A solinvivirus induces cellular antiviral responses and reduces the lifespan of adult black soldier flies

TL;DR

A virus called HiSvV causes disease and shortens the lifespan of black soldier flies, which are important for sustainable insect farming.

Contribution

This study is the first to confirm HiSvV as a pathogen in black soldier flies and demonstrates its replication and transmission.

Findings

HiSvV replicates in black soldier flies and causes premature mortality.

The virus is transmitted both horizontally and vertically.

Infected flies trigger antiviral responses despite not activating small RNA pathways.

Abstract

Viral pathogens pose an emerging threat to the sustainability of insect mass-rearing systems, yet they remain understudied in key species like the black soldier fly (BSF; Hermetia illucens). Although multiple viral sequences have been reported in BSF, their role in disease has not been established until now. Here, we provide the first in vivo characterization of H. illucens solinvivirus (HiSvV), confirming its role as a viral entomopathogen of BSF. Metatranscriptomic analysis of a diseased colony revealed a high viral load attributable to HiSvV. We successfully isolated the virus and developed injection- and oral-based infection assays to investigate replication, tissue tropism, transmission and risk of mortality. HiSvV replicated in inoculated adults, induced premature mortality in flies and was transmitted both horizontally and vertically. Infected flies also mounted a broad antiviral response, which supported active pathogenesis, even if the small RNA pathways were not activated. These findings establish HiSvV as the first confirmed viral pathogen of BSF and underscore the urgent need for viral surveillance and experimental tools to safeguard industrial insect rearing.

Figures

Detection of viral sequences in the BSF samples. (a) Visualization of read classifications based on result outputs generated by Lazypipe2 analysis. For both the HA and DA, the minor taxa classifications (encompassing mostly bacteria and archaea) totalled less than 0.03% of transcripts. (b) Mapping of sequencing reads to the HiSvV genome. Annotated polyprotein open reading frame (ORF) (orange) with putative conserved regions of proteins (purple) consists of an RH, P, RdRP, BS and MC. The upper coverage panel displays the HiSvV coverage in the HA, and the lower coverage panel shows the coverage of HiSvV in the DA. Note the difference in y-axis scales for the HA and DA samples, ‘k’ is a place holder for ‘value × 1000’. BS, dsRNA-binding site; DA, dead adult; MC, major capsid; P, protease; RdRP, RNA-dependent RNA polymerase 1; RH, RNA helicase.

HiSvV purification and visualization by TEM. (a) Enrichment of HiSvV genomes after virus partial purification. Stages of the prepurification protocol compared were the initial homogenate filtered through gauze (FH) and the PPV. ‘nd’ represents the lower threshold of detection for HiSvV. (b–d) TEM observation of HiSvV particles in the PPV samples. The scale bars in the bottom right corner, coloured in yellow, represent 100 nm (b and c) and 200 nm (d). PPV, partially purified virus.

Detection of HiSvV in adults and egg clusters after prepupal injection. (a) Viral relative abundance at different days after emergence from pupae and (b) detection of HiSvV in different tissues of adults processed 3 dpe. The tissues from the head and antennae (H.&A.), wings and halteres (W.&H.), thoracic muscles (T.m), mid- and hindgut with Malpighian tubules (M.&M.t.), reproductive organs (R.o.) and fat bodies (F.b.) were tested as separate samples. Individuals injected with PBS were used as negative controls during inoculation. The error bars represent the sd of the mean on a log scale. Only the upper bars and caps of the sd are shown. (c and d) Testing horizontal and vertical transmission of HiSvV between BSFs. Adults and eggs were obtained during cohabitation experiments, and the level of HiSvV was assessed in the (c) adults and (d) egg clusters. Single male and female individuals were each represented by red dots and blue squares, respectively. The green points represent single egg clusters tested. Individuals inoculated with HiSvV during the prepupal stage are indicated using ‘+V’, while ‘−V’ specifies mock-infected adults. The lower threshold for detection of HiSvV in samples was indicated by ‘nd’ during qPCR testing.

The effect of HiSvV oral infection on the survival of adult BSFs. (a) Abundance of HiSvV in orally infected adults. Mean relative viral abundance in adults over 7 days post-infection. For time post-inoculation timepoints, ‘h’ indicates hours and ‘d’ indicates days. Different colours represent the females (blue) and males (red). The filled and empty shapes indicate HiSvV- and mock-infected individuals, respectively. Lines indicate HiSvV (solid) and mock (dotted) infections. The upper error bars represent the sd of the mean on a log scale and ‘nd’ indicates the lower detection threshold for HiSvV. (b and c) Impact of HiSvV infection on adult lifespan. Adults were orally infected with HiSvV PPV at two concentrations: 1.6×104 genome equivalents per microlitre for experiment 1 (b) and 1.6×101 genome equivalents per microlitre for experiment 2 (c). Males and females are differentiated by the colours red and blue, respectively, and a solid line separates HiSvV-infected from mock (PBS)-infected individuals. The P-value within each plot was established by a global comparison of strata within each experiment, indicating a statistical difference between all strata.

Analysis of transcriptional response to HiSvV infection in BSF adults. (a) Volcano plots depicting DEGs with negative and positive L2FCs. (b) Heatmap of immune-related genes with significant L2FCs contrasting HiSvV- vs mock (PBS)-infected males and females. Males and females underwent DEQseq2 analysis separately for heatmap visualization. The black dots represent genes with significant L2FC in males, and the black dashes for females. The green dashes indicate genes for which the gene count was ‘0’. Sizing indicates sRNA from the HiSvV genome degradation. (c) sRNA reads were mapped to the HiSvV genome in HiSvV- and mock-inoculated BSFs. Profiles were obtained for adults collected at 5 and 7 days post-infection, for both males and females. sRNA sequences which started with an adenine were coloured in green, uracil in blue, thymine in red, guanine in yellow and ambiguous ‘N’ in black. The percentage of mapped sRNA reads to the total number of reads in each sample was indicated in the top right corner of each subplot. The bars above zero represent positive-sense RNA mapping, and the bars below zero represent negative-sense mapping.

Tables

Table 1.. Putative Reactome and KEGG pathway annotations of significant (PAdj<0.05) DEGs related to immune-linked BPsBoth KEGG and Reactome annotations were statistically significant (P<0.05). Genes coloured in red had an L2FC above 0, and those in blue had an L2FC below 0. ‘n/o’ indicates no significant L2FC observed.

NCBI gene ID (LOC#)	Gene description	L2FC		Pathway
		Male	Female	KEGG	Reactome
119654411	Lysozyme-like	4.17	3.40	Salivary secretion	Undefined
119654410	Lysozyme-like	6.27	3.19
119654544	Lysozyme 1-like	8.16	7.46
119654435	Uncharacterized LOC119654435	2.1	n/o	Apoptosis – fly	Programmed cell death
119651483	Caspase-1	3.37	2.53
119659367	Caspase-8	2.18	n/o	Apoptosis – fly; Toll and IMD signalling pathway	Apoptosis and cytokine signalling
119648931	Baculoviral IAP repeat-containing protein 3-like	3.72	3.11	Apoptosis – multiple species; Toll and IMD signalling pathway; ubiquitin-mediated proteolysis	Programmed cell death and PTEN regulation
119649011	Death-associated inhibitor of apoptosis 1-like	4.76	n/o
119650124	Death-associated inhibitor of apoptosis 1-like	2.97	2.64	Apoptosis – multiple species; ubiquitin-mediated proteolysis
119656049	Procathepsin l-like	n/o	5.96	Autophagy – animal; lysosome; phagosome; apoptosis; antigen processing and presentation	Innate immune system
119658254	5'−3' exonuclease PLD3-like	2.01	n/o	Ether lipid metabolism	Phagocytosis and inositol phosphate metabolism
119657566	5'−3' exonuclease PLD3-like	2.49	n/o
119658345	Uncharacterized LOC119658345	2.07	n/o
119655431	Uncharacterized LOC119655431	2.64	n/o	Protein processing in endoplasmic reticulum	Antiviral mechanism by IFN-stimulated genes
119656223	Uncharacterized LOC119656223	3.05	n/o
119656686	Uncharacterized LOC119656686	6.08	5.3
119655383	Uncharacterized LOC119655383	5.24	3.56
119654507	Uncharacterized LOC119654507	4.49	3.83
119654505	Uncharacterized LOC119654505	6.33	5.42
119652716	Uncharacterized LOC119652716	4.71	3.15
119650308	Baculoviral IAP repeat-containing protein 7-like	1.83	n/o	Ubiquitin-mediated proteolysis	Programmed cell death and TNF signalling (cytokine)
119650016	E3 ubiquitin-protein ligase XIAP-like	3.02	3.28
119659057	Uncharacterized LOC119659057	2.41	n/o	Toll and IMD signalling pathway	Cytokine signalling; Toll receptor cascades; interferon response and death receptor signalling
119658693	Peptidoglycan-recognition protein LB-like	2.07	n/o		AMPs; innate immune system
119656959	Peptidoglycan-recognition protein LE	2.77	2.41		Undefined
119658006	Protein spaetzle-like	2.50	n/o
119656559	Protein Toll-like	3.34	2.58		Toll receptor cascades
119656561	Protein Toll-like	4.65	2.86
119656562	Protein Toll-like	3.17	n/o	Undefined	Undefined
119659331	Toll-like receptor 6	1.47	n/o
119648531	E3 ubiquitin-protein ligase XIAP-like	11.86	10.3		RIPK1-mediated regulated necrosis
119653685	Caspase-3-like	4.85	3.39		Death receptor signalling
119649716	Protein argonaute-2-like	2.09	n/o		Small interfering RNA biogenesis; gene silencing by RNA; microRNA biogenesis
119649605	Protein argonaute-2-like	2.83	2.76
119661431	Ribonucleoside-diphosphate reductase large subunit	2.08	n/o		Undefined
119659672	Ribonucleoside-diphosphate reductase subunit M2	1.78	n/o
119659869	Cytokine receptor-like	1.8	n/o
119659746	Cytokine receptor-like	1.92	2.51
119657116	Protein C19orf12 homologue	4.85	4.04
119657115	Protein C19orf12 homologue	2.53	n/o
119657249	Attacin-B-like	3.89	4.02
119657250	Attacin-B-like	6.26	6.26
119651604	Protein draper-like	3.46	2.84
119648532	Baculoviral IAP repeat-containing protein 1e-like	5.27	n/o
119646668	Inhibitor of nuclear factor kappa-B kinase subunit beta	2.09	n/o
119653883	Endoribonuclease dicer	1.52	n/o
119660914	Holotricin-3-like	n/o	2.42
119647010	Holotricin-3-like	−2.68	n/o
119647096	E3 ubiquitin-protein ligase MARCHF2-like	−2.21	n/o
119648269	Acanthoscurrin-1-like	−1.9	n/o
119654952	Acanthoscurrin-2-like	−4.28	n/o
119648727	Alaserpin-like	−1.57	n/o
119648753	Ctenidin-3-like	−3.09	n/o
119660624	Succinate-CoA ligase (ADP-forming) subunit beta	−1.92	n/o
119657830	Cecropin-like peptide 1	−6.58	n/o	Toll and IMD signalling pathway

Full text

Data availability

The sequencing data generated during this study can be found under the NCBI BioProject PRJNA1079553; NBCI SRA accession numbers are listed in Table S5.

Introduction

Black soldier flies (BSFs), Hermetia illucens, are increasingly used in waste management and as a source of protein in animal feed, making their health and productivity crucial for these industries [15]. Studies on entomopathogens affecting BSFs are limited, presenting a gap in BSF pathology that has been highlighted on multiple occasions [1,69]. Although there has been a growing number of studies on the effect of more general insect pathogens [1014], the first natural pathogen of BSFs was only recently described in 2023 [8]. The authors described a Paenibacillus sp*.*-related pathogen causing the death of BSFs with a syndrome previously described as ‘soft rot’.

Issues in insect mass-rearing can also arise from a less conspicuous agent, such as viruses [1516]. Often, the presence of viruses can be missed, particularly since signs can mimic those of sub-optimal rearing practices, such as smaller larvae, lower feeding rates and reduced fecundity [1517]. In other cases, viral infections may remain unnoticeable until certain factors trigger the infection to produce obvious symptoms, including bloated individuals or even death [16,1820]. Since 2022, eight exogenous viruses associated with BSFs have been discovered [792122]. Among them, four were classified into the families Dicistroviridae, Iflaviridae, Rhabdoviridae and Solinviviridae and are closely related to other known insect pathogens [16,2326]. However, no additional evidence was obtained on the potential pathogenicity of these viruses.

Solinviviridae is a recently described viral family of the Picornavirales with non-segmented, linear, positive-sense RNA genomes of ∼10–11 kb, suspected of including an array of arthropod pathogens [25]. According to the International Committee on Taxonomy of Viruses (ICTV) [25], there are ∼69 viral sequences closely related to Solinviviridae, but only two species are accepted within the family. Despite this high number, few studies describe the actual viruses or their interactions with their hosts [2730]. The most well-described viruses in the Solinviviridae family are Solenopsis invicta virus 3 (SINV3), Nylanderia fulva virus 1 (NfV1) and two closely related viruses, Apis mellifera solinvivirus 1 (AmSV1) and Penaeus vannamei solinvivirus (PvSV), which are not yet formally accepted as Solinviviridae. For SINV3, high virus titre has been linked to mortality and can cause colony collapse and decreased fecundity in ants [273132]. The ecology of the Solinviviridae is not very well established, since the bulk of the family characterization work has been done using mainly SINV3 and NfV1 [25,27, 28, 30, 3241]. A common feature in Picornavirales (including Iflaviridae and Dicistroviridae), closely related to Solinviviridae, is that infections in insect rearings can go unnoticed until an outbreak occurs, often resulting in a strong display of symptoms and colony collapse [1518324243].

Insects are capable of mounting a broad array of immune responses against viruses, such as melanization and RNA interference (RNAi) [4445]. Diptera, including mosquitoes and Drosophila, rely particularly on the Toll, immune deficiency (IMD), Janus kinase-signal transducer and activator of transcription (JAK-STAT), RNAi and RNA decay, as well as autophagy pathways for their antiviral defences [4446]. Although these pathways can act against a variety of viruses, their relative importance differs among virus groups [4648]. To date, transcriptomic studies of BSF immunity remain limited. While more than 600 immune-related genes have been identified and annotated in BSFs, viral transcriptional studies have relied heavily on homologues from Drosophila to infer immune responses in BSFs [214950].

The goal of this study was to unravel the potential pathogenic effect of viruses in BSF adults derived from a BSF rearing facility experiencing unexplained mortalities. After identifying H. illucens solinvivirus (HiSvV) as the primary candidate responsible for the mortality experienced in a BSF colony, the virus was isolated and its interactions with BSFs were further characterized. To do that, we studied HiSvV transmission, pathology and the transcriptional response of BSFs to the infection. This comprehensive approach allowed us to better understand HiSvV biology and assess its risk and relevance to the BSFs mass-rearing industry.

Methods

BSFs laboratory colony

A starter colony was provided by Entomotech S.L. (Almeria, Spain), and the rearing was originally installed at the Control Biotecnológico de Plagas (CBP) research group at the University of Valencia (Spain) in 2022. This colony was considered ‘virus free’ after initial screenings for use in HiSvV infection studies [22]. The individuals were reared at 28 °C on a Gainesville diet (50% wheat germ, 30% alfalfa and 20% corn flour) in a similar manner to Deruytter et al. [51], but adding 150% MilliQ water to the dry diet mix (0.3 kg of diet to 0.45 l of water). Essentially, 0.3 l (100%) of water was initially added to 0.3 kg of diet and kept at 4 °C overnight to allow the diet to absorb the moisture. Before the diet was provided to the neonates or larvae, the remaining 0.15 l (50%) of water was mixed into the premoistened diet. Eggs laid were collected and incubated in parafilm-sealed petri dishes until the neonates hatched and were then placed onto 60 g of diet in a 0.5 l container. Five- to seven-day-old larvae were then moved onto 250 g of fresh diet within a 2 l container until pupation. The prepupae and pupae were manually sorted from the substrate and placed into the 2 l containers with a dry sheet of paper towel until emergence. Adults were then placed in cages (47.5×47.5×47.5 cm), and a cotton ball soaked in MilliQ water was provided every 2 days.

Virus detection in transcriptomic data

Reads from two metatranscriptomic datasets (National Center for Biotechnology Information (NCBI) Sequence Read Archive accession numbers SRR28203243 and SRR28203244) obtained from BSF adults originally received from a BSFs rearing facility experiencing unexplained mortalities were analysed for the presence of viruses infecting H. illucens, as done by Pienaar et al. [22]. Through this approach, Krona charts generated using Lazypipe 2 were used to visualize viral diversity within the metatranscriptomic datasets relative to the BSF reference genome (GCF_905115235.1) [4952]. Additionally, raw reads were trimmed using fastp v0.23.2 [53] and mapped to the reference genome of HiSvV (PQ228193). Bowtie2 v2.4.2 [54] and samtools v1.9 [55] were used to map and filter the unmapped reads from the BAM files. The mean coverage was then assessed using Geneious Prime v2021.1 (https://www.geneious.com), and the coverage plots were generated using the same script as previously described [22].

Virus partial purification and transmission electron microscopy

Depending on the number of individuals used (2–55), pools of BSF individuals (either pupae or adults) were crushed and homogenized in 0.5 ml of Tris-EDTA (TE) buffer containing 0.06% SDS using a 1 ml pipette and tip, or up to 17 ml of TE buffer using a 5 ml pipette and tip, inside either a 1.5 ml centrifuge tube or a 50 ml falcon tube. Virus particles were partially purified using the virus prepurification protocol used by Hernández-Pelegrín et al. [56]; however, centrifugations were performed at 10,000 × g. When needed, an additional concentration step was performed by 10% PEG 6000 precipitation, and the pellet containing the viral particles was resuspended in 100 µl of 1× PBS for further experiments.

For transmission electron microscopy (TEM), partially purified virus (PPV) underwent negative staining using 2% phosphotungstate after samples were fixed to a carbon-coated grid. The grids were visualized, and images were captured using an HT7800 RuliTEM 120 kV TEM (Hitachi, Chiyoda City, Japan). The mean size of the viral-like particles was measured from 30 viral particles using ImageJ v1.54g [57].

Virus detection by reverse transcription quantitative PCR

RNA extractions were performed as in Pienaar et al. [22]. Post extraction, a DNase treatment was performed on the RNA using the DNase I, RNase-free kit (EN0521; Thermo Fisher Scientific, Waltham, MA, USA), followed by cDNA synthesis using the PrimeScript RT Reagent Kit (Perfect Real Time) (TAKRR037A; Takara Bio, Kusatsu, Japan). The reverse transcription quantitative PCRs (RT-qPCRs) were performed as in Pienaar et al. [22]. The relative abundance of the viral targets was obtained using the RPL8 housekeeping gene as described in Herrero et al. [58]. The plots were generated using the relative abundance values and ggplot2 v3.5.0 [59] in R v4.2.2 [60]. The quantification of isolated virus was measured as ‘genomic equivalents’, calculated using the equation ‘Amplification factor^(y-intercept − Ct-value)^’.

Viral infections by injection of prepupae and pipette feeding of adults

The effect of infection on the colony was initially assessed by viral injection during the prepupal stages. For this purpose, four experiments were performed on different generations using 100 prepupae for each treatment. Prepupae were sorted from the substrate, washed three times with MilliQ water and placed on paper towel to dry before injection. For inoculation, the needle was inserted dorsolaterally at the end at an angle close to 90° into the tegument connecting the second and third segments from the posterior end (Fig. S1A).

Individuals were injected with 5 µl of either 1× PBS solution or PPV containing 4×10^8^ HiSvV genomic equivalents per microlitre. Post-injection, the prepupae were placed into 2 l pupation boxes but without any substrate and left to pupate at 25 °C with a lighting routine set to 12:12 h. From 9 days post-injection, the containers were checked daily for emerged adults. Once emerged, to collect adults, the containers were placed at 5 °C for a maximum of 10 min, and the adult males and females were placed separately into fresh containers precooled on ice for counting and sex identification. Afterwards, they were placed at 28 °C for further experiments and were maintained in complete darkness, except for adults selected for cohabitation experiments. The rate of emergence and sex were recorded and visualized using ggplot2, and a Kruskal–Wallis rank-sum test [61] in R stats v4.2.2 [60] was used to test the overall statistical differences between emergences.

Adults were inoculated by pipette-feeding them a solution containing prepurified HiSvV. To do that, adults were collected within 24 h post-emergence, and males and females were then retained in two separate dehydration boxes overnight (Fig. S1B). Each dehydration box contained a 1–2 cm deep layer of wood shavings. For the inoculation, individuals were held inside a P20 pipette tip by gently applying pressure to their thorax against the lip of the tip opening, with their wings outside, and allowed to drink (Fig. S1C). The abdomen windows were observed for visible dye to confirm that the flies were drinking (Fig. S1D). The inoculation solution was preprepared as follows per 4 µl dose: 1 µl of 5% sucrose water, 1 µl of food colouring and 2 µl of HiSvV PPV or 2 µl of 1× PBS.

Each droplet-feeding experiment consisted of two treatments, HiSvV inoculated and mock infection (PBS control), and unless stated otherwise, the virus treatment group was inoculated with a HiSvV concentration of 1.6×10^4^ genomic equivalents per microlitre. Inoculated adults were placed individually in 120 ml cups, and 1 ml of MilliQ water was provided directly to a cotton ball placed between the cup lid and a piece of paper towel (Fig. S1B). The cups were stored in an incubator at 25 °C with a 12:12 h (light:dark) lighting regime.

Assessment of HiSvV replication in BSF adults

The replication of HiSvV was monitored in adults for both the prepupal injection assays and the oral injection assays. For prepupae inoculated by injection, adult males and females were collected separately at various time points post-emergence. Additionally, adults that died during cohabitation between 8 and 13 days post-emergence (dpe) were collected. In each case, pools of two individuals were prepared. For orally infected adults, two males and two females were collected at different time points post-inoculation for qPCR as described above.

Tropism of HiSvV in adult BSFs

Virus tropism was determined in adults derived from the injected prepupae (three out of four experiments). From each experiment, one male and one female were collected at 3 dpe. Adults were washed with TE buffer (10 mM Tris-HCl, pH 8–8.6, 1 mM EDTA) beforehand, then the head, wings and legs were dissected and placed into tubes according to body parts. Afterwards, the corpse was submerged into TE buffer and, from the abdomen, the reproductive organs were removed, followed by the fat bodies, mid- and hindgut with Malpighian tubules and lastly, the thoracic muscles. Each piece of anatomy was rinsed three times in TE buffer before transfer to 300 µl of TRItidy G reagent (AppliChem GmbH, Darmstadt, Germany).

HiSvV transmission

Cohabitation experiments were designed to evaluate the horizontal and vertical transmission of the virus. To implement this, 15 individuals of each sex were selected to place 30 adults in a ‘25×25×25 cm’ cage. For each experiment, four cages were set up, each with a different combination of males and females. The following combinations were used: virus-positive males and females, virus-negative males and females, virus-positive males with virus-negative females and finally, virus-negative males with virus-positive females. This step was repeated four times using individuals from different generations. Prepupal-injected males and females emerging from pupae as adults within 24–48 h of each other were used for each experiment, and adults inside cages were maintained in a 12 h light and 12 h darkness regime. The cages were checked daily for eggs and dead adults, which were subsequently stored at −80 °C to test for horizontal (adults) and vertical (eggs) transmission of HiSvV using RT-qPCR as described above.

Adult BSF survival bioassays after HiSvV infection

Adults were orally infected as described above, and their mortality was monitored. The adults were incubated at 25 °C for up to 30 days, and 1 ml of MilliQ water was provided daily. The experiments were performed using HiSvV concentrations of 1.6×10^1^ and 1.6×10^4^ genome equivalents per microlitre, and negative controls were mock infected with PBS. Thirty females and thirty males were used for each treatment. Mortality was recorded daily, and survival curves were plotted using survfit from the survival v3.5.8 R package [6263] and ggsurvplot from the ggplot2 R package. Type II ANOVAs from the car v3.1.2 R package [64] were then performed on Cox proportional hazard models (survival R package) to test factors such as sex and experiment.

Differential gene expression analysis

The impact of viral infection on BSF gene expression was analysed by RNA sequencing (RNAseq). To do that, RNA was extracted from adults orally infected with HiSvV at different time points. Library preparation and high-throughput sequencing were carried out previously by Pienaar et al. [22]. Additionally, from the same samples, small RNA (sRNA; single-end 50 bp reads) was also sequenced using a NovaSeq 6000 (Illumina, San Diego, USA) by Novogene (Beijing, China). Datasets can be found at NCBI (BioProject no. PRJNA1079553). To obtain BSF transcripts, reads from long non-coding RNA and mRNA datasets were trimmed, and poor-quality reads were removed using fastp. Then the reads were mapped to the BSF reference genome (GCF_905115235.1) using HISAT2 v2.2.1 [65]. Using samtools, the mapped reads were sorted, filtered and SAM files were converted to BAM files. An R script from Cerqueira de Araujo et al. [66], utilizing featureCounts from the Rsubread R package v2.8.2 [67], was used in combination with a GTF file for the BSF reference genome to obtain a gene count table from the BAM files. The DESeq2 R package v1.38.3 [68] was then used to contrast treatments HiSvV and mock (PBS) and obtain the log2-fold changes (L2FCs) of genes using the design model ~ ‘Treatment + Sex’. The gene counts for the whole transcriptome were normalized using DESeq, and the counts of genes with a significant L2FC (Adjusted *P-*value (PAdj)<0.05) were extracted. This was then repeated for males and females separately, only using treatment as a factor in the design model. To visualize the significant upregulated (UR) and downregulated (DR) differentially expressed genes (DEGs), the DESeq results were plotted in the form of volcano plots using R. The EggNOG mapper webserver [6970] (accessed on 29 February 2024) was used to annotate functional and protein family (PFAM) descriptions for the protein sequences related to the BSF reference genome. From this, the significant DEGs were labelled with their corresponding EggNOG descriptions and PFAM. STRING [71] was then used to annotate the RefSeq BSF proteome (GCF_905115235.1; string taxid STRG0A70YGC) using the STRING database v12.0. Following this, a protein interaction network was constructed for the male- and female-related DEG protein sequences using the STRING multiple protein search option with the default settings (STRG0A70YGC; accessed on 30 April 2025). The STRING protein search also included a gene ontology (GO) enrichment analysis, as well as Kyoto Encyclopedia of Genes and Genomes (KEGG) and Reactome pathway searches for the DEG proteins. The protein sequences with biological process (BP) GO terms related to immunity were highlighted, and the KEGG and Reactome pathway descriptions assigned to the related DEGs were extracted. Additionally, the relation to immunity for extra DEGs was putatively determined through text mining of the gene/protein names and a literature search. Lastly, the normalized counts from the male and female DESeq results were log2 transformed, and the transformed counts for the DEGs with immune-related BP terms were extracted and used to construct a heatmap with row-scaled Z-scores plotted using pheatmap v1.0.12 [72].

Viral small RNA profiling

The small RNA sequencing datasets were trimmed and mapped to HiSvV and processed using the same approach as above. We used sRNAplot [73] to generate sRNA profiles from the mapped and filtered BAM files, and the script was modified to also produce a csv datafile output. An R script then used the csv file to generate the profile plots and collations.

Results

High abundance of HiSvV reads in the diseased colony

A deep-sequencing shotgun transcriptomic approach was employed [22] to analyse the presence of virus-related reads in a BSFs colony that had high levels of premature adult mortality. Two samples were analysed: one asymptomatic and freshly emerged adult fly [healthy adult (HA)] and one adult fly that had died prematurely [diseased adult (DA)]. In the transcript data of the BSF adults, other than bacteriophage reads, only viral reads related to Solinviviridae were detected in both the HA and DA samples, although with a striking difference in abundances (Fig. 1a). When mapping the reads to the reference sequence for HiSvV (PQ228193), the profiles were similar for both datasets and spanned the entire reference sequence, although the coverage was 500 times higher in diseased samples (coverage in DA=147,343× and in HA=296×; Fig. 1b). This higher viral abundance in the sample derived from the dead adult suggests that HiSvV is responsible for the premature mortality observed in some adults.

Detection of viral sequences in the BSF samples. (a) Visualization of read classifications based on result outputs generated by Lazypipe2 analysis. For both the HA and DA, the minor taxa classifications (encompassing mostly bacteria and archaea) totalled less than 0.03% of transcripts. (b) Mapping of sequencing reads to the HiSvV genome. Annotated polyprotein open reading frame (ORF) (orange) with putative conserved regions of proteins (purple) consists of an RH, P, RdRP, BS and MC. The upper coverage panel displays the HiSvV coverage in the HA, and the lower coverage panel shows the coverage of HiSvV in the DA. Note the difference in y-axis scales for the HA and DA samples, ‘k’ is a place holder for ‘value × 1000’. BS, dsRNA-binding site; DA, dead adult; MC, major capsid; P, protease; RdRP, RNA-dependent RNA polymerase 1; RH, RNA helicase.

Icosahedral-like viral particles were identified in HiSvV-positive samples

To confirm that HiSvV sequences derived from viral particles were infectious to BSFs and to further characterize its viral structure, transmission and pathogenicity, we partially purified the viral particles. Homogenates filtered through gauze (FH) from infected adults were processed to discard cellular debris and generate a partially purified virus (PPV) sample. According to the viral quantification, the purification procedure enriched the sample by ∼4.3-fold, with an increase in the relative abundance of HiSvV from 8,679 to 37,152 (Fig. 2a). Given the prepurification protocol aims at selecting viral particles, finding an enrichment of HiSvV genomic equivalents indicated the viral nature of HiSvV.

HiSvV purification and visualization by TEM. (a) Enrichment of HiSvV genomes after virus partial purification. Stages of the prepurification protocol compared were the initial homogenate filtered through gauze (FH) and the PPV. ‘nd’ represents the lower threshold of detection for HiSvV. (b–d) TEM observation of HiSvV particles in the PPV samples. The scale bars in the bottom right corner, coloured in yellow, represent 100 nm (b and c) and 200 nm (d). PPV, partially purified virus.

Further exploration of the prepurified viral sample using TEM (Fig. 2b–d) led to the visualization of particles resembling the size and structure of other members of the Solinviviridae family [25]. These particles were observed either isolated or aggregating together in a chain-like fashion and presented a similar morphology to other insect-infecting solinviviruses [25272874]. The capsid structure appeared to be spherical with an icosahedral-like shape and a mean diameter of 33.9±3 nm (Fig. 2b–d).

HiSvV replicates in BSFs and has a systemic tropism

Initial trials to orally infect BSF larvae with PPV were unsuccessful, and HiSvV was not detected in adults derived from 2- and 7-day-old larvae exposed to concentrations of 6×10^9^ and 1.5×10^8^ HiSvV genomic equivalents per gram of diet, respectively. In an alternative approach, BSFs in the prepupal stage were injected with the virus, and the viral abundance and tissue tropism were subsequently determined by RT-qPCR in adults which emerged on average 17 days after injection (Fig. 3a and b). While HiSvV inoculation did not clearly impact adult emergence, HiSvV was detected consistently across the lifespan of both male and female adults (, available in the online Supplementary Material; Fig. 3a and b). Despite this, there was no drastic increase in the level of HiSvV genetic material within 13 dpe (Fig. 3a). Mock-infected adults showed no HiSvV presence whether sampled within 24 h or 5 dpe, confirming the specificity of the viral detection (Fig. 3a). Following these observations, adults were dissected at 3 dpe to assess HiSvV tropism in various tissues (Fig. 3b). Although HiSvV was found in all examined tissues, it was less prevalent in the reproductive tracts (including testis and ovaries). These results indicate a systemic infection.

Detection of HiSvV in adults and egg clusters after prepupal injection. (a) Viral relative abundance at different days after emergence from pupae and (b) detection of HiSvV in different tissues of adults processed 3 dpe. The tissues from the head and antennae (H.&A.), wings and halteres (W.&H.), thoracic muscles (T.m), mid- and hindgut with Malpighian tubules (M.&M.t.), reproductive organs (R.o.) and fat bodies (F.b.) were tested as separate samples. Individuals injected with PBS were used as negative controls during inoculation. The error bars represent the sd of the mean on a log scale. Only the upper bars and caps of the sd are shown. (c and d) Testing horizontal and vertical transmission of HiSvV between BSFs. Adults and eggs were obtained during cohabitation experiments, and the level of HiSvV was assessed in the (c) adults and (d) egg clusters. Single male and female individuals were each represented by red dots and blue squares, respectively. The green points represent single egg clusters tested. Individuals inoculated with HiSvV during the prepupal stage are indicated using ‘+V’, while ‘−V’ specifies mock-infected adults. The lower threshold for detection of HiSvV in samples was indicated by ‘nd’ during qPCR testing.

HiSvV is horizontally transmitted between adults and likely vertically transmitted

To determine HiSvV transmission routes, cohabitation experiments were performed using adults that had emerged after HiSvV injection (Fig. 3c and d). Infected males were reared with non-infected females and vice versa. The relative abundance of HiSvV was analysed in the post-mated adult cadavers collected between 14 and 19 dpe, as well as in eggs from these crosses.

All tested adult cadavers from the cohabitation experiments were found to be positive for HiSvV, independently of their sex and their original infection status (Fig. 3c). After cohabitation, the viral relative abundance in adults that were initially uninfected was found to be within two orders of magnitude of the viral titres observed in individuals that were originally infected. Moreover, HiSvV was detected in egg clusters where both parents were infected with HiSvV before cohabitation, and in egg clusters where at least one parent had been infected with HiSvV before cohabitation, independently of the infected sex (Fig. 3d). HiSvV was not detected in the egg clusters from mock-infected parents (Fig. 3d). These cohabitation results indicated that HiSvV is horizontally transmitted between adult BSFs. In addition, since HiSvV was detected in egg clusters, the results also suggested probable maternal as well as paternal vertical transmission.

HiSvV negatively impacts BSF adult survival after oral infection

Since premature mortality of BSF adults was one of the symptoms originally observed, the pathogenic nature of HiSvV was assessed by monitoring the survival of adults orally infected with HiSvV (Fig. 4). Adults were orally inoculated at 1 dpe to assess the effect on their lifespan. The relative abundance of HiSvV in both males and females increased similarly by almost four orders of magnitude, plateauing by 7 days post-inoculation (dpi) (Fig. 4a), confirming the successful oral infection of the adults. Focusing on adult survival, independent of the infective status, females exhibited shorter lifespans than males by 8 days on average (Fig. 4b and c, Tables S1 and S2). However, non-infected females were already exhibiting a hazard ratio of 6.5 (P<0.001) when compared with the lifespan of non-infected males (Table S1A). Despite this, the hazard ratios of infections for both concentrations of HiSvV ranged between 3.5 and 4.9 (P<0.001) when separating by sex, showing similar levels of effect between the two concentrations used (Fig. 4b and c; Table S1B). Overall, survival analyses also showed that for both males and females, HiSvV infection increased the probability of early death by 3.9–4.1 times (P<0.001) (Fig. 4b and c; Table S2).

The effect of HiSvV oral infection on the survival of adult BSFs. (a) Abundance of HiSvV in orally infected adults. Mean relative viral abundance in adults over 7 days post-infection. For time post-inoculation timepoints, ‘h’ indicates hours and ‘d’ indicates days. Different colours represent the females (blue) and males (red). The filled and empty shapes indicate HiSvV- and mock-infected individuals, respectively. Lines indicate HiSvV (solid) and mock (dotted) infections. The upper error bars represent the sd of the mean on a log scale and ‘nd’ indicates the lower detection threshold for HiSvV. (b and c) Impact of HiSvV infection on adult lifespan. Adults were orally infected with HiSvV PPV at two concentrations: 1.6×104 genome equivalents per microlitre for experiment 1 (b) and 1.6×101 genome equivalents per microlitre for experiment 2 (c). Males and females are differentiated by the colours red and blue, respectively, and a solid line separates HiSvV-infected from mock (PBS)-infected individuals. The P-value within each plot was established by a global comparison of strata within each experiment, indicating a statistical difference between all strata.

HiSvV infection triggered a broad immune response in male BSFs

To determine host response to HiSvV infection, differential gene expression analyses were carried out on pooled males and females collected at 5 and 7 dpi. As similar viral titres and gene expression profiles were obtained for 5 and 7 dpi samples, they were used together as biological replicates in subsequent analyses. Overall, DEG analysis showed a broad response to HiSvV infection, with more genes being UR than DR in HiSvV-infected adults (Fig. 5). Initially, comparing significantly DEGs (PAdj<0.05) in infected adults regardless of sex presented 799 UR genes and 486 DR genes (Fig. 5a). However, when separating the males and females, the HiSvV-infected males showed a stronger level of DEGs (∼900) when compared with females (∼250) (Fig. 5, Tables S3 and S4). Although females had 8.5× fewer independent DEGs than males, males and females still shared 128 UR and 27 DR genes (Fig. 5, Tables S3 and S4).

Analysis of transcriptional response to HiSvV infection in BSF adults. (a) Volcano plots depicting DEGs with negative and positive L2FCs. (b) Heatmap of immune-related genes with significant L2FCs contrasting HiSvV- vs mock (PBS)-infected males and females. Males and females underwent DEQseq2 analysis separately for heatmap visualization. The black dots represent genes with significant L2FC in males, and the black dashes for females. The green dashes indicate genes for which the gene count was ‘0’. Sizing indicates sRNA from the HiSvV genome degradation. (c) sRNA reads were mapped to the HiSvV genome in HiSvV- and mock-inoculated BSFs. Profiles were obtained for adults collected at 5 and 7 days post-infection, for both males and females. sRNA sequences which started with an adenine were coloured in green, uracil in blue, thymine in red, guanine in yellow and ambiguous ‘N’ in black. The percentage of mapped sRNA reads to the total number of reads in each sample was indicated in the top right corner of each subplot. The bars above zero represent positive-sense RNA mapping, and the bars below zero represent negative-sense mapping.

Closer examination of the DEGs found that 26.5% of male DEGs and 31.4% of the female DEGs were uncharacterized (Tables S3 and S4). Analysis of BPs through GO enrichment pointed to at least ten genes from the ‘response to viruses’ process (False discovery rate (FDR)=3.9e-04), including eight involved in ‘defence responses to viruses’ (FDR=6.8e-04) in females. Additionally, six of these ten genes were involved in the ‘negative regulation of viral genome replication’ (FDR=3.9e-04) (Fig. S3). For males, there were at least 17 genes involved in ‘responses to viruses’ (FDR=8.8e-05), including 14 involved in ‘defence response against viruses’ (FDR=4.8e-04) and 9 involved in the ‘negative regulation of viral genome replication’ (FDR=4.2e-05) (Figs S3 and S4; Tables S3 and S4). A deeper screening of BPs from the same GO enrichment analysis, as well as text and reference mining of male and female datasets, found that at least 55 DEGs were likely related to immunity (Tables 1, S3 and S4; Figs S3 and S4). The vast majority of these genes (45 in males and 25 in females) were UR, suggesting the triggering of immune-related pathway activations after infection (Fig. 5b, Table 1). Of these 55 DEGs, most were linked to immune-related processes such as the Toll, IMD and apoptosis pathways, with several encoding antimicrobial peptides (AMPs), including attacins and holotricins (Table 1). Infected flies showed sex-specific AMP regulation: two attacin-B-like genes were strongly UR in males and females, whereas holotricin-3-like genes displayed opposite regulation between sexes (Fig. 5b). Males exhibited a broader overall immune response, while the putatively immune-related procathepsin l-like gene was uniquely expressed in females. In addition, several DEGs were associated with apoptotic and autophagy-related processes, as well as interferon-like responses, suggesting that programmed cell death and other antiviral mechanisms may contribute to the defence against HiSvV (Table 1). Only a few DEGs were associated with the RNAi pathway, including argonaute-2-like and dicer, questioning the potential influence of the RNAi mechanism against HiSvV infection (Table 1).

Table 1.: Putative Reactome and KEGG pathway annotations of significant (PAdj<0.05) DEGs related to immune-linked BPsBoth KEGG and Reactome annotations were statistically significant (P<0.05). Genes coloured in red had an L2FC above 0, and those in blue had an L2FC below 0. ‘n/o’ indicates no significant L2FC observed.

Absence of siRNA response to HiSvV infection in adult BSFs

Although only argonaute-2 and endoribonuclease dicer, potentially involved in the RNAi defence pathway, were found to be regulated, we decided to explore if this defence mechanism was triggered by HiSvV infection. sRNA profiles were thus examined after HiSvV infection (Fig. 5c). Mapping of sRNA reads to the HiSvV genome showed an increase in counts and in the percentage of mapped reads from 5 to 7 dpi (Fig. 5c), further supporting active viral infection as detected earlier (Figs 3a and 4a). However, the broad size range from 19 to 31 nt, targeting the HiSvV viral genome, without a prominent peak at 21 nt, suggested that the typical small interfering RNA (siRNA) antiviral pathway was not activated (Fig. 5c). The widespread distribution of the sRNA profile observed rather indicated the degradation of HiSvV genomic material and suggested an alternative defence response.

Discussion

Pathology research in mass-reared insects for food and feed has garnered more attention within the last few years, and BSFs are no exception [1261675]. While advances in virus research have been made in other mass-reared insects such as crickets and mealworms, they are just developing in BSFs [116197677]. Our study was the first to isolate a virus from diseased BSFs and to demonstrate its pathogenicity in this species, a notable finding given the economic importance of BSFs [478].

We received BSF adult samples from a rearing facility experiencing high levels of adult mortality and reduced egg production. The presence of HiSvV genetic material was linked with the presence of viral particles, confirming it as a viral agent with a morphology consistent with members of the Solinviviridae family [25]. The finding of HiSvV infections presenting as systemic infections is consistent with other Solinviviridae studied so far [2774]. Horizontal transmission of HiSvV was demonstrated through successful experimental oral infection of BSF adults, as well as through cohabitation experiments in which HiSvV was detected in exposed adults and in the egg clusters. Although the presence of HiSvV in egg clusters suggests the potential for vertical transmission, the systemic tropism favouring the digestive tract indicates that the typical infection occurs mainly through the oral-faecal route. This was further confirmed by oral inoculation of adults, as also found for a closely related solinvivirus, SINV3 [3379]. Whether HiSvV can sustainably replicate in larvae remains unresolved, as is the case for other solinviviruses [272837].

HiSvV infection in adults reduced the lifespan of both males and females. This is the first study examining the effect of a solinvivirus infection on sexes, contrasting with prior studies performed on unsexed larvae of shrimp or on female castes and larvae in ants and bees [2830,32]. Other Solinviviridae, such as SINV3, AmSV1 and PvSV, also cause premature mortality that has been linked to colony collapse in ants, honeybees and whiteleg shrimp [293032]. In addition, the risk of colony collapse due to HiSvV infection may be reinforced if BSFs die before mating or oviposition, particularly if virus levels are allowed to build up within the rearing facility [16]. This risk would be more pertinent for continuous rearing setups, as mating can start at least 3–4 dpe [8081].

To better understand host–pathogen interactions and potential sex-specific effects, we explored the putative immune response in both males and females using DEG analysis and sRNA profiling. The DEG analysis revealed transcriptional changes in antiviral immunity-related genes (Figs S3 and S4, Tables S3 and S4), indicating a clear immune-related response to an entomopathogenic virus introduced at a relatively low concentration. Of note, melanization did not seem to be activated by HiSvV infection; however, such non-specific immune responses may not be induced or may not be very efficient against viral infections [468283]. The sRNA profiles from HiSvV-infected flies were broad rather than displaying specific peaks, suggesting that BSFs did not mount a specific silencing response against HiSvV. This pattern implies that the virus may evade sRNA pathways, a strategy known in other insect-infecting Solinviviridae and Dicistroviridae, which can avoid RNAi silencing through a dsRNA-binding domain on their genomes [28,4648]. HiSvV also encodes a homologue of this dsRNA-binding domain, supporting the interpretation that the broad sRNA profiles reflect general viral genetic material degradation rather than specific RNAi or P-element induced wimpy testis (PIWI) RNA responses. Males exhibited a broader immune response than females, and they generally survived longer, suggesting that a major activation of the immune defences may contribute to reducing the detrimental effects of the viral infection. Similarly, differences in immune activity between sexes have been observed in other flies, e.g. Drosophila species, reinforcing the need to examine the impacts of BSF viruses in both sexes when possible [46,8488].

It remained to be determined whether the UR AMPs in our study were directly involved in antiviral defence or instead targeted opportunistic microbiota [14284648]. While not all AMPs act against viruses, some cecropins, attacins and defensins are known to participate in antiviral responses [4689]. Both sexes showed significantly higher L2FC values for attacin-B-like genes. Holotricin-3, which has antimicrobial properties against bacteria and fungi but unclear antiviral activity [9091], was DR in males yet UR in females, possibly reflecting a sex-specific role in HiSvV defence. The upregulation of genes linked to recognition, signalling and other processes in the IMD and Toll pathways further supports a putative antiviral role for effectors linked to these pathways, such as attacin-B-like and holotricin-3-like (putative defensin), in BSFs–HiSvV interactions [469092]. The apparent number of genes associated with programmed cell death, autophagy and interferon-like responses suggests that these processes may play a role in degrading HiSvV-infected cells, which could explain the broad, degraded sRNA profiles. These responses are defence mechanisms likely effective against pathogens that evade RNAi and AMPs and are particularly relevant for insect-infecting picornaviruses such as Dicistroviridae and Solinviviridae [28,4648]. The putative pathway-level patterns identified here may benefit from future validation by RT-qPCR using established indicator genes. Despite experimental limitations, this study provides an initial framework to understand antiviral responses in BSFs and how the solinvivirus HiSvV may evade host immunity.

Contrary to the ‘popular belief’ that BSFs were insensitive to viral infections, this study showed for the first time that a recently discovered virus, HiSvV, can significantly induce premature mortality of BSF adults and is likely the culprit behind a reported loss of production within a BSF farm. Given the capacity of HiSvV to decrease production, this prompts the need to develop management plans and surveillance tools to prevent future outbreaks and interfacility transfer. Lastly, this work represents an important milestone in BSF pathology since it is the first study to show that viruses could be a threat to BSF rearing facilities.

Supplementary material

10.1099/jgv.0.002234Uncited Supplementary Material 1.

10.1099/jgv.0.002234Uncited Supplementary Material 2.