Truncating mutations of RB1CC1 in human breast cancer.

Abstract

The protein RB1CC1 (retinoblastoma 1 (RB1)-inducible coiled-coil 1) has been identified as a key regulator of the tumor-suppressor gene RB1 (ref. 1). RB1CC1 is localized in the nucleus and has been proposed to be a transcription factor because of its nuclear localization signal, leucine zipper motif and coiled-coil structure. The gene RB1CC1 is localized to a region of chromosome 8q11 (ref. 2) containing several loci of putative tumor-suppressor genes; however, its role in human cancers remains to be determined. Here we report that 20% (7 of 35) of primary breast cancers examined contained mutations in RB1CC1, including nine large interstitial deletions predicted to yield markedly truncated RB1CC1 proteins. Wildtype RB1CC1 and RB1 were absent or significantly less abundant than normal in the seven cancers with mutations in RB1CC1, but were abundant in cancers without such mutations. In all seven cancers, both RB1CC1 alleles were inactivated; two showed compound heterozygous deletions. Thus, RB1CC1 is frequently mutated in breast cancer and shows characteristics of a classical tumor-suppressor gene.

Full text

Page 1

letter
nature genetics • volume 31 • july 2002
285
Truncating mutations of RB1CC1 in human
breast cancer
Tokuhiro Chano1, Keiichi Kontani2, Koji Teramoto2, Hidetoshi Okabe1& Shiro Ikegawa3
Departments of 1Clinical Laboratory Medicine and 2Surgery II, Shiga University of Medical Science, Seta, Otsu, Shiga 520-2192, Japan. 3Laboratory for Bone and
Joint Diseases, SNP Research Center, RIKEN, Tokyo 108-8639, Japan. Correspondence should be addressed to T.C. (e-mail: chano@belle.shiga-med.ac.jp).
The protein RB1CC1 (retinoblastoma 1 (RB1)-inducible coiled-
coil 1) has been identified as a key regulator of the tumor-
suppressor gene RB1 (ref. 1). RB1CC1 is localized in the nucleus
and has been proposed to be a transcription factor because of
its nuclear localization signal, leucine zipper motif and coiled-
coil structure1,2. The gene RB1CC1 is localized to a region of
chromosome 8q11 (ref. 2) containing several loci of putative
tumor-suppressor genes3,4; however, its role in human cancers
remains to be determined. Here we report that 20% (7 of 35) of
primary breast cancers examined contained mutations in
RB1CC1, including nine large interstitial deletions predicted to
yield markedly truncated RB1CC1 proteins. Wildtype RB1CC1
and RB1 were absent or significantly less abundant than nor-
mal in the seven cancers with mutations in RB1CC1, but were
abundant in cancers without such mutations. In all seven can-
cers, both RB1CC1 alleles were inactivated; two showed com-
pound heterozygous deletions. Thus, RB1CC1 is frequently
mutated in breast cancer and shows characteristics of a classical
tumor-suppressor gene.
About 80% of human cancers are associated with dysregulation
of the RB1 pathway, and of RB1 itself5–7. Loss of heterozygosity
(LOH) and other alterations at the RB1 locus occur in 7–37% of
breast cancers5,8–12, suggesting that RB1 is important in the
tumorigenesis of breast cancer. But correlations between LOH at
the RB1 locus and expression of RB1 are not always consistent9,
and some breast cancers without detectable changes in RB1 show
no RB1 expression12. These findings imply that additional
unknown factors regulating RB1 expression are involved in
tumorigenesis of breast cancers.
Expression of RB1CC1 correlates closely with that of RB1 in
various cancer cell lines and normal human tissues, and intro-
ducing wildtype RB1CC1 into human leukemic cells can induce
expression of RB1 (ref. 1). Thus, RB1CC1 may be an upstream
regulator of RB1 expression. The gene RB1CC1 is located within
a region of chromosome 8q11 (ref. 2), and LOH for this region
has been associated with breast cancer3. About half of breast can-
cers show LOH in a DNA segment of about 1.5 Mb located
between D8S567 and D8S591 in this region (ref. 3). In a prelimi-
nary analysis, we identified LOH at this locus in 5 of 12 primary
breast cancers examined (Fig. 1). Together, these findings
strongly suggest that RB1CC1 is involved in the tumorigenesis of
breast cancer.
To detect RB1CC1 mutations in breast cancers, we analyzed
the complete coding region of RB1CC1 using cDNAs prepared
from 35 surgically resected primary breast cancers. We identified
nine mutations in seven of the cancers (20%; Table 1 and Fig. 2).
All nine mutations were deletions within exons 3–24 and were
predicted to yield truncated RB1CC1 proteins lacking the pat.4
nuclear localization signal, the leucine zipper DNA-binding
motif and the coiled-coil region—that is, the principal structural
motifs thought to be essential for RB1CC1 function1.
Two primary breast cancers (MMK3 and MMK6) contained
compound heterozygous deletions in both alleles, which were
predicted to yield highly truncated RB1CC1 proteins. MMK6
contained deletions of exons 3–24 (nt 534–5,322) and 9–23 (nt
1,757–5,187), which would result in frameshifts at codons 4 and
411, respectively (Fig. 3a,b). In MMK3, we detected deletions of
exons 3–24 (nt 535–5,324) and 5–11 (nt 849–2109); the former
would result in early termination at codon 4, and the latter would
cause a frameshift at codon 109 to yield a truncated protein of
122 amino acids.
These mutations were found to be somatic: PCR analysis of
genomic DNA from the tumor samples detected aberrant
products that corresponded to each deletion mutation, but the
Published online: 17 June 2002, doi:10.1038/ng911
Fig. 1 Loss of heterozygosity at the RB1CC1 locus (D8S567) in primary breast
cancers. Tumor DNA samples and selected matching genomic DNA samples
were amplified using primer pairs for the microsatellite marker. PCR products
were resolved using 8% denaturing PAGE and visualized by silver staining. T,
tumor; N, normal.
© 2002 Nature Publishing Group http://genetics.nature.com

Page 2

mutations were not seen in germline DNA (Table 1). We did
not detect RB1CC1 in either MMK3 or MMK6; we did not
detect RB1 in MMK6, and we detected significantly less RB1 in
MMK3 than in tumors without RB1CC1 mutations (Figs 3c
and 4). We did not detect LOH at the RB1 locus in either
MMK3 or MMK6 (Table 1). By contrast, both RB1CC1 and
RB1 were present in tumors without RB1CC1 mutations
(MMK12 and MMK29; Figs 3c and 4). These findings suggest
that inactivating mutations of RB1CC1 result in insufficient
expression of RB1, which promotes dysregulation of the RB1
pathway and subsequent tumorigenesis.
We identified intragenic RB1CC1 deletions that predicted
functionally null truncated
proteins in five other cancers
(MMK1, MMK15, MMK31,
MMK38 and MMK40; Table 1
and Fig. 2). Although these
mutations were all heterozy-
gous, we observed LOH at the
RB1CC1 locus (Fig. 1) and we
saw no expression of RB1CC1
in any of the cancers (Table 1),
indicating a functional loss of
both alleles. Expression of RB1
was markedly lower in these
cancers than in cancers with-
out mutations in RB1CC1 and
RB1 (MMK12 and MMK29;
Fig. 3c). We did not observe
LOH at the RB1 locus in any of
these five cancers (Table 1).
To determine whether muta-
tion of a single allele could
cause RB1CC1 inactivation
through a dominant-negative
mechanism, we introduced the
RB1CC1 mutations detected in
three samples
MMK38 and MMK40) into
HEK293T cells, which express
wildtype RB1CC1 abundantly.
No change in RB1CC1 and RB1
expression was detected (data
not shown), which indicates
that a dominant-negative
mechanism for inactivation of
RB1CC1 is unlikely.
Our results
homozygous inactivation of
RB1CC1 is probably involved
in the tumorigenesis of breast
cancer. Twenty percent of pri-
mary breast cancers examined
(MMK1,
show that
Fig. 2 Location and predicted outcome of RB1CC1 mutations. The genomic structure of RB1CC1 and its functional pro-
teins domains are shown. NLS, pat.4 nuclear localization signal; LZ, leucine zipper DNA-binding motif; wavy-lined box,
coiled-coil region. Exons are indicated by white boxes and introns by black lines. Deleted and preserved DNA regions are
indicated by dotted lines and gray boxes, respectively. Nine truncating mutations were detected in seven primary breast
tumors. Predicted protein products are indicated by hatched and vertical lines under each RB1CC1 mutation. Vertical
lines represent aberrant amino-acid residues caused by frameshift mutations. All mutant proteins are predicted to be
markedly truncated and to lack the principal structural motifs.
letter
286 nature genetics • volume 31 • july 2002
Table 1 • Mutations and sequence alterations of RB1CC1 in primary breast cancers
Sample
name
Nucleotide
changea
Location
(exon)
Predicted
effect
Germline
DNA
RB1CC1 status
RB1 status
LOHbprotein alleleprotein
Tyr/MMK3
Pro
c.11_4800del
c.325_1585del
3–24
5–11
Tyr/Y4fsX4
Pro/P109fsX122
wild type compound-heterozygous
deletions
——
↓
Tyr/MMK6
Asp
c.10_4798del
c.1233_4663del
3–24
9–23
Tyr/Y4fsX48
Asp/D411fsX431
wild type compound-heterozygous
deletions
———
Asn/MMK1c.957_4785del 7–24Asn/N319fsX368 wild typeLOH——
↓
—Ser/MMK15 c.1635_4719del 12–24Ser/S545fsX557wild typeLOH——
Ile/MMK31 c.212_4811del5–24Ile/I71fsX111wild type LOH———
↓
↓
Cys/MMK38c.241_4621del 5–22Cys/C81fsX99wild typeLOH——
Ser/MMK40
aThe A of the first ATG (the initiator methionine) of RB1CC1 (AB059622) is denoted as nucleotide +1. —, absent; ↓, expression significantly lower than in wild
type. bMicrosatellite markers AFM058xd6 and RB1.20 were used to analyze LOH of RB1 (L11910).
c.591_4678del7–23Ser/S197fsX212 wild type LOH ——
© 2002 Nature Publishing Group http://genetics.nature.com

Page 3

contained deletion mutations in RB1CC1 that were predicted to
produce markedly truncated, functionally null RB1CC1 pro-
teins. Two of these cancers contained compound heterozygous
deletions within RB1CC1, and the other five showed LOH of
RB1CC1. We analyzed all 35 tumor samples by western blotting
or immunohistochemistry and compared seven cases with
RB1CC1 mutations to the remaining 28 cases without RB1CC1
mutations. And RB1 expression was absent or less abundant in
all seven cancers (compared with those lacking RB1CC1 muta-
tions); these cancers showed no LOH at the RB1 locus. Together
with sequence and functional analyses suggesting that RB1CC1 is
an upstream regulator of RB1expression1,2, our findings indicate
that RB1CC1 may be a tumor-suppressor gene in breast cancer.
Fig. 3 Mutation analysis of RB1CC1.
a, RT–PCR products in MMK6 and
MMK29 tumor samples. A single
amplicon of 4.9 kb representing
wildtype RB1CC1
MMK29, which does not have a
mutation in RB1CC1. By contrast,
MMK6 contains amplicons of 98 bp
and 1,456 bp, representing com-
pound heterozygous
within RB1CC1. b, Sequence chro-
matogram of truncating RB1CC1
mutations in MMK6. Amplicons of 98
bp and 1,456 bp correspond to dele-
tions that encompass exons 3–24 and
exons 9–23, respectively. c, Western
blots probed with antibodies against
RB1CC1 and RB1 in MMK6, MMK40,
MMK38, MMK29
RB1CC1 and RB1 are barely detectable in samples with truncating mutations
of RB1CC1 (MMK6, MMK38 and MMK40), whereas their expression is pre-
served in those with wildtype RB1CC1 (MMK29 and MMK12).
is present in
deletions
and MMK12.
letter
nature genetics • volume 31 • july 2002
287
Fig. 4 Immunohistochemical analysis of RB1CC1 and RB1 in MMK3, 6 and 12.
Both RB1CC1 and RB1 proteins are expressed in the nuclei of breast cancer cells
in MMK12, which contains no RB1CC1 mutations. In contrast, they are barely
detectable in MMK 3 and 6, which contain RB1CC1 mutations in both alleles.
Protein localization is demonstrated as brown-stained dots, and nuclei are
counterstained with hematoxylin. RB1CC1 and RB1 immunostainings are indi-
cated left- and right-columns, respectively. Each scale bar shows 100 µm. Mag-
nification, ×200.
a
b
c
Methods
Primary breast cancer samples. Primary breast cancers were collected at
Shiga University of Medical Science (Shiga, Japan). When possible, we also
obtained matching blood samples from individuals who gave their consent
for use of their tumor samples for this study. The study was approved by
the Ethics Committee of the University. We flash-froze all tumors immedi-
ately after surgery.
Preparation of DNA, RNA and protein. We prepared serial sections on
slides from tumor samples for histological evaluation. We extracted DNA,
RNA and protein from samples that contained primarily tumor tissue. We
used DNAzol and TRIzol reagents (Gibco-BRL), respectively, according to
the manufacturer’s protocols to prepare DNA and total RNA. We synthe-
sized cDNA from total RNA using SuperScript II reverse transcriptase and
Oligo d(T)12–18primer (Gibco-BRL). We prepared genomic DNA from
leukocytes obtained from matching blood samples. We also prepared pro-
tein extracts in lysis buffer (0.1 mM Tris-HCl (pH 6.8), 4% SDS, 20% glyc-
erol and protease inhibitors) from several tumor samples.
LOH analyses of RB1CC1 and RB1. For RB1CC1, we amplified tumor
DNA and selected matching genomic DNA samples using primer pairs for
D8S567 and D8S591 (ref. 3). For RB1, we analyzed tumor DNA samples
with RB1CC1 mutations and matching genomic DNA samples using the
microsatellite markers AFM058xd6 (UniSTS: 13158) and RB1.20 (ref. 13).
We resolved PCR products using 8% urea-PAGE gels (GeneGel Clean;
Amersham) and visualized them by silver staining.
RB1CC1 mutation analysis. We screened cDNAs produced from each
sample to identify mutations in the RB1CC1 coding sequence. We ampli-
fied the complete RB1CC1 coding region using the primers CC1S2
(5′–TCAGAG GGTGAGGTATAAGC–3′) and CC1AS2 (5′–TGAGCACT-
GCAGGACAAATCA–3′). We purified and directly sequenced the amplifi-
cation products, including wildtype fragments of 4.9 kb and truncated
© 2002 Nature Publishing Group http://genetics.nature.com

Page 4

letter
288 nature genetics • volume 31 • july 2002
fragments resulting from mutated DNA, using additional primers
(sequences are available on request) and an ABI PRISM 310 genetic analyz-
er (Applied Biosystems). To confirm RB1CC1 mutations identified in
cDNA samples, we carried out analysis by PCR on relevant genomic DNA
using the ELONGASE system (Gibco-BRL) with minor modifications
(primer sequences are available on request). We carried out all PCR analy-
ses in duplicate and sequenced the PCR products directly to confirm their
identity. We analyzed genomic DNA from matching blood samples for
germline alterations.
Western blot analysis. We resolved equal amounts of protein extract in 5%
SDS–PAGE gels and blotted them to polyvinyl difluoride membranes using
standard techniques. We probed the blots with an antiserum against
human RB1CC1 (RBICC-642)1that recognizes amino acids 642–658 of
RB1CC1 protein. We re-probed the blots for RB1 using a mouse mono-
clonal antibody against RB1 (G3-245; PharMingen). All blots were devel-
oped using ECL reagents (Amersham).
Immunohistochemical analysis. To examine the expression of the
RB1CC1 and RB1 proteins, we used standard immunohistochemical stain-
ing with RBICC-642 and G3-245, respectively, on tissue sections prepared
from breast cancer samples. Sections from all seven tumors containing
RB1CC1 mutations and from several tumors with wildtype RB1CC1 were
cut from paraffin-embedded blocks.
GenBank accession number. RB1CC1 mRNA, AB059622.
Acknowledgments
We thank A. Mabuchi for computational searches and H. Honjo, H. Chen,
N. Takashima, M. Sugimoto, K. Kobayashi and J. Shimizu for experimental
assistance. This study was partially supported by grant-in-aids for Japan
Society for the Promotion of Science Fellows, Scientific Research (B), The
Ministry of Education, Science, Sports and Culture of Japan, and a grant
from the Japan Orthopaedics and Traumatology Foundation.
Competing interests statement
The authors declare that they have no competing financial interests.
Received 18 December 2001; accepted 9 May 2002.
1.Chano, T. et al. Identification of RB1CC1, a novel gene that can induce RB1 in
various human cells. Oncogene 21, 1295–1298 (2002).
Chano, T. et al. Isolation, characterization and mapping of the mouse and human
RB1CC1 genes. Gene (in press).
Dahiya, R., Perinchery, G., Deng, G. & Lee, C. Multiple sites of loss of
heterozygosity on chromosome 8 in human breast cancer has differential
correlation with clinical parameters. Int. J. Oncol. 12, 811–816 (1998).
Perinchery, G. et al. Loss of two new loci on chromosome 8 (8p23 and 8q12-13) in
human prostate cancer. Int. J. Oncol. 14, 495–500 (1999).
Kaelin, W.G. Jr. Functions of the retinoblastoma protein. BioEssays 21, 950–958
(1999).
Kamb, A. Cell cycle regulators and cancers. Trends Genet. 11, 136–140 (1995).
Taya, Y. RB kinases and RB-binding proteins: new points of view. Trends Biol. Sci.
22, 14–17 (1997).
T’Ang, A., Varley, J.M., Chakraborty, S., Murphree, A.L. & Fung, Y.K. Structural
rearrangement of the human retinoblastoma gene in human breast carcinoma.
Science 242, 263–266 (1988).
Bieche, I. & Lidereau, R. Loss of heterozygosity at 13q14 correlate with RB1 gene
underexpression in human breast cancer. Mol. Carcinog. 29, 151–158 (2000).
10. Thorlacious, S., Jonasdottir, O. & Eyfjord, J.E. Loss of heterozygosity at selective
sites on chromosomes 13 and 17 in human breast carcinoma. Anticancer Res. 11,
1501–1507 (1991).
11. Lemoine, N.R. Molecular biology of breast cancer. Ann. Oncol. 5 (Suppl. 4), 31–37
(1994).
12. Varley, J.M. et al. The retinoblastoma gene is frequently altered leading to loss of
expression in primary breast tumors. Oncogene 4, 725–729 (1989).
13. Wadayama, B. et al. The spectrum of the retinoblastoma gene in osteosarcomas.
Cancer Res. 54, 3042–3048 (1994).
2.
3.
4.
5.
6.
7.
8.
9.
© 2002 Nature Publishing Group http://genetics.nature.com