Dinoflagellates are an ecologically significant group related to ciliates and apicomplexans. Like many phytoplankton, this taxa is identified using traditional morphological archetype and over 2000 species have been described this way to date. However, Dinoflagellates pose a particular challenge as there are many cryptic species which remain unidentified. Recent molecular phylogeny has shown that Dinoflagellates, once thought to be of the same genera, are in fact paraphyletic e.g Amphidinum and Gymnodinium. Some taxa such as Symbiodinium, instead of being a single species by morphological standards, are now known to contain hundreds of cryptic taxanomic units; most of which have not been identified. In order to standardize the method of Dinoflagellate classification, a DNA-based identification system should be implemented. DNA-barcoding may hold the key for this concept to be conceived. This technique involves using a short stretch of DNA sequencing to identify a species; known as a barcode marker. This technique has been applied to two mitochondrial markers known as the Cytochrome Oxidase I (COI) and Cytochrome Oxidase B genes (COB). However, these barcodes have limitations as both cannot always be amplified from all Dinoflagellates and both cannot specifically identify species from the same genera. One of the main objectives of DNA-barcoding is to accurately and quickly identify toxic HAB species, though the modern barcode markers already in existence are unable to resolve many issues surrounding these hazardous taxa.
The aims of this study were to test a third barcode marker, this being the internal transcribed spacer (ITS) units 1 and 2. These markers were used to identify various Dinoflagellate taxa from a range of global algal collections. The ITS markers have previously been used to identify Eukaryotes, Metazoans and some Dinoflagellates.
The limitations of the ITS markers were previously described as being that they are found in multiple distinct copies, with high intra/intergenomic variation. This can make sequencing challenging and alignment difficult. However, to test the international applicability of the ITS markers this paper assessed nearly 400 dinoflagellate strains.
By collecting 669 dinoflagellates and eliminating low-quality sequences, 151 ITS barcodes were left remaining. 242 ITS sequences were collected from GenBank and added to this, equalling 393 ITS barcodes from 78 identified species. Each strain was centrifuged, snap-frozen and thawed three times before DNA extraction was conducted. PCR sequencing was then carried out resulting in products ranging from 500-600bp. These were then DNA-barcoded on the premises or sent to the Canadian Centre of DNA Bar-coding to be analysed. The sequences were screened for contamination using BLAST.
These markers were heavily biased toward larger assemblages with Symbiodinium, Alexandrium and Pfiesteriaceae being two-thirds of those identified. Some members of the Pfiesterial and Oxyrrhis were poorly represented whereas Lingulodinium, Symbiodinium, Gymnodinium, Gyrodinium and Karenia were represented at an above average level.
The ITS markers could identify 93% of known strains and was able to resolve previously genera/species arguments in Lingulodinium and Protoceratiumand taxa. Unusually, Gymnodinium sp, CCMP 424, was re-identified as Heterocapsa niei, in conflict with COI barcode markers. However, the strain was re-sequenced using ITS and SSU sequencing, confirming that this was not a case of PCR contamination as originally believed.
The results indicated that there were 21 cases of misidentification which may have arisen from two cryptic species being thought as one. It was discovered that there were four Dinoflagellate strains that did not match their given taxa and two other strains that, although given the same name, were not of the same species.
32 strains were shown to have been previously unknown cryptic species revealing a higher diversity than originally thought. This included strains from the Gyrodinium, Karlodium and Cryptoperidinopsis taxa.
The success rate in these studies varied between 2% and 53% suggesting that the ITS markers are too technical to work as a generic DNA-barcoding marker for Dinoflagellates, with COB and COI being more broadly comparable. Although, ti should be noted that COI primers were shown to fail when sequencing Amphidinium, Heterocapsa and Oxyrrhis. The causes of failure in ITS, COI and COB sequencing were thought to be a mixture of issues. Failed amplification, failed sequencing and poor sequencing quality were all thought to contribute toward failed sequencing. The ITS markers were shown to lack sequencing efficiency with an observed success rate of 50% compared to 66% for COI. This was attributed to poor sequencing quality and the mis-representation of Pfiesteriaceae and Alexandrium were thought to skew this data set. However, despite this, the ITS markers successfully identified 96% of strains including 3 genus that had no identity.
This study was considered significant as it highlights the genetic instability in long-term cultured strains worldwide and focuses on the need for re-examination into Heterocapsa taxa. DNA-barcoding is important for this reason as it helps to identify stability of cultured strains, with ITS bar-coding being particularly important as it is capable of identifying cryptic species and speciation events in long-cultured strains. ITS markers were also considered good markers to identify harmful Dinoflagellate groups.
Stern, RF, Anderson RA, Jameson I, Kupper FC, Coffroth MA, Vaulot D, Le Gall F, Veron B, Brand JJ, Skelton H, Kasai F, Lily EL & Keeling PJ. 2012 Evaluating the Ribosomal Internal Transcriber Spacer (ITS) as a Candidate Dinoflagellate Barcode Marker. PLOS One. 10.