Phylogenetic analysis of eukaryotic NEET proteins uncovers a link between a key gene duplication event and the evolution of vertebrates

Abstract

NEET proteins belong to a unique family of iron-sulfur proteins in which the 2Fe-2S cluster is coordinated by a CDGSH domain that is followed by the "NEET" motif. They are involved in the regulation of iron and reactive oxygen metabolism, and have been associated with the progression of diabetes, cancer, aging and neurodegenerative diseases. Despite their important biological functions, the evolution and diversification of eukaryotic NEET proteins are largely unknown. Here we used the three members of the human NEET protein family (CISD1, mitoNEET; CISD2, NAF-1 or Miner 1; and CISD3, Miner2) as our guides to conduct a phylogenetic analysis of eukaryotic NEET proteins and their evolution. Our findings identified the slime mold Dictyostelium discoideum's CISD proteins as the closest to the ancient archetype of eukaryotic NEET proteins. We further identified CISD3 homologs in fungi that were previously reported not to contain any NEET proteins, and revealed that plants lack homolog(s) of CISD3. Furthermore, our study suggests that the mammalian NEET proteins, mitoNEET (CISD1) and NAF-1 (CISD2), emerged via gene duplication around the origin of vertebrates. Our findings provide new insights into the classification and expansion of the NEET protein family, as well as offer clues to the diverged functions of the human mitoNEET and NAF-1 proteins.

Figure 1. zf-CDGSH domain organization and architecture across different lineages.

The conserved sequence C-X-C-X2-(S/T)-X3-P-X-C-D-G-(S/A/T)-H is a defining feature of the CDGSH protein family (3Cis-1His coordinates are in bold and the CDGSH motif is highlighted in yellow). The presence or absence of each protein type in bacteria, archaea, fungi, algae, plants and animals is indicated on the right. Human CISD1/CISD2 NEET proteins belong to groups 1 and 3, and human CISD3 NEET protein belongs to group 2. The three human CDGSH NEET proteins (represented by groups 1–3; dashed box) were used for all BLAST searches and phylogenetic tree analysis of NEET proteins in this work.

Figure 2. Occurrence of CISD, CISD1, CISD2 and CISD3 proteins in different species.

A taxonomy common tree of species was obtained from NCBI (NCBI Taxonomy). The tree was populated with one representative fully sequenced genome on each of its branches (Supplementary Table S2). Each genome was then subjected to a PSI-BLAST search with each of the different human CISD sequences and the presence or absence of human CISD homologs is indicated on right. When a clear distinction could not be made between homologs of CISD1 or CISD2 with a 50% similarity cutoff to the two different proteins, the homolog was annotated as CISD. All CISD, CISD1, and CISD2 homologs contain a single copy of the CDGSH domain per polypeptide chain, and all homologs of CISD3 contain two. Oxygen-evolving photosynthetic organisms are highlighted with a green background.

Figure 3. Phylogenetic tree constructed for all CISD, CISD1 and CISD2 proteins, with a single copy of the CDGSH domain per polypeptide chain, found in two different representative fully sequenced genomes for each of the branches of the taxonomy common tree of species presented in Fig. 2 (Supplementary Tables S2 and S3).

All protein sequences were obtained from the NCBI protein database using the PSI-BLAST algorithm with PSI threshold value 5 and e-value 10. The sequences were then aligned using the software MUSCLE. trimAL was employed to eliminate poorly aligned regions in the alignment. A maximum-likelihood phylogenetic tree with posterior probability support was then created using the PhyML program. The tree was finally edited with the software FigTree 1.4.0. Multiple sequence alignments and a version of the tree with complete protein annotations and posterior probabilities are included in Supplementary Figs S3 and S5, respectively.

Figure 4. Phylogenetic tree constructed for all CISD3 proteins with two copies of the CDGSH domain per polypeptide chain found in two different representative fully sequenced genomes for each of the branches of the taxonomy common tree of species presented in Fig. 2 (Supplementary Tables S2 and S3).

All protein sequences were obtained from the NCBI protein database using the PSI-BLAST algorithm with PSI threshold value 5 and e-value 10. The sequences were then aligned using the software MUSCLE and trimAL was used to delete regions with too many gaps. A maximum-likelihood phylogenetic tree with posterior probability support was then created using the PhyML program. The tree was finally edited with the software FigTree 1.4.0. Multiple sequence alignments and a version of the tree with complete protein annotations and posterior probabilities are included in Supplementary Figs S4 and S6, respectively.

Figure 5. Analysis of fungi CISD genes.

(a) Multiple sequence alignment of the CISD3 proteins from different fungi, bacteria and human generated using MUSCLE. Yellow boxes indicate the CDGSH domains. Bar graph under the aligned sequences indicates degree of conservation (%). Color legend: Background: White - Least conserved, Black - Most conserved; Font: Blue - Least conserved; Red - Most conserved. (b) Maximum-likelihood tree of CISD3 proteins of fungi, bacteria and human generated using PhyML. All the sequence have the same domain architecture as represented in the figure.