Species-Specific Duplication of Surface Antigen Genes in Paramecium

Abstract

Paramecium is a free-living ciliate that undergoes antigenic variation and still the functions of these variable surface antigen coats in this non-pathogenic ciliate remain elusive. Only a few surface antigen genes have been described, mainly in the two model species P. tetraurelia strain 51 and P. primaurelia strain 156. Given the lack of suitable sequence data to allow for phylogenetics and deeper sequence comparisons, we screened the genomes of six different Paramecium species for serotype genes and isolated 548 candidates. Our approach identified the subfamilies of the isogenes of individual serotypes that were mostly represented by intrachromosomal gene duplicates. These showed different duplication levels, and chromosome synteny suggested rather young duplication events after the emergence of the P. aurelia species complex, indicating a rapid evolution of surface antigen genes. We were able to identify the different subfamilies of the surface antigen genes with internal tandem repeats, which showed consensus motifs across species. The individual isogene families showed additional consensus motifs, indicating that the selection pressure holds individual amino acids constant in these repeats. This may be a hint of the receptor function of these antigens rather than a presentation of random epitopes, generating the variability of these surface molecules.

Keywords: Paramecium; antigenic variation; multigene family; surface antigen.

Figure A1

Neighbor-joining tree of group 5 with the SAg 51B, the B isogene family, and SAgs 51G, 51I, 51C, and 51A. The known genes for P. primaurelia (156S, 168G, and 156G) are spiked into the alignment to identify the orthologs in P. primaurelia strain AZ9-3. Identified proteins that correspond to known SAgs in P. tetraurelia are indicated (51B isogene family, 51A, and 51G). P. primaurelia is colored in yellow, P. biaurelia in pink; P. tetraurelia in blue, P. sexaurelia in red, and P. caudatum in black.

Figure 1

(A) Scheme of the evolutionary distance between the studied Paramecium species inferred from the mitochondrial COI gene sequence. The scale bar corresponds to 0.5 substitutions per site. (B) Cladogram of the neighbor-joining tree of the 548 identified putative SAgs. Scale bar represents substitutions per site.

Figure 2

Identification of recurring motifs in SAgs. Three motifs were identified, each with more than 1000 sites in the entire dataset and with E-values of $4\times {10}^{-1909}$ (red), $1.8\times {10}^{-1868}$ (green), and $1.6\times {10}^{-1299}$ (blue). The location of the motif sites is shown as exemplary for some SAgs involving D-, G-, B-, and S-SAgs from different species and also some unknown SAgs inferred from the genome data. Each block shows the position of a motif; the height of a block indicates significance—taller blocks are more significant, being proportional to the negative log of the positional p-value of each motif but truncated for a p-value of $1\times {10}^{-10}$. All motifs have p-values below that threshold. Between the accession numbers and the scheme, the combined p-value match for the entire SAg is given.

Figure 3

Characteristics of the individual SAg groups. The first column shows the length distribution of SAgs in amino acids. The bins are indicated at the bottom; the Y-axes show the percentages of SAgs that fall into one bin. The pie charts indicate the composition of each group in terms of species composition and membrane anchoring. The decision on transmembrane anchoring was the prediction of at least two transmembrane domains. GPI-anchoring prediction relies on the prediction of the C-terminal GPI-anchoring signal. The last columns refer to the analysis of the internal repeat motif (red motif) in the individual groups. SEA was used to analyze the enrichment of this motif in a dataset. The column “Consensus Repeat” indicates the percentage of primary sequences matching the motif. The p-value shows the optimal enrichment p-value of the motif but it is not adjusted for the number of motifs. The positional distribution of the motifs in the linear polypeptide sequence is shown in the respective column and the last column shows the number of matches per sequence as a percentage of the total sequences of a group.

Figure 4

Neighbor-joining tree of cluster 3, which contains the D and H surface antigens. The known genes for P. primaurelia (156D, 156D beta, the putative 156H) are spiked into the alignment to identify the orthologs in the P. primaurelia AZ9-3 strain. The identified proteins that correspond to known SAgs in P. tetraurelia are indicated (51D, 51J, 51D gamma1, 51D gamma2, 51H alpha, and 51H beta). P. primaurelia is colored in yellow, P. biaurelia in pink, P. tetraurelia in blue, P. sexaurelia in red, and P. caudatum in black. Scale bar represents substitutions per site; branch label indicates percentage consensus support of 100 pseudoreplicates.

Figure 5

A deeper analysis of cluster 5 containing the I, C, B, and G SAgs. (A,B) show subtrees of the cluster containing P. tetraurelia 51C and 51I SAgs (A) and the P. tetraurelia 51A, 51G, 51B family, and P. primaurelia G and S SAgs (B). (C) shows the motif of internal repeats for the genes in the subclusters in (A,B), and the positional distributions and motif frequencies in the I/C cluster and the B/G cluster. (D,E) show the motif of the internal repeats dissecting the G (D) from the B cluster (E). The motif is shown for the internal repeats only, which are indicated in blue for the G cluster and in green for the B cluster. Please note that the colors are used differently for the motif blocks in (D,E). Below the motif, the distribution is shown for the individual SAgs. In all three motifs in this figure, the common CTXNXXGTAC motif is indicated by the black arrows. The G-cluster-specific YTGTGLT motif is highlighted by the grey arrow.

Figure 6

Synteny between scaffolds ofP. primaureliaandP. tetraureliaharboring D isogenes (upper part) and B isogenes (lower part). The scaffolds are visualized by the individual rulers giving the length position in bp. The similarities among the scaffolds are shown by alignment scores within the highlighted boxes (green/teal). The positions of the identified SAg genes are shown by grey arrows. Putative orthologs are included according to the clustering of the SAgs of strain 156 in the neighbor-joining tree. Scale bar represents substitutions per site; branch label indicates the percentage consensus support of 100 pseudoreplicates.