Parasitol Res (1995) 81:696-699 9Springer-Verlag 1995 John Ellis- David Morrison Effects of sequence alignment on the phylogeny of Sarcocystis deduced from 18S rDNA sequences Received: 8 February 1995 / Accepted: 3 April 1995 Abstract The family Sarcocystidae contains a wide va- riety of parasitic protozoa, some of which are important pathogens of livestock and humans. The taxonomic rela- tionships between two of the genera in this family (Toxo- plasma and Sarcocystis) have been debated for a number of years and remain controversial. Recent studies, from comparisons of 18S rDNA-sequence data, have suggest- ed that Sarcocystis is paraphyletic, although a hypothesis supporting monophyly of Sarcocystis could not be reject- ed. The present study shows that the phylogenetically in- formative nucleotide positions within the 18S rDNA are primarily located in the regions that make up the helices in the secondary structure of the 18S rRNA. A phyloge- netic analysis of 18S rDNA-sequence data aligned by secondary structure constraints, or a subset of the data corresponding to all nucleotides found in the helices, provide unambiguous evidence supporting monophyly of Sarcocystis. One of the essential steps in any evolutionary study is to establish an hypothesis of homology among the taxa be- ing studied, and in molecular analyses this involves pair- wise alignment of the nucleotides or amino acids. The sequences are compared using a pattern-matching pro- cess that searches for correspondence between the ele- ments of the sequences, introducing gaps into the se- quences as required to maximize some criterion for opti- mality of the correspondence (Chan et al. 1992). There are many alignment algorithms currently available (Wa- terman 1989; Doolittle 1990; Chart et al. 1992), which maximize a wide variety of mathematical functions mea- suring correspondence between the sequences. However, J. Ellis ( ~ ) Department of Cell and Molecular Biology, University of Technology Sydney,Gore Hill, New South Wales, Australia D. Morrison Department of Environmental Biology and Horticulture, University of Technology Sydney, Gore Hill, New South Wales, Australia most of these algorithms use heuristic procedures, and hence do not make use of knowledge of the secondary structure of the molecules when such knowledge is avail- able (Chan et al. 1992). Analyses based on alignments that do not incorporate the known biological function of the nucleotide sequences are likely, in general, to be in- ferior to analyses where the nucleotides are aligned ac- cording to their known function within the sequences (Hillis and Dixon 1991). Furthermore, when alignment regions contain many gaps (insertions and deletions), these regions are often excised prior to phylogeny recon- struction on the grounds that they are phylogenetically uninformative, although there are few objective criteria for these judgements (Gatesby et al. 1993). These considerations suggest that superior phyloge- netic analyses should be provided using alignments based on secondary structure, taking into consideration the cladistic informativeness of different regions within the molecule (Wheeler and Honeycutt 1988; Smith 1989; Dixon and Hillis 1993). In the present study we tested this hypothesis for a group of taxa whose phylogeny has previously proved problematic so as to see whether the problems simply derive from inadequate alignments (previous analyses used heuristic techniques only) and/or the inclusion of uninformative data (previous analyses did not exclude any of the nucleotide positions). The phylogenetic relationships among protozoan par- asites of the family Sarcocystidae as inferred from com- parisons of 18S ribosomal RNA (rRNA) sequence data remain ambiguous. Generally speaking, it is now accept- ed that the genera Toxoplasma and Sarcocystis form a monophyletic group (i.e. are derived from a recent com- mon ancestor; Johnson et al. 1988; Tenter et al, 1992; El- lis et al. 1994a; Holmdahl et al. 1994; Ellis et al. 1995). However, a recent report concluded that the genus Sareo- cystis itself is paraphyletic, since the analysis of partial 18S rRNA sequences identified two mouophyletic sub- groups within Sarcocystis, one subgroup containing two species having felids as definitive hosts and the other subgroup containing four species having canids as defin- itive hosts, and these two clades together were monophy-