Advantages of Noncoding RNAs in Molecular Diagnosis

Abstract

Noncoding RNAs contribute to physiological processes by regulating many intracellular molecules participating in the life-supporting mechanisms of development, differentiation, and regeneration as well as by disrupting various signaling mechanisms such as disease development and progression and tumor growth. Because microRNAs (miRNAs) target and regulate the functions of key proteins, it is very useful to identify specific miRNAs that contribute to cellular functions and to clarify the roles of their target molecules as diagnostic and therapeutic strategies for cancer prognosis and treatment. In this section, the roles of miRNAs in various cancers and the processes leading to the identification of their target molecules are described, and the latest diagnostic strategies using miRNAs are discussed with specific examples.

1. Introduction

Noncoding RNAs (ncRNAs) comprise the majority of gene transcripts in eukaryotes. ncRNAs play little biological role, but microRNAs (miRNAs), first reported in 1993, have important roles in gene expression [1, 2]. One of the most well-known miRNAs, let-7, a heterochronic gene in C. elegans, is considered important for developmental differentiation and regulates the larva-to-adult transition during larval development [3]. As a summary of the biosynthetic process of miRNAs, double-stranded RNA, the source of RNA interference, is sequentially processed into single-stranded RNA fragments of 21–23 nucleotides (nt) by the action of various enzymes [4, 5]. Let-7 encodes a small 21-nt ncRNA that binds to the 3′-untranslated region (UTR) of lin genes, resulting in RNA–RNA interactions and negative regulation of mRNA [6]. More than 30,000 miRNAs have been identified from more than 250 species, ranging from prokaryotes to eukaryotes, invertebrates to vertebrates, and even viruses [7, 8]. More than 5000 of these miRNAs have been identified in humans [miRBase: The microRNA database; cited 2018 Aug; Available from: http://www.mirbase.org].

miRNAs function as posttranscriptional regulators (negative regulators) of various gene expressions by binding to complementary sequences, primarily in the 3′-UTR of their target genes. miRNAs can target a large number of genes, and the mRNAs targeted by miRNAs mRNAs targeted by miRNAs account for more than 60% of all genes. Most miRNAs are encoded within the introns of coding mRNAs, but some can also be found within the 3′-UTR sequences of noncoding or coding mRNAs but can also be found in the 3′-UTR sequence of noncoding or coding mRNAs [9]. When miRNAs bind to target mRNAs, these are destabilized by deadenylation factors becoming more susceptible to degradation. As a result, mRNA is degraded; however, sometimes the mRNA is not degraded and translation is regulated, a mechanism that is not yet unclear. Alterations in the genetic structure of miRNA-associated regions are known to occur in many pathologies, especially cancer. Such chromosomal aberrations include amplifications (e.g. miR-26a in gliomas), deletions (e.g. miR-15a/16–1 in chronic lymphocytic leukemia), mutations (e.g. miR-125a in breast cancer), translocations [e.g. miR-125b in acute myeloid leukemia with t(2;11)(p21;q23)], single-nucleotide polymorphisms (e.g. miR-608 in colorectal adenocarcinoma), and heterozygous deletions (such as the 14q32 cluster in acute lymphoblastic leukemia) [10, 11, 12, 13, 14, 15]. These changes can occur not only in the sequence of the mature miRNA itself but also in the promoter region/primary transcript (pri-miRNA) sequence or in the miRNA binding site of the target gene, and as somatic or germline mutations [16, 17, 18]. A number of transcription factors, as well as vital protein-coding genes, influence the expression levels of miRNAs, and this regulatory action is assumed to be particularly important for tissue specificity and developmental stage control. Conversely, loss of Dicer, DGCR8, Drosha, and Argonaute2(Ago2), components of pathways involved in the miRNA biosynthesis mechanism, is known to result in the inability to sustain life and early pregnancy death due to severe developmental defects. [19, 20, 21, 22]. Loss of reproductive function due to the deletion of components involved in miRNA biosynthesis, and developmental defects in the heart, lungs, and neuromuscular tissue have been reported [23, 24, 25, 26, 27, 28, 29, 30]. Dicer mutations have also been reported to cause tumor development and progression [31, 32]. Dysfunction of the miRNA biogenesis pathway causes the disruption of important life-support mechanisms associated with various diseases, among which cancer development and progression are strongly influenced. In particular, abnormal expression of Dicer and Drosha is known to lead to poor prognosis in various cancers [33, 34, 35, 36, 37, 38, 39]. Although there is a vast amount of evidence that miRNAs play fundamental roles in the mechanisms of many diseases, including cancer, unfortunately, they have not yet been fully exploited in diagnostic and therapeutic settings. It is therefore necessary to determine how miRNAs can potentially be utilized in clinical practice in the future. miRNAs likely have the greatest and most direct potential as novel biomarkers of diagnosis and prognosis, as well as predictors of therapeutic efficacy. It is also clear that miRNAs regulate or disrupt the expression of numerous target mRNAs, resulting in maintenance of vital activity or disease, which cannot be resolved by a single miRNA. Therefore, gene expression analysis by miRNA expression profiling in disease must accurately discriminate between disease diagnosis and tumor stage of development [40].

A more accurate diagnosis of tumors can be obtained by morphological histopathology using formalin-fixed paraffin-embedded (FFPE) tissues and profiling of expressed proteins via immunohistochemical staining. Genetic analysis using FFPE is also emerging as being important for therapeutic selection. However, the degradation of nucleic acids is a problem for FFPE, and nucleic acids extracted from aged tissues are degraded [41]. A particularly attractive property of miRNAs is their stability against chemical and enzymatic degradation. This implies that they can be purified from routinely prepared FFPE biopsy materials and measured robustly [42]. As a result, retrospective miRNA expression studies are underway, using FFPE as a conserved and valuable resource. The stability and clinical utility of miRNAs are due to their presence in extracellular biological fluids, including blood, as well as in tissues and cells. Tumor-associated miRNAs are found at higher levels in the serum and plasma of cancer patients than in healthy controls. Therefore, the use of miRNAs as potential noninvasive biomarkers of disease has attracted much attention, and miRNAs have been detected in many biological fluids, including plasma, serum, tears, urine, cerebrospinal fluid, breast milk, and saliva [43, 44]. The widespread use of biomarker testing, which is noninvasive and more reliable than the conventional histopathological diagnosis of cancer that invasive to the patient, could pave the way for screening programs, improve the early detection rate of disease, and enhance cancer prevention. Despite the growing functional importance of miRNAs as diagnostic and therapeutic strategies for cancer, there are still many knowledge gaps despite the potential to develop miRNA markers in the future.

2. Role of miRNAs as oncogenes and tumor suppressor genes

The role of miRNAs in regulating the genes involved in cancer growth and progression may be to suppress cancer progression primarily through suppressing target gene expression or to inhibit the function of genes that suppress cancer progression, therefore positively affecting cancer progression. The results obtained by detailed profiling of miRNA expression in cancer according to organ or cancer differentiation level revealed that the expression pattern of miRNAs varies depending on the organ, histological type, and differentiation level of cancer, indicating that the role of miRNAs is highly segmented [40]. Many miRNAs function in a suppressive manner in cancer. The expression of the aforementioned let-7 miRNA is downregulated in lung cancer, targeting and suppressing the expression of Ras, an important oncogene. This indicates that let-7 may function as a cancer suppressor gene [45]. The expression of miR-15 and miR-16 is absent or decreased in chronic lymphocytic leukemia, suggesting they act in a tumor-suppressive manner by targeting B cell lymphoma-2, an antiapoptotic factor [46]. Human miR-373 binds to the E-cadherin (CDH1) promoter and induces gene expression [47]. In addition to gene silencing through base pairing with DNA and mRNA target sequences, miRNAs are known to inhibit the function of regulatory proteins (decoy activity). miR-328 binds to poly C-binding protein 2 (PCBP2), also known as heterogeneous ribonucleoprotein E2. This binding does not involve the seed region of the miRNA, but it inhibits its interaction with the target mRNA. In chronic myeloid leukemia, downregulation of miR-328 inhibits myeloid differentiation via PCBP2, leading to tumor progression [48]. It is also important to understand the molecular mechanisms of miRNAs to define the mechanism of their suppressive properties against cancer. miRNAs primarily target the 3′-UTR of mRNAs and downregulate the cytoplasmic expression of protein-coding genes. As evidence for this, miRNAs are usually localized in the nucleus [49]. In addition to the 3′-UTR, other genetic regions at the DNA and RNA levels, that is, the 5′-UTR, promoter, and coding regions, as well as proteins can be targeted (Figure 1) [50, 51, 52]. By binding to various functional regions of genes, miRNAs can upregulate or downregulate gene expression, interact with other ncRNAs and RNA transcripts, and modulate biological networks [53, 54]. It has become clear that the actions of these miRNAs on their target molecules lead to the regulation of gene expression [55]. It has been shown that miRNAs not only bind to key sites in other RNAs but also to promoter regions at the DNA level, thereby affecting transcriptional activity. As a specific example, in terms of localization sites, miRNAs have been shown to function both in the cytoplasm and nucleus [56]. Investigation of the subcellular localization of miR-29b revealed it was mostly localized to the nucleus. The characteristic hexanucleotide terminal motif of miR-29b functions as a metastable nuclear localization element that induces the nuclear localization of the miRNA or small-interference RNA to which it binds [49]. It has been shown that miRNAs bind to adenylate-uridylate-rich elements and upregulate translation in cells that have undergone cell cycle arrest. This suggests that miRNAs can affect not only cell differentiation and proliferation stages in a broad sense but also the cell cycle [57].

Figure 1.

Predicted binding sites for miRNAs in genes. miRNAs have binding sites not only in the 3′UTR but also in every site of the gene.

3. Expression dynamics of miRNAs in cancer and expectations for clinical application

Cancer is a malignant disease in which a population of cells with genetically abnormal protein expression proliferate autonomously, destroying host tissues as they grow, metastasize, and settle in other organs via vascular flow. As a result, cancer causes the dysfunction of normal tissues and ultimately leads to organ failure. There is evidence that carcinogenesis is a multistep process in which malignant cells accumulate epigenetic and genetic changes that gradually transform normal cells into malignant cells. Triggers for carcinogenesis include tobacco, chemical exogenous stimuli such as chemicals and dust from air pollution, and accumulation of genetic damage due to the aging of individual cells [58]. The proliferation of cancer cells progresses by causing structural changes in genes that promote cancer cell proliferation (oncogenes), genes that prevent their proliferation (tumor suppressor genes), and abnormal gene expression [59, 60, 61, 62]. Malignant cells lose cell discipline in morphology and differentiation and acquire autonomous proliferation, antiapoptosis, and invasiveness properties. miRNAs have been analyzed in various human malignant tumors and can be involved in the pathophysiology of benign and malignant tumors in various ways. Although there are differences in miRNA expression profiles between normal and tumor tissues, many of these miRNAs are only indirectly altered by genetic changes that occur during carcinogenesis, epigenomic dynamics, and physiological changes in cell biology and do not directly trigger tumor development. Differentially expressed miRNAs between tumor and normal tissues have been identified in lymphomas, breast cancer, lung cancer, papillary thyroid cancer, glioblastoma, hepatocellular carcinoma, pancreatic tumor, pituitary adenoma, cervical cancer, brain tumors, prostate cancer (PCa), kidney and bladder cancer, and colon cancer [63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76]. The involvement of miRNAs in the carcinogenic process may involve changes in the components of miRNA biosynthesis pathway, and overall suppression of miRNA maturation by mutations in DROSHA, DGCR8, or DICER has been shown to affect the transcription and translation of multiple mRNAs, ultimately promoting cell transformation and tumorigenesis [77, 78, 79]. The fact that conditional loss of DICER or DGCR8 inhibits bone and cartilage growth and differentiation in mice and in vitro has been demonstrated, suggesting that components involved in the miRNA biosynthesis mechanism are also key molecules in carcinogenesis and cancer progression [80, 81]. The expression of Dicer, Exportin-5, TRKRA, TARBP2, DGCR8, and Argonaut-2 changes dynamically with changes in miRNA expression during the oncogenic phase in pancreatic cancer, suggesting that these miRNA biosynthesis-related molecules can cause cellular changes, including carcinogenesis [82]. Decreased expression of some components that play an important role in miRNA biogenesis, such as DICER, is associated with cancer prognosis and sensitivity to molecularly targeted drugs and has been shown to correlate with short survival in non-small cell lung cancer while increasing sensitivity to gefitinib [36, 83]. The methylation of DICER and DROSHA is also a potential biomarker for lung cancer [84]. Some miRNA precursors, known as mirtrons, have the unique feature of entering the miRNA biosynthesis pathway without undergoing Drosha-mediated cleavage while mimicking the structural features of pre-miRNAs. They are susceptible to epigenetic regulation and have been shown to exert epigenetic effects through interactions between various miRNAs and host genes in urothelial cell carcinoma [85, 86, 87]. Breast cancer susceptibility gene 1 (BRCA1) promotes the processing of primary miRNA transcripts and increases the expression of both precursors and mature forms of let-7a-1, miR-16-1, miR-145, and miR-34a [88]. BRCA1 interacts with Smad3, p53, and DHX9 RNA helicases by directly binding to DROSHA and DDX5 in the DROSHA microprocessor complex that regulates miRNA maturation and recognizes RNA secondary structures through its DNA-binding domain, which interacts with pri-miRNA via the DNA-binding domain.

Thus, from their discovery to the present, miRNAs have brought great promise and success in cancer diagnosis, prognosis, and treatment both in clinical practice and experimental medicine. Powerful technologies, such as miRNA arrays, short RNA deep sequencing, specific quantitative polymerase chain reaction (PCR) of miRNAs, and antisense technologies, have revealed the functions of numerous miRNAs. It is expected that they will continue to have a major impact on clinical oncology in terms of cancer diagnosis and prognostic treatment. Because miRNAs are important factors that define cellular characteristics, direct detection of miRNAs and miRNA target molecules from algorithms based on the vast amount of data obtained by profiling analysis could be used as a valuable tool for cancer diagnosis. miRNA expression profiling has been considered more efficient and informative than conventional mRNA profiling in classifying tumors with respect to their tissue of origin and differentiation [89, 90, 91, 92, 93]. These findings indicate that in clinical practice, miRNAs are useful biomarkers for tracking the tissue of origin of tumors [94, 95, 96, 97]. miRNA expression profiles can be an important source of information for cancer progression prediction and prognosis. It has been shown that high expression of miR-21 and miR-155 in lung cancer and colorectal cancer (CRC) predicts efficient recurrence, poor prognosis, and reduced survival [98, 99, 100, 101]. Functional analysis of miRNAs through experimental processes has been found to be valuable for therapeutic strategies based on the regulation of miRNA activity. As a therapeutic strategy against miRNAs acting as oncogenes (oncomirs), anti-miRNAs are expected to exert therapeutic effects by blocking oncomirs through the action of oligonucleotides that are complementary to miRNAs. For them to function stably, they must be chemically modified to enhance their stability and deliverability to the target organs. As is the case with many drugs used in cancer therapy, issues related to the delivery of miRNA-containing compounds, their stability in vivo, and biological defense measures against toxicity are important issues that must be resolved as miRNAs progress from experimental to clinical trials. However, because miRNAs have multiple targets, inhibition of target molecules by miRNAs may cause serious side effects. These issues are expected to be resolved and applied clinically in the future.

4. Functions of miRNAs in various organs

Given their properties, miRNAs have many target molecules and each performs different functions during cell differentiation and proliferation. Therefore, miRNAs are expected to perform characteristic functions in various organs. Here, the functions of miRNAs in typical cancers are introduced.

4.1 Breast cancer

miRNA profiling studies in breast cancer have been conducted from various perspectives. The first miRNA shown to be highly expressed in metastatic breast cancer was miR-10b (using mouse and human cells), with clinical correlation in primary breast cancer [102]. Subsequent studies confirmed the elevated levels of miR-10b, miR-34a, and miR-155 in patients with metastatic breast cancer [103]. Furthermore, miR-10b and miR-373 were recently shown to be upregulated in lymph node-positive breast cancer [104]. In addition, miRNAs from previous miRNA profiling studies have been shown to play potential roles as tumor suppressor genes or oncogenes in breast cancer [65, 105, 106]. The decreased expression of many genes involved in cancer development and progression has become evident, and miRNAs and their target molecules that are downregulated in breast cancer tissues and cell lines have become the focus of research. Breast cancer is a cancerous tissue with high histopathological heterogeneity. Identifying miRNAs and their target molecules that are specifically expressed and function in this context and have diagnostic value is therefore difficult. miRNA profiling should facilitate the accurate classification of tumor subtypes and provide useful additional information to complement current classification methods and guide therapeutic decisions. Studies have nested to predict miRNAs in accordance with previous subtype classifications. Recent analyses of miRNAs have identified predicted miRNAs corresponding to hormone receptor and HER2 subtypes, including ER (miR-342, miR-299, miR-217, miR-190, miR-135b, and miR-218), PR (miR-520g, miR-377, miR- 527-518a, miR-520f-520c), and HER2 (miR-520d, miR-181c, miR-302c, miR-376b, and miR-30e). These expression kinetics classified cases with 100% accuracy compared to immunohistochemistry results [107]. Further classification is expected to increase miRNA profiling accuracy and the development of new diagnostic markers and therapeutic strategies based on miRNA target molecules.

4.2 Prostate cancer

Prostate cancer (PCa) is one of the most common causes of cancer-related death in men. Although most PCa progress very slowly and remain within the primary tumor, some have an aggressive behavior that can spread to other organs via the vascular system. Prostate-specific antigen (PSA) and prostate acid phosphatase (PAP) are useful tumor markers for prostate cancer. However, PAP has low specificity and PSA, despite its superior specificity, is inadequate for predicting the tumor types in which it is not elevated and histological differentiation and progression. Many studies have indicated that miRNAs are involved in the development of PCa [108, 109, 110, 111, 112]. Aberrant expression of miRNAs has been detected in PCa cell lines, in experimental carcinogenesis using cancer xenografts, and in clinical samples obtained from patients [113, 114, 115, 116, 117, 118]. These changes suggest that miRNAs play an important role in the pathogenesis of PCa.

Many miRNAs are expressed as oncogenes. For example, miR-375, miR-106a, and miR-194 have been reported to play important roles in tumor formation, proliferation, and progression [119, 120, 121]. When the expression of tumor suppressor miRNAs decreases, the expression of oncogenes increases, promoting cancer growth and progression. Experimental introduction of precursors of these miRNAs into cells and overexpression of these miRNAs to reproduce normal expression patterns suppressed the growth, invasion, clonogenesis, and metastatic potential of PCa cells. The most well-known miRNAs are miR-149-5p, miR-7-5p, miR-122, miR-124-3p, miR-188, and miR-129 [122, 123, 124, 125, 126, 127]. Downregulation of these miRNAs is associated with increased cell proliferation, migration, and chemotherapy resistance, primarily through the regulation of cell growth and proliferation via the mitogen-activated protein kinase pathway.

Clinically practical markers in PCa include serum PSA measurement and digital rectal examination; however, their diagnostic value is limited and does not extend to prognostic prediction based on cancer progression or assessment of differentiation related to tumor growth potential. Human glandular kallikrein 2, urokinase plasminogen activator and its receptor, transforming growth factor-β1, and interleukin 6 and its receptor are biomarkers with high sensitivity and specificity for the early diagnosis of PCa. miRNAs are also clinically promising diagnostic markers for this cancer. miRNAs can be used as diagnostic and disease-monitoring markers for the presence of minimal residual disease and early detection of recurrence. Furthermore, the establishment of a diagnostic algorithm combining multiple oncomir and tumor suppressor miRNAs is expected to enable a more stepwise diagnosis of PCa. In this cancer, histopathological diagnosis by biopsy is highly invasive and often fails to detect microscopic cancers owing to the extremely small amount of visible tissue. To compensate for the shortcomings of local biopsy pathology diagnosis, the availability of profiling tests with circulating miRNAs has the potential to predict the onset of cancer and the degree of differentiation through its combination algorithm. From the time of diagnosis and initiation of treatment, the detection of circulating miRNAs could contribute to the early detection of hormone refractory PCa. The role of cancer stem cells in PCa has been elucidated in previous studies. The CD44-positive PCa stem cell population has tumorigenic and metastatic potential, which has been mentioned in conjunction with other useful markers for cancer stem cells [128]. Functional analysis of miRNAs has shown that miR-9-5p, miR-34a, miR-373, miR-708, and miR-520c directly target CD44 suppressing its expression [129, 130, 131, 132, 133]. The fact that the expression levels of these miRNAs are decreased in PCa suggests an increased potential for tumor invasion and metastasis. Profiling of miRNAs in PCa holds promise as a more precise diagnostic tool before performing prostate biopsy.

4.3 CRC

CRC is a common cancer worldwide and is the second most common cause of cancer-related deaths in the Western world. Although the 5-year survival rate for early-stage cancer exceeds 90%, it decreases to approximately 70% for advanced cancer and to approximately 10% when distant metastases are present. Although it is certainly a carcinoma for which early diagnosis is very useful, in reality early detection is less than 40% due to insufficient noninvasive screening. Although direct diagnosis by endoscopy is most effective in CRC, it often relies on a fecal occult blood test because of its invasive nature. However, the fecal occult blood reaction can be positive even for inflammatory or nonneoplastic polyps and lacks specificity in terms of cancer diagnosis. If CRC can be found to have miRNA expression characteristics, analysis of miRNAs using stool and the detection of circulating miRNAs may be useful in detecting marginally present tumor cells. Therefore, we searched for miRNAs that were significantly elevated in CRC using stool and plasma and found a number of miRNAs, including miR-17, miR-18a, miR-19a, miR-21, miR-92a, miR-200c, miR-221, and miR-106a as diagnostic markers [134, 135, 136, 137, 138]. miR-141 is a miRNA that functions in an inhibitory manner against colon cancer cells in terms of cell proliferation, invasiveness, and migration [139, 140, 141, 142]. As a circulating miRNA, miR-141, together with miR-200 and miR-143, could be an independent prognostic marker for advanced CRC [143, 144].

CRC progresses from adenoma to colorectal adenocarcinoma as precancerous lesions. Investigation of miRNA expression in staged histological types has revealed that the expression of several miRNAs is altered in precancerous lesions, adenomas, and CRC tissues compared to that in normal colorectal tissue. This suggests that some miRNAs are involved in the process of CRC and may be important biomarkers in the transition of tumor tissue from benign to malignant. In addition, miRNAs are now being extracted from stored FFPE tissue specimens and retrospectively profiled for miRNA expression using next-generation sequencing (NGS) to understand the origin of CRC development and metastatic disease in cases with multiple lesions. The most studied miRNAs with respect to clinical prognosis are miR-21, miR-143, and miR-145. Although miR-21 suppresses the expression of various tumor suppressor genes, it also affects cell proliferation, apoptosis, invasion, and tumor progression. Increased miR-21 expression in CRC tumor tissue correlates with increased metastatic potential of the primary tumor and is associated with shorter disease-free and overall survival. Owing to the antiapoptotic effect of miR-21, its overexpression was also found to affect therapeutic efficacy by reducing the efficacy of 5-fluorouracil (5-FU) chemotherapy [145, 146, 147, 148, 149, 150]. Decreased expression of miR-143 and miR-145 in CRC tumor tissue has been observed in several studies, suggesting that they behave as tumor suppressors [132, 151, 152, 153, 154]. Their decreased expression has been associated with increased tumor growth and angiogenesis, and it is clinically associated with shorter disease-free survival. The expression of miR-143 directly correlates with the sensitivity of the colon cancer cell line HCT116 to 5-FU treatment. miR-215 is involved in the cell proliferative capacity of CRC and is sensitive to chemotherapy and molecular-targeted therapy. miR-215 is also involved in the prognostic prediction of CRC [155, 156, 157, 158, 159, 160].

Although biomarker analysis using stool as a method to directly capture tumor cells in CRC is open for consideration, the establishment of a methodology for good-quality nucleic acid extraction that enables good analysis and sample quality maintenance is required to obtain reliable data due to the heterogeneity of sample processing. In contrast, for the prediction of treatment potential and prognosis for advanced cancer, miRNA profiling can be performed by nucleic acid extraction from FFPE of excised tumor tissue, and sufficient technology is being established for accuracy and reliability.

5. Analysis of circulating miRNAs for their usefulness as biomarkers

In addition to serum and plasma, various body fluids, such as saliva, cerebrospinal fluid, and body cavity fluid, are candidates as analyzable materials for the detection of circulating miRNAs. However, in samples of these fluids from cancer patients, not only tumor cells but also blood cells, macrophages, exosomes, and microparticles are present; where the miRNAs are present, there is also the issue of how many of the targeted tumor-derived miRNAs are included or how reliable the analysis is based on which miRNAs are included. Nevertheless, regardless of the members, each of whom may possess miRNAs, if the changes caused by the coexistence of cancer can be captured for cancer patients, the usefulness of profiling will be well maintained. It is well known that analysis of exosome-derived miRNAs can detect more cancer-relevant miRNAs than analysis using untreated serum. An interesting aspect of the quantitative analysis of miRNAs in exosomes and microparticles is that they may be involved in cell–cell interactions, stimulate intracellular signaling, and modulate metabolic functions and homeostasis in a small number of stem cells. Quantitative analysis of circulating miRNAs is most useful for miRNAs quantitative real-time PCR (qRT-PCR), because it detects miRNAs from a very small amount of nucleic acids. Although NGS is superior for expression profiling, it requires a sufficient amount of nucleic acids and is likely to lack reliability due to detection errors or low sensitivity. Therefore, a single-miRNA assay using the TaqMan method should be performed on miRNAs with presumed abnormal expression at the transcriptional level, as detected from microarray analysis or NGS analysis based on FFPE, followed by qRT-PCR validation. Capturing abnormal miRNA expression in cancer is a highly useful diagnostic strategy but, because capturing circulating miRNAs involves sampling a physiological diverse environment, we may be quantifying miRNAs derived not only from cancer but also from blood or other organs. In addition, miRNAs derived not only from cancer but also from coexisting diseases may be captured; therefore, it is not always the case that miRNAs of cancer cells are captured. In other words, the possibility of observing the dynamics of miRNAs in a broad sense, including in the cancer microenvironment, must be fully understood. Nevertheless, the analysis of circulating miRNAs is highly useful as a potential candidate for blood-based biomarkers that can be tested repeatedly. Circulating miRNAs are promising diagnostic biomarkers owing to their high information content and disease-specific regulation. miRNAs inhibit the translation of various proteins involved in disease-related signaling pathways, including malignant and benign diseases, and alterations in physiological states, thereby inhibiting specific signaling pathways. Therefore, the analysis of the complex profiling of multiple miRNAs involved in complex pathways is expected to provide more specific diagnostic algorithms.

Biomarker development involves several important issues. The use of various analytical technology platforms, serum or plasma samples, extraction of miRNAs from whole serum/plasma or exosomes/microvesicles separated by specific disease-related markers, and sample collection, storage, and processing, all need to be addressed. Depending on the time point of blood sample collection (presurgery, during surgery, and posttreatment) and patient treatment (surgery, chemotherapy, immunotherapy, and/or drug therapy), the expression profile of circulating miRNAs may change. It is also important to establish endogenous miRNA control and assess its consistency with the patient’s clinicopathological data to assess prognosis.

6. Conclusions

miRNA expression profiling for cancer diagnosis can be subdivided into miRNA expression patterns based on signaling common to cancer, such as cell proliferation, invasion, migration, and antiapoptotic effects, and miRNA expression patterns according to various organ or tumor differentiation levels and tissue types. In addition, the expression patterns of miRNAs in cancers of various organs are different according to cancer differentiation levels and tissue types. In addition, the degree of differentiation and tissue classification are often heterogeneous among cancers of various organs, and the estimation of miRNAs based on the differentiation of each organ or developmental stage and the miRNAs that may be targeted may be important determinants for organ identification in so-called primary unknown cancers. A vast number of miRNA expression analyses in various cancers are expected to be reported in the near future. As more cancer-specific miRNAs become known, they will be systematically classified and miRNA expression profiling consistent with conventional immunohistochemical staining of cancer tissues is expected to have practical application.

Acknowledgments

We thank Aya Sugimoto for her excellent research skills and support of our work.

Conflict of interest

The authors declare no conflict of interest.