Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2012 Jun 8.
Published in final edited form as: Genes Chromosomes Cancer. 2010 Dec 16;50(3):196–206. doi: 10.1002/gcc.20844

MicroRNAs and Colon and Rectal Cancer: Differential Expression by Tumor Location and Subtype

Martha L Slattery 1,*, Erica Wolff 1, Michael D Hoffman 1, Daniel F Pellatt 1, Brett Milash 2, Roger K Wolff 1
PMCID: PMC3370677  NIHMSID: NIHMS377448  PMID: 21213373

Abstract

MicroRNAs are thought to have an impact on cell proliferation, apoptosis, stress responses, maintenance of stem cell potency, and metabolism and are, therefore, important in the carcinogenic process. In this study, we examined 40 colon tumors, 30 rectal tumors, and 30 normal tissue samples (10 proximal colon, 10 distal colon, and 10 rectal paired with cancer cases) to examine miRNA expression profiles in colon and rectal tumors. MiRNA expression levels were adjusted for multiple comparisons; tumor tissue was compared with noncancerous tissue from the same site. A comparison of normal tissue showed 287 unique miRNAs that were significantly differentially expressed at the 1.5-fold level and 73 with over a two-fold difference in expression between colon and rectal tissue. Examination of miRNAs that were significantly differentially expressed at the 1.5-fold level by tumor phenotype showed 143 unique miRNAs differentially expression for microsatellite instability positive (MSI+) colon tumors; 129 unique miRNAs differentially expressed for CpG Island Methylator Phenotype positive (CIMP+) colon tumors; 135 miRNAs were differentially expressed for KRAS2-mutated colon tumors, and 139 miRNAs were differentially expressed for TP53-mutated colon tumors. Similar numbers of differentially expressed miRNAs were observed for rectal tumors, although the miRNAs differentially expressed differed. There were 129 unique miRNAs for CIMP+, 143 unique miRNAs for KRAS2-mutated, and 136 unique miRNAs for TP53-mutated rectal tumors. These results suggest the importance of miRNAs in colorectal cancer and the need for studies that can confirm these results and provide insight into the diet, lifestyle, and genetic factors that influence miRNA expression.

Introduction

MicroRNAs (miRNA) are a class of small regulatory RNAs that mediate post-transcriptional silencing of specific target mRNAs. These small non-coding RNAs are 18–24 nucleotides long (Smalheiser and Torvik, 2004). Like protein-coding genes, miRNAs show complex patterns of tissue-and disease-specific expression that merge with their ability to regulate many targets that are crucial to the carcinogenic process. They have been shown to have an impact on cell proliferation, apoptosis, stress responses, maintenance of stem cell potency, and metabolism. miRNAs are thought to have both oncogene and tumor suppressor gene regulatory roles. Although studies now show the importance of miRNAs in the etiology of colorectal cancer (CRC), few individuals and few individual miRNAs have been evaluated. Often data are from cell lines rather than population-based studies. Studies have not evaluated colon and rectal cancer separately to determine associations with miRNAs expression.

In 2003, it was first reported that differential expression of miRNAs was associated with CRC (Michael et al., 2003). At that time, Michael and colleagues illustrated tumor suppressor-like activity for miR-143 and miR-145 in colon cancers and hypothesized that these miRNAs were targeting ERK5 and IRS1 (Michael et al., 2003). Over 100 miRNAs have been implicated in CRC since then (Yang et al., 2009). Generally, studies have focused on describing differential expression of miRNA and have yet to delve into the etiology of that differential expression. Akao and coworkers showed downregulation of let-7a-1 in colon tumors and in cell lines (Akao et al., 2006); miR-34a has been shown to inhibit cell proliferation and be downregulated in colon cancer cells (Yang et al., 2009), whereas miR-31, miR-96, miR-31, miR-135b, and miR-183 were upregulated in CRC cells in the same study (Yang et al., 2009).

MiRNA associations with specific tumor markers also have been reported. TP53, which is frequently mutated in CRC, has been a target of much of this research. Loss of miR-34 has been shown to impair TP53-mediated cell death while over-expressed miR-34 leads to apoptosis (He et al., 2007a; He et al., 2007b; He et al., 2007c). miRNAs also have been shown to be involved with oncogene regulation and with tumor suppressor gene regulation. MiR-143 and let-7 expression levels have been associated with KRAS2 mutations (Schetter et al., 2010).

The ability to use paraffin-embedded tissue is essential to study tumor changes at the population level. A study of breast cancer demonstrated the ability to use paraffin-embedded tissue to examine miRNA expression (Hui et al., 2009). In that study, 365 miRNAs were examined in 40 paraffin-embedded breast tumors. Technical reproducibility was high, with intrasample correlation above 0.9 with 93% accuracy. Paraffin-embedded tissue also was compared to fresh-frozen tissue from the same breast cancer samples, and a correlation of 0.94 was found between miRNA expression levels.

In this study, we evaluate 70 colon and rectal tumor samples from a population-based sample of colon and rectal cancer and compare them with a smaller subset of 30 normal (20 colon and 10 rectal) tissue samples obtained from paraffin-embedded tissue from the colon and rectal cancer cases. We evaluate differences in expression of miRNA by colorectal site as well as by differences in specific tumor alterations, ie. CIMP+, MSI+, KRAS2, and TP53. We compare these specific tumor alterations with normal tissue expression levels as well as with expression levels observed for other tumor alterations to identify unique miRNA expression that accompanies specific disease pathways as indicated by unique molecular changes in tumors.

Matherial and Methods

Samples come from participants in two case– control studies, one of colon cancer and one of rectal cancer. The first study, a population-based case–control study of colon cancer, identified cases of first primary colon cancer between October 1, 1991 and September 30, 1994 (Slattery et al., 1997b). The second study, with identical data collection methods, included cases with cancer of the rectosigmoid junction or rectum identified between May 1997 and May 2001. Eligible criteria included: being between 30 and 79 years old at time of diagnosis (cases only), English speaking, mentally competent to complete the interview, having had no previous history of CRC, and having no known (as indicated on the pathology report) familial adenomatous polyposis, ulcerative colitis, or Crohn's disease. All participants included in this study resided in Utah or were members of the Kaiser Permanente Medical Care Program (KPMCP) in Northern California. Tumor tissue was obtained for 97% of all diagnosed cases in Utah and 85% of diagnosed cases at KPMCP. Study details have been previously reported (Slattery et al., 1997a; Slattery et al., 2003b)

Tumor site within the colon was defined as proximal (cecum through transverse colon) or distal (splenic flexure, descending, and sigmoid colon) or rectal (rectosigmoid junction and rectum) as classified by the tumor registries. Samples were selected for this study based on known unique tumor characteristics so that comparison could be made for individuals with only CIMP+, MSI+, KRAS2, and TP53+ tumors. MSI+ tumors were not selected for rectal cancer given the rarity of these alternations in rectal cancer.

RNA Isolation From Paraffin-Embedded Tissue and Microarray Aanalysis

Formalin-fixed paraffin embedded (FFPE) tissue representing 30 matched normal tissue samples, 20 from colon cancer cases and 10 from rectal cancer cases, and 70 colon and rectal tumor samples were used. Five-micron tissue slices from these samples were mounted on glass slides and stained with either H&E for tumor reference or Aniline Blue for dissection. Depending on tumor size, we dissected one to four slides per FFPE tissue block to isolate enough RNA to generate expression data on the Agilent human miRNA microarray platform. The area on each slide dissected ranged from 2 mm × 3 mm to 10 mm × 4 mm. Total RNA containing miRNA was extracted, isolated, and purified using the RecoverAll Total Nucleic Acid isolation kit (Ambion, Austin TX). RNA yields were determined using a NanoDrop spectrophotometer. RNA yields ranged from 100 ng to 1 ug total RNA.

For all 100 tissue samples described above, total RNA preps were labeled with cy3 and hybridized to Agilent Human miRNA V3.0 Microarrays. Quality control reports from the arrays were examined, and one array was excluded due to high background noise on the array (approximately three times higher than all other arrays). Normalized intensity values from the remaining arrays were filtered to remove nonuniform outlier and population outlier features. The remaining intensity values were shifted (to eliminate negative values), log transformed (log base 2), and quantile normalized in R software using the BioConductor package (Gentleman et al., 2004). Hierarchical clustering using Ward's method in Spotfire DecisionSite software [http://spotfire.tibco.com] revealed a batch effect that corresponded to the date on which the microarrays were processed. The batch effect was corrected using ComBat software (Johnson et al., 2007).

We also prepared technical replicates by isolating two independent preps of RNAs from sequential sections derived from four normal colon blocks. These paired replicas were assessed for expression levels of the 866 human miRNAs present on the Agilent microarrays. All four pairs were highly reproducible with Pearson's correlations of greater than 0.95. It appears that our collections of tumor sections are able to provide high quality miRNAs, allowing us to generate global miRNA expression data.

Statistical Analysis

The log-scale normalized intensities where loaded into GeneSifter software for initial microarray analysis [http://www.geospiza.com]. Up- and downregulation of miRNA was determined based on site-specific, i.e., colon and rectal, expression levels in the normal tissue. For group comparisons, we determined differential expression in specific types of epigenetic and genetic tumor molecular phenotype based on site-specific normal tissue expression. We also assessed expression levels from specific tumor alterations to all other tumor alterations for colon and rectal cancer separately. We used the following criteria for all assessments: t-test P value of 0.05, adjusted P value by Benjamini and Hochberg (Benjamini, 1995), and a 1.5-fold difference in expression between comparison groups, i.e., tumor and normal and specific alterations versus other alteration. We report findings that are statistically significant using these criteria.

The Agilent human miRNA microarray V.3.0. targets 866 unique human miRNAs. The miRNA array contains 2370 unique human sequences and averages six features per probe sequence. miRNAs are referred to using standard nomenclature used in the miRBase database (Griffiths-Jones et al., 2006). Briefly, the first three letters signifies the organism, followed by a unique number. The number is followed by a dash and number (i.e., −1 or −2) if more than one loci codes for the miRNA. A lettered suffix denotes closely related miRNAs. If 2 miRNAs are coded for by the same precursor product, then the minor product is assigned the suffix (*). If predominant/minor product status is not known then the suffix −5p and −3p are used to denote 5′ arm and 3′ arm.

Results

We detected statistically significant 1.5-fold differences in expression for 287 miRNA probes comparing noncancerous normal tissue from the proximal and distal colon and rectal sites. Figure 1 shows the top 73 miRNAs with over a twofold difference in expression. It is of interest that differences in miRNA expression generally were not observed for distal colon and proximal colon tissue, but differences were common between rectal and colon tissue. The one exception to this was miR-150 that had more than a twofold decrease in expression for distal colon tissue, while rectal and proximal colon tissue showed similar levels of expression. Nineteen of the top 73 miRNA had multiple probes displaying similar site-specific differential expression. Because of these differences in expression levels of miRNAs between colon and rectal normal tissue, all subsequent tumor comparisons were based on-site specific comparisons.

Figure 1.

Figure 1

Open in a new tab

Comparison of miRNA expression from normal, proximal, and distal colon and rectal cancers.

One hundred forty-three (143) unique miRNA probes showed more than a 50% difference in expression when MSI+ colon tumors were compared with normal colon tissue. Of these, 48 miRNAs had two or more independent probes that were significantly up- or downregulated. The top 30 of these differentially expressed miRNAs are shown in Table 1; a complete listing of all statistically significant associations can be found online. Eight miRNAs appeared to be uniquely differentially expressed at a level of 1.5-fold or greater in MSI+ tumors compared to other tumor phenotypes.

Table 1. Associations Between MSI and miRNA Expression Levels.

Colon MSI vs. normal tissue Colon MSI vs. other tumor changes
Gene name Ratio Direction adj. P-value Gene name Ratio Direction adj. P-value
hsa-miR-7 4.08 Up 0.000 hsa-miR-139–3p 2.50 Up 0.028
hsa-miR-183 3.33 Up 0.001 hsa-miR-223 1.99 Up 0.028
hsa-miR-195 3.25 Down 0.001 hsa-let-7g* 1.87 Down 0.028
hsa-miR-21 2.98 Up 0.008 hsa-miR-552 1.78 Down 0.028
hsa-miR-183 2.95 Up 0.002 hsa-miR-223 1.72 Up 0.028
hsa-miR-21 2.91 Up 0.015 hsa-miR-424* 1.67 Down 0.029
hsa-miR-145 2.71 Down 0.002 hsa-miR-24–2* 1.58 Down 0.047
hsa-miR-93 2.69 Up 0.001 hsa-miR-370 1.57 Up 0.028
hsa-miR-451 2.67 Down 0.036
hsa-miR-1 2.67 Down 0.001
hsa-miR-497 2.58 Down 0.001
hsa-miR-17 2.57 Up 0.007
hsa-miR-195 2.57 Down 0.001
hsa-let-7i 2.57 Up 0.002
hsa-miR-34a 2.56 Up 0.005
hsa-miR-451 2.49 Down 0.045
hsa-let-7i 2.49 Up 0.003
hsa-miR-34a 2.47 Up 0.002
hsa-miR-650 2.42 Down 0.001
hsa-miR-215 2.39 Down 0.016
hsa-miR-223 2.37 Up 0.016
hsa-miR-150 2.35 Down 0.027
hsa-miR-224 2.32 Up 0.000
hsa-miR-215 2.3 Down 0.016
hsa-miR-20a 2.26 Up 0.019
hsa-miR-145 2.26 Down 0.002
hsa-miR-25 2.22 Up 0.009
hsa-miR-130b 2.19 Up 0.001
hsa-miR-20a 2.19 Up 0.019
hsa-miR-92a 2.15 Up 0.013
Open in a new tab

Among colon tumors that were CIMP+, 230 miRNA probes, 129 of which were unique miRNAs, had greater than 50% differential expression compared to normal colon tissue. The top 30 miRNAs based on ratio of difference in expression compared with normal are shown in Table 2, and a complete listing of all significant miRNAs is included in the Supporting Information Tables. Comparison of CIMP+ colon tumors with other mutation-containing colon tumors showed eight unique miRNAs associated with CIMP+ tumors. For rectal CIMP+ tumors, there were 129 unique miRNAs (204 total probes) with a 50% or greater difference in expression compared to normal tissue. Of these, miR-135-b had the highest upregulation, with a 5.18 and 4.74 ratio (independent probes) compared with normal tissue. Comparing CIMP+ rectal tumors with all other mutated rectal tumors, there were eight unique miRNAs. MiR-146a had the greatest upregulation at 3.21 and 2.69 ratios.

Table 2. Associations Between CIMP+ Tumors and miRNA Expression Levels.

Colon CIMP+ vs normal CIMP + vs. other tumor markers Rectal CIMP+ vs normal Rectal CIMP+ vs other tumor markers
Gene name Ratio Direction adj. P-value Gene name Ratio Direction adj. P-value Gene name Ratio Direction adj. P-value Gene name Ratio Direction adj. P-value
hsa-miR-183 3.81 Up 0.0000 hsa-miR-24-2* 3.25 Down 0.030 hsa-miR-l35b 5.18 Up 0.0001 hsa-miR-l46a 3.21 Up 0.026
hsa-miR-650 3.69 Down 0.0000 hsa-miR-3 1 1.92 Up 0.030 hsa-miR-224 4.81 Up 0.0001 hsa-miR-l46a 2.69 Up 0.047
hsa-miR-183 3.4 Up 0.0002 hsa-miR-892b 1.84 Down 0.035 hsa-miR-l35b 4.74 Up 0.0007 hsa-miR-625 1.74 Up 0.047
hsa-miR-224 3.37 Up 0.0001 hsa-miR-646 1.69 Down 0.030 hsa-miR-21 4.25 Up 0.0029 hsa-miR-55lb* 1.63 Down 0.047
hsa-miR-892b 3.27 Down 0.0004 hsa-miR-3 1* 1.62 Up 0.030 hsa-miR-375 3.70 Down 0.0477 hsa-miR-l 297 1.60 Up 0.047
hsa-miR-1305 3.21 Down 0.0002 hsa-miR-l5l-3p 1.55 Up 0.035 hsa-miR-7 3.65 Up 0.0132 hsa-miR-625 1.57 Up 0.047
hsa-miR-21 3.2 Up 0.0005 hsa-miR-338-3p 1.53 Up 0.094 hsa-miR-224 3.54 Up 0.0005 hsa-miR-769-3p 1.53 Down 0.047
hsa-miR-892b 3.18 Down 0.0003 hsa-miR-492 1.53 Up 0.030 hsa-miR-l46a 3.42 Up 0.0030 hsa-miR-l 87* 1.53 Down 0.047
hsa-miR-29a 3.17 Up 0.0001 hsa-miR-375 3.41 Down 0.0149 hsa-miR-l 537 1.53 Down 0.047
hsa-miR-34a 3.12 Up 0.0006 hsa-miR-l 3.41 Down 0.0043 hsa-miR-635 1.52 Up 0.047
hsa-miR-l39-3p 3.11 Down 0.0204 hsa-miR-21 3.40 Up 0.0146
hsa-miR-24-2* 3.11 Down 0.0335 hsa-miR-l 33b 3.30 Down 0.0008
hsa-miR-21 3.07 Up 0.0016 hsa-miR-145 3.28 Down 0.0094
hsa-miR-29b 3.01 Up 0.0003 hsa-miR-17 3.10 Up 0.0014
hsa-miR-1288 2.99 Down 0.0003 hsa-miR-17 3.07 Up 0.0017
hsa-miR-29a 2.96 Up 0.0002 hsa-miR-20a 3.04 Up 0.0014
hsa-miR-92a 2.94 Up 0.0003 hsa-miR-183 3.04 Up 0.0053
hsa-miR-7 2.91 Up 0.0006 hsa-miR-20a 2.98 Up 0.0013
hsa-miR-34a 2.88 Up 0.0004 hsa-miR-650 2.95 Down 0.0094
hsa-miR-25 2.88 Up 0.0004 hsa-miR-183 2.80 Up 0.0048
hsa-miR-92a 2.88 Up 0.0005 hsa-miR-20b 2.78 Up 0.0048
hsa-miR-21* 2.86 Up 0.0001 hsa-miR-145 2.75 Down 0.0139
hsa-miR-3 1 2.86 Up 0.0066 hsa-miR-l46a 2.73 Up 0.0115
hsa-miR-20a 2.84 Up 0.0007 hsa-miR-183 2.71 Up 0.0056
hsa-miR-17 2.83 Up 0.0007 hsa-miR-34a 2.69 Up 0.0013
hsa-miR-1305 2.76 Down 0.0006 hsa-miR-552 2.53 Up 0.0030
hsa-miR-93 2.73 Up 0.0006 hsa-miR-93 2.51 Up 0.0013
hsa-miR-96 2.72 Up 0.0002 hsa-miR-29a 2.47 Up 0.021 1
hsa-miR-1288 2.64 Down 0.0004 hsa-miR-96 2.44 Up 0.0043
hsa-miR-892b 2.64 Down 0.0002 hsa-miR-497 2.41 Down 0.0133
Open in a new tab

Comparison of KRAS2-mutated colon tumors with normal colon tissue showed 135 probes differentially expressed at or above the 1.5-fold level; the top 30 of these miRNAs based on ratio of differences in expression are shown in Table 3. MiR-650 had the highest level of upregulation, being 3.02 greater than normal tissue. There were no significant differences between KRAS2 mutations and other colon tumor mutations after adjusting for multiple comparisons. For rectal cancer, 87 of the 143 probes that showed statistically significant differences in miRNA expression were unique miRNAs. The top three miRNAs, all with duplicate probes, were miR-224 (upregulated ratio of 5.33 and 4.08), miR-135b (upregulated ratio of 4.42 and 4.29), and miR-451 (downregulated ratio of 4.27 and 4.15). Comparison of KRAS2-mutated rectal tumors with other rectal tumor mutations showed 10 unique miRNAs that were differentially expressed. MiR-146a had the largest degree of downregulation, with a ratio 2.63 and 2.43 compared to other tumor markers.

Table 3. Associations Between KRAS2 Mutations in Colon and Rectal Tumors and miRNA Expression Levels.

Colon KRAS2 vs. normal tissue Rectal KRAS2 vs. normal tissue Rectal KRAS2 vs. other tumor markers
Gene name Ratio Direction adj. P-value Gene name Ratio Direction adj .P-value Gene name Ratio Direction adj. P-value
hsa-miR-650 3.02 Down 0.001 hsa-miR-224 5.33 Up 0.0000 hsa-miR-l46a 2.63 Down 0.028
hsa-miR-150 2.99 Down 0.010 hsa-miR-l35b 4.42 Up 0.0000 hsa-miR-l46a 2.43 Down 0.028
hsa-miR-l39-3p 2.92 Down 0.039 hsa-miR-l35b 4.29 Up 0.0000 hsa-miR-451 2.09 Down 0.036
hsa-miR-195 2.75 Down 0.006 hsa-miR-451 4.27 Down 0.0001 hsa-miR-451 2.06 Down 0.036
hsa-miR-183 2.7 Up 0.001 hsa-miR-451 4.15 Down 0.0001 hsa-miR-6l5-5p 1.89 Down 0.050
hsa-miR-20a 2.63 Up 0.003 hsa-miR-552 4.13 Up 0.0001 hsa-miR-l46b-5p 1.78 Down 0.034
hsa-miR-20a 2.58 Up 0.003 hsa-miR-224 4.08 Up 0.0001 hsa-miR-572 1.75 Up 0.034
hsa-miR-17 2.55 Up 0.003 hsa-miR-l 4.05 Down 0.0005 hsa-miR-572 1.65 Up 0.028
hsa-miR-92a 2.54 Up 0.003 hsa-miR-552 3.93 Up 0.0003 hsa-miR-638 1.65 Up 0.036
hsa-miR-183 2.54 Up 0.003 hsa-miR-l 33b 3.79 Down 0.0001 hsa-miR-342-3p 1.59 Down 0.034
hsa-miR-215 2.53 Down 0.014 hsa-miR-183 3.66 Up 0.0007 hsa-miR-23b 1.58 Down 0.036
hsa-miR-34a 2.51 Up 0.006 hsa-miR-552 3.65 Up 0.0001 hsa-miR-49l-5p 1.55 Down 0.036
hsa-miR-34a 2.48 Up 0.003 hsa-miR-375 3.65 Down 0.0116 hsa-miR-638 1.51 Up 0.036
hsa-miR-92a 2.48 Up 0.003 hsa-miR-650 3.59 Down 0.0013 hsa-miR-342-3p 1.50 Down 0.036
hsa-miR-29a 2.46 Up 0.006 hsa-miR-195 3.57 Down 0.0017
hsa-miR-21 2.45 Up 0.018 hsa-miR-183 3.5 Up 0.0004
hsa-miR-195 2.43 Down 0.003 hsa-miR-195 3.39 Down 0.0017
hsa-miR-96 2.35 Up 0.001 hsa-miR-497 3.38 Down 0.0012
hsa-miR-224 2.33 Up 0.003 hsa-miR-l 45 3.27 Down 0.0015
hsa-miR-215 2.31 Down 0.018 hsa-miR-375 3.25 Down 0.0029
hsa-miR-l25b-2* 2.29 Down 0.054 hsa-miR-17 3.08 Up 0.0005
hsa-miR-451 2.29 Down 0.039 hsa-miR-92a 3.04 Up 0.0000
hsa-miR-150 2.28 Down 0.015 hsa-miR-150 2.97 Down 0.0005
hsa-miR-451 2.22 Down 0.039 hsa-miR-92a 2.93 Up 0.0000
hsa-miR-29a 2.22 Up 0.010 hsa-miR-l 45 2.91 Down 0.0016
hsa-miR-17 2.22 Up 0.003 hsa-miR-21 2.86 Up 0.0029
hsa-miR-552 2.17 Up 0.010 hsa-miR-20a 2.79 Up 0.0011
hsa-miR-l25b-2* 2.14 Down 0.054 hsa-miR-96 2.67 Up 0.0001
hsa-miR-21 2.13 Up 0.053 hsa-miR-183 2.64 Up 0.0016
hsa-miR-l39-3p 2.09 Down 0.049 hsa-miR-20a 2.62 Up 0.0017
Open in a new tab

There were no miRNAs with adjusted P value above 0.05 when comparing.

Colon TP53-mutated tumors compared with normal tissue demonstrated differential expression for 139 unique miRNA and 242 total probes; there were 11 that were unique to TP53 when compared with other colon tumor phenotypes. The highest level of upregulation when compared with normal was for miR-224 with a ratio of 5.97 above normal; miR-650 was downregulated at a 4.61 ratio to normal tissue. The top probes with differential expression comparing normal tissue with tumor tissue are shown in Table 4. There were 136 unique miRNA associated with rectal TP53-mutated tumors compared with normal tissue; 82 of these miRNA had duplicate probes that exhibited up- or downregulation. MiR-224, miR-183, and miR-21 showed the highest degree of upregulation. A comparison of TP53-mutated rectal tumors with other rectal tumor markers showed five unique miRNA that were up- or downregulated. MiR-187, which was down-regulated, showed the highest degree of difference from other tumor markers.

Table 4. Associations Between TP53 and miRNA Expression Levels.

Colon TP53 vs. normal tissue Colon TP53 vs. other tumor markers Rectal TP53 vs. normal tissue Rectal TP53 vs. other tumor markers
Gene name Ratio Direction adj. P-value Gene name Ratio Direction adj. P-value Gene name Ratio Direction adj. P-value Gene name Ratio Direction adj. P-value
hsa-miR-224 5.97 Up 0.0000 hsa-miR-224 2.20 Up 0.033 hsa-miR-224 4.90 Up 0.0000 hsa-miR-l 87 1.98 Down 0.017
hsa-miR-650 4.61 Down 0.0000 hsa-miR-224 2.00 Up 0.035 hsa-miR-21 4.16 Up 0.0002 hsa-miR-l 35b 1.94 Down 0.050
hsa-miR-17 4.54 Up 0.0000 hsa-miR-548a-3p 1.92 Down 0.030 hsa-miR-183 4.02 Up 0.0001 hsa-miR-l 35b 1.81 Down 0.041
hsa-miR-l39-3p 4.18 Down 0.0044 hsa-miR-17 1.91 Up 0.043 hsa-miR-224 3.87 Up 0.0000 hsa-miR-302b* 1.53 Down 0.006
hsa-miR-17 4.18 Up 0.0000 hsa-miR-1226 1.82 Down 0.050 hsa-miR-183 3.81 Up 0.0000 hsa-miR-541* 1.53 Down 0.022
hsa-miR-20a 4.08 Up 0.0000 hsa-miR-532-5p 1.69 Up 0.025 hsa-miR-l 3.78 Down 0.0004 hsa-miR-302b 1.52 Down 0.022
hsa-miR-224 4.02 Up 0.0000 hsa-miR-17 1.69 Up 0.034 hsa-miR-17 3.74 Up 0.0000
hsa-miR-20a 3.92 Up 0.0000 hsa-miR-573 1.68 Down 0.043 hsa-miR-21 3.68 Up 0.0014
hsa-miR-92a 3.64 Up 0.0001 hsa-miR-574-5p 1.67 Down 0.050 hsa-miR-l 33b 3.68 Down 0.0000
hsa-miR-92a 3.57 Up 0.0002 hsa-miR-424* 1.63 Up 0.043 hsa-miR-20a 3.48 Up 0.0000
hsa-miR-29a 3.53 Up 0.0000 hsa-miR-50l-3p 1.61 Up 0.030 hsa-miR-17 3.46 Up 0.0000
hsa-miR-93 3.18 Up 0.0000 hsa-miR-890 1.57 Down 0.043 hsa-miR-20a 3.42 Up 0.0000
hsa-miR-1305 3.06 Down 0.0001 hsa-miR-16-l* 1.54 Down 0.030 hsa-miR-375 3.40 Down 0.0107
hsa-miR-1288 3.02 Down 0.0002 hsa-miR-424* 1.54 Up 0.035 hsa-miR-92a 3.35 Up 0.0000
hsa-miR-150 3.01 Down 0.0030 hsa-miR-92a 3.26 Up 0.0000
hsa-miR-1305 3 Down 0.0002 hsa-miR-7 3.08 Up 0.0022
hsa-miR-29a 2.95 Up 0.0001 hsa-miR-l 45 2.98 Down 0.0017
hsa-miR-21* 2.95 Up 0.0000 hsa-miR-552 2.97 Up 0.0007
hsa-miR-552 2.93 Up 0.0001 hsa-miR-183 2.91 Up 0.0007
hsa-miR-183 2.9 Up 0.0000 hsa-miR-17 2.90 Up 0.0002
hsa-miR-221 2.9 Up 0.0001 hsa-miR-20b 2.88 Up 0.0002
hsa-miR-552 2.89 Up 0.0001 hsa-miR-650 2.86 Down 0.0014
hsa-miR-1288 2.86 Down 0.0001 hsa-miR-552 2.83 Up 0.0019
hsa-miR-17 2.86 Up 0.0000 hsa-miR-455-3p 2.81 Up 0.0001
hsa-miR-20b 2.86 Up 0.0000 hsa-miR-96 2.80 Up 0.0002
hsa-miR-29b 2.85 Up 0.0002 hsa-miR-96 2.78 Up 0.0002
hsa-miR-21 2.85 Up 0.0008 hsa-miR-375 2.78 Down 0.0035
hsa-miR-892b 2.77 Down 0.0010 hsa-miR-552 2.70 Up 0.0025
hsa-miR-183 2.76 Up 0.0001 hsa-miR-17* 2.67 Up 0.0002
hsa-miR-375 2.75 Down 0.0012 hsa-miR-424 2.59 Up 0.0018
Open in a new tab

Discussion

Numerous studies have shown that miRNAs are extensively involved in cancer pathogenesis of solid tumors and support their function as dominant or recessive genes (Volinia et al., 2006). We know that miRNAs are highly conserved across species, are tissue specific, and play a critical role in regulation of proliferation, differentiation, apoptosis, and stress response (Iorio and Croce, 2009). What is not known is the extent to which miRNAs are involved in colon and rectal cancer. Data from this study suggest that miRNA are differentially expressed in colon and rectal cancers.

It is believed that miRNAs are tissue specific (Aslam et al., 2009). Our work and that of others have shown that both disease-correlated genetic and lifestyle factors differ for colon and rectal cancers (Alexander and Cushing, 2010; Slattery et al., 2003a; Slattery et al., 2003b; Slattery et al., 2004a; Slattery et al., 2004b; Slattery et al., 2004c; Slattery et al., 2007; Iacopetta et al., 2009; Slattery et al., 2009). Additionally, other characteristics of these cancers suggest two separate diseases, such as differences in survival, younger age at diagnosis for rectal cancer, smaller risk associated with family history of CRC for rectal cancer and fewer inherited syndromes, such as Lynch syndrome, adenomatous polyposis coli (APC), and attenuated APC involving rectal cancer. The data we present add support for these disparate findings, in that over 200 miRNAs are differentially expressed in colon and rectal normal tissue. These differences in miRNA expression levels support the hypothesis that two diseases each arising from different tissues exist and that associations can be masked when studying them as one disease.

It has been suggested that regulation of miRNA expression in colon and rectal cancers may involve epigenetic mechanisms. Transfection of miR-342 into colon cancer cell lines has been shown to lead to apoptosis that can be inhibited by methylation (Grady et al., 2008). Four miRNAs (miR-124a, let-7a-3, miR-10a, and miR-342) have been shown to be silenced by DNA hypomethylation (Lujambio et al., 2007). The miR-17-92 miRNA family has been associated with MSI+ tumors (Lanza et al., 2007). Of these, only miR-10a was upregulated significantly and only for colon tumors compared with normal tissue. MiR-10a was upregulated with two probes being significantly associated with CIMP+ tumors (ratios of 2.5 and 2.3 when compared with normal tissue). However, it should be noted that while this finding is replicated, it was ranked number 54 and 69 in terms of miRNAs that were differently expressed in CIMP+ tumors.

Several studies have focused on TP53 and miRNA, in part because of the hypothesized role with inflammation (Schetter et al., 2010) Work focusing on inflammation has shown links between nuclear factor kappa B (NFκB), inflammatory cytokines, prostaglandins, reactive oxygen and nitrogen species and specific miRNAs (Schetter, et al.). MiR-155 and miR-146 have been associated with NFκB, suppressor of cytokine signaling 1 (SOCS1), interleukin 6 (IL6), and tumor necrosis factor-alpha (TNF-α); let-7i has been associated with IL6 pathways; miR-21 has been associated with IL6, signal transducer and activator of transcription 3 (STAT3), phosphatase and tensin homolog (PTEN), and Interleukin 12 (IL12). Polymorphisms in some of these inflammatory-related genes also have been associated with miRNAs (Schetter et al.; Taganov et al., 2006). Many of the miRNAs associated with inflammation-related genes also have been associated with TP53-mutated tumors. MiRNA expression levels have been associated with the vascular endothelial growth factor (VEGF) and angiogenesis signaling pathways in studies of nasopharyngeal carcinomas (Chen et al., 2009). VEGF and other angiogenesis genes are intertwined with an inflammation-related pathway that is hypothesized as being important to the etiology of colon and rectal cancer (Slattery and Fitzpatrick, 2009). Of these previously reported associations, only miR-21 was associated with both colon and rectal cancer, with a more prominent role in rectal cancer. For rectal cancer, miR-21 was upregulated at two probes, ranking number 2 and number 8 on the list, with a ratio of 4.16 and 3.74 compared with normal tissue. For colon cancers compared with normal two probes also were upregulated at rank number 17 and 66 and corresponding ratios of 2.85 and 2.45. Let-7i also was upregulated in rectal tumors, ranking number 74 and 117 (upregulated ratios of 2.18 and 1.89).

Most other studies have targeted a few preselected miRNAs with a hypothesized mechanism of action. In a field where there is rapid increase in knowledge, dismissing the potential spectrum of associations to only a handful of previously examined miRNAs will miss many important associations, as shown here. In this study, we evaluated all 866 known human miRNAs; we identified hundreds of miRNAs that were differently expressed in tumors relative to normal tissue. Many of these miRNAs have been linked to specific pathways and gene targets but have not been studied with colon and rectal cancer.

Assessing target genes of miRNAs that were the most differentially expressed showed several important targets (Betel et al., 2008). MiR-135b has been shown to target genes such as signal transducer and activator of transcription 6 (STAT6), insulin receptor substrate 2 (IRS2), bone morphogenetic protein receptor, type II (BMPR2) and bone morphogenetic protein receptor, type IA (BMPR1A) which are members of the TGF-β-signaling pathway and recently shown to be associated with colon and rectal cancer (Slattery et al., 2010). These genes were further linked to important phenotypes of ulcerative colitis, obesity, and CRC. MiR-224 was linked to important phenotypes of obesity and insulin resistance and targeted genes such as adiponectin (ADIPOQ), mothers against decapentaplegic 4 (SMAD4), breast cancer 1 (BRCA1), alcohol dehydrogenase 1D (ADH1D), and interleukin 1 receptor (IL1R). MiR-146a also appeared to be important for inflammation and obesity related mechanisms and targeted genes such as C-reactive protein (CRP), SMAD4, leptin (LEP), IRS2, BMPR1A, and prostaglandin-endoperoxide synthase 2 (PTGS2 or COX2). MiR-139-3p associated with colon cancer in our study appears to target adiposity, insulin, and colon cancer-related genes such as lipoprotein lipase (LPL), kinase insert domain receptor (KDR, a VEGF receptor), and cyclin D1 (CCND1). MiR-451 also appears to target important genes within obesity, inflammation, and insulin-related pathways, including IL6R and tumor necrosis factor ligand superfamily member 15 (TNFSF15). MiR-183 appears to target insulin receptor substrate 1 (IRS1), PTEN, methylenetetrahydrofolate reductase (MTHFR), NFKB1, PTGS2, insulin-like growth factor (IGF), transforming growth factor beta receptor 1 (TGFBR1), all important genes in colon and rectal cancer. MiR-187 also targets important genes involved in obesity, insulin, and inflammation-related pathways including tumor necrosis factor (TNF), ADIPOQ, LEPR, insulin receptor (INSR), peroxisome proliferator-activated receptor gamma (PPARG), and ornithine decarboxylase (ODC). Further examination of miRNAs at a more global level may provide important information into how genes and pathways are linked together to influence cancer risk.

The data presented here are limited in the number of samples examined and their selection was based on known tumor characteristics. Thus, this does not represent the importance of miRNAs to all colon and rectal cancers, but rather to colon and rectal cancers with specific tumor characteristics. In addition, although technical replicates and independent probes for specific miRNAs showed a high level of correlation of expression levels, it is important to validate these findings using methods such as TaqMan assays to verify quantitative miRNA expression levels. Another limitation to this study is that we used a subset of normal tissue from cases with tumor tissue for comparison. Ideally, all cases would have had normal tissue samples analyzed as paired samples, but such an approach was cost prohibitive and we believe that the normal tissues chosen were representative. However, despite these limitations, the study is the most extensive evaluation of miRNA in colon and rectal cancers to date. Given our ability to separate colon and rectal tumors, we can gain insight into important differences between these two diseases that can help substantiate our previous observations of differences in risk factors for the two cancers.

It has been stated that “Every cellular process is likely to be regulated by microRNAs, and an aberrant microRNA expression signature is a hallmark of several diseases, including cancer” (Iorio and Croce, 2009). The importance of understanding how miRNAs function as well as what causes their expression to be increased or decreased is only beginning to be understood. However, given their importance to colon and rectal cancer, as demonstrated here by the number of miRNAs that are differentially expressed in tumor tissue, is it imperative that we attempt to understand how they function in the carcinogenic process from disease development to survival after diagnosis, using powerful samples with the capability of exploring the spectrum of the disease process.

Acknowledgments

NCI, Grant numbers: CA48998, CA61757; Utah Cancer, Registry, Grant number: #N01-PC-67000 from the National Cancer Institute; State of Utah Department of Health, the Northern California Cancer Registry, and the Sacramento Tumor Registry.

The authors acknowledge Dr. Bette Caan and the Kaiser Permanente Medical Care Program for their contribution to the study, Dr. Wade Samowitz for early efforts in data collection, Dr. Hans Albertsen for his insightful comments regarding miRNA in epidemiological research, and Ms. Sandra Edwards for her efforts in obtaining tumor tissue. The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent the official view of the National Cancer Institute.

Footnotes

Additional Supporting Information may be found in the online version of this article.

References

  1. Akao Y, Nakagawa Y, Naoe T. let-7 microRNA functions as a potential growth suppressor in human colon cancer cells. Biol Pharm Bull. 2006;29:903–906. doi: 10.1248/bpb.29.903. [DOI] [PubMed] [Google Scholar]
  2. Alexander DD, Cushing CA. Red meat and colorectal cancer: A critical summary of prospective epidemiologic studies. Obes Rev. doi: 10.1111/j.1467-789X.2010.00785.x. Epub ahead of print 21 July 2010. [DOI] [PubMed] [Google Scholar]
  3. Aslam MI, Taylor K, Pringle JH, Jameson JS. MicroRNAs are novel biomarkers of colorectal cancer. Br J Surg. 2009;96:702–710. doi: 10.1002/bjs.6628. [DOI] [PubMed] [Google Scholar]
  4. Benjamini YaH, Y Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Stat Soc. 1995;57:289–300. [Google Scholar]
  5. Chen HC, Chen GH, Chen YH, Liao WL, Liu CY, Chang KP, Chang YS, Chen SJ. MicroRNA deregulation and pathway alterations in nasopharyngeal carcinoma. Br J Cancer. 2009;100:1002–1011. doi: 10.1038/sj.bjc.6604948. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Betel D, Wilson M, Marks DS, Sander C. microRNA org: A Resource for Predicted microRNA Target and Gene Expression. New York: Memorial Sloan-Kettering Cancer Center; 2008. [Google Scholar]
  7. Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, Ellis B, Gautier L, Ge Y, Gentry J, Hornik K, Hot-horn T, Huber W, Iacus S, Irizarry R, Leisch F, Li C, Maechler M, Rossini AJ, Sawitzki G, Smith C, Smyth G, Tierney L, Yang JY, Zhang J. Bioconductor: Open software development for computational biology and bioinformatics. Genome Biol. 2004;5:R80. doi: 10.1186/gb-2004-5-10-r80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Grady WM, Parkin RK, Mitchell PS, Lee JH, Kim YH, Tsuchiya KD, Washington MK, Paraskeva C, Willson JK, Kaz AM, Kroh EM, Allen A, Fritz BR, Markowitz SD, Tewari M. Epigenetic silencing of the intronic microRNA hsa-miR-342 and its host gene EVL in colorectal cancer. Oncogene. 2008;27:3880–3888. doi: 10.1038/onc.2008.10. [DOI] [PubMed] [Google Scholar]
  9. Griffiths-Jones S, Grocock RJ, van Dongen S, Bateman A, Enright AJ. miRBase: MicroRNA sequences, targets and gene nomenclature. Nucleic Acids Res. 2006;34(Database issue):D140–D144. doi: 10.1093/nar/gkj112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. He L, He X, Lim LP, de Stanchina E, Xuan Z, Liang Y, Xue W, Zender L, Magnus J, Ridzon D, Jackson AL, Linsley PS, Chen C, Lowe SW, Cleary MA, Hannon GJ. A microRNA component of the p53 tumour suppressor network. Nature. 2007a;447:1130–1134. doi: 10.1038/nature05939. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. He L, He X, Lowe SW, Hannon GJ. microRNAs join the p53 network—Another piece in the tumour-suppression puzzle. Nat Rev Cancer. 2007b;7:819–22. doi: 10.1038/nrc2232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. He X, He L, Hannon GJ. The guardian's little helper: microRNAs in the p53 tumor suppressor network. Cancer Res. 2007c;67:11099–11101. doi: 10.1158/0008-5472.CAN-07-2672. [DOI] [PubMed] [Google Scholar]
  13. Hui AB, Shi W, Boutros PC, Miller N, Pintilie M, Fyles T, McCready D, Wong D, Gerster K, Waldron L, Jurisica I, Penn LZ, Liu FF. Robust global micro-RNA profiling with formalin-fixed paraffin-embedded breast cancer tissues. Lab Invest. 2009;89:597–606. doi: 10.1038/labinvest.2009.12. [DOI] [PubMed] [Google Scholar]
  14. Iacopetta B, Heyworth J, Girschik J, Grieu F, Clayforth C, Fritschi L. The MTHFR C677T and DeltaDNMT3BC-149T polymorphisms confer different risks for right- and left-sided colorectal cancer. Int J Cancer. 2009;125:84–90. doi: 10.1002/ijc.24324. [DOI] [PubMed] [Google Scholar]
  15. Iorio MV, Croce CM. MicroRNAs in cancer: Small molecules with a huge impact. J Clin Oncol. 2009;27:5848–5856. doi: 10.1200/JCO.2009.24.0317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Johnson WE, Li C, Rabinovic A. Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics. 2007;8:118–127. doi: 10.1093/biostatistics/kxj037. [DOI] [PubMed] [Google Scholar]
  17. Lanza G, Ferracin M, Gafa R, Veronese A, Spizzo R, Pichiorri F, Liu CG, Calin GA, Croce CM, Negrini M. mRNA/micro-RNA gene expression profile in microsatellite unstable colorectal cancer. Mol Cancer. 2007;6:54. doi: 10.1186/1476-4598-6-54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Lujambio A, Ropero S, Ballestar E, Fraga MF, Cerrato C, Setien F, Casado S, Suarez-Gauthier A, Sanchez-Cespedes M, Git A, Spiteri I, Das PP, Caldas C, Miska E, Esteller M. Genetic unmasking of an epigenetically silenced microRNA in human cancer cells. Cancer Res. 2007;67:1424–1429. doi: 10.1158/0008-5472.CAN-06-4218. [DOI] [PubMed] [Google Scholar]
  19. Michael MZ, SM OC, van Holst Pellekaan NG, Young GP, James RJ. Reduced accumulation of specific microRNAs in colorectal neoplasia. Mol Cancer Res. 2003;1:882–891. [PubMed] [Google Scholar]
  20. Schetter AJ, Heegaard NH, Harris CC. Inflammation and cancer: Interweaving microRNA, free radical, cytokine and p53 pathways. Carcinogenesis. 2010;31:37–49. doi: 10.1093/carcin/bgp272. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Slattery ML, Fitzpatrick FA. Convergence of hormones, inflammation, and energy-related factors: A novel pathway of cancer etiology. Cancer Prev Res (Phila Pa) 2009;2:922–930. doi: 10.1158/1940-6207.CAPR-08-0191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Slattery ML, Potter J, Caan B, Edwards S, Coates A, Ma KN, Berry TD. Energy balance and colon cancer—Beyond physical activity. Cancer Res. 1997a;57:75–80. [PubMed] [Google Scholar]
  23. Slattery ML, Potter JD, Duncan DM, Berry TD. Dietary fats and colon cancer: Assessment of risk associated with specific fatty acids. Int J Cancer. 1997b;73:670–677. doi: 10.1002/(sici)1097-0215(19971127)73:5<670::aid-ijc10>3.0.co;2-a. [DOI] [PubMed] [Google Scholar]
  24. Slattery ML, Caan BJ, Benson J, Murtaugh M. Energy balance and rectal cancer: An evaluation of energy intake, energy expenditure, and body mass index. Nutr Cancer. 2003a;46:166–171. doi: 10.1207/S15327914NC4602_09. [DOI] [PubMed] [Google Scholar]
  25. Slattery ML, Edwards S, Curtin K, Ma K, Edwards R, Holubkov R, Schaffer D. Physical activity and colorectal cancer. Am J Epidemiol. 2003b;158:214–224. doi: 10.1093/aje/kwg134. [DOI] [PubMed] [Google Scholar]
  26. Slattery ML, Neuhausen SL, Hoffman M, Caan B, Curtin K, Ma KN, Samowitz W. Dietary calcium, vitamin D, VDR genotypes and colorectal cancer. Int J Cancer. 2004a;111:750–756. doi: 10.1002/ijc.20330. [DOI] [PubMed] [Google Scholar]
  27. Slattery ML, Samowitz W, Curtin K, Ma KN, Hoffman M, Caan B, Neuhausen S. Associations among IRS1, IRS2, IGF1, and IGFBP3 genetic polymorphisms and colorectal cancer. Cancer Epidemiol Biomarkers Prev. 2004b;13:1206–1214. [PubMed] [Google Scholar]
  28. Slattery ML, Samowitz W, Hoffman M, Ma KN, Levin TR, Neuhausen S. Aspirin, NSAIDs, and colorectal cancer: Possible involvement in an insulin-related pathway. Cancer Epidemiol Biomarkers Prev. 2004c;13:538–545. [PubMed] [Google Scholar]
  29. Slattery ML, Wolff RK, Herrick JS, Caan BJ, Potter JD. IL6 genotypes and colon and rectal cancer. Cancer Causes Control. 2007;18:1095–1105. doi: 10.1007/s10552-007-9049-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Slattery ML, Wolff RK, Curtin K, Fitzpatrick F, Herrick J, Potter JD, Caan BJ, Samowitz WS. Colon tumor mutations and epigenetic changes associated with genetic polymorphism: Insight into disease pathways. Mutat Res. 2009;660:12–21. doi: 10.1016/j.mrfmmm.2008.10.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Slattery ML, HJ, Lundgreen A, Wolff RK. Genetic variation in the TGF-Beta-signaling pathway and colon and rectal cancer risk. Cancer Epidemiol Biomarkers Prev. 2010 doi: 10.1158/1055-9965.EPI-10-0843. In press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Smalheiser NR, Torvik VI. A population-based statistical approach identifies parameters characteristic of human micro-RNA-mRNA interactions. BMC Bioinformatics. 2004;5:139. doi: 10.1186/1471-2105-5-139. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Taganov KD, Boldin MP, Chang KJ, Baltimore D. NF-kappaB-dependent induction of microRNA miR-146, an inhibitor targeted to signaling proteins of innate immune responses. Proc Natl Acad Sci USA. 2006;103:12481–12486. doi: 10.1073/pnas.0605298103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Volinia S, Calin GA, Liu CG, Ambs S, Cimmino A, Petrocca F, Visone R, Iorio M, Roldo C, Ferracin M, Prueitt RL, Yanaihara N, Lanza G, Scarpa A, Vecchione A, Negrini M, Harris CC, Croce CM. A microRNA expression signature of human solid tumors defines cancer gene targets. Proc Natl Acad Sci USA. 2006;103:2257–2261. doi: 10.1073/pnas.0510565103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Yang L, Belaguli N, Berger DH. MicroRNA and colorectal cancer. World J Surg. 2009;33:638–646. doi: 10.1007/s00268-008-9865-5. [DOI] [PubMed] [Google Scholar]

RESOURCES