On the genomic DNA level, the only early marker-based studies were focused
on the identification of internal transcribed spacers II (ITS II) of the nuclear ribo-
somal genes, as markers that are useful to the univocal discrimination ofCannabis
from other species. These studies either identified variant Restriction Fragment
Length Polymorphism (RFLP) fragments (Siniscalco Gigliano 1998 ), or developed
PCR-based assays (Siniscalco Gigliano et al. 1997 ).
However, all the above described studies focused on the search for sequences
with a minimum amount of intra- and maximum inter-specific variation (and hence
of information). In this search for constantCannabis-specific sequences, as opposed
to the interest and search for variability, provides the main difference between
forensic and breeding approaches to the development of molecular markers.
Before the advent of the next generation sequencing era, analysis of the
Cannabisgenome and mapping useful traits relied on the use of multi-locus
markers like Random Amplified Polymorphic DNA (RAPD), Amplification
Fragment Length Polymorphism (AFLP) or microsatellite, already widely exploited
by geneticists and breeders for other crops.
During the exploration of theCannabisgenome by multi-locus marker analysis,
once again forensic scientists made early ground-breaking work. Despite their low
reproducibility, RAPD markers have successfully been used to identifyC. sativa
samples, when High Performance Liquid Chromatography analysis failed (Gillan
et al. 1995 ), and to separate 51C. sativaandHumulus lupulussamples, based on
their origin (Jagadish et al. 1996 ).
Nevertheless, due to the huge variability present in mostCannabisaccessions,
the use of molecular markers for varietyfingerprinting has been relatively limited;
whereas, some multi-locus markers like RAPDs have been exploited to investigate
the degree of genetic variability, the relatedness betweenCannabispopulations or
accessions, and the effects of selection on the genome structure. Only later did
microsatellites (otherwise known as Single Sequence Repeats SSRs) become an
active field of study for both forensic scientists and geneticists working on
Cannabis.
Early RAPD-marker analyses on 54 samples belonging to 12 different
cultivars/accessions showed that, among the 205 amplification products detected, a
very high degree of polymorphism (98%) was present. The markers used were able
to group the different accessions/cultivars according to known common ancestors,
and to their geographical origin both by Principal Component Analysis (PCA) and
by Unweighted Pair Group Method with Arithmetic Mean (UPGMA) clustering
(Faeti et al. 1996 ).
In a later study using RAPD markers (Forapani et al. 2001 ) on sixCannabis
varieties with different expected degrees of genetic variability, overall 97.1% of the
102 loci identified were found polymorphic. The proportion of identified
inter-cultivar variation ranged from 12.8 to 76.8%; the latter was observed between
two highly selected divergent cultivars. The variance component analysis by
Analysis of Molecular Variance (AMOVA) revealed that the proportion of variance
caused by differences between male and female groups within dioecious cultivars
was not significant, whilst most of the observed variations (51.2%) were explained
15 Genomics and Molecular Markers inCannabis sativaL. 321