key: cord-0888084-u8r42cod
authors: Kalarikkal, Sreeram Peringattu; Sundaram, Gopinath M.
title: Edible plant-derived exosomal microRNAs: Exploiting a cross-kingdom regulatory mechanism for targeting SARS-CoV-2
date: 2021-01-29
journal: Toxicol Appl Pharmacol
DOI: 10.1016/j.taap.2021.115425
sha: cf0d17b16a90473ac195ad60c17d38111277fadb
doc_id: 888084
cord_uid: u8r42cod

BACKGROUND: The current COVID-19 pandemic is caused by SARS-CoV-2 which belongs to coronaviridae family. Despite the global prevalence, there are currently no vaccines or drugs. Dietary plant derived exosome-like vesicles are known as edible nanoparticles (ENPs). ENPs are filled with microRNAs (miRNAs), in bioavailable form. Recently, cross-kingdom regulation of human transcripts by plant miRNAs have been demonstrated. However, ENP derived miRNAs targeting SARS-CoV-2 has not been described. STUDY DESIGN: Mature ENP-derived miRNA sequences were retrieved from small RNA sequencing datasets available in the literature. In silico target prediction was performed to identify miRNAs that could target SARS-CoV-2. ENPs were isolated from ginger and grapefruit plants and the expression of SARS-CoV-2 targeting miRNAs were confirmed by qRT-PCR. RESULTS: From a total of 260 ENP-derived miRNAs, we identified 22 miRNAs that could potentially target SARS-CoV-2 genome. 11 miRNAs showed absolute target specificity towards SARS-CoV-2 but not SARS-CoV. ENPs from soybean, ginger, hamimelon, grapefruit, tomato and pear possess multiple miRNAs targeting different regions within SARS-CoV-2. Interestingly, osa/cme miR-530b-5p specifically targeted the ribosomal slippage site between ORF1a and ORF1b. We validated the relative expression of six miRNAs (miR-5077, miR-6300, miR-156a, miR-169, miR-5059 and miR-166 m) in ginger and grapefruit ENPs by RT-PCR which showed differential enrichment of specific miRNAs in ginger and grapefruit ENPs. CONCLUSION: Since administration of ENPs leads to their accumulation into lung tissues in vivo, ENP derived miRNAs targeting SARS-CoV-2 genome has the potential to be developed as an alternative therapy.

within ENPs, including nucleic acids, exhibit excellent bioavailability and resistance to GI digestion process (Yang et al., 2018) . ENPs are non-toxic and lack immunogenicity which has led to their exploitation as a nanocarrier for in vivo delivery of biological macromolecules such as plasmid DNA, antibodies, phytochemicals, siRNAs and chemotherapeutic drugs (Rome, 2019; Wang et al., 2015; Wang et al., 2013) .

Interestingly, ENPs also contain numerous plant-derived miRNAs which has been predicted to target human transcripts. Xiao et. al., (2018) isolated ENPs from 11 edible plants and profiled miRNAs in these ENPs using small RNA sequencing . The authors identified between 32 and 127 miRNAs per plant species. Further, the human target mRNAs of these ENP-derived miRNAs were predicted and validated through chimeric luciferase reporter assay (Xiao et al., 2018) . Moreover, ginger ENPs have been extensively studied and have been proven to have versatile therapeutic benefits against inflammatory diseases. Ginger

ENPs have also been used as a nanocarrier for delivery of biological cargo. Ginger ENP contains 125 different miRNAs within which 124 miRNAs have been predicted to have putative human targets (Zhang et al., 2016a) .

It has earlier been demonstrated that intranasal as well as intravenous administration of ginger ENPs led to their homing in mice lung tissues and was detectable up to 72hr postadministration. Since lung tissue is one of the major replication sites for SARS-CoV-2, antiviral bioactives in ginger ENPs can be used as an effective therapeutic modality against COVID-19 (Wang et al., 2013) . In coronaviruses, even though the viral genome is encapsidated by the N protein, the sub-genomic RNAs produced during active viral infection provides an excellent target for plant miRNAs to act upon. Hence, it is also likely that ENP with specific miRNAs against SARS-CoV-2 genes, will decrease disease burden by its spontaneous intracellular uptake and delivery of COVID-19 targeting miRNAs in alveolar epithelial cells.

Journal Pre-proof However, it remains to be determined if edible plant ENP-derived miRNAs can target SARS-CoV-2 genes. Due to the relative abundance and diversity of miRNAs present within the various ENPs isolated so far, we hypothesize that ENP/s may also contain SARS-CoV-2 targeting miRNAs which can be used as a cost-effective and target-specific therapeutic modality against COVID-19 infection.

To this end, we first extracted the mature miRNA sequences that have been identified in ENPs from prior literature. Using in silico analysis with stringent selection criteria, we predicted SARS-CoV-2 targeting miRNAs that are present in high abundance in ENPs.

Target specificity of SARS-CoV-2 targeting miRNAs was compared by sequence alignment of miRNA binding site with different isolates of SARS-CoV-2 and SARS-CoV. As a proof of concept, ENPs were isolated from grapefruit and ginger rhizomes and SARS-CoV-2 targeting miRNAs were validated in these ENPs by quantitative real-time PCR (qRT-PCR).

The genome sequences of SARS-CoV-2 isolated from different geological locations across the world were downloaded from NCBI as described earlier (Sardar et al., 2020 Xiao et al., (2018) , the top 20 miRNAs constitute more than 90% of total ENP-associated miRNAs and hence, we chose the top 20 miRNAs, based on their read count from each ENP for our study. We also included the top 20 miRNAs for ginger and grapes from two other independent studies (Ju et al., 2013; Teng et al., 2018) .The ENP associated miRNAs from grapes were identified by microarray analysis and hence the actual read counts could not be established. In addition, the read counts of ginger ENP miRNAs cannot be compared between Teng et al., (2018) and Xiao et al., (2018) due to the experimental variability, library construction and the exact sequencing pipeline used. The final list of 260 ENP-miRNAs selected from all the three studies along with their relative abundances is presented in Supplementary file 1. Since some of the miRNA sequences represented in Xiao et al. (2018) were partial, we also downloaded the actual miRNA sequence for each accession no/miRNA name from miRbase version 22 and used for target prediction.

Target prediction was carried out using the RNA hybrid software with stringent prediction criteria using MT050493.1 as a template sequence (Rehmsmeier et al., 2004) . The following critical parameters were set during target prediction to identify physiologically relevant SARS-CoV-2 targeting miRNAs. The minimum free energy (MFE) was fixed at -25kcal/mol to ensure that the binding of miRNA to target mRNA can supersede the local secondary structures (Kertesz et al., 2007; Mückstein et al., 2006) . In animal miRNAs, seed sequence complementarity is a major determinant for efficient miRNA binding to target mRNA. J o u r n a l P r e -p r o o f Journal Pre-proof Hence, helix constraint was placed in seed sequences between 2-8 bases from the 5' end of the miRNA. Since non-canonical G-U base pairing is well tolerated in miRNA mediated target mRNA suppression in animal miRNAs, this was allowed (Saxena et al., 2003) . The maximum internal bulge and bulge loop length were restricted between 0-2 bases to ensure specific binding and to avoid false positives. Conservation of miRNA binding site across homologous species is a parameter used to assess the specificity of miRNA target interaction.

However, RNA hybrid does not allow the evaluation of binding site conservation within different viral genomes. Hence, to ensure the specificity of miRNA binding to SARS-CoV-2 genome, we aligned the miRNA binding site sequence within the SARS-CoV-2 genome (MT050493.1) with global isolates of SARS-CoV-2 as well as with SARS-CoV sequence using MEGA software, version 10.1 (Stecher et al., 2020) . Local RNA secondary structures that may be present in miRNA target sites were predicted using the MFOLD web server (http://unafold.rna.albany.edu/?q=mfold/RNA-Folding-Form) (Zuker, 2003) .

ENPs were isolated from fresh ginger rhizomes and edible portion of grapefruits using a polyethylene glycol (PEG)-based precipitation method developed in the laboratory (Kalarikkal et al., 2020) . We have earlier demonstrated that this method yields ENPs of similar biophysical and biochemical characteristics in comparison to the gold standard differential ultracentrifugation protocol. Total RNA was isolated from 100mg equivalent of ENPs using TRI reagent (Sigma) as per manufacturer's instructions. Total RNA was quantified by Nanodrop spectrophotometer and RNA quality was assessed by resolving it through 1.5% Agarose gel electrophoresis as per Kalarikkal et al., (2020) . To ascertain the RNA nature of the total RNA isolated, 1g of total RNA was incubated with 0.5g of RNAse A for 10 min at 37C before resolving them through 1.5% Agarose gel electrophoresis.

J o u r n a l P r e -p r o o f Journal Pre-proof Quantification of mature miRNAs using oligonucleotide primers was carried out as described earlier (Balcells et al., 2011) . Briefly, mature miRNAs present in total RNA (100ng) were first polyadenylated in a total volume of 10l containing 1l of 10X poly (A) polymerase buffer, 1l of 10mM ATP, 2.5units of poly (A) polymerase (New England Biolabs) for 30 min at 37C. The polyadenylated miRNAs were reverse transcribed using a universal reverse (Rev) primer which contains an anchored oligo (dT) sequence and an adapter sequence at the 5' end. To the reaction mix (10l), 4l of 5X RT buffer, 1l of 25mM dNTP, 1l of 10M universal Rev primer and 200 units of M-MuLV reverse transcriptase were added and the volume was made up to 20l with nuclease-free water. Reverse transcription was performed by incubation at 42C for 45 min followed by incubation at 85C for 5 min to inactivate the reverse transcriptase. The complementary DNA (cDNA) thus obtained was diluted 10 fold with RNAse free water and 5l of the diluted cDNA was used in qRT-PCR reactions containing 10l of 2X iTaq Universal SYBR green master mix (BioRad) and 250nM each of miRNA specific forward and universal real-time PCR reverse (qRT) primer in a total volume of 20l. QRT-PCR was carried out in BioRad CFX96 PCR system under the following conditions; initial denaturation at 95C for 3 min followed by 40 cycles of denaturation (95C for 10 sec), annealing (55C for 30 sec) and extension (72C for 30 sec). Automatic threshold option was used for each miRNA primer sets for the acquisition of Ct values. No template controls (NTC) lacking the cDNA were amplified to verify the specificity of the primers. The sequences of miRNA specific forward primers, universal Rev primer and universal qRT primer are provided in Table 1 .

qRT-PCR data are presented as mean ± S.D. Statistical significance was determined by unpaired Student's two-tailed t-test analysis, with p <0.05 considered to be statistically J o u r n a l P r e -p r o o f Journal Pre-proof significant. The experiments were repeated at least three times with two independent batches of ENPs..

Edible plant-derived ENPs contain SARS-CoV-2 targeting miRNAs:

The entire process of miRNA selection from different ENPs, mature miRNA sequence retrieval, target prediction, the specificity of miRNA-target interaction and secondary structure analysis is depicted as a flow chart in Figure 1 . First, we shortlisted top 20 miRNAs from 13 edible plant ENPs based on their relative read counts obtained from small RNA sequencing data sets for each plant species (Supplementary file 1). A total of 260 ENPassociated miRNAs were taken as input for this study and subjected to target site prediction against SARS-CoV-2 genome sequence with RNAhybrid software. With the stringent selection criteria (described in methods ), we obtained a sum of 22 ENP-derived miRNAs that could target SARS-CoV-2 genome ( Table 2) We identified both spike and ORF8 genes were targeted by at least two different miRNAs (gma-miR-4995, mdm-miR_1511 targeting spike transcript and pvu-miR_482-5p, gma-miR-166m targeting ORF8 transcript). Membrane, Nucleocapsid and ORF10 mRNAs were targeted by aqc-miR-159, zma-miR-164-3p and sly-miR-1919a, respectively ( Table 2) .

In coronaviruses, the replicase complex is encoded by the ORF1ab gene which occupies the first two thirds of the viral genome from the 5' end. ORF1ab codes for two different ORFs, namely ORF1a and ORF1b, arranged in tandem manner. Intact synthesis of both ORF1a and J o u r n a l P r e -p r o o f

ORF1b requires a ribosomal frameshift at the junction between ORF1a and ORF1b. This is achieved by a sequence and structure dependent ribosomal slippage event which occurs during the translation of ORF1ab mRNA. This phenomenon in coronaviruses requires two cis RNA elements, namely, a slippery sequence (UUUAAAC, 13442 to 13448 nt in SARS-CoV-2) and a hairpin type pseudoknot structure located at the downstream sequence (Baranov et al., 2005) . It has been proposed, that pausing of ribosomes in this hairpin pseudoknot allows ribosomal slippage by one base at the -1 direction leading to the in-frame translation of ORF1b. Disruption of ribosomal slippage event at this site will lead to the premature termination of translation at the stop codon located within the hairpin pseudoknot, thereby inhibiting the synthesis of ORF1b. Interestingly, cme-miR-530b and osa-miR-530-5p both of which share similar mature miRNA sequence was found to target ribosomal slippage region between two overlapping ORFs, ORF1a and b ( Figure 3A) . The osa/cme-miR-530-5p binding region overlaps with the pseudoknot structure encompassing the stop codon (13448 to 13469). The MFE of miR-530-5p binding far exceeds the MFE of the pseudoknot structure (-27 vs -18 kcal/mol). Hence, we posit that even in the absence of a canonical miRNA mediated suppression, this unwinding of pseudoknot by miR-530-5p may indirectly inhibit the synthesis of ORF1b by not allowing ribosomal slippage and thus preventing the replication of SARS-CoV-2 ( Figure 3A ).

To extrapolate our results for COVID-19 treatment, we plotted the relative abundance of the SARS-CoV-2 targeting miRNAs as a heatmap in 11 edible plant ENPs, using the raw read counts obtained from Xiao et al., (2018) . Hami melon, Soybean, tomato and pear ENPs has 6-7 miRNAs while grapefruit, pea and blueberry ENPs have 3-4 miRNAs that could target SARS-CoV-2 ( Figure 3B ). Gma-miR-166m which targets ORF8 in SARS-CoV-2 belongs to a highly conserved miR-166 family of miRNAs with 21 sub-members encoded within . This could be due to the technical variation during library preparation, sequencing platform or the variety of ginger used for ENP isolation. It is also worth noting that, aly-miR-166a-3p, which share identical sequence with gma-miR-166m with an additional U at the 5' end was also found to be highly enriched in ginger ENPs (Supplementary file 1) .

To investigate the specificity of the selected miRNAs against SARS-CoV-2, we aligned the miRNA target sequences identified using the template strain (MT050493.1) with sequences of SARS-CoV-2 isolated from other countries such as USA, Nepal, France, Italy and China (Wuhan isolate) and SARS-CoV. It has been observed that global strains of SARS-CoV-2 share 99% sequence similarity with the original Wuhan isolate (Sardar et al., 2020) . Due to this, all the 22 miRNA target sites were highly conserved across all isolates of SARS-CoV-2 ( Figure 4) . Even though SARS-CoV and SARS-CoV-2 are highly homologous, six regions of differences (RD) has been identified earlier which are located within ORF1ab, S, ORF7b and ORF8 genes (Xu et al., 2020a) . Based on this, we segregated the miRNA target sites into 3 groups. Group I consisted of 12 miRNA target sites in which nucleotide mismatch/s were found between SARS-CoV-2 isolates and SARS-CoV in seed sequences ( Figure 4A) . Given that complete complementarity at the seed sequences is essential for miRNA mediated target suppression, miRNAs targeting these regions are expected to be specific against SARS-CoV-2 but may not bind to SARS-CoV (Witkos et al., 2011) . Group II consisted of 3 miRNA target sites which are conserved between SARS-CoV-2 and SARS-CoV suggesting that these miRNAs can equally suppress both species of viruses ( Figure 4B) . On the other hand, 7 J o u r n a l P r e -p r o o f Journal Pre-proof miRNA target sites had full conservation at the seed sequence with partial mismatches at the 5' end and these are classified as Group III ( Figure 4C) . However, successful binding of miRNAs from Group III to target sites in SARS-CoV-2 may depend on the miRNA-target mRNA duplex strength and the context of the intracellular environment in virus infected cells.

The accessibility of the target site to miRNA binding is one of the critical factors determining successful miRNA mediated target mRNA suppression (Didiano and Hobert, 2006; Long et al., 2007; Witkos et al., 2011) . It has been experimentally validated that the local secondary structure between 17nt upstream and 13nt downstream of the miRNA binding site can influence target site accessibility to miRNAs (Kertesz et al., 2007) . Hence, we assessed the local secondary structure formation at the miRNA target site along with 17nt upstream and 13nt downstream sequences from SARS-CoV-2 using MFOLD webserver. Out of the 22 ENP-miRNA binding sites in SARS-CoV-2, we found 4 miRNA target sequences had at least one predicted secondary structure with MFE ≥ 15 kcal/mol (Supplementary file 2) . The target sites of pvu-miR-482-5p and zma-miR-398b-5p overlapped with local secondary structures with MFE of -15.4 kcal/mol and 20.8 kcal/mol, respectively. As discussed earlier, the target sites of cme-miR-530b and osa-miR-530-3p displayed secondary structures with an MFE of 17.5 kcal/mol due to hairpin pseudoknot involved in ribosomal slippage. Since all the 22 miRNA binding sites within SARS-CoV-2 genome formed secondary structures with MFE < -25 kcal/mol, we anticipate that the ENP-derived miRNAs could supersede the local secondary structures to allow its own binding.

To further confirm that the SARS-CoV-2 targeting miRNAs are present in high abundance within ENPs, we first purified ENPs from ginger rhizomes and grapefruit using a cost-J o u r n a l P r e -p r o o f Journal Pre-proof effective PEG method we have developed earlier (Kalarikkal et al., 2020) . Total RNA was isolated and resolved through agarose gel electrophoresis. As shown in Figure 5A , both ginger and grapefruit ENPs showed the presence of small RNA population that are less than 100 bp in size, visualized by ethidium bromide staining. Treatment with RNAse A led to the disappearance of these bands confirming that these are indeed small RNAs likely representing miRNAs extracted from ENPs. Further, relative expression of mature miRNAs were confirmed by qRT-PCR in ginger and grapefruit ENPs. To confirm the differential enrichment of miRNAs in ginger and grapefruit ENPs, we selected six miRNAs based on the heat map in Figure 3B . These include, miRNA that are expressed equally in both ginger and grapefruit ENP (gma-miR-166m & mtr-miR-156a), miRNAs that show higher enrichment in grapefruit ENPs (bdi-miR-5059 & osa-miR-5077) and miRNAs that show higher enrichment in ginger ENPs (aqc-miR-159 & gma-miR-6300). In agreement with the heatmap, we observed similar Ct values (between 26-28) for gma-miR-166m in both ginger and grapefruit ENPs suggesting equal expression levels ( Figure 5B ). Both bdi-miR-5059 and osa-miR-5077 showed significantly lower Ct values in grapefruit compared to ginger ENPs suggesting their higher expression in grapefruit ENPs, corroborating with their relative expression in the heatmap (Figure 5D & F) . However, very high levels of gma-miR-6300 was detected in both ginger and grapefruit ENPs (Ct values < 20) without a significant difference in enrichment ( Figure 5E ). Aqc-miR-159 showed higher expression in grapefruit ENPs compared to ginger ENPs, evidenced by significantly lower Ct values ( Figure 5C ). On the other hand, mtr-miR-156a showed higher expression in ginger ENPs compared to grapefruit ENPs, evidenced by significantly lower Ct values in ginger ENPs (Figure 5G) . The discrepancy observed between our qRT-PCR results and heatmap data generated from Xiao et al., (2018) in Figure 3B , for aqc-miR-159, gma-miR-6300 and mtr-miR-156a, could be due to intrinsic differences in miRNA expression levels caused by variation in demographic and climate J o u r n a l P r e -p r o o f Journal Pre-proof conditions in which ginger and grapefruit are grown. Reaction mix lacking template cDNA did not show detectable qRT-PCR amplification in all the cases suggesting that the products formed are the result of specific amplification of mature miRNAs. This is also confirmed by resolving the PCR products through Agarose gel electrophoresis which showed single amplicon (size >100bp) in the gel images shown below the bar graphs in Figure 5B to G.

The cross-kingdom regulation of human mRNAs by dietary plant-derived miRNAs is an interesting hypothesis with far-reaching implications. One such example is the treatment of devastating diseases with a diet enriched with candidate miRNA/s that can suppress diseasecausing mRNAs. Although many recent studies have provided a hint at this cross-kingdom regulation of human transcriptome via diet/plant-derived miRNAs, this has also been a subject of intense controversy (Sundaram, 2019) . Cross-kingdom regulation between plant miRNAs and human transcriptome was first reported for rice derived miR-156a/miR-168 and soybean miR-159 with human targets (Chin et al., 2016; Zhang et al., 2012) . These reports were later disputed due to the lack of data reproducibility, the sequence similarity between some of the plant miRNAs with human counterpart and the stability of plant miRNAs during GI digestion process (Sundaram, 2019) . However, this is possible when these plant miRNAs are protected within ENPs during GI digestion, for which several recent reports provide substantial evidence. In a study by Teng et al., (2018) , miR-7267-3p derived from ginger ENP was stable during GI digestion process and was shown to suppress Lactobacillus rhamnosus monooxygenase ycnE mRNA, in the gut microbiome, thereby modulating the gut immunity (Teng et al., 2018) . Recently, the presence of miR-156c and miR-159a have been demonstrated in ENPs derived from walnuts and these miRNAs can regulate mammalian TNF- signalling pathway in adipocytes and regulate inflammation (Aquilano et al., 2019) .

MiRNAs have also been found in grape-derived ENPs in high abundance and whether these miRNAs target human genes, is yet to be validated (Ju et al., 2013) .

In this study, we further extend the possibility of exploiting the cross-kingdom regulation by miRNAs to treat the current COVID-19 pandemic. In this line, we identified ENP-derived miRNAs that could potentially target SARS-CoV-2 genes. So far, ENPs have been isolated and characterized from several edible plants and fruits (Ju et al., 2013; Xiao et al., 2018; Yang et al., 2018; Zhang et al., 2016b) . We first mined these datasets to identify miRNAs that are present in high abundance in ENPs. Next, using a target prediction pipeline with stringent selection criteria, we identified 22 miRNAs that bind to SARS-CoV-2 and 12 of these miRNA candidates do not bind to SARS-CoV, a close relative of SARS-CoV-2 which caused the SARS outbreak in 2003. Further, we have also independently verified the presence of six miRNAs in ginger and grapefruit ENPs.

To our knowledge, this the first study which has systematically evaluated the possibility of using plant ENP-derived miRNAs to target viral genome. Plant miRNAs targeting influenza genome has been reported earlier for miR-2911, a miRNA highly expressed in honeysuckle (HS) plant. When mice were administered orally with HS decoction, high levels of miR-2911 was detected in serum and this could suppress influenza virus (H5N1) infection in vitro and in vivo (Zhou et al., 2015) . The same authors recently showed SARS-CoV-2 targeting potential of HS derived miR-2911. SARS-CoV-2 possessed twenty eight miR-2911 binding sites within its genome and the specific binding of miR-2911 mimics with SARS-CoV-2 target sites were validated in vitro with chimeric luciferase assays. Exosome encapsulated with miR-2911 mimics, was able to suppress SARS-CoV-2 viral replication in vitro. Further, administration of HS decoction to human volunteers led to a favourable reduction in time taken to become SARS-CoV-2 negative (Zhou et al., 2020) . However, in both influenza and SARS-CoV-2 studies, encapsulation of miR-2911 in vesicles, is an essential process for J o u r n a l P r e -p r o o f Journal Pre-proof systemic delivery of miR-2911. Interestingly, miR-2911 appears to be intrinsically resistant to nucleases and also have a natural tendency to associate itself within exosomes after absorption in GI tract. However, it is not clear if all plant-derived miRNAs from the diet exhibit nuclease resistant and are encapsulated in serum exosomes. Hence, plant miRNAs targeting SARS-CoV-2 may require additional encapsulation in liposomal or nano formulations for intracellular delivery (Lee et al., 2019) . The innate ability of ENPs to fuse with any type of mammalian cells obviates the need for separate miRNA delivery methods to target COVID-19 . Further, ENPs derived from different host plants appears to have preferential target tissue which is likely attributed to their lipid composition (Yang et al., 2018) . It has been shown earlier that intranasal/venous delivery of grapefruit derived ENPs, was detected in murine lung tissue up to 72 hr. post-delivery, suggesting that SARS-CoV-2 targeting ENP miRNAs could be delivered safely via the intranasal/venous route for tissues affected in COVID-19. However, further experimental research supporting spontaneous uptake of ENPs in lung epithelial cells in vivo, direct binding of ENP derived miRNAs to SARS-CoV-2 transcriptome and suppression of SARS-CoV-2 infection in vitro and in vivo are warranted in future. Alternatively, candidate miRNAs which efficiently suppresses SARS-CoV-2 replication can be overexpressed in edible plants by transgenic approach and ENPs isolated from the transgenic plants can be used for COVID-19 therapy (Yang et al., 2017) .

For the successful therapeutic application of ENPs/ENP derived miRNAs, large scale compatible purification methods are essential. We have recently demonstrated a costeffective and scalable purification method to purify ginger ENPs using PEG-6000 (Kalarikkal et al., 2020) . Ginger ENPs purified by this method was structurally and functionally similar to ENPs purified by the gold standard differential ultracentrifugation method which requires expensive instrumental set up and lack scalability. The PEG method developed was J o u r n a l P r e -p r o o f advantageous in multiple aspects. Firstly, food-grade PEG is cheaper (<4USD per Kg), it is generally regarded as safe (GRAS) material and it is also employed as a food additive (Panel et al., 2018) . Hence, we speculate that exploiting ENP-derived miRNAs as a therapeutic modality against COVID-19 infection offers a safer, target-specific and cheaper solution.

Thus, it is evident from our study that ENP derived miRNAs may be developed as a nontoxic therapeutic modality against COVID-19. However, further validation of ENP derived miRNAs binding to target mRNA (SARS-CoV-2 genome/transcriptome) could not be performed due to the requirement of advanced biosafety level -3 laboratories to handle virus cultures.

In this manuscript, we have provided in silico evidence for the ability of ENP derived miRNAs to target SARS-CoV-2 transcriptome. Further, the presence of SARS-CoV-2 target miRNAs was experimentally confirmed in ginger and grapefruit ENPs. The predicted miRNAs have also been verified for target site accessibility, suggesting their higher likelihood to target SARS-CoV-2 genome/transcriptome within the intracellular environment.

By employing SARS-CoV-2 targeting miRNAs in the form of ENPs, these miRNAs can selectively be delivered in primary target tissues of SARS-CoV-2. However, further investigations on the intracellular stability of these miRNAs and their purported anti-viral activity under in vitro and in vivo conditions have to be performed. In future, comprehensive experimental validation of these issues will aid in the development of efficient and non-toxic therapy for COVID-19. J o u r n a l P r e -p r o o f J o u r n a l P r e -p r o o f 

Adipocyte metabolism is improved by TNF receptor

Development of remdesivir repositioning as a nucleotide analog against COVID-19 RNA dependent RNA polymerase

Specific and sensitive quantitative RT-PCR of miRNAs with DNA primers

Programmed ribosomal frameshifting in decoding the SARS-CoV genome. of plant-derived nanovesicles

Perfect seed pairing is not a generally reliable predictor for miRNA-target interactions

Discovering drugs to treat coronavirus disease 2019 (COVID-19)

Grape exosome-like nanoparticles induce intestinal stem cells and protect mice from DSS-induced colitis

A costeffective polyethylene glycol-based method for the isolation of functional edible nanoparticles from ginger rhizomes

The role of site accessibility in microRNA target recognition

MicroRNA delivery through nanoparticles

Potent effect of target structure on microRNA function

The Molecular Biology of Coronaviruses

Thermodynamics of RNA-RNA binding

Overview of MicroRNA Biogenesis

Refined exposure assessment of polyethylene glycol (E 1521) from its use as a food additive

Fast and effective reveals unique features potentially consequential to host-virus interaction and pathogenesis

Small RNAs with imperfect match to endogenous mRNA repress translation. Implications for off-target activity of small inhibitory RNA in mammalian cells

Molecular Evolutionary Genetics Analysis (MEGA) for macOS

Dietary non-coding RNAs from plants: Fairy tale or treasure? Noncoding RNA Res

Plant-Derived Exosomal MicroRNAs Shape the Gut Microbiota

Grapefruit-derived nanovectors use an activated leukocyte trafficking pathway to deliver therapeutic agents to inflammatory tumor sites

Delivery of therapeutic agents by nanoparticles made of grapefruit-derived lipids

Practical Aspects of microRNA Target Prediction

Identification of exosome-like nanoparticle-derived microRNAs from 11 edible fruits and vegetables

Systematic Comparison of Two Animal-to-Human Transmitted Human Coronaviruses: SARS-CoV-2 and SARS-CoV

High expression of ACE2 receptor of 2019-nCoV on the epithelial cells of oral mucosa

Advances in plant-derived edible nanoparticle-based lipid nano-drug delivery systems as therapeutic nanomedicines

Bioavailability of transgenic microRNAs in genetically modified plants

Exogenous plant MIR168a specifically targets mammalian LDLRAP1: evidence of cross-kingdom regulation by microRNA

Edible ginger-derived nanoparticles: A novel therapeutic approach for the prevention and treatment of inflammatory bowel disease and colitis-associated cancer

Plant derived edible nanoparticles as a new therapeutic approach against diseases. Tissue Barriers 4, e1134415

Absorbed plant MIR2911 in honeysuckle decoction inhibits SARS-CoV-2 replication and accelerates the negative conversion of

Honeysuckle-encoded atypical microRNA2911 directly targets influenza A viruses

Mfold web server for nucleic acid folding and hybridization prediction

Edible plant derived nanoparticles (ENPs) are non-toxic, exosome-like vesicles from plants 2. We identified 22 ENP miRNAs targeting both structural and non-structural SARS-CoV-2 genes 3

Expression of SARS-CoV-2 targeting miRNAs in ginger and grapefruit ENPs confirmed by qRT-PCR

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