Sub-grouping and sub-functionalization of the RIFIN multi-copy protein family

Abstract

Background: Parasitic protozoans possess many multicopy gene families which have central roles in parasite survival and virulence. The number and variability of members of these gene families often make it difficult to predict possible functions of the encoded proteins. The families of extra-cellular proteins that are exposed to a host immune response have been driven via immune selection to become antigenically variant, and thereby avoid immune recognition while maintaining protein function to establish a chronic infection.

Results: We have combined phylogenetic and function shift analyses to study the evolution of the RIFIN proteins, which are antigenically variant and are encoded by the largest multicopy gene family in Plasmodium falciparum. We show that this family can be subdivided into two major groups that we named A- and B-RIFIN proteins. This suggested sub-grouping is supported by a recently published study that showed that, despite the presence of the Plasmodium export (PEXEL) motif in all RIFIN variants, proteins from each group have different cellular localizations during the intraerythrocytic life cycle of the parasite. In the present study we show that function shift analysis, a novel technique to predict functional divergence between sub-groups of a protein family, indicates that RIFINs have undergone neo- or sub-functionalization.

Conclusion: These results question the general trend of clustering large antigenically variant protein groups into homogenous families. Assigning functions to protein families requires their subdivision into meaningful groups such as we have shown for the RIFIN protein family. Using phylogenetic and function shift analysis methods, we identify new directions for the investigation of this broad and complex group of proteins.

Figure 1

RIFIN proteins overview. (A) Alignment of a selection of A and B-type RIFINs. Conserved cysteines are highlighted in red; shading according to conservation. (B) Schematic RIFIN sub-group characteristics: Overall domain organization and classification into subtypes. The 25 amino acid stretch present only in the semi-conserved domain of A-type RIFINs is highlighted and depicted by a sequence logo. Grey arrows: common conserved cysteine residues; black arrows: sub-type specific cysteine residues; SP: signal peptide; PEXEL: Plasmodium export element; C1: semi-conserved domain, including the 25 AA insertion/deletion; C2: C-terminal conserved domain; TM1 and TM2: previously predicted transmembrane regions; V1: first variable domain; V2: second variable domain.

Figure 2

Phylogenetic tree of rif cDNA. The tree shows the segregation of A- and B-rif genes (gaps considered as complete deletions). The B-rif group is further subdivided into B1, B2 and B3 clusters. Stars indicate sequences that group atypically. Bootstrap support, after 1000 replicates, is only shown for the branches separating the different groups, dots at nodes indicate bootstrap values above or equal to 60%.

Figure 3

Non-congruence of phylogenetic trees of RIFIN conserved (C1) versus variable (V2) domains. (A) Neighbor Joining tree of the C1 domain (gaps considered as pairwise deletions) showing the segregation of A- from B-RIFIN sequences. (B) The same tree construction method applied to the V2 domain showing that B3-RIFIN sequences do not cluster with B1- and B2-sequences. Bootstrap support, after 1000 replicates, is shown for values above 50%.

Figure 4

Function shift analysis of A- and B- RIFIN proteins. (A) Sample sequences from the high stringency global alignment available as Additional file 4. Columns with Orange-Blue represent RSS; columns with yellow-green represent CSS; columns with Salmon-green represent both RSS and CSS. (B) Plots of Z-scores and U-values, for CSS (red curve) and RSS (blue curve) respectively, according to alignment position. The predicted consensus secondary structure is plotted with pink and green bars representing helices and loops, respectively. The heights of the bars indicate conserved predictions. Arrows correlate the highest scoring shifted sites with secondary structure predictions.