Tilting the balance between RNA interference and replication eradicates Leishmania RNA virus 1 and mitigates the inflammatory response

Naturally Abundant siRNAs Directed Against LRV1 of L. braziliensis and L. guyanensis.

Previous siRNA studies in Leishmania analyzed RNAs using a tagged Argonaute inserted into an δago1⁻ knockout of Lbr M2903, which lacks LRV1 (9, 29, 38, 39). Because the lines bearing LRV1 studied here had not been similarly modified, we sequenced total small RNAs (sRNAs) as an alternative. Lbr siRNAs bear a 5′-P and 3′-OH, reflecting their origin through the action of cellular Dicer nucleases (39), and we used these properties to make siRNA-focused sRNA (<42 nt) libraries for next-generation sequencing (Table S1). For Lgy, we chose the established LRV1⁺ Lgy M4147 strain (7), and three different Lbr shown to bear LRV1 by PCR and/or anti-dsRNA antibody tests (40).

For sRNAs from Lbr M2903 mapping to the Lbr reference genome, read length displayed a biphasic distribution, with a major peak centered around 23 nt (20 to 26 nt, 77.9% of total mapped reads) and a minor one around 33 nt (30 to 36 nt, 9.4% of total mapped reads) (Fig. 1A and Tables S1 and S2). The 33-nt peak reads mapped primarily to structural RNA loci (62% of mapped reads) (Table S2), similar to an sRNA class described in many eukaryotes, including trypanosomes and Leishmania lacking the RNAi pathway (41–44). In contrast, reads from the 23-nt peak showed properties similar to AGO1-bound siRNAs (39), including their size and the presence of one to two untemplated nucleotides at the 3′ end in about 21% of the reads (Fig. 1A and Table S1). The 3′ untemplated bases likely arise from the action of cellular terminal transferases because Leishmania sp. lack the HEN1 methyltransferase that normally blocks their action (39). When both AGO1-bound siRNAs and the 23-nt sRNA peak reads were mapped to the Lbr genome, their distributions were very similar, with the vast majority mapping to transposable elements (TEs) (Fig. 1B and Table S2; and see Fig. S1A) (39). We concluded that the 23-nt peak sRNAs (23-nt sRNAs) provide a reasonable proxy for siRNAs.

The properties of sRNAs from the LRV1-bearing Lgy M4147 and Lbr LEM2700, LEM2780, and LEM3874 mapping to the Lgy or Lbr reference genomes were similar to those of Lbr M2903, including the 23- and 33-nt sRNA peaks, genomic mappings, and the presence and level of 3′-nt extensions in the 23-nt sRNAs (Fig. 1, Fig. S1, and Tables S1 and S2). Importantly, a substantial fraction of sRNA reads obtained from the LRV1⁺ Lgy and Lbr lines mapped to the LRV1 genomes, ranging from 0.4 to 2.5% of the 23-nt mapped reads (Fig. 1B). Unlike those aligned to the nuclear genome, LRV1-mapped reads showed a single size distribution centered around 23 nt (Fig. 1A), with about 20% again showing short 3′ extensions (Table S1), typical of Lbr siRNAs and 23-nt sRNAs (39). LRV1-mapping 23-nt sRNAs showed no consistent strand- or region-specific biases in all four strains (Fig. S2), suggesting that they likely originated from the action of DICERs on the viral dsRNA genome.

We previously showed that LRV1 does not encode a trans-acting inhibitor of RNAi activity (30), and the presence of high levels of LRV1-directed sRNAs similarly suggests that it does not encode a strong cis-acting inhibitor. Importantly, the levels of 23-nt sRNAs mapping to LRV1s were in the same range as siRNAs mapping to an efficiently silenced Luciferase reporter (0.4 to 2.5% vs. 0.8% targeted by the long-hairpin/stem-loop transgene) (30, 39). Thus, LRV1 is able to persist in the face of a significant RNAi response, as judged by 23-nt sRNA levels.

LRV1 Can Be Efficiently Targeted by Transgenic RNAi.

These data are consistent with a model where RNAi activity and LRV1 replication have achieved a “balance” between viral synthesis and degradation, which might be shifted by increasing or decreasing RNAi activity. With an eye toward virus elimination, we focused on increasing LRV1-targeting siRNA levels through the use of transgenic RNAi methods developed previously (30), in which long-hairpin RNA is expressed at high levels from a stem-loop (StL) construct containing LRV1 sequences integrated into the ribosomal RNA locus (Fig. 2A). We targeted regions of LRV1 from the capsid or RDRP ORFs (Lgy M4147, Lbr LEM2700, and LEM2780), or a region that spanned them (Lbr LEM3874), ranging in length from 794 to 1,143 bp (Fig. 2B and Table S3); because the two viral genes reside within the same RNA segment, targeting either one should lead to degradation of the entire LRV1 RNA. Because LRV1 sequences diverge substantially between parasite strains (69 to 90% nt identity), “stems” specific for each species/strain were used. To assess nonspecific effects, we integrated an StL construct for an AT-rich GFP (GFP65 StL), which efficiently silences expression of GFP65 (30). The untransfected parental lines served as LRV1⁺ controls, and Lbr M2903 or Lgy M4147/HYG (12) served as LRV1-negative controls.

To screen for loss of LRV1, StL transfectants were analyzed by flow cytometry of fixed, permeabilized cells using an antibody raised against the Lgy M4147 LRV1 capsid (45), which cross-reacts with Lbr LRV1. For both Lgy M4147 (Fig. 3, Top) and Lbr LEM2780 (Fig. 3, Bottom), there was a clear separation in capsid staining between the LRV1-positive (Fig. 3, red) and LRV1-negative controls (Fig. 3, green). Although control GFP65 StL lines (Fig. 3, purple) had capsid protein levels similar to WT, capsid protein was undetectable in LRV1-targeted StL lines (Fig. 3, light and dark blue), indistinguishable from the LRV1-negative control. This result was observed whether the capsid or RDRP was targeted (Fig. 3). Similar results were obtained with LRV1 StL transfectants from Lbr LEM2700 and Lbr LEM3874. In support of the flow cytometry data, Western blot analysis with an anti-capsid antibody showed high LRV1 levels in the Lgy parental line and GFP65 StL transfectants whereas capsid protein was undetectable in the capsid StL transfectants (Fig. S3).

StL Constructs Result in High Levels of siRNAs Mapping to the LRV1 Stem.

Despite the insensitivity of LRV1 to “natural” levels of RNAi, as judged by the abundance of 23-nt sRNAs, introduction of StL constructs targeting LRV1 resulted in great reduction in LRV1 levels. To understand the basis for this reduction, we analyzed 23-nt sRNA peak reads mapping to the nuclear and LRV1 genomes, for one LRV1 StL transfectant of each species (Fig. 4). Remarkably, the percentage of total 23-nt sRNAs mapping to LRV1 had increased greatly from that seen in the WT parent, from 2.5 to 86.7% for Lgy and from 1.8 to 73.0% for Lbr LEM3874 (Fig. 4A and Fig. S1B). Concomitantly, the percentages of 23-nt sRNAs mapping to the nuclear genome was proportionately reduced, with some variability among loci and/or lines (for example, rRNA reads were unchanged in both species whereas tRNA reads decreased in Lgy) (Fig. 4A and Fig. S1B). Although we did not measure the absolute levels of sRNAs, previous studies have shown that these levels are tightly controlled by the level of Argonaute 1 and thus are unlikely to differ significantly (39). Essentially, all LRV1-mapping sRNAs in LRV1 StL lines now mapped only to the RNAi-targeted stem region (Fig. 4B, dark gray), as expected, because LRV1 had been eliminated (as discussed in Complete Virus Elimination Following RNAi of LRV1), which also argues against the occurrence of “transitive” siRNA formation (46, 47).

The levels of LRV1 23-nt sRNAs (76 to 87%) in LRV1 StL transfectants were much greater than those seen with siRNAs mapping to the LUC ORF/stem targeted using the same StL transfection construct (0.8%) (39). To rule out the possibility that this result arose from reliance on 23-nt sRNAs, we analyzed these 23-nt sRNAs from a line bearing the LUC StL RNAi reporter used in the siRNA studies [IR2-LUCStL(b)-LUC(a)]. From this analysis, 1.14% of the 23-nt sRNA peak reads mapped to the LUC ORF/stem, suggesting that use of 23-nt sRNAs vs. siRNAs did not significantly impact quantitation. To assess the target-specific effects, we compared these results with those quantitating 23-nt peak sRNAs after RNAi StL targeting of a panel of 10 chromosomal genes. For this group, 1.5 to 34% of 23-nt sRNAs mapped to the RNAi-targeted gene, compared with less than 0.02% basally. Thus, the StL-bearing vectors generate a high but variable level of sRNAs for all genes tested, with the LUC reporter being at the low end and LRV1 at the high end, which may reflect the fact that, although the LRV1 target is typically eliminated by RNAi (Fig. 3 and Complete Virus Elimination Following RNAi of LRV1), chromosomal RNAi targets continuously transcribe mRNAs. In other organisms, studies have shown that the presence of a cognate target facilitates the turnover of sRNAs; thus, the absence of an LRV1 target may lead to higher levels of siRNAs (48, 49). Future studies should address the factors contributing to the differences in sRNA levels among genes and to the very high steady-state levels of LRV1-directed 23-nt sRNAs seen here.

Complete Virus Elimination Following RNAi of LRV1.

RNAi-mediated LRV1 knockdown would be most useful as a tool if it resulted in a complete elimination of LRV1. To achieve a sensitivity beyond that of flow cytometry (∼20-fold) or Western blotting (∼100-fold), we validated a sensitive quantitative RT-PCR assay (qRT-PCR) for LRV1, using strain- and LRV1-specific primers to amplify a region located outside the stem regions (Fig. 2B and Table S4). Because the melting temperatures of PCR amplicons are sequence- and length-dependent, comparison of dissociation (melt) curves facilitated discrimination between specific and nonspecific amplification.

Because LRV1 copy number was estimated to be ∼100 per cell (50), a cutoff for classification as LRV1-negative was set at 10⁴-fold below WT. Analysis of Lbr qPCR data by the ΔΔC_t method (51) showed that most LRV1 StL transfectants had LRV1 RNA levels more than 10⁵-fold lower than WT (Fig. 5A and Fig. S4 A and B). Raw C_t values for LRV1 StL lines with LRV1-specific primers were indistinguishable from mock cDNA preparations, and ΔC_t values were indistinguishable from those of negative controls. Melt curves showed that products seen at C_t arose from nonspecific amplification (Fig. 5A and Fig. S4; white bars). As expected for control GFP65 StL lines, LRV1 RNA levels were similar to those in WT (Fig. 5A and Fig. S4; black bars).

Similar results were obtained with RNAi of LRV1 in Lgy M4147, with most transfectants showing reductions below the 10⁴-fold cutoff (Fig. 5B). However, low levels of LRV1 remained in two lines where the RDRP was targeted, ∼300- to 500-fold less than the parent line (Fig. 5B, black bars); here, melt curve analysis suggested that these products were LRV1-specific. Alternate primers targeting other regions across the virus gave similar results, suggesting the presence of intact LRV1. We hypothesized that this result was due to heterogeneity in viral load, with most but not all cells lacking LRV1. In support of this hypothesis, we generated and showed that all clonal lines arising from one of the “weakly positive” lines were negative for LRV1 by flow cytometry and satisfied the 10⁴-fold cutoff by qPCR (Fig. S4C). The occasionally incomplete LRV1 elimination is consistent with our prior observation that RNAi was somewhat less efficient in Lgy than in Lbr (30). Nonetheless, even for weakly positive Lgy transfectants, RNAi was sufficiently efficient for the ready isolation of LRV1-negative lines (Fig. 3, top; Fig. 5B; and Fig. S3).

LRV1 Knockdowns Induce Less Cytokine Production in in Vitro Macrophage Infection Assays.

Previous reports showed that LRV1⁺ Lgy stimulated the TLR3-dependent release of higher levels of cytokines from bone marrow-derived macrophages (BMDMs) than LRV1-negative strains (7). The availability of defined RNAi-derived LRV1-negative lines now allowed tests of this finding in Lbr, as well as confirmation of prior results obtained with a single isogenic LRV1- Lgy. Briefly, BMDMs were infected in vitro with LRV1 StL and GFP65 StL Lbr and Lgy transfectants, as well as positive and negative control lines, and the levels of two cytokines known to be induced by LRV1 (TNF-α and IL-6) (7, 10) were measured.

Capsid StL and RDRP StL LRV1-negative lines of both Lbr and Lgy induced significantly lower levels of cytokine production than did the LRV1-positive lines (both parental and GFP65 STL) (Fig. 6 and Fig. S5). Additionally, when macrophages from TLR3-deficient mice were infected with Lbr LEM2700, the LRV1-positive parasites no longer elicited higher levels of cytokine release (Fig. S5). Of note, all Lgy LRV1 StL lines induced background levels of cytokine release, including the two lines that retained low levels of LRV1 (Figs. 5B and 6B and Fig. S5), consistent with the observation that high levels of LRV1 were necessary for cytokine stimulation (7, 10).

Leishmania LRV1 and the Endogenous RNAi Pathway.

We identified two populations of sRNA in Lbr and Lgy. The less abundant 33-nt sRNAs mapped primarily to genes encoding structural RNAs (Table S2), as seen in other organisms, including trypanosomatids (41–44). In contrast, the more abundant 23-nt sRNA fraction exhibited properties similar to authentic AGO1-bound Lbr siRNAs (39), including size, the presence of 3′ untemplated bases at the same frequency (∼20%), and mapping primarily to transposable elements and repetitive sequences (Fig. 1 and Tables S1 and S2). Only 23-nt sRNA reads mapped to the LRV1 dsRNA genome (Fig. 1A), and these reads also bore 3′ nucleotide extensions at the same frequency, again consistent with an origin via the RNAi pathway (Table S1). Importantly, the levels of 23-nt sRNAs mapping to LRV1 constituted a substantial fraction of total aligned 23-nt sRNAs (Fig. 1B and Table S1), comparable with those targeting an efficiently silenced LUC reporter gene (30, 39). Thus, LRV1 can persist in the face of RNAi pressure that gives rise to sRNA levels comparable with that which efficiently silences a chromosomal target gene.

In other organisms, sRNA/siRNA levels provide a gauge of RNAi pathway recognition and targeting of viruses: When RNAi controls virus replication, as in plants, fungi, and insects (26, 32, 33), high levels of siRNAs accompany viral infections, leading to eradication of the virus. In mammals, quantitatively fewer siRNAs are present, which do not effectively control virus levels, at least in adult somatic tissues (34, 36, 37). In contrast, high levels of siRNA-like 23-nt sRNAs in Leishmania suggest an attack on LRV1 by the RNAi pathway, but the virus persists. Although many viruses encode trans-acting RNAi suppressors, mediating their survival (52), this seems unlikely for LRV1. There is no obvious coding potential for an RNAi suppressor in the compact LRV1 genome. Our studies here suggest that there is no strong cis-acting inhibitory activity, and we showed previously that a luciferase reporter was equally silenced in the LRV1⁺ and LRV1-negative Lgy studied here (30), which suggests a third model where LRV1 is targeted strongly by the RNAi pathway, but the RNAi-mediated degradation is “balanced” by virus replication or other factors. This balance is perhaps most likely mediated by the slicer activity of Argonaute; however, previous studies examining the role of RNAi in control of viruses frequently raise the possibility of Dicer-mediated control as well (53–55). It is likely that the sequestration of the LRV1 dsRNA genome within the capsid may also contribute by limiting the exposure of the LRV1 dsRNA to the RNAi machinery and other degradative pathways. In yeast, SKI genes act to prevent deleterious effects of L-A viruses toward its fungal host through alterations in mRNA degradation and/or surveillance (27), and homologs for several mRNA degradation and surveillance genes are evident in the Leishmania genome.

In other organisms, persistent viruses can also be maintained in the face of an active RNAi pathway, but at considerably reduced levels (26, 56). Over evolutionary time, this strong pressure likely accounts for the inverse relationship in fungi between virus levels and the activity and/or presence of the RNAi pathway, especially when associated with a selective advantage for viral retention, as seen with the yeast killer factors that are dependent on the L-A virus (26, 57). Similarly, in Leishmania, we had originally proposed that RNAi pressure would be sufficiently strong as to in some cases provide a driving force for loss of RNAi, to maintain LRV1-dependent increases in pathogenicity (30). Given the greater ability of LRV1 to survive in the presence of an active RNAi pathway, our data suggest that the magnitude of this effect may be considerably less than envisioned. However, even small pressure could prove a significant force toward down-regulating pathways impacting on LRV1 levels during evolution.

RNAi as a Tool for Generating LRV1-Negative Lines for Biology.

Following the predictions of the balance hypothesis, we aimed to increase activity against LRV1 through the increased synthesis of siRNAs targeting LRV1, which proved quite successful: The fraction of 23-nt sRNAs targeting LRV1 rose dramatically in lines expressing StL constructs targeting LRV1 (Figs. 1B and 4A). Correspondingly, the fraction of 23-nt sRNAs mapping to the Leishmania genome dropped proportionately, most of which again mapped to TEs and repeats (Fig. 4A). Importantly, LRV1 levels were dramatically reduced for all LRV1 StL transfectants, and, in most cases, the virus was eliminated, as judged by protein and RNA methods (Figs. 3 and 5 and Figs. S3 and S4). Targeting of either the capsid or RDRP gene eliminated LRV1, as was expected given that both are encoded by the same RNA (Fig. 2B). Only in Lgy were some transfectants found that retained low levels of LRV1, which could reflect less RNAi activity in this species, as was seen with reporter genes (30). However, most transfectants had completely lost LRV1.

Viral infection has been reported for Giardiavirus (58), and stable viral transfer for several fungal Totiviruses (59). However, de novo infection and stable viral transfer have been unsuccessful with Lgy (29), and reverse genetic systems have yet to be reported for any Totivirus. Therefore, the ability to reproducibly mediate viral cure by RNAi is of great value for biological studies of LRV1. Previous work used an LRV1-negative Lgy, which was obtained after transfection with an episomal Leishmania vector expressing resistance to hygromycin B, followed by a long period of growth under selection (12); however, this method seems to have been successful only once. Neither have we succeeded with several “stress-related” treatments that have proven effective in curing mycoviruses, such as yeast L-A (60). Our studies establish RNAi as a viable strategy for cure of LRV1 and perhaps other viruses in RNAi-competent Leishmania species.

LRV1⁺, but not LRV1-negative, Lgy, induces a “hyperinflammatory” cytokine response in infections of BMDMs in vitro, which is TLR3-dependent (6, 7). Infectivity tests of mouse BMDMs in vitro showed that RNAi-generated LRV1-negative Lgy lines likewise failed to induce a substantial cytokine response, as shown for two cytokines (TNF-α and IL-6) known to be diagnostic for an LRV1-driven innate immune response. Interestingly, this result occurred with RNAi-derived lines where LRV1 loss was substantial but incomplete (RDRP StL c3 and 4; 500- and 300-fold below parental levels, respectively) (Figs. 5B and 6B and Fig. S4C), consistent with data from natural Lgy showing low LRV1 levels (7). Thus, a partial reduction in LRV1 levels, similar to that seen in some natural isolates, is sufficient to ameliorate LRV1-dependent virulence, which may facilitate future efforts targeting LRV1 in human disease. Importantly, the continued presence of the integrated StL constructs seemed to have no “off target” effect in the BMDM infections, despite the high levels of transgene-derived 23-nt sRNAs present in these lines (Fig. 4); the LRV1 StL “cured” lines induced the release of cytokines at a level similar to that of StL-negative, LRV1-negative controls, and control GFP65 StL lines that maintained LRV1 induced the release of cytokines at a level similar to the StL-negative, LRV1⁺ parent (Figs. 3, 5, and 6). Additional studies will be necessary to determine whether this also pertains to other cell types or host infections.

LRV1-Dependent Virulence in L. braziliensis.

Previous studies of LRV1-dependent virulence focused primarily on Lgy; however, in humans, Lbr is associated with the larger share of MCL (14, 15). Our studies extend the generality of LRV1-dependent virulence to Lbr because LRV1⁺ Lbr likewise induces strong TLR3-dependent cytokine responses. These findings are especially important in light of published work on the association of LRV1 with MCL, with mixed results depending on the geographic region and methods used (6, 16–19). Our data show that, in in vitro infections, LRV1 contributes strongly to the proinflammatory phenotype associated with elevated pathogenicity, as seen in Lgy, which suggests that, in human infections, it may be informative to seek for correlations between LRV1 and the severity of CL in Lbr infections. Indeed, recent studies show that LRV1 in Lbr clinical isolates correlates with drug treatment failure (16), as was also seen in Lgy (20). Thus, although other parasite or host factors may play a significant role in the development of MCL (21, 22), current data now bolstered by our studies of isogenic LRV1⁺/negative lines support a role for LRV1 in the severity of human leishmaniasis caused by Lbr.

Parasites and in Vitro Culture.

Lbr LEM2700 (MHOM/BO/90/AN), LEM2780 (MHOM/BO/90/CS), and LEM3874 (MHOM/BO/99/IMT252 n°3) were from Patrick Bastien, Université de Montpellier, Montpellier, France; Lbr M2903 (MHOM/BR/75/M2903) was from Diane McMahon Pratt, Yale School of Public Health, New Haven, CT; and Lgy M4147 (MHOM/BR/78/M4147) and its derivative Lgy M4147/HYG were from Jean Patterson, Southwest Foundation for Biomedical Research, San Antonio, TX. Before introduction of StL constructs, parasites were transfected with the linear SSU-targeting SwaI fragment from B6367 pIR2SAT-LUC(B) (30), and clonal lines were derived, validated, and used. The luciferase-expressing clone of Lbr LEM2780 contained only LRV1-LbrLEM2780(b). Parasites were grown in fresh Schneider’s Insect Medium supplemented with 10% (vol/vol) heat-inactivated FBS, 100 μM adenine, 10 μg/mL hemin, 2 μg/mL biopterin, 2 mM l-glutamine, 500 units/mL penicillin, and 50 μg/mL streptomycin, and selective drugs as indicated below.

RNAi Stem-Loop Constructs.

Regions of interest from LRV1 were screened using the RNAit target selection tool to ensure that there was no homologous sequence in the parasite genome (61), amplified from cDNA by PCR using KlenTaq-LA polymerase, and cloned into the pCR8/GW/TOPO cloning vector (Thermo Fisher Scientific) using the protocol recommended by the manufacturer and a 20-min ligation. The stem segments and PCR primer sequences can be found in Table S3. The stems were transferred from the pCR8/GW/TOPO donor vector to the pIR2HYG-GW(A) (B6365) destination vector (which contains sequences from the parasite rRNA locus to enable integration into the genome and inverted LR recombinase sites for the generation of inverted repeat through Gateway technology) using LR Clonase II (Thermo Fisher) in an overnight reaction at room temperature. Reactions were terminated by incubating with proteinase K for 1 h at 37 °C. Constructs were verified by restriction digest.

Transfections.

Stable transfections were performed as previously described (30, 62). Clonal lines were obtained by plating on semisolid media with 50 µg/mL hygromycin B. After colonies formed, cells were grown to stationary phase in 1 mL of media and passaged thereafter in 10 mL of media with 30 µg/mL hygromycin B.

RNA Preparation and Quantitative Real-Time PCR.

Total RNA was prepared from log-phase cells dissolved in TRIzol reagent (Thermo Fisher) at 3 × 10⁸ cells per milliliter using the Direct-zol kit (Zymo Research) and eluted in 50 µL of nuclease-free water. The RNA was DNaseI-treated (Thermo Fisher) in a 200-µL reaction using the provided buffer and 20 units of enzyme for 1 h at 37 °C, purified using the RNA Clean & Concentrator-25 kit (Zymo Research), and eluted in 50 µL of nuclease-free water. Reverse transcription was performed using the SuperScript III first-strand synthesis kit (Thermo Fisher) according to the manufacturer’s instructions in a 20-µL reaction containing 0.25 μg of purified RNA. Control reactions contained the same amount of RNA but lacked the reverse transcriptase enzyme. For qRT-PCR, primers were designed to amplify ∼100-bp regions of the LRV1 genome that lie outside the stem regions (Table S4). qPCR reactions were performed with cDNA templates in 20 µL of total reaction volume using the Power SYBR Green Master Mix (Thermo Fisher), 5 µL of 10-fold diluted cDNA, and final primer concentrations of 0.2 µM. Reactions were run on the ABI PRISM 7000 Sequence Detection System (Applied Biosystems, Thermo Fisher). PCR amplification conditions were as follows: 50 °C for 2 min and 95 °C for 10 s followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. PCR products were confirmed to be specific by melt curve analysis. All experiments were performed in triplicate. Amplification of KMP-11 was used as an internal control to normalize parallel reactions.

Small RNA Sequencing.

sRNA libraries were generated from total RNA as described (39); briefly, a primer (5′-rApppATCTCGTATGCCGTCTTCTGCTTG/ddC for all samples except Lgy M4147, which used primer rApppTGGAATTCTCGGGTGCCAAGG/ddC) was ligated first to the 3′ end using truncated mutant T4 RNA Ligase (New England Biolabs), and then a second riboprimer (5′-GUUCAGAGUUCUACAGUCCGACGAUC) to the 5′ end with T4 RNA Ligase. cDNA was generated using reverse transcriptase and primer 5′-CAAGCAGAAGACGGCATACGA, and then PCR was performed with this primer in conjunction with primer 5′-AATGATACGGCGACCACCGACAGGTTCAGAGTTCTACAGTCCGA. Products corresponding to inserts of 10 to 50 nt were purified and taken for sequencing with Illumina HiSeq2500 technology. Sequences have been deposited in the National Center for Biotechnology Information (NCBI) Short Read Archive (accession SRP082553).

Bioinformatic Analysis of sRNAs.

The 5′ and 3′ adapter sequences were removed from the sRNA reads, those fewer than 15 nt were removed, and the trimmed reads were mapped to homologous LRV1 or Leishmania genomes [Lbr M2904 (63) or a draft Lgy M4147 genome (Bioproject PRJEB82; accession CALQ01000001–CALQ01004013)] using Novoalign software (www.novocraft.com); parameters were set as -F ILMFQ; -H; -g 40; -x 6; -R 5; -r; and -e 1000). A random strategy was used to align reads mapping to multiple regions, and hard clipping of low coverage bases at the 3′ end was performed. sRNA abundance was assessed directly, or after “collapsing” to remove duplicate reads using algorithms within the FASTX toolkit (hannonlab.cshl.edu/fastx_toolkit/index.html). To annotate transposable or repeated elements, we used RepeatMasker (www.repeatmasker.org) to identify known elements and/or BLAST to identify regions corresponding to Leishmania specific elements (SLACS, TAS, and TATE) (63). The annotations were collected in .bed file format for further use. Coverage was calculated by counting the number of reads that align to each strand of the LRV1 genome.

LRV1 Sequences.

From the sRNA sequences we assembled whole or partial LRV1 contigs, which were confirmed and completed by PCR amplification and sequencing. The sequences for LRV1-LbrLEM2700, LRV1-LbrLEM2780(a) and -(b), LRV1-LbrLEM3874, and a revision of the LRV1-LgyM4147 (formerly LRV1-4) (64) genome sequences were deposited in GenBank (accession numbers KX808483–KX808487).

LRV1 Capsid Flow Cytometry.

The development and optimization of the LRV1 capsid flow cytometry protocol will be described elsewhere. Briefly, 1 × 10⁷ cells were fixed at room temperature using 2% (vol/vol) paraformaldehyde (Thermo Fisher) in PBS for 2 min and then incubated in blocking/permeabilization buffer (BPB) [10% (vol/vol) normal goat serum (Vector Laboratories) and 0.2% Triton X-100 in PBS] for 30 min at room temperature. Anti-Lgy LRV1 capsid antibody (45) was added (1:20,000 dilution) and incubated at room temperature for 1 h. After two washes with PBS, cells were resuspended in 200 μL of BPB with Alexa 488-labeled goat anti-rabbit antibody (Thermo Fisher) (1:2,000 dilution) and incubated 1 h at room temperature. After two additional washes with PBS, cells were subjected to flow cytometry, and the data were analyzed using CellQuest software (BD Bioscience).

Western Blot, Macrophage Infections, and Cytokine Assays.

After an initial wash with PBS, 5 × 10⁷ parasites were resuspended in 100 µL of 1× PBS. Then, 1 × 10⁷ cells (20 µL) were lysed with 7 µL of 4× Laemmli’s gel sample buffer. After heating for 5 min at 95 °C, cell lysates were loaded and separated on a 10% polyacrylamide denaturing gel, transferred to a nitrocellulose membrane, and visualized by Ponceau Red staining. The membrane was blocked for 1h in 5% powdered milk diluted in TBS plus 0.05% Tween 20, incubated overnight at 4 °C with the g018d53 anti-capsid polyclonal antibody (1:5,000 in 1% milk TBS-Tween 20), washed 4 × 15 min at room temperature, incubated for 1 h with an anti-rabbit IgG antibody coupled to peroxidase (Promega) (1:2,500 in 1% milk TBS-Tween 20), washed again 4×, and finally revealed by ECL chemiluminescence (Amersham). Infections of BL6 mouse BMDM and cytokine assays were performed as previously described (7, 10).

Statement Identifying Institutional and/or Licensing Committee Approving Animal Experiments.

Animal handling and experimental procedures were undertaken with strict adherence to ethical guidelines relevant in both host countries. These guidelines are set out by the Swiss Federal Veterinary Office (SFVO) and are under inspection by the Department of Security and Environment of the State of Vaud, Switzerland. Experiments were carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals (65) of the US National Institutes of Health. Animal studies were approved by the Animal Studies Committee at Washington University (protocol 20090086) in accordance with the Office of Laboratory Animal Welfare's guidelines and the Association for Assessment and Accreditation of Laboratory Animal Care International.