Phylogenetic utility of molecular markers—ITS dataset. The use of the nuclear ribosomal gene region ITS has been very important in the reconstruction of phylogenetic relationships in the Passifloraceae (Muschner et al., 2003; Hansen, 2004; Hearn, 2004; Krosnick and Freudenstein, 2005; Krosnick et al., 2006; Chapter 3). In these studies, ITS data has provided information sufficient to resolve relationships at the species to subgeneric levels. However, the caveats associated with the use of ITS as a molecular marker are significant (Alvarez and Wendel, 2003; Bailey et al., 2003). In most cases, phylogenetic studies of Passiflora that use ITS as a marker also include other loci, thus allowing for tests of congruence among the topologies produced by the individual datasets.

Multiple copies of ITS have been identified in subgenus Passiflora (Hansen, 2004), as well as in subgenus Decaloba (Kay, 2003). In Kay’s analysis, ITS was cloned for several species, including P. murucuja and P. tulae. Multiple copies were obtained of equal length, but they were not alignable with one another. In the analysis presented here, both P. murucuja and P. tulae were also difficult to sequence directly for ITS. Therefore, sequences were obtained from Genbank from the study of Muschner et al. (2003). In almost all other species sampled thus far, sequencing of the ITS region was not problematic. The continued use of ITS as a tool for phylogenetic reconstruction in Passiflora is appropriate as long as a cautious approach to its use is taken. The high number of available sequences in Genbank for Passiflora, Adenia, and other genera in Passifloraceae, Turneraceae, and Malesherbiaceae further increases the value of this locus for future studies within the family.

Chloroplast intron and intergenic spacer trnL-F dataset. The chloroplast region trnL-F was first utilized by Muschner et al. (2003) as a molecular marker in Passiflora. Krosnick and Freudenstein (2005), Krosnick, Harris and Freudenstein (2006), and Krosnick (2006) used the trnL-F regionfor reconstruction at the subgeneric level and below with good success. The number of informative characters in the trnL-F datasetis only slightly higher at the subgeneric level than at the at species level . However, as with ITS, the use of chloroplast markers in Passiflora is somewhat problematic. This is because the mode of inheritance of the chloroplast is maternal, paternal, and biparental in Passiflora (Hansen, 2004). The three modes of inheritance have the potential to complicate interpretation of phylogenetic relationships in Passiflora due to intra-individual variation. These anomalies not withstanding, the trnL-F data are valuable at the subgeneric, sectional and species levels within Passiflora.

Chloroplast-expressed glutamine synthetase (ncpGS) dataset. Besides the use of ITS, the next most promising locus available for use in Passiflora has been the single copy nuclear gene ncpGS (Emshwiller and Doyle, 1999; Yockteng and Nadot, 2004a). This gene is most effective in phylogenetic reconstruction at the sectional level and above. Yockteng and Nadot (2004a) and Krosnick (2006) were able to code indel characters for ncpGS that provided additional information content for species reconstruction.

Cytosolic-expressed glutamine synthetase (cytGS) dataset. The chloroplast-expressed copy of glutamine synthetase (ncpGS) has been shown to be a good source of information for phylogenetic reconstruction in several angiosperm families (Emshwiller and Doyle, 1999, 2002; Doyle et al., 2003; Perret et al., 2003; Yockteng and Nadot, 2004a). NcpGS is informative at lower levels, and is advantageous because it is a nuclear gene that is present in the genome as a single copy (Emshwiller and Doyle, 1999). The ncpGS gene is part of a low-copy nuclear family of glutamine synthetase genes (GS) that encode enzymes involved in nitrogen metabolism (Doyle, 1991). Only one copy of GS is expressed in the chloroplast, and this copy (ncpGS) diverged before the evolutionary split between the monocots and dicots (Doyle, 1991). There are at least five other copies of cytosolic-expressed GS (Doyle, 1991; Li et al., 1993), and orthologues are difficult to identify because there is a high degree of sequence similarity within the cytosolic-expressed copies (Doyle, 1991; Biesiadka and Legocki, 1997). The similarity among the cytosolic copies was suggested to be indicative of either concerted evolution among the copies, or of multiple recent duplication events (Doyle, 1991).

While attempting to amplify the chloroplast-expressed copy for phylogenetic analysis in Passiflora, amplifications using primers 687F and 994R (Emshwiller and Doyle, 1999; Figure 4.5) consistently yielded a single cytosolic-expressed copy of GS (cytGS). Yockteng and Nadot (2004b) obtained the same results independently using the same primers. They reported sequences of cytGS as ca. 549 bp, while sequences of cytGS here ranged between 675-850 bp. Sequences from species that overlapped in both studies were aligned and intraspecific variation was very low, suggesting that the same copy of GS was being amplified independently.

The cytosolic-expressed copy of glutamine synthetase is more variable than the chloroplast-expressed copy. Yockteng and Nadot (2004b) reported 5.5% informative characters for the Passiflora cytGS matrix, compared to only 3.67% for ncpGS. The rate of variation for cytGS is comparable to that of ITS; thus, cytGS may be a promising molecular marker for use in Passiflora in species-level analyses.

Material from: