Controlled chaos of polymorphic mucins in a metazoan parasite (Schistosoma mansoni) interacting with its invertebrate host (Biomphalaria glabrata)

Abstract

Invertebrates were long thought to possess only a simple, effective and hence non-adaptive defence system against microbial and parasitic attacks. However, recent studies have shown that invertebrate immunity also relies on immune receptors that diversify (e.g. in echinoderms, insects and mollusks (Biomphalaria glabrata)). Apparently, individual or population-based polymorphism-generating mechanisms exists that permit the survival of invertebrate species exposed to parasites. Consequently, the generally accepted arms race hypothesis predicts that molecular diversity and polymorphism also exist in parasites of invertebrates. We investigated the diversity and polymorphism of parasite molecules (Schistosoma mansoni Polymorphic Mucins, SmPoMucs) that are key factors for the compatibility of schistosomes interacting with their host, the mollusc Biomphalaria glabrata. We have elucidated the complex cascade of mechanisms acting both at the genomic level and during expression that confer polymorphism to SmPoMuc. We show that SmPoMuc is coded by a multi-gene family whose members frequently recombine. We show that these genes are transcribed in an individual-specific manner, and that for each gene, multiple splice variants exist. Finally, we reveal the impact of this polymorphism on the SmPoMuc glycosylation status. Our data support the view that S. mansoni has evolved a complex hierarchical system that efficiently generates a high degree of polymorphism-a "controlled chaos"-based on a relatively low number of genes. This contrasts with protozoan parasites that generate antigenic variation from large sets of genes such as Trypanosoma cruzi, Trypanosoma brucei and Plasmodium falciparum. Our data support the view that the interaction between parasites and their invertebrate hosts are far more complex than previously thought. While most studies in this matter have focused on invertebrate host diversification, we clearly show that diversifying mechanisms also exist on the parasite side of the interaction. Our findings shed new light on how and why invertebrate immunity develops.

Figure 1. SmPoMuc polymorphism at the protein and transcript levels.

Positional differences between SmPoMuc from compatible (C) and incompatible (IC) strains on silver stained 2D-gels shown with a pH 3–10 non-linear (NL) gradient or a pH 3–6 linear (L) gradient (A). Positions of spots corresponding to SmPoMuc are indicated by arrows. Supplementary spots found in the present study using the pH 3–6 linear gradient are indicated by dotted arrows. (B) shows the precursor structure and polymorphism of SmPoMuc described in a previous study . Three kinds of repeats were identified in SmPoMuc cDNAs (r1, r1' and r2); the fourth repeat r3 was only identified at the genomic level only in this study. (C) Agarose gel separation of RT-PCR amplicons obtained from 11 individual sporocysts (1–11) of both strains (compatible: C and incompatible: IC). Amplification was performed using consensus primers amplifying the complete coding sequence of all SmPoMuc. C-: negative control of amplification.

Figure 2. Southern blot of S. mansoni genomic DNA and BACs containing SmPoMuc genes.

Southern Blot of adult worm genomic DNA from IC (lanes 1, 2 and 3) and C (lanes 4, 5 and 6) strains and BAC clones 45D24 (lanes 7, 13, 19), 47P6 (lanes 8, 14, 20), 51E8 (lanes 9, 15, 21), 62J10 (lanes 10, 16, 22), 41B11 (lanes 11, 17, 23), 62F12 (lanes 12, 18, 24). Genomic DNA from both strains is undigested (lanes 1 and 4), digested by EcoRV (lanes 2 and 5) or digested by BclI (lanes 3 and 6). All genomic DNA lanes were hybridized with a DIG labelled probe corresponding to the 1 kbp genomic repeat shared by all SmPoMuc genes. BAC DNAs were digested with EcoRV. Lanes 7 to 12, lanes 13 to 18 and lanes 19 to 24 correspond to the same blots hybridized successively with the 1 kbp genomic repeat, r1 and r2 probes, respectively. The membrane was stripped between two successive hybridization procedures.

Figure 3. Schematic representation of a complete SmPoMuc gene.

The complete SmPoMuc genes are composed of 15 exons. Exon 2 is included in a genomic repeat that can be repeated several times (a maximum of 20 repeats in SmPoMuc 2 genes). These genomic repeats of approximately 1 kilobase are separated by imperfect polypurine tracts (PPT). Positions of genomic primers used for SmPoMuc gene amplification (Intron2/3F1 – Exon15R) are indicated by arrows. PCR amplicon position used for gene copy number quantification is indicated by a bold line (–) and the position of a ribozyme between exon 9 and 10 is indicated by an asterisk. Triangles and chevrons indicate complementary sequence positions (12 and 13 nucleotides, respectively) identified in introns of the genomic repeats containing exon 2.

Figure 4. The SmPoMuc multigene family is organized in four paralogous groups that frequently recombine.

SmPoMuc genomic DNA sequences corresponding to the 3′ portion of SmPoMuc genes/alleles (exon 2/exon 15) were obtained by long range PCR and aligned to construct a cladogram with PAUP. Tree branches corresponding to C and IC strains are in red and black, respectively. SmPoMuc genes are identified as follows: first the strain (C or IC), then the last exon 2 (r1, r1' or r2) and finally the group (1, 2, or 4) or sub-group (3.1a, 3.1b, 3.2, 3.4, 3.5). This analysis reveals four paralogous sequence groups (gr.1–gr.4). In the right-hand part of the figure, a schematic representation of aligned SmPoMuc genomic sequences is given. We annotated the sequences by a color code that uses a different color for sequence fragments of less than 95% identity: gr.1 (red), gr.2 (blue), sub-gr.3.1a (purple), sub-gr.3.1b (pink), sub-gr.3.2 (sky-blue), sub-gr.3.3 (dark-green), sub-gr.3.4 (green) and gr.4 (yellow). Traces of retrotransposon insertion events (solo-LTR) are present in sub-gr.3.4 and gr.2. Large gaps necessary to obtain alignments are represented by dark lines. Short gap (<28 nucleotides) positions are indicated by rhombi. Short tandem repeats are indicated by (>). Frequent recombination events between SmPoMuc family members are apparent.

Figure 5. SmPoMucs contain a putative full-length hammerhead ribozyme between exon 9 and 10.

Alignment of putative ribozymes found in all SmPoMuc genes with a functional hammerhead ribozyme of S. mansoni (Sm5, AF036742). Asterisks indicate conserved positions in the alignment. Boxes A and B delimit sequences necessary for transcription by RNA polymerase III. The catalytic core nucleotides composed of domains I, II and III are underlined. The conserved nucleotides are numbered using the standard convention . The nucleotide position corresponding to G12 essential for ribozyme activity is indicated by a dotted arrow. The scissile bond is indicated by an arrow.

Figure 6. K_S/K_N comparison of SmPoMuc coding sequences.

The analysis was performed using SNAP (see Material and Methods) on 15, 71 and 56 sequences from groups 1, 2 and 3 respectively. The closed rhombi, open triangles and dashed lines are used for a pair of SmPoMuc sequences from groups 1, 2 and 3, respectively. The bisecting dotted line corresponds to K_S/K_N = 1.

Figure 7. Intermingled repeats (r1/r2) are present in C and IC genomic DNA but not in BACs.

PCR experiments were performed on BACs 45D24 – 47P6 – 51E8 – 62J10 – 41B11 – 62F12 (lanes 1 to 6, respectively) and on DNA from C and IC strains (lanes 7 and 8, respectively); lane 9 corresponds to the PCR negative control. Amplicons were separated on TAE 1% agarose gels and revealed by ethidium bromide staining. The primers used reveal two r2 exons (A), two r1 exons (B), r2r1 exons (C) or r1r2 exons (D) in two successive genomic repeats.

Figure 8. FISH mapping of SmPoMuc BACs clones.

Metaphase chromosome spreads showing positive signals (arrowheads) hybridized with biotinylated SmPoMuc BAC clone DNAs. (a) BAC clone 41B11 gave strong signals in the regions near the centromere of chromosome 3 and on the long arm of chromosome 4; two weaker signals were also detected on the short and on the long arm of chromosome 3. (b) BAC clone 45D24 hybridized to the same regions on chromosome 3 and 4, and yielded a strong supplementary signal at the large heterochromatic pericentromeric region of chromosome 2. This last signal is probably due to repetitive sequences in this BAC and not to the presence of SmPoMuc genes.

Figure 9. Aberrant splicing events during SmPoMuc gene expression.

The six aberrant splicing variants (AbS) obtained at the cDNA level are shown and numbered (AbS 1 to 6). The SmPoMuc genomic sequence areas subject to aberrant splicing are shown: introns are represented as thick lines and rectangles represent exons numbered as described in Figure 3. Normal splice donor and acceptor sites are in uppercase above the schematic gene representation. Aberrant splice sites are in uppercase, underlined, numbered (in brackets) below the schematic representation of SmPoMuc genes or in AbS 1–6. The different splicing events leading to them are indicated by dotted or full lines linking the different splice sites. These events are identified by circled numbers corresponding to the AbS they produce. Resulting aberrant splicing leads to exclusion (1-2-3) or inclusion (4-5-6) of DNA, leading to frame-shifts that create non-sense codons in all cases. The different aberrant splicing events observed correspond to cDNA variants given in Table S1: AbS 1 (individuals C3/2, IC/2-4-6-7-8-9-11/2); AbS 2 (individual IC11/2/7r2); AbS 3 (individual IC2/2/4r2); AbS 4 (individual IC5/3.1/10r1); AbS 5 (individual IC5/3.1/11r1) and AbS 6 (individual IC5/3.1/12r1).

Figure 10. Western blot of SmPoMuc proteins from C and IC strain before and after deglycosylation.

S. mansoni sporocyst extracts from C (lanes 1-2) and IC (lanes 3-4) strains were treated with TFMSA (lanes 2–4) or not (lanes 1–3) and submitted to a western blotting using anti-SmPoMuc antibodies. The shift in molecular weight observed in lanes 2 and 4 is related to the loss of carbohydrate chains associated with SmPoMuc proteins.

Figure 11. Controlled chaos of SmPoMuc polymorphism.

SmPoMuc polymorphism is controlled at the genomic (A), transcript (B), protein (C) and population (D) levels.