CHAPTER 1: Role of Nutrition, the Epigenome, and MicroRNAs in Cancer Pathogenesis

Good nutrition and the daily consumption of fresh fruit and vegetables have long been encouraged to maintain human health and extend lifespan. In recent years, accumulating experimental evidence supports the notion that dietary factors confer protective effects against diseases such as cancer. Scientists have identified specific bioactive factors present in dietary foods such as green tea, soy, whole grains, fruits, and cruciferous vegetables (broccoli, kale, cauliflower) that impact epigenetic mechanisms regulating chromatin remodeling and gene expression, and which are closely associated with human cancers. An important mechanism employed by these nutritional agents to mediate their cellular effects is the regulation of microRNAs (miRNAs). miRNAs are a major class of ∼22-nucleotide long non-coding RNAs that generally function to block protein translation and/or degrade their messenger RNA (mRNA) targets. These small RNAs direct many essential processes related to cellular growth, apoptosis, differentiation, metabolism, and the immune response. miRNAs are often aberrantly expressed in human tumors. Many of these cancer-associated miRNAs act as tumor suppressors or pro-oncogenic factors that directly impact cancer progression and metastasis. In this chapter, we review how nutritional factors influence the epigenome and miRNA expression to confer their cancer-protective effects based on emerging in vitro data and animal studies. The diet–epigenome interactions and their influence on cancer-associated miRNA-mediated gene regulation are also explored in the context of the gut microbiota. Finally, the controversial field of ingested xenomiRs, particularly plant-based miRNAs, is discussed as a novel method of miRNA delivery to control tumorigenesis. A complete understanding of how miRNAs respond to nutritional cues and cancer-related signaling pathways will lead to promising diagnostic biomarkers as well as affordable and easily obtained therapeutic dietary supplements and/or edible vaccines to improve human health.

A hallmark of cancer is the uncontrolled growth and survival of damaged cells. This is caused by inappropriate activation or inhibition of RNA and protein factors residing within signaling pathways that control proliferation, differentiation, and apoptosis. These pathways can be altered by exposure to environmental factors, such as stress, drugs, and nutrition, leading to genomic mutations or alterations of the epigenome. Epigenetic modifications are heritable and often reversible changes in gene expression that do not alter the DNA sequence. These epigenetic alterations control gene expression in both positive and negative ways, commonly involving DNA hypo- and hypermethylation (i.e., CpG islands within promoters), chromatin remodeling, histone protein modifications (e.g., acetylation, methylation, phosphorylation), and non-coding RNAs (i.e., miRNAs). Epigenetic modifications can influence DNA stability as well as the ability of transcription factors to interact with genomic elements, and thus ultimately determine if genes will be active or silenced. Therefore, the epigenome can dictate the overall protein profile in a cell that can have important and lasting biological and/or pathological ramifications. Epigenetic control of developmental events and differentiation processes include X-chromosome inactivation, genomic imprinting, genomic reprogramming, and stem cell maintenance.

An epigenetic link to cancer was first proposed in 1983 by Feinberg and Volgelstein, who observed that certain genes in tumor cells of cancer patients were hypomethylated compared to cells from normal adjacent tissues.¹  The theory that changes in gene expression due to epigenetic alterations (e.g., DNA methylation status) predisposed individuals to cancer was verified with the discovery in 1989 that hypermethylation of the tumor suppressor gene Retinoblastoma (RB) was the driver of disease initiation and spontaneous regression.²  It is now well established that DNA methylation and histone modifications that result in gene silencing occur in a variety of human cancers.^3,4  Recently, epigenetic mechanisms associated with miRNAs have gained considerable attention.⁵  These small ∼22-nucleotide non-coding RNA transcripts are often aberrantly expressed in a wide array of human cancers.⁶  A growing subset of this non-coding RNA class, designated as “oncomirs” (miRNAs associated with cancer), mediate tumor formation and disease progression. A greater understanding of how cancer-associated miRNAs intersect with the epigenome and the nutritional axis, discussed in this chapter, will aid in their development as novel diagnostic and therapeutic tools for cancer.

There are more than 2600 miRNAs in the human genome to date, largely identified via cloning and RNAseq methods (miRBase, release 22). miRNAs do not encode for proteins, but exert their biological effects as non-coding RNAs, and generally act to block target gene expression post-transcriptionally. The biogenesis of miRNAs is complex.⁷  miRNAs are transcribed in the nucleus by RNA polymerase II (although Pol III transcription has been observed) to generate a precursor, pri-miRNA, which are often 5′ capped and 3′ polyadenylated. A pri-miRNA transcript can encode for multiple miRNA genes and each gene is processed into a ∼70-nucleotide pre-miRNA hairpin precursor by the RNase III enzyme Drosha and its cofactor, DiGeorge syndrome critical region 8 (DGCR8). An independent subclass of pre-miRNAs, termed “mirtrons”, do not rely on Drosha processing and rather are generated from mRNA transcripts as by-products of exon splicing and intron disbranching events.⁸  Pre-miRNAs are subsequently exported out of the nucleus by RAN-GTP and Exportin 5. Once in the cytoplasm, pre-miRNAs are processed further by the RNase III enzyme, Dicer, and its co-factor, TAR RNA binding protein (TRBP) to generate a ∼22-nucleotide double-stranded RNA duplex. One strand of this duplex is preferentially loaded into a large multiprotein miRNA-associated RNA-induced silencing complex (miRISC). Argonaute (AGO) is the key catalytic component of this complex responsible for miRNA strand selection and mediating miRNA-based alterations of gene expression. The loaded miRNA serves as a guide to escort the AGO/miRISC complex to the targeted site via miRNA binding to complementary sequences within the mRNA transcript, and ultimately results in target mRNA degradation and/or blocked protein translation.

Human miRNAs have been well characterized to bind with imperfect complementarity to the 3′ untranslated region (3′ UTR) of the target mRNA, although miRNA binding within the 5′ UTR and coding regions of the mRNA target can also modulate target gene expression.⁷  Perfect base pairing of a highly conserved “seed sequence” (nucleotides 2–8) in the miRNA is important for proper miRNA targeting.⁹  A few notable exceptions exist where miRNAs activate target gene expression via epigenetic regulation of enhancer regions in the nucleus or post-transcriptionally to induce protein translation.^10,11  Bioinformatic estimates indicate that a single miRNA can recognize upwards of 100 distinct targets, and therefore can regulate multiple non-overlapping biological pathways simultaneously. From a therapeutic standpoint, single miRNA therapy could be a powerful tool to treat cancer in patients that harbor an accumulation of genetic alterations and be effective without identifying the key mutations that led to tumor formation. This strategy could be particularly useful in diseases that exhibit extremely heterogeneous tumor populations carrying multiple genetic mutations within the same cancer patient.

miRNAs often exhibit abnormal expression profiles in fluids and tissues obtained from cancer patients compared to non-cancer patients.⁶  A large proportion of these cancer-associated miRNAs are located in genomic loci referred to as “fragile sites”, which are unstable regions of human chromosomes subject to gaps, breaks, or DNA rearrangements during replicative stress, and are closely associated with cancer.¹²  For example, miR-125 and let-7 family members reside within fragile sites often deleted in lung, breast, ovary, and cervical cancers.¹²  miRNA dysregulation in cancerous tissues can also be caused by genetic defects outside of fragile sites, epigenetic alterations, or disruptions in miRNA biogenesis or miRNA–target interactions. Research using in vitro cell culture systems and animal disease models indicates that aberrant miRNA expression in turn can directly impact chromatin remodeling, tumor initiation and progression, epithelia to mesenchymal (EMT) transitions related to cell migration and invasion, metastasis and drug resistance via dysregulation of miRNA cancer-related targets, such as RAS and TP53.^6,13,14  In this way, patterns of miRNA expression detected in diseased patients compared to normal control populations likely reflect abnormalities in mechanistic pathways related to cancer progression, patient survival, and response to clinical therapies. These miRNA expression signatures are promising biomarkers for detection of early-stage cancer as well as discriminatory markers to identify patients at highest risk for aggressive and lethal forms of cancer and who may benefit most from therapeutic and/or surgical interventions.

Calin et al. first proposed that miRNAs play a direct role in cancer progression in 2002, and found that the clustered miRNAs miR-15a and miR-16-1 were often deleted in patients with B-cell chronic lymphocytic leukemia.¹⁵  Indeed, these miRNAs have been well characterized and belong to a subclass of tumor suppressor miRNAs that act in vitro and in vivo to block proliferation, invasion, and tumorigenesis of leukemias (as well as epithelial cancers) by repressing downstream miRNA targets, which include anti-apoptotic gene B-cell lymphoma 2 (BCL2), G1/S cell cycle progression factor Cyclin D2 (CCND2), and cell growth and survival factor insulin-like growth factor 1 receptor (IGF1R).^16–18  Conversely, pro-oncogenic miRNAs are overexpressed in cancer patients and experimentally verified to promote cancer progression pathways and accelerate tumor formation and metastasis. miR-21 is one of the most commonly overexpressed miRNAs in human cancers and promotes tumor growth by blocking the expression of tumor suppressor genes, including phosphatase and tensin homologue (PTEN) and programmed cell death protein 4 (PDCD4).¹⁹  Furthermore, miRNA expression profiling more accurately classifies human cancers based on tissue origin than the expression of protein-coding genes, reflecting the utility of miRNAs as clinical diagnostic tools.²⁰  In addition, miRNAs show promise as prognostic biomarkers.²¹  For example, lung cancer patients exhibiting reduced let-7 expression and elevated miR-155 expression have worse post-operative survival than patients exhibiting reciprocal expression patterns.^22,23  Thus, miRNAs have emerged as major regulators of cancer progression with immense clinical potential. The ability of dietary factors to modulate miRNA expression will have a huge impact on disease prevention and therapeutic development.

miRNA activity is not confined to the cells in which they are produced. Rather, miRNAs are found to circulate in the blood and can be taken up by other cells and tissues at distant sites.²⁴  These small non-coding RNAs are resistant to degradation in body fluids, presumably when complexed with proteins such as AGO2.²⁵  Mitchell et al. were the first to isolate cell-free miRNAs from human serum and found that miR-141 was elevated in the blood of metastatic prostate cancer patients compared to controls.²⁶  miRNAs have also been reported in human urine and Bryant et al. reported significantly higher miR-107 and miR-573-3p levels in urine of patients with prostate cancer, compared with controls.²⁷ 

The majority of miRNAs found in body fluids are concentrated in small membrane vesicles referred to as “exosomes” or “microvesicles”.²⁸  Exosomes are 40–100 nm vesicles generated by a large range of cell types, including tumor cells and are released extracellularly via fusion of multivesicular bodies with the plasma membrane.²⁹  These secreted microvesicles have been detected in a variety of human body fluids, e.g., blood (serum and plasma), bronchoalveolar lavage, urine, bile, ascites, breast milk, and cerebrospinal fluid.³⁰  Delivery of exosomes to other cells is not well understood and may occur by receptor-mediated internalization, docking and fusion to target cells, or endocytotic mechanisms. Exosomes probably represent a novel mechanism of miRNA transfer from one cell to another and evidence suggests that these miRNAs can mediate downregulation of gene expression and influence cellular activities within the target cell.^31,32  Therefore, miRNAs selectively packaged in exosomes and secreted by tumor cells into the circulation may be useful fluid-based biomarkers reflecting a patient's cancerous state and disease prognosis.

From the perspective of this chapter, exosomal miRNAs delivered through the diet may be used to mediate cellular effects in the human body (Figure 1.2). Exosomes containing miRNAs are in fact present in human breast milk and bovine milk.^33,34  Novel (and somewhat controversial) evidence indicates that exosomes from bovine milk are taken up by human intestinal cells and endothelial vascular cells via endocytosis.^35,36  Wolf et al. showed that bovine exosomes are transported into human colon carcinoma Caco-2 cells in culture and result in elevated miR-29b and miR-200c levels.³⁵  Interestingly, both of these miRNAs have been implicated in human cancers. Exosomal miRNAs are unusually stable and are resistant to acidic conditions and RNase degradation in the gut.³⁷  Bovine exosomes can enter into the body's circulation and are transported to peripheral tissues—likely due to exosomal uptake by circulating immune cells (i.e. human macrophages) in the blood.^36–38  Taken together, dietary uptake of exosomes could have major ramifications in our discussion of how miRNAs are ingested through the diet (milk, plants, microbiota) and regulate gene expression in recipient cells. Continued work to understand the mechanisms of exosome-based delivery of cancer-protective miRNAs is required to translate this research into the clinic.

Dietary factors can impact both the epigenome and miRNA expression to mediate cancer progression pathways in human cells. It is therefore important to briefly review the overall role miRNAs play in epigenetics and chromatin remodeling (Figure 1.1).

Epigenetic mechanisms involving DNA modifications such as methylation and changes to chromatin structure can have a significant impact on gene expression in both positive and negative ways by regulating access of the transcriptional machinery to the DNA (for a comprehensive review, see³⁹ ). For example, DNA methylation is a key means of silencing gene expression in human cells (Figure 1.1A). This involves the addition of a methyl group exclusively on cytosine residues adjacent to guanines (clustered in CpG islands) and generally located within gene promoter regions. The DNA methyltransferase (DNMT) enzymes, DNMT1, DNMT3A and DNMT3B, catalyze this reaction de novo by transferring a methyl (–CH₃) group from S-adenosyl methionine (a methyl donor) to the 5′ position on the cytosine ring. The promoters of active genes are sparsely methylated (hypomethylated), whereas silenced genes are commonly associated with hypermethylation.

Another method to regulate gene expression is the control of chromatin structure (Figure 1.1B). DNA is compactly packaged into the nucleus. Approximately two turns of DNA are wrapped around a nucleosome, which is composed of an octamer of histone core proteins. Modifications to these histone proteins (e.g., methylation, acetylation, phosphorylation, ubiquitination, SUMOylation, and ADPribosylation) will alter the compactness of the DNA and thus the ability of DNA-binding factors and cellular machinery to access the chromatin and initiate/maintain gene transcription. For instance, acetylation of histones is one of the primary mechanisms used to open up the chromatin and activate gene transcription. These modifications are dynamic. Specific enzymes called histone acetyltransferases (HATs), which add acetyl (–CH₃CO) groups onto histone tails (i.e., H3K9ac), and histone deacetylases (HDACs), which catalyze the removal of these acetyl groups, and are used to fine-tune gene expression. The effects of histone methylation, mediated by the enzymes histone methyltransferases (HMTs) and histone demethylases (HDMs) are more complex then histone acetylation. In some cases, histone methylation can enhance gene transcription (i.e., H3K4me3); however, in other cases, histone methylation represses transcription (i.e., H3K9me3, H3K27me3). In this section, we focus on how miRNAs associated with human cancers can directly alter the enzymes responsible for such epigenetic modifications, as well as how miRNAs are themselves regulated by the epigenetic machinery, and in turn impact cancer progression.

miRNAs can contribute to cell transformations and cancer progression by targeting DNA methyltransferase enzymes. Hypermethylation of DNA promoters is a common feature in human cancers.⁴  The DNMT enzymes DNMT1, DMNT3A, and DNMT3B, are often elevated in liver, prostate, breast, colorectal, and lung cancers.^40–43  Expression of the miR-29 family (miR-29a, miR-29b, and miR-29c) is downregulated in tumors from non-small cell lung cancer patients and inversely correlates with DNMT3A and DNMT3B expression. Work by Fabbri et al. validated that these DNMT enzymes are direct miR-29 family targets.⁴⁴  Enforced miR-29 expression in human A549 lung cancer cell lines resulted in global reduction of DNA methylation patterns and blocked tumorigenesis in vitro and in vivo due to the re-expression of the tumor suppressor genes fragile histidine triad protein (FHIT) and WW domain-containing oxidoreductase (WWOX).⁴⁴  Overexpression of miR-29b also causes global DNA hypomethylation in acute myleloid leukemia by repressing DNMT3A and DNMT3B.¹³  Interestingly, miR-143 (but not miR-29) targets DNMT3A in colorectal tissues, indicating that these miRNAs function in a tissue-specific manner.⁴⁵ 

In addition, miRNAs can repress enzymes involved in histone modifications. For example, miR-101 directly targets the histone methyltransferase enhancer of Zeste homolog 2 (EZH2), a subunit of the polycomb-repressive complex 2 (PRC2) which catalyzes methylation of lysine 27 of histone H3 (H3K27me3) and lysine 9 of histone H3 (H3K9me3) to promote transcriptional silencing.⁴⁶  EZH2 is often overexpressed in cancers (i.e. prostate, breast, bladder, gastric) and functions to promote cell proliferation, invasive disease, and stem cell maintenance.⁴⁶  In prostate cancer, reduction of miR-101 expression inversely correlates with increased EZH2 levels, and in many cases this is due to miR-101 genomic loci deletions.^46,47  Furthermore, miR-1 and miR-140 have both been validated as direct targets for HDAC4. miR-1 targets HDAC4 in muscle and chondrocyte tissues and is responsible for promoting differentiation of these tissues, whereas miR-140 is cartilage-specific and targets HDAC4 to regulate long bone development.^48,49  Interestingly, an estimated 93% of human chordoma bone tumors show low levels of miR-1 expression compared to normal tissue, indicating its use as a biomarker for this cancer type.⁵⁰  In addition, Duan et al. revealed that forced expression of miR-1 in human chordoma cancer cell lines decreased HDAC4 expression and inhibited cell growth—implicating the importance of tumor-suppressor miRNAs to fine-tune chromatin remodeling.⁵⁰ 

The epigenetic machinery can regulate miRNA expression and, in turn, impact cancer progression pathways. 5–10% of miRNA loci reside within CpG islands and are proposed to be epigenetically regulated. The expression of miRNAs such as miR-148a, miR-34b/c, miR-9, and let-7a-3 are dependent on their methylation status and DNMT1 and DNMT3b DNA methyltransferases.^51–55 

Promoter CpG hypermethylation is associated with a global loss of miRNA expression in colon cancer cell lines.³  Many of these miRNAs play tumor-suppressive roles. For example, DNA hypermethylation of miR-124a leads to increased activation of cyclin dependent kinase 6 (CDK6) and the phosphorylation of RB, both of which promote cell cycle progression and tumor growth.⁵⁶  Furthermore, in a screen using T24 bladder cancer cells, Saito et al. compared miRNA expression profiles in the presence and absence of DNMT inhibitor 5-aza-2′-deoxycytidine (5-AZA) and HDAC inhibitor 4-phenylbutyric acid (4-PBA).⁵³  17 out of 313 miRNA genes studied were upregulated following this treatment. This group noted that the miR-127 gene resides within a CpG island and miR-127 expression was upregulated due to DNA demethylation as well as histone H3 acetylation and histone H3 lysine 4 trimethylation (H3K4me3). Therefore, epigenetic mechanisms control miRNA expression and ultimately regulate the miR-127 target, pro-oncogene B-cell lymphoma 6 (BCL6).⁵³  In addition, miR-148a, miR-34b/c, and miR-9 hypermethylation have all been linked to metastasis in patients with melanoma, colon, lung, breast, head, and neck carcinomas, and thus could be employed as a prognostic tool. In animal models, these miRNAs function as anti-metastatic agents to reduce migration, tumor formation, and metastasis.⁵⁵ 

miRNA promoter hypomethylation has also been linked to cancer progression and can result in overexpression of pro-oncogenic miRNAs to promote tumor formation. miRNA promoter demethylation is attributed to increased expression of miR-126 and miR-128 in acute myeloid leukemia and acute lymphocytic leukemia, respectively.⁵⁷  Therefore, chromatin-modifying drugs may be a powerful approach to regulate oncomiR expression and treat a large range of human cancers.

We have discussed the impact miRNAs and epigenetic modifications have on cancer progression. Diet is also a key player in the development of cancer. In this section, we discuss how the dietary intake of fresh fruits, vegetables, nuts, and beverages (i.e., tea, wine), which are rich in polyphenols modulate miRNA expression, and in some cases the epigenome, in order to mediate their cancer-protective effects.

Polyphenols are a major class of metabolites that possess antioxidant, free-radical scavenging properties, and notable cancer-protective effects. More than 500 polyphenols have been identified, including classes of phenolic acids, flavonoids, stilbenes, tannins, lignans, quinones, and curcuminoids. Many of these compounds directly regulate miRNAs that are closely associated with human cancers.^58,59  Specifically, the dietary polyphenols resveratrol, ellagitannins and ellagic acid, epigallocatechin-3-gallate, genistein, curcumin, and diindolylmethane have been studied intensely and shown mechanistically to exert their tumor-suppressive and cancer-protective effects by regulating miRNA expression (Table 1.1). These edible compounds are promising therapeutic drugs and disease-prevention agents for administration to the general population, since foods rich in these anti-cancer polyphenols are easily accessible, non-toxic, and affordable.

Resveratrol belongs to the stilbenoid class of polyphenol compounds commonly found in grapes, wine, peanuts, cocoa, blueberries, and cranberries. This polyphenol has antioxidant and protective effects against a variety of malignancies, including pancreas, colon, and breast cancers.⁶⁰  Resveratrol has been shown to inhibit pancreatic and bladder cancer cell growth and induce apoptosis by suppressing pro-oncogenic miR-21 expression and anti-apoptotic factor BCL2.^61,62  In the context of colon cancer, treatment of human SW480 colon cancer cells with resveratrol decreased several pro-oncogenic miRNAs, including miR-17, miR-21, miR-25, and miR-92a-2, which are often overexpressed in colorectal cancer patients.⁶³  Resveratrol delivery also elevated the expression of tumor-suppressive miRNAs, including miR-34a and miR-663, in colon cancer cells.^63,64  Reciprocal repression of miR-663 target transforming growth factor-β receptor type I (TGFβR1) in treated cells probably mediates the protective effects of resveratrol against cancer progression phenotypes associated with angiogenesis, cell motility, invasion, and metastasis.⁶³  Furthermore, Hagiwara et al. found that resveratrol exposure increased tumor-suppressive miRNAs miR-141 and miR-200c in MDA-MB-231-luc-D3H2LN breast cancer cells in vitro, resulting in the inhibition of cancer cell invasiveness.⁶⁵  Consequently, miR-200c targeted and repressed pro-oncogenic transcription factor zinc finger E-box binding homeobox 1 (ZEB1), an E-cadherin regulator and promoter of tumor invasion and metastasis. Interestingly, resveratrol is likely to have a more global impact on miRNA activity in cancer cells. Resveratrol treatment can increase transcription of AGO2, the catalytic component of miRISC, in breast cancer cells.⁶⁵  Since cancer cells often exhibit widespread miRNA underexpression, this work indicates that resveratrol treatment could normalize and restore miRNA activity indirectly via AGO2 to confer its protective cancer effects.

Ellagitannins and ellagic acid are the main bioreactive polyphenol compounds found in fruits and nuts, such as pomegranates, black raspberries, raspberries, strawberries, walnuts, and almonds. Colon cancer cells treated with pomegranate extract containing ellagitannins and ellagic acid indicated that these polyphenol compounds have anti-proliferative and apoptotic effects in vitro and in vivo.⁶⁶  Nunez-Sanchez et al. investigated miRNA gene modulation in 35 colorectal cancer patients who consumed 900 mg of pomegranate extract daily before surgery compared to patients who did not consume the extract.⁶⁷  Specifically, miR-646, miR-1249, miR-135b-5p, miR-135b-3p, miR-92b-5p, miR-765, miR-496, miR-181c-3p, and miR-18a-3p were differentially expressed in colon tumors versus normal adjacent colon tissues in the treated patients. Unfortunately, the functional significance of these miRNAs was not further explored, and thus the contribution of non-coding RNAs in mediating the protective effects of pomegranate derived-ellagitannins and ellagic acids are unknown.

Epigallocatechin-3-gallate (EGCG) is a polyphenol catechin found in green tea (Camellia sinensis), which is a popular beverage in China and Japan. Epidemiologically, EGCG ingestion is attributed to the prevention of human oral and lung cancers.^68,69  EGCG effectively blocks tumor growth and promotes apoptotic cell death for a range of cancers, including oral, lung, digestive tract, skin, liver, prostate, and breast cancers.⁷⁰  EGCG has emerged as an important regulator of cancer-associated miRNAs. In the context of liver cancer, EGCG treatment of human hepatocellular carcinoma HepG2 cells in vitro resulted in the upregulation of tumor-suppressive miRNA miR-16 and reciprocal downregulation of the miR-16 target anti-apoptotic protein BCL2. Indeed, treatment of HepG2 cells with miR-16 antimiRs blocked EGCG-mediated apoptotic effects and increased BCL2 expression.⁷¹  In addition, EGCG impacts prostate cancer progression. EGCG can block the nuclear translocation of the androgen receptor, a major signaling pathway for prostate growth. EGCG suppressed prostate tumor growth in mouse xenograft models, resulting in the downregulation of oncogenic miRNA miR-21 as well as elevated expression of tumor suppressive and pro-apoptotic miR-330.⁷²  Therefore, EGCG controls the expression of androgen receptor-regulated miRNAs and holds promise as a clinical tool for castration-resistant prostate cancer, which remains an incurable and lethal disease. In lung cancer, Wang et al. showed that EGCG exposure in human H1299 and H460 lung cancer cell lines caused hypoxia-induced factor 1α (HIF-1α) accumulation, as well as upregulation of the hypoxia-inducible miRNA miR-210.⁷³  As predicted, miR-210 treatment mirrored the ability of EGCG to reduce cell proliferation and block anchorage-independent growth in lung cancer cells. Furthermore, EGCG treatment of human lung cancer A549 cells resulted in selective repression of miR-98-5p.⁷⁴  Interestingly, Zhou et al. found that EGCG enhanced the efficacy of cisplatin-induced apoptosis in these cells, which was mediated via miR-98-5p downregulation.⁷⁴  Therefore, the combined treatment of chemotherapeutic agents, such as cisplatin, with green tea extracts could be a novel clinical strategy for lung cancer.

Genistein is an isoflavinoid present in soybeans with antioxidant, antimicrobial, and protein tyrosine activation enzyme inhibitor activities. This plant polyphenol shares a similar chemical structure to estrogen and is shown to modulate cancer progression of hormone-responsive tissues such as the breast, ovary, and prostate. Although genistein is reported to have oncogenic estrogenic properties in estrogen receptor positive (ER+) breast cancers, dietary levels of genistein inhibited cell growth and induced apoptosis in estrogen receptor negative (ER−) human breast cancer cell lines MDA-MB-435 and Hs578t (but not in ER + MCF-7 cells).^75,76  de la Parra et al. found that anti-cancer activity correlated with pro-oncogenic miR-155 expression in the ER− cell lines and genistein treatment of MDA-MB-435 and Hs578t cells decreased endogenous miR-155 levels.⁷⁶  Reciprocally, treated ER− cells showed increased expression of miR-155 targets, i.e., tumor suppressors forkhead box O3 (FOXO3) transcription factor, PTEN, serine-threonine kinase β-catenin inhibitor casein kinase 1alpha (CK1α), and cyclin dependent cell cycle inhibitor p27.⁷⁶  Additional miRNAs expressed in the mammary gland, notably miR-23b, were induced upon genistein treatment and acted to inhibit breast cancer cell growth.⁷⁷  In the context of ovarian cancer, Xu et al. found that genistein downregulated miR-27a levels in ovarian carcinoma SKOV3 cells and subsequently blocked cell growth and migration.⁷⁸  This was accompanied by reciprocal elevation of putative miR-27a target sprouty homologue 2 (SPRY2), an extracellular receptor kinase 1/2 (ERK1/2) signaling inhibitor.⁷⁸  Furthermore, genistein reduced miR-221 and miR-222 expression in human prostate cancer PC3 cells and conversely elevated levels of the validated miR-221/miR-222 target aplysia ras homology member I (ARH1), a tumor suppressor that inhibits prostate proliferation, invasion, and promotes apoptosis.⁷⁹  Genistein delivery to human PC3 and LNCaP prostate cancer cell lines also induced miR-1296 expression, which in turn suppressed the DNA replication factor minichromosome maintenance 2 (MCM2).⁸⁰ 

Interestingly, a link between genistein's tumor suppressive effects and its ability to regulate both miRNA expression and the epigenome has been well established in the prostate. For example, Chiyomaru et al. showed that genistein treatment elevated miR-574-3p levels in prostate cancer cells, and reciprocally downregulated the pro-oncogenic miR-574-3p targets that control prostate cell proliferation, migration, invasion, and metastasis. These included Rho family GTPase RAS-related C3 botulinum toxin substrate 1 (RAC1), epidermal growth factor receptor (EGFR), and histone acetyltransferase p300.⁸¹  In another study, Zaman et al. noted that treatment with genistein or demethylation drug 5-azacytidine (5-AZA) of human PC3 prostate cancer cells induced expression of miR-145 via demethylation of the miR-145 promoter, and subsequently resulted in cell cycle arrest and apoptosis.⁸²  Similarly, Rabiau et al. compared genistein treatment in a panel of human prostate cancer cell lines with 5-AZA exposure.⁸³  They found that either condition resulted in similar miRNA expression profiles. Specifically, either genistein or 5-AZA treatment significantly increased miR-548b-5p levels (and decreased miR-125a, miR-125b, miR-15b, and miR-320 expression) in PC3 cells, increased the tumor suppressor miR-15a (and decreased miR-494, miR-520g, and miR-542) in LNCaP cells; and reduced miR-155, miR-208b, miR-211, miR-376a, and miR-411 levels in DU145 cells. Therefore, the soy-based isoflavone genistein relies on DNA methylation and histone protein alterations to regulate miRNA expression and cancer progression.

In addition, genistein is an important cancer-associated miRNA regulator in tissues outside of the reproductive system. For example, application of genistein to human A549 lung cancer cells resulted in increased caspase-3/9 driven apoptosis and suppressed cell proliferation by activating miR-27a expression and reciprocally shutting down the miR-27a target MET, a pro-oncogenic receptor tyrosine kinase.⁸⁴  (Potentially, miR-27a has oncogenic activities in the ovary, and genistein had opposite effects on miRNA expression in this cell type, as discussed earlier.) In colorectal cancers, genistein was found to reduce cell growth and induce apoptosis in HCT-116 cells both in vitro and in vivo.⁸⁵  These effects were mediated by downregulation of miR-95 as well as the oncogenic serine/threonine-protein kinases Ak strain transforming (AKT) and serum glucocorticoid regulated kinase 1 (SGK1).⁸⁵  Furthermore, genistein reduced oncogenic miR-21 in renal cancer cells in order to inactive NOTCH signaling.⁸⁶  Taken together, genistein has potential as a natural anti-cancer agent and may prove especially useful in drug-resistant ER− breast cancers.

Curcumin is a natural curcuminoid phenol derived from turmeric root (Curcuma longa), a plant native to South Asia that has been used as a cooking spice and an herbal medicine for centuries. Scientific evidence demonstrates that curcumin has anti-oxidant, anti-inflammatory, and anti-cancer activities. This dietary compound is particularly toxic to tumor cells, but ineffective in normal cells and cellular uptake is notably higher in cancer cells for reasons not well understood.⁸⁷  Curcumin regulates miRNA expression in a large range of tumors. Curcumin treatment of breast cancer cell lines, for instance, induces miR-181b, miR-34a, miR-16, miR-15a, and miR-146b-5p expression, and downregulates miR-19a and miR-19b.⁸⁸  This treatment resulted in apoptosis, suppressed tumor formation, and metastasis. Many of these modulated miRNAs have been functionally characterized. For example, Kronski et al. observed that curcumin treatment of human breast cancer MDA-MB-231 cells elevated miR-181b and consequently suppressed the miR-181b pro-metastatic targets cytokines CXCL1 and CXCL2.⁸⁹  Indeed, forced miR-181b expression in breast cancer cells similarly blocked proliferation and invasion, promoted apoptosis, and inhibited metastasis in mouse models.⁸⁹  Conversely, miR-181b antimiRs abolished curcumin's anti-cancer effects.⁸⁹  Curcumin appears to use miR-15-based regulation as a common mechanism to suppress cancer cell growth and induce apoptosis. Curcumin-induced miR-15a expression and reciprocal downregulation of miR-15a target anti-apoptotic factor BCL2 was employed in human breast cancer and laryngeal cancer cells to curb their growth.^90,91  In leukemia cells, curcumin elevated both miR-15 and miR-16 levels to suppress growth via the miR15/miR-16 target, oncogene Wilms' tumor 1 protein (WT1).⁹² 

Pro-oncogenic miRNA miR-21 is another important mechanism used by curcumin to mediate cancer protective effects in colorectal cancers. Evidence indicates that curcumin blocks miR-21 expression via transcriptional regulation. Curcumin treatment of colon cancer Rko and HCT116 cells reduced miR-21 promoter activity by inhibiting activator protein 1 transcription factor (AP1) binding to the miRNA promoter.⁹³  This resulted in elevated expression of miR-21 targets, programmed cell death protein 4 (PDCD4) and PTEN, tumor suppressors known to block colon cancer cell growth, invasion and metastasis.^93,94  In chemotherapy (5-flurouracil + oxaliplatin)-resistant HCT116 and HT-29 colon cancer cell lines, curcumin-mediated reduction of miR-21 restored PTEN levels, reduced oncogenic AKT-phosphorylation, and inhibited colon cell growth.⁹⁴  Therefore, curcumin-based control of pro-oncogenic miR-21 may be a promising approach to treat drug-resistant colorectal cancers.

Curcumin may also use epigenetic mechanisms to elevate tumor suppressor miRNA expression. Roy et al. specifically investigated how curcumin restored tumor suppressor miR-34a expression in colon cancer.⁹⁵  miR-34a is silenced in colon cancer cells due to CpG island hypermethylation within the miR-34 promoter. They noted that treatment of human colon cancer SW620 cells with curcumin or demethylating drug azacitidine (Aza-dC) reverted the miR-34a promoter to a CpG hypomethylated state. Both treatments increased miR-34a levels and reciprocally decreased miR-34a target NOTCH1 expression in order to inhibit colon cancer growth.⁹⁵  Interestingly, curcumin can suppress human prostate cancer DU145 and 22Rv1 proliferation and invasion in vitro by inhibiting miR-770-5p and miR-1247 expression. These miRNAs are located within the DLK1-DIO3 imprinted gene cluster and further indicate epigenetic control by curcumin in this cell type.⁹⁶ 

Curcumin also regulates non-coding RNA expression at the level of miRNA biogenesis. Work by Liu et al. in A549 lung cancer cells showed that curcumin upregulated expression of the let-7 family member, miR-98, and reciprocally downregulated miR-98 target LIN28.⁹⁷  Suppressed lung cancer cell migration and invasion following curcumin treatment were mediated by blocking LIN28A-induced matrix metalloproteinase proteins MMP2 and MMP9 expression. Intriguingly, LIN28 inhibits let-7 family pri-miRNA and pre-miRNA processing.^98,99  It is conceivable that the let-7/miR-98/LIN28 negative feedback loop is broken in cancerous colon cells, allowing for elevated LIN28 protein and thereby blocking let-7/miR-98 activity via interference with miRNA biogenesis. The re-establishment of let-7/miR-98 expression by curcumin treatment is probably cancer-protective by keeping LIN28 levels in check.

Therapeutic development of curcumin to regulate miRNA expression in order to suppress tumor progression and metastasis is currently being explored in clinical trials. Use of chemically modified curcumin analogs, such as difluorinated curcumin (CDF) designed to increase stability and bioavailability, will facilitate these studies.⁹⁴  Combined treatment approaches involving curcumin and miRNA delivery may also be on the horizon. Indeed, studies by Li et al. to treat glioblastoma verified that co-delivery of miR-138a (an established tumor suppressive cell cycle inhibitor in glioblastoma) with curcumin enhanced the sensitivity of curcumin treatment in human U87 glioblastoma cells by targeting the p38 mitogen-activated protein kinase (MAPK) signaling pathway.¹⁰⁰ 

Indole-3-carbinol (I3C) is a glucosinolate found in cruciferous vegetables (i.e., broccoli, cauliflower, and cabbage). Upon ingestion and contact with gastric acids in the stomach, I3C undergoes in vivo self-condensation to form the bioactive metabolite 3,3′-diindolylmethane (DIM). DIM is well characterized for its antioxidant and anti-inflammatory activities via inhibiting the nuclear factor-κB (NF-κB) and AKT pathways to reduce oxidative stress.¹⁰¹  In addition, DIM is shown to have anti-tumor properties in various tissues, in part by regulating the expression of cancer-associated miRNAs. For example, Li et al. performed miRNA profiling on gemcitabine-sensitive and gemcitabine-resistant pancreatic cancer cell lines. They noted significant downregulation of tumor suppressor miRNAs miR-200b, miR-200c, let-7b, let-7c, let-7d, and let-7e in gemcitabine-resistant cells that possessed pronounced EMT phenotypes.¹⁰²  Treatment of DIM using a modified bioresponse-formulated DIM (BR-DIM) (conferring greater bioavailability) to gemcitabine-resistant MiaPaCa-2, Panc-1, and Aspc-1 human pancreatic cancer cell lines significantly increased the expression of miR-200 (miR-200a, miR-200b, miR-200c), the let-7 family, and miR-146a.^102,103  Forced expression of miR-200 using miRNA mimics in gemcitabine-resistant MiaPaCa-2 cells reversed EMT phenotypes, marked by increased epithelial E-cadherin and suppressed mesenchymal vimentin proteins, in a similar manner as cells exposed to DIM alone.¹⁰² 

Similarly, Li et al. verified that upregulation of miR-146a by DIM plays an important role in pancreatic cancer progression. miR-146a is commonly repressed in human pancreatic cancer cells compared to normal human pancreatic duct epithelial cells.¹⁰³  When pancreatic cancer cells were either treated with DIM or with miR-146a mimics, cancer cell invasion was suppressed through inhibition of the NF-κB and EGFR pathways. Anti-cancer effects were mediated via downregulation of the miR-146a targets EGFR, NF-κB, NF-κB regulatory kinase interleukin 1 receptor-associated kinase 1 (IRAK-1), and metastasis-associated protein 2 (MTA-2).¹⁰³  DIM was previously reported to sensitize pancreatic cancer cells to the EGFR inhibitor erlotinib and induce apoptotic cell death.¹⁰⁴ 

These miRNA studies may provide insight into how DIM mechanistically sensitizes cancer cells to chemotherapeutic agents (i.e., erlotinib) through miRNA-based regulation (e.g., miR-146a–EGFR target interactions). Indeed, Li et al. found that treatment of miR-200b or DIM could increase the sensitivity of human MiaPaCa-2 cells to gemcitabine and significantly decrease pancreatic cancer cell growth.¹⁰²  The combined efficacy of DIM and miR-200 with the chemotherapeutic drug Herceptin (trastuzumab) was also noted. DIM treatment of breast cancer cells induces miR-200 family (miR-200a, miR-200b, and miR-200c) levels.¹⁰²  These tumor-suppressive miRNAs are often decreased in breast cancer tumors and shown to block EMT phenotypes associated with migration and invasion.¹⁰⁵  Interestingly, although Herceptin treatment alone could not modulate miR-200 levels in human MBA-MB-468 breast cancer cells, DIM and Herceptin treatment combined had synergistic effects to significantly increase miR-200a and miR-200b expression.¹⁰⁵ 

DIM is also shown to have cancer-protective effects in the prostate using both miRNA and epigenetic mechanisms. Kong et al. found that treatment of formulated BR-DIM to human LNCaP, C4-2B, and PC3 prostate cancer cell lines increased let-7a, let-7b, let-7c, and let-7d levels and reciprocally reduced let-7 target EZH2, a histone methyltransferase that induces cancer stem cell maintenance, EMT transitions, and cancer progression.¹⁰⁶  The let-7-EZH2 target interaction in BR-DIM-treated cells resulted in reduced prostate cancer cell growth in clonogenic assays and suppressed sphere-forming capacity. Subsequently, this group tested prostate tissue from patients enrolled in phase II clinical trials for prostate cancer prior to radical prostatectomy to test BR-DIM intervention, and confirmed let-7 upregulation and reciprocal EZHA expression following DIM treatment occurred in a clinical setting.¹⁰⁶ 

In addition to the ability of DIM to downregulate miRNA expression in cancer cells, this polyphenol agent is reported to repress the expression of pro-oncogenic miRNAs. DIM treatment decreased miR-21 expression in liver cells.¹⁰⁷  Since miR-21 is commonly overexpressed in a broad range of human cancers, this work could impact therapeutic treatment in other tissues.¹⁹  In fact, DIM may be a natural and potent agent to specifically antagonize miR-21 activity. Molecular docking and dynamics simulations revealed that DIM binds covalently to the precursor pre-miR-21 transcript at a guanine at the sixth position, which is postulated to interfere with DICER processing and miR-21 maturation.¹⁰⁸  Intriguingly, Jin reported that DIM increased miR-21 in MCF-7 breast cancer cells and functioned to inhibit cell growth via suppression of miR-21 target Cell division cycle 25A (CDC25A), a G1/S cell cycle factor.¹⁰⁹  This result underscores the complexity of miR-21 regulation in breast cancer. Contrary reports exist that indicate forced miR-21 expression in MCF-7 cells is oncogenic and promotes breast cancer cell proliferation, migration, and tumor formation.^110–112  DIM is also shown to regulate the oncogenic miR-221, which is elevated in most epithelial cancers (e.g., liver, lung, breast, prostate, brain, stomach, and pancreas).¹¹³  Sarkar et al. found that BR-DIM treatment of MiaPaCa-2 and Panc-1 pancreatic cancer cells resulted in decreased miR-221 expression.¹¹⁴  The subsequent block in pancreatic cell growth and migration following treatment was attributed to an increase in the miR-221 target proteins PTEN, cyclin dependent kinases (CDKs) p27^kip1 and p57^kip2, as well as p53 upregulated modulator of apoptosis (PUMA). Interestingly, this study showed that a synthetic curcumin analogue produced similar results via miR-221-target regulation. Therefore, distinct subclasses of natural polyphenol agents could employ overlapping miRNA mechanisms to achieve cancer-protective effects. These findings further support use of dietary compounds and miRNAs as effective treatment regimens to sensitize cancer cells and inhibit drug-resistant cancers.

We have discussed how dietary factors regulate miRNAs and impact cancer progression in a large range of tissues. Research also indicates that dietary factors have a pronounced influence on the gut microbiota. In turn, gut microbial metabolites, such as butyrate, exert physiological effects on the host via miRNA expression and the epigenome (Figure 1.2). Whole-genome sequencing studies of clinical fecal samples revealed that human individuals across nationalities harbor 10–100 trillion microorganisms, which make up a surprisingly diverse colonization of microbes that include fungi, archaea, protists, and viruses. The vast majority of the microbiome, however, is composed of bacteria, predominantly Firmicutes and Bacteroidetes.^115,116  Selective enrichment of particular subgroups of microbes appears to vary across individuals and likely correlates with diet, drug intake, and health status. Bacterial enterotypes analyzed by Qin et al. indicate that gut-harbored bacterial strains are those that predominantly utilize fermentable substrates in the colon for energy.¹¹⁷  This relationship is symbiotic and the human host benefits from microbial colonization in the gut due to the need for microbial conversion of complex carbohydrates into absorbable substrates, vitamin production, and maintenance of physiological homeostasis. Humans exhibit a wide range of metabolic, immune, neurological, and cognitive disorders when these synergistic interrelationships are disrupted. There is a strong association between gut microbiota dysregulation and human disease, i.e., obesity, diabetes, autoimmune disorders, and, of course, cancer.¹¹⁸ 

The interplay of miRNA regulation and the microbiota has only recently been appreciated. Evidence indicates that human fecal miRNAs, such miR-515-5p and miR-1226-5p, specifically target bacterial genes that control bacterial growth, and thus directly shape the overall composition of the gut microbiota.¹¹⁹  Conversely, studies reveal that gut microbes can control host miRNA expression and influence cancer progression. For instance, gut microbiota generated butyrate induces the expression of miRNAs, such as miR-22, and decreases others, including miR-106b and miR-92a.^120–122  Therefore, non-coding RNA mechanisms involving inter-species gene regulation are extremely important to maintain homeostasis and human health. In this section, we explore how nutrition, the gut microbiota, and the epigenome are closely associated with the role miRNAs play during disease progression, primarily in colorectal cancer.

Colorectal cancer (CRC) is the third most common cancer and possesses the fourth highest cancer mortality rate worldwide.¹²³  Epidemiological studies show that Western-style diet (consumption of energy-dense foods, red and processed meats and low consumption of fresh fruits, vegetables, or whole grains) is one of the highest risk factors for this disease.^124,125  Conversely, adherence to the Mediterranean diet correlates with a 14% reduced CRC risk and diets high in plant-based fiber contribute to decreased rates of CRC.^126,127  Dietary fibers are broken down in the distal intestine by microbial anaerobic fermentation into short-chain fatty acids (SCFAs), namely butyrate, acetate, and propionate. The colon and liver absorb these metabolites as a nutrient energy source. Specifically, high levels of butyrate strongly correlate with decreased CRC risk, and this dietary factor protects against cancer progression.¹²⁸  Indeed, cancer patients harbor a lower number of butyrate-producing bacteria than healthy volunteers.^129,130 

Butyrate is the principle source of energy in normal colon epithelial cells, rather than glucose, and promotes cell growth.¹³¹  Due to the Warburg effect in colorectal cancer cells however, butyrate is not metabolized. Instead, malignant cells accumulate butyrate in the nucleus and this microbial metabolite functions as a HDAC inhibitor to impact gene expression.¹³²  Butyrate influences the epigenome of tumor cells to suppress colon cancer cell proliferation and induce apoptosis via controlling gene expression of the histone deacetylase sirtuin 1 (SIRT1), caspase 3, and NF-κB, as well as the wingless/integrated (WNT) and ERK/MAPK cancer signaling pathways.^118,133  Histone deacetylases (HDAC 1,2,3) are elevated in gastric and colorectal cancers, and therefore inhibition of HDACs by butyrate likely confers protective anti-cancer effects.^134,135 

The anti-tumor effects of the gut microbiota generated butyrate are also associated with the regulation of miRNA expression in CRC cells. Research by Schlormann et al. showed that treatment of LT97 colon adenoma cells with physiological doses of butyrate reduced proliferation and resulted in downregulated miRNA expression of miR-135a/b, let-7a, miR-24, and miR-106b.¹³⁶  Hu et al. found that HCT-116 human colon cancer cells treated with butyrate significantly repressed miR-106b family levels and this correlated with elevated expression of the miR-106b target p21, an established cell cycle progression inhibitor.¹²² 

Hu et al. also investigated the correlation between patients with sporadic colon cancer and increased expression of the oncogenic miR-17-92a cluster.¹²¹  Butyrate treatment of HCT116 and HT29 human colon cancer cells reduced precursor miR-17-92a pri-miRNA transcript levels and mature miR-17-92a cluster members, miR-17, miR-18a, miR-19a/b, miR-20a, and most predominantly miR-92a. (Similar results were obtained using the HDAC inhibitors suberoylanilide hydroxamic acid and valproic acid.) This study indicated that butyrate regulates the miR-17-92a cluster indirectly by epigenetically suppressing MYC expression. MYC is an established transcriptional activator that binds directly to the miR-17-92a promoter C13orf25 transcription.¹³⁷  In vitro experiments to overexpress miR-92a in colon cancer cell lines revealed that this miRNA effectively blocked butyrate's ability to slow the growth and induce apoptosis by targeting the tumor suppressor p57.¹²¹  p57 regulation by butyrate was noted to be extraordinarily complex. p57 transcription is epigenetically modulated due to butyrate's HDAC inhibitor activity but butyrate also promotes p57 translation by reducing miR-92a activity (via c-MYC suppression).¹²¹ 

Butyrate is also shown to play anti-cancer activities in the liver by regulating miRNA expression. Pant et al. found that butyrate incubation significantly increased miR-22 expression in human hepatic cancer Huh 7 cells and resulted in suppression of the miR-22 target SIRT1.¹²⁰  Indeed, miR-22 acts as a tumor suppressive miRNA in a variety of cell types to block proliferation, tumor formation, and metastasis.¹³⁸  Interestingly, the miR-22 target SIRT1 is a histone deacetylase directly suppressed by butyrate.¹³⁹  SIRT1 is highly expressed in hepatoma and colorectal adenocarcinomas and promotes tumor growth in mouse models.¹⁴⁰  The pro-oncogenic activity of SIRT1 is mediated by superoxide dismutase (SOD), which serves to quench reactive oxygen species (ROS) production and suppress apoptosis. Researchers found that miR-22 suppresses SIRT1 expression in hepatic cancer cells resulting in increased ROS levels and apoptosis (via mitochondrial cytochrome c/caspase pathway) and decreased proliferation (via PTEN/p-AKT pathway). Therefore, the HDAC inhibitor activity of butyrate is likely mediated by miR-22 in the liver. Taken together, gut microbial metabolites are potent cancer protective agents against colon and liver cancers by employing epigenetic and non-coding RNA regulation, and which could be translated into effective therapeutics. This work highlights the complex interplay between nutrition (high-fiber diet), the gut microbiota, and their impact on human health.

As discussed, extrinsic factors, such as diet and the gut microbiota, have a huge impact on endogenous miRNAs expression in human cells and cancer progression pathways. Recent evidence indicates that the intake of exogenous miRNAs—those ingested from the diet as food (termed xenomiRs)—can also impact human health and disease pathology (Figure 1.2).¹⁴¹  Foods, including rice, potatoes, cabbage, and breast milk, contain miRNAs that are hypothesized to survive digestion and be available for uptake into recipient cells.¹⁴²  It is proposed that exogenous plant miRNAs enter the body's circulation in their free form or are incorporated into exosomes, and once delivered into recipient cells, plant xenomiRs regulate endogenous mRNA targets to modulate cancer progression. Plant miRNAs are particularly stable because of their unique 2′O-methylated 3′terminal nucleotide modifications, and therefore are well suited to withstand gastric low-pH environments and remain stable in fluids such as the blood.¹⁴³  Studies by Philip et al. verified that prolonged food storage, processing, cooking, and simulated digestion did not destroy plant miRNA activity.¹⁴³  However, the concept of exogenous miRNA delivery from dietary foods remains controversial. O'Neill et al. found that oral administration of plant miRNA resulted in rapid degradation upon digestion.¹⁴⁴  Other groups have reported that plant miRNAs are not expressed in recipient cells at high enough levels to be reliably detected in the blood or tissues and failed to modulate target gene expression in their mouse models.^145–148 

There is growing experimental evidence to support the notion that exogenous miRNAs can be delivered through the diet, which we review in this section. An early study in Chinese subjects reported high levels of plant miRNAs in human sera, including miR-168a, which is enriched in rice (Oryza sativa).¹⁴⁹  Zhang et al. tested in vivo if exogenous rice miRNAs could be delivered to mice through their diet.¹⁴⁹  Animals were fed a chow diet of fresh rice or were gavage-fed with total RNA extracted from fresh rice. As predicted, animals exhibited elevated levels of miR-168a in the stomach, small intestine, and liver, but not in the kidney. Furthermore, in vitro and in vivo evidence indicated that dietary plant miR-168a was selectively loaded into microvesicles (exosomes), entered the blood via the animal gastrointestinal tract, and targeted the low-density lipoprotein receptor adapter protein 1 (LDLRAP1) mRNA in liver cells, which ultimately increased plasma LDL levels.¹⁴⁹ 

In a similar study, Liang et al. administered cruciferous miRNAs extracted from Brassica oleracea (a member of the cabbage family) in the food or as oral gavage to mice.¹⁵⁰  Brassica oleracea miR-172 was resistant to gastrointestinal tract degradation and this miRNA was recovered in the blood, stomach, intestines, spleen, and feces of treated animals up to 72 h after feeding. This study is particularly intriguing from the perspective of nutritional therapeutics, because epidemiological data support a cancer-protective role for cruciferous vegetables such as broccoli.¹⁵¹  Additional studies indicate that miRNAs from Zea mays (corn), Solanum lycopersicum (tomato), Vitis vinifera (grapes), and Glycine max (soybean) are present in the blood of animals exposed to these plants in their diet.^152–154  Therefore, dietary intake of exogenous plant miRNAs may be more widespread than initially appreciated and contribute to recipient gene expression and human health.

Introduction of exogenous miRNAs via the diet is very intriguing both in regards to maintaining human health and as a novel therapeutic approach, particularly plant xenomiRs carrying 2′-O-methylated modifications, which are unusually stable. A study by Chin et al. revealed that plant miRNA miR-159 (present in Arabidopsis thaliana and Glycine max) was detected in microvesicles of human sera.¹⁵⁵  miR-159 levels were inversely correlated with breast cancer incidence and progression in human subjects, highlighting a use for miR-159 as a prognostic marker for breast cancer risk. In vitro and in vivo studies revealed that miR-159 played a tumor suppressive role to block breast cancer cell growth by targeting the WNT signaling transcription factor 7 (TCF7) and lowering MYC oncogenic protein levels.¹⁵⁵  Furthermore, Chin et al. employed an MDA-MB-231 breast cancer xenograft mouse model, in which animals were gavage-fed synthetic 2′-O-methylated miR-159 daily. The treated animals exhibited significantly decreased tumor growth compared to controls. These findings could have major implications for the development of oral vaccines against cancer.

Mlotshwa et al. tested the feasibility of xenomiRs as therapeutics.¹⁵⁶  Delivery of a cocktail of total plant RNA spiked with three established tumor suppressive mammalian miRNAs, miR-34a, miR-143, and miR-145, that were chemically modified to contain 2′-O-methylated 3′-terminal nucleotide plant modifications. This cocktail was delivered by oral gavage to Apc^Min/+ mice, an established animal model for colon cancer. Treated animals showed elevated levels of these miRNAs in the intestine and exhibited significant reduction in tumor load. In short, these results indicate that edible plants could be specifically engineered to produce tumor-suppressive miRNAs for treatment of a wide range of cancers.

The growing list of promising “nutraceuticals” (pharmaceutical-grade and standardized nutrients) discussed in this chapter all have the remarkable ability to protect against a large range of human cancers. These natural dietary agents, which include edible polyphenols, gut microbial metabolites, and plant xenomiRs, mediate their anti-cancer effects via epigenetic regulation and miRNA expression. A greater molecular understanding of how nutritional compounds influence cancer progression pathways is needed as these nutraceuticals move into clinical trials.¹⁵⁷  Development of nutritional compounds and metabolites as disease biomarkers and therapeutic drugs will be transformative in the clinic due to their overall affordability, proven non-toxicity, and ease of administration.

Support was provided to AE-K (PI) by the National Institutes of Health (R21CA175894), Department of Defense (PC131691), EVMS Grant Enhancement Fund, Breedan Adams Foundation, Edmondson Fund, Coach Ray Barlow Prostate Cancer Research Funds, and EVMS Prostate Cancer Research Funds.