Characterization of a Novel Mitovirus of the Sand Fly Lutzomyia longipalpis Using Genomic and Virus-Host Interaction Signatures

Abstract

Hematophagous insects act as the major reservoirs of infectious agents due to their intimate contact with a large variety of vertebrate hosts. Lutzomyia longipalpis is the main vector of Leishmania chagasi in the New World, but its role as a host of viruses is poorly understood. In this work, Lu. longipalpis RNA libraries were subjected to progressive assembly using viral profile HMMs as seeds. A sequence phylogenetically related to fungal viruses of the genus Mitovirus was identified and this novel virus was named Lul-MV-1. The 2697-base genome presents a single gene coding for an RNA-directed RNA polymerase with an organellar genetic code. To determine the possible host of Lul-MV-1, we analyzed the molecular characteristics of the viral genome. Dinucleotide composition and codon usage showed profiles similar to mitochondrial DNA of invertebrate hosts. Also, the virus-derived small RNA profile was consistent with the activation of the siRNA pathway, with size distribution and 5' base enrichment analogous to those observed in viruses of sand flies, reinforcing Lu. longipalpis as a putative host. Finally, RT-PCR of different insect pools and sequences of public Lu. longipalpis RNA libraries confirmed the high prevalence of Lul-MV-1. This is the first report of a mitovirus infecting an insect host.

Keywords: Mitovirus; RNA interference; codon usage; mitochondrial viruses; nucleotide frequency; positive-sense single-stranded RNA viruses; profile Hidden Markov Model; seed-driven progressive assembly.

Figure 1

Small RNA size distribution of putative viral sequences assembled from sandfly RNA libraries. Clustering is shown as a heatmap based on Pearson correlation of the sRNA profile. Clusters are defined by a Pearson correlation above 0.8. Viral contigs were classified according to the small RNA size distribution expected for siRNA and piRNA populations in insects. Shaded columns represent small RNAs with length between 19 and 23 nt.

Figure 2

Phylogenetic analysis. Phylogenetic tree inferred by maximum likelihood using full-length amino acid sequences of the RNA-dependent RNA polymerase (RdRp). Phylogenetic reconstruction was performed with IQ-TREE using the evolutionary model Blosum62 + F + R6 and 1000 bootstrap pseudoreplicates. The tree was rooted using two sequences of ssRNA bacteriophages belonging to Leviviridae family. Colored clades correspond to the familis Mitoviridae (blue) and Narnaviridae (green), and to Botourmiaviridae/ourmia-like viruses (red).

Figure 3

Comparative genome length. Comparison of genome and RNA-dependent RNA polymerase (RdRp) lengths of viruses of the genera Narnavirus and Mitovirus.

Figure 4

Dinucleotide frequency of Lul-MV-1 and nuclear genomes of fungi that commonly infect insects. Yellow circles represent dinucleotide odds ratios which are highly biased regarding the expected frequencies. Purple circles refer to dinucleotide odds ratios within the unbiased region (delimited by dashed lines). Dashed lines indicate cutoffs values of 0.78–1.25.

Figure 5

Relationship between codon-usage of mitochondrial viruses and mitochondrial genomes (MT) of putative hosts. Hierarchical clustering of codon usage profiles of viruses and mitochondria of their respective hosts. Clusters were defined by Pearson correlation above 0.8. The numbers indicate the main clades referred in the text.

Figure 6

Characterization of small RNAs derived from Lu. longipalpis and fungi-related viruses. (A) Small RNA size profiles and 5′ base preference of small RNAs derived from Lul-MV-1 and (B) fungi-related viruses (from top to bottom: Aspergillus fumigatus mycovirus, Botrytis paeoniae virus and Sclerotinia sclerotiorum hypovirus). 5′ base preferences of small RNAs are indicated by color. Significant differences are also indicated. (C) Cumulative frequency according to the size distribution of small RNAs varying from 15 and 35 nt, derived from viruses infecting Lutzomyia (Lul-MV-1 and VSV) and fungi. Statistical significance among cumulative frequencies was determined using the Kolmogorov-Smirnov test.

Figure 7

Detection of Lul-MV-1. (A) RT-PCR amplification of a Lul-MV-1 target using Lu. longipalpis RNA from different pools of individuals derived from the same colony used to prepare the small RNA deep sequencing libraries. The arrow indicates the amplification product. (B) Reads from different RNA and DNA public libraries of Lu. longipalpis mapped onto the genome of Lul-MV-1 with Bowtie2, quantified and resulting numbers normalized by Reads Per Million (RPM).