Genome Analysis of Latin American Cervical Cancer: Frequent Activation of the PIK3CA Pathway
Supplementary Data Files in this Data Supplement:
Supplementary Figure 1- Supplementary Figure 1. Sample and analyses diagram.
Supplementary Figure 2- Supplementary Figure 2. Aggregate mutation spectrum of Guatemalan cervical
tumors.
Supplementary Figure 3- Supplementary Figure 3. Highly amplified regions of chromosome (Chr.) 2, 6
and 10 in tumor A4.
Supplementary Figure 4- Supplementary Figure 4. Chromosome breakpoints in cervical tumors.
Supplementary Figure 5- Supplementary Figure 5. PIK3CA mutation and pathology.
Supplementary Figure 6- Supplementary Figure 6. Co-occurring mutations in the PIK3CA and PTEN
genes.
Supplementary Figure 7- Supplementary Figure 7. Sanger sequence concordance of mutant read
percentage.
Supplementary Figure 8- Supplementary Figure 8. Mutation Fraction of PIK3CA mutations.
Supplementary Figure 9- Supplementary Figure 9. Age range associations with PIK3CA mutation and
HPV type.
Supplementary Figure 10- Supplementary Figure 10. PIK3CA and HPV E6 and E7 mRNA expression.
Supplementary Figure 11- Supplementary Figure 11. Model of cervical cancer progression.
Supplementary Figure 1. Sample and analyses diagram.
Venezuela
Inv. Cancer=41
Pre-Cancer
In situ=34
Guatemala
Mexico
Inv. Cancer=296
Inv. Cancer=330
Exome=23
Paired normal DNA 23
HPV=36
No seq=5
HPV=31
No seq=3
HPV=279
No seq=17
HPV=324
No seq=6
Target seq=40
Fail=1
Target seq=30
Fail=4
Target seq=280
Fail=8
Target seq=325
Fail=5
Paired normal DNA=120
Legend: The diagram displays the number and type of sample from each population, and the number of
samples used in exome sequence (Exome), HPV sequencing and targeted sequencing (target seq). Samples
that failed to provide a definitive HPV type or adequate reads for targeted sequence are displayed as No seq
and Fail
Supplementary Figure 2. Aggregate mutation spectrum of Guatemalan cervical tumors.
Legend: The aggregate mutation spectrum of mutations of different type and context is
shown. Predicted somatic mutations were categorized by mutation type and flanking
sequence and displayed in aggregate for 23 tumors with exome sequence.
Supplementary Figure 3. Highly amplified regions of chromosome (Chr.) 2, 6 and 10 in tumor
A4.
Chr. 2
Chr. 6
Chr. 10
Legend: Chromosome plots are displayed for a tumor (A4) with three amplified
chromsosomal regions.
Supplementary Figure 4. Chromosome breakpoints in cervical tumors.
Chromosome Abnormalites
300
250
200
150
100
50
A9
A1
A8
B6
A12
B5
B9
A7
B7
A3
B3
A5
B10
B4
B2
A4
A10
A2
B8
A11
B1
B11
0
ASC
AC
SC
Legend: Total number of chromosomal breakpoint was determined by ngCGH (See
Methods) for each tumor is displayed. ASC represent Adenosquamous, AC represent
Adenocarcinoma and SC represent Squamous.
Supplementary Figure 5. PIK3CA mutation and pathology
Legend: The percentage of PIK3CA mutations are shown separately for each major histological
type. Squamous tumors have significantly higher levels of helical mutations (residues E542 and
E545); X2=5.7, P=0.017 for squamous and adeno tumors vs. helical mutations and non-helical
mutations.
Helical
other
Squamous
Adeno
132
14
20
7
Supplementary Figure 6. Co-occurring mutations in the PIK3CA and PTEN genes.
ABD
1
RBD
108
191
C2
291 330
Helical
480
525
Kinase
696
1068
Supplementary Figure 6. Co-occurring mutations in the PIK3CA and PTEN genes
A
B
E542Q_E545K cis
E542K_E545Q cis
Supplementary Figure 6. Co-occurring mutations in the PIK3CA and PTEN genes
C
D
E542K/E545K trans
E542K/Q546R trans
Supplementary Figure 6. Co-occurring mutations in the PIK3CA and PTEN genes
E
E542K/E547D cis
Supplementary Figure 6. Co-occurring mutations in the PIK3CA and PTEN genes
F
G
E542K/E545K cis
E453K_D454Y cis
H
PTEN R130L/F154I trans
Legend: Eight examples are shown of unique combinations of adjacent mutations in the
PIK3CA helical domain involving residues 542-547 (A-G) and the PTEN gene (H). The
sequence reads are displayed above the reference DNA and protein sequence (red reads are
forward; blue, reverse). Mutations are observed that are on the same strand (cis) or different
strands (trans). A, E542Q_E545K in cis; B, E542K_E545Q in cis; C, E542K/E545K in trans; D,
E542K_Q546R in trans (blue circle is the rare E542K found in 1% of reads in this sample, and
red circles the Q546K reads found at 6%); E, E542K_E547D in cis; F, E542K and E545K in cis;
G, E453K and D454Y in cis; H, PTEN R130L and F154I in trans.
Supplementary Figure 7. Sanger sequence concordance of mutant read percentage.
A
R38H 41% mutant by AmpliSeq; 31-37% by Sanger
Supplementary Figure 7. Sanger sequence concordance of mutant read percentage.
B
C
E545K
NG=15%, Sanger 17-20%
E545K
NG=5%, Sanger ~5%
Legend Supplementary Figure 7: Three examples are shown of mutations in the
PIK3CA gene at differing percentages of mutant reads, and the corresponding
Sanger sequencing traces. A, A R38H mutation seen in 41% of mutant reads by
AmpliSeq was observed to have a 31% drop in wt signal in the forward
direction and 37% drop in the reverse direction (from the Mutation Surveyor
software.) B, E545K mutation seen at 15% by Next-generation sequence (NG)
and 17-20% by Sanger sequencing; C, E545K detected in 5% of mutant reads
and called in the next-generation sequence, not called by Mutation Surveyor,
but visible in the sequence traces (red circles).
Supplementary Figure 8. Mutation Fraction of PIK3CA mutations.
A
Mutation fraction (%)
25
3-10%
20
11-20%
20-33%
15
>33%
10
5
0
E542K Mut
B
E545K Mut
Other
Mutation Fraction
0.25
0.2
0.15
0.1
0.05
0
E542K
E545K
Legend: Mutation fraction of PIK3CA mutations. A) The number of tumors with PIK3CA
mutations at different mutation fraction is shown for E542K, E545K, and others. B) The
mutation fraction of each mutation is displayed for tumors with E542K-E545K double
mutants.
Supplementary Figure 9. Age range associations with PIK3CA mutation and HPV type.
A.
B.
Supplementary Figure 9. Age range associations with PIK3CA mutation and HPV type.
C.
Legend: Age groups of cervical tumor patients. A) The frequency of PIK3CA mutation by age
group is displayed. The mean ages are 49.1 for non-mutated and 53.8 for mutated, which is a
statistically significant difference (P=0.001). B. Relation of major HPV type with age. C.
Percentage of total somatic driver mutations by age group and stage for squamous
carcinomas. Subjects sampled under age 50 have a significantly lower percentage [X2=31, P=
2.1 E-08].
Supplementary Figure 10. PIK3CA and HPV E6 and E7 mRNA expression.
A
B
Legend: A, PIK3CA mRNA expression levels are significantly higher in HPV16 positive
tumors with mutated PIK3CA than with wild-type PIK3CA, as determined by quantitative
RT-PCR (P = 0.029). B, Differences in E6 and E7 mRNA expression levels between
wild-type PIK3CA (PIK3CA _WT) and mutant PIK3CA (PIK3CA _MT) in HPV16 positive
cervical tumors. The E7 expression level in PIK3CA_MT is significantly higher than in
PIK3CA_WT (P = 0.040). Primer sequences were GAGCGACCCAGAAAGTTACCA (forward) and
AAA TCC CGA AAA GCA AAG TCA (reverse) for E6 HPV16; TCC AGC TGG ACA AGC AGA AC
(forward) and CAC AAC CGA AGC GTA GAG TC (reverse) for E7 HPV16; AAT AAG GTG CCT GCG
GTG (forward) and CTT GTG TTT CTC TGC GTC GT (reverse) for E6 HPV18; AAC ATT TAC CAG CCC
GAC GA (forward) and TCG TCT GCT GAG CTT TCT AC (reverse) for E7 HPV18.
Supplementary Figure 11. Model of cervical cancer progression.
HPV Infection
E6, E7
PIK3CA
1. Gain ̶ ˃ Mutation
3q Gain
HPV Integration/
Chromosome 3q Duplication/
PIK3CA Mutation
W
mutation
W
W
W
PIK3CA
M ≤33% reads
2. Mutation ̶ ˃ Amplification
mutation
amplification
PIK3CA
PI3K-downstream
Pathway
Activation
Metastasis
W
WM
PIK3CA
WM
M> 66% reads
Legend: Model for the role of PIK3CA up-regulation and
mutation in invasive cervical cancer. The expression of the HPV
E6/E7 is crucial for HPV-induced cervical cancer
transformation, and leads to elevated PIK3CA expression.
Duplication of one copy of chromosome 3q may precede
PIK3CA mutations or duplication may occur on a wild-type
allele (1) and yield a tumor with <33% PIK3CA mutant fraction;
or PIK3CA mutation can occur before amplification of
chromosome 3q. (2) leading a >66% mutant fraction.
Permanent activation of the PI3K-downstream signaling
pathway by mutation helps drive invasion and eventual
metastatic spread.