Rare Mutations in the PIK3CA Gene Contribute

to Aggressive Endometrial Cancer

Darina Konstantinova,
1,2
Radka Kaneva,
1,2
Roumen Dimitrov,
3
Alexey Savov,
4
Stephan Ivanov,
5
Tzvetanka Dyankova,
5
Ivo Kremensky,
2,4
and Vanio Mitev
1,2
The molecular basis of endometrial cancer (EC), a common gynecologic malignancy, often includes mutational
activation of the PIK3CA and KRAS genes. We aimed to determine the distribution of mutations in the two genes
depending on patient clinocopathological characteristics. We sequenced exon 1 of the KRAS gene and exons 9
and 20 of the PIK3CA gene in 108 consecutive EC tumor samples. PIK3CA mutations were present in 24 of the
108 (22.2%) cases and KRAS mutations in 18 of the 108 (16.7%) cases. PIK3CA mutations occurred more fre-
quently in KRAS-mutated samples (7=18, 38.9%; p 0.06) than in KRAS wild type (17=90, 18.9%) and showed a
very high frequency in metastatic tumors (4=9, 44.4%; p 0.1) and in samples displaying serous differentiation
serous and mixed endometrioid=serous tumors (6=12, 50.0%; p 0.021)where KRAS mutations were rare
(11.1% and 16.7%, respectively) and did not exist independently of a PIK3CA mutation. Non-hotspot (i.e., non-
E542K, -E545K, and -H1047R) mutations in the PIK3CA gene showed higher frequency in metastatic cases (3=9,
33.3%; p 0.05). Tumors displaying serous differentiation showed a particular patternthey harbored exclu-
sively mutations in PIK3CA exon 20 (5=12, 41.7%; p 0.005) and most of these were non-hotspot (4=12, 33.3%;
p 0.02). In all other comparisons exons 9 and 20 mutation distribution did not differ. These results suggest the
need for further exploration of the signicance of PIK3CA mutations in respect to aggressive EC.
Introduction
T
he phosphatidylinositol 3-kinases (PI3Ks) are a
family of lipid kinases that generate intracellular phos-
phatidylinositol 3,4,5-triphosphate by phosphorylating its
precursor, phosphatidylinositol 4,5-bisphosphate, upon ac-
tivation. PI3Ks are important regulators of cellular growth
and proliferation, transformation, adhesion, apoptosis, sur-
vival, and motility (Cantley, 2002). PI3Ks are composed of
catalytic and regulatory subunits encoded by separate genes.
The PIK3CA gene encodes the catalytic subunit (p110 a) of
class I A PI3Ks, which are involved in cellular signaling
from receptor tyrosine kinases upon growth factor binding.
An upstream effector of the PI3K is RAS. Mutations in
the PIK3CA and KRAS genes are known to activate the
corresponding proteins and lead to downstream cascade
signaling.
Gain-of-function somatic mutations in the PIK3CA gene
have been reported in a wide variety of human malignancies
including endometrial cancer (EC) (Hayes et al., 2006; Cata-
sus et al., 2008)one of the most common gynecologic ma-
lignancies in the industrialized world. Mutations cluster to
two hotspot regionsthe helical (exon 9) and kinase (exon
20) domains of the protein (Samuels et al., 2004).
The clinicopathological signicance of PIK3CA mutations
in EC remains largely unexplored. One study has reported
that mutations in exon 20 have been associated with adverse
prognostic factors in EC (Catasus et al., 2008).
We have sequenced the mutational hotspot regions of the
PIK3CA gene (exons 9 and 20) in a series of 108 EC tumor
samples and have related presence and type of mutations to
patients clinicopathological characteristics. We have also
examined their coincidence with mutations in the KRAS
gene.
Materials and Methods
Sampling
A cohort of 108 unselected patients with EC were inves-
tigated for mutations in the PIK3CA gene (exons 9 and 20).
Patients were treated at two institutionsthe University
Hospital of Obstetrics and Gynecology Maichin Dom and
the Clinic of Oncogynecology at the National Centre of
Oncology. Tissue retrieval was done in compliance with the
1
Department of Chemistry and Biochemistry and
2
Molecular Medicine Center, Medical UniversitySoa, Soa, Bulgaria.
3
Clinic of Operative Gynecology and
4
National Genetics Laboratory, Maichin Dom University Hospital of Obstetrics and Gynecology,
Soa, Bulgaria.
5
Clinic of Oncogynecology, National Centre of Oncology, Soa, Bulgaria.
DNA AND CELL BIOLOGY
Volume 29, Number 2, 2010
Mary Ann Liebert, Inc.
Pp. 6570
DOI: 10.1089=dna.2009.0939
65
Ethics Committee of the Medical UniversitySoa. Histolo-
gical diagnosis was performed by the hospitals pathology
units. Grade and stage were determined according to the
International Federation of Gynecology and Obstetrics
(FIGO) guidelines. Depth of myometrial inltration was
separately classied into less than 1=3, less than 2=3, or more
than 2=3 of the thickness of the wall. Tumors containing both
endometrioid and nonendometrioid components were clas-
sied as mixed whenever the second component constituted
more than 10% of the tumor volume.
Availability for analysis specimens included fresh, frozen
tissue samples (108 cases) and ve formalin-xed parafn-
embedded noncancerous tissue samples, which corresponded
to tumor samples having shown a previously unknown
mutation.
DNA isolation was performed following standard proce-
dures. Tissue sections of fresh, frozen tumor samples were
digested overnight with proteinase K, and then DNA was
phenolchlorophorm extracted. Formalin-xed parafn-
embedded tissue samples were initially deparafnated by
two to ve xylol washes, and subsequently DNA was ex-
tracted following the same procedure.
The cases were also subjected to KRAS mutation analysis,
which had identied mutations in exon 1 of the gene con-
taining hotspot codons 12 and 13.
Mutation analysis
PCR amplication of exons 9 and 20 of the PIK3CA gene
was carried out using primers PIK3CA20 F: 5
0
-CATTTGCTC
CAAACTGACCA-3
0
and PIK3CA20 R: 5
0
-CCTATGCAAT
CGGTCTTTGC-3
0
; PIK3CA9 F:5
0
-TTGCTTTTTCTGTAAAT
CATCTGTG-3
0
and PIK3CA9 R: 5
0
-CTGCTTTATTTATTCC
AATAGGTATG-3
0
. Cycling conditions were 948C for 3 min,
followed by 32 cycles of denaturation at 948C for 30 s, an-
nealing for 30 s, elongation at 728C for 60 s, and a nal
elongation step at 728C for 5 min. Annealing temperature
was 638C and 618C for exons 20 and 9, respectively.
PCR products were puried using exonuclease I=shrimp
alkaline phosphatase and subjected to direct sequencing
using ABI PRISM BigDye Terminator V3.1 Cycle Sequencing
Kit and an ABI PRISM 3130xl Genetic Analyzer (Applied
Biosystems, Foster City, CA).
KRAS mutational status was analyzed using direct cycle
sequencing as described. Briey, the primers used were as
follows: F: 5
0
-GGTACTGGTGGAGTATTTGATAGT-3
0
and
R: 5
0
-CCTTTATCTGTATCAAAGAATG-3
0
. Cycling condi-
tions were the same with the exception that annealing was
done at 588C.
Statistical methods
Distribution of mutation frequencies among groups was
compared using Fishers exact test. A p-value of 0.05 was
considered statistically signicant. All p-values were two
sided.
Results
Of the 108 endometrial tumors, 90 were endometrioid and
8 were nonendometrioid (5 serous and 3 clear cell). Ten
tumors were classied as mixedseven of them contained
endometrioid and serous components and three contained
endometrioid and clear cell components. Nine patients
had metastasis, which included three patients with regional
lymph node spread and six patients with distant spread.
We identied PIK3CA mutations in 22.2% (24 of 108) of
the samples16 were in exon 20 and 10 were in exon 9
(Table 1). All were missence point mutations and revealed
the mutated and wild-type allele.
We did not detect an association between presence of
PIK3CA mutations and patients grade, stage, or depth of
myometrial inltration (Table 2).
PIK3CA mutation frequency varied signicantly depend-
ing on tumor histological subtype. One of the ve purely
serous cancers (20%) and ve of the seven (71.4%) mixed
endometrioid=serous cancers harbored a mutation, which
brought the frequency of PIK3CA mutations among tumors
having components with serous differentiation (50.0%) to
statistical signicance ( p 0.021). Particularly, involvement
of mutations in exon 20 of the gene (41.7%, p 0.005) was
typical for these tumors. None of the six tumors having clear
cell components contained a mutation.
Except for histology type, we were not able to make an as-
sociation of the aforementioned parameters to the location of
mutations in either exon 9 or in exon 20 of the gene (Table 2).
The three most common PIK3CA mutations in human
cancerE542K, E545K, and H1047Rrepresented 50.0% of
all the mutations we found. We classied all other mutations
as rare (Table 1). Among them, three amino acid changes
p.P539T (c.1615C>A), L989I (c.2964C>A), and T1025I
(c.3074C>T)and two nucleotide changesc.3129G>A and
g.70434C>Ahave not been reported previously to our
knowledge (Catalogue of Somatic Mutations in Cancer).
Their somatic nature was conrmed by sequencing a non-
cancerous patient sample. The mutation at L989 was dis-
covered in a low-grade, stage I, purely serous carcinoma. The
Table 1. Mutations in the PIK3CA and KRAS Genes
Nucleotide change Codon change Domain No. of cases
PIK3CA
c.1615C>A p.P539T
a
Helical 1
c.1616C>G p.P539R Helical 1
c.1624G>A p.E542K Helical 3
c.1633G>A p.E545K Helical 3
c.1636C>G p.Q546E Helical 1
g.70434C>A
a
n=a Helical 1
c.2964C>A p.L989I
a
Kinase 1
c.3073A>G p.T1025A Kinase 2
c.3074C>T p.T1025I
a
Kinase 1
c.3129G>A
a
p.M1043I Kinase 1
c.3140A>G p.H1047R Kinase 7
c.3139C>T p.H1047Y Kinase 1
c.3145G>C p.G1049R Kinase 1
c.3145G>A p.G1049S Kinase 1
c.3146G>C p.G1049A Kinase 1
KRAS
c.35G>C p.G12A 3
c.35G>A p.G12D 6
c.35G>T p.G12V 2
c.37G>T p.G13C 1
c.38G>A p.G13D 4
g.6253T>G
a
n=a 1
a
Previously undescribed substitutions.
n/a, not applicable.
66 KONSTANTINOVA ET AL.
intron-situated g.70434C>A coexisted with p.P539T in a low-
grade, stage I endometrioid tumor.
PIK=Akt pathway activation is known to be involved in
cancer metastatic spread. Consistently, PIK3CA mutations
were more common in advance staged tumors (30.8%,
p 0.48) and in tumors with metastatic spread (44.4%,
p 0.1), which further supports previous reports of their
involvement in EC invasion. We found that the high PIK3CA
mutation frequency in metastatic cases was largely not due
to the group of hotspot mutations as only one (11.1%,
p 1.000) of these cases harbored such a mutation (Table 2).
Thus, the group of rare mutations in the gene showed a
borderline statistically signicant overrepresentation in
metastatic EC (33.3%, p 0.047). Among tumors with me-
tastasis we found the E542K, G1049R, M1043I (c.3129G>A),
and the T1025I coexisting with P539R.
In tumors with serous components, non-hotspot muta-
tions also showed a much higher frequency (33.3%, p 0.02).
Among them we found Q546E, and P539R coexistent with
T1025I, H1047R, L989I, and G1049S.
Mutations in the KRAS gene were present in 18 (16.7%) of
the tumors (Table 1) including one new somatic intron mu-
tation, g.6253T>G, which was found in a highgrade, stage I
endometrioid carcinoma. KRAS mutations were not found in
any of the eight nonendometrioid samples. They often co-
existed with a PIK3CA mutation in seven cases (38.9% of all
the KRAS mutations, p 0.06).
Mutated KRAS has been shown to activate several
downstream pathways, one of which is the PIK=Akt; how-
ever, KRAS mutations did not show the distribution pattern
we found for PIK3CA mutations. KRAS mutations showed
higher frequency in grade 1 (30.0%, p 0.07), stage I=II
(21.0%, p 0.35), and noninltrative EC (33.3%, p 0.06).
The contribution of PIK3CA mutations to the characteristics
of KRAS-mutated samples is presented in Table 3. Although
statistical signicance could not be reached except for the
high grade (5=20, 25%; p 0.01), KRAS-onlymutated sam-
ples were not present in any of the 13 stage III=IV ( p 0.35),
9 metastatic ( p 0.59), or 12 serous ( p 0.6) samples.
Discussion
Class I A PI3 kinases are key components of the PIK=Akt
pathway. Upon tyrosine kinase activation, the PI3K is acti-
vated and recruited to the plasma membrane where it pro-
duces phosphatidylinositol 3,4,5-triphosphate. This triggers
membrane recruitment, phosphorylation, and activation of
downstream tyrosine kinases Akt and PDK1, which further
regulate gene translation through downstream targets.
In EC, PIK3CA mutations have been reported from around
10% (Miyake et al., 2008) to nearly 40% (Hayes et al., 2006) of
the cases and their presence has been related to invasion
(Hayes et al., 2006; Catasus et al., 2008). We have sequenced
hotspot exons 9 and 20 of the PIK3CA gene in 108 EC sam-
ples and have detected a mutation frequency of 22.2%.
Consistent with previous reports (Hayes et al., 2006; Catasus
et al., 2008), mutations in PIK3CA were more common among
tumors demonstrating invasion as evidenced by the ad-
vanced stage of the disease (30.8%, p 0.48) and the presence
of metastasis (44.4%, p 0.1). However, we did not detect a
difference in distribution depending on grade or degree of
myoinltration, which is also consistent with a recent report
Table 2. Distribution of Mutations in Exon 9 and Exon 20 of the PIK3CA Gene
Depending on Patient Clinicopathological Parameters
Variable
No. of
cases
No. of cases
with PIK3CA
mutation (%) p
No. of cases
with exon 20
mutation (%) p
No. of cases
with exon 9
mutation (%) p
No. of cases
with E545K,
E542K, and
H1047 (%) p
No. of cases
with rare
mutation (%) p
Grade
G1 20 4 (20.0) 3 (15.0) 1 (5.0) 2 (10.0) 2 (10.0)
G2 62 14
a
(22.6) 10 (16.1) 5 (8.1) 8 (12.9) 6 (9.7)
G3 22 6 (27.3) 3 (13.6) 3 (13.6) 3 (13.6) 3 (13.6)
Not graded
b
4 0 0 0 0
Stage
I and II 95 20 (21.0) 13 (13.7) 7 (7.4) 12 (12.6) 8 (8.4)
III and IV 13 4
a
(30.8) 3 (23.1) 2 (15.4) 1 (7.7) 3 (23.1)
Myometrial inltration
No inltration 15 3 (20.0) 1 (6.7) 2 (13.3) 2 (13.3) 1 (6.7)
Up to 1=3 48 12
a
(25.0) 9 (18.7) 4 (8.3) 4 (8.3) 8 (16.7)
Up to 2=3 28 5 (17.8) 4 (14.3) 1 (3.6) 3 (10.7) 2 (7.1)
More than 2=3 17 4 (23.5) 2 (11.8) 2 (11.8) 4 (23.5) 0
Histology type
EM 90 18 (20.0) 11 (12.2) 7 (7.8) 11 (12.2) 7 (7.8)
Ser or mixed Ser=EM 12 6
a
(50.0) 0.021 5 (41.7) 0.005 2 (16.7) 2 (16.7) 4 (33.3) 0.02
CC or mixed CC=EM 6 0 0 0 0 0
Metastatic status
9 4
a
(44.4) 3 (33.3) 2 (22.2) 1 (11.1) 3 (33.3) 0.047
99 20 (20.2) 13 (13.1) 7 (7.1) 12 (12.1) 8 (8.1)
a
One sample bears a mutation in both exons of the gene.
b
Four of the nonendometrioid samples have not been graded.
EM, endometrioid; Ser, serous; CC, clear cell.
RARE MUTATIONS IN THE PIK3CA GENE AND ENDOMETRIAL CANCER 67
(Salvesen et al., 2009). As this is the rst study documenting
PIK3CA mutational frequency in metastatic EC, we suggest
the need for further evaluation of the possible implications of
PIK3CA mutations, particularly for distant spread.
The hotspot H1047R, E542K, and E545K substitutions are
the most common PIK3CA mutations in human cancer. In
our sample they represent 50.0% of all the mutations.
Functional characterization of H1047R, E542K, and E545K
has demonstrated that they led to an increased lipid kinase
activity (Samuels et al., 2004; Ikenoue et al., 2005; Kang et al.,
2005) and were able to activate the Akt pathway in the ab-
sence of growth factors (Isakoff et al., 2005; Kang et al., 2005).
Experiments involving overexpression of the three hotspot
mutants have demonstrated that they were able to induce
cellular oncogenic transformation (Ikenoue et al., 2005; Kang
et al., 2005; Bader et al., 2006) and were sufcient for tumor
formation in vivo (Bader et al., 2006). Analysis using mutant
knock-in techniques, rather than overexpression, has sug-
gested that these mutations might not be sufcient for in vitro
or in vivo transformation (Gustin et al., 2009) while con-
rming their malignant potential.
At the same time, mutations in the helical and kinase re-
gions have been observed to have different functional
properties suggestive of a different mechanism for p110a
activation (Carson et al., 2008). In vivo studies have shown
that E542K and E545K (both situated in exon 9) led to the
formation of tumors less frequently than does the H1047R
(situated in exon 20) (Bader et al., 2006). Consistently, the
most frequently observed PIK3CA mutation in our patient
group was the H1047R. Part of the observations that helical
and kinase mutations act through different molecular
mechanisms has come from the analysis of constructs bear-
ing mutations in both exons (Zhao and Vogt, 2008). These
constructs have shown a strong synergistic effect in their
transforming potential compared to double mutants situated
in the same exon. We described a tumor that harbored mu-
tations in both domains (P539R and T1025I), and which
was a grade 2, stageIII, endometrioid=serous carcinoma
with distant metastasis and a tumor that harbored two
somatic substitutions in the helical domainp.P539T and
g.70434C>A. The latter was a grade 3, stage I endometrioid
EC. We found a high incidence (41.7%, p 0.005) of exon 20
mutations in tumors involving serous differentiation, but we
were otherwise not able to observe distinctive distribution
pattern of mutations situated in the two domains in all other
comparisons.
Activating mutations in the KRAS gene are another com-
mon molecular abnormality in EC, and the active form of
RAS is known to interact and thus activate the p110 catalytic
subunit of PI3K directly (Rodriguez-Viciana et al., 1996), as
well as indirectly (Toulany et al., 2007).
We have determined that 16.7% of the cases harbored a
somatic KRAS mutation. KRAS mutations showed higher
incidence in grade 1 (30%, p 0.07), stages I=II (21%, p
0.35), and in noninltrative cancer (33.3%, p 0.06). A num-
ber of previous studies have reported similar results (Tsuda
et al., 1995; Pappa et al., 2006), whereas others have not
detected overrepresentation depending on stage, grade, or
depth of myometrial invasion (Caduff et al., 1995; Esteller
et al., 1997).
Although mutations in two key players of a cascade sel-
domly coexist in a tumor, KRAS and PIK3CA double mutants
have been described in EC (Ollikainen et al., 2007; Catasus
et al., 2008). Mutations in the two genes have also been re-
ported to occur in a mutually exclusive manner (Velasco et al.,
2006; Kang et al., 2008). In our cohort, KRAS mutations oc-
curred more frequently in PIK3CA-mutated (7=24, 29.2%)
than in PIK3CA wild-type tumors (11=84, 13.1%; p 0.06),
suggesting a high degree of coexistence and a nonidentical
functional signicance.
KRAS mutations have been suggested to occur in pre-
dominantly noninvasive EC samples and to necessitate
Table 3. Characteristics of KRAS-Mutated Samples
Variable
No. of
cases KRAS (%) p
No. of KRAS cases
without PIK3CA (%) p
No. of KRAS cases
with PIK3CA (%) p
Grade
G1 20 6 (30.0) 0.076 5 (25.0) 0.015 1 (5.0)
G2 62 9 (14.5) 5 (8.1) 4 (6.4)
G3 22 3 (13.6) 1 (4.5) 2 (9.1)
Not graded
a
4 0 0
Stage
I and II 95 17 (21.0) 11 (11.6) 6 (6.3)
III and IV 13 1 (7.7) 0 1 (7.7)
Myometrial inltration
No inltration 15 5 (33.3) 0.061 3 (20.0) 2 (14.3)
Up to 1=3 48 7 (14.6) 3 (6.2) 4 (9.5)
Up to 2=3 28 4 (14.3) 3 (10.7) 1 (3.6)
More than 2=3 17 2 (11.8) 2 (11.8) 0
Histology type
EM 90 15 (16.7) 10 (11.1) 5 (5.5)
Ser or mixed Ser=EM 12 2 (16.7) 0 2 (16.7)
CC or mixed CC=EM 6 1 (16.7) 1 (16.7) 0
Metastatic status
9 1 (11.1) 0 1 (11.1)
99 17 (17.2) 11 (11.1) 6 (6.1)
a
Four of the nonendometrioid samples have not been graded.
68 KONSTANTINOVA ET AL.
additional abnormalities for enhancement of their oncogenic
potential (Oda et al., 2008). Consistently, coexpression of
mutant KRAS together with PIK3CA has shown more potent
p-Akt activation and transformation potential than either of
the two expressed separately (Oda et al., 2008). In our cohort,
KRAS-mutated samples in which a PIK3CA mutation was
absent showed an association to high grade (5=20, 25%;
p 0.01) and were not found among advance staged, meta-
static or serous tumor samples.
PIK=Akt signaling is known to be implicated in cancer
invasion and metastasis mainly through Akt activation (re-
viewed in Qiao et al., 2008) and also through PI3K activation,
which has been suggested to have an independent role in
promoting metastasis (Hutchinson et al., 2001). We have
evaluated the contribution of mutations in the PIK3CA and
KRAS genes to metastatic EC. Although we found no KRAS
mutations, nearly half of the metastatic tumors harbored
a mutation in the PIK3CA gene (44.4%; p 0.1). Previously,
mutations in the gene have been related to lymphovascular
involvement in EC (Catasus et al., 2008), which supports the
importance of PIK3CA mutation for distant spread and un-
derlies the need for further investigation of its prognostic as
well as therapeutic implications. Further, we found that the
group of common mutations did not contribute to this trait
( p 1.000), whereas rare mutations involving both helical
and kinase domains of the gene showed an association to
metastatic disease ( p 0.047).
Rare mutations in the two hotspot domains of the PIK3CA
gene have not been as extensively functionally characterized
as common ones. They not only occur less frequently but also
most of them fail to show the same nonrandom pattern of
appearance that common mutations show. Functional anal-
ysis of some of the rare PIK3CA mutants has demonstrated
that they also possess an increased lipid kinase activity
leading to constitutive signaling to Akt and ultimately to
tumorigenicity (Gymnopoulos et al., 2007). However, their
oncogenic potential has been graded as inferior to that of the
three common mutants (Gymnopoulos et al., 2007). Our re-
sults suggest that further exploration of rare mutations in the
gene is warranted to provide explanation to their biological
activities and their signicance for transformation and in-
vasion. Rare mutations outside exons 9 and 20 have also
been found to confer p110a gain of function (Gymnopoulos
et al., 2007; Oda et al., 2008).
Activation of Akt in human endometrium as a result of
the straightforward scheme RAS=PI3K=p-Akt has been
questioned in favor of a complex and yet not well-dened
signaling network controlling its expression, as p-Akt ex-
pression has not shown direct relation to PIK3CA mutational
status (Velasco et al., 2006; Mori et al., 2007). Moreover,
H1047R-mutated tumors (two samples) have not shown
p-Akt expression, whereas rare mutations in the PIK3CA
gene, including the H1047L (two samples), have done so
(Mori et al., 2007). If conrmed on additional samples, these
observations could reveal new aspects of p110a tumorige-
nicity and the potential signicance of more than one muta-
tion affecting the pathway.
Our subgroup of tumors displaying serous differentiation
(serous and mixed carcinomas) showed a very high fre-
quency (50.0%) of PIK3CA mutations. We cannot speculate
which of the two components of mixed tumors might bear
the mutation; however, the distribution was nonrandom
( p 0.02). A high frequency of PIK3CA mutations (15%) has
recently been described in a large cohort of serous ECs
(Hayes et al., 2009). The sum of serous carcinomas from both
cohorts (n 39) gives a mutation frequency of 15.4% (6=39),
the majority of mutations being non-hotspot (5=6, 83%) and
situated in exon 20 (5=6, 83%), suggesting a particular mu-
tational spectrum in these tumors. Mixed endometrioid=
serous tumors from our cohort (n 7) also contained an ex-
cess of exon 20 (4=5, 80%) and rare mutations (3=5, 60%).
Serous EC, although rare, is an aggressive entity and further
evaluation of the possible contribution of PIK3CA mutations
for this type of cancer is warranted.
It is interesting to parallel our results regarding the distri-
bution of PIK3CA and KRAS mutations depending on EC
histology type with that in ovarian cancer. Although PIK3CA
mutation is not common in ovarian cancer, it has been re-
ported to occur preferentially in clear cell and in endometrioid
(Willner et al., 2007) rather than in serous ovarian cancer
(Campbell et al., 2004; Nakayama et al., 2006; Willner et al.,
2007). Exactly the opposite is valid for mutations in the KRAS
gene, which are typical for serous ovarian cancer, and in
particular for low-grade serous ovarian cancer. These obser-
vations suggest that interpretation of the signicance of
PIK3CAmutations should be done in a tissue-specic manner.
In conclusion, PIK3CA mutations are common in EC and
often coexist with a KRAS mutation. Tumors containing se-
rous components show a particular mutational pattern. Rare
mutations in the gene, but not hotspot, show an association
to aggressive ECserous differentiation and presence of
metastasissuggesting the need of further exploration of the
molecular basis of rare mutations. None of the observations
can be conclusive until further conrmed.
Acknowledgments
Financial support for this study was provided by a re-
search grant VU-1-07=2005 from the National Science Fund
and 63=2009 from the Medical UniversitySoa.
Disclosure Statement
No competing nancial interests exist.
References
Bader, A.G., Kang, S., and Vogt, P.K. (2006). Cancer-specic
mutations in PIK3CA are oncogenic in vivo. Proc Natl Acad Sci
USA 103, 14751479.
Caduff, R.F., Johnston, C.M., and Frank, T.S. (1995). Mutations
of the Ki-ras oncogene in carcinoma of the endometrium. Am J
Pathol 146, 182188.
Campbell, I.G., Russell, S.E., Choong, D.Y., Montgomery, K.G.,
Ciavarella, M.L., Hooi, C.S., Cristiano, B.E., Pearson, R.B., and
Phillips, W.A. (2004). Mutation of the PIK3CA gene in ovarian
and breast cancer. Cancer Res 64, 76787681.
Cantley, L.C. (2002). The phosphoinositide 3-kinase pathway.
Science 296, 16551657.
Carson, J.D., Van Aller, G., Lehr, R., Sinnamon, R.H., Kirkpa-
trick, R.B., Auger, K.R., Dhanak, D., Copeland, R.A., Gontarek,
R.R., Tummino, P.J., and Luo, L. (2008). Effects of oncogenic
p110alpha subunit mutations on the lipid kinase activity of
phosphoinositide 3-kinase. Biochem J 409, 519524.
Catasus, L., Gallardo, A., Cuatrecasas, M., and Prat, J. (2008).
PIK3CA mutations in the kinase domain (exon 20) of uterine
RARE MUTATIONS IN THE PIK3CA GENE AND ENDOMETRIAL CANCER 69
endometrial adenocarcinomas are associated with adverse
prognostic parameters. Mod Pathol 21, 131139.
Esteller, M., Garcia, A., Martinez-Palones, J.M., Xercavins, J., and
Reventos, J. (1997). The clinicopathological signicance of
K-RAS point mutation and gene amplication in endometrial
cancer. Eur J Cancer 33, 15721577.
Gustin, J.P., Karakas, B., Weiss, M.B., Abukhdeir, A.M., Lauring,
J., Garay, J.P., Cosgrove, D., Tamaki, A., Konishi, H., Konishi,
Y., Mohseni, M., Wang, G., Rosen, D.M., Denmeade, S.R.,
Higgins, M.J., Vitolo, M.I., Bachman, K.E., and Park, B.H.
(2009). Knockin of mutant PIK3CA activates multiple onco-
genic pathways. Proc Natl Acad Sci USA 106, 28352840.
Gymnopoulos, M., Elsliger, M.A., and Vogt, P.K. (2007). Rare
cancer-specic mutations in PIK3CA show gain of function.
Proc Natl Acad Sci USA 104, 55695574.
Hayes, M.P., Douglas, W., and Ellenson, L.H. (2009). Molecular
alterations of EGFR and PIK3CA in uterine serous carcinoma.
Gynecol Oncol 113, 370373.
Hayes, M.P., Wang, H., Espinal-Witter, R., Douglas, W., Solo-
mon, G.J., Baker, S.J., and Ellenson, L.H. (2006). PIK3CA and
PTEN mutations in uterine endometrioid carcinoma and
complex atypical hyperplasia. Clin Cancer Res 12(20 Pt 1),
59325935.
Hutchinson, J., Jin, J., Cardiff, R.D., Woodgett, J.R., and Muller,
W.J. (2001). Activation of Akt (protein kinase B) in mammary
epithelium provides a critical cell survival signal required for
tumor progression. Mol Cell Biol 21, 22032212.
Ikenoue, T., Kanai, F., Hikiba, Y., Obata, T., Tanaka, Y., Im-
amura, J., Ohta, M., Jazag, A., Guleng, B., Tateishi, K., Asaoka,
Y., Matsumura, M., Kawabe, T., and Omata, M. (2005).
Functional analysis of PIK3CA gene mutations in human co-
lorectal cancer. Cancer Res 65, 45624567.
Isakoff, S.J., Engelman, J.A., Irie, H.Y., Luo, J., Brachmann, S.M.,
Pearline, R.V., Cantley, L.C., and Brugge, J.S. (2005). Breast
cancer-associated PIK3CA mutations are oncogenic in mam-
mary epithelial cells. Cancer Res 65, 1099211000.
Kang, S., Bader, A.G., and Vogt, P.K. (2005). Phosphatidylino-
sitol 3-kinase mutations identied in human cancer are on-
cogenic. Proc Natl Acad Sci USA 102, 802807.
Kang, S., Seo, S.S., Chang, H.J., Yoo, C.W., Park, S.Y., and Dong,
S.M. (2008). Mutual exclusiveness between PIK3CA and KRAS
mutations in endometrial carcinoma. Int J Gynecol Cancer 18,
13391343.
Miyake, T., Yoshino, K., Enomoto, T., Takata, T., Ugaki, H., Kim,
A., Fujiwara, K., Miyatake, T., Fujita, M., and Kimura, T.
(2008). PIK3CA gene mutations and amplications in uterine
cancers, identied by methods that avoid confounding by
PIK3CA pseudogene sequences. Cancer Lett 261, 120126.
Mori, N., Kyo, S., Sakaguchi, J., Mizumoto, Y., Ohno, S., Maida,
Y., Hashimoto, M., Takakura, M., and Inoue, M. (2007).
Concomitant activation of AKT with extracellular-regulated
kinase 1=2 occurs independently of PTEN or PIK3CA muta-
tions in endometrial cancer and may be associated with fa-
vorable prognosiss. Cancer Sci 98, 18811888.
Nakayama, K., Nakayama, N., Kurman, R.J., Cope, L., Pohl, G.,
Samuels, Y., Velculescu, V.E., Wang, T.L., and Shih Ie, M.
(2006). Sequence mutations and amplication of PIK3CA and
AKT2 genes in puried ovarian serous neoplasms. Cancer Biol
Ther 5, 779785.
Oda, K., Okada, J., Timmerman, L., Rodriguez-Viciana, P., Sto-
koe, D., Shoji, K., Taketani, Y., Kuramoto, H., Knight, Z.A.,
Shokat, K.M., and McCormick, F. (2008). PIK3CA cooperates
with other phosphatidylinositol 3
0
-kinase pathway mutations
to effect oncogenic transformation. Cancer Res 68, 81278136.
Ollikainen, M., Gylling, A., Puputti, M., Nupponen, N.N.,
Abdel-Rahman, W.M., Butzow, R., and Peltomaki, P. (2007).
Patterns of PIK3CA alterations in familial colorectal and en-
dometrial carcinoma. Int J Cancer 121, 915920.
Pappa, K.I., Choleza, M., Markaki, S., Giannikaki, E., Kyroudi, A.,
Vlachos, G., Voulgaris, Z., and Anagnou, N.P. (2006). Consistent
absence of BRAF mutations in cervical and endometrial cancer
despite KRAS mutation status. Gynecol Oncol 100, 596600.
Qiao, M., Sheng, S., and Pardee, A.B. (2008). Metastasis and AKT
activation. Cell Cycle 7, 29912996.
Rodriguez-Viciana, P., Marte, B.M., Warne, P.H., and Down-
ward, J. (1996). Phosphatidylinositol 3
0
kinase: one of the
effectors of Ras. Philos Trans R Soc Lond 351, 225231; dis-
cussion 231222.
Salvesen, H.B., Carter, S.L., Mannelqvist, M., Dutt, A., Getz, G.,
Stefansson, I.M., Raeder, M.B., Sos, M.L., Engelsen, I.B., Tro-
vik, J., Wik, E., Greulich, H., Bo, T.H., Jonassen, I., Thomas,
R.K., Zander, T., Garraway, L.A., Oyan, A.M., Sellers, W.R.,
Kalland, K.H., Meyerson, M., Akslen, L.A., and Beroukhim, R.
(2009). Integrated genomic proling of endometrial carcinoma
associates aggressive tumors with indicators of PI3 kinase
activation. Proc Natl Acad Sci USA.
Samuels, Y., Wang, Z., Bardelli, A., Silliman, N., Ptak, J., Szabo,
S., Yan, H., Gazdar, A., Powell, S.M., Riggins, G.J., Willson,
J.K., Markowitz, S., Kinzler, K.W., Vogelstein, B., and Velcu-
lescu, V.E. (2004). High frequency of mutations of the PIK3CA
gene in human cancers. Science 304, 554.
Toulany, M., Baumann, M., and Rodemann, H.P. (2007). Sti-
mulated PI3K-AKT signaling mediated through ligand or
radiation-induced EGFR depends indirectly, but not directly,
on constitutive K-Ras activity. Mol Cancer Res 5, 863872.
Tsuda, H., Jiko, K., Yajima, M., Yamada, T., Tanemura, K.,
Tsunematsu, R., Ohmi, K., Sonoda, T., and Hirohashi, S.
(1995). Frequent occurrence of c-Ki-ras gene mutations in well
differentiated endometrial adenocarcinoma showing inltra-
tive local growth with brosing stromal response. Int J Gy-
necol Pathol 14, 255259.
Velasco, A., Bussaglia, E., Pallares, J., Dolcet, X., Llobet, D.,
Encinas, M., Llecha, N., Palacios, J., Prat, J., and Matias-Guiu,
X. (2006). PIK3CA gene mutations in endometrial carcinoma:
correlation with PTEN and K-RAS alterations. Hum Pathol 37,
14651472.
Willner, J., Wurz, K., Allison, K.H., Galic, V., Garcia, R.L., Goff,
B.A., and Swisher, E.M. (2007). Alternate molecular genetic
pathways in ovarian carcinomas of common histological
types. Hum Pathol 38, 607613.
Zhao, L., and Vogt, P.K. (2008). Helical domain and kinase do-
main mutations in p110alpha of phosphatidylinositol 3-kinase
induce gain of function by different mechanisms. Proc Natl
Acad Sci USA 105, 26522657.
Address correspondence to:
Darina Konstantinova, M.Sc.
Department of Chemistry and Biochemistry
Molecular Medicine Center
Medical University
2 Zdrave St.
Soa 1431
Bulgaria
E-mail: darivko@gmail.com
Received for publication July 6, 2009; received in revised
form August 31, 2009; accepted September 1, 2009.
70 KONSTANTINOVA ET AL.
Copyright of DNA & Cell Biology is the property of Mary Ann Liebert, Inc. and its content may not be copied
or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.