F ull-length-enriched cD N A libraries from E chinococcus granulosus contain separate populations of oligo-capped and trans-spliced transcripts and a high level of predicted signal peptide sequences

T he tissue-dwelling larval stages of the cestode E chinococcus granulosus are intimately associated with the host, implying that a range of molecular mediators may be secreted by the parasite into the host environment. T hese mediators are being sought through a transcriptome-based analysis, using recombinant cD N A libraries. C onventional cD N A libraries of E . granulosus contain high levels of mitochondrial transcripts, as well as host (bovine) genomic D N A . I n particular, 60% of a conventional protoscolex stage cD N A library corresponds to the large subunit (L SU ) of mitochondrial rR N A . W e attribute the presence of L SU rR N A copies to its polyadenylation in E . granulosus . T o circumvent this problem, we adapted the 5 R apid A mplification of cD N A E nds (R N A -ligase mediated R A C E ) technique that excludes all polynucleotides missing the 7-methyl-guanosine (7M G ) cap specific to the 5 end of full-length mR N A . B y ligating a specific oligonucleotide (oligo-cap) to 7M G -bearing mR N A , three cD N A libraries were made by PC R from oligo-cap and oligo-dT primers. A nalysis of these libraries showed that mitochondrial R N A contaminants had been excluded. M oreover, no bovine genomic sequences were detected. I n parallel, we constructed three cD N A libraries using the newly described trans-spliced leader (SL ) from E chinococcus . A lthough these represent a smaller subset of parasite genes, mitochondrial and genomic contributions were again excluded. I n both cases, a majority of cD N A s (61 /92%) were judged to contain the initiation A T G codon, and 11 /27% of inserts included potential N -terminal signal sequences. T he 5 U T R tracts of most oligo-capped cD N A s were /100 nt, although /8% were longer than this. A mong the trans-spliced cD N A s, 43% potentially utilise the A U G donated by the SL , and in only 6% was the SL separated from an endogenous putative start site by /60 nt. Sequence analysis of randomly selected clones shows virtually no overlap between the oligo-capped and SL libraries, indicating that trans-spliced E . granulosus mR N A s appear to be insensitive to the enzymatic treatments used to ‘oligo-cap’ unspliced mR N A s. T he oligo-capped and SL strategies represent efficient and complementary pathways to isolate full-length cD N A clones from this cestode parasite and, possibly, from related parasitic flatworms. # 2002 E lsevier Science B .V . A ll rights reserved. K eywords: C estode; G enome; H elminth; Open reading frames; U ntranslated regions