br Results and discussion A systematic search for genes enco
Results and discussion A systematic search for genes encoding H+-PPases was carried out by PCR using genomic DNA from a wide range of protozoa as templates and degenerate oligonucleotides, designed from amino Fluorescein-12-dUTP domains conserved in H+-PPases of higher plants and the proteobacterium R. rubrum, as primers. This approach allowed the specific amplification of DNA fragments of around 0.6kb, comprising ca. 25–30% of the complete open reading frame of typical H+-PPases genes, from most of the genomes tested (Fig. 1). Fragments were expected to correspond to a region near the C-terminus of the protein, covering cytosolic loops V–VII, which has been previously shown to be a good molecular marker for phylogenetic studies . A single DNA band of ca. 0.56kb was obtained for all trypanosomatids, a group of protists for which no introns have been described so far, for the ciliates H. cavicola (hypotrich) and V. microstoma (peritrich), the photosynthetic heterokont (golden-brown alga) O. danica, the amoeboid slime mold D. discoideum, and the apicomplexans P. falciparum and T. gondii. A fragment of the same size was obtained when a cDNA clone containing the full Arabidopsis thaliana AVP1 gene was used as template (data not shown). Slightly larger bands (ca. 0.62–0.63kb) were amplified in the cases of the hymenostomatid ciliate P. tetraurelia and the heterotrophic euglenoid A. longa. Clear exceptions to this pattern were the two larger DNA fragments of ca. 0.7 and 0.83kb obtained for the hymenostomatid ciliate T. pyriformis (Fig. 1). This result was not unexpected since large introns (mean size, ca. 250bp) have been reported to be present in the genomic macronuclear DNA of this protozoan strain . Ligation of these bands into p-GEM-T plasmid and subsequent transformation of Escherichia coli resulted in artifacts due to bacterial recombination, making the cloning of the fragments especially difficult and impossible in some cases (see below); therefore, direct sequencing of the PCR-amplified bands was normally carried out. Sequences with no ambiguities were usually obtained (all trypanosomatids, O. danica, A. longa, the ciliates H. cavicola and V. microstoma, and the two bands of T. pyriformis), indicating that the bands were not mixtures of several DNA fragments of the same size and suggesting that only one gene had been amplified from most of the respective genomes. Since many ambiguities did appear when the bands from P. tetraurelia and D. discoideum were directly sequenced, subcloning of the PCR-amplified fragments in p-GEM-T was attempted. This approach proved to be successful only for P. tetraurelia, yielding two different clones (see below). Sequence analysis and BLAST search on databases unequivocally showed that all DNA bands were fragments of genes encoding H+-PPase; even the sequence from D. discoideium, although incomplete, hinted H+-PPase signatures, in agreement with previous biochemical studies . Moreover, they exhibited the expected genomic structures of the various protozoan strains tested, thus, no introns were found for the single-copy H+-PPase gene fragments of trypanosomatids, apicomplexans, the heterokont O. danica, and the euglenoid A. longa. The sequences of the ciliate strains tested—hymenostomatids (Tetrahymena, Paramecium), peritrichs (Vorticella) and hypotrichs (Histriculus)—showed the peculiar features for the genes of this protozoan group (Fig. 2): in-frame stop codons coding for Gln  and canonical introns of very different length depending on the strain—very small (27bp) in Paramecium and longer introns (0.11–0.23kb) with a very high A+T content in Tetrahymena. The identification of two DNA fragments with different sequences both in P. tetraurelia and T. pyriformis indicated the occurrence of two paralogous H+-PPase genes in these microorganisms. Sequence analyses further demonstrated that plant-like H+-PPase genes, very similar to those of human parasitic trypanosomes , , , are present in parasitic trypanosomatids of insects and other animals (Leptomonas, Herpetomonas, Endotrypanum) and plants (Phytomonas) and in free-living strains (Crithidia). These results demonstrate that the occurrence of this protein is not necessarily associated to the pathogenic or even the parasitic character of these protozoa, actually, plant-like H+-PPase genes were also found in free-living protozoa of other phylogenetic groups (ciliates, heterotrophic euglenoids, heterokonts). The finding of H+-PPase genes similar to those of higher plants in heterokonts of such diverse behavioral and physiological features as the mixotrophic golden-brown alga O. danica and the plant-pathogenic oomycete Phytophthora infestans may have interesting evolutionary implications (see below).