Ixodes scapularis salivary gland microRNAs are differentially expressed during Powassan virus transmission

Successful tick feeding is facilitated by an assortment of pharmacologically-active factors in tick saliva that create an immunologically privileged micro-environment in the host’s skin. Through a process known as saliva-assisted transmission, bioactive tick salivary factors modulate the host environment, promoting transmission and establishment of a tick-borne pathogen. This phenomenon was previously demonstrated for Powassan virus (POWV), a North American tick-borne flavivirus that is the causative agent of a severe neuroinvasive disease in humans. Here, we sought to characterize the Ixodes scapularis salivary gland microRNAs (miRNAs) expressed during the earliest period of POWV transmission to a mammalian host. POWV-infected and uninfected I. scapularis females were fed on naïve mice for 1, 3, and 6 hours, and Illumina next generation sequencing was used to characterize the salivary gland miRNA expression profiles of POWV-infected versus uninfected ticks. 379 salivary miRNAs were detected, of which 338 are reported here as putative novel I. scapularis miRNAs. 35 salivary gland miRNAs were significantly up-regulated and 17 miRNAs were significantly down-regulated in response to POWV infection. To investigate the potential role of salivary gland miRNAs in POWV replication in-vitro, we transfected miRNA inhibitors into VeroE6 cells to profile temporal POWV replication in mammalian cells. Together, the small RNA sequencing data and the in vitro miRNA inhibition assay suggest that the differentially expressed tick salivary miRNAs could act in regulating POWV replication in host tissues.


Results
Profile characteristics of small RNA libraries. POWV-infected and uninfected I. scapularis adult females were fed on naïve mice for 1, 3, and 6 hours (1 tick per mouse), and tick salivary glands were dissected at each time point. RNA (including small RNA molecules < ~200 nucleotides) was extracted from the salivary glands of every fed tick, and a separate small RNA library was prepared for each pair of tick salivary glands (21 small RNA libraries total). POWV infection was confirmed in individual tick salivary glands by qPCR. The average POWV titers in infected tick salivary glands after 1 hour, 3 hours, and 6 hours of tick feeding was 4.15 × 10 6 focus forming units (FFU)/µg RNA, 1.20 × 10 6 FFU/µg RNA, and 2.46 × 10 6 FFU/µg RNA, respectively (Table 1).
Four biological replicates were included for each of the 3 POWV-infected tick feeding time points (i.e. 4 pairs of POWV-infected tick salivary glands were harvested at 1 hour resulting in 4 individual small RNA libraries, 4 pairs of infected salivary glands were harvested at 3 hours resulting in 4 small RNA libraries, and 4 pairs of infected salivary glands were harvested at 6 hours resulting in 4 small RNA libraries). Three biological replicates were included for each of the 3 uninfected tick feeding time points (i.e. 3 pairs of uninfected tick salivary glands were harvested at 1 hour resulting in 3 individual small RNA libraries, 3 pairs of uninfected salivary glands were harvested at 3 hours resulting in 3 small RNA libraries, and 3 pairs of uninfected salivary glands were harvested at 6 hours resulting in 3 small RNA libraries). Four biological replicates (instead of 3) were included for each of the POWV-infected tick feeding time points to better account for any potential variation of POWV titers between salivary gland samples. In total, 21 small RNA libraries were included in this study.
142.8 million, and 97.9 million clean reads with passing filters for the POWV-infected 1 hour, 3 hour, and 6 hour fed salivary gland libraries, respectively. Likewise, from the uninfected 1 hour, 3 hour, and 6 hour fed salivary gland libraries, 62.8 million, 49.9 million, and 61.7 million reads with adapter sequences were acquired, respectively. Of the total reads processed in different uninfected salivary gland libraries, 0.23% -2.96% were discarded. This resulted in 62.8 million, 49.7 million, and 61.0 million clean reads with passing filters for the uninfected 1 hour, 3 hour, and 6 hour fed salivary gland libraries, respectively. The predominant size distribution of these small RNA reads was 20-30 nucleotides (nt). Based on the size distribution of the clean reads with passing filters at each experimental time point, a bimodal size distribution was detected, with one peak around 21-23 nt representing miRNAs and another peak around 26-29 nt likely representing piwi-interacting RNAs (piRNA) (Fig. 1). For the 1 hour POWV-infected salivary gland libraries, the highest percentage of clean reads (19.43% of library) were 28 nt long, followed by 14.90% that were 22 nt (Fig. 1a). Similarly, for the 1 hour uninfected salivary gland libraries, 28 nt reads were most abundant (18.51%), followed by 22 nt reads (14.05%). For all infected and uninfected 3 hour salivary gland libraries, the most abundant reads were 28 nt, followed by 27 nt, then 22 nt. Specifically, 28 nt reads (18.42%), 27 nt reads (13.30%), and 22 nt reads (13.28%) were most abundant for the 3 hour POWV-infected libraries, while 28 nt reads (21.80%), 27 nt reads (13.21%), and 22 nt reads (10.95%) were most abundant for the 3 hour uninfected libraries (Fig. 1b). Similar to the read abundance pattern displayed for the 1 hour libraries, 28 nt reads (19.34%) followed by 22 nt reads (14.48%) were most abundant for the 6 hour POWV-infected libraries, and 28 nt reads (19.90%) followed by 22 nt reads (13.01%) were most abundant for the 6 hour uninfected libraries (Fig. 1c). For all three experimental time points, the peaks at 21 and 22 nt were greater for the POWV-infected salivary glands than the uninfected salivary glands, suggesting that miRNAs are more abundant and potentially more diverse in the POWV-infected salivary glands versus the uninfected salivary glands (Fig. 1a-c).
For each experimental time point and infection condition, clean reads were matched to various categories of small noncoding RNAs, including rRNA, snoRNA, and snRNA (Table 2 and Fig. 1d-i). Additionally, small RNA reads were mapped to the POWV genome (POWV LB Strain, NCBI Reference Sequence: NC_003687.1). In total, 1746, 7759, and 4655 reads obtained from the POWV-infected 1 hour, 3 hour, and 6 hour salivary gland libraries, respectively, mapped to the POWV genome. 22 nt reads were most abundant for the virus-derived small RNAs Figure 1. Summary of next generation sequencing of tick-derived small RNAs obtained from POWV-infected and uninfected Ixodes scapularis salivary glands at 1 hour, 3 hours, and 6 hours post tick attachment. Salivary glands were dissected from POWV-infected female I. scapularis ticks and uninfected female ticks that were fed on mice for 1, 3, or 6 hours. RNA was extracted from the salivary glands, and 21 small RNA libraries were generated then subjected to Illumina small RNA sequencing. (a-c) Size distribution of tick-derived small RNA reads with passing filters mapped to the I. scapularis genome. (d-f) Summary of tick-derived reads that match various small RNA categories from POWV-infected tick salivary glands at 1 hour, 3 hours, and 6 hours post tick attachment. (g-i) Summary of tick-derived reads that match various small RNA categories from uninfected tick salivary glands at 1 hour, 3 hours, and 6 hours post tick attachment. across all three time points (Fig. 2) Identification of novel tick miRNAs. The miRDeep2 prediction software package was used to identify potential miRNA precursors from the NGS data. The software aligns the reads to the I. scapularis reference genome and looks for locations where potential miRNA reads stack up. The regions immediately surrounding the mapped reads are examined for miRNA biogenesis features, including mature miRNA, Star and precursor reads, and stem-loop folding properties. Essentially, the miRDeep2 program models the miRNA biogenesis pathway, using a probabilistic algorithm to score compatibility of the position and frequency of NGS reads with the secondary structure of the miRNA precursor 31 . A total of 379 miRNAs were detected in the salivary gland libraries, of which 338 are reported here as potential novel I. scapularis miRNAs, and 41 miRNAs matched a known, previously annotated miRBase I. scapularis mature miRNA (Table 3). Supplemental Table 1 includes the consensus mature, star, and precursor sequence, the total read counts, and the miRDeep2 score and probability for every miRDeep2-predicted miRNA with a miRDeep2 score of 4 or greater. The stem-loop structure, the star sequence and premature miRNA sequence are shown for a select few miRNAs that were highly expressed in I. scapularis salivary glands, including isc-miR-5307, nDS978597_16878, and nDS625977_65388 (Fig. 3).
Expression profiles of miRNAs in POWV-infected and uninfected tick salivary glands. For each experimental time point (1 hour, 3 hours, or 6 hours of tick feeding) and infection condition (POWV-infected or uninfected tick salivary glands), the normalized read counts were totaled for every miRNA to generate a list of the top 50 most abundant miRNAs (Table 4). In this text and in the corresponding figures, any miRDeep2-predicted novel miRNA that does not match an already annotated I. scapularis miRNA is denoted with "nDS" at the beginning of the identifier. The online tool, InteractiVenn, was used to generate an Edwards Venn diagram representing the 50 most abundant miRNAs expressed for each time point and infection condition 32 . Among the lists of top 50 expressed miRNAs, 39 were shared between all 6 experimental groups (Fig. 4). 28 of the 39 shared miRNAs matched a known, previously annotated miRBase I. scapularis mature miRNA and 5 matched a previously annotated I. ricinus miRNA, while the remaining 6 miRNAs were predicted to be novel by miRDeep2. Since these 39 shared miRNAs were highly expressed in the uninfected as well as in the POWV-infected tick salivary glands after 1, 3, and 6 hours of tick feeding, the expression of these miRNAs is likely to be independent of POWV infection of the salivary glands.

Differential expression of I. scapularis salivary gland miRNAs in response to POWV infection.
To examine the global differential expression of I. scapularis salivary gland miRNAs in response to POWV infection, read counts were normalized as reads per million. A Log 2 fold-change and corresponding adjusted P-value (FDR) were generated for the differential expression of miRNAs from POWV-infected salivary glands versus uninfected salivary glands. Log 2 fold-changes were derived from 4 biological replicates at each of the 3 POWV-infected tick feeding time points and from 3 biological replicates at each of the 3 uninfected tick feeding time points. miRNAs with a Log 2 fold-change expression > |1| and FDR ≤ 0.1 were considered significantly differentially expressed.
Small RNA libraries from POWV-infected I. scapularis salivary glands showed significant alteration of miRNA expression profiles compared to the uninfected I. scapularis salivary glands after 1, 3, and 6 hours of tick feeding. 6 salivary gland miRNAs were significantly up-regulated in response to 1 hour of POWV-infected I. scapularis feeding, all of which showed Log 2 fold change > 4 (Fig. 5). The maximum significant up-regulation was found in nDS645604_93679, with Log 2 fold change of 21.35. nDS777850_38819, which was significantly up-regulated after 1 hour of POWV-infected tick feeding (Log 2 fold change = 7.52), was also among the 50 most abundant miR-NAs from the POWV-infected 1 hour salivary gland libraries. Two miRNAs were significantly down-regulated in response to 1 hour of POWV-infected tick feeding, both of which showed Log 2 fold change = −4.50 (Fig. 5). nDS752087_3745, which was among the 50 most abundant miRNAs from the uninfected 1 hour salivary gland libraries, was significantly down-regulated in response to 1 hour of POWV-infected tick feeding (Log 2 fold change = −4.50).
In response to 6 hours of POWV-infected I. scapularis feeding, 6 salivary gland miRNAs were significantly upregulated, and all 6 showed Log 2 fold change > 3 (Fig. 5). The miRNA with maximum significant up-regulation in response to 6 hours of POWV-infected tick feeding was nDS642227_71264 (Log 2 fold change = 26.65). 4 significantly up-regulated miRNAs (nDS642227_71264, nDS748105_49761, nDS775456_13019, and nDS857739_57334) were among the 50 most abundant miRNAs from the POWV-infected 6 hour salivary gland libraries. 3 salivary gland miRNAs were significantly down-regulated in response to 6 hours of POWV-infected I. scapularis feeding, and the maximum significant down-regulation was for nDS868494_62448 with a Log 2 fold change of −7.89 (Fig. 5). All 3 of the significantly down-regulated 6 hour miRNAs showed Log 2 fold change > |3|.
Upon comparing the lists of salivary gland miRNAs significantly up-regulated in response to POWV infection across the 3 experimental time points, it is evident that the majority of significantly up-regulated miRNAs were unique to each time point and not shared (Fig. 6a). No significantly up-regulated miRNA was shared between all three time points; however, a single miRNA, nDS748105_49761, was common between 3 hours and 6 hours of POWV-infected tick feeding. nDS748105_49761 was up-regulated at all three experimental time points, with significant up-regulation in response to both 3 and 6 hours of POWV-infected tick feeding (Log 2 fold changes = 6.48 and 6.88, respectively). Analysis of the significantly down-regulated miRNAs across the 3 experimental time points also shows that the majority of miRNAs were unique to each time point and not shared between time points (Fig. 6b). None of the significantly down-regulated miRNAs were shared between all three time points. Although 14 miRNAs were significantly down-regulated in response to 3 hours of POWV-infected tick feeding, none were shared between the 1 hour and 6 hour time points. Two significantly down-regulated miRNAs (nDS752087_3745 and nDS973976_102858) were common between 1 hour and 6 hours of tick feeding. www.nature.com/scientificreports www.nature.com/scientificreports/ qRT-PCR validation of mature miRNA differential expression. To    www.nature.com/scientificreports www.nature.com/scientificreports/ small RNA library preparation, a portion of the RNA sample was converted to cDNA. Mature miRNAs were selectively converted into cDNA, and a miRNA-specific qRT-PCR reaction was performed for each sample. The qRT-PCR patterns of differential expression matched the NGS patterns in the majority of evaluated miRNAs across the three experimental time points (Fig. 7). Although it is not surprising to find discrepancies between these two methods of quantifying miRNA expression 33 , in 24 out of 33 instances (72.7%), the pattern of miRNA regulation was the same between the NGS data and the qRT-PCR data (ie. both NGS and qRT-PCR datasets were up-regulated or both datasets were down-regulated). The qRT-PCR validation data for the 8 known I. scapularis miRNAs showed that in 17 out of 24 instances the trend for the NGS data matched that of the qRT-PCR data (Fig. 7). Discrepancies were found for isc-miR-12 at 1 and 6 hours, isc-miR-124 at 6 hours, isc-miR-133 at 6 hours, isc-miR-184 at 3 hours, and isc-miR-5312 at 3 and 6 hours. Of these discrepancies, the only miRNA with significant differential expression in both the NGS and qRT-PCR datasets was isc-miR-5312 at 3 hours. The qRT-PCR Log 2 fold change expression values for isc-miR-184 at 6 hours and isc-miR-5312 at 1 hour were very small (Log 2 fold change = −0.0063 and −0.0031, respectively), and although these values do not clearly appear on the graphs, these patterns of differential expression matched the down-regulation pattern shown for the NGS data. The qRT-PCR validation data for the 3 miRDeep2-predicted novel I. scapularis miRNAs showed that in 7 out of 9 instances the trend for the NGS data matched that of the qRT-PCR data (Fig. 7). Discrepancies between the patterns of differential expression were found for nDS630914_20990 at 6 hours and nDS752087_3745 at 3 hours. For the 6 hour qRT-PCR versus NGS comparison of nDS630914_20990, the unmatched pattern of differential expression is likely due to the non-significant NGS data point. Overall, the qRT-PCR validation confirmed that all 11 selected miRNAs were detected in the POWV-infected and uninfected I. scapularis salivary glands and that the qRT-PCR patterns of differential expression matched the NGS patterns in the majority of evaluated miRNAs.

I. scapularis salivary gland miRNAs regulate POWV replication in mammalian cells. To investi-
gate the potential role of I. scapularis salivary gland miRNAs in POWV replication, miRNA inhibitors were transfected into VeroE6 cells to profile POWV replication in mammalian cells over time. In vitro data from testing 9 miRNA inhibitors demonstrated their role in regulating POWV infection. POWV titers in mock-transfected cells peaked at 96 hpi (hours post infection) at 6.63 Log 10 FFU equivalents. At 24, 48, and 72 hpi POWV titers were significantly higher (p < 0.05) in cells transfected with isc-miR-315, isc-miR-5307, and nDS630914_20990 inhibitors (Fig. 8) than in mock-transfected cells. At 72 hpi, cells transfected with isc-miR-1, nDS625977_65388, and nDS752087_3745 inhibitors displayed significantly higher (P < 0.05) POWV titers than mock-transfected cells, and at 96 hpi, POWV titers were significantly higher in cells transfected with isc-miR-5307, nDS630914_20990, and nDS752087_3745 inhibitors compared to mock-transfected cells. In this in vitro study, there were two instances in which the POWV titers were significantly lower in response to miRNA inhibitor transfection compared to mock transfection. At 96 hpi, cells transfected with either isc-miR-124 or isc-miR-184 inhibitors displayed significantly (P < 0.05) lower POWV titers (6.11 Log 10 FFU equivalents and 6.17 Log 10 FFU equivalents, respectively) compared to mock-transfected cells (6.63 Log 10 FFU equivalents). Though the titer differences were not significant, cells transfected with the isc-miR-5312 inhibitor displayed lower POWV titers at 24, 72, and 96

Discussion
The complex assortment of molecules in tick saliva reflects the multifaceted host defense mechanisms and enables ticks to remain attached to the host, to successfully complete blood feeding, and to evade the host immune response 16,34,35 . Through saliva-assisted transmission (SAT), bioactive tick salivary factors modulate various host defense mechanisms, creating an environment that is more favorable for pathogen transmission and establishment. The phenomenon of SAT has been demonstrated for several TBFVs. Tick saliva is composed of hundreds of proteins that are differentially expressed over the course of the tick's lengthy feeding process 34 . Many studies have demonstrated that various tick saliva proteins modulate host hemostasis, wound repair, and the innate and adaptive host immune responses 14 ; however, few non-proteinaceous tick salivary gland molecules have been identified. Prostaglandin was the first identified non-protein component of tick saliva 36 and was later found to have immunomodulatory properties 37 . Non-proteinaceous molecules such as prostaglandin E2, prostacyclin, purine nucleoside adenosine, endocannabinoids, and related fatty acids have all been shown to be immunomodulatory components of tick saliva 37-39 . miRNAs are yet another category of non-proteinaceous molecules that have only recently been detected in the saliva and salivary glands of several tick species 23,24,27 . These short, ~22 nucleotide, non-coding RNAs are involved in post-transcriptionally regulating host target genes in various physiological and pathological processes of eukaryotes. Detection of I. ricinus saliva-specific miRNAs provided the first evidence of tick salivary miRNAs that regulate host gene expression 27 . Unlike tick salivary gland proteins, no immune response can be mounted against these non-proteinaceous miRNA molecules. By gaining an understanding of how tick miRNAs immunomodulate the tick-host interface in the presence of a pathogen, we can build a foundation for developing tick miRNA-based therapeutic candidates.
In this study, we present the first data on tick salivary gland miRNA expression profiles in relation to virus infection and transmission. The tick miRNA profiles in I. scapularis salivary glands were composed of two peaks, including reads of 21-23 nt and reads of 26-29 nt (Fig. 1). The presence of these two peaks is consistent with previous reports of small RNA sequencing in other tick and mosquito miRNA studies. After 1, 3, and 6 hours of tick feeding, the size distribution peaks at 21 and 22 nt were greater for the POWV-infected salivary glands than for the uninfected salivary glands, suggesting that there is more abundance of miRNAs in the POWV-infected tick salivary glands compared to the uninfected salivary glands. Although the present study focused on I. scapularis miRNAs (21-23 nt), it would be beneficial for a future study to examine the potential role of other small RNA populations in regulating viral replication and transmission during tick feeding.
Using miRDeep2 prediction software, various novel tick miRNAs were predicted from the I. scapularis salivary gland libraries. The presence of select novel predicted miRNAs (nDS625977_65388, nDS630914_20990, and nDS752087_3745) was confirmed via their detection with qRT-PCR (Fig. 7). Additional validation of other predicted novel miRNAs is needed, especially for those predicted miRNAs with high miRDeep2 probability scores (Supplemental Table 1). The top 50 most abundant miRNAs were compared for each of the 6 experimental groups. 39 miRNAs were shared between all 6 groups (Fig. 4), of which 28 matched a known, previously annotated miR-Base I. scapularis mature miRNA and 5 matched a previously annotated I. ricinus miRNA. Previous reports indicate that evolutionarily-conserved miRNAs are usually highly expressed 23,24 . Since these shared, highly expressed www.nature.com/scientificreports www.nature.com/scientificreports/ miRNAs were detected in both the uninfected and POWV-infected tick salivary glands, the majority are likely constitutively expressed I. scapularis salivary gland miRNAs that are not involved in regulating POWV infection.
In I. scapularis female ticks, infection with POWV altered the salivary gland miRNA profile, with multiple instances of temporal variation in miRNA expression observed between 1, 3, and 6 hours of tick feeding. NGS analysis identified novel I. scapularis salivary gland miRNAs as well as known, previously annotated miRNAs that were differentially expressed upon POWV infection (Fig. 5). At each time point, more miRNAs showed significant up-regulation compared to down-regulation in response to POWV-infected tick feeding. The following 7 known I. scapularis miRNAs were significantly up-or down-regulated in response to POWV-infected tick feeding: isc-miR-1, isc-miR-12, isc-miR-315, isc-miR-317, isc-miR-5307, isc-miR-5312, and isc-miR-79.
Although no previous study has examined tick miRNAs in relation to pathogen infection and transmission, lipopolysaccharide (LPS)-induced differential expression of miRNAs in R. haemaphysaloides ticks has been documented. miR-1 and miR-315 were in the top 10 most abundant miRNAs in PBS-and LPS-injected R. haemaphysaloides ticks, but neither met the cut-off for significant up-or down-regulated expression in response to LPS 26 . miR-79-3p was significantly down-regulated in female R. haemaphysaloides ticks after LPS injection 26 . In the present study, isc-miR-1, isc-miR-315, and isc-miR-79 were all significantly up-regulated in response to 3 hours of POWV-infected tick feeding (Fig. 5). Discrepancies in the patterns of LPS-induced versus POWV-induced differential expression for miR-1, miR-315, and miR-79 could be attributed to the following: different types of samples processed (ie. individual tick salivary glands from blood-fed POWV-infected ticks versus whole tick bodies from LPS-injected unfed ticks), specific pathogen-tick species interactions, the physiology of unfed ticks versus feeding ticks. Figure 7. qPCR validation of select mature miRNAs that are differentially expressed in response to POWV infection. Each graph depicts the differential expression (Log 2 fold change) of a select miRNA from POWVinfected salivary glands versus uninfected salivary glands based on NGS data (shown in orange) and qRT-PCR data (shown in blue). Significance for qRT-PCR-based differential expression is determined by the 2-tailed Student's t test where *P < 0.05, **P < 0.01, ***P < 0.001. Significance for NGS-based differential expression is determined by a DESeq2-derived adjusted P-value (P-adj) where # P-adj < 0.1, ## P-adj < 0.01, ### P-adj < 0.001.
Recently, NGS of uninfected adult I. ricinus saliva-derived libraries provided the first evidence of miRNAs present in pure tick saliva, and molecular indicators suggest that that these saliva-derived miRNAs are secreted in exosomes 27 . Several of the same known (miRBase annotated) miRNAs that were significantly modulated in I. scapularis female salivary glands in response to POWV-infected tick feeding were also detected in the salivary glands and saliva of uninfected feeding I. ricinus females (miR-1, miR-12, miR-317, and miR-5307). In the I. ricinus study, iri-miR-5307-3p was identified as a putative feeding-regulated miRNA because its expression levels increased as tick feeding progressed 27 , and in the present study, isc-miR-5307 was significantly up-regulated in response to 3 hours of POWV-infected I. scapularis feeding (Fig. 5). In uninfected I. ricinus females, iri-miR-317-3p was 249-fold overrepresented in saliva but virtually absent in the salivary glands 27 . This study showed that in response to 3 hours of POWV-infected I. scapularis feeding, isc-miR-317 and isc-miR-5312 were significantly down-regulated. Interestingly, rmi-miR-317 and rmi-miR-5312 were significantly down-regulated when R. microplus larvae exposed to a host for six hours (but not allowed to feed) were compared to unexposed larvae, providing evidence that host-odor recognition triggers changes in rmi-miR-317 and rmi-miR-5312 expression 23 .
Prior to this publication, only one other study had characterized arthropod salivary miRNA profiles in response to virus infection. Exogenous miRNAs were identified in A. aegypti and A. albopictus mosquito saliva, several of which were expressed only upon CHIKV infection 29 . The patterns of virus-induced differential expression between POWV-infected I. scapularis salivary gland miRNAs and CHIKV-infected mosquito saliva miR-NAs aligned in several instances. For example, isc-mir-315 was significantly up-regulated in response to 3 hours of POWV-infected tick feeding, corresponding with the highly up-regulated expression levels of aae-miR-315 in CHIKV-infected A. aegypti and A. albopictus saliva 29 . The significant down-regulation of isc-miR-317 in 3 www.nature.com/scientificreports www.nature.com/scientificreports/ hour-fed POWV-infected I. scapularis salivary glands is complemented by aae-miR-317 down-regulation in CHIKV-infected A. albopictus saliva 29 . Additionally, miR-317-5p was significantly down-regulated in pools of whole A. aegypti and A. albopictus mosquitoes infected with Dengue virus 2 40,41 . Aae-miR-1-5p was significantly up-regulated in whole A. aegypti that were 7 days post-Zika virus infection 33 , an expression pattern that was mirrored, albeit at different timelines post-infection, with significant up-regulation of isc-miR-1 in 3 hour fed POWV-infected I. scapularis salivary glands.
Our in vitro data from testing 9 miRNA inhibitors demonstrated their potential role in regulating POWV infection (Fig. 8). POWV titers were significantly decreased in mammalian cells at 96 hpi when isc-miR-124 and isc-miR-184 were inhibited, suggesting that the presence of these miRNAs in tick salivary glands (and potentially in tick saliva) could enhance POWV infection and later dissemination from the cutaneous site of infection. In contrast, for several of the miRNAs inhibited, such as isc-miR-315, isc-miR-5307, miR-nDS630914_20990, and miR-nDS752087_3745, there were significant increases in POWV titers for at least two time points examined. Because inhibiting some miRNAs resulted in increased POWV titers in our in vitro assay, certain miRNAs present in tick saliva could potentially limit POWV replication and transmission to a host. Our data demonstrates that during the initial hours of tick feeding, unique expression patterns of specific tick salivary gland miRNAs are detected. Specifically, 52 salivary gland miRNAs were significantly differentially expressed in response to 1, 3, or 6 hours of POWV-infected tick feeding. It is plausible that specific tick salivary gland-derived miRNA signatures could temporally regulate POWV replication in certain host tissues, and it is the subject of our future study. Now that virus-induced differential expression has been identified for multiple tick and mosquito salivary miRNAs, it will be beneficial to determine the functional role of these miRNAs in the transmission and establishment of infection during vector blood feeding. This process has begun for mosquito miRNAs, and a recent comprehensive review outlined the miRNAs with validated functions in regulating West Nile virus, Dengue virus, and CHIKV infection 20 . The I. scapularis salivary gland miRNAs that displayed significant differential expression upon POWV-infected tick feeding will be the subject of our future in vivo studies. By gaining an understanding of how non-proteinaceous tick saliva factors such as miRNAs immunomodulate the tick-host interface in the presence of a virus, we will build a foundation for developing potential tick saliva-based therapeutic candidates.

cells and viruses. African green monkey kidney (VeroE6) cells were purchased from the American Type
Culture Collection (ATCC) and maintained in culture with Modified Eagle's Medium (MEM) supplemented with 10% fetal bovine serum, 1% non-essential amino acids, and a 1% antibiotic mixture of penicillin/streptomycin incubated at 37 °C with 5% CO 2 . The World Reference Center for Emerging Viruses and Arboviruses at UTMB provided stock of the prototypic LB strain of POWV, which had previously been passaged 7 times in suckling mice brains. The stock was then passaged 6 times on VeroE6 cells. Stock virus titers were determined by focus-forming immunoassay as described previously 42 . Animals. Five-week-old female BALB/cj mice were received from The Jackson Laboratory (Bar Harbor, ME).
Mice were allowed to acclimate to the local environment before incorporation into the experiments, at which point the mice were six weeks of age.

Tick infection and infestation on mice.
Uninfected adult Ixodes scapularis ticks were maintained in our lab within the UTMB arthropod containment level 2 (ACL-2) facilities. Tick vials were stored inside an incubator at 22 °C and 90-93% relative humidity. The incubator photoperiod was set on a 16:8 hour cycle. Adult, female I. scapularis were synchronously infected with POWV-LB strain (7 suckling mouse brain passages followed by 5 Vero cell passages) as previously described 43 . Control female ticks were mock-infected with media. POWV-infected and mock-infected I. scapularis females were co-housed with uninfected male I. scapularis and stored inside a desiccator at 22 °C to allow POWV replication. At 40 days post-synchronous infection, the I. scapularis females were infested on mice. Viral infection in each pair of tick salivary glands was confirmed by q-RT-PCR on a CFX96 real-time PCR (BioRad) using 10 µM POWV-specific forward and reverse primers (Table 5) and the iTaq Universal SYBR Green One-Step kit (BioRad). The following cycling protocol was used: 10 minutes at 50 °C; 1 minute at 95 °C; 10 seconds at 95 °C followed by 30 seconds at 60 °C for 45 cycles; and an 81-cycle (+0.5 °C/cycle) 55-95 °C melt curve.
Tick containment capsules were prepared as previously described 43 . Two days prior to tick infestation, mice were anesthetized with isoflurane and the torsos were shaved. One capsule was adhered to the dorsum of each mouse using livestock Kamar adhesive (Kamar Inc., Steamboat Springs, CO). On the day of tick infestation, a single I. scapularis female was placed inside each capsule and allowed to feed on the mouse for 1, 3, or 6 hours. At the experimental time points, mice were euthanized with CO 2 inhalation according to IACUC protocols. The feeding tick was removed and tick salivary glands were dissected in sterile phosphate-buffered saline (PBS) on a sterile microscope slide. Each pair of salivary glands was stored in a 100 µL aliquot of QIAzol lysis reagent (Qiagen).
RNA extractions from tick salivary glands. Tick salivary glands were homogenized in QIAzol reagent using a motorized hand-held pestle. An additional 600 µL of QIAzol reagent was added to each homogenized salivary gland sample. RNA extractions were performed as outlined in the miRNeasy Mini kit handbook (Qiagen), (2019) 9:13110 | https://doi.org/10.1038/s41598-019-49572-5 www.nature.com/scientificreports www.nature.com/scientificreports/ resulting in purification of total RNA, including small RNAs. Viral infection in tick salivary glands was confirmed by quantitative real-time PCR (qPCR) using primers specific to the NS5 gene of POWV as previously described 17 . Next Generation Sequencing. Small RNA libraries were made using the NEBNext smallRNA Multiplex kit (New England Biolabs, In.) following the manufacturer's protocol. Briefly, short adapter oligonucleotides are ligated to each end of the small RNAs in the sample, a cDNA copy is made with reverse transcriptase, and PCR is used to add sample specific barcodes and Illumina sequencing adapters.
The final concentration of all NGS libraries was determined using a Qubit fluorometric assay and the DNA fragment size of each library was assessed using a DNA 1000 high-sensitivity chip on an Agilent 2100 Bioanalyzer. After purification by polyacrylamide gel-electrophoresis, the sample libraries were pooled and sequenced on an Illumina NextSeq550 (single end 75 base) using TruSeq SBS kit v3 (Illumina) and protocols defined by the manufacturer.
Data analysis. The miRDeep2 software package, version 2.0.0.8, was used to process the sequencing data. For the novel miRNA prediction step, the reads from all the samples were combined. The mapper function of miRD-eep2 first trims the adapter sequences from the reads and converts the read files from fastq to fasta format. Reads shorter than 18 bases were discarded. These reads were then mapped to the Ixodes scapularis reference sequence using the default miRDeep2 mapper function parameters. For the final miRDeep2 steps, the trimmed reads were mapped to the known Ixodes scapularis miRNAs from the miRBase database (Version 22) and quantified. Next, the program used the reads previously mapped to the I. scapularis reference to predict potential novel miRNAs. The output file included a score reflecting the likelihood that the predicted miRNA is real, the sequence and location of the possible miRNA, and the number of reads that mapped to it. A total of 323 unique novel miRNAs had a miRDeep2 score of 4 or greater.
For differential expression analysis, reads from each sample were individually mapped to the I. scapularis reference with bowtie (version 1.2.2, parameters: -v2 -l 18 -a -M 10 -best -strata) and reads mapping to the known and novel miRNAs were quantified with featureCounts version 1.6.2 44,45 . Differential expression of miR-NAs in salivary glands upon POWV infection was analyzed using DESeq2, version 1.20.0, following the DESeq2 vignette 46 . Validation of mature miRNAs by qRT-PCR. The miScript II RT kit (Qiagen) was used to convert RNA samples to cDNA. Mature miRNAs were selectively converted into cDNA by using the kit's miScript HiSpec Buffer for the reverse transcription reaction. 4 µL of RNA was used per 20 µL reverse transcription reaction. An additional reaction was prepared with no RNA template. The miScript II RT reaction was heated on a Mastercycler-Pro S thermos cycler (Eppendorf). The cDNA was diluted by adding 40 µL of nuclease-free water to bring the final cDNA volume to 60 µL. Real-time PCR was performed with a miScript SYBR Green PCR kit (Qiagen), using the kit's proprietary-sequence universal primer as the reverse primer and 10 µM of a miRNA-specific forward primer (Table 5). 2 µL of cDNA was used per PCR reaction, and the volumes of master mix and primer used were those recommended by the manufacturer. 25 µL reactions were loaded into the wells of a 96-well plate to create customized miScript PCR arrays. Each real-time PCR array was run on a CFX96 real-time PCR (BioRad) with the following cycle: 95 °C for 15 minutes, followed by forty cycles of 94 °C for 15 seconds, 55 °C for 30 seconds, and 70 °C for 30 seconds. The ramp rate was set at 1 °C/second. The iCycler's software was used to calculate the Ct values for all analyzed miRNAs. The ΔΔCt method was used to calculate fold-changes in gene expression between test groups (POWV-infected tick salivary glands) and control groups (uninfected tick salivary glands). The miRNA expression for each sample was normalized to the expression of RNU6-2 small nuclear RNA (Qiagen). Statistical significance was calculated by the 2-tailed Student's T-test comparing the miRNA expression in POWV-infected salivary glands versus uninfected salivary glands. Significant values were those where P < 0.05.

Primer name
Sequence (5′ -3′)  www.nature.com/scientificreports www.nature.com/scientificreports/ In vitro miRNA inhibition assay. The following miRNAs were selected for inclusion in the in vitro miRNA inhibition assay: isc-miR-1, isc-miR-124, isc-miR-184, isc-miR-315, isc-miR-5307, isc-miR-5312, isc-miR-nDS625977_65388, isc-miR-nDS630914_20990, and isc-miR-nDS752087_3745. The miRNA inhibitors were designed based on the miRNA sequences (Table 6) and synthesized by Integrated DNA Technologies. VeroE6 cells were used for this in vitro miRNA inhibition assay as they are an epithelial cell line, and fibroblasts are an early skin cell target of POWV at the tick feeding site 12 . 24-well plates of VeroE6 cells were seeded at 5 × 10 4 cells/well. 24 hours post-seeding, the VeroE6 cells were transfected in triplicate with 55 pmol of each miRNA inhibitor via Lipofectamine RNAiMax transfection reagent (ThermoFisher Scientific). As a control, cells were mock-transfected without a miRNA inhibitor. 24 hours post-transfection, cells were infected with POWV at a multiplicity of infection of 0.05. Mock-transfected cells were also infected with POWV. Cell supernatant aliquots of 50 µL were harvested from each replicate every 24 hours until 96 hours post-infection, and 50 µL of fresh infection media was added to each well to replace the sample volume.
Each cell supernatant sample was mixed with TRIzol LS reagent (ThermoFisher Scientific), and viral RNA extractions were performed using a combination of TRIzol LS and QIAamp Viral RNA mini (Qiagen) protocols as previously described 17 . Virus titer for each cell supernatant sample was determined by q-RT-PCR with the parameters outlined above for the tick salivary gland viral load analysis. Briefly, POWV genomic RNA was assayed using POWV-specific forward and reverse primers (Table 5) for q-RT-PCR. RNA was extracted from a POWV sample of known titer (expressed in focus-forming units), and serial dilutions were used to generate a standard curve for viral load quantification as previously described 17 . Ct values for each supernatant sample were converted to focus-forming unit equivalents based on the in-house generated standard curve. A 2-tailed Student's T-test was used to analyze the significance of POWV titer differences between miRNA inhibitor-transfected versus mock-transfected VeroE6 cells at each time point. Significant values were those where P < 0.05.