In vivo multiscale analyses of spring viremia of carp virus (SVCV) infection: From model organism to target species

Abstract

Spring viremia of carp virus (SVCV) has a broad fish host spectrum and is responsible for a disease that generally affects juvenile fishes with a mortality rate of up to 90%. In the absence of treatments or vaccines against SVCV, the search for prophylactic or therapeutic solutions is thus relevant, particularly to identify solutions compatible with mass vaccination. In addition to being a threat to aquaculture and ecosystems, SVCV is a unique pathogen to study virus-host interactions in the zebrafish model. Establishing the first reverse genetics system for SVCV and the design of recombinant SVCV (rSVCV) expressing fluorescent or bioluminescent proteins adds a new dimension for the study of these interactions using innovative imaging techniques. The infection by bath immersion of zebrafish larvae with rSVCV expressing mCherry allows us to define the first SVCV replication sites and the host innate immune responses using different transgenic lines of zebrafish. The fins were found as the main initial sites of infection in both zebrafish and carp, its natural host. Hence, new insights into the physiopathology of SVCV infection have been described. We report that neutrophils are recruited at the sites of infection and persist up to the death of the animal leading to an uncontrolled inflammation correlated with the expression of the pro-inflammatory cytokine IL1β. Tissue damage was observed at the site of initial replication, a likely consequence of virus-induced injury or the pro-inflammatory response. Interestingly, SVCV infection by bath immersion triggers a persistent pro-inflammatory response rather than activation of the antiviral IFN signaling pathway as observed following intravenous injection, highlighting the importance of the route of infection on the progression of pathogenicity. Thus, this model of zebrafish larvae infection by rSVCV offers new perspectives to study in detail virus-host interactions and to discover new prophylactic or therapeutic solutions.

Fig 1. SVCV genome and infectious cDNA construct.

A. Whole-genome SVCV phylogenetic analysis. The complete genomic sequences of 16 SVCV isolates (Sprivivirus cyprinus) and Pike Fry Rhabdovirus (Sprivivirus esox) were obtained from NCBI Genbank. The SVCV sequence highlighted in red corresponds to the Fijan strain (Genbank #AJ318079.1) and was used to establish the SVCV reverse genetics system. Accession numbers are provided in each branch of the phylogenetic tree and in Table 1. The genomes were aligned using MUSCLE and the choice of substitution model was performed using MEGA X. A Maximum-Likelihood phylogenetic tree was inferred from the alignment using PhyML with the GTR+G substitution model and 1000 bootstrap replicates. Numbers at nodes indicate percent bootstrap support. The tree was drawn to scale using FigTree 1.4.4 and rooted using Pike fry rhabdovirus. Branch lengths correspond to the number of substitutions per site. The host species, date and country of origin are shown when found in the sequence metadata. B. Construction of the pSVCV plasmid encoding the complete antigenomic RNA of the SVCV Fijan strain. Five overlapping cDNA fragments (numbered 1 to 5) covering the complete Fijan strain antigenome were generated by RT-PCR with primers described in Table 2 and were assembled in pBluescript SK− following appropriate restriction enzyme digestion. In the final construct, named pSVCV, the leader end of the antigenome cDNA is flanked by the T7 promoter (T7), and the trailer end is fused to the hepatitis delta virus ribozyme followed by a terminator for T7 RNA polymerase (T7t). The complete antigenomic cDNA was designed to create XhoI, SmaI and EagI unique restriction sites in the M-G, G-L intergenic and trailer regions, respectively (see S1 Fig). In addition, this genome differs from the genome of the Fijan reference strain by three nucleotide substitutions in N and G genes.

Fig 2. Insertion of a transcription cassette encoding tracer proteins in the SVCV genome.

A. As described in the Materials and methods, an expression cassette encoding mCherry, firefly luciferase (ffLUC) or akaluciferase (akaLUC) was inserted in the M-G and G-L intergenic regions flanked by additional GS and GE transcription signals of SVCV using the unique XhoI and SmaI restriction enzyme sites, respectively, leading to the final construct pSVCV-mCherry M/G, pSVCV-ffLUC M/G, pSVCV-mCherry G/L, pSVCV-ffLUC G/L and pSVCV-akaLUC G/L (not to scale). Together with an expression cassette encoding GFPmax between M and G genes, a second expression cassette, as described above, encoding mCherry was inserted in the G-L intergenic region using the unique SmaI restriction enzyme site, leading to the final construct pSVCV-GFPmaxCherry that encodes simultaneously both green and red fluorescent proteins. B-C. EPC cells were infected with rSVCV-mCherry (B) or rSVCV-GFPmaxCherry (C) at a final MOI of 0.1. The cells were incubated at 25°C for 24 hours. Live cell monolayers were then visualized with a UV-light microscope after nuclei staining with Hoechst solution in the cell culture medium. 63× objective. Scale bars, 10 μm. D. EPC cells were infected with rSVCV-ffLUC at a MOI of 0.1. At 10 hours post-infection (hpi) and 17 hpi live cells were washed three times with sterile Phosphate Buffer Saline (PBS) and the D-luciferin substrate was added at a concentration of 250 μM. The luminescence was measured using the IVIS Spectrum BL imaging system.

Fig 3. Experimental carp infection with rSVCVs.

Virulence in carp of rSVCV and rSVCV bearing one or two additional expression cassettes compared to the parental Fijan strain. Juvenile carp [mean weight, 0.81 g (n = 50 per group)] were infected by bath immersion for 2 h in 3 L with 5 × 10⁴ PFU/mL of each of the indicated rSVCVs and maintained at 10°C. rSVCV is a wild-type recombinant virus derived from the Fijan strain. Mortality was recorded daily and is presented as the percent of survival. Mock, non-infected carp.

Fig 4. Kinetics of bioluminescence emission of rSVCV-akaLUC in vivo in carp.

Carp (mean weight, 1.91 g) were divided into two groups with one group infected by immersion with rSVCV-akaLUC G/L (A) as described in the Materials and methods and the other group had their caudal fins cut prior to infection by immersion with rSVCV-akaLUC G/L (B). At 3, 7, and 24 days post infection, fish were randomly harvested and transferred in a small tank with water containing Akalumine-HCl substrate for 2 h at 10°C. Two hours later, anesthetized fish were imaged using an IVIS Spectrum BL imaging system. Mock, non-infected carp (C). Stars indicate carp with detectable bioluminescent foci. Blue stars indicate the same fish before and after dissection.

Fig 5. Zebrafish larvae model of rSVCV-mCherry infection by bath immersion.

A. Nacre zebrafish larvae (a total of n = 147 from three independent experiments) at 3 dpf were infected by bath immersion with rSVCV-mCherry (2 × 10⁶ PFU) and incubated at 24°C. Fish were observed at different times post infection for the detection of mCherry fluorescent foci with a fluorescent stereomicroscope and mortalities were recorded daily. Mortality is presented as the mean of cumulative percent of dead larvae recorded in three independent experiments as well as the percent of mCherry positive larvae. No mortalities were recorded in the mock-infected group (a total of n = 30 from three independent experiments). B. Colocalization of mCherry protein and SVCV nucleoprotein (N). Mock infected and SVCV infected larvae were co-immunostained with antibodies raised against mCherry (in red) and SVCV nucleoprotein (in green). Scale bars: 200 μm. C. Virus load in individual larva. At different times post infection, individual larvae (n = 10 in 2 independent infections) were randomly harvested and virus load was determined by plaque titration in EPC cells. Means are shown together with standard errors. D. rSVCV-mCherry replication in zebrafish larvae. Five groups of 5 larvae from 2 independent experiments were randomly harvested at different times post infection and gene expression was analyzed by RT-qPCR. Virus loads are expressed as the ratio of mRNA copy of SVCV nucleoprotein to ef1a housekeeping gene.

Fig 6. Detection of mCherry fluorescent foci in larvae at early stages of SVCV infection.

A. Example of a zebrafish larva at 4 dpf in which the anatomical segments corresponding to the head, trunk, and tail are indicated. Scale bars: 200 μm. B. Details of the different foci of infection that can be observed in larvae infected at 24 hpi with 2 × 10⁶ PFU of rSVCV-mCherry. Of the 26 photographed larvae, 5 larvae had fluorescent foci in the head, 3 larvae in the trunk, and 16 larvae in the tail (2 larvae without detectable fluorescence). Pictures are presented in the following order: brightfield, red fluorescence, and merge of the two previous acquisitions. Magnification 6.3×. Scale bars: 200 μm.

Fig 7. Viral spread in larvae after bath infection.

A. Whole-larvae infection pattern observed in rSVCV-mCherry infected larvae at 48 hpi. Of the 20 photographed larvae, 2 larvae had only fluorescent foci in the tail, 5 larvae in the trunk and tail, and 7 larvae in the head, trunk, and tail, (6 larvae without detectable fluorescence in the stereomicroscope). Scale bar: 200 μm. B. Tracking of the distribution of the mCherry signal in larvae body (head, trunk, tail) recorded in the same experiment at 24, 48, and 72 hpi (a total of n = 127, data are presented as the mean of 3 independent experiments with standard errors).

Fig 8. Quantification of fluorescence intensity in rSVCV-Cherry infected larvae using the COPAS system.

A. Total red fluorescence signal of rSVCV-mCherry infected and mock-infected larvae measured with COPAS system at different times of infection (mean of 30–39 larvae with standard errors). Statistically significant differences are displayed as follows: ***, p value < 0.001; **, p value < 0.01. B. Virus load in larvae analyzed with the COPAS. After passing through the COPAS system, the larvae were harvested in 5 groups of 6 larvae, and virus load was determined by plaque titration in EPC cells. Means are shown together with standard errors. C. Correlation between the two methods: titration in EPC cells and COPAS analysis in larvae infected with rSVCV-mCherry. D. Profiles of mCherry distribution obtained with the COPAS system and segmented as in Fig 7. Examples of mock-infected and rSVCV-mCherry infected larvae profiles at 24 hpi are shown. Blue line: optical density of the larvae, Red line: profile of red fluorescence signal. Scale bar: 200 μm.

Fig 9. Analysis of the innate immune response in 3 dpf larvae infected by bath or microinjection in the duct of Cuvier with rSVCV-mCherry.

Expression analysis of the principal genes involved in the anti-inflammatory response (il1b, tnfa), antiviral response (ifnphi1, isg15), and marker genes of the two major innate immune cells (mpeg1 for macrophages, and mpx for neutrophils) measured by RT-qPCR when larvae were infected by immersion in the viral suspension (A, C, and E, respectively) or by IV microinjection (B, D, and F, respectively). 5 groups of 5 larvae each were sampled for rSVCV- and mock-infected larvae at 0, 6, 24, and 48 hpi and repeated twice for the bath infection experiment. 5 groups of 5 larvae each of IV rSVCV-infected and non-infected larvae were sampled at 0, 6, and 24 hpi. Each sample was normalized to the reference gene ef1a. The normalized expression values were standardized against their respective controls (Control fold change (FC) = 1, basal line). The graphs represent the mean of the fold changes ± SEM of the biological replicates, except for the il1b gene which is represented as the log₁₀ of the fold change expression ratio. Statistically significant differences are displayed as follows: ***, p value < 0.001; **, p value < 0.01; *, p value < 0.05.

Fig 10. Neutrophil recruitment to the initial SVCV replication site in the caudal fin.

A. Example of ROI delimitation done for all the analyzed larvae. Scale bar: 200 μm. B. Micrographs of merged GFP (MPX, neutrophils) and mCherry (rSVCV) signals of ROI delimited areas in SVCV-infected and mock-infected larvae at 24 hpi. C. Micrographs of merged GFP (MPX, neutrophils) and mCherry (rSVCV) signals of ROI delimited areas in SVCV-infected and mock-infected larvae at 48 hpi. D. Representation of the number of neutrophils present in the fin fold of the caudal fin of SVCV-infected and non-infected larvae at 24 and 48 hpi. The graph represents the mean ± SEM. Statistically significant differences are displayed as *, p value < 0.05.

Fig 11. Visualization of IL1β expression in SVCV-infected larvae.

IL1β expression in mock-infected and SVCV-infected larvae at 24 hpi (A) and 48 hpi (B-C). Confocal images of GFP (IL1β, green) and mCherry (rSVCV, red) signals observed in entire Tg(il1β:GFP-F) larvae were taken under a 10× objective. Scale bars: 200 μm. The blue box shows the magnified area of the caudal fin of SVCV-infected (A-B) and mock-infected (C) larvae taken with a 25× objective and zoom 2. Scale bars: 200 μm. The yellow box shows the magnified area of the caudal fin wound taken with a 25× objective and zoom 3. Scale bars: 20 μm.

Fig 12. Neutrophil trafficking in induced injuries.

A. Experimental design. 1. Photoablation (PA): Localization of the injury done by photoablation in 3 dpf larvae and zoom of the dorsal area at the level of urogenital aperture at the moment of PA (0 hour post PA, 0 hpPA). Scale bar: 200 μm. 2. Larvae infection: larvae were rSVCV- or mock-infected by immersion for 1 h, rinsed, and then incubated for 2 h before recording. 3. Time-lapse recording: Images taken with the stereomicroscope during the mounting of the larvae in agarose just before starting the recording of the time-lapse (4 hpPA, 2 hpi) of rSVCV- or mock-infected larvae. Images of the GFP signal of the neutrophils in green, and photoablation area marked with a yellow dotted crescent in the brightfield channel. B. Representation of the percentage of GFP mean intensity on the first frame of the time-lapse record in the photoablation area of rSVCV- or mock-infected larvae from 4 hpPA and 2 hpi to 22 hpPA and 20 hpi. The graph represents the mean ± SEM. C. Stereomicroscope images of GFP (mpx, neutrophils) and mCherry (rSVCV) signals of photoablation site of rSVCV- or mock-infected larvae at 26 hpPA and 24 hpi. The photoablation area is marked with a yellow dotted crescent. Scale bars: 200 μm.