Sympatric and allopatric evolutionary contexts shape differential immune response in 2 Biomphalaria / Schistosoma interaction

Selective pressures between hosts and their parasites can result in reciprocal evolution or adaptation of specific life history traits. Local adaptation of resident hosts and parasites should lead to host-parasite systems performing better in sympatry when compared to allopatry. Between-population variations in parasite infectivity/virulence and host defence/resistance, referred to as compatibility phenotype, were often the proxy used to analyse sympatric or allopatric adaptation. Nevertheless, some reported cases exist where allopatric host-parasite systems demonstrate compatibility phenotypes similar or greater than the one observed in sympatry. In these cases, the role of local adaptation is worth considering. Here, we study the interaction between Schistosoma and its vector snail Biomphalaria in which such a discrepancy in local versus foreign compatibility phenotype has been observed. Herein, we developed an integrative approach to investigate sympatric and allopatric interaction processes and link the underlying molecular mechanisms to the resulting phenotypes. Using comparative “omics” approaches joined to analysis of life history traits (immune cellular response, mortality, prevalence and compatibility) we tried to bridge the gap of knowledge that exists for connecting local adaptation observations to molecular phenotypes in Schistosoma/Biomphalaria interactions. We found that despite displaying similar prevalence phenotypes, parasite infection triggered an immune suppression in snails living in sympatry, while it activated an immune response for those living in allopatry. Dual-comparative molecular analyses revealed that parasite infection causes immune suppression in sympatry. miRNAs were used to hijack the host’s immune response, allowing sympatric parasites to initiate their developmental program earlier and more efficiently. We show that despite having similar prevalence phenotypes, sympatric and allopatric snail-Schistosoma interactions displayed a strongly different immunobiological molecular dialogue. The ability of allopatric pathogens to adapt rapidly and efficiently to new hosts could have critical consequences on disease emergence and risk of schistosomiasis outbreaks. These observations would have important consequences in term of schistosomiasis disease control.


116
requirements of our country. The experimenter possesses an official certificate for animal 117 experimentation from both of the above-listed French ministries (Decree # 87-848, October 19, 1987).

118
The various protocols used in this study have been approved by the French veterinary agency of the 119 DRAAF Languedoc-Roussillon (Direction Régionale de l'Alimentation, de l'Agriculture et de la    High-quality reads (Phred score >29) were aligned to the de novo transcriptome using Bowtie2 159 (v2.0.2), which was run locally on a Galaxy server. To compare the host responses during the 160 sympatric or allopatric interactions, we used the DESeq2 (v2.12) software to analyse the differential 161 transcript representation between BgBRE control strains (uninfected BgBRE1 and BgBRE2) to the 162 sympatric and allopatric conditions (p-value < 0.1) (48). A Venn diagram was generated using the 163 7 Venny 2.1 software to highlight which differentially expressed transcripts were specific or common to 164 the different interactions. A heatmap was then created to analyse transcript expression patterns, using 165 log2 Fold Change with Hierarchical Ascending Clustering (HAC) and Pearson correlation 166 (uncentered) as applied by the Cluster (v3.0) and Java TreeView (v1.1.6r4) software packages. The 167 differentially represented transcripts were functionally classified using a BlastX analysis with the cut-168 off set to e-value < 1e -3 (NCBI dataset; thanks to the Roscoff Data center Cluster, UPMC) and gene 169 ontology was assigned using an automatic annotation, Blast2GO (v3.0.8

229
The hemolymph was pooled from the three snails, and 100 µl were subjected to analysis with the

236
For each sample, 10,000 events were counted. The results were analysed with the FlowJo V 10.0.8 237 software. Between-group differences in the percent of proliferation were tested using the Mann-  were generated using the xlstats Mac software and the log-rank test was applied with significance 292 accepted at p < 0.05.

295
A whole-snail transcriptomic approach for investigating the molecular basis of the innate immune 296 response in the sympatric and allopatric contexts

297
The global transcript representation was analysed using the previously described RNAseq pipeline 298 developed in our laboratory (6,47,48). The BgBRE snail host had the same prevalence and intensity 299 phenotypes when exposed to SmBRE, SmVEN or Srod parasites (see Supplementary Table 1

313
Of the selected transcripts, 40 were shared by the BB, BV and BR interactions (

342
Among the transcripts distributed into these groups, 89.0% were underrepresented. In the BV 343 allopatric interaction, the three groups corresponded respectively to 10.8%, 12.9% and 8.6% and 344 among them 65.9% were underrepresented. Finally, in the BR allopatric interaction, they 345 corresponded to 12%, 12.5% and 7.8% and 62.5% of them were underrepresented. When we 346 compared the percentage of each immunological group in the sympatric and allopatric interactions, no 347 specific functional subset was particularly repressed in the BB sympatric interaction ( Fig. 1C; Fig. 2).

348
In fact, most of the time, the same immune functions were affected in sympatric and allopatric 349 infections but different immune transcripts (grey and black diamond in Fig. 2) showed differential 350 regulation following infections (Fig. 2).

351
The differentially regulated transcripts belonging to the three immunological groups (

357
Immune cellular responses in the sympatric and allopatric contexts

358
As the transcriptomic approach indicated that multiple immunobiological functions seem to be altered 359 in the host response during the interaction with the parasites, we next investigated whether differences 360 in immune responses in sympatric versus allopatric infections could be seen at the cellular level.

361
Hemocytes, the snail immune cells, participate directly in the immune response against the parasites,

364
Thus, to investigate such immune cell activation, cell proliferation was quantified using EdU 365 nuclear labelling (Fig. 3). EdU is a nucleoside analogue of thymine incorporated into DNA during 366 DNA synthesis. This incorporation reflects the mitotic activity of hemocytes. As the two tested

376
Hemocyte proliferation was assessed using flow cytometry after in vivo EdU-labelling ( Fig

395
In addition to the mitotic activity, hemocyte morphology was estimated using flow cytometry  Table 3). According to 439 the global Gene Ontology (GO): 70% of the annotated genes were involved in general metabolism and 440 growth, translation processes, regulation of cellular processes and RNA biosynthesis; 25% were 441 involved in molecular transport or cell organization; and 5% were involved in organism defence or 442 response to stimuli. In all these clusters, we identified six parasite gene products that had been 443 involved in parasite modulation or suppression of snail immunity. These molecules correspond to heat 444 shock proteins (Fig. 6, Clusters 1 and 2) (27); glutathione-S-transferase, NADH dehydrogenase 445 subunit, and calreticulin (Fig. 6, Cluster 2) (20, 53, 54); Alpha-2-macroglobulin (Fig. 6, Cluster 4)

447
Supplementary Table 3). These particular genes could be involved in the strategy used by the parasite 448 to ensure immunosuppression and favour its development and growth within host tissues.

449
In allopatric interactions, an activation of the immune response is observed. Interestingly, 450 allopatric parasites did not overexpress any transcripts that could have immunosuppressive function or 451 impair the activation of the immune response ( Fig. 6 and Supplementary   (Fig. 1B, Fig. 7).To explain the similar prevalence between sympatric and allopatric infections we 483 focus our attention on miRNAs that were shared between sympatric and allopatric interactions and we 484 identified one miRNA:sma-miR-190-3p ( Fig. 7 Fig. 2).

503
In the natural environment, it is assumed that the parasitic genes responsible for infectivity will evolve

20
The present RNAseq analysis demonstrated that in sympatric interaction (BB) a huge 527 immunosuppression occurs. Twenty-four hours after the infection, the three immunological processes 528 were downregulated: (i) immune recognition, (ii) effector and (iii) signalling pathways ( Fig. 1 and 2).

529
Conversely, in allopatric interactions (BV and BR), host immune response was activated ( Fig. 1 and   530 2). Differentially regulated transcripts mostly belong to immune cellular activation, migration, 531 proliferation or differentiation (Fig. 2). An EdU-labelling was used to detect proliferation and to 532 confirm that more hemocyte proliferation was observed in allopatry compared to sympatry (Fig. 3, 4A 533 and 4B). Using flow cytometry, we demonstrated that a new hemocyte subpopulation (named P2) was 534 observed exclusively following allopatric infection at 24 h ( Fig. 4C and 4D). P2 was EdU-positive and 535 characterized by an increase in granularity, indicating that the new P2 cell subtype could proliferate.

536
In Biomphalaria snails, we know three main hemocyte morphotypes: the Blast-like cells, the 537 type I hyalinocytes and the granulocytes (66). However, in absence of specific hemocyte markers, it is

548
The strong immunosuppression observed in sympatric interaction compared to the immune 549 cell activation in allopatry certainly results in differences in the capabilities of sympatric and allopatric 550 parasites to infect and grow in such hosts. This hypothesis was confirmed by our histological analysis 551 of sporocyst development in snail tissues. We observed a significant difference in sporocyst size 24 h 552 after infection (Fig. 5), when sympatric sporocysts were one-third bigger than allopatric sporocysts. At 553 96 h after infection, in contrast, there was no size difference between sympatric and allopatric 554 parasites (Fig. 5). This indicates that the allopatric parasites seem to overcome this early delay in 555 development. However, the consequences of this delay for duration of cercariae development, number 556 of cercariae produced, cercariae infectivity and pathogenicity in vertebrate host, will deserves further 557 investigation to determine the potential fitness cost between sympatric and allopatric parasites.

558
To go further and understand how parasites immunosuppress the host or circumvent the host 559 immune system, we used a dual-RNAseq approach to investigate transcripts expression of the 560 sympatric and allopatric parasite intra-molluscal stages (Fig. 6). As the histological differences were

564
We also identified some transcripts with GO terms or functions related to immunomodulation or 565 immunosuppression ( Fig. 6 and Supplementary Table 4). Nearly all of the identified transcripts were 566 overrepresented in the sympatric interaction compared to the allopatric interactions. Our results 567 therefore suggest that the installation, development and growth of the parasite occurred much more 568 rapidly in the sympatric combination, as sympatric parasites seemed to interfere more efficiently with 569 the host immune system. However, RNAseq data did not give any clear information about how 570 allopatric parasites succeed in circumventing the host immune system. We thus next examined the 571 generated dual-RNAseq libraries in an effort to identify whether sympatric and/or allopatric 572 schistosomes could hijack the host immune system using microRNAs. h after infection, several differentially expressed parasite miRNAs were identified. We investigated 580 the potential role played by these miRNAs on host immune response. We identified some predicted 581 targets of such schistosome miRNAs in the Biomphalaria immune reference transcriptome. We found 582 that schistosome miRNAs identified in the snails 24 h after sympatric infection may target 43.5% of 583 the differentially regulated immune transcripts (Fig. 7). In contrast, far fewer correspondences were 584 identified for the allopatric interactions (Fig. 7). The higher proportion of targeted genes in the 585 sympatric interaction may result in the observed efficient immunosuppression, demonstrating a 586 specific co-evolution or adaptation in the transcriptional regulation between sympatric host and 587 parasite. However, even if more host immune genes appeared to be targeted in the sympatric 588 combination compared to the allopatric combination (Fig. 7), both sympatric and allopatric 589 interactions displayed the same phenotype of compatibility. In the allopatric condition, despite the 590 activation of the snail immune response, the parasite is able to develop and succeed to infect the host.

591
This similarity in compatibility phenotype between sympatric and allopatric parasites could potentially 592 results from their ability to target host immune weapons or host genes that regulate the snail innate 593 cellular response. In this context, we were expecting to find miRNAs shared between sympatric and 594 allopatric parasites, that is, miRNAs that tend to weaken the immune system. A unique miRNA was 595 found to be shared by allopatric and sympatric parasites, sma-miR-190-3p, which has been identified

607
Compatibility reflects the outcome of complex immunobiological interactions and depends on:

608
(i) the ability of the snail immune system to recognize and kill the parasite; and (ii) the ability of the 609 parasite to circumvent or evade the host immune response (20,53,75). Based on the present 610 observations, we propose that sympatric and allopatric interactions trigger totally different responses.

611
In the sympatric interaction, the parasite is able to induce a host immunosuppression within the first 612 day of infection, enabling it to infect quickly the host and readily begins its development. In the 613 allopatric interaction, the parasite is not able to neutralize quickly the host immune system, and 614 therefore the parasite is recognized by the host defence system, which mounts a potent immune 615 response. The need for allopatric parasites to resist the immune system seems to disrupt the activation 616 of their developmental program during the first day of infection. However, they seemed to be rapidly

668
If we hope to draw conclusions regarding the existence of emerging or outbreak risks, we need 669 to develop integrative approaches to explore fine-scale patterns of host-parasite interactions. We must 670 consider the spatial scale at which comparisons are conducted, the patterns of disease occurrence, the 671 population genetics, and the involvement of physiological, immunological and molecular processes.

672
Studying the relevant factors at the relevant timing would be of critical importance in terms of 673 schistosomiasis control. Understanding further, how these allopatric parasites efficiently infect host 674 snails would be mandatory to identify markers and develop new tools to predict or to quantify risks of 675 schistosomiasis outbreaks.