C. sanguinolentus and C. conco in the mass range m/z 480–5000 corresponding
to the molecular masses of most known conopeptides
(Conoserver; Kaas et al., 2010). Each signal was generally found
in only a few specimens (mean = 3.27) reflecting a high level of
intra-specific venom variation, and only 30 were found in more
than 10 specimens. Each specimen of C. lividus, C. sanguinolentus
and C. conco included a mean of 35.6, 31.5 and 37.19 different
The sequencing of these Conus venom glands libraries yielded
around 2 million reads that were then assembled into 100,000 contigs
on average. Transcriptome analyses confirmed the high diversity
of toxins found in the three species. In total, 88 different toxin
precursors were found in the six specimens analyzed, attributable
to 12 different superfamilies using the signal sequences (following
the classification in Puillandre et al., 2012), plus one set of 8
sequences (found in the three species) characterized by a signal
sequence different from the known superfamily of conotoxins
(Supplementary Data 3). Most precursor amino-acid sequences
are unique to a single species, but three are shared by the three
species, four by C. lividus and C. sanguinolentus and nine by C. conco
and C. lividus, in agreement with the fact that C. conco and C. lividus
are closely related and diverged from each other more recently
than from C. sanguinolentus. Correlating the MALDI-TOF-MS signals
and the transcriptome data to infer potential mature conotoxin
sequences was judged to be too hypothetical due to the elevated
number of matches resulting from unknown precursor cleavages
sites, undefined post-translational modifications and the lack of
any previously characterized toxins from these species.