AUTHOR=Lawrence Travis J. , Datwyler Shannon L. 

TITLE=Testing the Hypothesis of Allopolyploidy in the Origin of Penstemon azureus (Plantaginaceae)

JOURNAL=Frontiers in Ecology and Evolution

VOLUME=4

YEAR=2016

URL=https://www.frontiersin.org/journals/ecology-and-evolution/articles/10.3389/fevo.2016.00060

DOI=10.3389/fevo.2016.00060

ISSN=2296-701X

ABSTRACT=<p>Polyploidy plays a major role in the evolution of angiosperms with multiple paleopolyploid events having been suggested making it plausible that all angiosperms have a polyploid event in their history. Polyploidy can arise by genome duplication within a species (i.e., autopolyploidy), or whole genome duplication coupled with hybridization (i.e., allopolyploidy). <italic>Penstemon</italic> subsection <italic>Saccanthera</italic> contains a species complex of closely related diploids and polyploids. The species in this complex are <italic>P. heterophyllus</italic> (2x, 4x), <italic>P. parvulus</italic> (4x), <italic>P. neotericus</italic> (8x), <italic>P. laetus</italic> (2x), and <italic>P. azureus</italic> (6x). Previous studies have hypothesized that <italic>P. azureus</italic> is an allopolyploid of <italic>P. parvulus</italic> (4x) X <italic>P. laetus</italic> (2x). To test the hypothesis of allopolyploidy in the origin of <italic>P. azureus</italic> and to determine possible progenitors, two nuclear loci (<italic>Adh</italic> and <italic>NIA</italic>) and three chloroplast spacer regions (t<italic>rnD</italic>-<italic>trnT, rpoB</italic>-<italic>trnC, rpl32</italic>-<italic>trnL</italic>) were sequenced from <italic>P. azureus, P. heterophyllus, P. laetus, P. parvulus</italic>, and <italic>P. neotericus</italic>. These data were analyzed in a phylogenetic framework and a network analysis was used on the nuclear data. Both nuclear datasets supported the allopolyploid origin of <italic>P. azureus</italic> with divergent <italic>Adh</italic> orthologs recovered in all five accessions of <italic>P. azureus</italic> and divergent <italic>NIA</italic> orthologs recovered in three of the <italic>P. azureus</italic> accessions. Furthermore, the <italic>Adh</italic> and <italic>NIA</italic> datasets support three hypotheses for the possible progenitors of <italic>P. azureus</italic>: (1) <italic>P. heterophyllus</italic> (2x) X <italic>P. parvulus</italic>; (2) <italic>P. heterophyllus</italic> (4x) X <italic>P. laetus</italic>; and (3) <italic>P. heterophyllus</italic> (2x) X <italic>P. heterophyllus</italic> (4x). Markedly, all hypotheses support <italic>P. heterophyllus</italic> (2x) as a progenitor. In addition, the <italic>Adh</italic> and <italic>NIA</italic> trees also suggest that <italic>P. neotericus</italic> is an allopolyploid, <italic>P. parvulus</italic> is an autopolyploid with two distinct origins, and <italic>P. heterophyllus</italic> (4x) is an autopolyploid or allopolyploid. The cpDNA analysis resolved geographically structured clades. This pattern seems best explained by local adaptation of the chloroplast genome and chloroplast capture. However, multiple origins of the polyploid species and gene flow could also explain this pattern. This study provides the first molecular phylogenetic evidence of the allopolyploid origin of <italic>P. azureus</italic> and has given insight into the origin of the other polyploids in this species complex.</p>