Content uploaded by Scott Powers
Author content
All content in this area was uploaded by Scott Powers
Content may be subject to copyright.
A microRNA polycistron as a potential human
oncogene
Lin He
1
*, J. Michael Thomson
2
*, Michael T. Hemann
1
, Eva Hernando-Monge
4
, David Mu
1
, Summer Goodson
2
,
Scott Powers
1
, Carlos Cordon-Cardo
4
, Scott W. Lowe
1
, Gregory J. Hannon
1
& Scott M. Hammond
2,3
To date, more than 200 microRNAs have been described in
humans; however, the precise functions of these regulatory, non-
coding RNAs remains largely obscure. One cluster of microRNAs,
the mir-17–92 polycistron, is located in a region of DNA that is
amplified in human B-cell lymphomas
1
. Here we compared B-cell
lymphoma samples and cell lines to normal tissues, and found that
the levels of the primary or mature microRNAs 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 tumour development in a mouse B-cell
lymphoma model. Tumours 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. Together, these studies
indicate that non-coding RNAs, specifically microRNAs, can
modulate tumour formation, and implicate the mir-17–92 cluster
as a potential human oncogene.
MicroRNAs (miRNAs) have emerged relatively recently as a new
class of small, non-coding RNAs that regulate gene expression.
Nascent primary miRNA transcripts (pri-miRNAs) are processed
sequentially by two RNase III enzymes, Drosha and Dicer
2,3
, to yield
mature miRNAs, ranging from 18 to 24 nucleotides (nt) in length.
miRNAs are incorporated into the RNA interference (RNAi) effector
complex, RISC, and target specific messenger RNAs for translational
repression or mRNA cleavage
4–6
. Although hundreds of miRNAs
have been cloned and/or predicted, only a handful have been
functionally characterized. For example, lin-4 and let-7 regulate the
timing of larval development in C. elegans
7,8
. Left/right asymmetric
gene expression in C. elegans chemosensory neurons is controlled by
lsy-6 and miR-273 (refs 9, 10). Bantam stimulates cell growth and
prevents apoptosis in Drosophila
11
, and miR-181 potentiates B-cell
differentiation in mammals
12
. These findings, in combination with
computational target predictions, are consistent with miRNAs regu-
lating a broad spectrum of physiological and developmental
processes.
Microarray-based expression studies have indicated specific
alterations in human miRNA expression profiles that correlate with
particular tumour phenotypes (J.M.T. and S.M.H., unpublished
data). Among those that show altered expression, the mir-17–92
cistron is located at 13q31, a genomic locus that is amplified in cases
of diffuse large B-cell lymphoma, follicular lymphoma, mantle cell
lymphoma, primary cutaneous B-cell lymphoma and several other
tumour types
1,13
. There are only two annotated genes in the epicentre
of this amplicon, c13orf25 and GPC5. Previous studies have shown
that c13orf25 is the only one of the two genes for which increased
expression correlates with the presence of the amplicon
1
. Therefore,
c13orf25 had been implicated as a target of the 13q31 amplicon
1
.Itis
unlikely that c13orf25 actually encodes a protein, as predicted open
reading frames (ORFs) encode only short peptides (,70 amino
acids), which are not conserved in closely related species. Instead, the
c13orf25 transcript appears to be the functional precursor of a series
of seven microRNAs: miR-17-5p, miR-17-3p, miR-18, miR-19a,
miR-20, miR-19b-1 and miR-92-1 (Fig. 1a). Additionally, this cluster
is related to the homologous mir-106a–92 cluster on chromosome X
and the mir-106b–25 cluster on chromosome 7 (ref. 14; Fig. 1a).
Alignment of the human c13orf25 locus and its murine orthologue
revealed extensive sequence conservation only within the mir-17–92
polycistron and its immediate flanking sequence. Several of the
expressed-sequence-tags (ESTs) derived from c13orf25 and its
mouse orthologue terminate at the 3 0end of mir-17–92 cluster,
consistent with the presence of a Drosha processing site at this
location (Fig. 1b).
A principal consequence of 13q31-q32 amplification could be
an increase in the abundance of mature miRNA species from the
mir-17–92 cluster. We acquired four cell lines previously described as
carrying amplifications in the 13q31-q32 region
1
and confirmed the
gene dosage increase at the c13orf25 locus in three of those cell lines
using quantitative polymerase chain reaction (PCR) analysis.
The abundance of 191 mature miRNAs was assessed in these four
cell lines and compared to normal B-cells, and to five leukaemia and
lymphoma cell lines lacking the amplicon (Fig. 1c and Supplemen-
tary Fig. 1). Using a significance analysis of microarrays (SAM)
analysis
15
, we identified six miRNAs for which high-level expression
correlated with increased gene dosage of c13orf25 (see Supple-
mentary Table 1). Five were from the mir-17–92 cistron, and
the sixth, miR-106a, was identified as a probable result of cross-
hybridization to miR-17-5p, from which it differs at only two
terminal nucleotides (Fig. 1c). This hypothesis is supported by the
observation that the mir-106a–92 locus does not show copy number
alterations in these cell lines (not shown). In each cell line, expression
levels correlated with the copy number of the mir-17–92 locus (Fig.
1c, lower panel).
We also examined the expression of pri-mir-17–92 in a series of
human tumour samples comprising both lymphomas and colorectal
carcinomas. Of 46 lymphoma samples, including 13 diffuse large
B-cell lymphomas and 6 follicular lymphomas, we saw significant
(greater than fivefold) overexpression of pri-mir-17–92 in 65% of the
samples. Considering all of the B-cell lymphoma samples analysed,
the average increase in pri-miRNA expression was about tenfold
(Fig. 1d). In contrast, colorectal carcinomas rarely showed over-
expression of the pri-miRNA. Increases in expression from this locus
were less common (15% of samples showed greater than fivefold
LETTERS
1
Cold Spring Harbor Laboratory, Watson School of Biological Sciences, 1 Bungtown Road, Cold Spring Harbor, New York 11724, USA.
2
Department of Cell and Developmental
Biology and
3
Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, North Carolina 27599, USA.
4
Memorial Sloan-Kettering Cancer Center, Division
of Molecular Pathology, Memorial Sloan-Kettering Cancer Center, 1275 York Ave, New York, New York 10021, USA.
*These authors contributed equally to this work.
Vol 435|9 June 2005|doi:10.1038/nature03552
828 © 2005 Nature Publishing Group
Figure 1 |The mir-17–92 cluster shows increased expression in B-cell
lymphoma samples and cell lines. a, Genomic organization of three
polycistronic miRNA clusters is shown. There are five paralogous groups
located in three homologous clusters (mir-17–92,mir-106a–92 and
mir-106b–25) with a conserved order: miR-17-5p/miR-106a/miR-106b;
miR-18;miR-19a/miR-19b-1/miR-19b-2;miR-20/miR-93;andmiR-92-1/
miR-92-2/miR-25 (yellow boxes, pre-miRNAs; purple boxes, mature
miRNAs). b, The level of conservation between human and mouse
homologues is represented using an mVista plot
28
. Two alternative isoforms
have been detected for the human gene, and these are shown schematically
1
(dark blue, exons; light blue, introns; orange, the mir-17–92 cluster).
c, MicroRNA expression levels in cell lines carrying the 13q31-q32 amplicon
(including Karpas 1718, OCI-Ly4 and OCI-Ly7) were compared to those in
leukaemia and lymphoma cell lines lacking this genetic lesion, and to normal
B-cells isolated from cortical blood (top panel). We included in this analysis
the OCI-Ly8 cell line, which has previously been identified as a cell line
carrying the 13q31-32 amplicon, but showed no gene dosage increase at the
c13orf25 locus in our study. Normalized one-channel measurements for 191
human miRNAs were hierarchically clustered for all miRNA genes
represented on the array. An excerpt of the data is shown, and the full cluster
analysis is presented in Supplementary Fig. 1. The expression map node that
correlates with the amplification is shown. The let-7 cluster node is also
shown for comparison (middle panel). In the cell lines examined, expression
levels of the mature microRNAs from the mir-17–92 polycistron correlate
with the copy number at the mir-17–92 locus (bottom panel). d, The level of
mir-17–92 pri-miRNA was determined by real-time quantitative PCR in 46
lymphomas and 47 colorectal carcinomas, and compared to levels found in
corresponding normal tissues from five individuals. In cand d, error bars
indicate standard deviation (s.d.).
NATURE|Vol 435|9 June 2005 LETTERS
829
© 2005 Nature Publishing Group
upregulation), and the degree of overexpression was substantially
lower (Fig. 1d).
Considered together, our data prompted the hypothesis that mir-
17–92 might contribute to tumour development. To test this idea
directly, we used a mouse model of human B-cell lymphoma.
Transgenic animals carrying a c-myc oncogene, driven by the
immunoglobulin heavy-chain enhancer (Em), develop B-cell lym-
phomas by 4–6 months of age
16
. Similarly, haematopoietic stem cells
(HSCs) derived from fetal livers of Em-myc transgenic mice generate
B-cell lymphomas with comparable latency when transplanted into
lethally irradiated recipients
17–20
(Fig. 2a). We therefore infected
Em-myc/þHSCs with a murine stem cell virus (MSCV) retrovirus
that directs expression of a truncated cluster comprising mir-17–19b-1
(hereafter mir-17–19b), the vertebrate-specific portion of the
mir-17–92 miRNA cistron (Fig. 1a). This virus also contained a
green fluorescent protein (GFP) transgene, allowing us to follow
infected stem cells in vitro and in vivo (Fig. 2a). Mice reconstituted
with Em-myc/þHSCs carrying a control MSCV vector developed
lymphomas after the expected latency (3–6 months), with incomplete
penetrance (Fig. 2b). Similarly, we examined .40 animals reconsti-
tuted with Em-myc/þHSCs expressing subsets of 96 different, single
microRNAs (see Supplementary Table 2). Although we did not
confirm miRNA overexpression for every case, we did not observe
any significantly accelerated onset of disease. In contrast, 100% of
the animals co-expressing the mir-17–19b polycistron and c-myc
developed leukaemias at an average of 51 days following transplan-
tation (s.d. ^13 days, P,0.0001 compared with MSCV controls
using the logrank test), and died of B-cell lymphomas at an average age
of 65 days (s.d. ^13 days, P,0.0001 compared to MSCV controls;
Fig. 2b). In all but one case, primary lymphomas could be
visualized by virtue of the linked GFP marker (Fig. 2c and Table
1). The mature miRNAs from the mir-17–19b cluster show high-
level expression in these tumours, compared with miRNAs from the
paralogous mir-106a–92 locus, and also have similar mir-17–19b
expression levels compared to the Karpas 1718 lymphoma cell line,
which has increased c13orf25 gene dosage (see Supplementary
Fig. 2).
The full mir-17–92 cistron was also tested in a small cohort of
animals. Although similar results were obtained compared to those
animals reconstituted with HSCs expressing mir-17–19b, studies in
cell lines indicated that the construct used to express the entire cluster
gave lower levels of mature miRNAs. We therefore focused most of
our study on the truncated mir-17–19b cluster. In these ongoing
studies, we have yet to find any individual miRNA from the
Figure 2 |Overexpression of the mir-17–19b cluster accelerates
c-myc-induced lymphomagenesis in mice. a, Schematic representation of
the adoptive transfer protocol using Em-myc HSCs. b, Mice reconstituted
with HSCs expressing mir-17–19b in an MSCV retroviral vector (MSCV
mir-17–19b) or infected with a control MSCV virus were monitored by
blood smear analysis starting 5 weeks after transplantation. The Kaplan-
Meier curves represent the percentage of leukaemia-free survival or overall
survival. c, External GFP imaging of tumour-bearing mice with Em-myc/mir-
17–19b or Em-myc/MSCV shows the overall distribution of tumour cells.
Em-myc/mir-17–19b tumours show a more disseminated phenotype
compared with control tumours. These animals are representative of their
genotype.
LETTERS NATURE|Vol 435|9 June 2005
830 © 2005 Nature Publishing Group
mir-17–19b cluster that can accelerate tumour formation to the
extent seen with the intact polycistron (not shown).
The Em-myc/mir-17–19b lymphomas are true malignancies rather
than hyperplasias, because primary tumour cells, when transplanted
into C57B6/J recipients, induce B-cell lymphomas in 2–3 weeks
that result in lethality after 4–5 weeks (data not shown). The
secondary tumours show pathological features indistinguishable
from the original tumours, and retain tumorigenic potential after
two additional rounds of serial transplantation (data not shown).
Therefore, a microRNA cluster can accelerate Em-myc induced
tumorigenesis in mice.
The pathological hallmarks of Em-myc/mir-17–19b mosaic animals
included massive enlargement of lymph nodes, splenic hyperplasia,
infiltration of the thymus by lymphoma cells, and leukaemia (Fig. 2c).
Animals with advanced lymphomas showed extramedullary haema-
topoiesis due to functional failure of the bone marrow. Furthermore,
6 out of 14 animals showed hind limb paralysis, associated with
substantial tumour growth at the lumbar node. Tumours resulting
from combined c-myc and mir-17–19b expression consistently
invaded visceral organs outside the lymphoid compartment, includ-
ing liver, lung and occasionally kidney (Figs 2c, 3b and Table 1).
Additionally, Em-myc/mir-17–19b lymphomas show a high mitotic
index without extensive apoptosis (Fig. 3a). This contrasts with the
Em-myc/MSCV tumours lacking the microRNA cluster, which show a
high degree of apoptosis (Fig. 3a). These findings indicate that
cooperation between Em-myc and mir-17–19b gives rise to highly
malignant, disseminated lymphomas capable of evading normal
apoptotic responses to inappropriate proliferation.
Em-myc-induced lymphomas originate from the B-lymphoid line-
age. However, the developmental characteristics of these tumour cells
are not stage-specific, as they can resemble either mature B cells or
pre-B cells. To examine the cell lineage of the Em-myc/mir-17–19b
lymphomas, we assessed the expression of cell-surface markers,
including the B-cell-specific marker B220 and the T-cell-specific
markers CD4 and CD8a. All tumours were found to be of B-cell
origin, staining positive for B220 and negative for both CD4 and
CD8a (Fig. 3c and Table 1). We next analysed these tumours for
CD19 and IgM expression to distinguish pre-B from mature B cells.
With one exception, Em-myc/mir-17–19b lymphomas were derived
purely from the pre-B cell lineage (low to absent Thy1.2 staining,
CD19
þ
B220
þ
IgM
2
) (Table 1), suggesting that overexpression of
mir-17–19b strongly favours transformation of B-cell progenitors
under our experimental conditions.
Exactly how overproduction of mir-17–19b might promote onco-
genesis is unclear. At a minimum, studies of tumour pathology
suggest that increased expression of this cluster mitigates the pro-
apoptotic response to elevated myc expression in vivo. Furthermore,
we have previously shown that this miRNA cistron is highly
expressed in embryonic stem cells, with expression levels decreasing
during embryonic development in mice
21
. It is therefore tempting to
speculate that these miRNAs promote ‘stem cell’ properties or specify
characteristics of early developmental lineages. A detailed, mechan-
istic understanding of how this non-coding RNA cluster acts as an
oncogene is at present hampered by the lack of a validated biochemi-
cal strategy for identifying miRNA targets.
Previous circumstantial evidence has indicated the potential
involvement of a number of miRNAs in tumorigenesis. Although
miRNAs only represent 1% of the mammalian genome, more than
50% of miRNA genes are located within regions associated with
amplification, deletion and translocation in cancer
22
. Expression
studies of various tumour types have also revealed specific alterations
in miRNA profiles
22–25
. For example, mir-15 and mir-16 are fre-
quently deleted and/or downregulated in B-cell chronic lymphocytic
leukaemia
26
, miR-143 and miR-145 show decreased abundance in
colorectal neoplasia
25
, and miR-155 and its non-coding RNA host
gene, BIC, are upregulated 100-fold in Burkitt’s lymphoma
patients
24
. Here, we have shown that one miRNA polycistron is not
only the subject of tumour-specific amplification, but that it is also
overexpressed in tumours and tumour cell lines, and can act as an
oncogene in vivo. Our results indicate that non-coding RNAs might
act as integral parts of the molecular architecture of oncogene and
tumour suppressor networks. Such oncogenic microRNAs might be
designated ‘oncomiRs, with mir-17–92 being oncomiR-1.
METHODS
miRNA expression profiling. Five micrograms of total RNA was labelled with
RNA ligase and a Cy3-conjugated dinucleotide, and hybridized to custom
oligonucleotide microarrays as described in ref. 21. Cy3 median intensity values
Table 1 |Phenotypic analysis of a subset of Em-myc/mir-17–19b tumours
Animal GFP Immunophenotyping Cell type Ki67 staining Apoptosis* Pathological features
1þB220
þ
, Thy1.2
2
, IgM
2
,
CD19
þ
, CD4
2
, CD8
2
Pre-B cell þþ(70–80%) Low Tumour cells invaded liver and spleen;
mild infiltrations observed in lung
and kidney; spleen enlarged; hindlimb
paralysis
2þB220
þ
, Thy1.2
low
, IgM
2
,
CD19
þ
, CD4
2
, CD8
2
Pre-B cell þþ(80–90%) Low Tumour cells invaded liver, lung and
spleen; spleen enlarged
3þB220
þ
, Thy1.2
2
, IgM
2
,
CD19
þ
, CD4
2
, CD8
2
Pre-B cell þþ(80–90%) Low Tumour cells invaded liver, lung and
spleen; spleen enlarged
4þB220
þ
, Thy1.2
low
, IgM
2
,
CD19
þ
, CD4
2
, CD8
2
Pre-B cell þþ(70–80%) Low Tumour cells invaded liver and spleen;
spleen enlarged
5þB220
þ
, Thy1,2
2
, IgM
2
,
CD19
þ
, CD4
2
, CD8
2
Pre-B cell þþ(80–90%) Slightly less
than control
Highly disseminated lymphoma; tumour
cells invaded liver, spleen, lung and
kidney; spleen enlarged
6þB220
þ
, Thy1,2
2
, IgM
2
,
CD19
þ
, CD4
2
, CD8
2
Pre-B cell þþ(80–90%) Low Highly disseminated lymphoma; tumour
cells invaded liver, spleen, lung and
kidney; spleen enlarged
7þB220
þ
, Thy1.2
2
, IgM
2
,
CD19
þ
, CD4
2
, CD8
2
Pre-B cell þþ(70–80%) Low Spleen enlarged; mild infiltration of
tumour cells into liver only
8þB220
þ
, Thy1.2
low
, IgM
2
,
CD19
þ
, CD4
2
, CD8
2
and
B220
þ
, Thy1.2
low
, IgM
2
,
CD19
þ
, CD4
2
, CD8
2
†
Pre-B cell and
mature B cell
þþ(70–80%) Low Highly disseminated lymphoma; tumour
cells invaded liver, spleen, lung and
kidney; enlarged spleen
92ND ND þþ(80–90%) Low Tumour cells invaded liver, lung and
spleen; enlarged spleen
*Levels of apoptosis in Em-myc/mir-17–19b lymphomas are compared to control Em-myc/MSCV lymphomas on the basis of haemotoxylin and eosin staining, and TUNEL staining.
†The two clones of tumour cells with different cell type specificity might reflect independent transformation events or maturation of a single primary clone.
ND, not determined.
NATURE|Vol 435|9 June 2005 LETTERS
831
© 2005 Nature Publishing Group
were filtered to remove data points that did not exceed background levels by
twofold. Values were log
2
-transformed and median-centred by array. Clustering
was performed with Cluster (Stanford University), using values that were
median-centred by gene. Dendrograms and expression maps were generated
using Treeview (Stanford).
Cell lines. The measurement of miRNA abundance was carried out using the
following human cell lines: Karpas 1718 (derived from splenic lymphoma with
villous lymphocytes, provided by A. Karpas), OCI-Ly4, OCI-Ly7 and OCI-Ly8
(derived from diffuse large B-cell lymphoma, provided by R. Dalla-Favera). The
cell lines lacking the 13q31-q32 amplicon were Raji (B-cell, derived from
Burkitt’s lymphoma, ATCC); Namalwa (B-cell, derived from Burkitt’s lym-
phoma, ATCC); HG 1125 (EBV-transformed human lymphoblastoid, provided
by B. Stillman); Manca (lymphoblast-like, derived from chronic myelogenous
leukaemia); Jurkat; proliferating B-cells (spleenic B-cells isolated from a C57B6/
Ly5.2 mouse and stimulated to proliferate in culture with lipopolysaccharide,
provided by I. Weissman); and normal B cells (derived from cord blood,
Cambrex).
PCR and copy number analysis. Tumour samples were obtained from the
Cooperative Human Tissue Network (http://www-chtn.ims.nci.nih.gov). Cor-
responding normal tissue RNA from five individuals was purchased from
Biochain Institute Inc. For fluorogenic real-time PCR, primers that amplify
mir-17–92 pri-miRNA and b-actin mRNA (control), and probes were designed
using Primer Express software (V.2): mir-17–92 forward primer, 30-CAGTAAA
GGTAAGGAGAGCTCAATCTG-50; reverse primer, 30-CATACAACCACTAA
GCTAAAGAATAATCTGA-50;mir-17–92 probe, (6-FAM)-TGGAAATAAGATC
ATCATGCCCACTTGAGAC-(TAMRA); b-actin forward primer, 30-GCAAAG
ACCTGTACGCCAACA-50; reverse primer, 30-TGCATCCTGTCGGCAATG-50;
b-actin probe, (6-FAM)-TGGCGGCACCACCATGTACC-(TAMRA). The ratios
of RNA species detected by mir-17–92 primers and b-actin primers in each RNA
sample were determined in triplicate by reverse-transcription, quantitative PCR
Figure 3 |Pathological and immunological analysis of lymphomas produced
by cooperation between mir-17–19b and c-myc.a, Haemotoxylin and eosin
(H&E), Ki67, B220 and TUNEL staining of Em-myc/mir-17–19b lymph node
tumours. The ‘starry sky’ morphology is a hallmark of cell clusters
undergoing apoptosis (black arrows). Scale bar, 10
m
m. b, Invasion of
visceral organs (liver, spleen, lung and kidney) by Em-myc/mir-17–19b
tumour cells, shown by H&E and B220 staining. Invasion was observed both
perivascularly and parenchymally in liver. Scale bar, 50
m
m.
c, Immunophenotyping of Em-myc/mir-17–19b lymphomas. Tumour cells
stained positively for the B-cell-specific marker B220, but not for the
T-cell-specific markers CD4, CD8a and Thy1.2. Tumour cells bore cellular
characteristics of pre-B cells, staining positively for CD19 but not for IgM, a
marker of mature B-cells.
LETTERS NATURE|Vol 435|9 June 2005
832 © 2005 Nature Publishing Group
using an ABI 7900HT Taqman sequence detector, following the ‘standard curve’
method. All values were subsequently normalized to the averaged ratio of the five
corresponding normal samples. For DNA copy number determination using the
ABI 7900HT sequence detector, we performed quantitative PCR analysis using
the same mir-17–92 primer set described above, and normalized the data against
a reference probe corresponding to chromosomal region 6p22 (forward primer,
30-GGTCTCTATTTGCACTTGGCTGAT-50; reverse primer, 30-TTTTCA
TTGTTGACCAAGCTAGACA-50; probe, (6-FAM)-TAGGGCATACTGCCTG
CATATTTCCTGCT-(TAMRA)) or a b-actin probe. The reported values rep-
resent the ratios of DNA copy number at the mir-17–92 locus to the normal
reference probe.
Adotptive transfer of Em-myc HSCs. Fetal liver-derived HSCs were isolated
from Em-myc/þmouse embryos at embryonic day (E)13.5–E15.5, and were
transduced with MSCV alone or MSCV expressing the mir-17–19b cluster. To
exclude the possibility that the observed acceleration of lymphomagenesis was
due to insertional mutagenesis, independent experiments were carried out using
fetal liver cells isolated from eight Em-myc/þembryos. Cells from each isolation
were separately infected with MSCV mir-17–19b or control MSCV. The MSCV
retroviral vector used in our studies contains a PGK-puromycin-IRES-GFP
(PIG) cassette
18
. Infection rates, assessed by fluorescence-activated cell
sorting (FACS), were typically 40% of bulk fetal liver cells. HSCs infected with
MSCV-mir17–19b-PIG and MSCV-PIG (control) were subsequently trans-
planted into 6–8-week-old, lethally irradiated C57BL/6 recipient mice
17
. Tumour
onset was monitored by blood smear analysis, and tumour samples were either
collected into 4% paraformaldehyde for histopathological studies, or prepared as
single-cell suspensions for FACS.
We also carried out a screen of 96 miRNAs, to look for miRNA(s) that
accelerate Myc-induced lymphomagenesis. In this experiment, each pre-miRNA
(and ,50 bp of flanking sequence 50and 3 0of the pre-miRNA) was cloned
downstream of the cytomegalovirus (CMV) promoter in a MSCV vector
containing SV40–GFP. Eight individual MSCV constructs, each overexpressing
a specific miRNA, were pooled at equal concentrations. Twelve pools of DNA
were used individually to produce virus to infect Em-myc/þfetal liver cells for
adoptive transfer into at least three recipient animals, as described above.
Recipient animals were monitored for tumour growth for at least six months.
For those that developed lymphoma, tumour cells were prepared from the
enlarged lymph nodes and then subjected to FACS analysis for GFP expression.
Histopathology. Tissue samples were fixed in 4% paraformaldehyde, embedded
in paraffin, sectioned into 5-
m
m slices, and stained with haemotoxylin and eosin.
For Ki67 detection (rabbit anti-Ki67 antibody, NovoCastra), representative
sections were deparaffinized, rehydrated in graded alcohols, and processed
using the avidin–biotin immunoperoxidase method. Sections were then sub-
jected to antigen retrieval by microwave oven treatment using standard
procedures. Diaminobenzidine was used as the chromogen and haematoxylin
as the nuclear counterstain. For B220 immunohistochemistry, a rat anti-mouse
CD45R/B220 antibody (clone RA3-6B2, BD Biosciences Pharmingen) was used;
pretreatment for antigen retrieval was not required. Analysis of the rate of
apoptotis by TUNEL assay was performed according to a published protocol
27
.
Received 15 February; accepted 16 March 2005.
1. Ota, A. et al. Identification and characterization of a novel gene, C13orf25,asa
target for 13q31-q32 amplification in malignant lymphoma. Cancer Res. 64,
3087–-3095 (2004).
2. Lee, Y. et al. The nuclear RNase III Drosha initiates microRNA processing.
Nature 425, 415–-419 (2003).
3. Bernstein, E., Caudy, A. A., Hammond, S. M. & Hannon, G. J. Role for a
bidentate ribonuclease in the initiation step of RNA interference. Nature 409,
363–-366 (2001).
4. Bartel, D. P. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell
116, 281–-297 (2004).
5. Ambros, V. The functions of animal microRNAs. Nature 431, 350–-355 (2004).
6. He, L. & Hannon, G. J. MicroRNAs: small RNAs with a big role in gene
regulation. Nature Rev. Genet. 5, 522–-531 (2004).
7. Ruvkun, G. & Giusto, J. The Caenorhabditis elegans heterochronic gene lin-14
encodes a nuclear protein that forms a temporal developmental switch. Nature
338, 313–-319 (1989).
8. Ambros, V. A hierarchy of regulatory genes controls a larva-to-adult
developmental switch in C. elegans.Cell 57, 49–-57 (1989).
9. Chang, S., Johnston, R. J. Jr, Frokjaer-Jensen, C., Lockery, S. & Hobert, O.
MicroRNAs act sequentially and asymmetrically to control chemosensory
laterality in the nematode. Nature 430, 785–-789 (2004).
10. Johnston, R. J. & Hobert, O. A microRNA controlling left/right neuronal
asymmetry in Caenorhabditis elegans.Nature 426, 845–-849 (2003).
11. Brennecke, J., Hipfner, D. R., Stark, A., Russell, R. B. & Cohen, S. M. bantam
encodes a developmentally regulated microRNA that controls cell proliferation
and regulates the proapoptotic gene hid in Drosophila.Cell 113, 25–-36 (2003).
12. Chen, C. Z., Li, L., Lodish, H. F. & Bartel, D. P. MicroRNAs modulate
hematopoietic lineage differentiation. Science 303, 83–-86 (2004).
13. Knuutila, S. et al. DNA copy number amplifications in human neoplasms:
review of comparative genomic hybridization studies. Am. J. Pathol. 152,
1107–-1123 (1998).
14. Tanzer, A. & Stadler, P. F. Molecular evolution of a microRNA cluster. J. Mol.
Biol. 339, 327–-335 (2004).
15. Tusher, V. G., Tibshirani, R. & Chu, G. Significance analysis of microarrays
applied to the ionizing radiation response. Proc. Natl Acad. Sci. USA 98,
5116–-5121 (2001).
16. Adams, J. M. et al. The c-myc oncogene driven by immunoglobulin enhancers
induces lymphoid malignancy in transgenic mice. Nature 318, 533–-538 (1985).
17. Schmitt, C. A. et al. Dissecting p53 tumor suppressor functions in vivo.Cancer
Cell 1, 289–-298 (2002).
18. Hemann, M. T. et al. An epi-allelic series of p53 hypomorphs created by stable
RNAi produces distinct tumor phenotypes in vivo.Nature Genet. 33, 396–-400
(2003).
19. Hemann, M. T. et al. Suppression of tumorigenesis by the p53 target PUMA.
Proc. Natl Acad. Sci. USA 101, 9333–-9338 (2004).
20. Wendel, H. G. et al. Survival signalling by Akt and eIF4E in oncogenesis and
cancer therapy. Nature 428, 332–-337 (2004).
21. Thomson, J. M., Parker, J., Perou, C. M. & Hammond, S. M. A custom
microarray plateform for analysis of microRNA gene expression. Nature
Methods 1, 47–-53 (2004).
22. Calin, G. A. et al. Human microRNA genes are frequently located at fragile sites
and genomic regions involved in cancers. Proc. Natl Acad. Sci. USA 101,
2999–-3004 (2004).
23. Calin, G. A. et al. MicroRNA profiling reveals distinct signatures in B cell
chronic lymphocytic leukemias. Proc. Natl Acad. Sci. USA 101, 11755–-11760
(2004).
24. Metzler, M., Wilda, M., Busch, K., Viehmann, S. & Borkhardt, A. High
expression of precursor microRNA-155/BIC RNA in children with Burkitt
lymphoma. Genes Chromosom. Cancer 39, 167–-169 (2004).
25. Michael, M. Z., O’Connor, S. M., van Holst Pellekaan, N. G., Young, G. P. &
James, R. J. Reduced accumulation of specific microRNAs in colorectal
neoplasia. Mol. Cancer Res. 1, 882–-891 (2003).
26. Calin, G. A. et al. Frequent deletions and down-regulation of micro- RNA genes
miR15 and miR16 at 13q14 in chronic lymphocytic leukemia. Proc. Natl Acad. Sci.
USA 99, 15524–-15529 (2002).
27. Di Cristofano, A., De Acetis, M., Koff, A., Cordon-Cardo, C. & Pandolfi, P. P.
Pten and p27
KIP1
cooperate in prostate cancer tumor suppression in the mouse.
Nature Genet. 27, 222–-224 (2001).
28. Mayor, C. et al. VISTA: visualizing global DNA sequence alignments of
arbitrary length. Bioinformatics 16, 1046–-1047 (2000).
Supplementary Information is linked to the online version of the paper at
www.nature.com/nature.
Acknowledgements We thank members of the Hannon, Lowe and Hammond
laboratories for discussions and input. We also thank Z. Xuan, N. Chen, N. Sheth
and R. Sachidanandam for bioniformatic analysis. C. Perou and J. Leib provided
advice and support for microarray methodologies, and A. Barnes and B. Boone
gave assistance with mouse tissue procurement. We are grateful to H. Wendel,
C. Scott, C. Marsden and C. Rosenthal, R. Karni, P. Moody and R. Whitaker, who
provided advice and technical support. F. Slack coined the oncomiR
nomenclature. L.H. and M.T.H. are Fellows of the Helen Hay Whitney
Foundation. S.W.L. and C.C.-C. are supported by a program project grant from
the NCI. G.J.H is supported by an Innovator Award from the US Army Breast
Cancer Research Program and by grants from the DOD and NIH. S.M.H. is a
General Motors Cancer Research Foundation Scholar, and J.M.T is a Frederick
Gardner Cottrell Postdoctoral Fellow.
Author Information Microarray data have been deposited in NCBI-GEO under
accession numbers GSM45026–GSM45065 and GSE-2399. Reprints and
permissions information is available at npg.nature.com/reprintsandpermissions.
The authors declare no competing financial interests. Correspondence and
requests for materials should be addressed to G.J.H. (hannon@cshl.edu) or
S.M.H. (hammond@med.unc.edu).
NATURE|Vol 435|9 June 2005 LETTERS
833
© 2005 Nature Publishing Group
