key: cord-340635-8wki7noy
authors: Yu, Bin; Ikhlas, Shoeb; Ruan, Chunsheng; Zhong, Xingxing; Cai, Dongsheng
title: Innate and adaptive immunity of murine neural stem cell-derived piRNA exosomes/microvesicles against pseudotyped SARS-CoV-2 and HIV-based lentivirus
date: 2020-11-13
journal: iScience
DOI: 10.1016/j.isci.2020.101806
sha: 
doc_id: 340635
cord_uid: 8wki7noy

Through testing pseudotyped SARS-CoV-2 and HIV-based lentivirus, this study reports that exosomes/microvesicles (Ex/Mv) isolated from murine hypothalamic neural stem/progenitor cells (htNSC) or subtype htNSCPGHM as well as hippocampal NSC have innate immunity-like actions against these RNA viruses. These extracellular vesicles also have a cell-free innate antiviral action through attacking and degrading viruses. We further generated the induced versions of Ex/Mv through prior viral exposure to NSCs and found that these induced Ex/Mv were stronger than basal Ex/Mv in reducing the infection of these viruses, suggesting the involvement of an adaptive immunity-like antiviral function. These NSC Ex/Mv were found to be characterized by producing large libraries of piRNAs against genomes of various viruses, and some of these piRNAs were enriched during the adaptive immunity-like reaction, possibly contributing to the antiviral effects of these Ex/Mv. In conclusion, NSC Ex/Mv have antiviral immunity and could potentially be developed to combat against various viruses.

The brain is the "headquarters" of the body and normally should be strictly protected from infections especially by viruses which are small and could penetrate into the brain parenchyma with less difficulty than large pathogens such as bacteria. This antiviral protection theoretically requires brain immunity, but it is much less clear how the brain fights against viral infection compared to the immune operation in the periphery. Recently, as research has appreciated the neural control of immunity (Pavlov et al., 2018; Rosas-Ballina et al., 2011; Wang et al., 2020) , some evidence was obtained to help understand brain immunity, for instance, the brain has the lymphatic system (Da Mesquita et al., 2018; Louveau et al., 2015; Moseman et al., 2020; Papadopoulos et al., 2020) and olfactory stem cells have an immune defense effect under chronic inflammation . Related to this research field, we discovered that the hypothalamus has a role in linking innate immunity to adaptive immunity (Kim et al., 2015) and hypothalamic neural stem/progenitor cells (htNSC) abundantly secrete exosomal miRNAs into the cerebrospinal fluid (CSF) while some of these miRNAs target immunity components (Zhang et al., 2017) . Here in this study, we report that murine NSCs produce Ex/Mv to provide innate and adaptive antiviral actions against viral infection, using recombinant viruses as examples.

Further, we found that these NSC Ex/Mv are characterized by producing large libraries of piRNAs against the genomes of various viruses, and some of these piRNAs can be further enriched in the induced versions of NSC Ex/Mv, likely contributing to the antiviral effects of these vesicular particles. All these findings suggest a value of developing NSC-derived piRNAscontaining Ex/Mv for combating against various viruses possibly including SARS-CoV-2.

J o u r n a l P r e -p r o o f Glycoprotein-deficient vesicular stomatitis virus (ΔG-VSV) is a standard tool to create pseudotyped viruses (Whitt, 2010) , and it has frequently been used to generate pseudotyped viruses and recently pseudotyped SARS-CoV-2 virus (Nie et al., 2020a) . Because ΔG-VSV was designed to accurately report viral infection, we incorporated wildtype SARS-CoV-2 spike protein into luciferase-expressing ΔG-VSV to generate a pseudotyped SARS-CoV-2 virus. The infection of host cells by this virus was quantitatively reflected through measuring the activity of luciferase which was encoded by this viral genome. We employed this pseudotyped SARS-CoV-2 to study if murine NSC-released extracellular vesicles might have an action to fight against this virus. As we previously revealed (Zhang et al., 2017) , murine NSC abundantly produce and release exosomes (Ex) which are the extracellular vesicles of small size (usually below 200 nm in diameter); this study was based on these Ex but also included microvesicles (Mv) which are the extracellular vesicles of large size (usually above 200 nm in diameter). We designed to comparatively study Ex/Mv from NSCs including htNSC, hpNSC as well as htNSC PGHM , a htNSC subtype which we recently established, because Ex/Mv from this htNSC subtype can importantly help animals survive from fatal conditions besides the benefit that this NSC subtype can be applied peripherally (Tang et al., 2020) . To do so, we isolated and purified Ex/Mv that were released from these NSCs in culture and then tested if treatment with these Ex/Mv could provide a therapeutic effect against the infection of these pseudotyped SARS-CoV-2 viruses in vitro. To mimic its infection in various organs, we employed two human respiratory cell models, human alveolar basal epithelial cell A549 and human bronchial epithelial cell Calu3, and human hepatocyte cell model HepG2, all of which express hACE2 and have been used to study SARS-CoV-2 recently (Hoffmann et al., 2020; Ma et al., 2020; Nie et al., 2020b) . We infected these cells with pseudotyped SARS-CoV-2 and in the meanwhile treated these cells with Ex/Mv J o u r n a l P r e -p r o o f derived from an NSC type vs. vehicle control. The results showed that treatment of NSC Ex/Mv regardless of NSC types provided a significant antiviral effect in A549 cells and HepG2 (Fig. 1A,   B ). However, infection in Calu3 cells was not affected by this treatment (Fig. 1C) , suggesting that some cells are less sensitive than others in responding to the treatment. We further verified these antiviral effects in A549 and HepG2 cells through immunostaining of luciferase protein. As represented in Fig. 1D and detailedly quantified in Fig. 1E , while luciferase protein was strongly present in cells with vehicle control group, it was only weakly present in cells that were treated with either type of NSC Ex/Mv. Hence, NSC Ex/Mv have innate antiviral actions for some human cells of different tissue origins.

We noted that compared to A549 and HepG2, Calu3 cells were not responsive to the treatment of NSC Ex/Mv (Fig. 1C) , and thus wondered if these NSC Ex/Mv could be optimized for an antiviral effect for targeting insensitive cells such as Calu3. We asked if an initial exposure of a specific virus to NSC could lead to an enhancement or enrichment of certain immunity features of NSC-derived Ex/Mv which could boost the antiviral actions. Because infection of pseudotyped SARS-CoV-2 requires human angiotensin-converting enzyme 2 (hACE2) as the receptor, we generated hACE2-NSC cell lines with overexpression of hACE2 through the process of lentiviral induction and selection ( Fig. 2A ). After these cell lines were stably maintained for several generations in culture, we treated these cells with the pseudotyped SARS-CoV-2 for 2 generations, and then maintained them under normal culture for about 5 generations. After that, we isolated and purified Ex/Mv from these virally pre-exposed NSCs vs. matched basal NSCs, leading to the generation of induced vs. basal versions of NSC Ex/Mv, J o u r n a l P r e -p r o o f respectively. Since infection of Calu3 cells was insensitive to basal NSC Ex/Mv treatment, this cell line represented a unique experimental model for comparing the effects of induced vs. basal Ex/Mv. As shown in Fig. 2B , while basal NSC Ex/Mv were ineffective, we excitingly found that the induced NSC Ex/Mv, regardless of NSC types, became consistently effective in reducing the infection of pseudotyped SARS-CoV-2 in Calu3 cells. Thus, NSC Ex/Mv can be induced via an adaptive immunity-like process to provide improved effects against viral infection.

Ex/Mv and RNA viruses are similarly small-sized membranous particles, and both consist of RNA sequences. We were provoked to question if NSC Ex/Mv might have an innate immunity-like action to directly interact with viruses in a cell-free environment. To test this question, we incubated NSC Ex/Mv with pseudotyped SARS-CoV-2, and then examined how this incubation could affect the virus. To test this idea directly, we employed transmission electron microscope (TEM) to observe if there could be any physical interaction between NSC Ex/Mv and viruses. In this experiment, we designed to focus on exosomes (below 200 nm in diameter) from htNSC Ex/Mv which were isolated through a filter. We then mixed these htNSCderived exosomes with pseudotyped SARS-CoV-2 in a buffer for 0.5 hour at room temperature and then overnight at 4 °C before fixed and processed with negative staining for TEM analysis.

Exosomes or viruses alone in the same reaction condition were included to provide technical controls. Clearly, exosomes and viruses had distinctly different morphologies, as viruses were rod-shaped while exosomes were round and cup-shaped. We observed that while typically viruses and exosomes were both spreadingly and randomly distributed in the TEM fields, mixture with NSC exosomes led to clustered aggregations and ring-like structures (Fig. 3A ).

Through high-magnification TEM imaging, we identified morphological evidence suggesting that exosomes physically attached, surrounded or engulfed viruses, and some of these behaviors were apparently associated with breakdown and degradation of viruses (Fig. 3B ). Independently, we performed experiments in which pseudotyped SARS-CoV-2 viruses were mixed with NSC Ex/Mv versus vehicle in a buffer and then subjected to Western blotting for SARS2-CoV-2 spike protein. The results showed that the mixture with htNSC or hpNSC Ex/Mv both led to degradation of this spike glycoprotein (Fig. 4A ). Furthermore, we performed an experiment in which SARS-CoV-2 and NSC Ex/Mv were maintained separately or were mixed together for an overnight period, and then were used to infect HepG2 cells. As shown in Fig. 4B , pre-mixture of these viruses with either type of NSC Ex/Mv led to an evident reduction in the abilities of these viruses to infect the host cells. Altogether, these results suggest that NSC Ex/Mv have a cellindependent innate antiviral effect in cell-free environment.

We have previously revealed that murine NSCs abundantly produce exosomal miRNAs (Zhang et al., 2017) . Because miRNAs are known to target mainly mRNAs, we directed our attention to P-element induced wimpy testis (PIWI)-interacting RNAs (piRNAs), a different type of small RNAs which can target many other types of RNA/DNA and there are huge numbers of piRNA species in various animals including rodents (Wang et al., 2019) , although research often focuses on germ cells for piRNA biology since the discovery. The activities of piRNAs are mediated by PIWI proteins, mainly PIWI1 and PIWI2 (PIWIL1 and PIWIL2 in mammals, respectively). We decided to employ PIWIL1/2 as the biomarkers to analyze if the piRNA machinery could be contained in NSC Ex/Mv which we have found antiviral. First, we examined J o u r n a l P r e -p r o o f if PIWIL1/2 proteins could be detectable in NSCs at the cellular level, and for comparison, we included MSC in this experimental assay. While PIWIL1 blot did not yield clear signals in Western blots, we found that PIWIL2 was evidently expressed in htNSC and htNSC PGHM as well as hpNSC, but comparatively its expression level was much lower in MSC (Fig. 5A ). In parallel, we employed PIWIL2 immunostaining to compare the cellular distribution of this protein in NSC and MSC. We found that PIWIL2 was present mainly in the nuclei of MSC, in contrast, PIWIL2 was present strongly in the cytoplasm of all NSC types in addition to the nuclear distribution of this protein (Fig. 5B ). The cytoplasmic distribution of PIWI protein is theoretically consistent with the source of this protein for piRNA machinery in NSC Ex/Mv. In this context, we directly examined PIWIL2 in antiviral NSC Ex/Mv focusing on exosomes through immunostaining and imaging under high magnifications. As represented in Fig. 5C , we consistently confirmed that these NSC-derived exosomes contained PIWIL2. Under co-staining condition, the immunostaining signals of PIWIL2 were comparable to the levels as revealed by exosomal markers TSG101 and CD81, suggesting that PIWIL2 in these vesicles was significant.

Hence, NSC Ex/Mv were characterized by containing piRNA machinery while being antiviral.

Thus, we searched for mouse piRNAs which could match against the genomic RNA sequence of SARS-CoV-2 virus including the encoding sequences for spike protein, envelope protein, membrane protein, nucleocapsid protein, open reading frame (Orf) sequences Orf1ab, 3a, 6, 7a, 7b, 8 and 10, untranslated region (UTR) sequences at 5' end and 3' end, and 3 gap structural sequences (gap 1-3), as elucidated in Fig. S1A . Although SARS-CoV-2 is singlestrand RNA virus, we analyzed both sense and antisense sequences since production and J o u r n a l P r e -p r o o f replication of RNA viruses in host cells involve the synthesis of both strands. A piRNA has the primary lead sequence made of nucleotides from position 2 to 11 and the secondary lead sequence made of nucleotides from position 12 to 21; while the targeting rules of piRNAs in mammals are still not clearly established, it has recently been described in C. elegans that matching of primary lead and particularly nucleotides 2-8 with target RNA is important, the secondary lead sequence could have a few mismatches with target RNA (Shen et al., 2018; Zhang et al., 2018) . To begin with, we screened for piRNAs which have at least 16 nucleotides matching with SARS-CoV-2 RNA and found over thousand unique piRNA species which met this requirement; diagram in Fig. S1A presented a portion of these piRNAs with sequence information detailed in Table S1 . For piRNAs in this Table, we further required perfect match of the primary lead nucleotides 2-11 and no more than 4 mismatches for the secondary lead nucleotides 12-21 (Criteria #1) and found that about 170 piRNA species met these criteria ( Fig.   S1A , Table S1 ). Given that SARS-CoV-2 viruses have many variants and mutations, we slightly adjusted the criteria by allowing 5 mismatches in the secondary lead sequence while the primary lead sequence was still required for perfect matches (Criteria #2). This led to identification of ~150 additional piRNA species that could target SARS-CoV-2 ( Fig. S2A , Table S1 ). Hence, mouse species has piRNAs against sense and antisense sequence of SARS-CoV-2 genome. It should be mentioned that piRNAQuest database is not a complete collection of piRNAs. We also did the same way for human piRNAs through this database but noted that there are much fewer human piRNA species meeting these criteria of targeting SARS-CoV-2. If this difference is verified through analysis of complete database, it might point to the importance of piRNAs for the different resistance to viral infection between rodents and humans.

We then quantitatively measured some of these piRNAs in NSC Ex/Mv. To do so, we studied Ex/Mv from NSC compared to Ex/Mv from MSC. We randomly examined about 60 piRNA species that could target different regions of this genome based on Criteria 1 or 2. The results showed that most of these piRNAs were clearly present in NSC Ex/Mv (using htNSC PGHM and hpNSC as the representative); in contrast, these piRNAs were much less detectable in MSC Ex/Mv. Relatively, the levels of some piRNAs were 100 to 30,000-fold higher in NSC Ex/Mv than in MSC Ex/Mv, based on the same quantity of Ex/Mv total small RNA (Fig. S1B ). In addition, we confirmed that these piRNAs were similar or slightly higher in htNSC PGHM Ex/Mv when compared to the levels in htNSC Ex/Mv; thus, as a subtype of htNSC, htNSC PGHM capture the feature of NSCs in producing Ex/Mv piRNAs. On the other hand, we noted several cases that the cellular levels of piRNAs were relatively comparable between NSC and MSC, which can, however, further point to the special feature of NSC for actively assembling piRNAs into the secretory Ex/Mv of these neural cells.

The genome of the pseudotyped SARS-CoV-2 in this study was based on VSV elements, including the RNA sequences encoding nucleocapsid protein (N sequence), phosphoprotein (P sequence), matrix protein (M sequence), RNA polymerase (R sequence) and structural nonencoding sequences (S1-6), as elucidated in Fig. S2A . We applied each of these sense and antisense sequences to screen for piRNA species through mouse piRNAQuest database and found a long list of piRNAs with at least 15-16 nucleotides matching with the viral sequence, and among them, about 40 piRNA species were identified to meet Criteria 1 or 2 (Fig. S2A , Table S2 ). We then measured some of these piRNAs in NSC Ex/Mv (using htNSC PGHM and J o u r n a l P r e -p r o o f hpNSC as the representative) compared to the levels in MSC Ex/Mv. As shown in Fig. S2B and S3, most of these piRNAs were present in these NSC Ex/Mv but were much less detectable in MSC Ex/Mv. Our additional assays showed that htNSC PGHM were comparable or slightly stronger than htNSC in producing these piRNAs. Thus, based on the information from wildtype and pseudotyped SARS-CoV-2, NSC Ex/Mv contain piRNAs against the genomic sequences of both viruses, although these NSC were not previously exposed to either virus, suggesting that mouse species has evolved to establish large antiviral piRNA libraries in NSC Ex/Mv.

We asked if an initial pre-exposure of a specific virus to NSC could lead to an enhancement or enrichment of specific antiviral piRNAs in NSC Ex/Mv. Thus, we treated these NSCs with pseudotyped SARS-CoV-2 virus for 2 generations, and then maintained them under normal culture for about 5 generations. After that, we isolated and purified Ex/Mv from these virally pre-exposed NSCs vs. matched basal NSCs and then measured their piRNA expression levels via qPCR. Among about 80 piRNAs which we examined, roughly half of them were significantly upregulated due to the viral pre-exposure, typically 5 to 10 folds higher than the basal levels (Fig. S4 ). Of note, the increased expression levels in these specific piRNAs were often due to relative downregulation in non-target small RNAs (such as U6 miRNA as an example) in these induced Ex/Mv compared to basal Ex/Mv, suggesting that at least a negative selection process occurred. We also evaluated this effect based on luciferase-encoding sequence in this viral genome, because we found a short list of piRNA species which even matched against the sense or antisense sequence of luciferase RNA (Table S2) , although luciferase is a non-viral protein. We measured 7 of these piRNA species and found that 5 of them were present in NSC Ex/Mv but much less detectable in MSC Ex/Mv (Fig. S5A) , indicating that murine NSC Ex/Mv have piRNA libraries against various types of foreign genomic sequences which are not limited J o u r n a l P r e -p r o o f to viral genomes. We examined if these piRNAs could be upregulated by viral pre-exposure and found that 4 of them showed significant upregulation (Fig. S5B, C) . Thus, viral genome-specific piRNAs in NSC Ex/Mv can be enriched and enhanced in response to viral pre-exposure to possibly contribute to the better antiviral effect of these vesicles against this particular virus.

To further assess the predicted antiviral role of murine NSC Ex/Mv, we extended the study to examine an RNA virus unrelated to SARS-CoV-2, i.e., HIV-based lentivirus. This viral model was based on a VSV glycoprotein (VSVG)-enveloped recombinant lentivirus, and it also 

Given that NSC Ex/Mv showed a cell-free innate antiviral effect against pseudotyped SARS-CoV-2 ( Fig. 3 and 4) , we further examined if this finding could apply to HIV-based lentivirus. To do so, we incubated the purified lentiviruses with a type of purified NSC Ex/Mv in a buffer for overnight time and then fixed them for electron microscope analysis. We obtained the morphological evidence suggesting that these NSC Ex/Mv physically interacted with lentiviruses leading to changes indicative of viral breakdown. These suggestive results are not presented, given that it was not always clear-cut to distinguish them since lentiviruses and exosomes were both round although exosomes were often cup-shaped. Alternatively, we designed a biochemical assay to quantitatively assess this relationship in cell-free incubation condition. To do so, we incubated lentiviruses with NSC Ex/Mv versus vehicle in a buffer and then examined how this incubation could affect the viruses through Western blotting for VSVG which was the envelop protein of these lentiviruses. Also, to provide a time course profile, we designed several time points including 0.5, 4 and 24 hours for Ex/Mv and lentivirus reaction in the buffer, compared to the virus or Ex/Mv alone in the same reaction condition. As shown in Fig. 7A & B, the mixture with basal NSC Ex/Mv whether from the hypothalamus or hippocampus both rapidly led to lentiviral degradation. This effect occurred as quickly as 0.5 hour at room temperature, but interestingly longer time of incubation at 4 °C did not further increase viral degradation. We similarly assessed the induced versions of NSC Ex/Mv and observed that both were comparable to basal NSC Ex/Mv in leading to degradation of VSVG-J o u r n a l P r e -p r o o f lentiviruses. This similarity between basal and induced NSC Ex/Mv suggested that the cell-free antiviral actions of NSC Ex/Mv are innate immunity-like. Hence, these results in conjunction with findings in Figs. 1-4 further support the conclusion that murine NSC Ex/Mv have antiviral immunity actions intracellularly as well as extracellularly.

Finally, we also analyzed piRNAs which might target the genome of the HIV-based lentivirus which contained a few HIV genomic RNA segments including HIV long terminal repeats (LTR), psi, RRE, and ΔU3 sequences, as elucidated in Fig. S6A . In addition, this lentiviral genome contained CMV promoter sequence and GFP encoding sequence. Despite that the RNA sequence of this recombinant lentivirus is rather short, we identified a list of piRNAs with at least 15 nucleotides matching with these viral sequences and found quite a few which met Criteria 1 or 2 (Table S3) . Through qPCR, we examined some of these piRNAs and found that they were present in NSC Ex/Mv but much less detectable in MSC Ex/Mv (Fig. S6B) . Further, some of these NSC Ex/Mv piRNAs were also upregulated after a lentiviral pre-exposure to these NSC (Fig. S7A, B) . We also asked if GFP-encoding RNA sequence could have target piRNAs and if so, whether they could be induced through a lentiviral pre-exposure. As shown in Table S3 , we found several mouse piRNAs which matched against either the sense or antisense sequence of GFP encoding RNA. Using qPCR, we then measured a few of them and found that each was detectable in NSC Ex/Mv but much less detectable in MSC Ex/Mv (data not shown). Moreover, two of these GFP piRNAs were significantly upregulated in induced NSC Ex/Mv compared to basal NSC Ex/Mv (Fig. S7C ). All these results based on HIV-lentivirus were consistent with piRNA adaptive enrichment as described above for pseudotyped SARS-CoV-2. Also, based on J o u r n a l P r e -p r o o f upregulation of piRNAs against RNAs that encoded luciferase (Fig. S5) and GFP (Fig. 7C) , The adaptive reaction of NSC Ex/Mv piRNA can target a newly incorporated sequence in a viral genome. Taken together, through sequence analysis and qPCR for multiple viral types and components, we obtained information suggesting that piRNAs could be involved in the antiviral actions of these NSC Ex/Mv.

In this study, we showed that murine NSC Ex/Mv can provide innate and adaptive antiviral actions through testing pseudotyped SARS-CoV-2 and HIV-based recombinant lentivirus. As elucidated in Fig. 8 , the antiviral effects of these Ex/Mv are predicted to comprise an intracellular action in host cells as well as an extracellular action through which Ex/Mv particles directly interact with viruses. While these Ex/Mv contain other non-coding RNAs (such as miRNAs and lncRNAs) and proteins/peptides which might have contributions, this study uncovered the possible importance of piRNAs for the antiviral immunity of these Ex/Mv.

In clinical studies, largely due to the limited availability of human NSC, the option of using human MSC has been considered for exosomal therapeutics, and recent studies of using MSC or their derived exosomes to target COVID-19 showed positive outcomes (Bari et al., 2020; Borger et al., 2020; Gupta et al., 2020; Jayaramayya et al., 2020; Pinky et al., 2020; Tsuchiya et al., 2020) , although it is unclear if these effects were related to a direct antiviral action or indirectly due to improved physiology against viral infection. Given that NSCs are capable of abundantly producing Ex/Mv (Zhang et al., 2017) , it further points to our proposal that NSCs could be important for Ex/Mv therapeutics. Because of the various advantages of NSC exosomes including the pro-survival effects in fatal disease conditions (Tang et al., 2020) , we predict that J o u r n a l P r e -p r o o f adding the application of NSC Ex/Mv will increase the coverage and effectiveness of exosome therapeutics. Furthermore, we observed that murine NSC Ex/Mv can provide a cell-free antiviral action against viruses, raising up the possibility that murine NSC Ex/Mv could be used for humans through an external application such as nasal spray to break viruses such as SARS-CoV-2, despite that little is known about the differences between the murine and human NSCs to date.

In this work, we further discovered that murine NSC Ex/Mv are characterized by producing a vast collection of diverse piRNA species including those which could potentially target genomes of several RNA viruses. These piRNA libraries might contain species for targeting against many other viruses including DNA viruses which this study did not examine.

Probably, this level of vast diversity is related to the heterogeneity of NSC population, which had been established in the history of life evolution. Indeed, we have previously appreciated that NSC is very heterogenous (Tang et al., 2020) ; thus, approaches such as single-cell analysis will be valuable to reveal in-depth information and understandings. The notion of piRNA immunity and its enrichment through viral pre-exposure could lead to the extension of classical antibodybased adaptive immunity to small RNA-based disciplines. This feature can be clinically relevant, since vaccine development for RNA viruses is challenged by the situation that variants and mutations of RNA viruses are high, for instance, mutational and protein profile analyses revealed a large number of amino acid substitutions in SARS-CoV-2 indicating that the viral proteins are heterogeneous (Islam et al., 2020) . Given this reality, the observed antiviral actions of NSC Ex/Mv and the possible role of piRNAs in antiviral immunity can provide the initial clues and ideas to promote future research for studying if NSC Ex/Mv and piRNAs could be developed to complement with antibody-based vaccine strategy and development.

Finally, this research provided a long list of representative piRNAs with the sequences matching against 3 different kinds of viral genomes and we predict that at least some of them are likely important for the related antiviral actions of NSC Ex/Mv. The pool of piRNA sequences which match against viral genomes can be further expanded and examined based on the methodologies in this work. The sequence information of these piNRAs if verified for being antiviral will help guiding RNA-based vaccination strategies as well as exploring the underlying effects and mechanisms of these strategies. In this context, although we showed that NSC Ex/Mv are more important than MSC Ex/Mv for containing these piRNAs, we should point out that NSC might not be the only cell type for this feature, and to explore and find out other types of cells and even peripheral cells with this feature will be valuable for increasing the width and opportunity of potential applications against viral infection and diseases.

One limitation of this study is that our experiments evaluated only two recombinant laboratory viruses. To test wildtype viruses such as coronavirus and HIV is important for future research. The second limitation is, this study was based on in vitro cell culture models, thus it is important to call for research of using animal models and human disease conditions to evaluate the predicted translational and pharmacological values of these findings. Also, while the antiviral actions of the induced versions of NSC Ex/Mv are stronger, basal versions of these extracellular vesicles seem effective only for some cell types, thus it is necessary to understand the underlying process which is still unclear in this study. Third, the potential role of piRNAs was predicted based on sequence analysis and correlative experimental results (thus these results are presented in the supplemental section), so future studies are needed to verify if piRNAs per se can provide J o u r n a l P r e -p r o o f important antiviral actions. Finally, as this study focused on murine NSC Ex/Mv ̵ which does its own significance such as the possible application, it should be more informative to compare them with human NSC Ex/Mv in terms of piRNAs and antiviral immunity.

Author contribution: B.Y. performed experiments including cell models and culture, virus and Ex/Mv production and purification, western blot, immunostaining, and biochemical assays, did data analysis, and contributed to data interpretation. S.I. co-performed cellular, viral and Ex/Mv experiments, did western blot, prepared plasmids, and contributed to data analysis and interpretation. C.R. performed small RNA isolation and qPCR and analyzed these data; X.Z.

analyzed viral and piRNA sequences and designed piRNA primers. D.C. conceived all the ideas and concepts, designed the project (hypotheses, specific aims, experimental strategies, and methodologies), provided resources, organized and supervised the study, evaluated approaches, guided and led data analysis, led and finalized data interpretation, and wrote the paper. Material and data availability: All materials, methodologies, protocols and data information which were generated in this study are available upon request. Tables   Table S1. indicated NSC Ex/Mv were maintained along (labeled as "without pre-mixture") or mixed together (labeled as "with pre-mixture") for an overnight period (0.5 hour at room temperature followed by 4 °C), and were then used to infect HepG2 cells for 24 hours. Subsequently these cells were harvested and lysed for measurement of luciferase activities. *p < 0.05, **p < 0.01; ANOVA/post-hoc, compared between indicated groups, n = 3 independent biological samples per group, values represent the mean ± s.e.m. DAPI nuclear staining (blue) was included as a reference. Scale bar 20 μm. C, Coimmunostaining of PIWIL2 (red) and exosomal marker TSG101 (green) or CD81 (green) for Ex/Mv focusing on exosomes released from cultured htNSC P (representing other NSCs). Left panel shows a representative high-magnification view of individual exosomes. Right panel shows a representative higher-magnification view of single exosome particle. Scale bars, 500 nm.

A, Cultured A549 cells were infected with VSVG-incorporated GFP lentivirus in the presence or absence of Ex/Mv that were isolated from NSC types and MSC as indicated, and 2 days later, these cells were fixed for GFP immunostaining. DAPI nuclear staining was included as a technical control. Scale bar, 100 μm. B & C, Cultured A549 cells were infected with a standard (B) or high (C) dose of VSVG-enveloped GFP lentivirus in the presence or absence of Ex/Mv that were isolated from basal vs. induced NSC types as indicated, and 2 days later, these cells were harvested and lysed for western blot for GFP. Blot for beta tubulin or beta actin for the same membrane was included to provide a reference. n.s., non-specific.

VSVG-enveloped GFP lentiviruses were mixed with basal Ex/Mv from htNSC PGHM as labelled as htNSC P (A) or hpNSC (B) for 0.5 hour at room temperature and then 4−24 hours at 4 °C.

These lentiviruses or Ex/Mv alone maintained under the same conditions were included as two control groups. These samples were then lysed and processed for western blot using an antibody against VSVG. Ponceau staining of duplicated gels under the same procedure was used as a technical control to reflect the protein distribution in these gels.

A diagram describes the antiviral immunity of NSC-released Ex/Mv against various viruses (the same image is used as the graphic abstract of this paper to provide the overall idea). These antiviral effects are suggested to comprise an intracellular action in infected host cells which take Ex/Mv particles and their contents as well as an extracellular action through which Ex/Mv particles directly interact with viruses despite that the underlying intracellular and extracellular mechanisms are still unclear. While NSC Ex/Mv contain other non-coding RNAs (such as miRNAs and lncRNAs) and proteins/peptides which might provide antiviral contributions, this study revealed that these Ex/Mv have a key feature of innately and even adaptively producing large families of piRNAs against genomes of various viruses, suggesting a possible important role from these special small RNAs for antiviral immunity. This study addressed a recombinant retrovirus and a negative-strand RNA virus and although it did not investigate a positive-strand RNA virus such as wildtype SARS-CoV-2, this type of virus is proposed in this diagram based on piRNA sequence analysis. 

Mesenchymal Stromal Cell Secretome for Severe COVID-19 Infections: Premises for the Therapeutic Use

ISEV and ISCT statement on EVs from MSCs and other cells: considerations for potential therapeutic agents to suppress COVID-19. Cytotherapy

Chronic Inflammation Directs an Olfactory Stem Cell Functional Switch from Neuroregeneration to Immune Defense

Functional aspects of meningeal lymphatics in ageing and Alzheimer's disease

Mesenchymal stem cells and exosome therapy for COVID-19: current status and future perspective

SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Genome-wide analysis of SARS-CoV-2 virus strains circulating worldwide implicates heterogeneity

Immunomodulatory effect of mesenchymal stem cells and mesenchymal stem-cell-derived exosomes for COVID-19 treatment

Rapid linkage of innate immunological signals to adaptive immunity by the brain-fat axis

Structural and functional features of central nervous system lymphatic vessels

Expression of SARS-CoV-2 receptor ACE2 and TMPRSS2 in human primary conjunctival and pterygium cell lines and in mouse cornea

T cell engagement of crosspresenting microglia protects the brain from a nasal virus infection

Establishment and validation of a pseudovirus neutralization assay for SARS-CoV-2

Establishment and validation of a pseudovirus neutralization assay for SARS-CoV-2

Meningeal Lymphatics: From Anatomy to Central Nervous System Immune Surveillance

Molecular and Functional Neuroscience in Immunity

Mesenchymal Stem Cell Derived Exosomes: a Nano Platform for Therapeutics and Drug Delivery in Combating COVID-19

Acetylcholine-synthesizing T cells relay neural signals in a vagus nerve circuit

Identification of piRNA Binding Sites Reveals the Argonaute Regulatory Landscape of the C. elegans Germline

Multifaceted secretion of htNSC-derived hypothalamic islets induces survival and antidiabetic effect via peripheral implantation in mice

piRBase: a comprehensive database of piRNA sequences

Neuronal, stromal, and T-regulatory cell crosstalk in murine skeletal muscle

Generation of VSV pseudotypes using recombinant DeltaG-VSV for studies on virus entry, identification of entry inhibitors, and immune responses to vaccines

The piRNA targeting rules and the resistance to piRNA silencing in endogenous genes

Hypothalamic stem cells control ageing speed partly through exosomal miRNAs

Murine NSC-released exosomes/microvesicles have an innate antiviral action

Murine NSC-released exosomes/microvesicles can be induced adaptively to be antiviral

Murine NSC-released exosomes/microvesicles can target viruses in cell-free environment

Murine NSC exosomes/microvesicles produce piRNAs that are potentially antiviral

J o u r n a l P r e -p r o o f