MATERIALS AND METHODS
Sample collection, treatment and toxicity test. The samples used in this study have been tested for the TTX toxicity in our previous published research 22. Based on the toxicity test,N. succinctus and N. variciferus were selected for transcriptional analysis since they included both toxic and non-toxic specimens. The liver and pancreas were treated for RNA extraction and sequencing since they contained more toxin. For RNA-seq protocol, the toxic and non-toxic specimens were sequenced as treatment group and control group respectively for each species. We selected three toxic specimens and three non-toxic specimens as triple repeat for treatment group and control group respectively, for each of N. succinctusand N. variciferus . The detailed sample collection was shown in Fig. 1.
RNA library preparation, transcriptome and gene assembly. After RNA extraction the mRNA were fragmented with fragmentation buffer and were used to synthesize cDNA. Short cDNA fragments were purified and then resolved with end reparation and adapter connection, from which fragments with suitable length were selected for PCR enrichment. Agilent 2100 Bioanaylzer and ABI StepOnePlus Real-Time PCR System was used for qualification. Finally, the library was sequenced using Illumina HiSeq 400023.
The Trinity (v2.0.6) 24 was used to perform de novo assembly after filtering out raw reads. The assembly quality was evaluated with N50 and ortholog hit ratios (OHR). Paired reads passing the filter were then concatenated using Concatenate datasets (version 1.0.0) in both the right and left direction. Transcript abundancies were calculated by RSEM version 1.1.175425-26 using the pool of non-normalized reads with default settings.
Differentially expressed gene analysis. The functional annotation was performed from databases of NT, NR, GO, COG, KEGG, SwissProt and InterPro. With BLASTX the final assembly was submitted to these databases with the threshold of E-value ≤ 10−527. The NR and InterProScan5 were used to get GO annotation and InterPro annotation respectively. Unigenes were obtained from mapping clean reads by Bowtie228. The gene expression level was calculated with RSEM. The PCA analysis was performed with all samples by princomp, a function of R. We detected DEGs (Differentially Expressed Genes) with Possion Dis as requested29.
Q-PCR confirmation. Real-time PCR was performed with the SG Fast qPCR Master Mix (High Rox) using the cDNA synthetized above. Cycle threshold values were normalized to the gene 18S. A total of 18 DEGs closely related with coding toxin-binding protein were selected as target Q-PCR functional genes. Fold changes were determined with the relative expression software tool.
Mutations in sodium channels genes. Based on the transcript unigenes of all samples, we retrieved the s odium channels genes to which the TTX binds. We downloaded available s odium channels genes from mollusca, human, mouse and others in NCBI as reference database. Then all unigenes of N. succinctus and N. variciferus were mapped to the reference database. While the unigenes were best matched to the reference sequences specific primers were designed to sequence them for confirmation.