Nucleomorph genome sequence of the cryptophyte alga Chroomonas mesostigmatica CCMP1168 reveals lineage-specific gene loss and genome complexity

Abstract

Cryptophytes are a diverse lineage of marine and freshwater, photosynthetic and secondarily nonphotosynthetic algae that acquired their plastids (chloroplasts) by "secondary" (i.e., eukaryote-eukaryote) endosymbiosis. Consequently, they are among the most genetically complex cells known and have four genomes: a mitochondrial, plastid, "master" nuclear, and residual nuclear genome of secondary endosymbiotic origin, the so-called "nucleomorph" genome. Sequenced nucleomorph genomes are ∼1,000-kilobase pairs (Kbp) or less in size and are comprised of three linear, compositionally biased chromosomes. Although most functionally annotated nucleomorph genes encode proteins involved in core eukaryotic processes, up to 40% of the genes in these genomes remain unidentifiable. To gain insight into the function and evolutionary fate of nucleomorph genomes, we used 454 and Illumina technologies to completely sequence the nucleomorph genome of the cryptophyte Chroomonas mesostigmatica CCMP1168. At 702.9 Kbp in size, the C. mesostigmatica nucleomorph genome is the largest and the most complex nucleomorph genome sequenced to date. Our comparative analyses reveal the existence of a highly conserved core set of genes required for maintenance of the cryptophyte nucleomorph and plastid, as well as examples of lineage-specific gene loss resulting in differential loss of typical eukaryotic functions, e.g., proteasome-mediated protein degradation, in the four cryptophyte lineages examined.

Fig. 1.—

Chroomonas mesostigmatica CCMP1168 nucleomorph genome map. The genome is comprised of three linear chromosomes, shown broken artificially at their midpoints, with genes on the left indicating transcription from bottom to top, and genes on the right indicating transcription from top to bottom. Colors of the blocks correspond to assigned functional categories, and multicopy genes are highlighted in pink. An asterisk beside the gene name indicates the gene contains an intron. Genes for which there are currently no known homologs (ORFans) are shown in black, genes that have homologs only in other cryptophyte nucleomorphs (nORFs) are shown in orange, and motif-containing genes whose identity cannot be determined with confidence are shown in light green. ORFan genes that retain conserved positions within syntenic regions between one or more other cryptophyte nucleomorphs are shown in gray (syntenic ORFans).

Fig. 2.—

Karyotype analysis of the C. mesostigmatica nucleomorph. (a) Ethidium-bromide-stained C. mesostigmatica chromosomes separated by PFGE for 24 h for three approximate total cell counts: 1) 5 × 10⁶ cells, 2) 1 × 10⁷ cells, and 3) 5 × 10⁷ cells. Nucleomorph chromosomes are indicated by arrowheads. Lambda DNA is used for size markers. (b) Ethidium-bromide-stained C. mesostigmatica nucleomorph chromosomes separated by PFGE for 60 h (left) and Southern-blot hybridization using a C. mesostigmatica nucleomorph-specific 18S probe.

Fig. 3.—

Percentage of all cryptophyte nucleomorph ORFs per genome as a function of length. Each of the four nucleomorph genomes examined in this study has a different distribution of ORF sizes. The smaller nucleomorph genomes are enriched in shorter ORFs, and as the size of the ORF increases, the percentage of those ORFs decreases. Larger nucleomorph genomes are slightly enriched in longer ORFs.

Fig. 4.—

Four-way cryptophyte nucleomorph gene content comparison. There are 311 genes of known or predicted function annotated in cryptophyte nucleomorph genomes, 216 of which (∼70%) are present in all four cryptophyte nucleomorph genomes presently sequenced, forming a highly conserved core gene set. Aside from the lineage-specific photosynthesis-related, spliceosome, and proteasome gene loss, the distribution of missing genes appears to be random with respect to each species and functional gene category.

Fig. 5.—

Hypothetical proteins inferred from complete cryptophyte nucleomorph genomes. The graph shows the proportion of cryptophyte nucleomorph-specific hypothetical protein-coding genes (nORFs) and hypothetical protein-coding genes unique to each individual nucleomorph genome (ORFans) relative to the total number of hypothetical protein-coding genes as additional nucleomorph genomic sequences become available. The leftmost bar compares the proportions of the two types of hypothetical proteins for Guillardia theta and Hemiselmis andersenii. The second bar compares these proportions with the addition of Chroomonas paramecium. The proportions do not change substantially until the addition of the fourth genome, C. mesostigmatica, shown in the third bar, where the proportion of ORFan genes drops by 21%.

Fig. 6.—

Verification of spliceosomal intron removal in the Chroomonas mesostigmatica nucleomorph genome. The figure shows agarose gel electrophoresis of PCR amplicons generated using cDNA and genomic DNA template and site-specific primers. Genes examined were (a) rps24 and (b) rpl9, rps16, and rps23. The cDNA amplicons (indicated by arrowheads) for each gene are shorter in length compared with their respective PCR-generated genomic DNA amplicons. Intron removal was verified by sequencing.

Fig. 7.—

ORF degradation in cryptophyte nucleomorph genomes. Schematic shows degenerating ORFs in syntenic regions between C. mesostigmatica, H. andersenii, and C. paramecium. Homologous genes are shown in gray, with gray highlights indicating the syntenic positions of the genes on the chromosome of each species. Genes shown in black are ORFan genes. Genes shown in red are those where one or more ORFan genes occupy the same syntenic position in a stretch of genes that have conserved order in another nucleomorph genome, which are highlighted in red. (a) An ORFan gene of 59 amino acids in C. mesostigmatica occupies the same syntenic position as the proteasome subunit gene prsB5 in H. andersenii. (b) ORFan genes occupy the same syntenic positions in H. andersenii as the spliceosomal genes prp2-like and prp4-like in C. mesostigmatica. (c) Three ORFan genes on chromosome two in C. paramecium occupy the same syntenic position and sum to be a similar size as the splicing factor gene sf3b3-like in C. mesostigmatica.