3.1. P. cf. Vaubani Displayed a High Diversity of Nonneuropeptide Venom Components
While all the VG transcriptomes appeared largely similar in terms of putative conotoxin and turripeptide diversity, Pvau2 showed a definitively higher number of other venom components. Because prior conid literature has focused on neuropeptides, few works discussed the presence and diversity of other molecular types of venom components. Therefore, it is difficult to determine if the high percentages of non-neuropeptide sequences observed in P. vaubani can be considered ubiquitous among conids or peculiar to this species.
P. cf. vaubani venom appeared to be composed, in large part, of putative toxins potentially involved in haemostasis impairing, affecting prey muscular and immune system, and a smaller proportion of enzymes responsible for correct toxin folding or facilitating their spreading into prey tissues. For example, RBLs and chymotrypsins, which impact the coagulation cascade, were found exclusively in P. cf. vaubani specimens and enriched in Pvau2 VG, suggesting potential use for impairing prey haemostasis. These two compounds are both found in other animal venoms, and were overexpressed in P. cf. vaubani VG, which suggests they may have an important role in the envenomation process. Of note is the P. cf. vaubani chymotrypsin, which is similar to chymotrypsins active at low pH, and could represent an adaptation to a specific hunting strategy or prey type. Another interesting protein overexpressed in the VG of P. cf. vaubani is PH-4. P. cf. vaubani PH-4 showed a different structure and mature peptide sequence from those found in other conids and in honeybees, and possessed two types of mature peptides, one of which in multiple copies. In honeybees, PH-4 was linked to foraging-related behaviour, however the lack of information concerning Profundiconus ecology and hunting strategy hampers the evaluation of PH-4, RBLs, and chymotrypsins and their potential effects on P. cf. vaubani prey.
3.2. Limited Conotoxin Diversity May Indicate a Narrow Worm or Molluscan Diet
Analysis of our transcriptome assemblies recovered different numbers of putative conotoxin and profunditoxin precursors: 75 for P.
and 55 for P. neocaledonicus
. If the three specimens are considered independently, the number of putative conotoxin and profunditoxin precursors are: 37 (Pvau1
), 53 (Pvau2
), and 55 (Pneo
). If only the conid species sequenced in conditions similar to the ones used in the present work (one to three VG samples, sequenced by Illumina HiSeq2000 platform) are taken into account, P. neocaledonicus
emerges as the species showing the lowest number of conotoxin and contoxin-like precursors. P.
showed instead a higher number of conotoxin precursors, equal to the one in Conus (Gastridium) geographus
Linnaeus, 1758 [52
] and C. (Virroconus) ebraeus
], from which, however, only one specimen was sequenced. Therefore, if we take into consideration only the result of one specimen for each Profundiconus
, and the most numerous Pvau2
with 53 conotoxins), P. neocaledonicus
may be included among those conids with a less diversified neuropeptide arsenal.
Conversely, when the numbers of gene superfamilies produced in the same sequencing condition are considered, P.
and P. neocaledonicus
show a high number of superfamilies (24 and 21, respectively) comparable to species with higher numbers of conotoxins, like Conus (Virgiconus) virgo
Linnaeus, 1758 with 25 gene superfamilies and 113 conotoxins [53
] and C. (Gastridium) geographus
Linnaeus, 1758 with 21 gene superfamilies and 75 conotoxins [52
]. The two Profundiconus
species display a similar profile in terms of types and abundance of gene superfamilies. In fact, although with different percentages, the superfamilies O3-like, M, O1, and P appeared to be the most diversified in all the three Profundiconus
specimens. The superfamilies M and O1 are among the most common and abundant conid gene superfamilies (e.g., [49
]), so it is not unexpected to find them well represented in Profundiconus
. The O3 superfamily, the most closely related to the newly described O3-like superfamily, is present in several species, but not abundant throughout conids. It was found in small numbers (up to seven different conotoxins) in 13 conid species with different feeding habits such as fish-, mollusc-, and worm-hunting [53
]. Interestingly, this gene superfamily, along with the J and T ones, were overexpressed in the distal part of the VG of C. (Gastridium) geographus
, accounting for ~50% of the total conotoxin reads found in this segment that is supposed to be the one producing predation venom, but for only ~5% in the proximal segments of the VG, the one producing defence-evoked venom [13
]. The P superfamily was found in several fish-, mollusc-, and worm-hunting Conus
species but always with low diversity (up to 14 different conotoxins), with the exception of Conus (Turriconus) praecellens
A. Adams, 1855, in which this superfamily was the most abundant and diversified [52
]. Remarkably, conotoxins belonging to the T gene superfamily, which are frequently common and diversified in the other conid species, are lacking in the venom gland transcriptome of the Profundiconus
species analyzed here.
The overexpressed conotoxin and profunditoxin fraction, as calculated by differential expression analysis on TMP values, was quite similar in composition between P. cf. vaubani and P. neocaledonicus. In fact, 40% of the gene superfamilies found were overexpressed in both species, and overall, 50–70% of them included at least one transcript overexpressed in the VG. These results suggest that the venom cocktails of P. cf. vaubani and P. neocaledonicus may be not identical but for a large part similar, at least for what concern the conotoxin composition.
Shannon’s diversity and evenness indexes with those of other conid species sequenced in similar conditions, like C. tribblei
(H’ = 3.30, E = 0.90) and C. lenevati
(H’ = 3.30, E = 0.89) [56
], confirmed what was already suggested by conotoxins abundance. P.
and P. neocaledonicus
showed slightly lower conotoxin diversities compared with other species and a less homogeneous distribution of transcripts among gene superfamilies. However, the pipeline we used to identify putative toxins in Profundiconus
tissues was largely based on similarity with conotoxins, potentially leaving highly divergent profunditoxins undetected, and leading to an underestimation of the real toxin diversity in Profundiconus
. Pipelines that do not rely mainly on similarity with previously described toxins but more on, e.g., their structural properties, are required to detect the eventual new toxins of this divergent genus that may be different from what found in Conus
up to now. This is particularly true when it is taken into account that the range of conotoxin precursors and gene superfamilies reported, the number of specimens used (1–20), and the types of sequencing platforms (five types) vary greatly among the conid VG transcriptomes produced to date: From 30 conotoxin precursors and 6 gene superfamilies in Conus (Textilia) bullatus
Linnaeus, 1758 [14
], to 401 conotoxin precursors in Conus (Harmoniconus) sponsalis
Hwass in Bruguière, 1792 [53
] and 55 gene superfamilies in Conus (Chelyconus) ermineus
Born, 1778 [54
] (Table S3
). Two studies reported even higher numbers of conotoxin precursors found in a single species: 662 in Conus (Rhizoconus) miles
Linnaeus, 1758 [57
] and 3305 in Conus (Darioconus) episcopatus
da Motta, 1982 [58
] (Table S3
). These studies, reporting exceptionally high putative conotoxins numbers, may be the only ones that have been able to detect minor conotoxin variants already found in proteomic studies, but may also be a result of sequencing data processing [52
]. This lack of homogeneity, along with the use of different bioinformatic pipelines and threshold criteria, the physiological intraspecific variation of the venom composition [11
], and the lack of corroborating proteomic data for much of the reported findings, demands caution in comparing numbers resulting from different projects.
In conid literature, venom complexity has always been related to prey preference. However, recent studies [53
] pointed out that conotoxin diversity is correlated to diet breadth more than prey type, revealing the tendency of species with more generalized diets to have more complex venoms and more predation-evoked venom genes. Almost nothing is known about Profundiconus
ecology, but the limited data available suggest a worm-hunting diet with at least one species able to prey on fast moving molluscs [24
]. The venom insulins found in Profundiconus
are similar to those of worm- and mollusc-hunting Conus
species and have an additional cysteine residue characteristic these species groups. As a result, based on transcriptomic analyses, P.
and P. neocaledonicus
diets may include only a limited diversity of worm and/or mollusc preys.
3.4. Turripeptides Retained in Profundiconus Venom
, turripeptide-related transcripts were among the most diversified neuropeptide classes. Turripeptides and conotoxins are thought to have a common evolutionary origin as they show similar precursor organization of signal, pro mature, and post regions [35
]. However, when the first turrid venom peptides were discovered, little overlap was found with conotoxins, and the large majority of them belonged to new gene superfamilies, not yet found in conids [31
]. Later, the discovery of peptides similar to conotoxins or turripeptides in distant taxa, likely as the results of convergent evolution, underlined the broader benefits obtained by recruiting them as venom or secretion components. For example, turripeptide-like toxins were found among the feeding-related proteins of the vampire snail Colubraria reticulata
(Blainville, 1829) [63
], in the hunting venom of bloodworms [64
], and in the defensive one of fireworms [65
], while conotoxin-like peptides were found in the mussel Mytilus galloprovincialis
Lamarck, 1819 [66
]. If the present distribution of turripeptides among taxa is considered in the context of the conoidean phylogeny, they might be present in the Conoidea common ancestor, and perhaps even earlier. Eventually, in the Conidae, some turripeptides (what we currently call conotoxins, found in both turrids and conids) started to rapidly diversify, becoming the prevalent component of conid venom, up to the complete loss of turripeptides at least in the genus Conus
. The early divergence of Profundiconus
from the rest of Conidae may explain why some turripeptide gene superfamilies are still retained in this genus. However, this hypothesis needs to be corroborated by characterizing the VG components of more Profundiconus
species, and by a more extensive comparison of turripeptides from other groups belonging to the conid radiation, such as Conasprella
, and Pygmaeoconus
, for which limited, or no data, are available.