping. Genetic markers associated with divergent cannabinoid phenotypes were identified.Although phenotypic segregation and a major quantitative trait locus (QTL) for the THCA/ CBDA ratio were consistent with a simple model of codominant alleles at a single locus, the diversity of THCA and CBDA synthase sequences observed in the mapping population, the position of enzyme coding loci on the map, and patterns of expression suggest multiple linked loci. Phylogenetic analysis further suggests a history of duplication and divergence affecting drug content.Marijuana is distinguished from hemp by a nonfunctional CBDA synthase that appears to have been positively selected to enhance psychoactivity. An unlinked QTL for cannabinoid quantity may also have played a role in the recent escalation of drug potency.

No abstract

No abstract

Segregation for leaf trichome density was observed in a cold-hardy hybrid grape population GE1025 (N = ∼125, MN1264 × MN1246) that was previously used to detect a quantitative trait locus (QTL) underlying foliar phylloxera resistance on chromosome 14. Our hypothesis was that high trichome density was associated with resistance to phylloxera. Existing literature found trichome density QTL on chromosomes 1 and 15 using a hybrid grape population of “Horizon” × Illinois 547-1 and suggested a few candidate genes. To validate the reported QTL and our hypothesis, interval mapping was conducted in GE1025 with previous genotyping-by-sequencing (GBS) single nucleotide polymorphism (SNP) genotype data and phenotypic scores collected using a 0–6 trichome density scale at several leaf positions. Evaluations were done on replicated forced dormant cuttings in 2 years and on field-grown leaves in 1 year. There was no strong relationship between trichome density and phylloxera resistance except for a Pearson’s correlation (r) of about -0.2 between a few trichome density traits and phylloxera severity traits at 2 and 3 weeks after infestation. Two genetic regions were repeatedly detected for multiple trichome density traits: from 10 to 20.7 Mbp (∼10 Mbp) on chromosome 1 for ribbon and simple density traits and from 2.4 to 8.9 Mbp on chromosome 10 for ribbon density traits, explaining 12.1–48.2 and 12.6–27.5% of phenotypic variation, respectively. To fine map, we genotyped a larger population, GE1783 (N = ∼1,023, MN1264 × MN1246), with conserved rhAmpSeq haplotype markers across multiple Vitis species and phenotyped 233 selected potential recombinants. Evaluations were conducted on field-grown leaves in a single year. The QTL for ribbon trichome density on adaxial vein and adaxial leaf and simple density on abaxial vein was fine mapped to 12.63–13.38 Mbp (747 kb) on chromosome 1. We found variations of MN1264 and MN1246 at candidate genes NAC transcription factor 29, EF-hand protein, and MYB140 in this region and three other surrounding candidate genes proposed previously. Even though no strong relationship between foliar phylloxera resistance and trichome density was found, this study validated and fine mapped a major QTL for trichome density using a cold-hardy hybrid grape population and shed light on a few candidate genes that have implications for different breeding programs.