Carotenoids & LGT

Rius, M., Rest, J. S., Filloramo, G. V., Novák Vanclová, A. M., Archibald, J. M., & Collier, J. L. (2023). Horizontal gene transfer and fusion spread carotenogenesis among diverse heterotrophic protists. Genome Biology and Evolution15(3), evad029.

Phylogeny of Phytoene and Squalene Synthase and Related Proteins

We have identified an unusual gene in both Thraustochytrid genomes (JGI protein ID 150841 in Aurantiochytrium limacinum) that is likely involved in carotenoid pigment biosynthesis, an ecologically relevant feature of many labyrinthulomycetes. This appears to be a unique fused gene containing domains for both phytoene synthase, phytoene desaturase, and lycopene cyclase.  These domains each encode successive steps in the biosynthesis of carotendoids. They are typically encoded as separate proteins, or as fused proteins with only two of these functions. The only other such fused gene with all three activities fused together is found in Thecamonas.  In order to investigate the origin and evolution of this fusion, our first step has been a study of the evolution of the putative phytoene synthase domain. Phytoene synthase is  involved in the first committed step of carotenoid biosynthesis, where it condenses two molecules of geranylgeranyl pyrophosphate to form phytoene.

We searched for phytoene synthase in a protein database composed of the UNIPROT proteome reference database (March, 2016) including Archaea, Bacteria, and Eukaryotes, supplemented with additional predicted proteomes from a variety of protists, as well as predicted proteins from the MMETSP project. Our search strategy was to find matching sequences in this database to PFAM HMM PF00494, which is a statistical characterization of phytoene and squalene synthase. These matches were identified using HMMER, and the alignment is based on this HMM. Our resulting alignment of 5,849 proteins is available at academic commons (pending). The phylogeny was constructed using ML with FastTree. The resulting phylogeny (Figure 1) is available in NEWICK format at academic commons (pending).

Our preliminary trees suggests that the phytoene synthase domain in Thraustochytrid genomes group with bacterial domains, suggestive of horizontal gene transfer. However, this preliminary result needs to be subjected to a hypothesis testing statistical framework to address alternative possible origins. Similar analyses are in progress to identify the origin of the phytoene desaturase and lycopene cyclase; preliminary analyses suggest that they also do not group with other stramenopiles.

Figure 1. Phylogeny of 5,849 phytoene and squalene synthase and related proteins. The phylogeny is visualized by Dendroscope. Randomly selected taxon names are shown. Branches and labels are colored by major taxonomic group. Thraustochytrid sequences are highlighted. In this phylogeny, Thraustochytrid sequences do not group with other stramenopile sequences, suggestive of horizontal gene transfer, although this hypothesis remains to be statistically tested.

Currently, we are working on knocking out this gene and analyzing the phenotype in Aurantiochytrium.  Because this should disrupt the carotenoid biosynthesis pathway, this should result in an easily screened visible phenotype (loss or reduction of carotenoid production in mutant colonies, assuming the phenotype is not completely recessive). Under appropriate growth conditions (specifically, in a rich soft agar medium exposed to light) wild-type Aurantiochytrium limacinum ATCC MYA-1381 colonies develop strong orange/pink carotenoid pigmentation, so we expect to be able to select and screen simultaneously for both antibiotic resistance (indicating presence of the construct) and modification of pigment in a colony (indicating site-specific integration). We will confirm the genetic structure (integration of the inactivation cassette by double homologous recombination) of putative knockouts by a combination of PCR and Southern blot analyses.