SEARCH

SEARCH BY CITATION

Keywords:

  • microRNAs;
  • noncoding RNAs;
  • microarray;
  • quantitative RT-PCR;
  • cancer

Abstract

  1. Top of page
  2. Abstract
  3. Cancer initiation and progression involve microRNAs alterations
  4. Genome-wide high-throughput miRNA expression profiling
  5. Micro RNA expression profiling by real-time quantitative PCR
  6. Variations in the processing machinery involved in the production of miRNAs
  7. Profiling as a tool to identify new noncodingRNAs
  8. Conclusion
  9. Acknowledgements
  10. References

MicroRNAs (miRNAs) represent a new class of small noncoding RNAs (ncRNAs, RNAs that do not codify for proteins) that can regulate gene expression by targeting messenger RNAs of protein coding genes and other ncRNA transcripts. miRNAs were recently found to be involved in the pathophysiology of all types of analyzed human cancers mainly by aberrant gene expression. This is characterized by abnormal levels of expression for mature and/or precursor miRNA transcripts in comparison to the corresponding normal tissues. miRNA profiling has allowed the identification of signatures associated with diagnosis, prognosis and response to treatment of human tumors. Therefore, miRNAs fingerprinting represents a new addition to the tools to be used by medical oncologists. © 2007 Wiley-Liss, Inc.


Cancer initiation and progression involve microRNAs alterations

  1. Top of page
  2. Abstract
  3. Cancer initiation and progression involve microRNAs alterations
  4. Genome-wide high-throughput miRNA expression profiling
  5. Micro RNA expression profiling by real-time quantitative PCR
  6. Variations in the processing machinery involved in the production of miRNAs
  7. Profiling as a tool to identify new noncodingRNAs
  8. Conclusion
  9. Acknowledgements
  10. References

Noncoding RNAs (ncRNAs) range in size from 19 to 25 nucleotides (nt) for the large family of miRNAs that modulate development in several organisms including mammals, up to more than 10,000 nt for RNAs involved in gene silencing in higher eukaryotes.1 First described in C. elegans more than a decade ago,2 over 5,000 members of a new class of small ncRNAs, named microRNAs (miRNAs),3, 4 have been identified in the last 5 years in vertebrates, flies, worms and plants, and even in viruses.5 In humans, the microRNoma (defined as the full complement of miRNAs present in a genome) contains more than 500 experimentally or “in silico” cloned miRNAs and the total number is expected to overpass the one thousand mark.6 miRNAs, ranging in size from 19 to 25 nt, are typically excised from a 60 to 110 nt hairpin precursor (fold-back) RNA (pre-miRNA) structure that is transcribed from a larger primary transcript (pri-miRNA).3 No open reading frame can be identified in the small piece of genome codifying for miRNAs. Functionally, it was shown that miRNAs reduce the levels of many of their target transcripts and the amount of protein encoded by these transcripts by direct and imperfect miRNA::mRNA interaction.7 It has been speculated that miRNAs could regulate ∼30% of the human genome.4

Initially identified by Calin et al. in B cell chronic lymphocytic leukemia (CLL)8, alterations of miRNAs have been detected by different groups in any type of analyzed human tumors. The examples of miRNA alterations in human cancers are growing exponentially with time, and miRNAs were proposed to contribute to oncogenesis functioning as tumor suppressor genes (TSGs, as is the case of miR-15a-16-1 cluster, of miR-143-145 cluster or the let-7 family) or as oncogenes (OGs, as is the case of miR-155, miR-21 or miR17-92 cluster). In a specific tumor, abnormalities in both protein coding genes and in miRNAs were abundantly identified in the recent years (extensive reviews covers the broad topics of miRNAs and cancer, see Refs.4, 9–21) (Fig. 1). Homozygous mutations, the combination of deletion plus mutation or hypomethylation in miRNA genes were recently described.22 Furthermore, the roles of polymorphisms in the complementary sites of target messenger RNAs (mRNAs) in cancer patients23 or individuals with predisposition to other hereditary diseases24 have just started to be understood. Therefore, the main mechanism of microRNoma alteration in cancer cells seems to be represented by aberrant gene expression, characterized by abnormal levels of expression for mature and/or precursor miRNA transcripts in comparison with the corresponding normal tissues. Their expression profiles can be used for the classification, diagnosis and prognosis of human malignancies.

thumbnail image

Figure 1. MicroRNAs as oncogenes and tumor suppressors. The abnormalities found to influence the activity of miRNAs in human tumors are the same as that described to target protein coding genes, including chromosomal rearrangements, genomic amplifications or deletions, mutations and hypermethylation. Activation of oncogenic microRNAs reduced the levels of proteins blocking proliferation and activating apoptosis; by contrast, inactivation of suppressor miRNAs is followed by accumulation of proteins that stimulates proliferation and decrease apoptosis. Upper panel shows the 7q23.2 amplification identified in gliomas, with the consequent overexpression of miR-21and downregulation of the tumor suppressor PTEN gene. MiR-21overexpression is correlated with block of apoptosis and consequent tumor growth. In the lower panel, the 13q14.3 deletion found in chronic lymphocytic leukemia is shown. The down-regulation of miR-15a and miR-16-1 induces over expression of the antiapoptotic BCL2 protein in leukemia cells with reduced apoptosis. The functional consequences are presented on the right side.

Download figure to PowerPoint

The main paradigms for the miRNA involvement in human cancers could be summarized in the following way: (i) miRNAs are altered in every type of analyzed human cancer; (ii) miRNAs act as oncogenes and tumor suppressors; (iii) miRNAs alterations may cause cancer predisposition; (iv) miRNAs profiling is a new diagnostic tool for cancer patients and (v) miRNA profiling represents a prognostic tools for cancer patients. In the present review we will focus on the recent advances (less covered by other reviews on the field) obtained as a result of genome-wide profiling of miRNAs by the 2 most used techniques: miRNA microarray25 and quantitative RT-PCR (qRT-PCR) of precursor26 and mature miRNAs27 and show how these data support the earlier paradigms.

Genome-wide high-throughput miRNA expression profiling

  1. Top of page
  2. Abstract
  3. Cancer initiation and progression involve microRNAs alterations
  4. Genome-wide high-throughput miRNA expression profiling
  5. Micro RNA expression profiling by real-time quantitative PCR
  6. Variations in the processing machinery involved in the production of miRNAs
  7. Profiling as a tool to identify new noncodingRNAs
  8. Conclusion
  9. Acknowledgements
  10. References

miRNA expression profiles derived from large-scale analysis of tumor samples have recently been shown to serve as phenotypic signatures of a particular cancer type.28, 29 With the advent of microarray and analysis of gene expression data, there have been a multitude of efforts to correlate mRNA expression signature with tumor pathology and disease prognosis. Diagnostic potential has also been recognized for miRNA profiling in CLL, B-cell lymphoma, breast cancer, papillary thyroid carcinoma, colorectal cancer and other solid tumors (for an extensive review see Ref.10) and most of the clinical application will depend on accurate assessment of miRNA profiles in human samples. A variety of platform have recently been developed for miRNA expression analyses such as: Northern blot30 (advantages: possibility for discovery of new miRNAs), microarrays31 (advantages: low cost and high throughput), real-time PCR27 (advantages: low cost, superior detection of low-abundance species and high throughput), bead-based hybridization28 (advantages: superior specificity), cloning (miRAGE)32 (advantages: possibility of discovery new miRNAs), single molecule detection33 (advantages: rapid assay), in situ hybridization34 (advantages: ability to visualize miRNA levels in tissue context) and sequencing (miRNA expression atlas based on small RNA library sequencing, see Ref.35). For example, to explore the possibility that additional miRNAs are present in the human genome, Cummins et al.,32 have developed an experimental approach called miRNA serial analysis of gene expression (miRAGE) and used it to perform the largest experimental analysis of human miRNAs to date. Sequence analysis of 273,966 small RNA tags from human colorectal cells allowed the authors to identify 200 known mature miRNAs, 133 novel miRNA candidates, and 112 previously uncharacterized miRNA* forms (the star means the minor products of miRNA biogenesis). In agreement with Michael et al.,36 Cummins et al. showed that the levels of miR-143 and miR-145 were significantly lower in colorectal tumor cells compared to normal colonic cells.

Commonly deregulated miRNAs in human cancers

MiRNAs expression is tissue specific,37 different sets of miRNAs are upregulated or downregulated in tumors of different cellular origins, although it has been reported that the miRNA signatures of different cancer types could share some individual miRNAs. Porkka et al.38 reported that some of the miRNAs that were detected to be up-down-regulated in prostate carcinoma samples are also deregulated in other cancers, for example: downregulation of miR-125a and miR-125b in breast cancer,39 the let-7 miRNAs in lung cancer,40, 41 and miR-143 and miR-145 in several different cancer types.32, 36, 42 Lu et al.28 reported on miRNAs expression profiles (217 miRNAs in 334 samples) and found that the majority of the differentially expressed miRNAs were downregulated in tumor samples compared to normal samples. More recently, Volinia et al.29 showed that miRNAs differently expressed in solid cancer were mostly overexpressed. They describe a large-scale detailed analysis of the miRNA profiles in 540 samples from 6 solid tumors including carcinomas of breast, colon, lung, prostate and stomach and endocrine pancreatic cancers. The authors study the clustering of miRNA expression profiles derived from 228 miRNAs in 363 solid cancer and 177 normal samples. One of the interesting finding of this report is the significant over expression of a large number of miRNAs (such as miR-21, miR-17-5p, mir-155 or miR-191) in at least 2 of the 6 types of analyzed tumors. The exact meaning of these findings is not completely deciphered, but Kulshreshtha et al.43 found that a significant proportion of the hypoxia-regulated miRNAs are also among the list of genes overexpressed in human cancers,29 suggesting a new link between hypoxia and tumorigenesis. Several members exhibit induction in response to hypoxia inducible factor (HIF) activation, thus extending the repertoire of HIF targets beyond translated genes.44

The utility of miRNAs profiling in CLL, lung cancer, breast cancer, thyroid cancer, prostate cancer and B-cell lymphoma is now apparent. For example, in lung cancer and CLL, specific expression signatures are associated with either favorable or poor prognoses. In Table I some miRNAs and their clinical correlation in different types of human malignancies are shown.

Table I. Genome-Wide Microrna Expression Profiling in Human Cancers
Cancer typemiRNA expression
IncreasedDecreasedUnfavorable prognosisReferences
  1. CLL, chronic lymphocytic leukemia; DLBCL, diffuse large lymhoma.

CLL23b15a[DOWNWARDS ARROW] 15a/16-131, 45, 46
24-116-1
14629
155223
195
221
 DLBCL155 [UPWARDS ARROW] 15547, 48, 49
17-92
 B cell lymphoma155 [UPWARDS ARROW] 15547, 48, 49
17-92
19a
92
142
155
221
 Breast cancer21125b 42, 50, 51
155145
 Lung cancer17-92Let-7 family[UPWARDS ARROW] 155; [DOWNWARDS ARROW] let-741
19a
21
92
155
191
205
210
 Prostate cancer2115a/16-1 29, 38, 51
143
145
 Colorectal cancer19a143 29, 32
21145

Chronic lymphocytic leukemia

MiRNA expression profiles can be used to distinguish normal B cells from malignant B cells in patients with CLL. Marton et al.52 showed a global reduction in miRNA expression levels in CLL cells associated to a consistent under expression of miR-181a, let-7a and miR-30d and observed over expression of miR-155 and a set of 5 miRNAs (including miR-15a and miR-16) that are differentially expressed between patients with different clinical outcomes.

The largest study on miRNA profiling in CLL to date was performed by Calin et al.,45 and investigated whether miRNA profiles are associated with known prognostic factors in CLL. The authors evaluated the miRNA expression profiles of 94 samples of CLL cells for which the level of expression of 70-kDa zeta-associated protein (ZAP-70), the mutational status of the IgV(H) gene, and the time from diagnosis to initial treatment were known. They also investigated the genomic sequence of 42 miRNA genes to identify abnormalities. A unique miRNA expression signature composed of 13 genes (of 190 analyzed) differentiated cases of CLL with low levels of ZAP-70 expression from those with high levels and cases with unmutated IgV(H) from those with mutated IgV(H). The authors also identified a germ-line mutation in the miR-16-1/miR-15a primary precursor, which caused low levels of miRNA expression in vitro and in vivo, and was associated with deletion of the normal allele. They showed that germ-line or somatic mutations were found in 5 of 42 sequenced miRNAs in 11 of 75 patients with CLL, but no such mutations were found in 160 subjects without cancer (p < 0.001). This was the first report that a unique miRNA signature is associated with prognostic factors and disease progression in CLL, and mutations in miRNA transcripts are common and may have functional importance. Strongly supporting this paradigm are the results by Raveche et al. describing abnormal miR-16 locus with synteny to human 13q14 linked to CLL in New Zealand Black (NZB) mice.53 Together these 2 studies, one in human CLL and the second in a mouse model of the human disease confirm that miR-16 is the first example of a miRNA involved in cancer predisposition.22, 54 The frequency of this predisposing event in various kindreds with familial cancers should be investigated by large association studies.

Diffuse large B cell lymphoma

Diffuse large B cell lymphoma (DLBCL) is a heterogeneous disease that can be divided into germinal center B cell-like (GCB) and activated B cell-like (ABC) subtypes by gene expression and immunohistochemical profiling. Using microarray analysis on prototypic cell lines, Lawrie et al.47 identified 3 miRNAs, miR-155, miR-21 and miR-221, more highly expressed in ABC-type than GCB-type cell lines. These miRNAs were overexpressed in de novo DLBCL, transformed DLBCL and follicular center lymphoma cases compared to normal B cells. Moreover, using multivariate analysis the authors found that expression of miR-21 was an independent prognostic indicator in de novo DLBCL (p < 0.05). Interestingly, expression levels of both miR-155 and miR-21 were also higher in nonmalignant ABC than in GCB cells.

One cluster of miRNAs, the miR-17-92 polycistron, is located in a region of DNA that is amplified in human B-cell lymphomas at chromosome 13q. He et al.48 compared B-cell lymphoma samples and cell lines to normal tissues, and found that the levels of the primary or mature miRNAs derived from the miR-17-92 locus are often substantially increased in these cancers. Enforced expression of the miR-17-92 cluster acted with c-myc expression to accelerate tumor development in a mouse B-cell lymphoma model. Tumors derived from haematopoietic stem cells expressing a subset of the miR-17-92 cluster and c-myc could be distinguished by an absence of apoptosis that was otherwise prevalent in c-myc-induced lymphomas. Furthermore, the oncogene product and transcription factor, c-MYC, is an activator of the 17-92 miRNA cluster, and this mechanism plays an important role in tumor formation. Similarly, E2F transcription factor family was also found to regulate this cluster.55

Breast cancer

In a recent study, Blenkiron et al.50 report the analysis of 93 primary breast tumors, 5 normal breast samples and 33 breast cancer cell lines using a bead-based flow cytometry miRNA expression profiling technique. From 309 human miRNAs assayed, 133 miRNAs are expressed in human breast tumors, and the authors also tested if miRNAs are differently expressed among cancer subtype. To this purpose, a single sample predictor was used to classify breast tumors into 5 subtypes: Luminal A, Luminal B, Basal-like, HER2+ and Normal-like. On the basis of Agilent and Illumina platforms of mRNA expression data for 86 tumor samples, they were able to classify 51 of the 93 tumors as 16 Basal-like, 15 Luminal A, 9 Luminal B, 5 HER2+ and 6 Normal-like tumors. Moreover, the 9 miRNAs that were found to be differently expressed between Luminal A and Luminal B tumors represent 7 miRNA families. These tumor subtypes show relevant difference not only in mRNA expression profiling but also display distinct clinicopathological characteristics.50 Notably, one miRNA, miR-155, was differentially expressed in ER (estrogen receptor proteins)-negative versus ER-positive tumors, overexpressed in breast tumors as compared to normal controls. They found that the majority of miRNA clusters are coregulated in human breast cancers.

Another interesting recent study performed on miRNA profiling in breast cancer come from Mattie et al.51 In this report the authors demonstrated the feasibility and utility of measuring miRNA profiles from clinically relevant biopsy samples using an optimized high-throughput microarray assay platform. They analyzed a panel of 20 breast cancer samples representing 3 clinically important breast cancer subtypes, defined by ErbB2 and ER status. The important possibility that miRNA signatures may prove to be novel cancer biomarkers is apparent from the study's preliminary finding that unique sets of miRNAs are associated with breast cancers currently defined by their ErbB2 status (let-7f, let-7g, miR-107, mir-10b, miR-126, miR-154 and miR-195) or their ER/PR status (miR-142-5p, miR-200a, miR-205 and miR-25).

The founder study on the topics of miRNAs and breast cancer was published by Iorio et al.42 analyzing 76 breast cancer and 10 normal breast samples to identify miRNAs whose expression is significantly deregulated in cancer versus normal breast tissues. They have indeed identified 29 miRNAs whose expression is significantly deregulated and a smaller set of 15 miRNAs that were able to correctly predict the nature of the analyzed samples (i.e., tumor or normal breast tissue) with 100% accuracy. Among the differentially expressed miRNAs, miR-10b, miR-125b, miR145, miR-21 and miR-155 emerged as the most consistently deregulated in breast cancer. Three of them, miR-10b, miR-125b and miR-145, were downregulated and the remaining two, miR-21 and miR-155, were upregulated, suggesting that they may potentially act as TSGs or oncogenes, respectively. In support to these findings it was recently shown that miR-10b is highly expressed in metastatic breast cancer cells and positively regulates cell migration and invasion.56

Lung cancer

In 2004 after the examination of miRNA profiling in lung cancer, Takamizawa et al.40 identified a significant reduction (>80%) in the expression levels of let-7 observed in 60% (12 of 20) of lung cancer cell lines. Expression levels of let-7 in primary human lung cancer tissues taken directly from surgically treated patients, 44% (7 of 16) of the cases examined were found to exhibit >80% reduction in let-7 expression when compared with that in the corresponding normal lung tissues. A more frequent occurrence of reduced let-7 expression in cell lines in vitro may be related to the inevitable contamination of normal stromal/inflammatory cells in tumor tissues in vivo or, alternatively, this may reflect in vitro selection of cells with reduced let-7 in the process of the establishment of cell lines. The authors were able to classify 143 cases of lung cancer into 2 major groups according to the let-7 expression. Those showing reduced let-7 expression had significantly shorter survival after surgical resections. Impressively, let-7 levels were more powerful predictors of patient survival than age, tumor histology and smoking history. Giving a molecular background to these correlations, are the studies by Johnson et al.57 showing that RAS oncogenes, previously proved to be linked to lung cancer prognosis,58 are targeted by let-7 genes and that the let-7 miRNAs represses cell proliferation pathways in lung cancers.59

In a more recent report Yanaihara et al.41 confirmed the correlation of let-7 levels with disease outcomes in lung cancer; moreover, they were able to identify the involvement of miR-155 as a prognostic marker in lung tumors. In this study the authors investigated the miRNA expression profiles in human lung cancer and miRNA regulation by epigenetic mechanisms and found that the miRNA molecular profile of lung adenocarcinoma correlates with patient survival. They analyzed the miRNA expression in 104 pairs of primary lung cancers and corresponding noncancerous lung tissues. When miRNA expression among lung cancer tissues versus corresponding noncancerous lung tissues was compared, 43 miRNAs had statistical differences in expression between the 2 groups.41

Prostate cancer

Using microarray and qRT-PCR, Mattie et al.,51 measured miRNA levels from 2 different prostate cancer cell lines (LNCaP and PC3 cells) and showed concordance between the 2 platforms. Microarray analyses identified differentially expressed miRNAs when comparing the expression patterns of the hormone sensitive LNCaP and hormone insensitive PC3 prostate cancer cell lines. To verify the accuracy of these microarray results, the authors measured the expression of individual miRNAs by a real-time quantitative TaqMan qRT-PCR method designed to detect mature miRNA sequences. Analysis of the fold changes for these miRNAs showed good concordance (R2 ∼ 0.81) between the microarray and RT-PCR platforms; the only discordance observed was for let-7 miRNA family members, which were found upregulated by microarray measurement and not significantly changed by qRT-PCR measurement. In this work they concluded that the optimized high-throughput miRNA expression profiling offers novel biomarker identification from typically small clinical samples such as breast and prostate cancer biopsies.51

To identify the miRNA signature specific for prostate cancer, Porkka et al.38 analyzed the miRNA expression profiling of 319 miRNAs in 6 prostate cancer cell lines, 9 prostate cancer xenografts samples, 4 benign prostatic hyperplasia (BPH) and 9 prostate carcinoma samples by using an oligonucleotide array hybridization method. When they compared the expression of miRNAs between the BPH samples and the carcinoma samples, 51 miRNAs were found to be differentially expressed, and of these 51 miRNAs, 37 were downregulated. Twenty-two of these 37 miRNAs showed decreased expression in all carcinoma samples, whereas 15 of them were only downregulated in the hormone-refractory carcinomas compared to BPH samples. In addition to these 37 downregulated miRNAs, 14 miRNAs were upregulated: 8 of them showed increased expression in all carcinomas (miR-202, miR-210, miR-296, miR-320, miR-370, miR-373, miR-498 and miR-503).

Micro RNA expression profiling by real-time quantitative PCR

  1. Top of page
  2. Abstract
  3. Cancer initiation and progression involve microRNAs alterations
  4. Genome-wide high-throughput miRNA expression profiling
  5. Micro RNA expression profiling by real-time quantitative PCR
  6. Variations in the processing machinery involved in the production of miRNAs
  7. Profiling as a tool to identify new noncodingRNAs
  8. Conclusion
  9. Acknowledgements
  10. References

Real-time qRT PCR is the gold standard of nucleic acid quantification due to the sensitivity and specificity of the PCR. miRNAs are challenging molecules to quantify by PCR since the miRNA precursors exist as a stable hairpin and the ∼21 nt mature miRNA is roughly the size of a standard PCR primer. Despite these challenges, successful PCR methods have been developed to amplify and quantify both the precursor and mature miRNA. In 2004, the Schmittgen laboratory developed the first real-time PCR assay for miRNA.26 This technique amplified and quantified the miRNA precursors as means to predict the mature miRNA expression. Using gene specific primers to prime the reverse transcription reaction and a thermostable reverse transcriptase, successful qRT PCR was performed on the stable hairpins.

In 2005 Chen et al., developed a method to quantify the mature miRNA.27 This technique used a loop primer to prime the reverse transcription reaction. The purpose of the loop primer is several fold (i) increased secondary structure to enhance the priming of small miRNA, (ii) prevents binding of the primer to the miRNA precursors and/or genomic DNA and (iii) increases the size of the cDNA to enhance the PCR amplification. The assay uses TaqMan MGB probes to detect the PCR product. Use of TaqMan probes is desirable to discriminate individual miRNA isoforms from nearly identical miRNAs (e.g., members of the let-7 family).27, 60 The TaqMan probe also improves the sensitivity of the assay by reducing the background from low levels of detection. Quantification of miRNAs using real-time qRT PCR has been used for several applications including gene expression profiling, validation of cDNA array data, quantification of individual miRNAs in a given study and quantifying precursor and mature miRNA levels. Examples of studies that have used real-time RT PCR as a screening tool for miRNA in cancer are discussed later.

Profiling of miRNA expression using real-time RT PCR

Several studies have applied the concept of a low-density array (i.e., profiling of several hundreds of miRNAs using real-time PCR) in both tumor tissues and in cancer cell lines. We profiled the expression of 222 miRNA precursors in 32 human cancer cell lines.60 Hierarchical clustering analysis of the miRNA precursor expression data revealed that most of the cell lines clustered into their respective tissues from which each cell line was derived. Another study utilized real-time PCR to profile 241 mature miRNA in the NCI60 panel of cancer cell lines.37 These authors found that most miRNAs were expressed at lower levels in cancer cell lines compared to the corresponding normal tissue. Hierarchical clustering analysis of miRNA expression revealed that cell lines of the hematologic, colon, central nervous system and melanoma classes clustered in a manner that reflected their tissue of origin. Thus both the studies37, 60 demonstrated that the expression data from a relatively small number of genes are sufficient to cluster the cell lines to the tissue of origin. This emphasizes the degree of informational content that is unique to miRNAs.

Lee et al., profiled the expression of over 200 miRNA precursors in a number of pancreas tumors, normal pancreas and adjacent benign tissues.61 These data indicated that a large number of miRNAs has increased expression in pancreatic cancer compared to normal and adjacent benign pancreas (Table II). Furthermore, the miRNA precursor expression pattern may be used to predict whether a given tissue is a tumor or not. The expression of over 200 precursor and mature miRNAs was profiled in specimens of hepatocellular carcinoma, and in normal and adjacent benign liver.65 The results of our study demonstrated that miRNA expression was predominately reduced in hepatocellular carcinoma compared to adjacent benign liver. Patients with reduced miRNA expression had a poorer survival and a set of 19 miRNAs significantly correlated with disease outcome (Table II).

Table II. miRNA Expression Profiles by Quantitative RT-PCR in Human Cancers
Cancer typemiRNA expression
IncreasedDecreasedUnfavorable prognosisReferences
  1. SE, seminoma; SS, spermatocytic seminoma.

Pancreas221375  61
301
100
376a
21
155
212
 Hepatocellular18101[DOWNWARDS ARROW] let-7c[DOWNWARDS ARROW] 125b65
21139[DOWNWARDS ARROW] let-7g[DOWNWARDS ARROW] 139
33150[DOWNWARDS ARROW] 26b[DOWNWARDS ARROW] 148a
130b199a/199a*[DOWNWARDS ARROW] 29c[DOWNWARDS ARROW] 150
135a199b[DOWNWARDS ARROW] 30e-3p[DOWNWARDS ARROW] 200c
221200b[DOWNWARDS ARROW] 31[DOWNWARDS ARROW] 220
301214[DOWNWARDS ARROW] 99a[DOWNWARDS ARROW] 221
223[DOWNWARDS ARROW] 99b[DOWNWARDS ARROW] 345
[DOWNWARDS ARROW] 100[DOWNWARDS ARROW] 372
[DOWNWARDS ARROW] 377
 CLL2115a[UPWARDS ARROW] 29b62
15516[UPWARDS ARROW] 29c
92[UPWARDS ARROW] 150
150[UPWARDS ARROW] 223
 Colorectal19a30-3p 63
20124a
21129
31133b
96145
135b328
183
 Germ cell21 (SE)133a (SE, SS) 64
155 (SE, SS)145 (SE, SS)
17-5p (SS)146 (SE, SS)
92 (SS)
106 (SS)
19a (SE, SS)
29a (SS)

The miRNA profile in CLL was studied using 2 different independent and complimentary methods; real-time PCR and cloning.47 Both approaches show the significant over expression of miR-21 and miR-155 in patients with CLL. In patients with CLL, they were also able to identify significant deregulation of miR-150 and miR-92. Moreover, they detected in 11% of cases a marked decrease of miR-15a and miR-16. In this study it was identified a set of miRNAs whose expression correlates with biologic parameters of prognostic relevance, particularly with the mutational status of the rearranged immunoglobulin heavy-chain variable-region (IgVH) gene. A small RNAs analysis in CLL reveals a deregulation of miRNA expression and novel miRNA candidates of putative relevance in CLL pathogenesis.52

A group of 13 miRNAs were altered in colorectal tumors compared to adjacent normal colorectal tissues (Table II).63 Among the most significantly reduced miRNA in colorectal cancer is miR-145. MiR-145 and another miRNA located on 5q32 (miR-143) were reduced at the mature miRNA level, but not the miRNA precursor level, in colorectal cancer.66 MiR-31 expression was able to distinguish Stage III from Stage IV tumor samples.65

Testicular germ cell tumors (GCTs) may be subdivided into Type II GCTs and Type III GCTs. The expression of 156 miRNAs was profiled by real-time PCR in 69 Type II and Type III GCTs.64 For the most part, the GCTs clustered separately among each subgroup. An overall reduced expression was found for the majority of the miRNAs in the spermatocytic seminomas and seminomas compared to normal testes. However, miRNAs are downregulated upon differentiation towards, for example, teratoma. MiR-372 and miR-373 were previously shown to enhance the proliferation and tumorigenesis of primary human cells that harbor both oncogenic RAS and active wild-type p53 by neutralizing p53-mediated CDK inhibition.67 In agreement with this previous study, the miR-371-373 cluster was expressed in GCTs that harbor wild-type p53.66

Variations in the processing machinery involved in the production of miRNAs

  1. Top of page
  2. Abstract
  3. Cancer initiation and progression involve microRNAs alterations
  4. Genome-wide high-throughput miRNA expression profiling
  5. Micro RNA expression profiling by real-time quantitative PCR
  6. Variations in the processing machinery involved in the production of miRNAs
  7. Profiling as a tool to identify new noncodingRNAs
  8. Conclusion
  9. Acknowledgements
  10. References

The causes for the widespread disruption of miRNA expression in tumoral cells are only partially known and, very probably, in each tumor various abnormalities could contribute to the microRNoma expression profile. Until now, at least 3 different mechanisms correlate with and could explain the abnormal expression of specific miRNA genes: (i) the miRNAs location at cancer associated genomic regions (CAGR) (for details, see Refs.68–70); (ii) epigenetic regulation of miRNA expression (for details, see Refs.71, 72) and (iii) abnormalities in miRNA processing genes and proteins. We will further detail only the last mechanism.

miRNA processing involves a large number of genes and proteins, and only part of them has been identified.3, 4 The long primary transcripts produced by the RNA polymerase II, are then processed by 2 ribonucleases III. One is located in the nucleus and called Drosha and the other called Dicer in the cytoplasm. The final miRNA duplex is incorporated into a large protein complex called RISC (RNA-induced silencing complex), whose core includes components of the Argonaute protein family. Recently it was found that the impaired miRNA processing enhances cellular transformation and tumorigenesis.73 Further strengthening these findings, a large study of non-small-cell-lung cancer (NSCLC) patients, found that Dicer, but not Drosha, expression levels were reduced in a fraction of lung cancers and correlated with shortened postoperative survival and poorly differentiated status, which is a mark of bad prognosis.74, 75 Furthermore, Chiosea et al.76 by using various techniques including gene array analysis showed a transient up-regulation of Dicer along with down-regulation of most genes encoding miRNA machinery proteins in bronchioloalveolar carcinomas and in adenocarcinomas. Immunohistochemically, Dicer was upregulated in atypical adenomatous hyperplasia and bronchioloalveolar carcinoma and downregulated in areas of invasion and in advanced adenocarcinoma. A fraction of adenocarcinomas lose Dicer as a result of deletions at the Dicer locus.76 The same group,77 found by using a gene array analysis of 16 normal prostate tissue samples, 64 organ-confined, and 4 metastatic prostate adenocarcinomas, an up-regulation of major components of the miRNA machinery, including Dicer, in metastatic prostate adenocarcinoma. Immunohistochemical studies on a tissue microarray consisting of 232 prostate specimens confirmed up-regulation of Dicer in prostatic intraepithelial neoplasia and in 81% of prostate adenocarcinoma. The increased Dicer level in prostate adenocarcinoma correlated with clinical stage, lymph node status, and Gleason score. Western blot analysis of benign and neoplastic prostate cell lines further confirmed Dicer up-regulation in prostate adenocarcinoma. Dicer up-regulation may explain an almost global increase of miRNA expression in prostate adenocarcinoma. The presence of upregulated miRNA machinery may predict the susceptibility of prostate adenocarcinoma to RNA interference-based therapy.77

The miRNA precursor26, 60 and mature miRNA assays27 were used to quantify the precursor and mature miRNA expression, respectively, in a number of cancer cell lines and tissues from patients with hepatocellular and pancreas cancer.78 This comparison allowed us to determine the degree to which the miRNA precursors are being processed to mature miRNA. The data demonstrated that the miRNA precursors were processed to a greater extent in pancreas tumors, while in hepatocellular cancer a greater percentage of the precursor miRNA were not processed to mature miRNA.78 These findings suggest that the increased expression of miRNA in pancreatic cancer61 is regulated at the level of transcription with no apparent alterations in processing but the reduced miRNA expression in hepatocellular cancer65, 79 may result from a reduction in miRNA processing.

Profiling as a tool to identify new noncodingRNAs

  1. Top of page
  2. Abstract
  3. Cancer initiation and progression involve microRNAs alterations
  4. Genome-wide high-throughput miRNA expression profiling
  5. Micro RNA expression profiling by real-time quantitative PCR
  6. Variations in the processing machinery involved in the production of miRNAs
  7. Profiling as a tool to identify new noncodingRNAs
  8. Conclusion
  9. Acknowledgements
  10. References

Using a custom designed microarray, Calin et al. recently found that a new class of ncRNAs, the ultraconserved genes (UCGs), are altered in human leukemias and carcinomas.80 One of the most intriguing characteristics of miRNAs is the almost complete conservation of orthologous genes6: for example, the active molecules of miR-16-1/miR-15a cluster are completely conserved in human, mouse and rat and highly conserved in 9 out of 10 primate species sequenced.6 Comparative sequence analysis represents an essential tool in identification of highly conserved genomic sequences. Several of these are considered not genic (not producing a transcript) and were named conserved noncoding sequences. A special subset of conserved sequences named ultraconserved regions (UCRs) includes, by definition, pieces of human genome located both intragenic and intergenic that are absolutely conserved (100% identity with no insertions or deletions) between orthologous regions of the human, rat and mouse genomes.81 Because of the huge conservation, the UCRs may have fundamental functional importance for the ontogeny of mammals and other vertebrates. Recently we reported that a large fraction of genomic UCRs of about 200 bp in length are encoding a particular set of ncRNAs and are altered in human CLLs. We show that UCRs are frequently located at fragile sites and genomic regions involved in cancers, and genome-wide UCRs profiling reveals distinct signatures in human leukemias and carcinomas. Furthermore, we identified UCRs whose expression may be regulated by miRNAs abnormally expressed in CLL (Fig. 2). These findings argue that noncoding genes are involved in tumorigenesis at a greater extent as thought before and offer the perspective of identification of signatures associated with diagnosis, prognosis and response to treatment composed by various categories of ncRNA genes.80

thumbnail image

Figure 2. The molecular architecture of CLL deciphered by profiling. Deletion or down-regulation of miR-15a/miR-16-1 cluster located at chromosome 13q14.3 represents an early event directly involved in the regulation of BCL2 expression.82 During the evolution of malignant clones, other miRNAs are deleted (such as miR-29) or overexpressed (such as miR-155) contributing to the aggressiveness of B-CLL. Such abnormalities influence the expression of other protein-coding genes, as was reported that miRNA-29 and miRNA-181 directly down-modulate TCL1 oncogene62 or, by targeting other ncRNAs, such as ultraconserved genes (UCGs), as was reported for miR-155.80 The consequences of this steady accumulation of abnormalities are represented initially by a low apoptosis and high survival of malignant B cells and lately by the evolution of more aggressive clones with a higher proliferative capacity.

Download figure to PowerPoint

Conclusion

  1. Top of page
  2. Abstract
  3. Cancer initiation and progression involve microRNAs alterations
  4. Genome-wide high-throughput miRNA expression profiling
  5. Micro RNA expression profiling by real-time quantitative PCR
  6. Variations in the processing machinery involved in the production of miRNAs
  7. Profiling as a tool to identify new noncodingRNAs
  8. Conclusion
  9. Acknowledgements
  10. References

MiRNAs are involved in tumor development and progression. Genome-wide miRNA profiling by various techniques just started to the identification of new diagnostic and prognostic tools for cancer patients. The theoretical rationale for using miRNAs in cancer treatment is based on the fact that miRNAs are natural antisense interactors that regulate many genes involved in eukaryotic survival and proliferation, and by the proven fact that miRNA over expression in cancer cells has a pathogenic effect. It is time to look forward for the first ncRNAs as clinically in-use tumor markers and therapeutic targets as well as new drugs!

Acknowledgements

  1. Top of page
  2. Abstract
  3. Cancer initiation and progression involve microRNAs alterations
  4. Genome-wide high-throughput miRNA expression profiling
  5. Micro RNA expression profiling by real-time quantitative PCR
  6. Variations in the processing machinery involved in the production of miRNAs
  7. Profiling as a tool to identify new noncodingRNAs
  8. Conclusion
  9. Acknowledgements
  10. References

Dr. Barbarotto is supported by an American-Italian Cancer Foundation Fellowship. Dr. Calin is supported by the CLL Global Research Foundation, and, in part, as a University of Texas System Regents Research Scholar and as a Fellow of The University of Texas, MD Anderson Research Trust. Dr. Calin highly acknowledges the 1997–1998 Yamagiwa-Yoshida Memorial International Cancer Study Grant by the International Union Against Cancer (Geneva, Switzerland) at the University of Ferrara, Italy, which helped him to start the career in molecular genetics.

References

  1. Top of page
  2. Abstract
  3. Cancer initiation and progression involve microRNAs alterations
  4. Genome-wide high-throughput miRNA expression profiling
  5. Micro RNA expression profiling by real-time quantitative PCR
  6. Variations in the processing machinery involved in the production of miRNAs
  7. Profiling as a tool to identify new noncodingRNAs
  8. Conclusion
  9. Acknowledgements
  10. References