Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and triosephosphate isomerase (TPI) are essential to glycolysis, the major route of carbohydrate breakdown in eukaryotes. In animals and other heterotrophic eukaryotes, both enzymes are localized in the cytosol; in photosynthetic eukaryotes, GAPDH and TPI exist as isoenzymes that function in the glycolytic pathway of the cytosol and in the Calvin cycle of chloroplasts. Here, we show that diatoms--photosynthetic protists that acquired their plastids through secondary symbiotic engulfment of a eukaryotic rhodophyte--possess an additional isoenzyme each of both GAPDH and TPI. Surprisingly, these new forms are expressed as an TPI-GAPDH fusion protein which is imported into mitochondria prior to its assembly into a tetrameric bifunctional enzyme complex. Homologs of this translational fusion are shown to be conserved and expressed also in nonphotosynthetic, heterokont-flagellated oomycetes. Phylogenetic analyses show that mitochondrial GAPDH and its N-terminal TPI fusion branch deeply within their respective eukaryotic protein phylogenies, suggesting that diatom mitochondria may have retained an ancestral state of glycolytic compartmentation that existed at the onset of mitochondrial symbiosis. These findings strongly support the view that nuclear genes for enzymes of glycolysis in eukaryotes were acquired from mitochondrial genomes and provide new insights into the evolutionary history (host-symbiont relationships) of diatoms and other heterokont-flagellated protists.

We have isolated a gene from a library of nuclear DNA for a chlorophyll a/c-binding protein (named Cac for chl a/c by analogy with Cab for chl a/b) of a chromophyte alga, Giraudyopsis stellifer, and sequenced it. The comparison of the deduced amino acid sequence with other chl a/c- and chl a/b-binding protein sequences shows that structural and functional features, i.e. the arrangement 'en X' of the two A and B transmembrane helices and the putative chl a-binding sites, are shared by both Chlorophyta and Chromophyta. Moreover, in contrast to Chlorophyta, a very strong identity is found among Chromophyta in the C helix suggesting a major function associated to this specific region. Nevertheless, the primary structure of the apoprotein does not seem affected by the pigment composition in Chromophyta. As in the few other examples currently known, we confirm that the cac genes are nuclear-encoded and are part of a multigenic family. Northern blots, performed on poly(A)+ mRNA from G: stellifer, give evidence that the cac gene is light-induced at a transcriptional level and that no expression can be observed in the dark.

No abstract