The Human let-7a-3 Locus Contains an Epigenetically

Regulated MicroRNA Gene with Oncogenic Function

Bodo Brueckner, Carlo Stresemann, Ruprecht Kuner, et al.

Cancer Res 2007;67:1419-1423.

Updated version Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/67/4/1419

Supplementary Access the most recent supplemental material at:

Material http://cancerres.aacrjournals.org/content/suppl/2007/02/14/67.4.1419.DC1.html

Cited Articles This article cites by 20 articles, 7 of which you can access for free at:
http://cancerres.aacrjournals.org/content/67/4/1419.full.html#ref-list-1

Citing articles This article has been cited by 48 HighWire-hosted articles. Access the articles at:
http://cancerres.aacrjournals.org/content/67/4/1419.full.html#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Subscriptions Department at pubs@aacr.org.

Permissions To request permission to re-use all or part of this article, contact the AACR Publications
Department at permissions@aacr.org.

Downloaded from cancerres.aacrjournals.org on September 11, 2014. 2007 American Association for
Cancer Research.
 Priority Report

The Human let-7a-3 Locus Contains an Epigenetically Regulated

MicroRNA Gene with Oncogenic Function
1 1 2 1 1
Bodo Brueckner, Carlo Stresemann, Ruprecht Kuner, Cora Mund, Tanja Musch,
3 2 1
Michael Meister, Holger Sultmann, and Frank Lyko
Divisions of 1Epigenetics and 2Molecular Genome Analysis, Deutsches Krebsforschungszentrum; 3Sektion Translationale Forschung,
Thoraxklinik am Universitatsklinikum Heidelberg, Heidelberg, Germany

Abstract be reverted by DNA methyltransferase inhibitors, such as 5-aza-2-

deoxycytidine (DAC). Treatment of cancer cells with these
MicroRNAs (miRNAs) are small noncoding RNAs that repress
compounds followed by expression profiling has been used to
their target mRNAs by complementary base pairing and
systematically unmask tumor suppressor genes that are hyper-
induction of the RNA interference pathway. It has been shown
methylated in human cancer cells (8, 9). Recently, Saito et al. (10)
that miRNA expression can be regulated by DNA methylation
have used a similar approach to identify miRNA genes that are
and it has been suggested that altered miRNA gene methyl-
affected by DNA hypermethylation: combinatorial treatment of T24
ation might contribute to human tumorigenesis. In this study,
bladder cancer cells with DAC and a second epigenetic drug, the
we show that the human let-7a-3 gene on chromosome
histone deacetylase inhibitor 4-phenylbutyrate, caused >3-fold up-
22q13.31 is associated with a CpG island. Let-7a-3 belongs to
regulation in 17 of 313 miRNA genes analyzed. The strongest effects
the archetypal let-7 miRNA gene family and was found to be
(49-fold up-regulation) were observed for miR-127 and the
methylated by the DNA methyltransferases DNMT1 and
corresponding gene was found to be embedded in a CpG island.
DNMT3B. The gene was heavily methylated in normal human
Epigenetic activation of miR-127 resulted in a detectable down-
tissues but hypomethylated in some lung adenocarcinomas.
regulation of the BCL6 proto-oncogene, which suggested a tumor
Let-7a-3 hypomethylation facilitated epigenetic reactivation of
suppressor function for this miRNA (10). However, the miR-127
the gene and elevated expression of let-7a-3 in a human lung
gene was found to be methylated in many human tissues and no
cancer cell line resulted in enhanced tumor phenotypes and
methylation changes could be detected in the three primary human
oncogenic changes in transcription profiles. Our results thus
tumor samples that were analyzed in this study.
identify let-7a-3 as an epigenetically regulated miRNA gene
We have analyzed the methylation of the human let-7a-3 gene
with oncogenic function and suggest that aberrant miRNA
that belongs to the archetypal family of let-7 miRNA genes. We
gene methylation might contribute to the human cancer
found that let-7a-3 methylation is prevalent in normal human
epigenome. [Cancer Res 2007;67(4):141923]
tissues but can be lost in lung cancers. The characterization of
the mechanisms mediating let-7a-3 methylation and the function
Introduction of let-7a-3 expression in a human lung adenoma cell line suggests
that epigenetic activation of let-7a-3 might contribute to human
MicroRNAs (miRNAs) are small noncoding RNAs that repress
lung tumorigenesis.
their target mRNAs by complementary base pairing and induction
of the RNA interference pathway (1). Several miRNAs have been
shown to play specific roles in development and differentiation Materials and Methods
and, correspondingly, their expression patterns seem to be highly Cell culture and patient samples. HCT116 and HCT116 knockout cells
regulated. miRNA expression profiling has provided indications (11) were cultured in McCoys 5A medium supplemented with 10% FCS. A549
for aberrant expression patterns in human cancer cell lines and in cells were cultured in DMEM supplemented with 10% FCS. For inhibitor
primary cancers, which suggested that aberrant miRNA expression treatment, cells were incubated for 72 h with 500 nmol/L 5-aza-
might contribute to tumorigenesis (24). Consistent with this 2-deoxycytidine (DAC; Calbiochem), and/or 1 mmol/L valproic acid (VPA;
notion, the expression pattern of 200 human miRNA genes can be Merck). Patient DNA was prepared from fresh-frozen tissues after
of greater prognostic value than the expression pattern of 13,000 institutional review board approval. For all experiments, genomic DNA was
protein-encoding genes (3). prepared using the DNeasy kit (Qiagen).
Human tumorigenesis is characterized by specific changes in DNA methylation analysis. Genomic DNA was deaminated with sodium
bisulfite using standard procedures and let-7a-3 was amplified using nested
genomic DNA methylation patterns. Compared with nonmalignant
primers as follows: let7a_out_for (GTTAGAATTAGGGTTTTTGGGGAGG)
cells, tumor cells show hypermethylation or hypomethylation in and let7a_out_rev (ACCTATCAAACTTCTCAATATAAAC), 95jC for 3 min,
CpG islands of genes that are functionally important for tumor followed by 34 cycles (95jC for 30 s, 54jC for 45 s, and 72jC for 1 min), and
development (57). Hypermethylation-induced gene silencing can 72jC for 4 min. Primers and PCR conditions for the second amplification
were let7a_in_for (GGGAGGGATGTTTGTTTGTTTAGTG) and let7a_in_rev
(AACTACCCCCAAACCTAACCCTACC), 95jC for 3 min, followed by
34 cycles (95jC for 30 s, 64jC for 30 s, and 72jC for 45 s), and 72jC for
Note: Supplementary data for this article are available at Cancer Research Online 4 min. The PCR product (723 bp) was gel purified and digested with
(http://cancerres.aacrjournals.org/). BstUI (New England Biolabs). Digested PCR products were separated on
Requests for reprints: Frank Lyko, Deutsches Krebsforschungszentrum, Im agarose gels and visualized by ethidium bromide staining. For bisulfite
Neuenheimer Feld 580, 69120 Heidelberg, Germany. Phone: 49-6221-423800; Fax: 49-
6221-423802; E-mail: f.lyko@dkfz.de.
sequencing, PCR products were gel extracted and cloned using the
I2007 American Association for Cancer Research. TOPO TA cloning kit (Invitrogen) according to the manufacturers
doi:10.1158/0008-5472.CAN-06-4074 instructions.

www.aacrjournals.org 1419 Cancer Res 2007; 67: (4). February 15, 2007

Downloaded from cancerres.aacrjournals.org on September 11, 2014. 2007 American Association for
Cancer Research.
Cancer Research

Let-7a-3 expression analysis. Total RNA was isolated from cells using let-7a-3 CpG island in HCT116 cells and the results showed that the
Trizol (Invitrogen) and cDNA was synthesized using the ThermoScript gene was densely (90%) methylated (Fig. 1B). The analysis of
reverse transcription-PCR (RT-PCR) system (Invitrogen). Twenty microliters genomic DNA from isogenic DNA methyltransferase knockout cell
of PCRs contained 2 AL cDNA template, 1 ReddyMix buffer (Abgene), lines (11) showed that methylation seemed unchanged in DNMT3B
1 Amol/L of each primer, 1 mmol/L deoxynucleotide triphosphates (Stra-
knockout cells (Fig. 1B, 3BKO). Let-7a-3 methylation seemed
tagene), and 1.5 units of Thermoprime polymerase (Abgene). The primers
and PCR conditions for let-7a-3 cDNA amplification were let7a_RT2_for somewhat lower in DNMT1 knockout cells (Fig. 1B, 1KO) and the
(CTCTGGAAGCCACGGAGTC) and let7a_RT2_rev (GTTCCAGACGCT- gene was almost completely demethylated in DNMT1;DNMT3B
CTGTCCAC), 95jC for 3 min, followed by 34 cycles (95jC for 30 s, 62jC double knockout cells (Fig. 1B, DKO). This showed that let-7a-3
for 30 s, and 72jC for 30 s), and 72jC for 3 min. Primers and PCR conditions methylation is cooperatively maintained by the DNA methyltrans-
for tissue inhibitor of metalloproteinase-3 and h-amyloid have been ferases DNMT1 and DNMT3B and suggests that miRNA genes are
described elsewhere (12). PCR amplicons were separated on agarose gels methylated by the same mechanisms that govern methylation of
and visualized by ethidium bromide staining. other genomic regions (11).
Establishment of stably transfected let-7a-3 cell lines. The precursor We then sought to further characterize the methylation pattern of
sequence encoding let-7a-3 was PCR amplified from human genomic DNA. let-7a-3 in human tissues by combined bisulfite restriction analysis
A product of 179 bp was subcloned into the expression vector pZeoSV2
(COBRA). COBRA detects DNA methylation by probing restriction
(Invitrogen) by using EcoRI and KpnI restriction sites contained in the PCR
primers and verified by DNA sequencing. The primers and PCR conditions enzyme sites in PCR amplicons from bisulfite-deaminated DNA.
for let-7a-3 amplification were let-7a_forcl (ATGAATTCCTCTGGAAGC- Methylated genes generate small restriction fragments, whereas
CACGGAGTC) and let7a_revcl (ATGGTACCGTTCCAGACGCTCTGTCCAC), unmethylated genes generate PCR fragments that cannot be cut
95jC for 3 min, followed by 34 cycles (95jC for 30 s, 62jC for 30 s, and 72jC by restriction enzymes. Let-7a-3 COBRA analysis of genomic DNA
for 30 s), and 72jC for 3 min. from normal placenta, brain, bone marrow, blood, colon, and skin
Constructs were transfected into A549 cells using Fugene 6 transfection samples revealed strong methylation of let-7a-3 in all tissues
agent (Roche), according to the manufacturers protocol. Cells were grown analyzed (Fig. 2A). Because let-7 miRNAs have been linked to lung
for 3 days in transfection medium and then selected in cell culture medium tumorigenesis (1518), we also analyzed let-7a-3 methylation in a
containing 200 Ag/mL zeocin (Invitrogen). set of eight lung adenocarcinomas and eight matched nonneo-
Colony formation assays. Cells (2  105) in 1.5 mL medium
plastic lung tissue samples from the same patients. The results
supplemented with 0.3% agarose were layered on a 3-mL base of 0.5%
agarose with medium. Soft agar assays were done in 60-mm dishes and in
triplicate. After colonies became visible, cells were stained with 200 AL
p-iodonitrotetrazolium violet solution (5 mg/mL).
Microarray experiments. Hybridizations were done on genome-wide
cDNA microarrays. Gene expression analysis comprised four microarray
hybridizations: four let-7a-3transfected cell culture replicates and four
parental cell culture replicates (controls). Each let-7a-3expressing
replicate was hybridized against a control replicate of the same cell line,
including a dye swap design. One-round linear amplification of 2 Ag total
RNA was done using the Low RNA Input Fluorescent Linear Amplification
kit (Agilent Technologies) according to the manufacturers instructions.
Hybridization and washing were done as described previously (13).
Hybridized arrays were scanned with the GenePix 4000B microarray
scanner (Axon Instruments), and analyzed using GenePix Pro 4.1.
Microarray raw data processing was done using software ArrayMagic
(14) and the data set was deposited to National Center for Biotechnology
Information Gene Expression Omnibus (GEO) database4 by GEO series
accession number GSE6474. We considered only genes that fulfilled the
cutoff criteria of a P value V0.05 and a linear fold change z1.5. Functional
annotation of the differentially expressed genes was done using the Gene
Ontology software GOstat.

Results and Discussion

Recently, 17 of 313 human miRNA genes were found to be
up-regulated in T24 human bladder cancer cells that had been
treated with the DNA methyltransferase inhibitor DAC and the
histone deacetylase inhibitor 4-phenylbutyric acid (10). The
observation that gene reactivation required a DNA methyltransfer-
ase inhibitor raised the possibility that a larger proportion of
miRNA genes might be methylated in human cells. We noticed that
the let-7a-3 gene on chromosome 22q13.31 is associated with a
well-defined CpG island (200 bp length, 55% CG content, 0.65 Figure 1. Methylation of let-7a-3 is cooperatively maintained by DNMT1 and
DNMT3B. A, let-7a-3 is embedded in a CpG island. The percentage of G+C over
observed/expected CG ratio; Fig. 1A). We used bisulfite sequencing a 100-bp window. Vertical bars, position of individual CpG dinucleotides; black
to analyze the methylation status of 33 CpG dinucleotides of the box, position of the let-7a-3 pri-miRNA; arrowheads, bisulfite PCR primers.
B, bisulfite sequencing analysis of let-7a-3 methylation with genomic DNA from
HCT116 cells and isogenic DNA methyltransferase knockout cell lines. o,
unmethylated CpG dinucleotides; ., methylated CpG dinucleotides. Percentages
4
http://www.ncbi.nlm.nih.gov/projects/geo/ indicate the fraction of methylated CpG dinucleotides.

Cancer Res 2007; 67: (4). February 15, 2007 1420 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on September 11, 2014. 2007 American Association for
Cancer Research.
 miRNA Gene Methylation

Having shown that let-7a-3 can be hypomethylated in human

lung cancer and that hypomethylation can lead to epigenetic
activation of the gene, we then sought to determine the
consequences of let-7a-3 expression in a human lung cancer
model. Let-7a-3 is strongly methylated and transcriptionally
silenced in the A549 adenocarcinoma cell line (Fig. 4A). To restore
let-7a-3 expression, we stably transfected A549 cells with a
construct that drives let-7a-3 pri-miRNA expression from a
cytomegalovirus promoter (L7-A549). For controls, we also trans-
fected A549 cells with the empty vector (Co-A549). Let-7a-3
expression was analyzed by semiquantitative RT-PCR in all trans-
genic cell lines and was found to be clearly increased in let-7a-3
transfected cells (Fig. 4A). To explore the functional consequences
of let-7a-3 expression, we analyzed the anchorage-independent cell
growth in soft agar assays. Staining and counting of colonies from
several independent experiments showed an 11-fold increase in
colony numbers for let-7a-3expressing cells when compared with
controls (Fig. 4B). These data provide a strong indication that
let-7a-3 can have oncogenic functions in lung cancer cells.
To further substantiate this observation, we obtained gene
expression profiles from L7-A549 and Co-A549 cells. Hybridization
of total RNA from both cell lines to cDNA microarrays revealed a
substantial let-7a-3associated transcriptional deregulation. Linear

Figure 2. Let-7a-3 is methylated in normal human tissues and can be

hypomethylated in lung cancer. A, COBRA analysis of let-7a-3 methylation in
normal human tissues. B, COBRA analysis of let-7a-3 methylation in eight
sample pairs from lung adenoma patients. N, samples from normal lung tissue;
T, samples from tumors. C, bisulfite sequencing analysis of let-7a-3 in patient D.
Percentages indicate the fraction of methylated CpG dinucleotides.

indicated that let-7a-3 was methylated in all normal lung samples,

with a methylation pattern similar to other human tissues (Fig. 2B).
Patients A and D showed a clear difference between tumor and
control tissue and the restriction pattern indicated strong
hypomethylation in the tumor (Fig. 2B). This observation was
confirmed by bisulfite sequencing of multiple independent clones,
which showed almost complete let-7a-3 demethylation (2%) in a
lung adenoma sample from patient D, whereas the locus was
substantially methylated (55%) in normal lung tissue from the same
patient (Fig. 2C). We therefore concluded that let-7a-3 is strongly
methylated in normal human tissues and that the gene can be
hypomethylated in lung tumors.
To analyze the role of DNA hypomethylation in the regulation of
let-7a-3, we treated A549 lung adenocarcinoma and HCT116 cells
with the DNA methyltransferase inhibitor DAC. COBRA analysis
of let-7a-3 showed that drug treatment caused effective gene
demethylation (Fig. 3A). Bisulfite sequencing of multiple indepen-
dent let-7a-3 clones also showed a decrease from 92% methylation
in untreated control cells to 42% in DAC-treated cells (Fig. 3B). Figure 3. Hypomethylation of let-7a-3 facilitates epigenetic reactivation.
A, COBRA analysis of let-7a-3 methylation in A549 and HCT116 cells indicates
Combinatorial treatment with DAC and the histone deacetylase effective demethylation following DAC treatment. B, bisulfite sequencing
inhibitor VPA caused a significant up-regulation of let-7a-3 analysis of let-7a-3 methylation with genomic DNA from untreated and
DAC-treated cells, respectively. o, unmethylated CpG dinucleotides; .,
expression, whereas treatment with either DAC or VPA alone had methylated CpG dinucleotides. Percentages indicate the fraction of methylated
little effect (Fig. 3C). These results agree with the observations CpG dinucleotides. C, epigenetic reactivation of let-7a-3 expression by
made for the human miR-127 gene (10) and suggest that combinatorial treatment with DAC and VPA. The results were derived from three
independent experiments, and all values were normalized to the expression
hypomethylation facilitates epigenetic reactivation of let-7a-3 level of h-amyloid in untreated HCT116 cells. Columns, relative let-7a-3
expression. expression; bars, SD.

www.aacrjournals.org 1421 Cancer Res 2007; 67: (4). February 15, 2007

Downloaded from cancerres.aacrjournals.org on September 11, 2014. 2007 American Association for
Cancer Research.
Cancer Research

In addition, down-regulation was observed for several genes that

have been described to inhibit lung cancer cell proliferation, such as
PPARG, TGFB2, and SFRP1. In addition, genes that are functionally
related to cell adhesion processes, a gene ontology category, which
is relevant for tumor progression and metastasis formation, were
predominantly down-regulated in our experiments (Supplementary
Table S2). The gene expression profiles are therefore in agreement
with the increased anchorage-independent growth of let-7a-3
expressing A549 cells and provide further confirmation for an
oncogenic role of let-7a-3 in lung carcinogenesis.
Various reports have linked let-7 miRNAs to human cancers
(1518). These observations, combined with the fact that the
human let-7a-3 gene was embedded in a well-defined CpG island,
prompted us to analyze the methylation and the function of this
gene in detail. It should be noted that the human let-7 family
encompasses at least 12 genes with distinct and potentially diverse
contributions to tumorigenesis and that only some of the previous
studies have discriminated between individual human let-7 genes.
This might explain why we have found a consistent oncogenic role
of let-7a-3, whereas others have described tumor-suppressing
activities for other let-7 miRNAs (1518). An oncogenic function of
let-7a-3 is also supported by our gene expression profiling results
that indicate increased oncogenic characteristics following expres-
Figure 4. Increased let-7a-3 expression causes oncogenic changes in human sion of let-7a-3 in A549 cells. Our microarray data support a role of
cancer cells. A, semiquantitative RT-PCR analysis of let-7a-3 pri-miRNA let-7a-3 in the regulation of RAS signaling, as described by others
expression in let-7a-3transfected (L7) and empty vector-transfected (Co ) A549 (1518). However, the observed effects were complex and not
cells. B, colony numbers of control and let-7a-3 overexpressing (L7) A549
cells in soft agar assays. The results were derived from three independent limited to down-regulation of RAS effector genes.
experiments. Columns, colonies > 100 Am; bars, SD. C, gene expression In agreement with an oncogenic function of let-7a-3 in a lung
profiling of L7-A549 and Co-A549 cells using cDNA microarrays.
Hierarchical clustering of 197 differentially expressed genes in four replicates.
cancer model, we found let-7a-3 to be substantially hypomethy-
D, let-7a-3dependent deregulation in the expression of RAS-responsive genes. lated in some lung cancer samples. To our knowledge, this finding
fc, fold changes in transcript levels in L7-A549 versus Co-A549 cells. represents the first example for an epigenetic mutation affecting a
miRNA gene. Global hypomethylation of genomic DNA has been
modeling of microarray data with cutoff values at P V 0.05 and the first epigenetic change described in human cancers, and several
linear fold change z1.5 revealed that let-7a-3 expression caused oncogenes, including BCL-2 (19) and R-RAS (20), are hypomethy-
deregulation of 197 genes (Fig. 4C; Supplementary Table S1). Gene lated in primary human tumors. More extensive studies will be
ontology statistics revealed overrepresentation of genes involved in required to comprehensively address the significance of aberrant
cell proliferation, adhesion, and differentiation (Supplementary miRNA gene methylation for human tumorigenesis.
Table S2). In agreement with a previous study showing a role of
let-7 in the regulation of RAS (17), we found 19 RAS-responsive Acknowledgments
genes to be deregulated in L7-A549 cells (Fig. 4D). RAS mRNA levels
Received 11/3/2006; revised 12/15/2006; accepted 12/29/2006.
were only weakly affected by let-7a-3 overexpression (data not Grant support: German National Genome Research Network grant NGFN-2
shown), which is consistent with previous data, suggesting that (F. Lyko).
The costs of publication of this article were defrayed in part by the payment of page
let-7dependent RAS regulation occurs at the translational level charges. This article must therefore be hereby marked advertisement in accordance
(17). We also identified several potentially oncogenic genes that with 18 U.S.C. Section 1734 solely to indicate this fact.
were up-regulated by let-7a-3 expression in A549 cells and that We thank Andreas Buness and Markus Ruschhaupt for their help in the statistical
analysis of microarray experiments; Bert Vogelstein for DNMT knockout cells; Heike
have been described previously to be associated with lung cancer Allgayer, Petra Boukamp, Amir Eden, and Thomas Muley for DNA samples; and the
progression. These genes include CDK6, PCNA, PRDX1, and CXCL5. tissue bank of the Thoraxklinik Heidelberg for the lung tumor tissue.

References 5. Jones PA, Baylin SB. The fundamental role of histone deacetylase inhibition in human colorectal
epigenetic events in cancer. Nat Rev Genet 2002;3: cancer. Nat Genet 2002;31:1419.
1. Esquela-Kerscher A, Slack FJ. OncomirsmicroRNAs 41528. 10. Saito Y, Liang G, Egger G, et al. Specific
with a role in cancer. Nat Rev Cancer 2006;6:25969. 6. Ehrlich M. DNA methylation in cancer: too much, but activation of microRNA-127 with downregulation of
2. Calin GA, Ferracin M, Cimmino A, et al. A MicroRNA also too little. Oncogene 2002;21:540013. the proto-oncogene BCL6 by chromatin-modifying
signature associated with prognosis and progression in 7. Esteller M. Relevance of DNA methylation in drugs in human cancer cells. Cancer Cell 2006;9:
chronic lymphocytic leukemia. N Engl J Med 2005;353: the management of cancer. Lancet Oncol 2003;4: 43543.
1793801. 3518. 11. Rhee I, Bachman KE, Park BH, et al. DNMT1 and
3. Lu J, Getz G, Miska EA, et al. MicroRNA expression 8. Yamashita K, Upadhyay S, Osada M, et al. Pharmaco- DNMT3b cooperate to silence genes in human cancer
profiles classify human cancers. Nature 2005;435:8348. logic unmasking of epigenetically silenced tumor cells. Nature 2002;416:5526.
4. Volinia S, Calin GA, Liu CG, et al. A microRNA suppressor genes in esophageal squamous cell carcino- 12. Brueckner B, Boy RG, Siedlecki P, et al. Epigenetic
expression signature of human solid tumors defines ma. Cancer Cell 2002;2:48595. reactivation of tumor suppressor genes by a novel small-
cancer gene targets. Proc Natl Acad Sci U S A 2006;103: 9. Suzuki H, Gabrielson E, Chen W, et al. A genomic molecule inhibitor of human DNA methyltransferases.
225761. screen for genes upregulated by demethylation and Cancer Res 2005;65:630511.

Cancer Res 2007; 67: (4). February 15, 2007 1422 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on September 11, 2014. 2007 American Association for
Cancer Research.
 miRNA Gene Methylation

13. Schneider J, Buness A, Huber W, et al. Systematic and genomic regions involved in cancers. Proc Natl microRNA molecular profiles in lung cancer diagnosis
analysis of T7 RNA polymerase based in vitro linear RNA Acad Sci U S A 2004;101:29993004. and prognosis. Cancer Cell 2006;9:18998.
amplification for use in microarray experiments. BMC 16. Takamizawa J, Konishi H, Yanagisawa K, et al. 19. Hanada M, Delia D, Aiello A, Stadtmauer E, Reed JC.
Genomics 2004;5:29. Reduced expression of the let-7 microRNAs in human bcl-2 gene hypomethylation and high-level expression in
14. Buness A, Huber W, Steiner K, Sultmann H, lung cancers in association with shortened postopera- B-cell chronic lymphocytic leukemia. Blood 1993;82:
Poustka A. ArrayMagic: two-colour cDNA microarray tive survival. Cancer Res 2004;64:37536. 18208.
quality control and preprocessing. Bioinformatics 17. Johnson SM, Grosshans H, Shingara J, et al. RAS is 20. Nishigaki M, Aoyagi K, Danjoh I, et al. Discovery of
2005;21:5546. regulated by the let-7 microRNA family. Cell 2005;120: aberrant expression of R-RAS by cancer-linked DNA
15. Calin GA, Sevignani C, Dumitru CD, et al. Human 63547. hypomethylation in gastric cancer using microarrays.
microRNA genes are frequently located at fragile sites 18. Yanaihara N, Caplen N, Bowman E, et al. Unique Cancer Res 2005;65:211524.

www.aacrjournals.org 1423 Cancer Res 2007; 67: (4). February 15, 2007

Downloaded from cancerres.aacrjournals.org on September 11, 2014. 2007 American Association for
Cancer Research.