Non-retroviral Endogenous Viral Element Limits Cognate Virus Replication in Aedes aegypti Ovaries

Abstract

Endogenous viral elements (EVEs) are viral sequences integrated in host genomes. A large number of non-retroviral EVEs was recently detected in Aedes mosquito genomes, leading to the hypothesis that mosquito EVEs may control exogenous infections by closely related viruses. Here, we experimentally investigated the role of an EVE naturally found in Aedes aegypti populations and derived from the widespread insect-specific virus, cell-fusing agent virus (CFAV). Using CRISPR-Cas9 genome editing, we created an Ae. aegypti line lacking the CFAV EVE. Absence of the EVE resulted in increased CFAV replication in ovaries, possibly modulating vertical transmission of the virus. Viral replication was controlled by targeting of viral RNA by EVE-derived P-element-induced wimpy testis-interacting RNAs (piRNAs). Our results provide evidence that antiviral piRNAs are produced in the presence of a naturally occurring EVE and its cognate virus, demonstrating a functional link between non-retroviral EVEs and antiviral immunity in a natural insect-virus interaction.

Keywords: Aedes aegypti; CRISPR-Cas9; RNA interference; cell-fusing agent virus; endogenous viral element; insect immunity; insect-specific flavivirus; mosquito; piRNA pathway; siRNA pathway.

Graphical abstract

Figure 1

CFAV-Derived Endogenous Viral Elements Interact with Natural CFAV Infection through the piRNA Pathway (A) The schematic represents two potential CFAV EVE structures detected in publicly available Ae. aegypti sequences. (B) The presence of putative CFAV-EVE1 and CFAV-EVE2 in eight mosquitoes from the same outbred colony was verified by PCR with primers specific to CFAV-EVE1 (left), CFAV-EVE2 (middle), and rps7 gene control (right). (C and D) Size distribution of sRNAs mapping to the CFAV genome from naturally CFAV-uninfected (C) and CFAV-infected (D) mosquitoes from the outbred colony. (E and F) Analysis of CFAV-derived piRNAs from naturally CFAV-uninfected (E) and CFAV-infected (F) mosquitoes from the outbred colony. Mapping of 26- to 30-nt sRNAs (top), sequence logos of 26- to 30-nt sRNAs (bottom left), and overlap probability of 26- to 30-nt sRNAs (bottom right) is shown. Sequence logos and overlap probability for CFAV-EVE1 were restricted to the NS2 region. In (C)–(F), positive- and negative-sense reads with respect to the reference CFAV genome are shown in yellow and blue, respectively. Uncovered nucleotides are represented by gray lines. See also Data S1, Figure S1, Table S1, Table S2, and Table S5.

Figure 2

CFAV-EVE1 Interacts with Experimental CFAV Infection through the piRNA Pathway (A) Schematic of the CFAV-EVE1 structure in the CFAV-free isofemale line represented as the alignment of the EVE locus in the Ae. aegypti genome assembly AaegL3 (top) to the genome of the CFAV-KPP isolate (bottom). CFAV-EVE1 comprises four different regions of the CFAV genome. Yellow and blue colors indicate forward and reverse strands, respectively, according to the transcription direction in the supercontig. (B) Production of piRNAs from CFAV-EVE1 in the CFAV-free isofemale line, represented as the size distribution (left) and alignment to the CFAV-EVE-1 locus (right). Blue color corresponds to negative-sense reads with respect to the mapping reference. (C and D) Size distribution of sRNAs mapping to the CFAV genome from experimentally CFAV-uninfected (C) and CFAV-infected (D) mosquitoes from the isofemale line. (E and F) Analysis of CFAV-derived piRNAs from experimentally CFAV-uninfected (E) and CFAV-infected (F) mosquitoes from the isofemale line. Mapping of 26- to 30-nt sRNAs (top), sequence logos of 26- to 30-nt sRNAs (bottom left), and overlap probability of 26- to 30-nt sRNAs (bottom right) is shown. Sequence logos and overlap probability were restricted to the NS2 region. In (C)–(F), positive- and negative-sense reads with respect to the reference CFAV genome are shown in yellow and blue, respectively. Uncovered nucleotides are represented by gray lines. See also Figure S2 and Table S5.

Figure 3

CRISPR-Cas9-Mediated Genome Editing of CFAV-EVE1 in Aedes aegypti (A) Deletion of the CFAV-EVE1 from the Ae. aegypti genome of the CFAV-free isofemale line using CRISPR-Cas9. The upper bar represents the CFAV-EVE1 with the flanking regions, and the three sgRNA target sites are shown with scissors. The lower bar represents the merged flanking regions without the CFAV-EVE1, where the short repeat sequences in the flanking regions (yellow segments on both bars) are merged into one. (B) Generation of the CFAV-EVE1 (+/+) and (−/−) Ae. aegypti lines after CRISPR-Cas9-mediated genome editing. A single G0 male mosquito heterozygous for the CFAV-EVE1 deletion (+/−) was outcrossed with wild-type females harboring the CFAV-EVE1. The resulting heterozygous male G1 progeny was outcrossed with wild-type females harboring the CFAV-EVE1. The G2 heterozygotes of both sexes were intercrossed to produce a mixed G3 progeny that was sorted into pure homozygous CFAV-EVE1 (+/+) and (−/−) lines. The letter G denotes the generation of mosquitoes originating from the CFAV-EVE1 heterozygous male and wild-type females. The letter F denotes the generation of the CFAV-EVE1 homozygous lines. The agarose gel picture represents a fraction of samples genotyped at G3, where the pure homozygous individuals were selected by PCR genotyping of a single leg. See also Table S1 and Table S3.

Figure 4

Ablation of CFAV-EVE1 Prevents CFAV-Derived piRNA Amplification (A, B, E, and F) Size distribution of sRNAs mapping to the CFAV genome in ovaries (A and B) and heads (E and F) from experimentally infected CFAV-EVE1 (+/+) (A and E) and CFAV-EVE1 (−/−) (B and F) mosquitoes 7 days post-injection. (C, D, G, and H) Analysis of CFAV-derived piRNAs in ovaries (C and D) and heads (G and H) from experimentally infected CFAV-EVE1 (+/+) (C and G) and CFAV-EVE1 (−/−) (D and H) mosquitoes 7 days post-injection. Mapping of 26- to 30-nt sRNAs (top), sequence logos of 26- to 30-nt sRNAs (bottom left), and overlap probability of 26- to 30-nt sRNAs (bottom right) is shown. Sequence logos and overlap probability were restricted to the NS2 region. In all panels, positive- and negative-sense reads with respect to the reference CFAV genome are shown in yellow and blue, respectively. Uncovered nucleotides are represented by gray lines. See also Figures S3, S4 and Table S5.

Figure 5

CFAV-EVE1 Ablation Results in Increased CFAV RNA Levels upon Viral Infection Relative CFAV RNA levels (normalized by the rp49 housekeeping gene) in heads and ovaries of the CFAV-EVE1 (+/+) (black boxplot) and CFAV-EVE1 (−/−) (white boxplot) Ae. aegypti lines on day 4 (A) and day 7 (B) post-CFAV-inoculation. Data are shown for six separate experiments represented by color- and symbol-coded data points. Relative viral RNA loads are represented by boxplots in which the box denotes the median and interquartile range (IQR) and the whiskers extend to the highest and lowest outliers within 1.5 times the IQR from the upper and lower quartiles, respectively. Multivariate analysis of variance (MANOVA) was performed for each time point and tissue separately, accounting for the experiment, mosquito line, and interaction effects. Stars indicate statistical significance of the mosquito line main effect accounting for the experiment effect (^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001; ns, not significant). The full MANOVA results are provided in Table S4. See also Table S1.

Figure 6

Model for the Antiviral Role of Non-retroviral EVEs in Mosquitoes Both a naturally occurring EVE (left panel) and exogenous viral infection (middle panel) produce primary piRNAs in antisense and sense orientation, respectively. Only when EVE and virus are present in the same mosquito do piRNAs acquire antiviral activity (right panel) through EVE-derived piRNAs targeting the viral genome. Under this model, integration of non-retroviral sequences into the host genome, their transcription into piRNA precursors, and their processing into antiviral piRNAs are mechanisms by which EVEs confer heritable, sequence-specific host immunity.