EVIDENCE FOR POST‐TRANSLATIONAL MODIFICATION OF TRIOSE PHOSPHATE ISOMERASE (TPI) IN ISOËTES (ISOËTACEAE)

Abstract

As part of an electrophoretic study on Isoites, a number of Neotropical and North American species were examined for allozyme variation in TPI. Three of these species-L storkii, L flaccida, and L mexicana-exhibit three distinct zones of TPI activity. The two most anodally migrating zones are comparable to the two zones found in most angiosperms and in several other species of Isoetes. The single or three-banded phenotypes produced at these loci correspond, respectively, to the homozygous and heterozygous pattems typical of a dimeric enzyme. The most cathodal zone (zone III) differs in producing either single or two-banded phenotypes. Analyses of these three zones indicate a nearly perfect correlation between zones II and III in putative allelic constitution and relative allelic mobility. Explanations involving TPI gene duplications and/or null alleles fail to account for the peculiar banding characteristics and origin of activity zone III. An altemative hypothesis involving a protease duplication and differential post-translational modification is postulated. This hypothesis adequately explains the zone III phenotypes and fixation of the third activity zone in the species examined. Amino acid sequencing is suggested as the most direct test of this hypothesis. The taxonomic distribution of TPI III generally supports a previous, morphologically-based, hypothesis on species relationships in Isoetes. The presence of this zone is regarded as an independent synapomorphy for a major clade of Neotropical Isoetes. TRIOSE PHOSPHATE ISOMERASE (TPI) is a glycolytic enzyme mediating the interconversion of dihydroxyacetone phosphate and glyceraldehyde 3-phosphate. A majority of diploid angiosperms and certain pteridophytes typically exhibit two zones of banding activity on starch gels indicating that TPI is encoded by two loci. Genetic studies (Gottlieb, 1982) have shown that these two zones do in fact represent discrete loci, both of which are nuclear. Additional studies employing pollen exudates and cell fractionations (Gottlieb, 1982; Gastony and Darrow, 1983; Soltis, 1986) have shown that the products of these two genes are differentially expressed in the cytosol and the plastids. Because this enzyme is dimeric, individuals heterozygous at a TPI locus normally resolve as a three-banded pattern using standard starch gel electrophoretic techniques; individuals I Received for publication 1O November 1987; revision accepted 6 June 1988. This research was supported by NSFgrant BSR 86-0672. The authors thank Drs. Susan Bamum, Jan Jaworski, and Jack Vaughn for comments and assistance during the formulation of this paper. We also thank Elizabeth Southworth for her help in the laboratory. The authors are grateful to the three reviewers who provided helpful suggestions on and insights into novel interpretations of the TPI phenotypes observed. homozygous at TPI loci produce only a single band at each locus. Interlocus heterodimers are not known to occur in plants between TPI loci in different compartments (Gottlieb, 1982). Although TPI is normally encoded by two loci, a number of cases are known in which at least one of these TPI loci has been duplicated. Gottlieb (1982), for example, has shown that both cytosolic and plastid TPI have been duplicated subsequent to the origin of the genus Clarkia and that one or both of these duplications represent distinguishing features for most species in the genus. A characteristic of gene duplications in TPI and in other dimeric enzymes is the presence of one or more sets of three nonsegregating bands: a case of apparent fixed heterozygosity. Examples ofthis apparent fixed heterozygosity have been found in a large number of pteridophytes including Lycopodium (Soltis and Soltis, 1988), Equisetum (Soltis, 1986), and nearly all members of the Polypodiaceae sensu lato examined to date (see Haufler and Soltis, 1986). In Lycopodium and Equisetum, the additional TPI bands are compartmentalized in the plastids, whereas the extra bands of the polypodiaceous ferns are cytosolic. Three hypotheses have been put forth to explain the peculiar three-banded pattern