Cellular and Molecular Life Sciences (2022) 79:209 https://doi.org/10.1007/s00018-022-04229-x Cellular and Molecular Life Sciences ORIGINAL ARTICLE Selective cleavage of ncRNA and antiviral activity by RNase2/EDN in THP1‑induced macrophages Lu Lu1,2 · Jiarui Li1 · Ranlei Wei3 · Irene Guidi1 · Luca Cozzuto4 · Julia Ponomarenko4 · Guillem Prats‑Ejarque1 · Ester Boix1 Received: 5 October 2021 / Revised: 21 February 2022 / Accepted: 3 March 2022 © The Author(s) 2022 Abstract RNase2 is the member of the RNaseA family most abundant in macrophages. Here, we knocked out RNase2 in THP-1 cells and analysed the response to Respiratory Syncytial Virus (RSV). RSV induced RNase2 expression, which significantly enhanced cell survival. Next, by cP-RNAseq sequencing, which amplifies the cyclic-phosphate endonuclease products, we analysed the ncRNA population. Among the ncRNAs accumulated in WT vs KO cells, we found mostly tRNA-derived frag- ments (tRFs) and second miRNAs. Differential sequence coverage identified tRFs from only few parental tRNAs, revealing a predominant cleavage at anticodon and d-loops at U/C (B1) and A (B2) sites. Selective tRNA cleavage was confirmed in vitro using the recombinant protein. Likewise, only few miRNAs were significantly more abundant in WT vs RNase2-KO cells. Complementarily, by screening of a tRF & tiRNA array, we identified an enriched population associated to RNase2 expression and RSV exposure. The results confirm the protein antiviral action and provide the first evidence of its cleavage selectivity on ncRNAs. Graphical abstract Keywords RNase · Virus · RSV · miRNA · tRNA · tRF Lu Lu and Jiarui Li both authors contributed equally to this work. Extended author information available on the last page of the article 13 Vol.:(0123456789) 209 Page 2 of 18 L. Lu et al. Introduction native and RNase2-KO cell lines exposed to RSV. The comparative study indicated that the knockout of RNase2 Human RNase2 is a secretory protein expressed in leu- in THP1-derived macrophages resulted in a heavier RSV kocytes with a reported antiviral activity against single titre and reduced cell survival. Next, we analysed the total stranded RNA viruses [1, 2]. RNase2 is one of the main non-coding RNA (ncRNA) population by amplification of components of the eosinophil secondary granule matrix. 2′3′-cyclic phosphate ends using the cP-RNAseq method- The protein, upon its discovery, was named the Eosinophil ology and by screening a library array of tRNA-derived Derived Neurotoxin (EDN), due to its ability to induce fragments. Results proved that RNase2 expression in mac- the Gordon phenomenon when injected into Guinea pigs rophage correlates to a selective ncRNA cleavage pattern. [3–8].Apart from eosinophils, RNase2 is also expressed in other leukocyte cell types, such as neutrophils and monocytes, together with epithelial cells, liver and spleen Results [8–10](https://​www.​prote​inatl​as.​org/). The protein belongs to the ribonuclease A superfamily, a family of secretory RSV activated the expression of RNase2 RNases that participate in the host response and combine in macrophages a direct action against a wide range of pathogens with diverse immunomodulation properties [2, 11]. RSV virus stock was obtained at a titration of 2.8 × ­106 RNase2 stands out for its high catalytic activity against ­TCID50/mL, as previously described [25] and THP-1 mac- single stranded RNA and its efficiency against several viral rophages were exposed to the RSV at a selected MOI of 1:1 types, such as rhinoviruses, adenoviruses and retroviruses, up to 72 h post of inoculation (poi). Following, we exam- including HIV [12–14]. Recently, presence of eosinophils ined whether RSV addition induced the RNase2 expres- and their associated RNases has been correlated to the sion in THP1-derived macrophages. The GAPDH gene was prognosis of COVID patients [15–17]. On the contrary, no used as a housekeeping gene control. Figure 1A shows that action is reported against the tested bacterial species [12, unstimulated macrophage cells had a constant and stable 18, 19]. In particular, among respiratory viruses, which transcriptional expression of RNase2 and it was significantly activate eosinophil recruitment and degranulation, the upregulated in a time-dependent manner upon RSV expo- human Respiratory Syncytial Virus (RSV), which is the sure. The significant RNase2 gene levels upregulation could principal cause of death in infants [20], is probably the be detected as early as at 4 h poi, with a sevenfold increase most studied model for RNase2 antiviral action. Indeed, at 72 h poi. Furthermore, to determine whether the induc- RNase2 levels have predictive value for the develop- tion of RNase2 mRNA levels correlated with an increase in ment of recurrent wheezing post-RSV bronchiolitis [21]. protein expression, ELISA and WB were conducted to detect RNase2 was proposed to have a role in the host response intracellular and secreted RNase2 protein of THP1-derived against the single stranded RNA virus [22] and early macrophages. At indicated poi time, culture medium and studies observed that RNase2 can directly target the RSV whole-cell extracts of macrophages exposed to RSV were virion [12]. Interestingly, the protein ribonucleolytic activ- collected for ELISA and WB, respectively. As indicated in ity is required to remove the RSV genome, but some struc- Fig. 1B, the secreted RNase2 protein was detected in human tural specificity for RNase2 is mandatory, as other family macrophages stimulated with RSV and was enhanced in homologues endowed with a higher catalytic activity are response to RSV in a time-dependent manner, while no devoid of antiviral activity [23]. significant change of secreted RNase2 was detected in In our previous work we observed that RNase2 is the control macrophages. However, the maximum concentra- most abundantly expressed RNaseA superfamily member tion of secreted RNase2 protein in macrophage culture was in the monocytic THP-1 cell line [24]. To broaden the detected at 48 h poi, with a slight decrease at 72 h poi. Like- knowledge of the immunomodulatory role and potential wise, a similar profile was obtained by WB (Fig. 1C). Taken targeting of cellular RNA population by RNase2 in human together, our results suggest that RSV induces both RNase2 macrophages, we built an RNase2-knockout THP-1 mono- protein expression and secretion in human THP1-induced cyte cell line using CRIPSR/Cas9 (clustered regularly macrophages. interspaced short palindromic repeats) gene editing tool. Transcriptome of the RNase2 knockout with the unedited RNase2 protects macrophages against RSV THP1-derived macrophage cells revealed that the top dif- ferently expressed pathways are associated to antiviral host To further characterize the protein mechanism of action, we defence (Lu et al., in preparation). Here, we explored the knocked out RNase2 gene by the CRISPR/Cas9 methodol- protein antiviral action by characterization of both THP-1 ogy. RNase2 gene was successfully knocked out in 2 out of 32 single cells THP-1 derived cell lines, named as KO18 and 13 Selective cleavage of ncRNA and antiviral activity by RNase2/EDN in THP1‑induced macrophages Page 3 of 18 209 the ribonuclease activity staining assay to evaluate the rib- onucleolytic activity of the samples from cell lysates and culture supernatants. According to the activity staining elec- trophoresis, two main activity bands can be visualized in wild-type macrophage lysate sample, with molecular weight sizes around 15 and 20 kDa, corresponding to previously reported native forms [2, 7]. Compared to the WT control, both activity bands were missing in the RNase2 knockout cell lines (Fig. 2D). In addition, as recent studies suggested that CRISPR/Cas9 frequently induces unwanted off-target mutations, we evaluated the off-target effects on these trans- duced monocytes. Here, we examined the top four potential off-target sites for the sgRNA1 (Table S1) and did not detect off-target mutation in the T7EI assay (Fig. S1). Overall, we confirmed that the RNase2 gene has been both structurally and functionally knocked out. The KO28 THP-1 cell line, which achieved full RNase2-knockout, was selected for all the downstream experiments. Next, we investigated whether RNase2 expression within macrophages contributes to the cell antiviral activity. First, THP-1 cells (WT and RNase2 KO) were induced into macrophages as described above. Macrophages were then exposed to RSV at MOI = 1 to investigate the kinetics of infection by monitoring both intracellular and extracellular RSV amplicon using probe RT-qPCR. Intracellularly, RSV levels increased during the first 24 h but decreased at longer periods (48–72 h), with a slow increase (0–4 h) followed by an exponential increase (4–24 h) and reaching a maximum at 24 h (Fig. S2A). At 24 h, RNase2 KO macrophages had Fig. 1  RSV activates the expression of RNase2 in THP1-induced macrophages. ­106 THP-1 cells/well were seeded in 6-well tissue cul- significantly more intracellular RSV than WT macrophages. ture plate and induced by 50 nM of PMA treatment. After induction, The extracellular RSV titre was also determined (Fig. S2B). macrophages were exposed to RSV under MOI = 1 for 2 h and then We observed that RSV increased in KO macrophage cell cul- cells were washed and replaced with fresh RPMI + 10%FBS (0 h time tures until 48 h and then stabilized at 48–72 h. While in WT point post of inoculation). At each time point post of inoculation, the supernatant and cells were collected to quantify the expression of macrophages, RSV profile showed an increase that reached RNase2. A qPCR detection of relative expression of RNase2 gene; a peak at 48 h and was followed by a decrease, significantly B concentration of the cells was controlled as 1­ 06 cells/mL, secreted higher RSV levels were detected in KO macrophage cul- RNase2 in culture supernatant was measured by ELISA and normal- tures at 24–72 h. Moreover, we monitored cell death using ized with alive cell number detected by MTT assay; C intracellular RNase2 protein in macrophage was detected by WB; “ + ” and “ − ” MTT assay and our results confirmed that RSV exposure indicate with or without RSV, respectively; * and ** indicate the sig- increased cell death in either WT or KO macrophages. nificance of p < 0.05 and p < 0.01, respectively However, significant differences were detected, where KO macrophage cell death upon RSV addition was higher than KO28. Both of KO18 and KO28 express a 15 residue-pep- in WT macrophage cells (Fig. 3). Altogether, we concluded tide in comparison to the wild type THP-1, which encodes that RNase2 KO macrophages burdening and cell death fol- the full-length protein of 134 residues (Fig. 2A). lowing RSV exposure is significantly higher in comparison After confirmation of successful gene deletion by Sanger to WT macrophages. The present results confirm the direct sequencing, we ensured that the expression of functional involvement of the macrophage endogenous RNase2 in the RNase2 has effectively been abolished. According to west- cell antiviral activity. ern blot (WB) assay, we can barely detect RNase2 in the KO18 cell lysate sample compared to the control sample Selective cleavage of RNase2 on ncRNA population and a total absence of signal is achieved in KO28 sample (Fig. 2B). Moreover, the total absence of secreted RNase2 Following, we aimed to identify the potential changes in by THP-1 cells was confirmed by ELISA assay in culture small RNA population associated to the expression of supernatant for KO28 line (Fig. 2C). Finally, we conducted RNase2 by comparison of WT and KO cell lines. Toward 13 209 Page 4 of 18 L. Lu et al. Fig. 2  CRISPR/Cas9 mediated knock out of RNase2 gene in THP1- ured by ELISA, the RPMI + 10%FBS complete culture medium was derived macrophages. A Scheme of the mutation of RNase2 caused used as a negative control, the supernatant was concentrated 50 × ; D by sgRNA1, the sequence was validated by Sanger sequencing; ribonuclease activity staining assay was used to confirm the removal replacement is indicated: red labelled sequence in wild type was of catalytic function. Cells were collected and resuspended in water replaced by the green labelled sequence, resulting in the coding frame and sonicated, cell lysates were loaded in each well at the indicated change and stop codon insertion; B western blot assay was applied to quantity detect RNase2 protein; C secreted RNase2 in supernatant was meas- analysis. Principal Component Analysis (PCA) confirmed good clustering within WT and KO and appropriate discri- minant power between the two groups (Fig. S3). Results revealed that RNase2 expression in THP-1 cells is mostly associated to significant changes in small RNAs and in particular in tRNA fragments and miRNAs popu- lation (see Figs. S4–S6 and additional files 1–3). Overall, we observed an overabundance of only few specific tRNA- derived fragments (Fig. 4, Table S2 and Fig. S5) and miR- NAs (Table S3 and Fig. S6). Fig. 3  Knocking out of RNase2 reduced the macrophages cell viabil- Analysis of differential sequence coverage between WT ity upon RSV inoculation. Macrophage cell viability was monitored and KO samples (Table S2) indicated that the preferen- by registering the absorbance at 570 nm using the MTT assay from 0 tial cleavage sites for RNase2 on tRNAs were CA and UA to 72 h post of RSV inoculation; “ + ” and “ − ” indicate in the pres- ence or absence of RSV, respectively; the star refers to the signifi- (Fig. 4). cance between KO + and WT + (*p < 0.05) Figure 5A illustrates the base preferences for B1 and B2 deduced from the differential sequence coverage analysis of bam files. Results highlighted a selectivity for pyrimidines this end, we applied the cP-RNA-seq methodology that is at B1 and preference for purines at B2, with a U/C ≥ 1 at the able to exclusively amplify and sequence RNAs containing 5′ side of the cleavage site, and a pronounced predilection a 2′3′-cyclic phosphate terminus, product of an enzymatic for A at the 3′ side. We also explored whether the cleavage endonuclease activity [26]. Total small RNA from WT and preference was dependent on the RNA adopted secondary. KO-THP-1 cells was purified and the 20–100 nt fraction We can see how RNase2 preferentially cuts at tRNA loops, was extracted and processed as described in the method- mostly at the anticodon loop and secondarily at the D-loop, ology section. Sequence quality control indicated that for as well as stem regions near the anticodon loop (Fig. 5B). all samples more than 95% sequences achieved an average On the other hand, analysis of ncRNA products by cP- value > 30 M reads. Following RNAseq amplification, the RNAseq revealed abundance of miRNA products. Interest- sequence libraries were inspected by differential enrichment ingly, out of the more than 2000 miRNAs in human genome 13 Selective cleavage of ncRNA and antiviral activity by RNase2/EDN in THP1‑induced macrophages Page 5 of 18 209 Fig. 4  Predicted cleavage sites of the most significantly abundant of the five prime fragments (blue arrows) and the 5′ terminal posi- tRNA-derived fragments identified by cP-RNAseq. Parental tRNAs tions of three prime fragments (green arrows) according to the dif- with significant coverage differences between WT and RNase2-KO ferent coverage. Only sequences with a ­log2fold > 0.5 and adjusted p macrophage cells are depicted and the identified tRFs are marked in value < 0.05 are included red. The possible cleavage sites are based on the 3′-terminal positions 13 209 Page 6 of 18 L. Lu et al. Fig. 5  Overall cleavage prefer- ence of RNase2 on tRNAs in THP-1 macrophage cells of WT vs RNase2-KO identified by cP-RNA-seq. A Estimated base preference at 5′ and 3′ of cleav- age site (B1 and B2, respec- tively) deduced from analysis of differential sequence coverage. B Percentage of predicted cleavage location is depicted from low (white) to high (blue) values there were only 14 miRNA products significantly altered, at the 3′ side of the cleavage target, we did not observe a of which several derived from the same parental miRNA clear distinct preference, with only a slight tendency for precursor (additional file 3). Inspection of miRNA overrep- U, followed by A. resented in WT vs RNase2-KO by comparison of sequence Overall, most of the parental tRNAs precursors identi- coverage indicated a marked cleavage preference at the end fied by the array screening matched the identified by the of stem regions with a less defined base preference (see cP-RNAseq assay (> 70%), although some differences Fig. 6 and Table S3). Overall, we can infer for RNase2 a were observed in the identity of the accumulated frag- slight preference for U and C at B1 site, followed by G, and ments and their relative fold change. To note, few of the no defined consensus for B2. top listed precursors by the library screening ­(LysCTT​ and In parallel, we decided to explore the changes in tRNA- ­MetCAT​) were also present in the amplified sequences by derived fragments population associated with RNase2 the cP-RNAseq, but with a lower abundance (Additional expression in THP-1 cells by screening a tRNA-derived frag- file 2). On the contrary, few of the top listed fragments ment library. The nrStar™ Human tRF&tiRNA PCR array spotted by the later methodology were absent from the includes a total of 185 regulatory tRNA-derived fragments, commercial library array. of which 110 are taken from the tRF and tiRNA databases Notwithstanding, we must bear in mind that the results and 84 have been recently reported in the bibliography. obtained from the tRNA array screening are determined Using the nrStar™ Human tRF&tiRNA library, we by the intrinsic library composition. The tiRNA & tRFs found that out of a total of 185 tRNA fragments, only 5 list is based on the previously reported tRNA fragments, were significantly decreased in RNase2 KO macrophage in which have been identified mainly by the characteriza- comparison to the WT control group in non-treated samples tion of other RNases [27]. On the contrary, by cP-RNAseq and 22 under RSV exposure: 6 tiRNAs, 4 itRFs, 9 tRF-5, 4 method only the RNA by-products of an endonuclease tRF-3, 1 tRF-1 (see Table 1). Overall, the most significant enzymatic cleavage are amplified. changes associated to RNase2 in both cell cultures (p < 0.01) In addition, we confirmed tRNA cleavage by RNase2 are observed in release products from few parental tRNAs. in vitro by assaying the recombinant protein activity on Upon inspection of sequence of the putative cleavage synthetic tRNAs. The most representative sequences iden- sites we observed most sites at or near any of the tRNA tified by tiRNA&tRFs list (p < 0.01) screening and two loops, with predominance of anticodon loop (~ 50%), fol- additional sequences found using the cP-RNAseq method, lowed by D-loop (Table S4). Moreover, we observed that were incubated with RNase2 and tRNA cleavage prod- the preferred target sequences in WT samples were sig- ucts were visualized by urea-PAGE electrophoresis (see nificantly enriched with U at the 5′ position of cleavage Fig. S9). The results highlighted a selective cleavage by site, although not enough data is available for a proper RNase2. Comparative analysis using RNase5 as a refer- statistical analysis. On the other hand, when analysing ence showed a similar but differentiated cleavage pattern all the conditions, in the presence or absence of RSV, we for both RNases. Besides, we obtained a distinct distribu- also observed overall a significant preference for U at B1 tion of tRFs products as a function of the parental tRNAs at loop sequences. Moreover, the most significant tRNA substrates. fragments associated to RNase2 presence showed a U/C The present study is the first characterization of the cata- cleavage target for B1 at the anticodon loop (Table S4). On lytic activity of RNase2 on cellular ncRNAs and represents the contrary, the main base at B1 was G when the cleavage the first evidence of a specific release of tRNA fragments site was located at a stem region. As for the base located associated to RNase2 expression. 13 Selective cleavage of ncRNA and antiviral activity by RNase2/EDN in THP1‑induced macrophages Page 7 of 18 209 Fig. 6  Representative miRNA overrepresented in WT vs RNase2-KO precursor miRNA were predicted based on the 3′-terminal positions samples ­(log2fold > 2 and p < 0.05) and predicted cleavage sites on of each miRNA in bam files precursor miRNA to release mature miRNAs. Cleavage sites in the Discussion is particularly abundant in monocytes [11], which are key contributors to host defence against pathogens. A number Expression of human RNase2 is widely distributed in of host defence-associated activities have been proposed for diverse body tissues, such as liver and spleen, together with RNase2, mostly associated to the targeting of single stranded leukocyte cells [2]. Among the blood cell types, RNase2 RNA viruses [1]. In particular, the protein has been reported 13 209 Page 8 of 18 L. Lu et al. Table 1  tRNA fragment Condition Transcript name* p value Fold change Type tRF&tiRNA precursor population changes upon RNase2 knockout and/or RSV WT vs KO 3'tiR_088_LysCTT (n) 0.0085 4.11 3-half LysCTT (n) exposure tiRNA-5033-LysTTT-1 0.0104 16.54 5-half LysTTT​ 1039 0.0183 80.58 tRF-1 Pre-ArgCCT​ 3002A 0.0108 8.03 tRF-3 ProAGG​ 5028/29A 0.0352 2.41 tRF-5 GluTTC (n) KO + RSV vs 3'tiR_060_MetCAT (n) 0.0091 3.62 3-half MetCAT (n) WT + RSV TRF62 0.0360 2.39 itRF MetCAT​ TRF315 0.0186 3.29 itRF LysCTT​ TRF353 0.0243 3.90 itRF GlnTTG​ TRF419 0.0214 2.16 itRF LeuTAG​ tiRNA-5030-LysCTT-2 0.0322 2.94 5-half LysCTT​ tiRNA-5031-HisGTG-1 0.0345 2.79 5-half HisGTG​ tiRNA-5031-GluCTC-1 0.0155 3.05 5-half GluCTC​ 1001 0.0475 2.92 tRF-1 SerTGA​ 1013 0.0488 4.36 tRF-1 AlaCGC​ 1039 0.0105 2.39 tRF-1 ArgCCT​ 3004B 0.0017 5.38 tRF-3 GlnTTG​ 3006B 0.0473 2.13 tRF-3 LysTTT​ 3016/18/22B 0.0116 2.20 tRF-3 LysCTT/ MetCAT​ 5024A 0.0118 3.84 tRF-5 LeuTAA​ 5032A 0.0346 2.42 tRF-5 AspGTC​ TRF356/359 0.0398 3.63 tRF-5 ArgTCG (n) TRF366 0.0285 2.85 tRF-5 ThrTGT​ TRF365 0.0253 7.96 tRF-5 ThrTGT​ TRF396 0.0462 2.39 tRF-5 AlaAGC​ TRF457 0.0420 4.43 tRF-5 SerAGA​ TRF550/551 0.0398 4.84 tRF-5 GluTTC/ GluCTC​ P value < 0.05 and absolute value of the fold change > 2 was set as the significant threshold *185 tRNA fragments from four groups (WT, wild-type macrophage; KO, RNase2-knockout macrophage; WT + RSV, WT macrophage exposed to RSV; KO + RSV, KO macrophage exposed to RSV) were detected by using nrStar Human tRF&tiRNA PCR Array and compared between each other to reduce the infectivity of the human respiratory syncytial levels of RNase2 and upon 4 h of RSV addition the protein virus (RSV) in cell cultures [2, 12, 23]. transcription is significantly activated, showing a time-line Here, we studied RNase2 expression in THP1-derived correlation between RSV population and RNase2 expres- macrophages exposed to RSV. Previous work indicated sion. Besides, a significant increase of the secreted protein is that RNase2 is the most abundant RNaseA family member detected by ELISA after 24 h of inoculation, reaching a peak expressed in this human monocytic cell line [24, 28] (https://​ at 48 h (Fig. 1). It was previously reported that human mono- www.p​ rotei​ natla​ s.o​ rg/). Viruses can manipulate cell biology cyte-derived macrophages challenged with a combination of to utilize macrophages as vessels for dissemination, long LPS and TNF-α produced RNase2 in a time-dependent man- term persistence within tissues and virus replication [29]. ner [33]. However, we did not find any significant change Although the epithelial airway is the principal RSV target, of transcriptional expression of RNase2 upon Mycobacteria macrophages mostly contribute to the overall host immune aurum infection [24]. Discrepancy of expression induction is response [30–32]. In our working model, we observed how also found for RNase2 secretion by eosinophils upon distinct RSV enters human THP1-induced macrophages within the bacterial infections. For example, Clostridium difficile and first 2 h of inoculation. A fast proliferation of RSV takes Staphylococcus aureus infection caused release of RNase2, place after 4 h post of inoculation and titre of intracellular while Bifidobacteria, Hemophilus, and Prevotella species RSV viruses reaches the highest peak at 24 h. Non-treated infection did not [2]. In agreement, our previous work on human THP1-derived macrophages stably expressed basal THP1-derived macrophages infected by mycobacteria also 13 Selective cleavage of ncRNA and antiviral activity by RNase2/EDN in THP1‑induced macrophages Page 9 of 18 209 discarded any induction of RNase2 expression [24]. In con- which stands out among the identified tRFs by our library trast, using the same working model and experimental pro- array screening (Table 1), is reported to block retroviruses tocol, we observe here how RSV exposure significantly acti- replication, such as in HIV-1, by direct binding to the virus vate both the expression and protein secretion of RNase2 in priming binding site [49–51]. Other regulatory tRNA frag- THP1-derived macrophage cells (Fig. 1). The present study ments underrepresented by RNase2-KO (such as 5′ ­LysCTT​ corroborates previous reports on RNase2 involvement in and ­GluCTC​ halves) (Table 1) were previously reported to be host response to viral infections [1, 2, 34]. In particular, released by RSV infection [36, 37]. our work highlights the protein role in macrophage cells The present experimental data constitute the first evidence challenged by RSV. More importantly, we observe how the on the specific cleavage by RNase2 of cellular ncRNA. knockout of RNase2 in macrophage derived cells results in Results obtained by both cP-RNAseq and tRFs array screen- enhanced cell death (Fig. 3). ing indicate that RNase2 expression in macrophages is asso- Considering previous reports on the contribution of ciated to the significant enrichment of selective miRNAs RNase2 enzymatic activity in the protein antiviral activity and tRNA-derived fragments (Figs. 4 and 6; Tables S2 and [12] and the evidence that RSV infection alters the cellular S3). In addition, a higher frequency of cleavage takes place RNA population, including the specific release of regula- at tRNA single stranded regions, with predilection for the tory tRFs [35–37], we decided here to analyse the contribu- anticodon loop, followed by the D arm (Fig. 5). tion of macrophage endogenous RNase2 on cellular small Exhaustive analysis of differential sequence coverage in RNAs. Increasing data demonstrates that small noncoding WT and RNase2-KO THP-1 cells suggests that RNase2 pref- RNAs (ncRNAs) play important roles in regulating antivi- erentially targets at UA and CA sequences at tRNA loops. ral innate immune responses [38–40]. In particular, ncR- Recently, Bartok and co-workers reported a RNase2 selec- NAs derived from tRNAs, such as tRNA halves (tiRNAs) tivity for U at B1 site in synthetic RNAs [28]. Interestingly, and tRNA-derived fragments (tRFs), have been identified according to Hornung and co-workers, the release of U > p and proven as functional regulatory molecules [41]. RSV ends by RNase2 would participate in the activation of TLR8 infection together with other cellular stress processes can at the endolysosomal compartment and will contribute to regulate the population of tiRNAs and tRFs. For example, sense the presence of pathogen RNA [52]. To note, we find it has been demonstrated that RSV infection and hepatitis a good agreement between RNase2 substrate specificity viral infection can induce the production of tRFs and tRNA- identified in the present cell assay study on tRNAs and the halves, and their release has been related to RNase5 activ- previously reported for synthetic single stranded oligonucle- ity [36, 42]. RNase5, also called Angiogenin (Ang) due to otides [53, 54] (see Table 2). However, some differences are its angiogenic properties, is one of the most well-known evidenced at the miRNAs cleavage and in particular at the ribonucleases that are responsible for endonucleolytic cleav- B2 site specificity, which does not fully match the reported age of tRNA [43–45]. Surprisingly, the release of a specific on synthetic substrates. This discrepancy is also evident for tRF, tRF5-GluCTC​, which targets and suppresses the apoli- the other two RNaseA family members described to release poprotein E receptor 2 (APOER2), can also promote the specific tRFs [55–58], i.e., RNase5/Ang and Onconase, an RSV replication [36, 37]. In the present study, although the tRF5-GluCTC​ was not significantly altered by RSV addition alone, we observed a significant reduction in the RNase2 Table 2  Comparison of RNase2 cleavage specificity on ncRNA in THP1-derived macrophages (this study) with synthetic RNA sub- knockout cell line challenged with RSV (Table 1). Besides, strates in vitro tiRNA-5034-ValCAC​-3, the 5′half originated from ­ValCAC​, is Synthetic Base prefer- Ref Cellular Base prefer- identified both in the present work (Table S4) and the previ- RNA ence RNA* ence ous mentioned study associated to RSV infection [36]. To note, we found that RSV induced the production of 4 tRFs B1 in WT macrophages, in agreement to the previous study that polyU/polyC U/C ~ 2 [98] tRNA (loops) U/C ~ 1.2 indicated that RSV infection induced the release of tRFs polyU/polyC U/C ~ 1 [97] tRNA (stems) in A549 epithelial cells [36], although most of the tRNA ORNs# U/C ~ 1.3 [28] miRNA U–U (loops) products differ, which may be attributed to the specific basal G–G (stems) composition of each source cell line [46]. Moreover, tRFs B2 production is also observed to be dependent on the specific UpA/UpG (A/G ~ 47) [97] tRNA (loops) A viral infection type; for example, human metapneumovirus, UpA/UpG (A/G > 100) [59] miRNA in contrast to RSV, did not induce tRFs but significantly ORNs# G≥A≥U≥C [28] altered miRNA population[40, 47]. Interestingly, release of # ORNs (short single-stranded oligonucleotides: ssRNA40, 9.2 s tRNA products is mostly associated to an antiviral defence RNA, and R2152) mechanism [48]. For example, the tRF3 from ­tRNALysTTT​, *This study 13 209 Page 10 of 18 L. Lu et al. RNase purified from Rana pipiens with antitumoral proper- replication. However, caution should be taken when extract- ties (Table S5). Previous kinetic studies on RNaseA family ing conclusions, as miRNA databases are strongly biased cleavage preference using single stranded RNA substrates from a predominance of previous clinical studies. We must revealed a specificity for pyrimidines at the main B1 site also bear in mind that our working model (THP-1) is a leu- and preference for purines at B2 [53, 54]. Among the fam- kaemia cell line. Accumulation of miRNAs can be toxic to ily members, we observe distinct preferences for U vs C the cells, due to their potential interference with the transla- and A vs G at B1 and B2 sites, respectively. Interestingly, tion of essential proteins. Raines and co-workers correlated RNase 2 shows a marked preference for U at B1 and A at B2 miRNA release by Ang with potential toxicity to cells [45]. on synthetic oligonucleotides [53, 59], which mostly corre- In any case, in our working model we do not observe any sponds to the observed preference identified by cP-RNAseq change on the viability of both WT and KO cell lines. for tRNA in this study (Fig. 5). Nevertheless, our analysis On the other hand, we should be aware not to over inter- on tRNA cleavage sites would suggest a U/C ratio for B1 a pret our results that could be also somehow biased by the bit lower than the estimated for some synthetic substrates applied methodologies. The RNAseq methodology might (Table 2). lead to underrepresentation of some fragments, due to their The present study on cellular ncRNA also highlights the relative size, short half-life or presence of posttranscrip- key role of the RNA 3D structuration. Overall, our data tional modifications. Therefore, caution must be taken when reveal a cleavage preference by RNase2 at single stranded conclusions are drawn from the analysis of tRNA cleavage sequences and secondarily at stem adjacent to loop regions. product population. Another important source of variability Besides, we also observe that location of the targeted site comes from the presence of posttranscriptional modifica- also influences the cleavage base preference. tions, which can influence both the RNase selectivity and Interestingly, previous kinetic and structural studies on the product amplification step [26, 70]. Unfortunately, the RNaseA highlighted the unusual enhancement of the protein current ncRNA databases are still incomplete and lack full affinity to a dinucleotide probe by addition of a phosphate information on the precise post-transcriptional modifications linkage insert that can adopt a contorted conformation close that take place in vivo and might intervene in the RNases to the cleavable 3′5′ phosphodiester bond [60–62]. Likewise, recognition target. previous work on Onconase nucleotide base selectivity also More importantly, the array screening technique is encountered significant differences in vitro among di and prone to be biased by the selection criteria used to build the tetranucleotides [63] and tRNA [57]. tiRNA&tRF array; a library composed on previously avail- In addition, the cleavage of tRNAs by RNases would able experimental data, i.e., products by RNases, such as probably be modulated by the presence of regulatory pro- Dicer, Angiogenin, RNaseP or RNaseZ. This might explain teins within the cell [64]. For example, tRNA can be pro- some of the differences observed in the identified fragments tected by Schlafen 2 (SLFN2) protein from cleavage by when comparing the screening of the tRFs array and the RNase5 during oxidative-stress response [65]. Onconase amplified sequences by the cP-RNAseq methodology, which selectivity for specific tRNAs was attributed to the presence only amplifies the products of an endonuclease cleavage. in vivo of RNA binding proteins that might protect RNA On the other hand, we should also take into account regions from RNase activity [56, 66]. In this context, we the protein traffic and accessibility to the distinct subcel- should consider the formation of regulatory complexes, such lular compartments in the assayed experimental condi- as the RISC formation, the binding of Argonaute (AGO) tions. For example, in contrast to RNase5, mostly located subunits [67] or interactions with the RNHI. On its turn, the at the nucleus, RNase2 is associated to the endolysosomal released tRNA products would regulate the formation of cel- compartment [28, 52, 67]. In addition, cleavage of cellular lular complexes. For example, tRF3 interaction with AGO2 mature tRNAs might occur during stress conditions, when mediates the cleavage of complementary Priming Binding leakage of the RNases to the cytosol is favoured. We should Sequences (PBS) in retroviruses and thereby can avoid the bear in mind that the cell cytosolic RNA is in normal condi- replication of endogenous virus elements [68]. Interestingly, tions protected by the presence of the RNHI, which would a recent study demonstrates a direct interaction between lose its functional conformation under stress conditions due AGO4 and the RSV-induced ­GluCTC​-tRF5 fragment [69]. to oxidation of surface exposed Cys residues [71]. Accord- Among the ncRNA population mostly altered upon ingly, in the case of RNase5/Ang, it has been described that RNase2 knockout, we find, together with tRFs, miRNAs. the selective tRNA cleavage takes place in oxidative condi- Interestingly, the identified miRNAs subproducts come tions [55, 72, 73]. This might also explain the much higher frequently from the same group of parental pre-miRNAs. number of tRNA fragments obtained in the present study Inspection on information related to these miRNAs entities in the RSV treated vs non-treated cells (Tables 1 and S4). reveals a predominance of miRNAs associated to cancer and Notwithstanding, recent bibliography indicates that other neurological disorders, although few are also related to virus proteins may intervene in the regulation of the RNase–RNHI 13 Selective cleavage of ncRNA and antiviral activity by RNase2/EDN in THP1‑induced macrophages Page 11 of 18 209 complex [67] and further studies are needed to fully interpret macrophages exposed to RSV confirms the RNase involve- our results. ment in the cell host antiviral defence. By amplification Interestingly, under certain cell conditions, such as nutri- of 2′3′cP end RNA products, we have identified the tRNA tion deficiency or oxidative stress, RNase5/Ang is reported fragments and miRNAs associated to RNase2 cleavage. The to stimulate the formation of cytoplasmic stress granules analysis of RNA recognition regions reveals the RNA base and produce tRNA-derived stress-induced RNAs (tiRNAs) composition at the 5’ and 3’ of cleavage site. To note, tRNA [74–76]. The released tiRNAs functionally enhance damage cleavage is mostly favoured at anticodon and d-loops at UA repair and cellular survival through suppressing the forma- and CA sites. Further work is mandatory for an unambigu- tion of the translation initiation factor complex or associat- ous pattern assignment towards the understanding of how ing with the translational silencer [44, 68]. Ivanov and co- RNase2 can shape the ncRNA population and its antiviral workers recently characterized the structural determinants role. that guide release of tiRNA population during stress condi- tions [58, 77]. Accessibility of tRNAs will also depend on their potential entrapment in Tbox riboswitch or RNA gran- Materials and methods ules, which are abundant in starvation situation and have an unequal propensity to protect tRNA from cleavage. Besides, Plasmid construction proteomic analysis revealed the presence of RNHI within stress granules [78], an inhibitor protein that can complex For long term consideration, we used the two-plasmid sys- to RNase5/Ang and other regulatory proteins to control cell tem to run the CRISPR gene editing experiments instead translation [67]. of using all-in-one CRISPR system. Thus, we constructed Another important source of variability comes from the a pLenti-239S coding Cas9, GFP and Puromycin resist- assayed cell type. Although it is widely accepted that the ance gene for the knockout assay by replacing the sgRNA levels of parental tRNAs differ significantly upon cell con- expression cassette of LentiCRISPRv2-GFP-puro (gifted ditions and tissues [27] and a very unequal tissue distribu- by Manuel Kaulich) short annealed oligos; for activation tion is observed for the more than 500 tRNAs listed in our assay, we cloned the eGFP into lenti-dCAS-VP64-Blast genome, little is still known of their relative expression rates. (Addgene61425, gifted by Manuel Kaulich), the new plas- Last but not least, the specific expression pattern of RNases mid was named pLenti-239G. Finally, plenti-239R, a new associated to each cell type and experimental condition will lentiGuide plasmid coding sgRNA expression cassette and definitely contribute to shape the ncRNA population. We Cherry fluorescence gene was created by using the Cherry must also take into consideration that our results are based gene (gifted by Marcos Gil, UAB) to replace the Cas9 of on a single monocytic cell line, which undoubtedly can- LentiCRISPRv2-GFP-puro (gifted by Manuel Kaulich). The not fully reproduce the alveolar macrophages that infiltrate primers used for PCR are listed in Table S6. within the lung epithelial barrier during virus respiratory tract infections. sgRNA design and clone into pLentiGuide Despite the inherent limitations of this study, our results (pLenti239R) vector confirm that RNase2 can target ncRNA and release specific miRNAs and tRFs. Particular interest should be drawn to N20NGG motifs in the RNase2 locus were scanned, and the new identified tRNA fragments associated to RNase2 candidate sgRNAs that fit the rules for U6 Pol III transcrip- and absent from the commercial library array, which repre- tion and the PAM recognition domain of Streptococcus pyo- sent potential new regulatory elements for future studies. A genes Cas9 were identified. From CRISPOR (http://​crisp​or.​ growing evidence emphasizes the key role of tRNA halves tefor.n​ et/) and CRISPR-ERA, the top 2 sgRNA were selected and tRFs in regulating cellular functions [48, 79–81]. Deci- for knockout RNase2. Using the same procedure, potential phering the contribution of the distinct RNases to shape the OT sites were also predicted. The sequences are listed in cell ncRNA population should be key to analyse the cell Table S1. Oligonucleotides were annealed and cloned into response to adapt to distinct stress conditions, such as viral BbsI-digested pLenti-239R. The resulting plasmids contain- infection. ing sgRNAs were further confirmed by Sanger sequencing. Cell culture Conclusions HEK293T cell line was kindly provided by Raquel Peq- This is the first report of RNase2 selective targeting of uerul (UAB). HEK293T cell line was maintained in ncRNA. We have engineered a THP-1 cell line defective in DMEM (Corning Life Science) supplemented with 10% RNase2. Comparison of native and knockout THP1-derived foetal bovine serum (FBS) (Gibco). The culture media was 13 209 Page 12 of 18 L. Lu et al. replaced every 2–3 days, and the cells were passaged using 95 °C for 10 min followed by 30 cycles of 95 °C for 30 s, Trypsin–EDTA Solution (Gibco, 25200056). annealing at 60 °C for 30 s, and extension at 72 °C for 30 s. Human THP-1 cells (NCTC #88081201) were maintained The pairs of primers were designed to amplify the region or passaged in 25 or 75 ­cm2 tissue culture flasks (BD Bio- that covers the two possible double breaking sites. The PCR sciences) using RPMI-1640 (Lonza, BE12-702F) medium product is 410 bp. After each reaction, 200 ng of the PCR with 10% heat-inactivated FBS at 37 °C in humidified 5% products were purified using a QIAquick PCR Purifica- ­CO2 conditions. The culture media was replaced every tion Kit (QIAGEN), subjected to T7E I assays, and then 3 days. analysed by agarose gel electrophoresis. The indel mutation was confirmed by Sanger Sequencing. Generation of lentiviral vectors Construction of RNase2‑KO THP‑1 cell line To make the cell reach 90% confluence for transfection, 7.5 × ­106 of HEK293T cells were seeded in T75 culture A lentiviral system was selected to deliver CRISPR com- flask with 15 ml DMEM + 10% FBS complete medium 1 ponents into the THP-1 monocytic cell line. Cas9 and sgR- day before transfection. The lentiviral plasmids were trans- NAs lentiviral particles were produced in HEK293T cells fected into HEK293T cells using calcium phosphate protocol by Calcium phosphate precipitation method as previously [82]. Briefly, 36 µg transfer plasmid, 18 µg psPAX2 packag- reported [82–84]. THP-1 cells were then transduced with ing plasmid (Addgene#35002, gifted by Marina Rodriguez) the concentrated lentiviral particles with 10 µg/mL of poly- and 18 µg pMD2G envelope plasmid (Addgene#12259, brene, the overall transduction efficiency is about 7% (Fig. gifted by Marina Rodriguez) were mixed. Next, 93.75 µL S7). We designed 2 single guide RNAs (sgRNAs) (Table S1) of 2 M ­CaCl2 was added and the final volume was adjusted targeting the RNase2 locus to generate double strand breaks to 750 µL with ­H2O. Then, 750 µL of 2 × HBS buffer was (DSBs) (Fig. S8A). T7EI assay was employed to select the added dropwise and vortexed to mix. After 15 min at room more active guide RNA, achieving the knockout efficiencies temperature, 1.5 mL of the mixture was added dropwise to of about 40% (Figs. S8B and S8C). The GFP and cherry red HEK293T cells and the cells were incubated at 37 °C at 5% double fluorescence positive cells were then sorted by FACS ­CO2 for 6 h; then the medium was replaced with pre-warmed into single cells and were further allowed to grow into a fresh medium. After 24 h, 48 h, and 72 h, the supernatant single cell derived clonal cell line. Following, the genomic was collected and cleared by centrifugation at 4000×g for DNA was extracted from the single cell derived cell lines 5 min and passed through 0.45 µm filter. Then, the super- and further used as a template to amplify RNase2 using the natant fraction was concentrated by PEG6000 precipitate primers covering the potential mutation sites. After Sanger method [83] and the concentrated virus stock was aliquoted sequencing validation, cell lines, where RNase2 gene knock and stored at − 80 °C. out was successful were identified. Cell transduction T7 endonuclease I assay‑gene editing detection 5 × ­105 THP-1 monocytes were infected with 20 µL concen- As illustrated above, 200 ng of the purified RNase2 PCR trated lentivirus in the presence of 8 µg/mL polybrene over- products were denatured and re-annealed in 1 × T7EI Reac- night. Next day, the cells were replaced with fresh medium tion Buffer and then were incubated with or without T7E I and cultured for 72 h. Fluorescence positive monocytes were (Alt-R® Genome Editing Detection Kit, IDT). The reaction checked by fluorescence microscopy and then sorted by Cell mixtures were then separated by 2% agarose gel electropho- sorter. After the transduction, cells were resuspended and resis. The knockout efficiency (KO%) was determined using fixed by 2% paraformaldehyde in PBS for 10 min prior to the following formula: KO% = 100 × (1 − [1 − b]/[a + b])1/2, flow cytometer. Single cells were sorted by Cell sorter BD where a is the integrated intensity of the undigested PCR FACSJazz. product and b is the combined integrated intensity of the cleavage products [85]. Sanger sequencing Protein detection by western blot and ELISA Briefly, the genomic DNA of THP-1 cells was extracted using GenJET Genomic DNA purification kit (Ther- For the western blot assays, 5 × ­105 cells with or without moFisher, K0721) and was further used to amplify RNase2 transduction and their supernatants were harvested with using NZYTaq II 2 × Green master mix (NZYTech, MB358). RIPA buffer and the protein concentration was determined Genomic DNA was subjected to PCR (BioRad) using prim- with the Pierce BCA Protein Assay Kit (Thermo Fisher Sci- ers listed in Table S1. The general reaction conditions were entific, 23225). Equal amounts of protein (50 µg) for each 13 Selective cleavage of ncRNA and antiviral activity by RNase2/EDN in THP1‑induced macrophages Page 13 of 18 209 sample were loaded and separated by 15% SDS–PAGE, and infected with RSV stock under multiplicity of infection transferred to polyvinylidene difluoride membranes. Then (MOI) of 0.1. After 3 h infection, the cells were washed and the membrane was blocked with 5% non-fat milk in TBST replaced with fresh medium (DMEM + 10%FBS) and incu- for 1 h at room temperature, and incubated with rabbit bated for 4 days at 37ºC, 5% C ­ O2. The cells and the virus source anti-RNase2 primary antibody (Abcam, ab103428) suspension were collected when the cytopathology appeared, overnight at 4 °C. After washing, the membranes were with scraping and vortexing of the cells to release more viral treated with horseradish peroxidase (HRP)-conjugated goat particles. The virus suspension was centrifuged for 10 min anti-rabbit IgG (Sigma Aldrich, 12-348) for 1 h at room at 1800×g to remove the cell debris. The virus suspension temperature (RT). Finally, the membranes were exposed to without cell debris were either frozen immediately and an enhanced chemiluminescent detection system (SuperSig- stored at − 80 °C as seeding stock and concentrated before nal West Pico Chemiluminescent Substrate, ThermoFisher use with Ultra15 Amicon 100 kDa cutoff filters. The pro- Scientific, 32209). As a control, GAPDH was detected duced viruses were titrated using the median tissue culture with chicken anti-GAPDH antibodies (Abcam, ab9483) and infectious dose ­(TCID50) method in HEK293T cells [88]. goat anti-chicken secondary antibody (Abcam, ab6877). Secretory RNase2 in cell culture was detected by using human RNASE2 ELISA Kit (MyBioSource, MBS773233). RSV treatment of THP‑1‑induced macrophages Beforehand, the supernatant of the culture was concentrated 50 times using 15 kDa cutoff centrifugal filter unit (Ami- Before RSV inoculation, THP-1 cells were induced to con, C7715). Following, the standard and the concentrated macrophage by 50 nM of PMA treatment for 48 h as previ- culture supernatants were loaded to wells pre-coated with ously described [24]. Cells were washed three time with anti-RNase2 antibody, then the HRP-conjugated reagent pre-warmed PBS and replaced with fresh RPMI + 10%FBS was added. After incubation and washing for the removal medium for 24 h incubation. After that, macrophages were of unbound enzyme, the substrate was added to develop the washed and incubated with RSV, mixing at every 15 min for colourful reaction. The colour depth or light was positively the first 2 h. All virus treatment tests were performed using correlated with the concentration of RNase2. Triplicates RSV at a MOI of 1 ­TCID50/cell. were performed for all assays. Zymogram/ Ribonuclease activity staining assay Real‑time quantitative PCR Zymograms were performed as previously described [86]. RSV quantification were detected by real-time quantitative 15% polyacrylamide–SDS gels were casted with 0.3 mg/mL PCR. After the indicated time post of inoculation, the extra- of poly(U) (Sigma Aldrich, P9528-25MG). Then, cells were cellular RSV virus were collected by PEG6000 precipitation collected by centrifugation and resuspended in ­106 cells/ml method and intracellular RSV virus were collected by lys- with 1% SDS buffer. After sonication, cell lysate with indi- ing the macrophage cells with the lysis buffer from mirVa- cated number of cells was loaded using a loading buffer that naTM miRNA Isolation Kit (Ambion, Life Technologies, does not contain 2-mercaptoethanol. Electrophoresis was run AM1560). Total RNA from RSV treated macrophage cells as at a constant current of 100 V for 1.5 h. Following, the SDS well as stock virus was extracted using mirVanaTM miRNA was removed by washing with 10 mM Tris/HCl, pH 8, and Isolation Kit according to the manufacturer’s instructions. 10% (v/v) isopropanol for 30 min. The gel was then incu- cDNA was synthesized using iScriptTM cDNA Synthesis bated for 1 h in the activity buffer (100 mM Tris/HCl, pH 8) Kit (Bio-Rad, 170-8891). The synthesis was performed to allow ribonuclease digestion of the embedded substrate using random hexamers, starting with 1 µg of total cell RNA. and then stained with 0.2% (w/v) toluidine blue in 10 mM The RT-qPCR was performed using ddPCR ™ Supermix for Tris/HCl, pH 8, for 10 min. Positive bands appeared white Probes (Bio-Rad, 1863024). Samples with a cycle threshold against the blue background after distaining. value of more than 40 were recorded as negative. A stand- ard curve was prepared using serially diluted RNA extracts RSV production from a known quantity and used to quantify RSV as T ­ CID50/ mL. In parallel with the RSV probe assays, an endogenous Human respiratory syncytial virus (RSV, ATCC, VR-1540) glyceraldehyde-3-phosphate dehydrogenase (GAPDH) con- stock was ordered from ATCC. Hela cells were used to pro- trol was used for relative quantification of the intracellular duce RSV under biosafety level II conditions [87]. Briefly, virus. The relative expression of GAPDH and RNase2 gene Hela cells were plated in 75 ­cm2 culture flask and incu- in macrophages was quantified by real-time PCR using iTaq bated at 37 °C degree in DMEM + 10%FBS until they were Universal SYBR Green Supermix (Bio-Rad, 1725120). The approximately 50% confluent. The cells were then washed primers and probe [89] used are listed in Table S6. 13 209 Page 14 of 18 L. Lu et al. Cell viability assay tRF&tiRNA PCR Array THP-1 monocytes (wild type or RNase2 knockout) were Total RNA was extracted by using mirVana miRNA Isola- seeded at 5 × ­104 cells/well in 96 well plates and differen- tion Kit (Ambion, AM1556). Next, 2 µg of purified total tiated into macrophages as described [24]. After treatment RNA was used to create cDNA libraries from small RNAs with RSV under MOI = 1 for different times, dynamic cell for qPCR detection using rtStar First-Strand cDNA Synthe- viability was measured using MTT assay. sis Kit (ArrayStar, AS-FS-003). This method sequentially ligates 3′-adaptor with its 5′-end to the 3′-end of the RNAs, and 5′-adaptor with its 3′-end to the 5′-end of the RNAs. Selective amplification and sequencing of cyclic After cDNA synthesis, 185 (tRNA-derived fragments) tRFs phosphate‑containing RNA (cP‑RNA‑seq) & (tRNA halves) tiRNAs were profiled and quantified by qPCR method using nrStar Human tRF&tiRNA PCR Array Selective amplification and sequencing of cyclic (ArrayStar, AS-NR-002). phosphate-containing RNAs was performed as previ- ously reported [26]. Briefly, small RNAs (< 200nt) Protein expression and purification were extracted using mirVanaTM miRNA Isolation Kit (Ambion, Life Technologies, AM1560) as described The recombinant RNase2 and RNase5 proteins were by the manufacturer. Following RNA extraction, 20- to expressed and purified as previously described [59]. Briefly, -100nt RNAs were purified from 8% TBE–PAGE gel. the genes cloned into the pET11c expression vector (Nova- Then, the purified RNAs were treated by calf intestinal gen) were expressed in E. coli BL21(DE3) cells (Novagen). alkaline phosphatase. After phenol–chloroform purifica- Then the protein was extracted from solubilized inclu- tion, the RNAs were oxidized by incubation in 10 mM sion bodies, refolded and purified by FPLC chromatogra- ­NaIO4 at 0 °C for 40 min in the dark, followed by ethanol phy using a Resource S (GE Healthcare) cation exchange precipitation. The RNAs were then treated with T4 PNK. column. The lyophilized protein purity was checked by After phenol–chloroform purification, directed ligation of SDS–PAGE and MS spectrometry. adapters, cDNA generation, and PCR amplification were performed using the Truseq Small RNA Sample Prep Kit In vitro Transcription (IVT) and analysis of tRNA for Illumina (NewEngland Biolabs, E7335S) according to degradation products the manufacturer’s protocol. The amplified cDNAs were sequenced using Illumina hiSeq2500 system at the Centre The preparation of synthetic tRNAs was performed using for Genomic Regulation, CRG, Barcelona). in vitro transcription (IVT) assay based on the T7 RNA pol- For small RNA-seq analysis, skewer (v0.22) was used ymerase reaction [96]. According to the sequence of inter- to remove the 5’adaptor sequences and discard low-qual- est, we first designed and amplified the dsDNA templates ity reads [90]. Reads have been size selected before being for IVT considering the preference of T7 RNA polymerase aligned to the reference genome (GRCh38) with short- by replacing the first three bases by GGG in the targeted Stack based on bowtie1 aligner [91, 92]. The mapped sequences. The corresponding synthetic primers (Sigma, reads were counted with HTSeq-count [93] using the Aldrich) are listed in Table S7. After PCR amplification, annotation from miRBase version 22.1 ((http://​w ww.​ the product was washed and concentrated by Gelpure kit mirba​se.​org/) [94]. For differential analysis, DESeq2 [95] (NZYtech, MB01101). was used on count matrices of tRNA-derived fragments A total of 1 µg of dsDNA templates was used for a 50 µl and miRNAs. Quantification of differential abundance of IVT reaction by following the manufacturer’s instructions small RNAs was estimated by tRAX (http://​trna.​ucsc.​edu/​ (NZYtech, MB08001). In short, after 2 h of IVT reaction, tRAX/). The selective cleavage was identified by com- DNase treatment was performed at 37 °C for 1.5 h. Next, parison of the differential coverage in WT and KO sam- we added the same volume of acid–phenol: chloroform ples. Complementarily, small RNAs were mapped based (Ambion™, AM9722) to the mixture and centrifuged at on NCBI (https://​www.​ncbi.​nlm.​nih.​gov/) and Gencode maximum speed for 10 min to remove the protein. Next, the 38 (https://​w ww.​g enco​d egen​e s.​o rg/). The Integrative supernatant was taken and 1/10 volume of 3 M acetic acid Genomics Viewer (IGV) was used for visualizing the (pH 5.5) as well as 3 volumes of 100% ethanol was added aligned bam file reads and check the nucleotide compo- to precipitate the RNA at − 20 °C for 1.5 h. Then, the pre- sition at the putative cleavage site, which was estimated cipitated RNA was centrifuged at the maximum speed for from the analysis of the differential sequence coverage of 30 min, stored at − 4 °C and washed twice with 75% ethanol. WT and RNase2-KO samples. Finally, the RNA was dissolved in RNase-free water. 13 Selective cleavage of ncRNA and antiviral activity by RNase2/EDN in THP1‑induced macrophages Page 15 of 18 209 The tRNA degradation reaction was done in 10 mM Tris- expanding roles in promoting host defense. Int J Mol Sci HCl (pH 7.4) and 100 mM NaCl using a fixed concentration 16:15442–15455. https://​doi.​org/​10.​3390/​ijms1​60715​442 3. Acharya KR, Ackerman SJ (2014) Eosinophil granule proteins: of tRNA (2 µg) and different concentrations of RNase2 (up form and function. J Biol Chem 289:17406–17415. https://​doi.​ to 60 nM) or RNase5 (up to 25 µM) at room temperature org/​10.​1074/​jbc.​R113.​546218 during 20 min or 30 min, respectively. Following, the same 4. Lee JJ, Jacobsen EA, McGarry MP et al (2010) Eosinophils volume of 2 × stop buffer (2 × TAE, 8 M urea, 0.1 M DTT in health and disease: the LIAR hypothesis. Clin Exp Allergy 40:563–575. https://​doi.​org/​10.​1111/j.​1365-​2222.​2010.​03484.x and 20% glycerol) was added and samples were incubated 5. Weller PF, Spencer LA (2017) Functions of tissue-resident eosino- at 95℃ for 15 min before gel loading. The 12% denaturing phils. Nat Rev Immunol 17:746–760. https://​doi.​org/​10.​1038/​nri.​ urea polyacrylamide (NZYtech, MB15501) gel and Sybr 2017.​95 gold (Life technologies, S11494) was used to run and stain 6. Davoine F, Lacy P (2014) Eosinophil cytokines, chemokines, and growth factors: emerging roles in immunity. Front Immunol the gel, respectively. 5:570. https://​doi.​org/​10.​3389/​fimmu.​2014.​00570 7. Gleich GJ, Loegering DA, Bell MP et al (1986) Biochemical and Supplementary Information The online version contains supplemen- functional similarities between human eosinophil-derived neuro- tary material available at https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 00018-0​ 22-0​ 4229-x. toxin and eosinophil cationic protein: homology with ribonucle- ase. Proc Natl Acad Sci USA 83:3146–3150. https://​doi.​org/​10.​ Acknowledgements We thank Yundong Peng (Max Planck Institute for 1073/​pnas.​83.​10.​3146 Heart and Lung Research, Bad Nauheim, Germany), Manuel Kaulich 8. Rosenberg H (2008) Eosinophil-derived neurotoxin/RNase 2: con- (Goethe University Frankfurt, Germany) for technical supporting with necting the past, the present and the future. Curr Pharm Biotech- CRIPSR design. We would also like to show our gratitude to Dr. Laura nol 9:135–140. https://​doi.​org/​10.​2174/​13892​01087​84567​236 Tussel, Anna Genescà, Marina Rodriguez, and David Soler from UAB 9. Boix E, Nogués MV (2007) Mammalian antimicrobial proteins for helping in lentiviral production. We heartily thank Prof. Helene and peptides: overview on the RNase A superfamily members Rosenberg for helpful discussions. involved in innate host defence. Mol Biosyst 3:317. https://​doi.​ org/​10.​1039/​b6175​27a Funding Open Access Funding provided by Universitat Autonoma 10. Monteseirín J, Vega A, Chacón P et al (2007) Neutrophils as a de Barcelona. This research was founded by Research work was sup- novel source of eosinophil cationic protein in IgE-mediated pro- ported by the Ministerio de Economía y Competitividad (SAF2015- cesses. J Immunol 179:2634–2641. https://d​ oi.o​ rg/1​ 0.4​ 049/j​ immu​ 66007P), co-financed by FEDER funds, by Agencia Estatal de Investi- nol.​179.4.​2634 gación (PID2019-106123GB-I00/AEI/10.13039/501100011033) and 11. Lu L, Li J, Moussaoui M, Boix E (2018) Immune modulation by by Fundació La Marató de TV3 (2080310). LL and JL were supported human secreted RNases at the extracellular space. Front Immunol by China Scholarship Council (CSC) predoctoral fellowships. 9:1–20. https://​doi.​org/​10.​3389/​FIMMU.​2018.​01012 12. Domachowske JB, Dyer KD, Bonville CA, Rosenberg HF (1998) Recombinant human eosinophil-derived neurotoxin/RNase 2 func- Data availability All data generated or analysed during this study are tions as an effective antiviral agent against respiratory syncytial included in this published article (see supplementary information files). virus. J Infect Dis 177:1458–1464. https://d​ oi.o​ rg/1​ 0.1​ 086/5​ 15322 13. Rosenberg HF, Dyer KD, Domachowske JB (2009) Respiratory Declarations viruses and eosinophils: exploring the connections. Antiviral Res 83:1–9. https://​doi.​org/​10.​1016/j.​antiv​iral.​2009.​04.​005 Conflict of interest The authors declare no conflicts of interest. 14. Malik A, Batra JK (2012) Antimicrobial activity of human eosino- phil granule proteins: involvement in host defence against patho- gens. Crit Rev Microbiol 38:168–181. https://​doi.​org/​10.​3109/​ Open Access This article is licensed under a Creative Commons Attri- 10408​41X.​2011.​645519 bution 4.0 International License, which permits use, sharing, adapta- 15. Rodrigo-Muñoz JM, Sastre B, Cañas JA et al (2021) Eosinophil tion, distribution and reproduction in any medium or format, as long response against classical and emerging respiratory viruses: as you give appropriate credit to the original author(s) and the source, COVID-19. J Investig Allergol Clin Immunol 31:94–107. https://​ provide a link to the Creative Commons licence, and indicate if changes doi.​org/​10.​18176/​jiaci.​0624 were made. The images or other third party material in this article are 16. Rosenberg HF, Foster PS (2021) Eosinophils and COVID-19: included in the article's Creative Commons licence, unless indicated diagnosis, prognosis, and vaccination strategies. Semin Immuno- otherwise in a credit line to the material. If material is not included in pathol 43:383–392. https://​doi.​org/​10.​1007/​s00281-​021-​00850-3 the article's Creative Commons licence and your intended use is not 17. Tan Y, Zhou J, Zhou Q et al (2021) Role of eosinophils in the permitted by statutory regulation or exceeds the permitted use, you will diagnosis and prognostic evaluation of COVID-19. J Med Virol need to obtain permission directly from the copyright holder. To view a 93:1105–1110. https://​doi.​org/​10.​1002/​jmv.​26506 copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. 18. Harrison AM, Bonville CA, Rosenberg HF, Domachowske JB (1999) Respiratory syncytical virus-induced chemokine expres- sion in the lower airways: eosinophil recruitment and degranula- tion. Am J Respir Crit Care Med 159:1918–1924. https://​doi.​org/​ References 10.​1164/​ajrccm.​159.6.​98050​83 19. Rosenberg HF, Dyer KD (1997) Diversity among the primate eosinophil-derived neurotoxin genes: a specific C-terminal 1. Li J, Boix E (2021) Host defence RNases as antiviral agents sequence is necessary for enhanced ribonuclease activity. Nucleic against enveloped single stranded RNA viruses. Virulence Acids Res 25:3532–3536. https://​doi.​org/​10.​1093/​nar/​25.​17.​3532 12:444–469. https://​doi.​org/​10.​1080/​21505​594.​2021.​18718​23 20. Borchers AT, Chang C, Gershwin ME, Gershwin LJ (2013) Res- 2. Rosenberg HF (2015) Eosinophil-derived neurotoxin (EDN/ piratory syncytial virus-a comprehensive review. Clin Rev Allergy RNase 2) and the mouse eosinophil-associated RNases (mEars): Immunol 45:331–379. https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 12016-0​ 13-8​ 368-9 13 209 Page 16 of 18 L. Lu et al. 21. Kim C-K, Seo JK, Ban SH et al (2013) Eosinophil-derived neuro- through a novel targeting mechanism. Mol Ther 23:1622–1629. toxin levels at 3 months post-respiratory syncytial virus bronchi- https://​doi.​org/​10.​1038/​mt.​2015.​124 olitis are a predictive biomarker of recurrent wheezing. Biomark- 38. Skalsky RL, Cullen BR (2010) Viruses, microRNAs, and host ers 18:230–235. https://​doi.​org/​10.​3109/​13547​50X.​2013.​773078 interactions. Annu Rev Microbiol 64:123–141. https://​doi.​org/​10.​ 22. Rosenberg HF, Domachowske JB (1999) Eosinophils, ribonucle- 1146/​annur​ev.​micro.​112408.​134243 ases and host defense: solving the puzzle. Immunol Res 20:261– 39. Ghildiyal M, Zamore PD (2009) Small silencing RNAs: an 274. https://​doi.​org/​10.​1007/​BF027​90409 expanding universe. Nat Rev Genet 10:94–108. https://​doi.​org/​ 23. Sikriwal D, Seth D, Parveen S et al (2012) An insertion in loop L7 10.​1038/​nrg25​04 of human eosinophil-derived neurotoxin is crucial for itsantiviral 40. Wu W, Choi E-J, Lee I et al (2020) Non-coding RNAs and their activity. J Cell Biochem 113:3104–3112. https://​doi.​org/​10.​1002/​ role in respiratory syncytial virus (RSV) and human metapneu- jcb.​24187 movirus (hMPV) infections. Viruses 12:345. https://​doi.​org/​10.​ 24. Lu L, Arranz-Trullén J, Prats-Ejarque G et al (2019) Human anti- 3390/​v1203​0345 microbial RNases inhibit intracellular bacterial growth and induce 41. Cech TR, Steitz JA (2014) The noncoding RNA revolution-trash- autophagy in mycobacteria-infected macrophages. Front Immu- ing old rules to forge new ones. Cell 157:77–94. https://​doi.​org/​ nol. https://​doi.​org/​10.​3389/​fimmu.​2019.​01500 10.​1016/j.​cell.​2014.​03.​008 25. Lu L, Wei R, Prats-Ejarque G et al (2020) Human RNase3 immune 42. Selitsky SR, Baran-Gale J, Honda M et al (2015) Small tRNA- modulation by catalytic-dependent and independent modes in a derived RNAs are increased and more abundant than microRNAs macrophage-cell line infection model. Cell Mol Life Sci 1:3. in chronic hepatitis B and C. Sci Rep 5:7675. https://​doi.​org/​10.​ https://​doi.​org/​10.​1007/​s00018-​020-​03695-5 1038/​srep0​7675 26. Honda S, Morichika K, Kirino Y (2016) Selective amplification 43. Fu H, Feng J, Liu Q et al (2009) Stress induces tRNA cleavage by and sequencing of cyclic phosphate-containing RNAs by the cP- angiogenin in mammalian cells. FEBS Lett 583:437–442. https://​ RNA-seq method. Nat Protoc 11:476–489. https://​doi.​org/​10.​ doi.​org/​10.​1016/j.​febsl​et.​2008.​12.​043 1038/​nprot.​2016.​025 44. Yamasaki S, Ivanov P, Hu GF, Anderson P (2009) Angiogenin 27. Anderson P, Ivanov P (2014) tRNA fragments in human health cleaves tRNA and promotes stress-induced translational repres- and disease. FEBS Lett 588:4297–4304. https://d​ oi.o​ rg/1​ 0.1​ 016/j.​ sion. J Cell Biol 185:35–42. https://​doi.​org/​10.​1083/​jcb.​20081​ febsl​et.​2014.​09.​001 1106 28. Ostendorf T, Zillinger T, Andryka K et al (2020) Immune sensing 45. Thomas SP, Hoang TT, Ressler VT, Raines RT (2018) Human of synthetic, bacterial, and protozoan RNA by toll-like receptor angiogenin is a potent cytotoxin in the absence of ribonuclease 8 requires coordinated processing by RNase T2 and RNase 2. inhibitor. RNA 24:1018–1027. https://​doi.​org/​10.​1261/​RNA.​ Immunity 52:591-605.e6. https://​doi.​org/​10.​1016/j.​immuni.​2020.​ 065516.​117 03.​009 46. Torres AG, Reina O, Attolini CS-O, de Pouplana LR (2019) Dif- 29. Nikitina E, Larionova I, Choinzonov E, Kzhyshkowska J (2018) ferential expression of human tRNA genes drives the abundance Monocytes and macrophages as viral targets and reservoirs. Int J of tRNA-derived fragments. Proc Natl Acad Sci U S A 116:8451– Mol Sci 19:2821. https://​doi.​org/​10.​3390/​ijms1​90928​21 8456. https://​doi.​org/​10.​1073/​PNAS.​18211​20116 30. Miller AL, Bowlin TL, Lukacs NW (2004) Respiratory syncytial 47. Deng J, Ptashkin RN, Wang Q et al (2014) Human metapneumovi- virus-induced chemokine production: linking viral replication to rus infection induces significant changes in small noncoding RNA chemokine production in vitro and in vivo. J Infect Dis 189:1419– expression in airway epithelial cells. Mol Ther Nucleic Acids. 1430. https://​doi.​org/​10.​1086/​382958 https://​doi.​org/​10.​1038/​MTNA.​2014.​18 31. Ennaciri J, Ahmad R, Menezes J (2007) Interaction of mono- 48. Pawar K, Shigematsu M, Sharbati S, Kirino Y (2020) Infection- cytic cells with respiratory syncytial virus results in activation induced 5’-half molecules of tRNAHisGUG activate toll-like of NF-κB and PKC-α/β leading to up-regulation of IL-15 gene receptor 7. PLoS Biol 18:e3000982. https://d​ oi.o​ rg/1​ 0.1​ 371/j​ ourn​ expression. J Leukoc Biol 81:625–631. https://d​ oi.​org/​10.​1189/​ al.​pbio.​30009​82 jlb.​08065​07 49. Das AT, Klaver B, Berkhout B (1995) Reduced replication of 32. Rosenberg HF, Domachowske JB (2012) Inflammatory responses human immunodeficiency virus type 1 mutants that use reverse to Respiratory Syncytial Virus (RSV) infection and the develop- transcription primers other than the natural tRNA(3Lys). J Virol ment of immunomodulatory pharmacotherapeutics. Curr Med 69:3090–3097. https://​doi.​org/​10.​1128/​JVI.​69.5.​3090-​3097.​1995 Chem 19:1424. https://​doi.​org/​10.​2174/​09298​67127​99828​346 50. Jin D, Musier-Forsyth K (2019) Role of host tRNAs and ami- 33. Yang D, Chen Q, Rosenberg HF et al (2004) Human ribonuclease noacyl-tRNA synthetases in retroviral replication. J Biol Chem A superfamily members, eosinophil-derived neurotoxin and pan- 294:5352–5364. https://​doi.​org/​10.​1074/​jbc.​REV118.​002957 creatic ribonuclease, induce dendritic cell maturation and activa- 51. Schorn AJ, Gutbrod MJ, LeBlanc C, Martienssen R (2017) LTR- tion. J Immunol 173:6134–6142. https://​doi.​org/​10.​4049/​jimmu​ retrotransposon control by tRNA-derived small RNAs. Cell nol.​173.​10.​6134 170:61-71.e11. https://​doi.​org/​10.​1016/j.​cell.​2017.​06.​013 34. Rugeles MT, Trubey CM, Bedoya VI et al (2003) Ribonuclease 52. Greulich W, Wagner M, Gaidt MM et al (2019) TLR8 is a sen- is partly responsible for the HIV-1 inhibitory effect activated by sor of RNase T2 degradation products. Cell 179:1264-1275.e13. HLA alloantigen recognition. AIDS 17:481–486. https://​doi.​org/​ https://​doi.​org/​10.​1016/j.​cell.​2019.​11.​001 10.​1097/​00002​030-​20030​3070-​00002 53. Sorrentino S (2010) The eight human “canonical” ribonucleases: 35. Zhou J, Liu S, Chen Y et al (2017) Identification of two novel molecular diversity, catalytic properties, and specialbiologi- functional tRNA-derived fragments induced in response to res- cal actions of the enzyme proteins. FEBS Lett 584:2194–2200. piratory syncytial virus infection. J Gen Virol 98:1600–1610. https://​doi.​org/​10.​1016/j.​febsl​et.​2010.​04.​018 https://​doi.​org/​10.​1099/​jgv.0.​000852 54. Boix E, Blanco JA, Nogués MV, Moussaoui M (2013) Nucleo- 36. Wang Q, Lee I, Ren J et al (2013) Identification and functional tide binding architecture for secreted cytotoxic endoribonucleases. characterization of tRNA-derived RNA fragments (tRFs) in res- Biochimie 95:1087–1097. https://​doi.​org/​10.​1016/j.​biochi.​2012.​ piratory syncytial virus infection. Mol Ther 21:368–379. https://​ 12.​015 doi.​org/​10.​1038/​mt.​2012.​237 55. Su Z, Kuscu C, Malik A et al (2019) Angiogenin generates specific 37. Deng J, Ptashkin RN, Chen Y et al (2015) Respiratory syncytial stress-induced tRNA halves and is not involved in tRF-3-mediated virus utilizes a tRNA fragment to suppress antiviral responses 13 Selective cleavage of ncRNA and antiviral activity by RNase2/EDN in THP1‑induced macrophages Page 17 of 18 209 gene silencing. J Biol Chem 294:16930–16941. https://d​ oi.o​ rg/1​ 0.​ 72. Rashad S, Tominaga T, Niizuma K (2021) The cell and stress- 1074/​jbc.​RA119.​009272 specific canonical and noncanonical tRNA cleavage. J Cell 56. Suhasini AN, Sirdeshmukh R (2006) Transfer RNA cleavages by Physiol 236:3710–3724. https://​doi.​org/​10.​1002/​JCP.​30107 onconase reveal unusual cleavage sites. J Biol Chem 281:12201– 73. Shigematsu M, Kirino Y (2020) Oxidative stress enhances the 12209. https://​doi.​org/​10.​1074/​jbc.​M5044​88200 expression of 2’,3’-cyclic phosphate-containing RNAs. RNA 57. Ardelt W, Ardelt B, Darzynkiewicz Z (2009) Ribonucleases Biol 17:1060–1069. https://​doi.​org/​10.​1080/​15476​286.​2020.​ as potential modalities in anticancer therapy. Eur J Pharmacol 17668​61 625:181–189. https://​doi.​org/​10.​1016/j.​ejphar.​2009.​06.​067 74. Li S, Hu G-F (2012) Emerging role of angiogenin in stress 58. Akiyama Y, Tomioka Y, Abe T et al (2021) In lysate RNA diges- response and cell survival under adverse conditions. J Cell tion provides insights into the angiogenin’s specificity towards Physiol 227:2822–2826. https://​doi.​org/​10.​1002/​jcp.​23051 transfer RNAs. RNA Biol. https://​doi.​org/​10.​1080/​15476​286.​ 75. Xin X, Wang H, Han L et al (2018) Single-cell analysis of the 2021.​19307​58 impact of host cell heterogeneity on infection with foot-and- 59. Prats-Ejarque G, Lu L, Salazar VA et al (2019) Evolutionary mouth disease virus. J Virol 92:e00179-e218. https://​doi.​org/​ trends in RNA base selectivity within the RNase A superfamily. 10.​1128/​JVI.​00179-​18 Front Pharmacol 10:1170. https://​doi.​org/​10.​3389/​fphar.​2019.​ 76. Zhang X, Cozen AE, Liu Y et al (2016) Small RNA modi- 01170 fications: integral to function and disease. Trends Mol Med 60. Leonidas DD, Shapiro R, Irons LI et al (1999) Toward rational 22:1025–1034. https://​doi.​org/​10.​1016/J.​MOLMED.​2016.​10.​ design of ribonuclease inhibitors: high-resolution crystal struc- 009 ture of a ribonuclease A complex with a potent 3’,5’-pyroph- 77. Akiyama Y, Kharel P, Abe T et al (2020) Isolation and initial osphate-linked dinucleotide inhibitor. Biochemistry 38:10287– structure-functional characterization of endogenous tRNA- 10297. https://​doi.​org/​10.​1021/​BI990​900W derived stress-induced RNAs. RNA Biol. https://​d oi.​o rg/​1 0.​ 61. Holloway DE, Chavali GB, Leonidas DD et al (2009) Influence 1080/​15476​286.​2020.​17327​02 of naturally-occurring 5’-pyrophosphate-linked substituents on 78. Jain S, Wheeler JR, Walters RW et al (2016) ATPase-modulated the binding of adenylic inhibitors to ribonuclease a: an X-ray stress granules contain a diverse proteome and substructure. crystallographic study. Biopolymers 91:995–1008. https://​doi.​ Cell 164:487–498. https://​doi.​org/​10.​1016/J.​CELL.​2015.​12.​038 org/​10.​1002/​bip.​21158 79. Shigematsu M, Kawamura T, Kirino Y (2018) Generation of 62. Shepard SM, Windsor IW, Raines RT, Cummins CC (2019) 2’,3’-cyclic phosphate-containing RNAs as a hidden layer of Nucleoside tetra- and pentaphosphates prepared using a tetrap- the transcriptome. Front Genet 9:562. https://​doi.​org/​10.​3389/​ hosphorylation reagent are potent inhibitors of ribonuclease A. fgene.​2018.​00562 J Am Chem Soc 141:18400–18404. https://​doi.​org/​10.​1021/​jacs.​ 80. Shigematsu M, Morichika K, Kawamura T et al (2019) Genome- 9b097​60 wide identification of short 2’,3’-cyclic phosphate-containing 63. Lee JE, Bae E, Bingman CA et al (2008) Structural basis for RNAs and their regulation in aging. PLoS Genet 15:e1008469. catalysis by onconase. J Mol Biol 375:165–177. https://​doi.​org/​ https://​doi.​org/​10.​1371/​journ​al.​pgen.​10084​69 10.​1016/j.​jmb.​2007.​09.​089 81. Schimmel P (2018) The emerging complexity of the tRNA 64. Yu X, Xie Y, Zhang S et al (2021) tRNA-derived fragments: world: mammalian tRNAs beyond protein synthesis. Nat Rev mechanisms underlying their regulation of gene expression and Mol Cell Biol 19:45–58. https://​doi.​org/​10.​1038/​nrm.​2017.​77 potential applications as therapeutic targets in cancers and virus 82. Al Yacoub N, Romanowska M, Haritonova N, Foerster J (2007) infections. Theranostics 11:461–469. https://​doi.​org/​10.​7150/​ Optimized production and concentration of lentiviral vectors THNO.​51963 containing large inserts. J Gene Med. https://​doi.​org/​10.​1002/​ 65. Yue T, Zhan X, Zhang D et al (2021) SLFN2 protection of jgm.​1052 tRNAs from stress-induced cleavage is essential for T cell– 83. Kutner RH, Zhang X-Y, Reiser J (2009) Production, concentra- mediated immunity. Science. https://​doi.​org/​10.​1126/​scien​ce.​ tion and titration of pseudotyped HIV-1-based lentiviral vectors. aba42​20 Nat Protoc 4:495–505. https://​doi.​org/​10.​1038/​nprot.​2009.​22 66. Saxena SK, Sirdeshmukh R, Ardelt W et al (2002) Entry into 84. Tiscornia G, Singer O, Verma IM (2006) Production and puri- cells and selective degradation of tRNAs by a cytotoxic member fication of lentiviral vectors. Nat Protoc 1:241–245. https://​doi.​ of the RNase A family. J Biol Chem 277:15142–15146. https://​ org/​10.​1038/​nprot.​2006.​37 doi.​org/​10.​1074/​jbc.​M1081​15200 85. Shen B, Zhang J, Wu H et al (2013) Generation of gene-mod- 67. Sarangdhar MA, Allam R (2021) Angiogenin (ANG)—ribonu- ified mice via Cas9/RNA-mediated gene targeting. Cell Res clease inhibitor (RNH1) system in protein synthesis and disease. 23:720–723. https://​doi.​org/​10.​1038/​cr.​2013.​46 Int J Mol Sci 22:1287. https://​doi.​org/​10.​3390/​IJMS2​20312​87 86. Prats-Ejarque G, Arranz-Trullén J, Blanco JA et al (2016) 68. Ivanov P, Emara MM, Villen J et al (2011) Angiogenin-induced The first crystal structure of human RNase 6 reveals a novel tRNA fragments inhibit translation initiation. Mol Cell 43:613– substrate-binding and cleavage site arrangement. Biochem J 623. https://​doi.​org/​10.​1016/j.​molcel.​2011.​06.​022 473:1523–1536. https://​doi.​org/​10.​1042/​BCJ20​160245 69. Choi E-J, Ren J, Zhang K et al (2020) The importance of AGO 87. Vissers M, Habets MN, Ahout IML et al (2013) An in vitro 1 and 4 in post-transcriptional gene regulatory function of tRF5- model to study immune responses of human peripheral blood GluCTC, an respiratory syncytial virus-induced tRNA-derived mononuclear cells to human respiratory syncytial virus infec- RNA fragment. Int J Mol Sci 21:8766. https://​doi.​org/​10.​3390/​ tion. J Vis Exp. https://​doi.​org/​10.​3791/​50766 IJMS2​12287​66 88. Sun Y, Jain D, Koziol-White CJ et al (2015) Immunostimula- 70. Kellner S, Burhenne J, Helm M (2010) Detection of RNA modi- tory defective viral genomes from respiratory syncytial virus fications. RNA Biol 7:237–247. https://​doi.​org/​10.​4161/​r na.7.​ promote a strong innate antiviral response during infection in 2.​11468 mice and humans. PLOS Pathog 11:e1005122. https://​doi.​org/​ 71. Pizzo E, Sarcinelli C, Sheng J et al (2013) Ribonuclease/angio- 10.​1371/​journ​al.​ppat.​10051​22 genin inhibitor 1 regulates stress-induced subcellular localiza- 89. Dewhurst-Maridor G, Simonet V, Bornand J et al (2004) Devel- tion of angiogenin to control growth and survival. J Cell Sci opment of a quantitative TaqMan RT-PCR for respiratory syn- 126:4308–4319. https://​doi.​org/​10.​1242/​jcs.​134551 cytial virus. J Virol Methods 120:41–49. https://​d oi.​o rg/​1 0.​ 1016/J.​JVIRO​MET.​2004.​03.​017 13 209 Page 18 of 18 L. Lu et al. 90. Jiang H, Lei R, Ding S-W, Zhu S (2014) Skewer: a fast and DESeq2. Genome Biol 15:550. https:// ​ d oi. ​ o rg/ ​ 1 0. ​ 1 186/​ accurate adapter trimmer for next-generation sequencing paired- s13059-​014-​0550-8 end reads. BMC Bioinformatics 15:182. https://​doi.​org/​10.​1186/​ 96. Lukavsky PJ, Puglisi JD (2004) Large-scale preparation and 1471-​2105-​15-​182 purification of polyacrylamide-free RNA oligonucleotides. RNA 91. Langmead B, Trapnell C, Pop M, Salzberg SL (2009) Ultrafast 10:889–893. https://​doi.​org/​10.​1261/​rna.​52648​04 and memory-efficient alignment of short DNA sequences to the 97. Sikriwal D, Seth D, Dey P, Batra JK (2007) Human eosinophil- human genome. Genome Biol 10:R25. https://​doi.​org/​10.​1186/​ derived neurotoxin: involvement of a putative non-catalytic gb-​2009-​10-3-​r25 phosphate-binding subsite in its catalysis. Mol Cell Biochem 92. Axtell MJ (2013) ShortStack: comprehensive annotation and 303:175–181. https://​doi.​org/​10.​1007/​s11010-​007-​9471-0 quantification of small RNA genes. RNA 19:740–751. https://​ 98. Sorrentino S (1998) Human extracellular ribonucleases: multi- doi.​org/​10.​1261/​r na.​035279.​112 plicity, molecular diversity and catalytic properties of the major 93. Anders S, Pyl PT, Huber W (2015) HTSeq—a Python frame- RNase types. Cell Mol Life Sci 54:785–794. https://​doi.​org/​10.​ work to work with high-throughput sequencing data. Bioinfor- 1007/​s0001​80050​207 matics 31:166–169. https://​d oi.​o rg/​1 0.​1 093/​B IOIN​F ORMA​ TICS/​BTU638 Publisher's Note Springer Nature remains neutral with regard to 94. Griffiths-Jones S (2006) miRBase: microRNA sequences, tar- jurisdictional claims in published maps and institutional affiliations. gets and gene nomenclature. Nucleic Acids Res 34:D140–D144. https://​doi.​org/​10.​1093/​nar/​gkj112 95. Love MI, Huber W, Anders S (2014) Moderated estima- tion of fold change and dispersion for RNA-seq data with Authors and Affiliations Lu Lu1,2 · Jiarui Li1 · Ranlei Wei3 · Irene Guidi1 · Luca Cozzuto4 · Julia Ponomarenko4 · Guillem Prats‑Ejarque1 · Ester Boix1 2 * Lu Lu College of Animal Science and Technology, Sichuan lu.lu@sicau.edu.cn Agricultural University, Chengdu, Sichuan, China 3 * Ester Boix National Frontier Center of Disease Molecular Network, Ester.Boix@uab.es West China Hospital, Sichuan University, Chengdu 610041, China 1 Department of Biochemistry and Molecular Biology, 4 Bioinformatic Unit, Centre de Regulació Genòmica (CRG), Faculty of Biosciences, Universitat Autònoma de Barcelona, Barcelona, Spain Cerdanyola del Vallès, Spain 13