BRAF mutations characterize colon but not gastric cancer with mismatch repair deficiency

Abstract

Genes from the RAF family are Ras-regulated kinases involved in growth cellular responses. Recently, a V599E hotspot mutation within the BRAF gene was reported in a high percentage of colorectal tumors and significantly associated to defective mismatch repair (MMR). Additionally, BRAF mutations were described only in K-Ras-negative colon carcinomas, suggesting that BRAF/K-Ras activating mutations might be alternative genetic events in colon cancer. We have addressed to what extent the tumorigenic-positive selection exerted by BRAF mutations seen in colorectal MMR-deficient tumors was also involved in the tumorigenesis of gastric cancer. Accordingly, BRAF mutations were detected in 34% (25/74) of colorectal MMR-deficient tumors and in 5% (7/142) of MMR-proficient colorectal cases (P=0.0001). All mutations found in the MSI cases corresponded to the previously reported hotspot V599E. Two D593K and a K600E additional mutations were also detected in three MSS cases. However, only one mutation of BRAF was found within 124 MSS gastric tumors and none in 37 MSI gastric tumors, clearly suggesting that BRAF mutations are not involved in gastric tumorigenesis. Nonetheless, a high incidence of mutations of K-Ras was found within the MSI gastric group of tumors (P=0.0005), suggesting that the activation of K-Ras-dependent pathways contributes to the tumorigenesis of gastric cancers with MMR deficiency. Accordingly, our results show evidences that BRAF mutations characterize colon but not gastric tumors with MMR deficiency and are not involved in the tumorigenesis of gastric cancer of the mutator phenotype pathway.

Full text

BRAF mutations characterize colon but not gastric cancer with mismatch
repair deficiency
Carla Oliveira
1
, Mafalda Pinto
1
, Alex Duval
2
, Caroline Brennetot
2
, Enric Domingo
3
, Eloi Espı
´n
3
,
Manel Armengol
3
, Hiroyuki Yamamoto
4
, Richard Hamelin
2
, Raquel Seruca
1
and
Simo
´Schwartz Jr*
,3
1
Instituto de Patologia e Imunologia Molecular da Universidade do Porto (IPATIMUP), Porto 4200-465, Portugal;
2
INSERM U434
CEPH, Paris 75010, France;
3
Molecular Pathology Program, Centre d’Investigacions en Bioquı
´
mica i Biologia Molecular
(CIBBIM), Passeig Vall d’Hebron 119-129, Barcelona 08035, Spain;
4
First Department of Internal Medicine, Sapporo Medical
University, S.1, W.16, Chuo-ku, Sapporo 060-8543, Japan
Genes from the RAF family are Ras-regulated kinases
involved in growth cellular responses. Recently, a V599E
hotspot mutation within the BRAF gene was reported in a
high percentage of colorectal tumors and significantly
associated to defective mismatch repair (MMR). Addi-
tionally, BRAF mutations were described only in K-Ras-
negative colon carcinomas, suggesting that BRAF/K-Ras
activating mutations might be alternative genetic events in
colon cancer. We have addressed to what extent the
tumorigenic-positive selection exerted by BRAF muta-
tions seen in colorectal MMR-deficient tumors was also
involved in the tumorigenesis of gastric cancer. Accord-
ingly, BRAF mutations were detected in 34% (25/74) of
colorectal MMR-deficient tumors and in 5% (7/142) of
MMR-proficient colorectal cases (P¼0.0001). All muta-
tions found in the MSI cases corresponded to the
previously reported hotspot V599E. Two D593K and a
K600E additional mutations were also detected in three
MSS cases. However, only one mutation of BRAF was
found within 124 MSS gastric tumors and none in 37 MSI
gastric tumors, clearly suggesting that BRAF mutations
are not involved in gastric tumorigenesis. Nonetheless, a
high incidence of mutations of K-Ras was found within the
MSI gastric group of tumors (P¼0.0005), suggesting
that the activation of K-Ras-dependent pathways con-
tributes to the tumorigenesis of gastric cancers with
MMR deficiency. Accordingly, our results show evidences
that BRAF mutations characterize colon but not gastric
tumors with MMR deficiency and are not involved in the
tumorigenesis of gastric cancer of the mutator phenotype
pathway.
Oncogene (2003) 22, 9192–9196. doi:10.1038/sj.onc.1207061
Keywords: genomic instability; BRAF; K-Ras; DNA
mismatch repair; mutator phenotype; gastrointestinal
cancer
Introduction
Approximately 15% of sporadic colorectal and gastric
tumors show microsatellite instability due to defects in
their DNA mismatch repair (MMR) system. In both
tumor types, this molecular phenotype is associated with
a particular clinicopathological behavior, and charac-
terized by an underlying genomic instability and a
specific profile of target gene mutations (Aaltonen et al.,
1993; Ionov et al., 1993; Thibodeau et al., 1993;
Perucho, 1996). Moreover, several evidences have
shown that stomach and colon tumors with microsa-
tellite instability share the same target genes of the
mutator phenotype, with similar mutational incidences
(Yamamoto et al., 1997, 1999; Schwartz et al., 1999;
Duval and Hamelin, 2002). Recently, a high incidence of
activating mutations in BRAF, a gene from the Ras-
regulated kinase-encoding RAF family, has been found
in colorectal tumors and associated to MMR-deficient
cases (Davies et al., 2002; Rajagopalan et al., 2002;
Yuen et al., 2002). Further, BRAF data showed a
mutational hotspot in nucleotide 1796 within exon 15,
accounting for a T:A transversion mutation and a valine
to glutamic acid substitution, which is the most frequent
somatic substitution ever identified in MMR-deficient
colon cancers. In addition, these mutations were
inversely associated to K-Ras mutations, reinforcing
the idea that colorectal tumors with defective MMR
progress through the same Ras/Raf/MAPK pathway
than MMR-proficient tumors (Rajagopalan et al.,
2002). However, it is not clear if colon and gastric
tumors of the mutator phenotype share the alterations
of the Ras/RAF/MAPK genes found in MMR-deficient
colon tumors, nor if K-Ras and BRAF genes are also
alternative genetic events in gastric cancer.
Here, we show evidences to support that K-Ras but
not BRAF mutations contribute to the tumorigenesis of
MMR-deficient gastric cancer and therefore that K-Ras
and BRAF mutations are not alternative genetic events
in gastric cancer. Further, we also conclude that
although MMR-deficient gastrointestinal tumors
share the same mutational profile of the target genes
of the mutator phenotype, they differ concerning
Received 15 April 2003; revised 30 July 2003; accepted 31 July 2003
*Correspondence: S Schwartz; E-mail: sschwartz@vhebron.net
Oncogene (2003) 22, 9192–9196
&
2003 Nature Publishing Group
All rights reserved 0950-9232/03 $25.00
www.nature.com/onc
ONCOGENOMICS

oncogenic mutations in genes from the Ras/RAF/
MAPK pathway.
Results and discussion
It has been recently shown that the V599E mutation of
BRAF, the most common BRAF mutation found
within colorectal tumors, has a transforming and
oncogenic activity in NIH3T3 cells 138 times over
wild-type BRAF (Davies et al., 2002). The presence of
BRAF mutations in a high percentage of colorectal
MMR-deficient carcinomas suggests its tumorigenic
involvement within these tumors, and also its capability
to induce tumor cell clonal expansions. Further, the
absence of concomitant K-Ras and BRAF mutations in
these tumors has suggested that both are alternative
genetic events in colorectal tumorigenesis, and also that
alterations within the Ras/RAF/MAPK pathway char-
acterize MMR-proficient and -deficient colorectal tu-
mors accordingly (Rajagopalan et al., 2002).
In agreement with these findings, BRAF and K-Ras
mutations were, respectively, detected in 34% (25/74)
and 18% (11/60) of tumors from our collection of MSI
colorectal cases (Figures 1 and 2). Further, all BRAF
mutations detected corresponded to the V599E hotspot
substitution reported by Davies et al. (2002) and were
clearly associated to the presence of MMR deficiency
(P¼0.0001). In fact, only 5% (7/142) of colorectal MSS
tumors showed BRAF mutations, from which four
(57%) were V599E, one K600E and two D593K
(Figures 1 and 2). As expected, the analysis of K-Ras
showed higher mutational frequencies in colorectal MSS
Figure 1 Mutation analysis of BRAF and K-Ras in colon and gastric carcinoma cases. SSCP/HA from BRAF are shown. DGGE
analysis of K-Ras is also shown. N, normal tissue counterpart; T, tumor. BRAF mutations and wild-type sequence (normal) are
indicated on top and abnormal bands pointed by arrows. (a) A representative SSCP/HA analysis of V599E BRAF mutations within
MSI colorectal tumors. The corresponding mutated sequence is also shown at the right side. (b) A representative SSCP/HA analysis of
BRAF mutations in MSS colorectal tumors. Two additional D593K and one K600E mutations were detected in the HA analysis and
also by automatic sequencing (right). (c) SSCP/HA analysis of BRAF in MSS (left panel) and MSI (right panel) gastric carcinomas.
Sequencing analysis of the single V599E BRAF mutation found in an MSS gastric tumor is also shown (right side). (d) Representative
DGGE analysis of K-Ras in MSI gastric tumors. HAs corresponding to G12D and G13D mutations were found in several MSI tumor
cases (arrows). A normal case is shown on the left. Two G12V mutations were also detected by automatic sequencing. Sequences are
also shown. Additional V599E BRAF mutations were also detected in our collection of gastrointestinal tumor cell lines, including three
MSI colon cell lines (Co115, RKO and LS411), two MSS colon cell lines (WIDR and HT29) and one MSI gastric cell line (St2957). No
mutations were detected in other cell lines from the colon, including nine MSI (LS174 T, LoVo, TC71, HCT15, HCT116, TC7, SW48,
HCT8 and KM12) and 19 MSS (GLY, EB, Isreco1, LS513, CBS, FET, V9P, ALA, FRI, LS1034, Isreco2, Colo320, Isreco3, SW480,
SW1116, Colo205, SW620, CaCo2 and T84), neither in the MSI gastric cell line SNU1 nor in the MSS gastric cell lines GTL16,
SNU16, KatoIII, MKN28, N87, TMK1, MKN1, MKN74, HGT1, AGS, MKN45, GP220, GP202 and L195
BRAF and K-ras mutations within the mutator phenotype model
C Oliveira et al
9193
Oncogene

(38%, 28/73) tumors than in MSI (18%, 11/60) color-
ectal cases (P¼0.01) (Figure 2). Mutations in both
genes were also detected in a variety of colorectal MSS
and MSI tumor cell lines (data not shown – see legend
of Figure 1). In the series of MSI (n¼60) and MSS
(n¼73) colon tumors that were analysed for both genes,
50% of MSI and 41% of MSS cases show activating
mutations of K-Ras or V599E BRAF mutations or both
(Figure 3). These results clearly link the activation of the
Ras/RAF/MAPK pathway in colorectal tumorigenesis
to the oncogenic activation of K-Ras or BRAF genes
(Figure 3).
Even though K-Ras and BRAF mutations have been
described as alternative events in these tumors (Raja-
gopalan et al., 2002), we found two concomitant
mutations in two Japanese cases of our collection
(Figure 3). These results might suggest either a possible
synergistic contribution of both genes to the tumorigenic
clonal expansion of these cases or a heterogeneous
clonal population of cells within these tumors. Further
analysis also revealed that the colorectal Japanese cases
(n¼22) from our collection showed higher levels of
BRAF mutations (11/22, 50%) than tumors from
European origin (14/52, 27%) (P¼0.05), even though
similar levels of K-Ras mutations were detected in both
groups (data not shown). These results might suggest a
modulation of the tumorigenic contribution from the
Ras/RAF/MAPK pathway activation by the genetic
ethnical background.
Further, because stomach tumors of the mutator
phenotype share similar target genes than colorectal
tumors (Yamamoto et al., 1997, 1999; Schwartz et al.,
1999; Duval and Hamelin, 2002), we did investigate
whether BRAF mutations are also involved in gastric
MMR-deficient cancer. However, no significant muta-
tions were found in our collection of stomach primary
tumors, including 37 MSI gastric primary tumors
(Figures 1 and 2). In fact, we only detected a V599E
mutation in an MSS tumor (1/124, 0.8%) within our
series. These data led us to the conclusion that BRAF
mutations are not involved in gastric tumorigenesis.
However, as we have recently reported (Brennetot et al.,
in press), a high incidence of K-Ras mutations was
found in tumors from the MSI group (10/36, 28%) of
our gastric tumor collection whereas no mutations were
detected in any of the MSS cases (P¼0.0005), clearly
suggesting that the activation of K-Ras contributes to
the tumorigenesis of gastric cancer of the mutator
phenotype, but not to MSS gastric cancer (Figures 2 and
3). Further, these data also support that K-Ras and
BRAF mutations are not alternative events in gastric
cancer.
Further, although MMR-deficient colon and gastric
tumors show a similar spectrum of mutations regarding
the typical target genes of the mutator phenotype,
they differ in the oncogenic mutational incidences
found in genes from the Ras/RAF/MAPK pathway
(Figure 4), raising the question of whether alternative
K-Ras-depending pathways, other than the classical
Figure 2 Comparative analysis of the mutation frequencies of
BRAF and K-Ras in MSS and MSI colon and gastric tumors. The
upper part of the figure represents the mutation frequency of
BRAF mutations in colon (light bars) and gastric (dark bars) in
MSS and MSI tumors. The lower part of the figure represents the
mutation frequency of K-Ras mutations in colon (light bars) and
gastric (dark bars) in MSS and MSI tumors
Figure 3 Comparative analysis of the percentage of cases harboring activating K-Ras and/or V599E BRAF mutations in colon and
gastric MSS and MSI tumors. The upper bar represents the percentage of MSS and MSI colon carcinoma cases harboring activating
K-Ras (light gray) and V599E BRAF (dark gray) or both (black). The lower bar represents the percentage of MSS gastric carcinoma
cases harboring the V599E BRAF (dark gray) and the percentage of MSI gastric carcinoma cases harboring activating K-Ras (light
gray). Cases without activating K-Ras and/or V599E BRAF, including K600E and D593K BRAF, are represented in white
BRAF and K-ras mutations within the mutator phenotype model
C Oliveira et al
9194
Oncogene

Ras/RAF/MAPK, might also be activated during
gastric MMR-deficient tumorigenesis. Future research
will have to answer to this possibility.
Materials and methods
Tissue samples and microsatellite instability analysis
Tumors were obtained from the Hospital of S. Joa
˜o (Porto,
Portugal), the Saint-Antoine Hospital (Paris, France), the
Sapporo Medical University (Sapporo, Japan) and from the
Centre d’Investigacions en Bioquimica i Biologia Molecular
Vall d’Hebron (CIBBIM) (Barcelona, Spain). Sample collec-
tion was carried out in accordance with previously established
ethical protocols. Collected tumors were immediately frozen in
liquid nitrogen for further analysis. Human cancer cell lines
were obtained from different sources, including the American
Type Culture Collection and the Japanese Cancer Research
Resources Bank in Osaka, Japan. Gastric cell lines GP202,
GP220 and L195 were established at the IPATIMUP. Some of
the cell lines were previously characterized for a number of
genetic alterations (Gayet et al., 2001). A family history was
obtained for every carcinoma case. None of the patients
included in the present study had a family history suggestive of
hereditary nonpolyposis colorectal cancer. Hematoxylin- and
eosin-stained sections were used to classify the tumors and
allowed their macrodissection. High molecular weight DNA
was isolated using standard methods from total sections of the
tumors whenever tumor cells occupied more than 50% of
tumor tissue or from macrodissected areas with at least 50% of
tumor cells.
Our collection of tumors and cell lines was analysed for
microsatellite instability, using the mononucleotide repeats
BAT-26 and BAT-25, and also a panel of dinucleotide repeat
sequences, as previously described (Hoang et al., 1997; Boland
et al., 1998; Oliveira et al., 1998; Yamamoto et al., 2001). We
gather for this study 124 tumors and cell lines with MMR
deficiency (74 sporadic colon carcinomas, including 22 cases
from Japan, and also 37 sporadic gastric carcinomas, 12 colon
cancer cell lines and one gastric cancer cell line), and 302
MMR-proficient tumors and cell lines (142 sporadic colon
carcinomas, 124 sporadic gastric carcinomas, 21 colon cancer
cell lines and 15 gastric cancer cell lines). All tumors and cell
lines classified as MSI showed additional mutations in several
target genes of the mutator phenotype, including hMSH3,
hMSH6, BAX and TGFbRII (not shown).
BRAF and K-Ras mutation screening
Mutational analysis of BRAF was performed by single-
stranded conformation polymorphism and heteroduplex ana-
lysis (SSCP/HA). The fragment encompassing exon 15 was
amplified by PCR in 376 carcinoma samples and 48 cell lines.
Primer sequences and PCR conditions were based on those
reported previously (Davies et al., 2002). Genomic DNA (25–
100 ng) was amplified by PCR using the following cycling
conditions: 30 s at 941C, 30 s at 601C and 45 s at 721C for 35
cycles. Reaction products were diluted with denaturing buffer
(formamide with 0.025% xylene cyanol and 0.025% bromo-
phenol blue) and heated up to 991C for 10 min before being
loaded onto 0.8 mutation detection enhancement gels
(MDE – Flowgen, Rockland, ME, USA), run at 81C for 12–
18 h and stained with silver nitrate. Selected bands were
recovered from the gels and submitted to PCR reamplification
with the original primer sets. Reamplified products were
purified and sequenced on an ABI Prism 377 automatic
sequencer (Perkin-Elmer, Foster City, CA, USA) using the
ABI Prism Dye Terminator Cycle Sequencing Kit (Perkin-
Elmer). All detected mutations were confirmed in a second
independent PCR. K-Ras mutations were screened in 133
colorectal tumors (73 MSS and 60 MSI) and in 77 gastric
carcinomas (42 MSS and 35 MSI). Mutational analysis of the
K-Ras gene was performed by denaturing gradient gel
electrophoresis (DGGE) as described (Gayet et al., 2001).
Acknowledgements
This work was supported by Grant FIS 01/1350 from the
Spanish Fondo de Investigaciones Sanitarias and Fundac¸a
˜o
para a Cieˆ ncia e a Tecnologia, Portugal (Project: POCTI/
35374/CBO/2000 and POCTI/CBO/40820/2001).
References
Aaltonen LA, Peltomaki P, Leach FS, Sistonen P, Pylkkanen
L, Mecklin JP, Jarvinen H, Powell SM, Jen J, Hamilton SR,
Petersen GM, Kinzler KW, Vogelstein B and de la Chapelle
A. (1993). Science,260, 812–816.
Boland CR, Thibodeau SN, Hamilton SR, Sidransky D,
Eshleman JR, Burt RW, Meltzer SJ, Rodrigues-Bigas MA,
Fodde R, Ranzani GN and Srivastava S. (1998). Cancer
Res.,58, 5248–5257.
Brennetot C, Pinto M, Oliveira C, Schwartz Jr S, Seruca R,
Duval A and Hamelin R: Gastroenterology. (in press).
Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg S,
Teague J, Woffendin H, Garnett MJ, Bottomley W, Davis
N, Dicks E, Ewing R, Floyd Y, Gray K, Hall S, Hawes R,
Hughes J, Kosmidou V, Menzies A, Mould C, Parker A,
Stevens C, Watt S, Hooper S, Wilson R, Jayatilake H,
Gusterson BA, Cooper C, Shipley J, Hargrave D, Pritchard-
Jones K, Maitland N, Chenevix-Trench G, Riggins GJ,
Bigner DD, Palmieri G, Cossu A, Flanagan A, Nicholson A,
Ho JWC, Leung SY, Yuen ST, Weber BL, Seigler HF,
Darrow TL, Paterson H, Marais R, Marshall CJ, Wooster
Figure 4 Mutation frequencies observed at BRAF, K-Ras and at
coding repeat sequences in MSI sporadic colorectal (dark gray) and
MSI gastric tumors (light gray). *Significant differences found
between MSI colon and gastric tumors
BRAF and K-ras mutations within the mutator phenotype model
C Oliveira et al
9195
Oncogene

R, Stratton MR and Futreal PA. (2002). Nature,417, 949–
954.
Duval A and Hamelin R. (2002). Cancer Res.,62, 2447–
2454.
Gayet J, Zhou XP, Duval A, Rolland S, Hoang JM, Cottu P
and Hamelin R. (2001). Oncogene,20, 5025–5032.
Hoang JM, Cottu PH, Thuille B, Salmon RJ, Thomas G and
Hamelin R. (1997). Cancer Res.,57, 300–303.
Ionov Y, Peinado MA, Malkhosyan S, Shibata D and Perucho
M. (1993). Nature,363, 558–561.
Oliveira C, Seruca R, Seixas M and Sobrinho-Simo
˜es M.
(1998). Am. J. Pathol.,153, 1211–1219.
Perucho M. (1996). J. Biol. Chem.,377, 675–684.
Rajagopalan H, Bardelli A, Lengauer C, Kinzler KW,
Vogelstein B and Velculescu VE. (2002). Nature,418,
934.
Schwartz Jr S, Yamamoto H, Navarro M, Maestro M,
Revento
´s J and Perucho M. (1999). Cancer Res.,59, 2995–
3002.
Thibodeau SN, Bren G and Schaid D. (1993). Science,260,
816–819.
Yamamoto H, Fukushima H, Itoh F, Horiuchi S, Min Y, Iku
S and Imai K. (2001). J. Exp. Clin. Cancer Res.,20, 553–559.
Yamamoto H, Perez-Piteira J, Yoshida T, Terada M, Itoh F,
Imai K and Perucho M. (1999). Gastroenterology,116, 1348–
1357.
Yamamoto H, Sawai H and Perucho M. (1997). Cancer Res.,
57, 4420–4426.
Yuen ST, Davies H, Chan TL, Ho JW, Bignell GR, Cox C,
Stephens P, Edkins S, Tsui WW, Chan AS, Futreal PA,
Stratton MR, Wooster R and Leung SY. (2002). Cancer
Res.,62, 6451–6455.
BRAF and K-ras mutations within the mutator phenotype model
C Oliveira et al
9196
Oncogene