Can selection for resistance to OsHV-1 infection modify susceptibility to Vibrio aestuarianus infection in Crassostrea gigas? First insights from experimental challenges using primary and successive exposures

Epigenetic variations are more substantial than genetic variations in rapid adaptation of oyster to Pacific oyster mortality syndrome

Disease emergence is accelerating with global changes. Understanding by which mechanisms host populations can rapidly adapt will be crucial for management practices. Pacific oyster mortality syndrome (POMS) imposes a substantial and recurrent selective pressure on oyster populations, and rapid adaptation may arise through genetics and epigenetics. In this study, we used (epi)genome-wide association mapping to show that oysters differentially exposed to POMS displayed genetic and epigenetic signatures of selection. Consistent with higher resistance to POMS, the genes targeted included many genes in several pathways related to immunity. By combining correlation, DNA methylation quantitative trait loci, and variance partitioning, we revealed that a third of phenotypic variation was explained by interactions between the genetic and epigenetic information, ~14% by the genome, and up to 25% by the epigenome alone. Similar to genetically based adaptation, epigenetic mechanisms notably governing immune responses can contribute substantially to the rapid adaptation of hosts to emerging infectious diseases.

Susceptibility variation to the main pathogens of Crassostrea gigas at the larval, spat and juvenile stages using unselected and selected oysters to OsHV-1 and/or V. aestuarianus

French commercial hatcheries are massively producing Crassostrea gigas selected for their higher resistance to OsHV-1, and soon should also implement selection for increasing resistance to Vibrio aestuarianus. The first objective of this study was to optimize the breeding programs for dual resistance to OsHV-1 and V. aestuarianus to determine the earliest life stage for which oysters are able to develop disease resistance. Wild stocks and selected families were tested using experimental infections by both pathogens at the larval, spat and juvenile stages. Oyster families could be evaluated for OsHV-1 as soon as the larval stage by a bath method, but this only highlighted the most resistant families; those that showed the highest resistance to V. aestuarianus could be determined using the cohabitation method at the juvenile stage. The second objective of this study was to determine if selection to increase/decrease the resistance to OsHV-1 and V. aestuarianus could have an impact on other major pathogens currently detected in hatchery at the larval stage, and in nursery and field at the spat/juveniles stages (V. coralliilyticus, V. crassostreae, V. tasmaniensis, V. neptunius, V. europaeus, V. harveyi, V. chagasi). No relationship was found between mortality caused by V. aestuarianus/OsHV-1 and the mortality caused by the other virulent bacterial strains tested regardless the stages, except between OsHV-1 and V. tasmaniensis at the juvenile stage. Finally, miscellaneous findings were evidenced such as (1) bath for bacterial challenges was not adapted for spat, (2) the main pathogens at the larval stage were OsHV-1 and V. coralliilyticus using bath, while it was V. coralliilyticus, V. europaeus, and V. neptunius at the juvenile stage by injection, and (4) variation in mortality was observed among families/wild controls for all pathogens at larval and juvenile stages, except for V. harveyi for larvae.

Comparative Proteomics of Ostreid Herpesvirus 1 and Pacific Oyster Interactions With Two Families Exhibiting Contrasted Susceptibility to Viral Infection

Massive mortality outbreaks affecting Pacific oysters (Crassostrea gigas) spat/juveniles are often associated with the detection of a herpesvirus called ostreid herpesvirus type 1 (OsHV-1). In this work, experimental infection trials of C. gigas spat with OsHV-1 were conducted using two contrasted Pacific oyster families for their susceptibility to viral infection. Live oysters were sampled at 12, 26, and 144 h post infection (hpi) to analyze host-pathogen interactions using comparative proteomics. Shotgun proteomics allowed the detection of seven viral proteins in infected oysters, some of them with potential immunomodulatoy functions. Viral proteins were mainly detected in susceptible oysters sampled at 26 hpi, which correlates with the mortality and viral load observed in this oyster family. Concerning the Pacific oyster proteome, more than 3,000 proteins were identified and contrasted proteomic responses were observed between infected A- and P-oysters, sampled at different post-injection times. Gene ontology (GO) and KEGG pathway enrichment analysis performed on significantly modulated proteins uncover the main immune processes (such as RNA interference, interferon-like pathway, antioxidant defense) which contribute to the defense and resistance of Pacific oysters to viral infection. In the more susceptible Pacific oysters, results suggest that OsHV-1 manipulate the molecular machinery of host immune response, in particular the autophagy system. This immunomodulation may lead to weakening and consecutively triggering death of Pacific oysters. The identification of several highly modulated and defense-related Pacific oyster proteins from the most resistant oysters supports the crucial role played by the innate immune system against OsHV-1 and the viral infection. Our results confirm the implication of proteins involved in an interferon-like pathway for efficient antiviral defenses and suggest that proteins involved in RNA interference process prevent viral replication in C. gigas. Overall, this study shows the interest of multi-omic approaches applied on groups of animals with differing sensitivities and provides novel insight into the interaction between Pacific oyster and OsHV-1 with key proteins involved in viral infection resistance.

Genetic variability in Ruditapes decussatus clam combined with Perkinsus infection level to support founder population selection for a breeding program

Clam farmers worldwide face several challenges, including irregular seed supply and high mortalities due to pathogenic organisms such as Perkinsus olseni. In Europe, there is a high unmet consumer demand for native clam species such as Ruditapes decussatus. The high market value of R. decussatus makes the culture of this species potentially more attractive than that culture of the alien species Ruditapes philippinarum. Thus, there is a market opportunity in breeding and producing R. decussatus at an industrial scale. A selective breeding program to improve R. decussatus performance will be carried out in Portugal; and the first critical step to develop such a breeding program is the establishment of a founder population. In this study, intra- and interpopulation genetic diversity was assessed using 13 microsatellite markers in eight natural beds located in Portugal, Spain and Italy. Also, allele and genotypic frequencies of each microsatellite locus were assessed discriminating between clams infected and non-infected by P. olseni. All locations showed similar values for several genetic diversity parameters. Analyses of population differentiation (F_ST, Bayesian clustering and AMOVAs) revealed five genetically differentiated regions: Rías Altas and Rías Baixas (NW Spain), North/Central Coast of Portugal, Gulf of Cadiz and Adriatic Sea. Significant differences in the allelic and genotypic frequency distribution between infected clams and non-infected ones at four microsatellite loci are reported suggesting that resistance to the disease could have a genetic basis. Moreover, a positive or negative relationship between the frequency of certain alleles and the parasite infection was inferred. Further studies should confirm the potential use of those alleles as genetic markers for P. olseni infection. Integrating results of genetic diversity within and between populations and Perkinsus infection levels, a founder population for a R. decussatus breeding program is proposed, composed by individuals from Barallobre (Rías Altas), Pontevedra or Cangas (Rías Baixas), Óbidos (North/Central Coast of Portugal), Algarve (Gulf of Cadiz) and Venice (Adriatic Sea).

Removal of pathogens by ultrafiltration from sea water

Among water treatment processes, ultrafiltration is known to be efficient for the elimination of micro-organisms (bacteria and viruses). In this study, two pathogens were targeted, a bacterium, Vibrio aestuarianus and a virus, OsHV-1, with the objective to produce high quality water from seawater, in the case of shellfish productions. The retention of those microorganisms by ultrafiltration was evaluated at labscale. In the case of OsHV-1, the protection of oysters was validated by in vivo experiments using oysters spat and larvae, both stages being highly susceptible to the virus. The oysters raised using contaminated seawater which was then subsequently treated by ultrafiltration, had similar mortality to the negative controls. In the case of V. aestuarianus, ultrafiltration allowed a high retention of the bacteria in seawater with concentrations below the detection limits of the 3 analytical methods (flow cytometry, direct seeding and seeding after filtration to 0.22 µm). Thus, the quantity of V. aestuarianus was at least, 400 times inferior to the threshold known to induce mortalities in oysters. Industrial scale experiment on a several months period confirmed the conclusion obtained at lab scale on the Vibrio bacteria retention. Indeed, no bacteria from this genus, potentially harmful for oysters, was detected in permeate and this, whatever the quality of the seawater treated and the bacteria concentration upstream of the membrane. Moreover, the resistance of the process was confirmed with a stability of hydraulic performances over time for two water qualities and even facing an algal bloom. In terms of retention and resistance, ultrafiltration process was validated for the treatment of seawater towards the targeted pathogenic microorganisms, with the aim of biosecuring shellfish productions.

Contribution of Viral Genomic Diversity to Oyster Susceptibility in the Pacific Oyster Mortality Syndrome

Juvenile Pacific oysters (Crassostrea gigas) are subjected to recurrent episodes of mass mortalities that constitute a threat for the oyster industry. This mortality syndrome named “Pacific Oyster Mortality Syndrome” (POMS) is a polymicrobial disease whose pathogenesis is initiated by a primary infection by a variant of an Ostreid herpes virus named OsHV-1 μVar. The characterization of the OsHV-1 genome during different disease outbreaks occurring in different geographic areas has revealed the existence of a genomic diversity for OsHV-1 μVar. However, the biological significance of this diversity is still poorly understood. To go further in understanding the consequences of OsHV-1 diversity on POMS, we challenged five biparental families of oysters to two different infectious environments on the French coasts (Atlantic and Mediterranean). We observed that the susceptibility to POMS can be different among families within the same environment but also for the same family between the two environments. Viral diversity analysis revealed that Atlantic and Mediterranean POMS are caused by two distinct viral populations. Moreover, we observed that different oyster families are infected by distinct viral populations within a same infectious environment. Altogether these results suggest that the co-evolutionary processes at play between OsHV-1 μVar and oyster populations have selected a viral diversity that could facilitate the infection process and the transmission in oyster populations. These new data must be taken into account in the development of novel selective breeding programs better adapted to the oyster culture environment.

Vibrio-bivalve interactions in health and disease

In the marine environment, bivalve mollusks constitute habitats for bacteria of the Vibrionaceae family. Vibrios belong to the microbiota of healthy oysters and mussels, which have the ability to concentrate bacteria in their tissues and body fluids, including the hemolymph. Remarkably, these important aquaculture species respond differently to infectious diseases. While oysters are the subject of recurrent mass mortalities at different life stages, mussels appear rather resistant to infections. Thus, Vibrio species are associated with the main diseases affecting the worldwide oyster production. Here, we review the current knowledge on Vibrio-bivalve interaction in oysters (Crassostrea sp.) and mussels (Mytilus sp.). We discuss the transient versus stable associations of vibrios with their bivalve hosts as well as technical issues limiting the monitoring of these bacteria in bivalve health and disease. Based on the current knowledge of oyster/mussel immunity and their interactions with Vibrio species pathogenic for oyster, we discuss how differences in immune effectors could contribute to the higher resistance of mussels to infections. Finally, we review the multiple strategies evolved by pathogenic vibrios to circumvent the potent immune defences of bivalves and how key virulence mechanisms could have been positively or negatively selected in the marine environment through interactions with predators.

Modeling the Transmission of Vibrio aestuarianus in Pacific Oysters Using Experimental Infection Data

Vibrio aestuarianus is a bacterium related to mortality outbreaks in Pacific oysters, Crassostrea gigas, in France, Ireland, and Scotland since 2011. Knowledge about its transmission dynamics is still lacking, impairing guidance to prevent and control the related disease spread. Mathematical modeling is a relevant approach to better understand the determinants of a disease and predict its dynamics in imperfectly observed pathosystems. We developed here the first marine epidemiological model to estimate the key parameters of V. aestuarianus infection at a local scale in a small and closed oyster population under controlled laboratory conditions. Using a compartmental model accounting for free-living bacteria in seawater, we predicted the infection dynamics using dedicated and model-driven collected laboratory experimental transmission data. We estimated parameters and showed that waterborne transmission of V. aestuarianus is possible under experimental conditions, with a basic reproduction number R₀ of 2.88 (95% CI: 1.86; 3.35), and a generation time of 5.5 days. Our results highlighted a bacterial dose–dependent transmission of vibriosis at local scale. Global sensitivity analyses indicated that the bacteria shedding rate, the concentration of bacteria in seawater that yields a 50% chance of catching the infection, and the initial bacterial exposure dose W₀ were three critical parameters explaining most of the variation in the selected model outputs related to disease spread, i.e., R₀, the maximum prevalence, oyster survival curve, and bacteria concentration in seawater. Prevention and control should target the exposure of oysters to bacterial concentration in seawater. This combined laboratory–modeling approach enabled us to maximize the use of information obtained through experiments. The identified key epidemiological parameters should be better refined by further dedicated laboratory experiments. These results revealed the importance of multidisciplinary approaches to gain consistent insights into the marine epidemiology of oyster diseases.

Variability in the Composition of Pacific Oyster Microbiomes Across Oyster Families Exhibiting Different Levels of Susceptibility to OsHV-1 μvar Disease

Oyster diseases are a major impediment to the profitability and growth of the oyster aquaculture industry. In recent years, geographically widespread outbreaks of disease caused by ostreid herpesvirus-1 microvariant (OsHV-1 μvar) have led to mass mortalities among Crassostrea gigas, the Pacific Oyster. Attempts to minimize the impact of this disease have been largely focused on breeding programs, and although these have shown some success in producing oyster families with reduced mortality, the mechanism(s) behind this protection is poorly understood. One possible factor is modification of the C. gigas microbiome. To explore how breeding for resistance to OsHV-1 μvar affects the oyster microbiome, we used 16S rRNA amplicon sequencing to characterize the bacterial communities associated with 35 C. gigas families, incorporating oysters with different levels of susceptibility to OsHV-1 μvar disease. The microbiomes of disease-susceptible families were significantly different to the microbiomes of disease-resistant families. OTUs assigned to the Photobacterium, Vibrio, Aliivibrio, Streptococcus, and Roseovarius genera were associated with low disease resistance. In partial support of this finding, qPCR identified a statistically significant increase of Vibrio-specific 16S rRNA gene copies in the low disease resistance families, possibly indicative of a reduced host immune response to these pathogens. In addition to these results, examination of the core microbiome revealed that each family possessed a small core community, with OTUs assigned to the Winogradskyella genus and the Bradyrhizobiaceae family consistent members across most disease-resistant families. This study examines patterns in the microbiome of oyster families exhibiting differing levels of OsHV-1 μvar disease resistance and reveals some key bacterial taxa that may provide a protective or detrimental role in OsHV-1 μvar disease outbreaks.

Revealing Mytilus galloprovincialis transcriptomic profiles during ontogeny

Mediterranean mussels are a worldwide spread bivalve species with extraordinary biological success. One of the reasons of this success could be the reproduction strategy of bivalves, characterized by the presence of trochophore larvae. Larval development in bivalves has been a topic of raising interest in the scientific community but it deserves much more attention. The principal objective of this work was to study the transcriptomic profile of the ontogeny of Mytilus galloprovincialis analyzing the gene expression in different developmental stages, from oocytes to juveniles. For this purpose, after conducting a 454 sequencing of the transcriptomes of mussel hemocytes, adult tissues and larvae, a new DNA microarray was designed and developed. The studied developmental stages: unfertilized oocytes, veliger, pediveliger, settled larvae and juveniles, showed very different transcriptomic profiles and clustered in groups defining their characteristic gene expression along ontogeny. Our results show that oocytes present a distinct and characteristic transcriptome. After metamorphosis, both settled larvae and juveniles showed a very similar transcriptome, with no enriched GO terms found between these two stages. This suggests: 1.- the progressive loss of RNA of maternal origin through larval development and 2.- the stabilization of the gene expression after settlement. On the other hand during metamorphosis a specific profile of differentially expressed genes was found. These genes were related to processes such as differentiation and biosynthesis. Processes related to the immune response were strongly down regulated. These suggest a development commitment at the expense of other non-essential functions, which are temporary set aside. Immune genes such as antimicrobial peptides suffer a decreased expression during metamorphosis. In fact, we found that the oocytes which express a higher quantity of genes such as myticins are more likely to reach success of the offspring, compared to oocytes poor in such mRNAs, whose progeny died before reaching metamorphosis.

Infection dynamics of a V. splendidus strain pathogenic to Mytilus edulis: In vivo and in vitro interactions with hemocytes

The pathogenic strain V. splendidus 10/068 1T1 has previously been reported for its virulence to the blue mussel and for its capacity to alter immune responses. In this study, we expanded the knowledge on hemocyte-pathogen interactions by using in vitro and in vivo assays. V. splendidus 10/068 1T1 severely inhibited cell adhesion and acidic vacuole formation unlike the innocuous phylogenetically related V. splendidus 12/056 M24T1 which had no effect on these cell functions. Furthermore, the virulent bacteria decreased hemocyte viability (59% of viability after 24 h). Infection dynamics were explored by using a model based on water tank cohabitation with septic mussels infected by GFP-tagged V. splendidus 10/068 1T1. Experimental infections were successfully produced (16.6% and 45% mortalities in 3 days and 6 days). The amount of GFP Vibrio in seawater decreased during the experiment suggesting its horizontal transfer from diseased animals to healthy ones. At the same time periods, bacteria were detected in hemocytes and in various organs and caused necrosis especially in gills. Total hemocyte count and viability were affected. Taken together, our results indicate that the pathogen V. splendidus 10/068 1T1 colonizes its host both by bypassing external defense barriers and impairing hemocyte defense activities.

Several strains, one disease: experimental investigation of Vibrio aestuarianus infection parameters in the Pacific oyster, Crassostrea gigas

This study investigated oyster infection dynamics by different strains of Vibrio aestuarianus isolated before and after the apparent re-emergence of this pathogen observed in France in 2011. We conducted experiments to compare minimal infective dose, lethal dose 50 and bacterial shedding for six V. aestuarianus strains. Whatever the strain used, mortality was induced in juvenile oysters by intramuscular injection and reached 90–100% of mortality within 5 days. Moreover, bacterial shedding was comparable among strains and reached its maximum after 20 h (≈10 EXP5 bacteria/mL/animal). Similarly, our first estimations of lethal dose 50 were comparable among strains (minimal infective dose around 0.4 × 10EXP5 bacteria/mL and LD50 around 10EXP5 bacteria/mL) by using seawater containing freshly shed bacteria. These results indicate that, at least with these criteria, despite V. aestuarianus strains genetic diversity, the disease process is similar. The strains isolated after the apparent re-emergence of the bacteria in 2011, do not present a more acute virulence phenotype than the reference strains isolated between 2002 and 2007. Finally, our study provides original and noteworthy data indicating that infected oysters shed bacteria at a level above the threshold of LD50 a few days before they die, meaning that infection is expected to spread in a susceptible population.

Genetic parameters of resistance to Vibrio aestuarianus, and OsHV-1 infections in the Pacific oyster, Crassostrea gigas, at three different life stages

Background

In France, two main diseases threaten Pacific oyster production. Since 2008, Crassostrea gigas spat have suffered massive losses due to the ostreid herpesvirus OsHV-1, and since 2012, significant mortalities in commercial-size adults have been related to infection by the bacterium Vibrio aestuarianus. The genetic basis for resistance to V. aestuarianus and OsHV-1 and the nature of the genetic correlation between these two traits were investigated by using 20 half-sib sire families, each containing two full-sib families. For each disease, controlled infectious challenges were conducted using naïve oysters that were 3 to 26 months old. In addition, siblings were tested under field, pond and raceway conditions to determine whether laboratory trials reflected mortality events that occur in the oyster industry.

Results

First, we estimated the genetic basis of resistance to V. aestuarianus in C. gigas. Susceptibility to the infection was low for oysters in spat stage but increased with later life stages. Second, we confirmed a strong genetic basis of resistance to OsHV-1 infection at early stages and demonstrated that it was also strong at later stages. Most families had increased resistance to OsHV-1 infection from the spat to adult stages, while others consistently showed low or high mortality rates related to OsHV-1 infection, regardless of the life stage. Our third main finding was the absence of genetic correlations between resistance to OsHV-1 infection and resistance to V. aestuarianus infection.

Conclusions

Selective breeding to enhance resistance to OsHV-1 infection could be achieved through selective breeding at early stages and would not affect resistance to V. aestuarianus infection. However, our results suggest that the potential to select for improved resistance to V. aestuarianus is lower. Selection for dual resistance to OsHV-1 and V. aestuarianus infection in C. gigas might reduce the impact of these two major diseases by selecting families that have the highest breeding values for resistance to both diseases.

Electronic supplementary material

The online version of this article (doi:10.1186/s12711-017-0297-2) contains supplementary material, which is available to authorized users.

Single or dual experimental infections with Vibrio aestuarianus and OsHV-1 in diploid and triploid Crassostrea gigas at the spat, juvenile and adult stages

French production of the Pacific cupped oyster, Crassostrea gigas, is currently threatened by two pathogens, OsHV-1 and V. aestuarianus. While oysters selected for their higher resistance to OsHV-1 are now available for the industry, the impact of V. aestuarianus on such oysters is unknown, especially for triploids. In addition, experimental infection has used the virus or the bacteria alone, but there have been no investigations of dual exposure to these pathogens. This study is the first report of single or dual exposure in spat (Spat1 and Spat2), juvenile and adult naïve oysters. For each of the two stocks evaluated, unselected oysters and oysters selected for their higher resistance to OsHV-1 infection were tested, as well as their triploid siblings of the selected oysters produced using cytochalasin B. We confirmed that resistance to OsHV-1 infection and susceptibility to V. aestuarianus increased with age and size, although selected oysters were not significantly impacted by OsHV-1 whatever their ploidy, size or age. We found different mortality patterns depending on the pathogen tested. The mortality pattern was similar for oysters exposed to OsHV-1 or to both pathogens in the Spat1 trial (4months old and 1.9g). The mortality pattern was similar for oysters exposed to V. aestuarianus or to both pathogens in the Adult trial (25months old and 63.1g). Surprisingly, mortality was much higher (ranging from 75.9% to 100%), in particular for the selected oysters, for the Spat2 (8months old/3.9g) and Juvenile trials (16months old/18.4g) given a dual exposure, regardless of the level of selection for OsHV-1 and the ploidy state. Our findings highlight an important threat for oyster farmers: oysters exposed to both pathogens could experience dramatic mortality rates, even in oysters selected for their higher resistance to OsHV-1. Finally, our study demonstrated for the first time that triploid oysters were more susceptible to experimental challenges with V. aestuarianus at the spat stage than their diploid siblings. However, the difference in mortality between the triploids and diploids remained limited and ranged from 22.9% to 6.6% for spat and adults, respectively with a relatively regularly decrease in the difference with increased age.

Distinct immune responses of juvenile and adult oysters (Crassostrea gigas) to viral and bacterial infections

Since 2008, massive mortality events of Pacific oysters (Crassostrea gigas) have been reported worldwide and these disease events are often associated with Ostreid herpesvirus type 1 (OsHV-1). Epidemiological field studies have also reported oyster age and other pathogens of the Vibrio genus are contributing factors to this syndrome. We undertook a controlled laboratory experiment to simultaneously investigate survival and immunological response of juvenile and adult C. gigas at different time-points post-infection with OsHV-1, Vibrio tasmaniensis LGP32 and V. aestuarianus. Our data corroborates epidemiological studies that juveniles are more susceptible to OsHV-1, whereas adults are more susceptible to Vibrio. We measured the expression of 102 immune-genes by high-throughput RT-qPCR, which revealed oysters have different transcriptional responses to OsHV-1 and Vibrio. The transcriptional response in the early stages of OsHV-1 infection involved genes related to apoptosis and the interferon-pathway. Transcriptional response to Vibrio infection involved antimicrobial peptides, heat shock proteins and galectins. Interestingly, oysters in the later stages of OsHV-1 infection had a transcriptional response that resembled an antibacterial response, which is suggestive of the oyster’s microbiome causing secondary infections (dysbiosis-driven pathology). This study provides molecular evidence that oysters can mount distinct immune response to viral and bacterial pathogens and these responses differ depending on the age of the host.