Preparation of public RNA-seq datasets

A total of 148 public RNA-seq datasets were downloaded from NCBI Sequence Reads Archive (SRA) with accession number SRP219966, [17] SRP050258 [18], SRP018112 [19], SRP077936 [20], SRP056407 [21], and SRP164784 [22]. All 148 RNA-seq libraries used in this study were listed in S1 Table. FASTQC was used to check the quality of raw reads. Adapters and low quality reads (<20 Phred score) were removed using Trimmomatic version 0.38 [23].

Mapping and transcriptome assembly

Reads were individually mapped onto the Ae. albopictus genome (Genome version: AalbF2, assembly: GCA_006496715.1, NCBI) using Hierarchical indexing for spliced alignment of transcripts (HISAT2) version 2.1.0 [24]. The resulting alignment files were used as input in Stringtie software (version 1.0.1) for transcriptome assembly [25]. Annotation file (GTF format) of Ae. albopictus from NCBI (Aalbo_primary.1_genomic.gtf) was used to guide the transcriptome assembly. A minimum of 200 bp was set for the length of the assembled transcripts. Then, the Stringtie output files were merged into a single transcriptome assembly using Stringtie merge. The resulting merged assembly was then compared to the reference annotation using Gffcompare (https://github.com/gpertea/gffcompare). Transcripts with class code “u” (intergenic), “i” (intronic), and “x” (antisense) were kept for lncRNA prediction. These class codes were chosen because lncRNAs were categorized based on their positions relative to protein-coding genes provided in the reference annotation [14,26–29].

Novel lncRNA prediction

TransDecoder [30] was used to predict the presence of long open reading frame (ORF) within transcripts with class code “u”, “i”, and “x”. A minimum of 100 amino acids was set as threshold of ORF length in TransDecoder analysis, and transcripts having long ORF were removed. Long ORF was defined as transcripts having more than 100 amino acids. The remaining transcripts were then subjected to coding potential assessment analysis. Coding Potential Calculator 2 (CPC2) [31], Predictor of long noncoding RNAs and messenger RNAs based on an improved k-mer scheme (PLEK) [32], and Coding Potential Assessment Tool (CPAT) [33] were used to evaluate coding potential of the transcripts. For CPAT analysis, a coding probability threshold of < 0.3 was set [14,15]. Transcripts that passed CPAT threshold, and detected as noncoding by CPC2 and PLEK, were kept for subsequent analyses. To avoid false positive prediction, BLASTX algorithm was used to search for sequence similarity of the potential lncRNA transcripts in Swissprot database. Transcripts with high sequence similarity (E-value < 1e-5) were removed. Transcripts without strand information, and those located less than 1 kb from protein-coding genes on the same strand were removed. Finally, to remove lowly expressed transcripts, transcripts that had an expression of more than 1 transcript per million (TPM) in at least 5 RNA-seq libraries were kept. The expression was determined by featureCounts [34], and alignment files (BAM format) generated by HISAT2 were used as input.

Cell culture and virus

C6/36 cells (ATCC: CRL-1660) were cultured in Leibovitz’s L-15 medium (Gibco, 41300039) supplemented with 10% Fetal Bovine Serum (FBS, Gibco, 10270) and 10% Tryptose Phosphate Broth (TPB) solution (Sigma, T9157). The C6/36 cells were grown at 25°C without CO₂. BHK-21 cells (ATCC: CCL-10) were used for virus titration. BHK-21 cells were cultured in Dulbecco’s modified Eagles Medium (DMEM, Gibco, 11995065) supplemented with 10% FBS (Gibco, 10270) at 37°C and 5% CO₂. In this study, DENV serotype 1 (DENV-1) of Hawaiian strain, and ZIKV of strain H/PF/2013 were used. Both viruses were propagated in C6/36 cells. Virus titer of DENV and ZIKV was determined using tissue culture infectious dose 50 (TCID50) assay. DENV-1 (Hawaiian strain) was a gift from David Perera, University Malaysia Sarawak, and ZIKV (H/PF/2013 strain) was a gift from Shee Mei Lok, Duke-NUS Medical School, Singapore.

Tissue Culture Infectious Dose 50 (TCID50) assay

TCID50 assay was conducted to determine infectious virus titre using a method that has been previously described [35–37]. For each determination, a 96-well plate culture of confluent BHK-2 was inoculated with a 150 μl mixture of 100 μl medium and 50 μl of serial 10-fold dilutions of clarified cell culture supernatant media. Each row included 6 wells of the same dilution and two negative controls. Plates were incubated at at 37°C and 5% CO₂ for 5 days. Each well was observed for the presence or absence of cytopathic effect (CPE). The determination of TCID50/ml was performed using the Reed and Muench method [38].

Virus infection, RNA extraction and sequencing

C6/36 cells were infected with DENV-1 and ZIKV at a multiplicity of infection (MOI) of 0.25. At 3 days post infection (dpi), the culture medium was removed, and total RNA extraction was carried out using miRNeasy Mini Kit 50 (Qiagen, 217004) according to the manufacturer’s protocol. RNA concentration was determined using ND-2000 instrument (NanoDrop Technologies), and RNA integrity was determined using 2100 Bioanalyzer (Agilent). Libraries were constructed using Illumina TruSeq RNA Sample Prep kit (Illumina, San Diego) using one microgram (μg) of total RNA as starting material. The prepared libraries were sequenced on the Illumina HiSeq 2000 platform to generate 150 bp paired-end reads. Raw reads were deposited in NCBI with accession SRP221722 (https://www.ncbi.nlm.nih.gov/sra/?term=SRP221722).

Identification of differentially expressed lncRNAs upon infection

RNA-seq libraries generated in this study were mapped to the Ae. albopictus genome (AalbF2, assembly: GCA_006496715.1, NCBI) using HISAT2 [24]. The expression was quantified using featureCounts [34]. GTF file annotated with the newly predicted lncRNA was used as reference. Read counts computed by featureCounts were then subjected to differential expression analysis using edgeR [39] in R/Bioconductor environment.

Reverse transcription and real-time quantitative PCR (qPCR)

cDNA synthesis was performed with one μg of total RNA in 20 μl reactions using the iScript Reverse Transcription Supermix (Bio-Rad Laboratories, California; cat. no. 1708840). Quantitative PCR was performed using iTaq™ Universal SYBR Green Supermix (Bio-Rad Laboratories, California; cat. no. 1725120), in a total of 20 μl reactions. All experiments were done in three biological and technical replicates. RPS17 and RPL32 were used as reference genes, and all qPCR runs were carried out on the BioRad CFX96 qPCR platform. The 2^−ΔΔCT [40] method was used to estimate the relative expression of each lncRNA, and the results were statistically analyzed using student’s t-tests. List of primers used in this study can be found in S8 Table.

Functional annotation and enrichment

Functional annotation was performed using Pannzer [41] using amino acid sequences of lncRNA-target genes as input. Gene enrichment analysis was done using g;Profiler [42] and g:SCS threshold was used for multiple testing correction.

CRISPR-Cas9 induced mutation of selected lncRNAs

Single guide RNA (sgRNA) targeting exon 1 of lncRNA_27639.2 and XR_003899061.1 were designed using CHOPCHOP [43]. The sgRNAs were cloned into the Drosophila CRISPR-Cas9 vector pAc-dU6-sgRNA-Cas9 [44], which was a kind gift from Ji-Long Liu (Addgene plasmid 49330). This plasmid has been shown to work well in Ae. aegypti (Aag2) cell [45]. Detailed protocol for sgRNA cloning was previously described [44]. C6/36 cells (~3x10⁶) were transfected with 2.5 μg of gRNA-containing plasmid using Lipofectamine 2000 (Life Technology). Plasmid without gRNA (empty vector) was used as control. After 72 hours post-transfection, Cas9-expressing cells were selected using puromycin (5 μg/mL) for 5 days. Cells were then resuspended and ~5x10⁵ cells were added to each well of 12-well plate. Cells were allowed to settle for ~2 hour, before being infected with DENV-1 (Hawaiian strain) and ZIKV (H/PF/2013 strain) at MOI of 0.25. After 3 dpi, cells were frozen and thawed twice, causing the cells to rupture and release viral particles. The samples were then centrifuged, and the cell culture supernatant containing viral particles were used for TCID50 assay. Expression level of lncRNA after CRISPR-Cas9 induced mutation was evaluated by qPCR. RNA isolation and reverse transcription were carried out as previously described. Meanwhile, indel mutation within lncRNA loci was confirmed by sequencing of the PCR products spanning the cleavage sites. The PCR products were cloned into pGEM-T Easy vector (Promega, cat. no: A1360), and ten colonies were selected for sequencing.

Identification of novel lncRNAs in Ae. albopictus

To identify novel lncRNAs in Ae. albopictus, a total of 148 RNA-seq libraries were used as raw inputs. The list of all 148 RNA-seq datasets used in this study can be found in S1 Table. The RNA-seq libraries used were derived from many tissues, organs, developmental stages and cell lines; hence, allowing our lncRNA prediction to be more comprehensive. The lncRNA prediction pipeline used here (Fig 1) were modified and adapted from previous works [14,15,26,46–51]. After mapping raw reads to the reference genome using HISAT2, we assembled and merged the assemblies into a single unified transcriptome. Using Gffcompare, we selected transcripts with class code “i”, “u”, and “x” (35,863 transcripts), which indicate intronic, intergenic, and antisense to reference genes respectively [14,26]. In this study, we were able to recover all Ae. albopictus annotated transcripts including known lncRNAs (7,609) using Stringtie; hence, validating our transcriptome assembly method.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Overview of lncRNA identification pipeline.

https://doi.org/10.1371/journal.pntd.0008351.g001

We then filtered the transcripts harboring open-reading frame (ORF) and coding potential. The ORF prediction was performed using TransDecoder [30], while coding potential assessment was performed using three different algorithms, namely CPC2 [31], CPAT [33], and PLEK [32]. A total of 22,057 transcripts did not have an ORF as determined from TransDecoder. Of these, 16,630 transcripts passed coding potential assessment by CPC2, CPAT, and PLEK. To avoid false positive prediction, we performed BLASTX on these 16,630 transcripts against Swissprot database. We removed transcripts having significant BLASTX hits (E-value < 10³). Next, we removed unstranded transcripts and those located less than 1 kb from protein-coding genes. Finally, we only retained transcripts that had minimum expression of 1 TPM in at least five out of the 148 RNA-seq libraries used in this study. This was done to remove lowly expressed transcripts which might be falsely identified as novel lncRNAs. In total, 10, 867 putative novel lncRNA transcripts, which derived from 9,298 genomic loci, were identified in Ae. albopictus genome. Of these 10,867 transcripts, 5,809 were located in the intergenic regions, 4,139 were intronic, while the remaining 919 were antisense to annotated genes. Full list of predicted novel lncRNAs and their genomic coordinates can be found in S2 Table. Current Ae. albopictus annotation catalogs 7,609 lncRNA transcripts, and we classified these transcripts as known lncRNAs. Therefore, in Ae. albopictus genome, the number of both novel and known lncRNAs were 18,476.

Ae. albopictus lncRNAs shared similar features with other species

Next, we determined the similarities of Ae. albopictus lncRNAs to other species, such as shorter in length, low GC content, and low sequence conservation [14,26,46,47]. Ae. albopictus lncRNAs were found to be significantly shorter in size compared to protein-coding transcripts (Kolmogorov-Smirnov test (KS-test) p-value < 2.2e-16, Fig 2A). The average length of novel and known lncRNAs in Ae. albopictus were 569.67 bp and 763.68 bp respectively, whereas protein-coding transcripts had a mean size of 2650.1 bp. Short in size is a typical feature of lncRNAs; hence, our findings are congruent with lncRNAs identified in other living organisms [14,46,52]. Besides, Ae. albopictus known and novel lncRNAs exhibited slightly lower GC content in comparison to protein-coding transcripts (Fig 2B). This lower GC or AT enriched characteristic is another notable feature of lncRNA transcripts [14,15,26,46,47]. Mean GC content of novel lncRNAs, known lncRNAs and protein-coding transcripts were 42%, 44%, and 48% respectively. We also analyzed GC content of other noncoding sequences in the genome such as 5’UTR and 3’UTR. We discovered that both of them had lower GC content in comparison to protein-coding transcripts. The average GC content of 5’UTR and 3’UTR were 44% and 36% respectively.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Length and GC content Ae. albopictus lncRNAs.

A) Sequence length distribution of lncRNA and protein-coding transcript. B) GC content of lncRNA, protein-coding transcript, 3’UTR, and 5’UTR.

https://doi.org/10.1371/journal.pntd.0008351.g002

As previous studies showed that lncRNAs demonstrate low sequence conservation even among closely related species [14,15,48], we investigated the validity of this in Ae. albopictus. To assess the level of sequence similarity, we performed used BLASTN algorithm to investigate the level of similarity among Ae. albopictus lncRNAs sequences with other closely related insect genomes such as Ae aegypti, D. melanogaster, Anopheles gambiae (An. gambiae), Anopheles darlingi (An. darlingi), and Culex quinquefasciatus (Cx. quinquefasciatus) [14,15]. BLAST algorithm bit score was used as an indicator for sequence similarity [14,15]. We used BLASTP algorithm to assess the level of similarity among Ae. albopictus proteins with other insects including Ae. aegypti, D. melanogaster, An. gambiae, An. darlingi, and Cx. quinquefasciatus In general, protein-coding transcripts were found to be more conserved than lncRNAs (Fig 3A). Some Ae. albopictus lncRNAs showed high degree of sequence similarity (E-value < 10−⁵) with Ae. aegypti, presumably because the two species are very closely related. (Fig 3B). However, the fraction of Ae. albopictus lncRNAs that displayed high level of sequence similarity (E-value < 10−⁵) with Ae. aegypti were quite low. For instance, only 3,676 (33.8%) novel lncRNAs and 2,053 (29%) known lncRNAs shared high sequence similarity with Ae. aegypti. On the other hand, 38,790 (96.7%) Ae. albopictus proteins demonstrated high degree of sequence similarity with Ae. aegypti protein sequences.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Conservation analysis of Ae. albopictus lncRNAs.

A) Similarity bit score of of lncRNA and protein sequences. In general, protein-coding mRNA is more conserved than lncRNA. B) The Venn diagram shows the number of Ae. albopictus lncRNAs that display high conservation with other species according to BLASTN algorithm (E-value < 10-e5).

https://doi.org/10.1371/journal.pntd.0008351.g003

Overview of sequencing data generated from DENV and ZIKV-infected Ae. albopictus cells

Studies on Ae. aegypti transcriptomes provide evidence that lncRNAs could be involved in DENV and ZIKV-mosquito interaction [15,16]. To examine the relevance of this to Ae. albopictus, we generated paired-end RNA-seq libraries from larval-derived Ae. albopictus cell line (C6/36) which had been infected with DENV and ZIKV. A total of 9 samples were sequenced using Illumina HiSeq 2000 platform, and they comprised of three replicates of mock, DENV, and ZIKV-infected C6/36 cells. In total, 426 million paired-end reads were generated in this study (S3 Table). More than 70% of the reads (average of 35.5 million reads per sample) mapped to Ae. albopictus reference genome (S3 Table), providing sufficient coverage for genome-wide transcriptomic analysis, specifically in detecting differentially expressed transcripts. Hierarchical clustering based on the expression values (CPM) of all the expressed genes showed that the replicates were grouped into their three distinct groups (S1 Fig). This unsupervised clustering also showed that the data formed two main clusters, i.e., uninfected, and virus-infected (DENV and ZIKV) samples.

Differential expression of Ae. albopictus lncRNAs upon DENV and ZIKV infection

In this study, we considered genes with fold change (FC) of more than or equal to 2 in either direction, and false discovery rate (FDR) of less than 0.05 as differentially expressed. A total of 433 and 338 differentially expressed lncRNAs (FC ≥ |2|, FDR < 0.05) were identified respectively in DENV and ZIKV samples. In DENV-infected cells, 298 lncRNAs were found to be upregulated, while the remaining 136 transcripts were downregulated. Meanwhile, 216 and 122 lncRNAs were upregulated and downregulated respectively in C6/36 cells infected with ZIKV. Volcano plot analysis was used to depict the distribution of lncRNAs with their corresponding P-value and log FC (Fig 4A), and hierarchical clustering analysis was performed to cluster the differentially expressed lncRNAs among different replicates (Fig 4B). Similar lncRNAs were found to be differentially expressed; 11 lncRNAs were upregulated and 12 lncRNAs were downregulated in both DENV and ZIKV infected cells (Table 1). The similar changes seen in these 23 lncRNAs post-infection suggest that they may be key players, regardless of the type of viral infection. On the other hand, the number of differentially expressed protein-coding mRNAs were higher, i.e., 1,718 and 1,098 mRNAs were differentially expressed (FC ≥ |2|, FDR < 0.05) in DENV and ZIKV-infected cells respectively. List of differentially expressed lncRNAs can be found in S4 Table.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. lncRNA expression profiles of DENV and ZIKV-infected Ae. albopictus cells.

A) Volcano plot of differentially expressed Ae. albopictus lncRNAs in DENV-1 and ZIKV-infected C6/36 cells. Dots with red color represent lncRNAs with P-value < 0.05. B) Hierarchical clustering of differentially expressed Ae. albopictus lncRNA was done in Morpheus (https://software.broadinstitute.org/morpheus), using one minus Pearson correlation metric with average linkage method.

https://doi.org/10.1371/journal.pntd.0008351.g004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Differentially expressed lncRNAs that are common in DENV-1 and ZIKV transcriptome.

https://doi.org/10.1371/journal.pntd.0008351.t001

We randomly selected ten differentially expressed lncRNAs from RNA-seq analysis for qPCR. The ten chosen candidates were upregulated or downregulated in both DENV and ZIKV-infected samples. We validated the differentially expressed lncRNAs using qPCR from the same RNA samples used for Illumina sequencing. RPS17 and RPL32 were used as reference genes [53,54]. Our preliminary tests showed that the expression level of the two reference genes were unchanged in DENV, and ZIKV-infected C6/36 cells; hence, qualifying them as suitable references in this study. We also performed qPCR on subsets of lncRNAs that were differentially expressed only in DENV-infected cells. Similar experiments were done for ZIKV-infected cells. Overall, the qPCR results of all lncRNAs were consistent with RNA-seq data; hence, validating the bioinformatic approach used in this study (S2 Fig).

Functional analysis of the differentially expressed lnRNAs

To determine the putative regulatory roles of Ae. albopictus lncRNAs upon virus infections, we searched for protein-coding genes located within 100 kb (upstream and downstream) from differentially expressed lncRNAs as potential cis-regulated genes [28,55] and found a total of 198 and 153 coding genes which fulfilled that criteria in the DENV and ZIKV transcriptome respectively. As lncRNAs also have the ability to regulate genes in trans or distal regions [28,55–57], we searched for protein-coding genes that were co-expressed with the differentially expressed lncRNAs based on Pearson correlation (coefficient > 0.95 or < - 0.95, P-value < 0.05). In the transcriptome of DENV-infected cells, 110 coding genes were co-expressed with the differentially expressed lncRNAs. Meanwhile, we detected 38 coding genes having high correlation in expression with differentially expressed lncRNAs in ZIKV transcriptome. We categorized the co-expressed genes as putative trans-regulated genes in Ae. albopictus cells upon DENV and ZIKV infection. Results of co-expression analysis between lncRNAs and protein-coding mRNAs can be found in S5 Table.

We selected the potential cis and trans-regulated genes for functional annotation and enrichment analyses. Previous reports demonstrated that upon virus infection in Aedes mosquitoes, there were changes in the expression level of host immune-related genes [16,20,58–60]. By conducting functional annotation using Pannzer [41], we discovered 13 lncRNAs that potentially regulated genes involved in immune-related functions (S5 Table). Based on gene ontology (GO), these genes were implicated in innate immune response (GO:0045087), defense response to Gram-positive (GO:0050830) and Gram-negative (GO:0050829), defense response to bacterium (GO:0042742), regulation of defense response to virus (GO:0050688), virus maturation (GO:0019075), and innate immune response (GO:0045087).

Previous studies reported that genes encoding clip-domain serine protease, and metalloproteinase were differentially expressed in Aedes mosquitoes upon virus infection. For example, clip-domain serine protease and metalloproteases were found to be significantly upregulated and downregulated respectively upon DENV infection in Ae. albopictus midguts [20]. Meanwhile, another study in Ae. aegypti reported that metalloproteinase and serine protease showed an increase in expression upon infection with ZIKV [16]. In this study, we discovered lncRNAs that potentially regulate genes that were also involved in serine-type endopeptidase activity (GO:004252, GO:004253, GO:004254) and metalloendopeptidase activity (GO:004222).

We then performed functional enrichment analysis of lncRNA-regulated genes using g;Profiler [42]. We discovered that 32 and 22 GO terms were significantly enriched (P-value < 0.05) in ZIKV and DENV-transcriptome respectively. A total of 8 GO terms were significant and shared by both transcriptomes, suggesting that lncRNAs may target genes that perform similar regulatory functions upon DENV and ZIKV infections (S7 Table). Among the significant GO terms were cellular processes, ion/protein binding, metabolic processes, and DNA binding.

Mutation of selected lncRNAs and enhancement of DENV and ZIKV replication

To verify the role of Ae. albopictus lncRNAs in virus infection, we used CRISPR-Cas9 to induce mutation of selected lncRNAs in C6/36 cells following DENV and ZIKV infection. We discovered that CRISPR-Cas9 induced mutations in two lncRNAs loci resulted in the increase of DENV-1 and ZIKV infectious titer by at least two folds after three dpi (Fig 5A). One of the lncRNAs was novel (transcript ID: lncRNA_27639.2, loci ID: XLOC_029733) and another was known lncRNA (XR_003899061.1, ncbi gene ID: LOC115270134). Indels for both lncRNAs were confirmed by sequencing the PCR products spanning the cleavage sites, and the expression level lncRNAs after CRISPR-Cas9 induced mutation were validated by RT-qPCR (S3 Fig). Reduction in the level of expression indicated that mutations induced by CRISPR-Cas9 resulted in the silencing of the two aforementioned lncRNAs in C6/36 cells.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Possible involvement of lncRNAs in DENV and ZIKV replication in C6/36 cell.

A) Titer of ZIKV and DENV-1 after 3 dpi as determined by TCID50 assay. In mock sample, C6/36 cells were transfected with transfection reagent only (Lipofectamine 2000). Empty vector refers to the Drosophila CRISPR-Cas9 vector (pAc-dU6-sgRNA-Cas9) without sgRNA targeting any gene. B) Expression level of lncRNA_27639.2 and XR_003899061.1 at 1, 3 and 5 dpi. At 1 dpi, the fold change between infected and uninfected samples were statistically insignificant. C) Expression level of LOC109431891 and LOC115269840 in C6/36 cells that had mutations in lncRNA_27639.2 and XR_003899061.1 respectively. Expression levels were normalized to empty vector samples. The expression of both genes (LOC109431891 and LOC115269840) increased in lncRNA-mutated C6/36 cells compared to empty vector samples. For A), B), and C), bars indicate mean +/- SEM of three independent experiments. Student’s t-test was used to assess the statistical significance. * p<0.05; **p<0.01.

https://doi.org/10.1371/journal.pntd.0008351.g005

A previous report showed that the expression of Ae. aegypti lincRNA_1317 increased substantially following the progression of infection [15]. When this lncRNA was silenced by RNAi, it led to an enhancement of DENV replication in Ae. aegypti cell, implying the involvement of certain lncRNAs in viral infection. We next analyzed our lncRNAs to idenfify if they also played a key role in anti-viral response. Based on RNA-seq and RT-qPCR analysis, both lncRNAs (lncRNA_27639.2 and XR_003899061.1) were upregulated following infection with DENV and ZIKV at 3 dpi (S2 Fig). We investigated the level of expression at early (1 dpi) and late (5 dpi) stage of DENV and ZIKV infection. Interestingly, the expression of both lncRNAs displayed an increasing pattern along the progression of infection (Fig 5B), indicating possible involvement of lncRNA_27639.2 and XR_003899061.1 in defense against viruses in Ae. albopictus.

lncRNA_27639.2 might potentially regulate a gene that encodes for solute carrier family 35 member G1 (LOC109431891). This gene was located 1.9 kb downstream of lncRNA_27639.2 on the same strand; hence, making it a potential cis-regulated gene. GO annotation of solute carrier family 35 member G1 (LOC109431891) showed that it participates in calcium sensing and homeostasis, and it is a part of endoplasmic reticulum [61]. Meanwhile, XR_003899061.1, was located in cis to and 61 kb away from a gene encoding clathrin interactor 1 protein (LOC115269840), which is involved in clathrin assembly and trafficking [62]. Interestingly, calcium homeostasis, clathrin-mediated endocytosis, and vesicular trafficking are essential for viral replication in host cells [63–65]; thereby, further supporting the idea that lncRNA_27639.2 and XR_003899061.1 might be involved in antiviral response.

We then investigated the expression of LOC109431891 and LOC115269840 in C6/36 cells that had CRISPR-Cas9 induced mutations in lncRNA_27639.2 and XR_003899061.1 respectively. Our qPCR results showed that the expression of both LOC109431891 and LOC115269840 increased nearly two-folds in C6/36 cells harboring mutations in the lncRNA loci (Fig 5C). Although the fold change of both genes was only approximately two-folds, the changes were statistically significant (P-value < 0.05). Therefore, we speculate that lncRNA_27639.2 and XR_003899061.1 may be involved in antiviral response by limiting the expression of LOC109431891 and LOC115269840 respectively.