MicroRNAs (miRNAs/miRs) are small single-stranded non-coding RNA molecules of ~24 nucleotides in length, discovered in Caenorhabditis elegans in 1993 (1, 2). MiRNAs play a key role in post-transcriptional gene expression regulation by their base complementary binding to mRNAs, which promote mRNA degradation or suppress translation (1, 3). MiRNAs participate in several biological functions in animals, such as metabolism (4), immune system development and immune responses (5), proliferation, cell death, differentiation (6) or embryogenesis (7). In plants, miRNAs are attributed a role in processes such as plant development (8) stress responses (9), metabolism or cell signaling (10, 11). Of note, mature plant miRNAs are characterized by a methylation in the 3-terminal nucleotide 2-hydroxyl group, which increases stability by protecting miRNAs from exonuclease degradation (12, 13). In vascular plants, RNA molecules, including miRNAs, could be transferred locally between cells and over long-distances, by the plasmodesmata and the phloem vasculature, which suggest a role for RNA molecules in transmitting information beyond their cell of synthesis (influencing processes such as stress responses or development) (14–16).

Indeed, plant miRNAs have gained special attention in recent years for their potential role as cross-kingdom gene expression regulators, influencing plant interactions with animals and microorganisms, which has been discussed in a number of reports (17–19). Thus, plant miRNAs have been extensively studied for their involvement in plant-microorganism interactions, influencing processes like the arm race co-evolution between host and parasites (20, 21). In animals, diet would be a primary source of phase plant miRNA uptake, subsequently being detected in fluids and tissues of a wide range of species such as mice (22), pigs (23) and humans (24), whereby have been termed as xeno-miRNAs (xenomiRs). The mechanisms through which xenomiRs would penetrate into the bloodstream to eventually reach animal tissues and organs have been reviewed by several authors (17, 25, 26). To sum up,

In prokaryotes, in particular bacteria, several small non-coding RNAs (snRNAs) from 50 to 200 nucleotides with post-transcriptional regulatory functions have been also identified. These prokaryotic snRNAs affect bacterial physiology by modulating many processes, such as microorganism virulence and communication, metabolism or stress response (32, 33). A wide range of those snRNAs exert their post-transcriptional effects through base complementary binding to mRNAs targets, being considered functional analogs to eukaryotic miRNAs (32, 34). However, these prokaryotic snRNAs present many differences with eukaryotic miRNAs, regarding their biogenesis, their mRNA binding site, the presentation to mRNA target and/or their mechanism of action, since they can inhibit or enhance mRNA translation, meanwhile eukaryotic miRNAs classically down-regulate gene expression (35, 36). Notably, a subtype of prokaryotic snRNA of similar size to eukaryotic miRNAs has been recently identified. This newly identified snRNA has received several denominations, such as micro-like size sRNAs or microRNA-like molecules, being thus considered as typical miRNAs (35, 37, 38). Indeed, it has been reported that prokaryotic microRNA-like molecules could play an important role as virulence factors (39, 40). For instance, Gu et al. (40) showed that Salmonella could take advantage of the machinery of human intestinal epithelial cells to produce microRNA-like molecules as virulence factors, thus promoting bacteria survival and proliferation in mice.

On the other hand, the influence of diet on microbiota composition has been extensively investigated and documented (41–43). Gut microbiota is involved in essential functions for host physiology such as metabolism (through, for example, the fermentation of indigestible carbohydrates and the production of metabolites important for organism homeostasis maintenance) (44, 45), immune system development and function (46), protection against pathogen colonization (47) or intestinal barrier shaping, been involved in their development, reparation and integrity and functionality maintenance (48–51). Additionally, a large number of studies reported that host-derived miRNAs could shape gut microbiota composition through modulation of expression of genes that affect microbe growth (52, 53). In turn, gut microbiota could regulate miRNA expression of host cells, such as intestinal epithelial cells (54–56). Furthermore, some authors pointed out that the bidirectional interaction between host miRNAs and gut microbiota could have an impact on intestinal permeability (52, 54). However, although host fecal miRNAs can modulate the microbiome (53), the role of non-absorbed dietary xenomiRs on microbiota composition modulation by regulating gene expression is scarcely known. Interestingly, it has been postulated that prokaryotic miRNA-like molecules could have an important role in cross-kingdom gene expression regulation, which has been proposed by several authors (35, 57). For example, Shmaryahu et al. (58) predicted pathogenic-bacteria miRNAs that may potentially regulate the expression of human genes involved in several diseases such as diabetes, colon cancer or leukemia. In fact, the authors demonstrated that the prokaryotic miRNAs could down-regulate target gene expression in human cells. The studies carried out by Choi et al. (39) revealed that prokaryotic miRNA-size molecules secreted by periodontal pathogens in vesicles, could enter human cells and down-regulate cytokine expression. However, most of the documented cases of cross-kingdom communication mediated by prokaryote miRNA-like molecules usually refer to pathogenic bacteria and an interaction between non-pathogenic bacteria, such as microbiota, and host cells, mediated by prokaryotic miRNA-like molecules is not well-known. Even so, it has been postulated that not only host miRNAs, but also prokaryotic miRNA-like molecules could also be involved in microbiota-host communication (59, 60). Indeed, and due to the lack of solid evidences regarding the influence of prokaryotic snRNAs, particularly miRNA-like molecules in human physiology, they are out of the scope of the current review.

In this review, we ought to provide new insights on the potential role of the gut microbiota influencing dietary plant xenomiR absorption and to summarize the evidences provided so far regarding the impact that the crosstalk between diet, gut microbiota and mammalian cells could exert on xenomiR absorption/uptake.

One of the potential fates of plant-derived xenomiRs incorporated through diet, once they have reached the gastrointestinal tract, is to penetrate into the bloodstream and eventually, to exert a regulation of the expression of target genes in several tissues and organs. However, the specific mechanisms involved in this hypothesis are, to date, still unclear.

Zhang et al. in 2011 (29) reported the presence of rice (Oryza sativa) miR-168a in the serum from humans and other animals. This xenomiR could inhibit the expression of the low-density lipoprotein receptor adapter protein 1 (LDLRAP1) in mouse liver, with physiological consequences in the organism, such as the increase of LDL in plasma. After this pioneer study, numerous publications have continued to report the presence of xenomiRs from diverse plants in the serum and tissues of mammals, including humans (22–24, 61, 62). Mlotshwa et al. (63) revealed that oral administered synthetic miR-34a, miR-143 and miR-145, which mimic the characteristics of plant miRNAs, were capable of displaying anti-tumor properties, thus preventing the development and lessen colon cancer progression in mice. Furthermore, Zho et al. (30) discovered that honeysuckle (Lonicera japonica) miR-2911, once it had crossed the gastrointestinal barrier and reached the lungs, was able to repress PB2 and NS1 mRNA expression, which encode for two proteins essential for H1N1 Influenza A virus replication, diminishing mortality rate in mice fed with honeysuckle. Subsequently, Chin et al. (64) detected plant miR-159 in human serum and tissues. Their data revealed that, upon oral uptake of synthetic miR-159, TCF7 gene expression was inhibited, resulting in repression of breast cancer tumor growth in mice. Recent studies are still providing data about the relevance of inter-species gene regulation by plant xenomiRs. Thus, Du et al. (65) showed that Rhodiola crenulata small RNA HJT-sRNA-m7 can reach the lungs via intratracheally administration and down-regulate the expression of genes involved in pulmonary fibrosis development in mice, which could culminate in the amelioration of the disease. Moreover, Svecia et al. (66) detected an increase in plant xenomiR levels in the plasma of humans who had taken Tuscany Sangiovese grape juice. They also revealed that, in mice, Tuscany Sangiovese miRNAs appeared to play a cardioprotective role maintaining the functionality of hearts that had suffered a myocardial infarction.

Despite the huge amount of emerging evidences of the capacity of plant miRNAs from food to be absorbed, travel through the bloodstream and modulate the expression of target genes in mammals, several studies do not fully support the xenomiR cross-kingdom regulation hypothesis due to the inability to detect xenomiRs in serum or tissues. In fact, Zhang et al. (67) questioned technical/methodological issues affecting miRNA detection in publically available databases and alleged that dietary plant miRNAs present in some databases of animal small RNAs could be cross-contaminations during sequencing processes. Some supportive and refuting evidences of the cross-kingdom regulation dietary xenomiR hypothesis have been summarized in Table 1 and have been also extensively summarized by other works (17).

One of the first studies that cast doubt on cross-kingdom uptake of xenomiRs was conducted by Dickinson et al. (68), who unsuccessfully tried to reproduce Zhang's results (29). These authors observed that xenomiRs levels in mouse plasma and liver after ingestion of rice chow were negligible. They also did not detect any change in LDLRAP1 expression in liver, even though LDL levels were increased. So, they postulated that the result obtained by Zhang et al. (29) could arise from sequencing errors, cross-contamination or the use of little sensitive and specific techniques, while the increase in plasma LDL could be a consequence of the impact of diet and not of the xenomiR gene expression modulation (68). However, although their data were controversial, Zhang et al. (29) did not retract their findings, so far.

Afterwards, Snow et al. (69) published a study in which xenomiRs were not detected in the plasma of healthy humans and mice that had been fed with a diet enriched in certain plant miRNAs. On the contrary, Witwer et al. (70) were able to detect a low amount of plant xenomiRs in plasma of macaques fed with plant miRNAs-rich substances, but they concluded that this relatively low xenomiR circulating levels could have been originated in a non-specific amplification. Later, Pastrello et al. (73, 74) retracted their report in which they detected miRNAs from broccoli (Brassica oleracea) in the serum of human subjects who had consumed this vegetable, alleging errors in primers design. In the experiments performed by Huang et al. (71) no significant differences were detected in either serum or tissues of mice that had been fed with chow diet supplemented with corn miRNAs. In light of these findings, they suggested that most part of xenomiRs might have been degraded during digestive process and the minimal amount of xenomiRs that could prevail and reach tissues would be insufficient to modulate the expression of endogenous host genes and cause physiological changes. Finally, Micó et al. (72) reported an increase of extra virgin olive oil miRNAs in the plasma of healthy humans but they considered that the negligible levels detected were artifacts resulted from cross-contamination. Analysis of plant miRNAs in human sequencing databases has led many authors to suggest that the underlying causes of xenomiR detection in mammal's body fluids and tissues would be technical artifacts due to cross-contamination during library preparation. The authors conclude that the abundance of xenomiRs in human tissues and fluids would be negligible and there would not be significant differences in terms of xenomiR levels in tissues exposed to food intake, such as the liver, compared to those that have no direct relationship with, like the cerebrospinal fluid (75, 76). Another study demonstrated that miRNA purification columns themselves could be a capture focus for exogenous miRNAs with an unknown origin, many of which appear reflected in human miRNA sequencing databases (77).

Efforts to explain the disparity of evidences for and against xenomiR hypothesis continue monopolizing the core of many investigations. It seems obvious that technical factors such as cross-contaminations, miRNAs purification methods, xenomiR degradation during digestion, and errors during sequencing processes (primer design and non-specific sequencing), could throw into question many of the studies that suggest that xenomiRs could modulate gene expression among different species and kingdoms. Nevertheless, recent reports such as the published by Zhao et al. (78) highlight the need to continue researching in this area and open the door to the exploration of other factors that could shed light and clarify the xenomiRs-host transfer hypothesis. After analyzing an enormous amount of small RNA sequencing data in public databases, Zhao et al. (78) reported that plant miRNAs could reach human fluids and tissues. Unlike previous studies, a remarkably thorough filtering of the sequencing reads was achieved in order to avoid any type of false positives, and cross-contamination was also strongly ruled out (78).

In this way, Zhao et al. (79) recently suggested that one of the factors that could explain the lack of detection of plant miRNAs in animals by certain studies would be the miRNA sequence. Thus, the authors pointed out that not all plant miRNAs could be absorbed, but it would take place a selective absorption reliant on the sequence, in which the high GC content, the shorten length and the motif “CAG” would be determinant in promoting xenomiR uptake (79). Another potential explanatory factor of the variance of xenomiR detection could be intestinal barrier permeability, which will be addressed in the following section.

Concerning current evidences, we propose that differences in the absorption of plant miRNAs between subjects could be an important factor that could explain, at least in part, the observed variability in the detection of plant miRNAs in mammalian plasma and tissues among different research groups. The permeability of the intestinal barrier could play a major role in the absorption efficiency of xenomiRs and further exploration of its potential role is required. In support of this notion, the studies of Raoof et al. (80) showed an increase of antisense oligonucleotide plasma levels when the epithelial intestinal barrier is leakier. The co-administration of phosphorothioate antisense oligonucleotide ISIS 104838 drug and the permeation enhancer sodium caprate, which promotes changes in intestinal barrier permeability, improved the bioavailability of the drug in dog blood (80, 81). In this context, Yang et al. (61) suggested that the alteration of gastrointestinal barrier permeability could enhance xenomiR uptake and entry to the bloodstream. They observed an increase in diet miRNAs in the bloodstream of mice who had their intestinal barrier integrity altered with cisplatin, suggesting that xenomiR uptake could be enhanced during certain pathologies or chemotherapeutic treatments that alter intestinal barrier integrity (61). Recently, Yang et al. (82) generated mouse models that displayed increased intestinal permeability through the administration of aspirin or anti-CD3 antibodies. In these models, plant-based small RNA uptake was higher as compared to controls. However, their data pointed out that changes in intestinal permeability could only have a significant impact on the absorption of the more stable small RNAs, since the less digestion-resistant RNAs would be degraded (82). Nevertheless, although modulation of intestinal permeability by itself could not have a relevant impact on the bioavailability of the most degradation-sensitive xenomiRs during cooking or digestion, we propose that it could entail special relevance in those cases where xenomiRs were transported in exosome-like extracellular vesicles, since they could be highly resistant to degradation (83, 84) (this issue is further developed and more widely documented in the following section) (section Involvement of Gut Microbiota in Release of Plant miRNAs From Extracellular Vesicles).

Recent studies have highlighted that microbiota could play a key role in maintaining integrity and functionality of intestinal barrier in physiological conditions and during adult lifelong, since it would be involved in determining paracellular permeability through mechanisms such as expression modulation of proteins that are part of tight junctions. Compared to controls, germ-free mice displayed lower paracellular permeability, and transplantation of fecal microbiota from healthy humans led to restoration of paracellular permeability up to levels considered normal, through the decrease of claudin-1 expression and the transient IL-18 production by enterocytes (85).

For this reason, we hypothesize that it would be possible that specific gut microbes could have different effects on intestinal barrier integrity in such a way that a given microorganism could increase permeability while others could be involved in its decrease or have any effect. Thus, there could be an inherent variability as regards of intestinal barrier permeability among different subjects that could depend on the specific composition of gut microbiota. In support of this hypothesis, several authors have reported that colonization of germ-free mice with Bacteroides thetaiotaomicron or Escherichia coli Nissle 1917 (EcN) led to up-regulation of genes that encode for proteins involved in maintaining adhesion between intestinal epithelial cells, such as small proline-rich protein-2 (sprr2a) and zonula occludens-1 (ZO-1), respectively, which for EcN was proved to result in decreased intestinal permeability when colitis mouse model was treated with this bacterium (86, 87). By contrast, other bacteria such as Escherichia coli MG1655 (K12) had no impact on intestinal barrier fortification (87). Results presented in other reports suggest that pathogens, such as enterohemorrhagic Escherichia coli (EHEC) O157:H7, enterotoxigenic Escherichia coli (ETEC) K88 and Salmonella typhimurium SL1344, could compromise intestinal barrier integrity. Infection of epithelial cell monolayers with EHEC O157:H7 promoted claudin-1 and ZO-1 redistribution and further down-regulation of ZO-1 expression, while toxins released upon the infection of epithelial cells with ETEC K88 and Salmonella typhimurium SL1344 resulted in disruption of the complexes that maintain adhesion between intestinal epithelial cells (88, 89). Remarkably, the increase in paracellular permeability elicited by these pathogenic bacteria could be ameliorated by certain Lactobacillus species (88, 89). It is worth mentioning that non-pathogenic commensal bacteria could also enhance intestinal barrier permeability, like Escherichia coli C25, which altered claudin-1 localization increasing intestinal permeability in vitro (90).

The influence of microbiota in modulating integrity and functionality of intestinal barrier has been demonstrated in a large bulk of studies showing that gut microbiota dysbiosis could contribute to the alteration of intestinal barrier with notable implications in development and/or progression of diverse pathologies. Cani et al. (91) suggested that microbiota dysbiosis could be the underlying cause of the inflammatory phenotype that characterizes obese and diabetic individuals by disrupting intestinal barrier. They also proposed that an increase of Bifidobacterium species abundance would conduce to higher glucagon-like peptide-2 (GLP-2) production which, among other effects, could led to up-regulation of occludin and ZO-1 expression, decreasing intestinal permeability and culminating in a decrease of inflammation (91). For example, hepatic damage caused by cadmium intake could be exacerbated by microbiota dysbiosis. Low-dose intake of this toxic metal for long periods reduced Akkermansia muciniphila and increased the relative abundance of Acidithiobacillaceae, Gammaproteobacteria, Methylobacterium, and Rhizobiales in mice. The imbalance in microbe proportion affected occludin, claudin-1 and ZO-1 expression, the down-regulation of which increased intestinal permeability, cadmium uptake and subsequent accumulation in the liver (92). Alteration of intestinal microbiota, specifically the decrease in butyrate-producing gram-positive bacteria and the increase in gram-negative bacteria such as Bacteroidetes, would be the cause of intestinal damage produced after skin burns in mice. The authors proposed that decreased butyrate levels would be primarily responsible for the increase of intestinal permeability and inflammation (93, 94).

It has also been reported that disruption of intestinal barrier integrity on the grounds of a severe physiological stress could be the result of microbiota dysbiosis. Specifically, decrease of Actinobacteria, such as Bifidobacterium and Collinsella, and increase of Proteobacteria would cause imbalance of several metabolites, like cysteine, thus increasing intestinal permeability (95). The increase of Proteobacteria has also been associated with the leaky gut phenotype that characterizes patients with acute coronary syndrome (96). Likewise, microbiota dysbiosis has been proposed as a mediator of intestinal permeability changes through dysregulation of immune system of children with beta cell autoimmunity at risk for type 1 diabetes (97). In other diseases such as non-alcoholic steatohepatitis, psychiatric disorders or cancer, it has been suggested that the enhancement of intestinal barrier permeability due to microbiota dysbiosis, could aggravate, and even in some cases be a requirement, for their development (98–102).

With all these supportive data, it appears that microbiota could be a key driving force controlling intestinal permeability, which in turn could modulate bioavailability of nutrients and therefore, their presence in blood circulation for potential biological action. It should be noticed that gut microbiota is not static and immutable but it varies between individuals and even within the same individual throughout life, under normal but also pathological states (103). Age, environment, genetics, diet and physiological status could render account for the aforementioned inter- and intra-individual variability observed in humans (103). Intestinal microbiota, together with genetic variability, are considered leading factors determining bioavailability and bioefficacy of plant bioactive compounds, explaining inter-individual variability concerning their digestion, absorption, distribution, metabolism, and secretion (104). For this reason, we propose that intestinal microbiota could be a major factor determining the amount of xenomiRs absorbed through modulation of intestinal barrier permeability under physiological and pathological conditions. We also suggest that divergence regarding specific configuration of microbiota communities among individuals could entail differences in xenomiR uptake efficiency via mechanisms like modulation of gastrointestinal barrier permeability. Some above evidences of the impact of bacteria on intestinal barrier integrity and functionality can be found in Table 2.

Extracellular vesicles have been discovered in both prokaryotic and eukaryotic organisms, including bacteria, archaea, fungi, plants, and mammals. They carry several types of bioactive molecules and thus, they are key factors in intercellular communications (105). In bacteria, extracellular vesicles range from 20 to 400 nm in diameter and they have been involved in bacterial functions such as biofilm formation, competition, antibiotic resistance or horizontal gene transfer (106). Extracellular vesicles of bacterial origin are not only involved in bacteria-bacteria communication, but also in bacteria-host cell communication, being able to elicit phenotypic changes in host cells altering their function with both detrimental and beneficial actions (107–109). Characterization studies of their cargo showed that they can contain several types of molecules, including structural and soluble proteins such as enzymes, secondary metabolites, virulence factors, glycolipids, lipopolysaccharides, bacterial antigens and a wide range of nucleic acids, including DNA and microRNA-like molecules, which have been extensively summarized in other reviews (110–112). Remarkably, gut microbiota can release extracellular vesicles that can influence host physiology, playing part in processes such as digestion (by carrying digestive enzymes), intestinal permeability regulation, metabolism or gut immunity (113–115).

In mammals, extracellular vesicles range in size from 50 to 100 nm in diameter and they modulate system biology as tissue repair, coagulation or embryonic development. Mammalian extracellular vesicles are also involved in disease progression, such as cancer, neurodegenerative and/or metabolic diseases (33, 116–118). The cargo of mammalian extracellular vesicles include lipids (being enriched in lipids such as cholesterol and sphingolipids), cytokines, proteins (in particular plasmatic and cytosolic proteins), nucleic acids (DNA and RNA species, such as miRNAs and mRNAs), amino acids, fatty acids and sugars, and they largely reflect the parental cell content (117, 119, 120). Mammalian extracellular vesicles are classified in microvesicles, exosomes and apoptotic bodies, which differ in their biogenesis and size (121). Contrary to microvesicles and exosomes, apoptotic bodies can contain organelles and chromatin, since they are released from cells undergoing programmed cell death (120). Notably, host extracellular vesicles could also influence gut microbiota, occurring a bidirectional communication mediated by microbiota and host extracellular vesicles (122).

In plants, extracellular vesicles range from 400 to 1,000 nm, depending on the species, and they have similar features (i.e., cargo, morphology and secretion) as mammalian exosomes (110, 123). Plant extracellular vesicles, frequently referred as “exosome-like” nanovesicles, contain lipids, proteins, metabolites and nucleic acids, including smalls RNAs (110, 123, 124). Interestingly, the lipids that constitute the nanovesicle surface confer specificity of uptake by different cell types (125). Plant extracellular vesicles are also known to contain antimicrobial compounds, including defense-signaling lipids and defense-related proteins, and they are involved in functions such as defense against pathogens, immune response, growth and development or symbiosis (124, 126). Plant exosome-like nanovesicles could have an important role in inter-kingdom communication. For example, they could be uptaken by gut microbiota or mammalian cells and their cargo, such as xenomiRs, might have a relevant impact, as it will be further detailed in the following sections (83, 125).

Plant cells can secrete miRNAs to the extracellular medium, as a part of ribonucleoproteins and/or encapsulated in vesicles of different origins (14), which has been reviewed elsewhere (26, 127). It has been observed that a wide range of edible plant can release exosome-like nanovesicles that yield biological effect in mammals. It has also been proposed that plant exosome-like nanovesicles could be an efficient mechanism of transporting bioactive compounds since they provide resistance to degradation, which together with its biocompatibility, led them to even apply as drug delivery system to treat diseases (128–131). The characterization of plant exosome-like nanovesicles has revealed that, among the small RNA cargo, they are enriched in miRNAs, which are selectively loaded in them (128, 132, 133).

Of note, a study carried out by Philip et al. (134) suggested that plant xenomiR bioavailability could be enhanced by processes that promote cell wall disintegration, such as cooking. The authors observed that miRNA levels were higher in cooked beans and brown rice compared to raw controls, and that cooking food promotes the miRNA release into the cooking water. Interestingly, several studies have documented that certain bacteria belonging to Firmicutes and Bacteroides phyla (such as Ruminococcus champanellensis, Bacteroides intestinalis or Bacteroides thetaiotaomicron), which are part of the human gut microbial community, are able to degrade cellulose, hemicellulose and pectins, major components of the cell wall (135–139). In fact, it has been attributed a role for gut microbiota in enhancing bioaccesibility of fiber-encapsulated nutrients, allowing its release through enzymatic activities capable of fermenting plant-cell wall component, which would be crucial for intestinal nutrient absorption (140, 141).

Furthermore, some studies suggested that some gut microbiota bacteria could hydrolyse lipids via lipase enzymes (142). Indeed, specific bacteria from bovine raw milk possess phospholipolytic activity capable of disrupting milk fat globule membranes, whose main constituents are phospholipids, as in cellular membranes and extracellular vesicles (128, 143, 144). Remarkably, it has been proposed that gut microbiota could also utilize lipolytic and phospholipolytic enzymes to digest milk fat, which in turn might led to changes in gut microbiome (145), which should be deeply explored and confirmed in further studies. Indeed, ongoing studies are exploring the biochemical pathway by which gut microbiota could digest milk fat globules through lipolytic and phospholipolytic enzymes activities (146) to further study the relevance of this newly proposed hypothesis.

Although this field of research has not been explored in depth yet, together, the summarized findings have led us to propose the following hypothesis: gut microbiota could contribute to degrade milk fat globule membrane lipids, as well as cellulose, hemicellulose and pectin fibers. In a similar manner, specific gut microbiota composition could be able to hydrolyse envelope lipids of miRNA-containing plant extracellular vesicles and/or cellulose, hemicellulose and pectin fibers (that might have become attached to plant extracellular vesicles during exocytosis and come out across the cell wall), which might enhance the amount of xenomiRs released in the gastrointestinal tract. Therefore, it should be explored if gut microbiota could play a relevant role in determining the bioaccessibility and bioavailability of plant miRNA encapsulated in extracellular vesicles, through degradation of extracellular vesicle envelope components.

Alto, it is worth to underline that plant extracellular vesicles can directly interact with target cells and release their cargo through several mechanisms such as macropinocytosis or endocytosis (28, 147). Thus, one of the issues that pose the above proposed hypothesis on the role of gut microbiota eliciting the release of miRNA from plant extracellular vesicles, is whether (a) it could be a complementary or facilitating mechanism to promote in vivo capture of plant extracellular vesicles, or on the contrary, (b) it could interfere with the suggested selective capture of plant nanovesicles that might depend on the specific lipid membrane composition (125). It would be also worthy to address whether the hypothetical microbiota-dependent plant xenomiR release could entail special relevance in those cases in which plant xenomiRs must be available to exert the documented non-canonical function of binding to cell surface receptor and thus, to modulate cellular signaling pathways (148, 149).

Another issue in which it should be important to dig into consists on determining whether inter-individual differences in gut microbiota composition between healthy subjects or dysbiosis associated to some metabolic diseases like obesity (150, 151) could lead to dissimilarities on the efficiency of the proposed plant xenomiR extracellular vesicles release mechanism. Finally, we hypothesize that gut microbiota could play a major role determining plant xenomiR intestinal absorption rate through the modulation of both miRNA extracellular vesicles release and intestinal barrier permeability, which might partly been explained by the disparity of results regarding the detection of plant xenomiRs in mammals and the eventual veracity of the still controversial plant xenomiR cross-kingdom regulation hypothesis.

Some studies have suggested that most dietary plant xenomiRs would be degraded during digestive processes and the amount that could reach host cells might not be enough to exert cross-kingdom gene expression regulation (71, 82, 152). However, substantial amount of reports has unveiled that plant xenomiRs would be able to withstand harsh conditions of processing, cooking and digestion (22, 23, 134), and thus, could reach a functionally relevant copy number per cell that might surpass a minimum threshold to effectively modulate endogenous host gene expression (153). Within this context, it has been well-documented that factors such as plant miRNA GC content (30), 3'terminal nucleotide 2'-O-metylation of plant miRNAs (13, 29), RNA-binding proteins (154, 155) and especially the packaging into exosome-like nanovesicles (83, 84, 147, 156, 157), are responsible for plant xenomiR stability, greatly endowing them with the capacity to remain unaltered and reach the gastrointestinal tract, to potentially modulate host gene expression.

Another aim of this review is to provide new insights into the potential role of plant xenomiRs in modulating gut barrier function, by interacting with gut microbiota, intestinal epithelial cells and intestinal immune system. The description of the global plant xenomiR effects, once they are absorbed and travel through the bloodstream reaching specific mammalian tissues and organs, goes beyond of the scope of this review and can be found summarized in recent works (17, 127, 158).

The current lack of consensus regarding the veracity of the plant-derived xenomiR hypothesis highlights the need to thoroughly identify the factors underlying the high variability of results obtained in the different studies, including optimizing technical methods. One of the factors that might potentially contribute to explain the lack of detection of plant miRNAs in animals by certain studies could be intestinal permeability, since it could play a relevant role in xenomiR absorption. In parallel, there is a strong support for the crucial role of gut microbiota controlling intestinal barrier integrity and function, standing as an important modulator of intestinal permeability. Remarkably, gut microbiota composition varies in an intrinsic manner between healthy individuals (inter-individual variability) and it is well-known that changes in gut microbiota composition and functionality could promote variability in the uptake of plant bioactive compounds through intestinal barrier modulation. Therefore, it would be necessary to explore whether the differences in microbiota composition between healthy subjects could lead to differences in intestinal permeability and whether these normal dissimilarities would be significant enough to endorse variability in xenomiR absorption rate. On the other hand, since gut microbiota dysbiosis could contribute to the onset and progression of a wide range of diseases by affecting intestinal barrier permeability, the link between the alteration of intestinal permeability (as result of gut microbiota specific composition in a pathological context/dysbiosis) and xenomiR bioavailability should also be comprehensively addressed in future researches.

Notably, diet is a key driver in establishing gut microbial communities composition, influencing microbiome diversity and ultimately determining inter-individual microbiome variability (204–207). As part of diet, plant xenomiRs are rising as promising candidates for the modulation of gut microbiota composition, intestinal immune system and intestinal epithelial cells functions, which in turn might affect intestinal barrier permeability and xenomiR absorption efficiency. Nonetheless, further investigation is mandatory to evaluate the potential role of plant xenomiRs on their own absorption and the bioavailability of other bioactive molecules, by shaping microbiota composition, microbe-derived metabolite production, intestinal epithelial cell and immune system function. A schematic representation of the potential interactions and mechanisms of action of plant xenomiRs with mammalian cells within the gastrointestinal tract is depicted in Figure 1.

It has been proposed that a large variety of microenvironments would exist throughout the gastrointestinal tract and gut microbes would thrive in those niches that best suit their nutritional and environmental requirements, leading to the establishment of diverse microbial communities along the gastrointestinal tract (208). Therefore, diet would be a crucial factor in determining variability of gut microbiota composition between individuals and, within an individual, in establishing different configurations of microbial communities throughout the intestine. In this context, recent studies have suggested that plant miRNAs could contribute to the dietary effect on gut microbiota community's assembly. Thus, Teng et al. proposed that ginger exosome-like nanoparticles small RNAs (and maybe other dietary small RNAs) could be involved in the spatial gut microbiota niche partitioning and in the selection of the bacteria that would be located near the intestinal epithelium (125). This finding provides a new avenue for further studying the following hypotheses:

Of note, along with proposed modulation of intestinal permeability mediated by the crosstalk between xenomiRs and gut microbiota, immune system and/or intestinal epithelial cells, other factors that would contribute to determine xenomiR bioavailability are the following:

A summary of the most important hypothesis concerning the effect of dietary plant xenomiRs on gut microbiota, intestinal immune system and gastrointestinal barrier presented within this work are summarized in Figure 2.

In conclusion, although it is well-established that gut microbiota plays an important role in determining the integrity and functionality of intestinal barrier, more studies are needed to verify the direct impact of gut microbiota-intestinal permeability modulation on xenomiR absorption efficiency, as well as the role of xenomiRs shaping this process. A better understanding of the interactions of xenomiRs with gut microbiota, immune system and intestinal epithelial cells, in the context of intestinal barrier modulation, could provide new insights on the mechanisms underlying variability of xenomiR intestinal absorption and emerge as new explanatory factors of the discrepancies regarding the hypothesis of xenomiR detection in animals.

ED-S, SL-C, and FM contributed to the ideas and hypotheses presented in the manuscript. ED-S wrote the paper, which was revised and edited by SL-C, FM, PA, JR-B, and JM. All the authors approved the final version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.