Milk‐Derived Extracellular Vesicles and microRNAs: Potential Modulators of Intestinal Homeostasis

ABSTRACT

Milk represents a complex pool of nutrients and bioactive components that are indispensable for the growth and development of the infant. The key well‐established roles played by milk's bioactive components in the infant are those at the level of the infant's intestinal and immune development. Through its bioactive components, including proteins, lipids, and oligosaccharides, milk helps the infant develop a mature intestinal identity with fully active digestive, absorptive, and barrier capacity and shapes both innate and adaptive immune responses. Recent evidence points to a new class of milk bioactive components including milk extracellular vesicles (EVs) and microRNAs, which are hypothesized to take part in contributing to infant development. The potential functions of milk microRNAs and EVs in the consumer's systems are not limited to the infant as these components can also be found in bovine milk, both raw and processed. Hence, adult consumers could also be influenced by milk EVs and microRNAs, which could affect their intestinal homeostasis particularly under pathological conditions. Nonetheless, the debate regarding the stability of milk EVs and microRNAs in the digestive tract persists, and their bioavailability and bioactivity in the consumer's tissues are still arguable. In this review, we discuss the potential functions mediated by milk microRNAs and EVs in the epithelial and immune components as well as the microbiome of the intestinal mucosa in health and disease. We also discuss the bioavailability, bioaccessibility, and bioactivity of milk EVs and microRNAs in consumer's tissues.

Upon ingestion, milk extracellular vesicles (EVs) carrying bioactive cargo like microRNAs and proteins may survive digestion and interact with intestinal epithelial cells via surface signaling or be internalized by intestinal epithelial cells to which they deliver their bioactive cargo. In either case, milk EVs can trigger signaling changes and alter gene expression through microRNA‐based translation inhibition and mRNA decay. Additionally, milk EVs may cross the intestinal barrier, enter the bloodstream, and distribute systemically, potentially influencing beyond‐intestinal processes in distant tissues. Created with BioRender.com

Abbreviations

ACVR1C

activin A receptor type 1C

ALIX

ALG2‐interacting protein X

AMPK

AMP‐activated protein kinase

AMR

antimicrobial resistance

ATM

ataxia‐telangiectasia mutated

BCL11b

B‐cell CLL/lymphoma 11B

C/EBPδ

CCAAT/enhancer binding protein δ

CD28

cluster of differentiation 28

CD3

cluster of differentiation 3

CD4

cluster of differentiation 4

CD63

cluster of differentiation 63

circRNAs

circular RNAs

CMV

cytomegalovirus

DGRC8

DiGeorge syndrome critical region 8

DNMT‐1

DNA‐methyl transferase 1

DSS

dextran sulfate sodium

E2F5

E2F transcription factor 5

EMT

epithelial‐to‐mesenchymal transition

EVs

extracellular vesicles

FcRn

neonatal Fc receptor

FHC

fetal human colon

FOXP3

forkhead box P3

GALT

gut‐associated lymphoid tissue

GI

gastrointestinal

GO

GeneOntology

HDAC4

histone deacetylase 4

HIECs

human intestinal crypt‐like epithelial cells

HIF‐α

hypoxia‐inducible factor α

HMGA2

high mobility group AT‐hook 2

Hsp70

heat shock protein 70

IBDs

intestinal bowel diseases

IEC‐6

intestinal epithelial cell 6

IECs

intestinal epithelial cells

IGF2BP1 and IGF2BP2

insulin‐like growth factor 2 mRNA binding protein 1 and 2

Igs

immunoglobulins

IL‐17

interleukin 17

IL‐1β

interleukin 1β

IL‐23

interleukin 23

IL‐6

interleukin 6

IL‐8

interleukin 8

IPEC‐J2

intestinal porcine epithelial cells—jejunum 2

KIF9‐AS1

kinesin family member 9 antisense RNA 1

Lgr5

leucine‐rich repeat‐containing G‐protein coupled receptor 5

lncRNAs

long noncoding RNAs

LNPs

lipid nanoparticles

LPS

lipopolysaccharide

MAPK

mitogen‐activated protein kinase

MCF10A

Michigan Cancer Foundation 10A

MCF7

Michigan Cancer Foundation 7

MFGs

milk fat globules

mRNAs

messenger RNAs

mTOR

mechanistic target of rapamycin

mTORC1

mechanistic target of rapamycin complex 1

MUC2

mucin 2

NAFLD

nonalcoholic fatty liver disease

NEC

necrotizing enterocolitis

NF‐κB

nuclear factor κ‐light‐chain‐enhancer of activated B cells

NLRP3

NOD‐like receptor family pyrin domain containing 3

NR6A1

nuclear receptor subfamily 6 group A member 1

PBMCs

peripheral blood mononuclear cells

rRNAs

ribosomal RNAs

sIgA

secretory immunoglobulin A

SIRT1

sirtuin 1

SOCS3

suppressor of cytokine signaling 3

SOX9

SRY‐box transcription factor 9

Sp1

specificity protein 1

TFF‐3

trefoil factor 3

TGF

transforming growth factor

Th17

T helper 17 cells

THP‐1

Tamm‐Horsfall protein 1

TLR4

toll‐like receptor 4

TNBS

2,4,6‐trinitrobenzene sulfonic acid

TNF‐α

tumor necrosis factor α

TRAF6

TNF receptor‐associated factor 6

Treg

regulatory T cells

tRNAs

transfer RNAs

TSG101

tumor susceptibility gene 101

Wnt

wingless‐related integration site

ZEB‐1 and ZEB‐2

zinc finger E‐box binding homeobox 1 and 2

ZO‐1

zonula occluden‐1

α‐SMA

α‐smooth muscle actin

1. Introduction

Milk represents a complex repertoire of nutritional and bioactive molecules that are essential for the growth and development of all newborn mammals. Milk bioactive molecules are those with a beyond‐nutritional value, meaning they may potentially mediate a specific biological activity in the infants or adult consumers. Examples of such milk bioactive molecules include many antimicrobial peptides (e.g., lactoferrin) [1, 2, 3], antioxidant peptides (e.g., casein‐derived peptides) [4], antihypertensive peptides (e.g., β‐casein) [5], immunoglobulins (e.g., IgA) [6, 7], growth factors (e.g., insulin, TGF) [8], and many other nonprotein bioactive molecules [9, 10]. Accordingly, milk directly contributes to specific processes in mammalian newborns. This is mostly known for the passive immunity imparted by the milk via its immune‐related components, especially the wide spectrum of antimicrobial peptides and immunoglobulins acting at the intestinal mucosal interface in the newborn. These are particularly crucial for the infant to fight against pathogens specifically right after exiting the sterile womb environment. Another critical function of milk bioactive constituents is to sustain the maturation and development of the intestinal barrier. After birth, milk is the only food consumed by the newborns up until 6 months (minimum) after which they become able to digest and assimilate other food types. Hence, milk supports the intestinal barrier throughout the developmental steps required for the acquisition of a fully functional adult intestinal identity. Several milk components, including milk fat globules (MFGs), proteins (e.g., lactoferrin), and oligosaccharides, have been proven to promote intestinal cell differentiation and proliferation (via Wnt pathway), tight junction formation (specifically claudins), and attenuation of intestinal cell apoptosis [11, 12, 13, 14], thereby playing a critical role in the intestinal barrier's digestive, absorptive, and permeability functions. In addition to those mentioned above, milk also contains other bioactive molecules, like microRNAs, which could potentially influence certain biological processes in the infants and adult consumers. Many of the bioactive components listed earlier, including microRNAs, exist as free molecules or are associated with milk extracellular vesicles (EVs). EVs are thought to serve as a vehicle to protect and transport these bioactive components in the consumer's systems. In this review, we will shed light on two milk elements, namely milk EVs and milk microRNAs. We will also highlight their potential roles in the functioning and the biology of the intestinal barrier in both infant (human breast milk as a source) and adult (bovine milk as a source) consumers.

2. Milk Extracellular Vesicles (EVs)

2.1. General Definition and Cellular Origin

Extracellular vesicles (EVs) encompass a heterogeneous collection of membranous structures which derive from cells through multiple biogenesis pathways including: (1) direct outward budding from the plasma membrane (releasing what is known as microvesicles or ectosomes); (2) inward budding within endosomes to form multivesicular bodies, which later fuse with the plasma membrane to release intraluminal vesicles (also referred to as exosomes) into the extracellular milieu [15]. Additionally, in a process that is spatially similar to the formation of microvesicles, cells undergoing apoptosis also produce vesicles called apoptotic vesicles or apoptotic bodies via membrane blebbing [16]. The aforementioned microscopic vesicles, ranging in size from 50 nm to 1000 nm, consist of a cytoplasmic component enclosed by a lipid membrane and represent a dynamic system of intercellular communication [17]. EVs mediate cell‐to‐cell communication by directly regulating signaling pathways through interactions with cell surface proteins and by acting as shuttle systems that transfer bioactive cargo, such as microRNAs capable of modulating epigenetic programs [18]. The cargo constituents carried by EVs include lipids, proteins, saccharides, and nucleic acids (DNA fragments, coding, and noncoding RNAs) which either exist within the EV core or are embedded in the EV lipid bilayer [17, 19]. These are secreted by various cell types including cells of the mammary gland [20].

EVs, like the majority of molecules secreted by the cells of the mammary tissue, are released into milk and become one of its components. Based on their protein content, milk EVs are believed to majorly derive from the epithelial cells of the mammary gland [21, 22]. Additionally, global analysis of human milk‐derived EV proteome predicted an immune origin, potentially deriving from mammary tissue resident immune cells [21, 23, 24]. Membranous sac‐like structures were first reported in fresh raw bovine milk in 1973 [25]. Although these structures were not initially recognized as EVs, milk‐derived EVs as we now know them were first isolated from fresh, non‐treated human colostrum and mature milk in 2007 [26]. Then, milk EVs were reported in milk from different other mammals including cows, buffalos, yak, goats, sheep, camels, pigs, pandas, donkeys, rats, and mice [27, 28, 29, 30, 31, 32, 33, 34, 35, 36]. Furthermore, milk EVs were reported in both raw and processed liquid milk as well as in powdered infant formulas [27, 37, 38].

2.2. Milk EV Isolation and Purification Approaches

As it is the case for EVs in other biological fluids, different approaches are employed for the isolation of EVs from milk including size exclusion chromatography, polymer‐based precipitation kits, differential ultracentrifugation, and density gradient [39, 40, 41, 42, 43]. Nevertheless, due to the high protein and fat content of milk, obtaining a pure milk EV preparation seems to be more challenging. Several steps have been optimized and included in the milk EV isolation pipeline in an attempt to reduce potential contaminants. This includes differential centrifugation steps for clearing larger protein complexes and cellular debris, as well as treatment with chemical reagents (e.g., hydrochloric acid, EDTA, and sodium citrate) to either precipitate out or disrupt casein micelles, avoiding their co‐precipitation with milk EVs in later steps [44, 45, 46]. Nonetheless, some of these pre‐isolation steps could influence EV integrity [45, 46]. Though there is no single pipeline that can produce an optimally pure milk EV preparation, integrating multiple isolation techniques, although it could be laborious and time‐demanding, offers the best current option for minimizing contaminants.

2.3. Milk EV Biological Activity

As it is the case for the other milk bioactive components, milk EVs were also proven to influence several processes in both physiological and pathological contexts [47]. Here we will focus on the roles of milk EVs from different species on the epithelial barrier functions of the intestines in health and disease (Figure 1). We will also highlight the immunoregulatory effects of these EVs, particularly on the immune component of the intestinal mucosa (Figure 1).

FIGURE 1.

Schematic illustrating the potential roles of milk microRNAs and milk extracellular vesicles (EVs) in the cells of the intestinal mucosa and commensal bacteria of the intestinal lumen. The left side represents adult intestinal tissue with dense epithelial layers, thick mucus, tight junctions, elongated intestinal epithelial cells (IECs or enterocytes), and mature Peyer's patches. The right side depicts infant‐like tissue with loosely packed intestinal lining, thinner mucus layer, fewer tight junctions, shorter enterocytes, and less developed Peyer's patches. The potential roles of milk EVs and microRNAs are depicted in the callouts: in pink are the functions that have been reported to be mediated by milk EVs, while in blue are the validated (as well as predicted) functions of endogenous microRNAs that are highly enriched in milk. The functions were randomly assigned to each side of the scheme and hence are not specific to one developmental stage (adult vs. infant). For references, refer to sections “2.3”, “3.7”, “4”, and Table 1. Created with BioRender.com.

2.3.1. Roles of Milk EVs in Intestinal Epithelial Homeostasis and Functions

2.3.1.1. In Vitro Studies

Studies on the effects of milk EVs on the biology of the cells of the intestinal epithelium showed that human milk EVs can protect intestinal epithelial cells (IECs) in vitro against H₂O₂‐induced oxidative stress [75]. Though the exact mechanism via which milk EVs reduced cellular toxicity was not investigated, it was shown that this was independent of de novo protein synthesis, indicating the ability of EVs to induce rapid responses in recipient cells [75]. Both yak and cow milk‐derived EVs, which demonstrated differential protein profiles, improved the survival of LPS‐stimulated IEC‐6 cells (LPS is often used to induce experimental injury and trigger inflammation [76, 77]) and reduced inflammation while increasing the expression of tight junction proteins [78]. Remarkably, the effects of yak milk EVs were more profound compared to those of cow milk [78]. The positive survival effects of yak and cow milk EVs were also confirmed in IEC‐6 models under hypoxic conditions [35, 79]. Notably, milk EVs of bovine and murine origins were also reported to enhance cell growth and viability of IECs in vitro [34, 80]. Similarly, porcine milk EVs were shown to induce the proliferation of Intestinal Porcine Epithelial Cells‐Jejunum 2 (IPEC‐J2, an intestinal cell line, often used to study barrier functions [81]) [82]. Another study indicated that bovine milk‐derived EVs increased the rate of apoptosis‐based cell death in a cell line of colon cancer, highlighting a potential antitumor effect [30]. These findings indicate a role of milk EVs in the control of the turnover of intestinal epithelial cells (proliferation‐apoptosis balance) which is an essential aspect for maintaining intestinal homeostasis and ensuring tissue regeneration upon injury or inflammation. Human milk EVs were also proven to influence other cellular aspects including migration and secretory functions. Milk EVs were reported to induce epithelial cell migration, leading to enhanced re‐epithelization of the oral epithelial cells (lining the upper part of the digestive mucosa) in vitro [83]. Functionally, bovine milk EVs promoted mucin secretion and favored goblet cell differentiation [84].

2.3.1.2. In Vivo Studies

At the in vivo level, healthy mice receiving porcine milk‐derived EVs showed higher intestinal proliferative capacity, which was proposed to be mediated by milk EV‐borne miR‐146a‐5p [85]. Similarly, healthy mice receiving a daily oral dose of cow milk‐derived EVs showed alterations in their microbiome communities and an upregulation of Mucin 2 (MUC2), IgAs, and secretory IgAs (sIgAs) in their intestinal mucosa compared to vehicle‐fed mice [86]. At the histological level, porcine milk‐derived EVs increased the intestinal volume (villus height and crypt depth) in healthy mouse intestines [82]. The above examples highlight an interspecies transfer of regulatory molecules through food ingestion and the ability of milk EVs to mediate cellular phenotypic changes in a cross‐species context.

While these findings highlight the multifaceted role of milk EVs in modulating epithelial cell function, milk EVs were also proven to regulate multiple immune processes, including those in the intestinal mucosa.

2.3.2. Milk EVs as Modulators of Immunity‐Related Processes

In addition to their roles in epithelial intestinal fitness discussed earlier, milk EVs have immunomodulatory functions and can act on several immune cells; they could also modulate the innate immunity of the intestinal epithelial cells. Although these examples pertain to systemic immunity, it is believed that milk EVs could act in a similar way on the immune cells of the gut‐associated lymphoid tissue (GALT). Examples of how milk EVs could regulate the innate immunity of intestinal epithelial cells as well as intestinal inflammation are also discussed.

2.3.2.1. In Vitro Studies

Several studies on milk EVs have reported their enrichment in bioactive components with immunomodulatory activities. For instance, commercial bovine milk EVs were shown to induce the differentiation of T cells into Th17 cells via TGF‐β expressed on their surface [87]. Furthermore, bovine and human milk EVs were shown to trigger cytokine release by activated peripheral blood mononuclear cells (PBMCs), induce lymphocyte differentiation into regulatory FOXP3+ cells (Forkhead box P3 (FOXP3)‐positive regulatory T cells (Treg cells), commonly recognized for their anti‐inflammatory and immunosuppressive functions [88]), and augment the proliferation of human macrophages under hypoxic conditions [26, 89]. Another study investigating a direct effect of milk‐derived EVs on immune cells indicated that EVs can counteract CD4+ T cell activation (induced by anti‐CD3/CD28 antibody) and their subsequent differentiation into memory cells [83]. Further work reported the regulation of macrophage phenotype by goat milk, whereby goat milk‐derived EVs triggered the differentiation of macrophages into a pro‐inflammatory M1 phenotype [90]. Additionally, similar to other milk antimicrobial components, EVs derived from bovine colostrum also acquire anti‐inflammatory and antimicrobial activities and could inhibit bacterial adhesion to IPEC‐J2 intestinal cells in vitro [91]. Human milk EVs were also shown to reduce inflammation in intestinal organoids upon experimental injury (LPS‐induced) by reducing the release of pro‐inflammatory cytokines and inhibiting toll‐like receptor 4 (TLR4) expression [92].

At the level of the innate immunity of intestinal epithelial cells, EVs from goat milk were also shown to regulate the innate immune processes in LPS‐ and H₂O₂‐stimulated IPEC‐J2 cells via the regulation of several TLR and cytokine levels, also favoring the release of several anti‐inflammatory mediators [93, 94]. Similar immunomodulatory functions were reported for cow milk‐derived EVs, which favored anti‐inflammatory cytokine secretion by stimulated Caco‐2 (transformed human colorectal adenocarcinoma intestinal epithelial cell line, acquire enterocyte‐like features when differentiated [95]) and THP‐1 (human acute monocytic leukemia cell line, commonly used as a model for monocytes and macrophages [96]) cells in a co‐culture set‐up [97].

2.3.2.2. In Vivo Studies

Goat milk‐derived exosomes were shown to protect the intestinal villi from excessive LPS‐induced damage and breakage in mice [98]. The orally administered EVs were proven to reduce inflammation by acting on the TLR4/NF‐κB pathway [98]. Notably, goat colostrum EVs had a more profound protective and anti‐inflammatory effect compared to EVs derived from mature goat milk [98]. Orally administered milk EVs from porcine origin also demonstrated an anti‐viral capacity as they reduced the replication and propagation of the porcine epidemic diarrhea virus and reduced infection symptoms in piglets' intestines [99]. Of note, the antiviral effect was mainly attributed to the microRNA content of these EVs, particularly miR‐27b and let‐7e [99].

2.3.2.3. In Silico Studies

Besides these in vitro and in vivo validated functions, in silico analysis of the milk EV proteome allowed the prediction of several functions in immune modulation. These studies indicated the enrichment of immune regulatory molecules and immune cell‐related Gene Ontology (GO) terms in milk EV pellets obtained from both bovine and human milk [22, 23, 100]. Furthermore, metabolomic analysis of milk EVs derived from goats, cows, and donkeys revealed a significant enrichment of various metabolites, including amino acids and vitamins, in the EV fraction compared to non‐EV fractions [31]. Notably, many of these metabolites were identified as key immune modulators with distinct anti‐inflammatory properties [31].

The above functions further stress the beneficial effects of milk EVs on the maturation and development of the immune components, particularly in the intestinal mucosa.

As mentioned in the introduction, milk is widely studied for its immune functions mediated by several proteins, lipids, and saccharides with immunomodulatory and anti‐microbial properties. The above‐reported immune‐related functions of milk EVs could thus complement the already established immunomodulatory, anti‐inflammatory, and antimicrobial activity of these other milk bioactive components. They could also act in synergy or modulate each other's signaling, potentially amplifying or attenuating immune responses. It would be thus valuable to assess such crosstalk in future studies.

2.3.2.4. Pro‐ Versus Anti‐Inflammatory Effects of Milk EVs: Context and Complexity?

As mentioned earlier, the immunomodulatory functions of milk EVs often appear to tilt the balance toward an anti‐inflammatory environment, suppressing pro‐inflammatory responses. This is critical for reducing the chronic inflammation that is a hallmark of intestinal bowel diseases. However, there are few examples where milk EVs induced a rather pro‐inflammatory response, especially in their action on macrophages triggering their transition to the pro‐inflammatory M1 phenotype [90]. This was also seen in non‐intestinal diseases (respiratory tract‐related) where the oral ingestion of milk EVs increased inflammation by a similar mechanism (favoring an M1 macrophage phenotype), suggesting that under certain conditions these EVs may exacerbate inflammation [101]. These contradictory findings highlight the need for caution in generalizing the immunomodulatory role of milk EVs and their effect on inflammation. Such contrasting results in milk EV outcomes on inflammatory responses might be context‐dependent (cell type, health status, or disease model) but might also stem from the differences in the EV preparations used in each study, given the variations in isolation protocols, which likely influence the EV populations obtained, and enrich for one population relative to the other. As highlighted above, milk EVs constitute heterogeneous populations (with differences in their protein profiles and microRNA content) [22, 102, 103, 104, 105]. This heterogeneity in EVs could also reflect a heterogeneity in their physiological outcomes including their effect on inflammation, as has been seen with the 100K and 35K cow milk EV subtypes, which resulted in different outcomes when administered to mice [102]. Moreover, the inherent molecular complexity of EVs—encompassing diverse lipid, RNA, protein, and surface marker compositions—further complicates their biological interpretation.

2.3.3. Milk EV Functions in Pathological Contexts

Interestingly, studies demonstrating the biological activity of milk EVs in pathological conditions were more prevalent than those in a normal physiological context. Given that breast‐milk fed infants are less vulnerable to necrotizing enterocolitis (NEC) compared to formula‐fed infants [106], and that milk EVs are believed to be bioactive especially at the intestinal mucosal interface, the biological activity of milk EVs was extensively studied in the context of experimentally induced colitis and NEC.

2.3.3.1. Necrotizing Enterocolitis (NEC)

In a rat model of experimental NEC, human breast milk‐derived EVs offered protection against NEC by significantly reducing disease incidence compared to controls upon oral and intraperitoneal EV administration [107]. Additionally, the same study showed that milk EVs mediated pro‐proliferative and antiapoptotic functions in IEC‐6 cells in vitro [107]. On a similar note, human milk EVs were also shown to attenuate the symptoms of NEC in a rat model and to enhance the proliferation and migration of LPS‐induced intestinal cells in vitro [108]. Bovine milk EVs promoted mucin secretion and favored goblet cell differentiation in vivo, mitigating NEC intestinal injury in mice [84]. Intraperitoneally administered bovine milk EVs also reduced intestinal inflammation (via the NF‐κB pathway) and promoted tight junction expression in NEC mice models [109]. In this study, and using microRNA mimics, the authors attributed these effects to miR‐148a‐3p, which was found to be highly enriched in bovine milk EVs [109].

2.3.3.2. Ulcerative Colitis

Studies on the effect of milk‐derived EVs, especially those of bovine origin, on the progression of experimentally‐induced colitis were comprehensive in that they analyzed the effects of milk EVs on multiple disease aspects. For example, using metabolomic, metagenomic, and transcriptomic analyses, it was shown that cow colostrum‐derived EVs decreased the severity of DSS‐induced (DSS or dextran sodium sulfate induces severe intestinal injury and inflammation and is thus often used to establish experimental colitis models [110]) colitis symptoms by acting on different disease components [111]. The authors showed that orally‐administered EVs can restore mucus levels, reduce the expression of the pro‐inflammatory cytokines (IL‐6 and TNF‐α), and increase the abundance of beneficial bacteria (Akkermansia) and their metabolites (hydroxybutyrate) [111]. Additionally, another report showed that orally‐administered cow milk‐derived EVs could alleviate DSS‐induced colitis symptoms at the level of the three barriers of the intestinal mucosa: epithelial, immune, and mucus [112]. The authors showed that milk EVs upregulated the expression of the epithelial tight junctions, suppressed pro‐inflammatory cytokine release, expanded macrophage and regulatory T cell (Treg) populations, and upregulated mucins, eventually reducing DSS‐induced colonic disintegration [112]. In another model of DSS‐induced colitis, milk EVs were able to mitigate colitis symptoms at the histological level, inhibit pro‐inflammatory cytokine release, augment the expression of both proliferation and migration factors, and enhance tight junction expression [102]. In line with this and using a similar colitis model, findings from another study demonstrated a similar attenuation of colonic inflammation upon gavage administration of milk EVs [113]. It is of relevance to note that both studies mentioned a role of EVs in regulating the NF‐κB pathway to mediate their positive effects on colitis progression [102, 113]. Additionally, cow colostrum‐derived EVs resulted in a significant reduction of the apoptotic rates in the intestines of mice with DSS‐induced colitis and a marked downregulation of the pro‐inflammatory cytokine IL‐10 [114]. Similarly, DSS‐induced inflammation in mice intestinal tissues was attenuated by the oral administration of porcine milk‐derived EVs [115]. Presumably through their let‐7c microRNA content, these EVs shifted the polarization of intestinal macrophages from the pro‐inflammatory M1 phenotype to the anti‐inflammatory M2 phenotype and augmented the expression of the Claudin 1 tight junction protein [115].

2.3.3.3. Other Pathologies

Besides those mentioned above in the context of the common intestinal pathologies, ulcerative colitis and NEC, crucial roles of milk EVs were documented in other disease conditions. One remarkable activity of milk EVs is the restoration of the permeability of the intestinal epithelial barrier in a mouse model of malnutrition [116]. Bovine milk EVs were able to reinstate the expression of Claudin‐3 tight junctions, trigger the proliferation of crypt stem cells (Lgr5+ cells), and sustain Wnt pathway activation [116].

Aside from their functions in intestinal pathologies, EVs were also shown to influence disease phenotypes in other tissues. For instance, human breast milk‐derived EVs improved the proliferation and attenuated apoptosis by targeting the IL‐17/caspase‐3 signaling pathway in the cells of the alveolar epithelium in a rat model of bronchopulmonary disease [117]. The latter is a condition commonly reported in premature infants. In another study on cancer, bovine milk‐derived EVs reduced tumor burden, but strikingly enough, increased both pancreatic cancer liver metastasis as well as breast cancer lung metastasis [118]. The opposing effects of milk EVs could be contributed to their complex cargo and membrane constituents, which could influence distinct pathways within the same cellular system.

Beyond offering a literature review, the compilation of these studies serves to illustrate that independent investigations using various models consistently reported the EV‐dependent regulation of key aspects in the intestines, like promoting an anti‐inflammatory environment and reinforcing the intestinal barrier. Through regulating intestinal tight junctions, proliferation, apoptosis, and key immune processes of the GALT, milk EVs could play a crucial role in shaping a mature intestinal identity during infant development. Additionally, as reviewed earlier, these same functions could also help in the prevention and protection against intestinal diseases in infant and adult consumers equally.

2.4. Milk EVs in Therapy

The use of milk EVs as molecules with therapeutic potential, especially for their immunomodulatory and inflammatory functions, or as vehicles for the delivery of therapeutic molecules has been already addressed [113, 116, 117, 118, 119, 120]. The biocompatibility and biodistribution of milk EVs or chemically modified milk EVs have been explored in vitro and in vivo upon their administration via multiple routes [117, 121, 122, 123]. Furthermore, studies have indicated superior drug delivery features of EVs compared to controls [124, 125, 126]. Nonetheless, when considering milk EVs as vehicles for the delivery of therapeutic molecules, siRNAs in particular, rigorous and in‐depth testing is required to anticipate any potential off‐target concerns that might arise, especially as the tissue homing profiles of milk EVs are still not clearly understood. This, once again, highlights the necessity for more standardized EV isolation protocols and more thorough characterization of EVs before their use in therapeutic applications.

2.5. Milk EVs: A Heterogeneous Population

2.5.1. Different EV Subtypes, Differential Stability?

Most of the studies on milk EVs listed earlier focus on EVs pelleted at high speeds (≥ 100, 000 × g) with or without prior steps for the elimination of larger and denser vesicles. Thus, EV preparations in some of these studies might entail heterogeneous EV subpopulations, each with a specific assortment of surface proteins, microRNAs, lipids, and sugars. The heterogeneity in milk EV populations was first reported in commercial cow milk [104]. This and other studies reported the presence of multiple milk EV subsets (those pelleting at 12,000 × g, 35,000 × g, and 70,000 × g; referred to as 12K, 35K, and 70K EVs, respectively) other than the often‐studied milk small EVs (sEVs) or ≥ 100,000 × g EVs (100K, also referred to as exosomes) [46, 103, 105]. The newly characterized subsets appeared to be heavier and contained higher amounts of microRNAs compared to the often‐studied milk sEVs [103, 127]. They also showed differential enrichment of the common EV markers (ALIX, TSG101, Hsp70, and CD63) [103]. It was also suggested that these subpopulations may exhibit additional heterogeneity and could be further subdivided [103, 127]. The physicochemical properties of EVs, including their surface lipid, sugar, and protein content, could potentially affect their stability against digestion as well as their uptake by intestinal cells. In a report on commercial cow milk, it was noted that 100K EVs were more resilient to in vitro digestion than the 12K and 35K subsets (see next) [127]. This was also corroborated by a recent study showing that human milk‐derived EVs which are > 200 nm in size are more sensitive to digestion‐based degradation compared to the smaller EVs (< 200 nm) [112].

2.5.2. Different EV PTMs, Differential Uptake Efficiencies, and Functions?

Though not much is known about the mechanisms mediating milk EV uptake by intestinal cells, few reports showed that intestinal cells might take up EVs through receptors like Galectin‐3 (member of the lectin family of carbohydrate‐binding proteins), neonatal Fc receptor (FcRn), or Class A scavenger receptor‐1/2 (macrophages) [121, 128, 129]. Reports also showed that the addition of a carbohydrate moiety can significantly reduce the uptake of milk EVs by intestinal cells in vitro, indicating the involvement of a glycosylated actor in the EV uptake [130, 131]. This might imply that EV subpopulations more enriched in immunoglobulins (ligands of FcRn, particularly IgAs and IgMs) and Galectin‐3 binding proteins (including glycosylated proteins) could potentially be more bioaccessible. A comparison of milk EV proteomic data from different laboratories revealed the consistent presence of the following protein components: distinct immunoglobulin heavy chains, immunoglobulin J chain, and Galectin‐3 binding protein [26, 105, 121, 132, 133, 134, 135, 136]. Although not statistically significant, the 100K subset described earlier showed greater enrichment in Galectin‐3 binding protein than the 35K subset (supplementary data from [22]). Moreover, predicted PTMs of proteins specific to the 35K (20 proteins) and 100K (41 proteins) fractions were thoroughly analyzed using the UniProt consortium and the raw data from [22]. The analysis showed that the majority of the 100K‐specific proteins were glycosylated (either the bovine, the human homologue, or both), while none of those which are 35K‐specific were glycosylated [22]. On a similar note, a study on the glycome profile of raw bovine milk‐derived EVs showed a marked enrichment of glycopeptides and glycoproteins [137]. Additionally, a more recent study investigating the glycome of milk EVs from four different species (human, caprine, bovine, and porcine) reported similar results and indicated species‐specific glycosylation patterns [138]. Of note, the EVs which were analyzed in this study were those pelleting at high speeds (small EVs pelleted at 135, 000 × g, sEVs, not including the 35, 000 × g fraction) [138], which is consistent with the enriched glycosylation predicted in the 100K‐specific proteins from commercial milk EVs [22].

Given the above differences between milk EV subsets, it is plausible that each subpopulation may influence the intestinal biology differently. It was shown that the 100K and 35K subpopulations manifested different biological effects in a model of DSS‐induced colitis, where the 100K subset, but not the 35K, influenced Zonula Occluden‐1 (ZO‐1, tight junction protein which plays central role in sealing space between intestinal epithelial cells creating tightly‐packed intestinal barrier [139]) expression and microRNA intestinal levels, while the 35K subset influenced more the innate immunity [102]. Additionally, given their distinct microRNA profiles [103, 104], differences in the microRNA‐mediated functions between milk EV populations are to be anticipated. In case smaller milk EVs actually have a higher chance of survivability and bioactivity after milk ingestion, this means that the microRNA profile detected in whole milk might not predict the effects on the consumer, as they might be majorly driven by those enriched in the small EVs.

3. Milk microRNAs

3.1. MicroRNAs: A General Definition

Different RNA species, both coding and noncoding, were reported in the milk samples from different species and of different sources. The different RNA biotypes that were reported in milk, of both human and nonhuman origins, include messenger RNAs (mRNAs) [140], long noncoding RNAs (lncRNAs) [141, 142], circular RNAs (circRNAs) [143, 144, 145], microRNAs [79, 146, 147], ribosomal RNA fragments (rRNAs) [148, 149], and transfer RNA fragments (tRNAs) [150, 151]. Though not the most abundant, microRNAs are the most commonly studied ncRNAs in milk. MicroRNAs are defined as small noncoding RNA species (~19–24 nucleotides) that regulate gene expression at the post‐transcriptional level and are estimated to regulate up to 60% of human genes [152]. MicroRNAs influence gene expression by binding to the microRNA‐response elements in the 3'UTR of the target mRNA [153]. Thereafter, this interaction may result in either the inhibition of protein translation and/or mRNA destabilization and eventual decay [154]. The presence of other noncoding RNAs, such as lncRNAs, in milk introduces an additional regulatory dimension to the activity of microRNAs [155]. Specific milk microRNA species could be sponged out by their cognate lncRNAs, which alters the functionally active pool of these microRNAs relative to other microRNAs. This interaction could influence the relative availability of specific microRNAs in milk, potentially affecting their downstream regulatory roles.

3.2. Cellular Origin of Milk microRNAs

MicroRNAs were abundantly reported in all three milk fractions: fat, cells, and whey (Figure 2). Moreover, the levels and profiles of microRNAs vary between fractions where the cellular fraction records the highest microRNA levels (Figure 2) [156, 157]. The primary cellular source of microRNAs in milk is predicted to be the epithelial mammary cells (lactocytes) of the mammary gland [103, 158]. Nonetheless, milk microRNAs are also predicted to derive from maternal circulation [158, 159]. Given that milk whey could be further fractioned by ultracentrifugation into (1) protein‐ or membrane‐rich fractions (sedimenting fractions, usually rich in EVs) and (2) milk serum (supernatant, usually depleted or poor in EVs), the profile of microRNAs was also compared between these two fractions and found to be more enriched in the EV‐rich fractions compared to milk serum [104]. In fact, the encapsulation of microRNAs within EVs or lipid droplets (e.g., MFGs) as well as their association with high density protein complexes might in part explain their survivability in milk, which represents a milieu rich in RNases and other enzymes with RNase‐like activity (e.g., abzymes) [2, 160, 161]. The observed variability in microRNA profiles across milk fractions raises intriguing questions about their functional specialization and selective packaging mechanisms.

FIGURE 2.

Scheme illustrating the different fractions of milk separated by low‐speed centrifugation (2000 × g–5000 × g). Each of the three fractions, fat, whey, and cellular pellet, contains specific components, though microRNAs could be found in all three fractions in various forms and associated with distinct elements. The depicted icons are illustrative and are not necessarily all to scale. Created with BioRender.com.

3.3. Detection in Processed Milk Products and Infant Formulas

Despite the degradative conditions associated with industrial milk processing and treatment, milk microRNAs were detected in processed dairy products like pasteurized commercial bovine milk [103], fermented milk products (cheese) [162], and pasteurized human milk (human milk banks) [163]. Nonetheless, a couple of studies have reported a negative effect of milk processing on the milk microRNA content. A comparison of the microRNA content in human milk before and after Holder pasteurization (heat‐based pasteurization, 62.5° for 30 min) indicated a significant loss of total microRNA content in whole milk and milk exosomes [163]. A similar observation was reported with bovine milk when miR‐200c and miR‐29b levels were investigated [164]. This is also similar to the effect of pasteurization on other milk bioactive components [165]. Nonetheless, some papers indicated no significant effect of pasteurization on total RNA content [42]. Regarding the biological activity of microRNAs following pasteurization, it was shown that the latter treatment reduced the immune activity of milk microRNAs by influencing their ability to induce cytokine secretion by THP‐1 cells [166]. Remarkably, microRNAs were also detected in infant milk formulas. However, their levels were either very close to detection limits or were significantly reduced compared to liquid milk (both raw and processed) [38, 158, 167, 168, 169]. The ability of milk microRNAs to remain stable following milk treatment might be attributed to their association with protein complexes and milk EVs.

3.4. Milk EV‐Associated microRNAs

As mentioned earlier, EVs purified from milk are often associated with noncoding RNAs, including microRNAs. The profile of microRNAs in milk EVs of human and bovine origins has been well‐documented [170, 171, 172, 173, 174]. Milk microRNAs are not equally enriched in the different subtypes of milk EVs [103, 104]. Importantly, the precise localization (whether present within the EV core, embedded in its surface, or sequestered between EV aggregates) of the majority of microRNAs in milk EVs is yet uncertain. One report assessed the resistance of microRNAs to RNase A before and after EV treatment with proteinase K and showed that some microRNAs seem to be more protected than others, which might reflect a difference in their localization within the EV [104]. The localization of microRNAs in EVs is critical as it can determine their ability to resist degradation and will affect their activity. Of note, some EV‐membrane‐embedded microRNAs do not need to transfer to the cellular cytoplasm; rather, they trigger specific cellular signaling pathways via direct interaction with cell surface molecules [175]. Though no reports, to our knowledge, clearly distinguish EV‐associated milk microRNAs from non‐EV‐associated milk microRNAs, the latter could potentially exist in EV‐free forms. They could be potentially associated with protein complexes or non‐EV lipid particles (like high density lipoproteins, HDL) as it is the case for microRNAs in other body fluids like plasma [176, 177]. These non‐EV‐associated microRNAs may differ in their stability, resistance to digestion, and mechanisms of cellular uptake compared to EV‐encapsulated microRNAs. This distinction is important when interpreting studies on the oral transfer of dietary microRNAs, as the bioavailability and bioactivity of milk microRNAs may depend on their mode of packaging.

3.5. Highly Abundant Milk microRNA Species

Remarkably, the same 10 microRNAs were almost consistently reported by different studies to be the most highly enriched in whole milk as well as milk EVs regardless of the species source, milk maturity (colostrum vs. mature), or milk fraction [178]. In some studies, however, some of these microRNAs were not in the top ten; they were still reported as relatively enriched microRNAs [170, 179]. These microRNAs include miR‐146b, miR‐148a‐3p, let‐7 (either family member, but mainly let‐7a, and let‐7b), miR‐200 (miR‐200a, miR‐200b, and miR‐200c), miR‐30 (miR‐30a, miR‐30b, and miR‐30d), miR‐181 (miR‐181b and miR‐181d), miR‐22, miR‐21, miR‐320, and miR‐26a [103, 147, 149, 151, 157, 163, 167, 171, 180, 181]. Note that certain studies did not specify whether it was the 3p or 5p arm of the microRNA that was enriched; hence, we preferred to list these microRNAs as they were reported in the paper. Of note, studies on these microRNAs in different biological contexts have well‐documented their roles in many cellular and molecular events [7, 182] (also see sections below and Table 1). MicroRNA enrichment and conservation throughout species might as well indicate a potentially conserved role of these microRNAs, which might be manifested in milk consumers.

TABLE 1.

A list of the functions reported to be mediated by a few endogenous microRNAs that are highly enriched in milk.

microRNA or microRNA family	References	Experimental setup and findings	Target gene or pathway
let‐7 family	[48]	let‐7 family repression induced adenocarcinoma development and favored a stem‐like phenotype of intestinal cells.	Hmga2
	[49]	let‐7 family repression in mice intestines induced adenocarcinoma emergence, supported the expansion of the stem cell population, diminished Paneth cell population.	Hmga2, Igf2bp1, Igf2bp2, E2f5, Acvr1c, and Nr6a1
	[50]	let‐7b‐5p‐low exosomes induced the expression of pro‐inflammatory cytokines by macrophages in vitro and aggravated colitis in vivo. let‐7b‐5p mimic‐encapsulated exosomes alleviated colitis in mice.	TLR4/NF‐κB pathway
	[51]	let‐7b‐knockout mice demonstrated lower expression of the intestinal tight junctions. let‐7b overexpression in NCM620 upregulated the expression of Occludin.	p38/MAPK pathway
miR‐200 family	[52]	miR‐200b‐rich extracellular vesicles reduced the expression of the mesenchymal markers vimentin and α‐SMA and augmented E‐cadherin expression in vitro and in mice with TNBS‐induced colitis.	ZEB1 and ZEB2
	[53]	miR‐200 antagomiR administration into mice with intestinal dysfunction reduced the expression of intestinal tight junctions.	SIRT1
	[54]	miR‐200‐rich LNPs reduced injury‐induced p53 activation, intestinal stem cell loss, and enhanced intestinal regeneration in mice.	p53
	[55]	miR‐200 upregulation in SPF mice diminished the expression of IL‐2 induced by intestinal lamina propria leukocytes [55].	BCL11b and ETS‐1
	[56]	miR‐200c‐3p antagomiR oral administration into mice with colitis upregulated occludin expression.
miR‐22‐3p	[57]	miR‐22 suppressed the proliferation, migration and invasion of colorectal cancer cell lines (SW620 and LoVo) and suppressed tumor growth and metastasis of colorectal cancer in vivo.	Sp1
	[58]	miR‐22‐3p mimic transfection induced proliferation of human intestinal crypt‐like epithelial cells (HIECs) in vitro.	C/EBPδ
	[59]	miR‐22‐3p mimic transfection positively regulated proliferation while inhibiting apoptosis in fetal human colon (FHC) cells in vitro following LPS‐induced injury. Reduced the expression of TNF‐α and IL‐1β.	NLRP3
	[60]	miR‐22‐3p mimic transfection reduced cellular apoptosis and pro‐inflammatory cytokine (IL‐6, IL‐8, TNF‐α, and IL‐1β) secretion by LPS‐treated IPEC‐J2 cells in vitro.	ATM/p53 and NF‐κB
	[61]	Overexpression of miR‐22‐3p promoted proliferation and reduced pro‐inflammatory cytokine (IL‐6, IL‐8, TNF‐α, and IL‐1β) in LPS‐treated IPEC‐J2 cells in vitro.	MAPK14
[62]	Antisense miR‐22 decreased pro‐inflammatory cytokine (IL‐6 and TNF‐α) release in intestinal mucosa and a decrease in Th17 cell count in mice with TNBS‐induced colitis.	HDAC4
miR‐148a‐3p	[63]	miR‐148a‐3p overexpression in HT‐29 cells abolished the elevated apoptosis rate induced by DSS treatment.	KIF9‐AS1 lncRNA and SOCS3
	[64]	miR‐148a overexpression in Caco‐2 cells augmented the expression of ZO‐1.	DNMT‐1
	[65]	miR‐148a‐3p ablation in mice exacerbated DSS‐induced colitis symptoms and subsequent colorectal cancer‐induced colitis. Restoration of miR‐148a‐3p levels reduced pro‐inflammatory cytokine release and tumor development.
miR‐181	[66]	Induced the proliferation of intestinal epithelial cells from DSS‐induced colitic mice.	Wnt pathway
miR‐30 family	[67]	miR‐30 overexpression induced the proliferation and inhibited the differentiation of Caco‐2 cells and intestinal epithelial cells from mouse intestines (ex vivo).	Sox9
	[68]	Oral administration of miR‐30d mimics induced growth of commensal bacteria in vivo. Promoted regulatory T cell differentiation.
miR‐515‐5p	[69, 70]	miR‐515‐5p mimics induced the growth of commensals in vitro.
miR‐21‐5p	[71]	miR‐21‐5p knockout mice demonstrated lower fecal load of Lactobacillus spp. miR‐21‐5p mimics inhibited the growth of commensal Lactobacillus spp. in vitro.
miR‐155‐5p	[72]	miR‐155 antagomiR elevated tight junction expression and enhanced intestinal barrier function in mice with DSS‐induced colitis.	HIF‐1α and TFF‐3
	[73]	miR‐155 inhibition in septic mice improved intestinal barrier function and reduced inflammation. miR‐155 mimic transfection into NCM460 cells in vitro reduced tight junction expression and increased pro‐inflammatory cytokine (TNF‐α and IL‐6) release.
miR‐223	[74]	miR‐223 antagomiR upregulated Claudin‐8 expression in colitic mice.	IL‐23/Th17

3.6. Milk microRNAs: A Dynamic Profile

The microRNA profile in milk is influenced by multiple factors, many of which are related to mammary gland conditions, pregnancy‐related factors, and overall maternal health status.

3.6.1. Pregnancy and Lactation‐Related Factors

The microRNA profile may vary across different lactational stages. MiRNome analyses highlighted differences in the levels of certain microRNAs between the whey fraction of human colostrum (milk expressed during the first 7 days post‐delivery) and that of the mature milk (milk expressed > 2 weeks post‐delivery, Figure 3A) [147]. Findings by other studies on colostrum and mature milk samples (whole and whey fractions) indicated the presence of colostrum‐specific and other mature milk‐specific microRNAs [167, 183, 184]. The colostrum‐specific microRNAs were predicted to have targets that are known actors in key metabolic pathways like mTOR and AMPK [167, 184]. Studies also showed an enrichment of immunomodulatory and development‐related microRNAs (e.g., miR‐155 and miR‐223) in bovine colostrum compared to mature milk (whey fraction, Figure 3B) [167, 168]. Additionally, specific immune‐related microRNAs (like miR‐155 and miR‐181a) were less abundant in the milk whey collected after 6 months following delivery compared to that collected before 6 months, indicating that microARNs with immunomodulatory functions are more enriched in the milk in the first months after delivery [185]. This coincides with the period during which the infant needs the most protection against external pathogens.

FIGURE 3.

Scheme showing the dynamic profiles of milk microRNAs across different conditions and in different milk samples. The listed microRNAs were selected from the lists of the differentially expressed microRNAs (DEMs) between two comparative states. The scheme was based on data collected from the main texts as well as supplementary data of the following studies [148, 167, 170, 179, 180, 186]. Created with BioRender.com. and GraphPad Prism version 10.4.1.

Similar differential microRNA enrichment was reported in the milk and milk EVs derived from mothers who delivered preterm babies compared to those who delivered term babies [187]. As for the functions that were overrepresented in preterm milk compared to term milk, these were related to specific metabolic processes, more specifically lipid biosynthesis (Figure 3C) [179]. Findings from another study indicated that milk from mothers who delivered prematurely was more enriched in neurodevelopment‐associated microRNAs than term milk [188]. These examples prompt the question about the mechanisms behind such variations in milk microRNA profiles: are they being tailored to the infant's specific needs (term vs. preterm, colostrum‐fed vs. mature milk‐fed), or are they primarily altered by maternal hormonal and health status?

3.6.2. Maternal Health Status‐Associated Factors

The influence of maternal metabolic diseases, particularly obesity and diabetes, on the microRNA profiles was investigated by multiple studies and, importantly, similar correlations were consistently reported for certain microRNAs. As per the influence of mother's weight, milk and milk EVs from obese mothers showed alterations in the levels of certain microRNAs (e.g., miR‐148a, miR‐30b, and miR‐146b) which are known to act in certain metabolic axes, including but not limited to nutrient sensing and adipogenesis (Figure 3D) [170, 186, 189, 190]. Additionally, compared to healthy controls, milk‐derived EVs from mothers with type 1 diabetes demonstrated a significant upregulation of nine microRNAs, many of which were implicated in immunity and cell biology (cell cycle and autophagy, Figure 3E) [180]. Other maternal‐related factors like stress experienced before and during pregnancy, maternal diet, and milk removal by the baby (pre‐ vs. post‐nursing, Figure 3F) were also reported to have an influence on the abundance of certain milk microRNAs [148, 191, 192]. The above examples indicate that microRNA abundance in milk seems to be highly dynamic and might be reflected by differences in their potential functions in the infant and hence might shape the latter's development.

3.7. Roles of Milk‐Borne microRNAs in the Intestinal Mucosa

Unlike milk EVs, the study of the biological activity of a specific milk‐borne microRNA, including those which are EV‐associated, is challenging owing to the complexity of the milk and milk EV components. In studies where a specific microRNA is differentially enriched between different milk samples, any differential biological outcome can not be causally attributed to that microRNA. Especially that these studies often do not account for variations in other milk bioactive components or even other microRNAs. Below are a few studies in which the biological activity of orally ingested milk microRNAs in the intestinal mucosa was reported. We also highlight in vitro studies in which intestinal cells were exposed to exogenous microRNAs that happen to be highly enriched in milk.

3.7.1. In Vivo Studies

A study on sheep milk indicated that the intestinal tissues of mice with DSS‐induced colitis showed elevated levels of miR‐148a and let‐7b following the oral administration of sheep milk‐derived EVs [193]. This increase was associated with colitis attenuation via direct targeting of TLR4 and TRAF1 by miR‐148a and let‐7b, respectively [193]. It is worth noting that sheep miR‐148a (oar‐miR‐148a) shares an identical sequence with that of the human (has‐miR‐148a‐3p) [194], and hence could potentially exert the same functions in human consumers. Nonetheless, and as mentioned earlier, though TLR4 and TRAF1 are validated targets of miR‐148a and let‐7b, respectively, the observed functions could not be attributed solely and exclusively to these two microRNAs, as the authors did not monitor the levels of other microRNAs. Nonetheless, milk microRNAs reaching the intestines at levels high enough and as intact functional molecules could assume the same functions as those mediated by endogenous microRNAs.

Remarkably, as mentioned earlier, many of the milk highly enriched microRNAs have well‐documented functions in the intestines (refer to Table 1 for the well‐established functions of these endogenously expressed microRNAs).

For example, let‐7 family members have profound roles in sustaining barrier integrity and intestinal mucosa immune homeostasis, including regulation of tight junction expression and the negative regulation of pro‐inflammatory cytokine secretion [48, 49, 51, 195] (Figure 1). As regards miR‐148a, it was reported to be associated with regulation of intestinal tight junction expression, regulation of intestinal macrophage inflammatory functions, and reduction of intestinal cell death [63, 64, 65].

An independent report also provided evidence indicating that milk may be an important source of microRNAs besides those which are endogenously expressed [196]. Mice receiving a milk exosome‐ and microRNA‐depleted diet had lower miR‐200a‐3p levels in their ceca and liver, along with more severe cecal intestinal bowel disease (IBD) symptoms, compared to those receiving a milk exosome‐ and microRNA‐sufficient diet [196]. Breastfeeding also enriched intestinal tissues with milk microRNAs whereby breast‐fed mice pups showed increased levels of miR‐375‐5p compared to formula‐fed mice pups [197]. Though the effect of this elevation was not investigated in mice tissues, miR‐375‐5p was proven to regulate intestinal cell proliferation and apoptosis in vitro [197].

The transfer of milk microRNAs, especially those associated with EVs, upon oral ingestion is similar to the use of lipid nanoparticles (LNPs) for the oral transfer of RNA therapeutics. Exogenous microRNAs packaged in LNPs were reported to transfer and enrich the intestinal mucosa with functional microRNAs (miR‐146b and miR‐200b, another two highly enriched milk microRNAs) which suppressed intestinal inflammation (miR‐200b) and reduced damage‐induced intestinal cell loss (miR‐146b) [54, 198].

3.7.2. In Vitro Studies

Several studies have investigated the uptake of milk EVs and the transfer of their microRNA content into recipient intestinal cells. However, the data on the bioactivity of the transferred microRNAs were either missing or not conclusive due to the complexity of EV components and the difficulty to pinpoint the EV component or microRNA responsible for the observed biological outcome. A study using raw cow milk demonstrated a significant accumulation of bta‐miR‐148a‐3p, bta‐miR‐146b‐5p, bta‐miR‐30a‐5p, bta‐miR‐26a‐5p, and bta‐miR‐22‐3p in different intestinal cell lines (Caco‐2 and HT‐29) after their incubation with bovine microRNA‐transfected milk‐derived EVs [199]. In this study, a minor alteration in the gene expression of certain mRNA targets was noted [199]. MiR‐22‐3p is considered a key microRNA for the regulation of key cellular processes in the intestines, especially in colorectal cancer, as it was shown to reduce proliferation, invasion, and migration of non‐transformed intestinal cells and colorectal cells and also reduced colorectal cancer metastasis in vivo [57, 58]. Similarly, the miR‐30 family was reported to limit cellular proliferation and favor the differentiation of Caco‐2 via the targeting of multiple signaling pathways [67, 200]. The scarcity of studies investigating the biological role of specific milk‐borne microRNAs is also evidenced by recent reviews on the potential roles of milk‐borne microRNAs in the intestinal mucosa [201, 202]. We believe, however, that the functions assumed by endogenous microRNAs could also apply to exogenous microRNAs, including those from milk, as they both share highly similar (and often identical) sequences and the same mechanisms of action (Figure 1).

3.8. Context‐Dependent Effects of Milk microRNAs

The majority of microRNA functions listed above (also functions listed in Table 1) seem to counteract the inflammatory profile induced by intestinal injury and to restore the barrier integrity (tight junction re‐establishment). This is of critical importance specifically in IBDs in which chronic inflammation and barrier dysfunction are considered disease hallmarks. Nonetheless, depending on the intestinal state, the levels of these microRNAs should be carefully controlled. For example, both miR‐22 (among the top ten highly enriched milk microRNAs) and miR‐155 (another commonly studied milk‐borne microRNA) were shown to aggravate IBD symptoms by increasing inflammation in mice intestinal tissues [62, 203, 204]. Thus, the transfer of the exogenous forms of these microRNAs through milk might aggravate IBD cases in consumers. The fact that the effects of a microRNA might be context‐dependent adds a layer of complexity to classifying it as a positive or a negative regulator of intestinal function, particularly under certain pathological conditions. For example, miR‐181b was shown to induce intestinal cell proliferation in mice with DSS‐colitis and to protect the mice from injury [66]. However, the question remains about whether miR‐181b would induce proliferation in normal intestinal tissues or in intestinal adenocarcinomas. Since in the latter case, miR‐181b would potentially contribute to tumor progression. It is thus important to deepen our understanding of the functions that could be mediated by dietary microRNAs and more specifically milk microRNAs. If these microRNAs are actually profoundly bioactive in consumers, consumption of milk products after weaning and later throughout adulthood should be reassessed.

4. The Interplay Between the Gut Microbiome and Milk microRNAs and EVs

The functions mediated by the intestinal microbiome are indispensable for the proper functioning of the intestines and the protection of the latter against pathogens. Additionally, the intestinal microbiome influences intestinal epithelial and immune cells through direct interactions or via secretion of metabolites that modulate cellular signaling pathways [7, 205, 206]. Mother's milk contains bacteria which could contribute to the infant's intestinal microbiome and supports microbiome establishment in the developing infant [207, 208]. Additionally, besides living bacteria, milk contains several components whose functions on the microbiome have been explored [208]. The findings of these studies were further corroborated by reports revealing differences in microbiome composition between breastfed and formula‐fed infants [209, 210, 211]. Formula‐fed infants might not benefit from the full spectrum of milk bioactive components deemed critical for the establishment of the standard microbiome profile. Among these components, milk EVs and microRNAs are hypothesized to potentially influence the intestinal microbiome and interfere in its functions under physiological and pathological conditions.

4.1. Milk EVs‐Microbiota Crosstalk

A few reports have investigated the effect of milk EVs on the intestinal microbiome. As mentioned earlier in Section 2.3.3, besides their roles in mitigating colitis symptoms via their effects on intestinal immune cells and epithelial cells, milk EVs can also modulate the intestinal microbiome. In fact, it goes without saying that any change at the level of the intestinal mucosa (inflammation, change in permeability and immune tolerance, cellular infiltration) will be associated with a change in the microbiome due to the reciprocal relationship between intestinal cells and the microbiome. Three similar studies using mouse models with DSS‐induced colitis reported that gavage feeding of bovine milk‐derived EVs restored the levels of gut commensal bacteria back to normal and rectified the colitis‐induced dysbiosis [102, 113, 212]. It is worth noting, however, that in these studies, the effects of milk EVs on the gut microbiota are likely the result of the restored immune and barrier functions rather than direct interaction between bacterial cells and EVs. Nonetheless, EV‐based communication between prokaryotes and eukaryotic cells has been already reported [213]. In a recent study, it was shown that bacterial cells from mice ceca could take up and internalize commercial cow milk‐derived EVs [214]. Moreover, in an ex vivo setup, the authors showed that treatment of cecal bacteria with milk EVs resulted in the selection of specific bacterial genomic variants leading to the restructuring of bacterial communities and altered bacterial metabolism [214]. Another study on healthy mice showed that the ingestion of a bovine milk EV‐rich diet affected the prevalence of multiple bacterial phyla and families compared to mice fed milk EV‐depleted diets [215]. Milk EVs were also shown to induce the proliferation of two common intestinal flora strains via the regulation of the gene expression of bacterial growth‐related genes [216]. Given the complex nature of EV cargo and membrane components, further studies are warranted to investigate the effects of milk EVs on the intestinal microbiome.

5. Milk EVs and microRNAs: Bioavailability and Bioactivity

While dietary microRNAs share similarities with the endogenous ones in terms of sequence and mechanism of action, their stability against digestion, uptake, and gene regulatory activities in the consumer's tissues remain debatable. In 2012, Zhang et al. proposed that bioactive plant microRNAs could be absorbed into mouse tissues via food intake [217]. Despite the controversy surrounding the findings from this study, several reports have emerged over the last 10 years investigating the oral transfer of milk‐derived EVs and microRNAs.

5.1. Findings From Bioavailability Studies

Studies on the resistance to digestion supported an overall stability of milk microRNAs to different in vitro digestion approaches [127, 174, 187, 218, 219], though to varying degrees [199]. This proof of stability was corroborated by the detection of milk EVs and microRNAs in fecal components of mice administered with milk or milk EVs orally [220, 221]. Additionally, milk EVs and microRNAs can be taken up and internalized by different cell types in culture [87, 117, 140, 172, 173], including multiple intestinal primary cells and cell lines (IEC‐6 and Caco‐2) [130, 218, 222, 223, 224] and via multiple uptake mechanisms including phagocytosis and caveolin‐mediated endocytosis [126, 129, 130, 131]. Some studies even identified some receptors as being responsible for the uptake of milk EVs by intestinal cells (e.g., Galectin‐3 and FcRn) [121, 128, 225]. As for their bioactivity, milk EVs were proven to influence different cellular processes (immune responses, proliferation, and differentiation) [26, 84, 87, 172]. Whereas microRNAs (EV‐associated or non‐EV‐associated) were proven to regulate gene expression in cells in vitro and in vivo [120, 226]. Pertaining to the in vivo tissue distribution, the use of multiple label‐based approaches and different transgenic models revealed that milk EVs and microRNAs have different distribution profiles over tissues upon their oral administration, with the intestines, liver, brain, kidneys, spleen, and lungs constituting the primary tissues of accumulation [118, 121, 128, 227, 228, 229].

5.2. Study Limitations and Biological Barriers to Bioavailability and Bioactivity

Nonetheless, due to microRNA sequence homology across species and the limitations associated with the label‐based approaches (false positive signals resulting from dye detachment from EVs or microRNAs and concerns regarding label stability during digestion) used for both EVs and microRNAs, there still exists a controversy and debate about their tissue bioavailability and their activities [199, 230, 231]. In fact, the main controversy surrounds the bioactivity of microRNAs in tissues, with doubts about their stability and whether they remain at sufficient levels after digestion to regulate gene expression [199]. Additionally, assuming that sufficient amounts of milk microRNAs resisted to digestion and reached the intestines, a substantial amount of these microRNAs will have to avoid being trapped in the mucus of the intestinal mucosa or be excreted in the stool [220, 228]. Even more so, dietary EVs and microRNAs need to be efficiently uptaken by intestinal cells to subsequently cross the gastrointestinal (GI) barrier and reach the circulation. Thereafter, the amounts of milk EVs and microRNAs could be further partitioned over different tissues, which is also governed by the uptake efficiencies of the different cell types of each tissue. MicroRNAs and EV cargo might also be subject to repackaging on their way, which can affect their homing to tissues. All these events could contribute to what is collectively referred to as the “dilution effect”, greatly diminishing the likelihood of milk microRNAs and EVs being bioactive in recipient tissues. Nonetheless, this influence of the “dilution effect” could be minimized when we consider the effect of microRNAs and EVs on the intestines, as this is the first tissue encountered by milk EVs and microRNAs after digestion.

6. Do Infants Benefit More From Milk microRNAs and EVs?

Since the functions addressed earlier could only be realized by the intestinal mucosa if EVs and microRNAs survive digestion at sufficient levels, it is worth commenting on the differences in the biology of the intestinal mucosa between infants and adults; as these may significantly impact bioactivity and bioavailability.

It is well‐established that the GI tract of the infants during the first 2–3 weeks postpartum represents a primitive version of that of an adult (higher gastric pH, lower enzymatic activity, and higher permeability) [232, 233, 234, 235, 236], implying that the digestion conditions are relatively gentler. In the first place, this could be considered as one factor that may ensure less degradation and loss of the milk components. A study on the intestinal mucus layer in pigs showed that the piglet mucus layer has a different microstructure, is less viscous, and is more permeable compared to that of the pig, which might provide better diffusion and access of diet components to villi cells [237]. When considering the structure and packing of the intestinal lining, infants tend to have shorter enterocytes and a more leaky barrier due to the differences in tight junction composition compared to adults [238]. Regarding the efficiency of absorption and uptake, compared to adults, the enterocytes of the newborn mammals are well known for their increased permeability as they possess a higher macropinocytosis‐based absorptive capacity, more lysosomes, and robust apical endocytosis [239, 240]. What is interesting, though, is the notion that molecules taken up by the lysosome‐rich cells are not necessarily all destined for degradation [240].

All these factors might help reduce the degradation of milk bioactive molecules, leaving them at levels high enough to influence gene expression in the intestinal cells of an infant.

In fact, the vertical transfer of intact bioactive molecules from the mother to the child is not a new concept. During the first 6 months of life, maternal immunoglobulins and immune cells are transferred to infants from their mothers, which can then transduce through the intestinal wall to the lamina propria [241, 242]. Similar to the uptake of immunoglobulins [243], the uptake of bovine milk‐derived EVs by intestinal cells in vitro and in vivo was proven to be FcRn‐dependent [121, 225]. Moreover, research has reported cases of the vertical transfer of several viruses, including cytomegaloviruses (CMV) via breastfeeding [244]. More recently, the H5N1 virus was detected in the dairy milk from infected cows and was able to transfer and transfect mice receiving oral gavage aliquots of the unprocessed milk [245]. If we consider the structural homology between viral particles and EVs, one could assume that EVs and their microRNA content might actually have a chance of being transferred as bioactive molecules to the infant.

7. Unresolved Questions and Research Priorities

Over the past 15 years, roughly 700+ studies have focused on milk EVs and milk microRNAs (an estimate based on PubMed search using the terms “milk extracellular vesicles” or “milk microRNAs”; each term used separately yielded 700 or so results). These studies have provided valuable information on the EV miRNome, biological activity of milk EVs, and their biodistribution profiles. Nonetheless, many questions await being addressed especially in the context of milk EV biogenesis and uptake mechanisms. In fact, these aspects are still underrepresented even for EVs from non‐milk origin.

7.1. Research on Milk EV Biogenesis and Milk EV‐Associated microRNA Packaging

As per the research on the biogenesis pathways, however, it seems to be more challenging in the milk context due to the limited availability of mammary gland cellular models. This also poses a challenge for the study of another aspect which also remains ambiguous in the field of milk EVs, namely the mechanisms governing the sorting of the different microRNAs into milk EVs. Pursuing research in this direction is especially valuable, given the prospect of using biocompatible milk EVs to deliver a particular therapeutic microRNA.

7.2. Research on Milk EV Uptake

Over the last 5 years, the number of studies focusing on the protein content of milk EVs from various species has increased markedly [28, 30, 105, 133, 134, 135, 246, 247, 248], compared to those published over the period extending from 2007, the year milk EVs were first reported [26]. Interestingly, the data on milk EV proteomes seem to be more consistent among different studies, unlike miRNome profiles. Albeit, information on the different PTMs of milk EVs, other than the glycome, are still lacking. In fact, a deeper understanding of the EV proteome, more specifically the surfaceome and the associated PTMs, would be the first step to understanding the potential uptake mechanisms involved. On the same note, much more focus should be invested in the analysis of the EV lipidomic. The investigation of the EV lipid content not only helps in the understanding of the uptake mechanism but could also help predict the fate of the EV bioactive components, including their microRNA content, in the recipient cells. A study showed that the lipid composition of the LNPs governs their behavior in response to endosomal pH [249]. The latter event could eventually determine whether LNP components are to be sorted to recycling endosomes for subsequent escape to the cytoplasm or to lysosomes for their eventual degradation [249]. Studies dissecting the microRNA cellular trajectory of the internalized EVs are needed for milk EVs to understand the final fate of their microRNA content, as has been done in studies on tumor‐derived EVs [250].

7.3. Tailored Models for the Study of Bioactivity

Given the differences among the developmental and physiological processes between adults and infants, milk EVs and microRNAs might not manifest the same biological outcome in both systems. Therefore, establishing model systems tailored to each developmental stage is crucial for accurately assessing the biological effects of milk EVs and microRNAs. Insights from such studies could shape adult nutritional recommendations and refine protocols for infant formula formulation and donor milk processing (human milk banks). These aspects are critical, especially when comparing formula‐fed infants versus breastfed infants, as current formulas seem to lack many of milk's bioactive components. Considering the potential roles of milk microRNAs and EVs in ensuring a proper intestinal development for the infants, failure to benefit from these molecules by the formula‐fed or processed donor milk‐fed infants might predispose the latter to future intestinal diseases.

8. Conclusion

Research on milk EVs and microRNAs is still growing at a noticeable rate and, interestingly, recent studies are beginning to break down critical aspects related to EV uptake and the factors involved, which were overrepresented for a while of time. Studies aiming at dissecting the different subpopulations of milk EVs, their microRNA content, their resistance to digestion and processing, and their behavior in recipient cells or tissues are warranted. By examining the public database portal (PubMed) for the recent reports on milk EVs and microRNAs, it is sort of noticeable that the scientific community is tending to lean into supporting a more localized role of milk‐derived microRNAs and EVs, primarily in the intestinal mucosa rather than in other distant tissues [251, 252, 253]. Given that the study of milk EVs and microRNAs transfer in the intestines is less complex compared to other distant tissues, recent findings appear to be more robust. The above findings could pave the way for future studies that may reveal novel therapeutic applications of milk‐derived EVs and microRNAs in maintaining and enhancing intestinal barrier function.

Author Contributions

Z.H. reviewed the literature, compiled the evidence, and wrote the first draft of the manuscript. C.G. revised and finalized the manuscript. All authors have read and approved the final manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Husseini Z. and Gilbert C., “Milk‐Derived Extracellular Vesicles and microRNAs: Potential Modulators of Intestinal Homeostasis,” The FASEB Journal 39, no. 16 (2025): e70947, 10.1096/fj.202501630R.

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated in this review. The authors confirm that the data supporting the findings are available within previously published studies. Proteomic data used for secondary analysis of protein glycosylation profiles were obtained from the Supporting Information of a previously published study, as cited in the manuscript.