2323

High Prevalence of p53 Exon 4 Mutations in Soft

Tissue Sarcoma
Parimal Das, PhD1
Dhanasekaran Kotilingam, MD2
Borys Korchin, MD, PhD2
Jeuhui Liu, PhD1
Dihua Yu, MD, PhD1
Alexander J. Lazar, MD, PhD3
Raphael E. Pollock, MD, PhD1
Dina Lev, MD2

BACKGROUND. p53 is the most commonly mutated gene in cancer, including soft
tissue sarcoma (STS). The authors characterized p53 alterations (protein accumulation and gene mutation) in STS to evaluate possible associations with patient
outcomes.

METHODS. Thirty-one STS specimens (multiple histologies) were analyzed by p53

immunohistochemistry (IHC) and direct DNA sequencing of p53 exons 211 and
then correlated with outcomes.

RESULTS. Direct p53 sequencing detected mutations in 10 of 31 STSs; 7 of 10

were missense mutations, whereas 3 of 10 were either insertions or frameshift

mutations, leading to nonfunctional truncated p53; 7 of these p53 mutations

Department of Surgical Oncology, University of

Texas M. D. Anderson Cancer Center, Houston,
Texas.

have not been previously described. Four p53 exon 4 mutations were identified, a

Department of Cancer Biology, University of

Texas M. D. Anderson Cancer Center, Houston,
Texas.

mens expressed p53 when the authors used the clinical IHC assay of their institu-

IHC, whereas 12 of 18 specimens exhibiting p53 protein expression by IHC har-

Department of Pathology, University of Texas

M. D. Anderson Cancer Center, Houston, Texas.

p53 region previously unknown to be mutation prone. Eighteen of the 31 specition. Interassay concordance of 48% was observed; only 6 of 10 sequencingidentified p53 mutated specimens exhibited nuclear p53 protein expression by
bored sequencing-identified wild-type p53. Decreased survival was observed in
STS patients bearing sequencing-identified mutated p53 versus wild-type p53, as
was a correlation between IHC-determined nuclear p53 protein expression and
decreased survival.

CONCLUSIONS. p53 protein stabilization and p53 mutation frequently occur in

STS, and both suggest worse outcomes for patients so affected. However,
increased p53 protein expression does not necessarily indicate p53 gene mutation. The high incidence of exon 4 mutations found in STS suggests that p53
sequencing should not be limited to the core DNA binding domain. Cancer
2007;109:232333.
2007 American Cancer Society.

KEYWORDS: soft tissue sarcoma, p53, mutations, exon 4, sequencing, immunohistochemistry.

Supported in part by the National Institutes of

Health grant RO1 CA 67802 (to R.E.P.)
Address for reprints: Dina Lev, MD, Department
of Cancer Biology, Unit 1104, University of Texas
M. D. Anderson Cancer Center, 1515 Holcombe
Blvd, Houston, TX 77030; Fax: (713) 563-1185;
E-mail: dlev@mdanderson.org
Received November 14, 2006; revision received
January 21, 2007; accepted February 14, 2007.

2007 American Cancer Society

oft tissue sarcoma encompasses a large family of more than 50

histologically distinct tumor subtypes, all of which share a putative mesenchymal origin. Overall survival is approximately 50% at 5
years depending upon tumor size, grade, location, the presence of
regional or distant disease, as well as responsiveness to surgical,
radiational, and/or chemotherapeutic approaches.1,2 Uncontrolled
metastasis is particularly ominous and accounts for approximately
80% of disease-specific sarcoma deaths.3,4 In contrast to the much
more common epithelial origin malignancies, less is known about
underlying molecular mechanisms leading to soft tissue sarcoma
progression and metastasis. Mutations in the p53 tumor-suppressor
gene have also been identified as the most common genetic alterations in soft tissue sarcoma.5 Such mutations are found more fre-

DOI 10.1002/cncr.22680
Published online 11 April 2007 in Wiley InterScience (www.interscience.wiley.com).

2324

CANCER

June 1, 2007 / Volume 109 / Number 11

quently in metastases than in primary tumors and in

high-grade versus low-grade sarcomas, and these
mutations are thought to have a significant negative
impact on both overall as well as sarcoma-specific
survival.68
Our own investigations of autologous human primary and metastatic sarcoma have demonstrated
that clonal expansion of p53-mutated cells in soft tissue sarcoma confers distinct metastatic advantages.9
Moreover, we have shown that reintroduction of
wtp53 into soft tissue sarcoma cells that harbor a p53
mutation inhibits cell proliferation, soft agar colony
formation in vitro, and tumorigenesis in severe combined immunodeficient (SCID) mice.10 We have also
demonstrated that p53 alterations in soft tissue sarcoma contribute to metastasis-promoting behaviors,
including loss of cell cycle control,11 enhanced angiogenesis,12 invasiveness,13 and chemoresistance.14,15
The p53 gene protein product is expressed and
activated in response to multiple signals of cellular
stress such as DNA damage, hypoxia, or cell cycle
perturbation, leading to induction of either transient
growth arrest or programmed cell death (apoptosis).16
Apoptosis is usually seen in response to a genotoxic
insult that exceeds the DNA repair potential of a
given cell; this response to genotoxic stress is
mediated by ATM. Under normal conditions, p53 has
multiple effectors including BAX/bcl-2 and p21 that
mediate the pathway to apoptosis and blockade of
cell progression, respectively. The cell can downregulate p53 levels through the MDM2 pathway,
which marks p53 for ubiquitination and destruction,
thereby preventing inappropriate apoptosis.16 A nonfunctional p53 pathway offers survival advantages to
the cell via unregulated progression through the cells
cycle, thereby facilitating cell growth, allowing for
more rapid mutational accumulation, and facilitating
the emergence of potentially more aggressive malignancy. The majority of p53 mutations alter either the
structural integrity of p53 protein and/or p53-DNA
interactions, leading to partial or complete loss of
p53 protein functions.17 Because p53 is a tetramer
when functional, a mutated monomer may be able to
disrupt the function of the complex by acting as a
dominant negative factor. Intriguingly, the preponderance of missense mutations lead to the expression of
a stable, but conformationally aberrant, protein rather
than absence of the wild-type protein activity by
itself, thereby suggesting a possible gain of function
as a result of some of these mutations.18,19 Indeed, it
has been demonstrated that mutp53 (mutant p53) is
a cause of tumor progression, as expression of
mutp53 in tumor cells having a p53 null background
can lead to an increase in tumorigenic potential in

vitro and in vivo.20 There is an emerging awareness

that different mutp53 proteins may have distinctly different roles in the tumorigenic phenotype, suggesting
the necessity of specifically characterizing the predominant p53 mutations in any specific type of malignancy and even within a given tumor. Little is known
about the spectrum of p53 mutations in soft tissue sarcoma, and it is likely that identifying precise p53
mutations would lead to a better understanding and
even the possibility of predicting the molecular behavior of a given soft tissue sarcoma.
Consequently, in this study, we analyzed p53
mutations in a panel of soft tissue sarcomas by using
extensive (exons 2 to 11) gene sequencing, and we
compared our results with previously identified p53
mutations recorded in published human p53 mutation databases.2124 We also wanted to determine
whether p53 immunohistochemistry could accurately
identify soft tissue sarcomas harboring a p53 mutation, thereby aiding the selection of specimens for
further p53 sequencing or even eliminating the need
for sequencing altogether if all that is needed is an
indication of whether a mutation is present or absent
or if the p53 regulatory network is possibly dysregulated. Further analysis of specific soft tissue sarcoma
p53 mutational impacts may expand our understanding of mutp53 gain of function, which, in turn, may
be relevant to possible future soft tissue sarcoma individually tailored molecular therapeutic approaches.

MATERIALS AND METHODS

Patient and Tumor Data Accrual
This study was compliant with the Health Insurance
Portability and Accountability Act of 1996 and was
approved by the institutional review board at the
University of Texas M. D. Anderson Cancer Center,
which also granted a waiver of informed consent for
use of patients tissue samples. Patient data, followup information, and tumor parameters were retrieved from a well established, soft tissue sarcoma,
prospective, relational, clinical database that includes
more than 150 fields of information for each patient.
Tissue Retrieval
Thirty-one human soft tissue sarcomas were selected
from all frozen soft tissue sarcomas available in the
University of Texas M. D. Anderson Cancer Centers
Surgical Oncology Sarcoma Tissue Bank. The majority of these tumors were high grade; all were at least
intermediate grade, and all were stage III. Histology
was verified by using H & E-stained frozen sections
interpreted by a soft-tissue pathologist, and only
cases with >75% tumor cellularity (most cases were

p53 Mutations in Soft Tissue Sarcoma/Das et al.

2325

TABLE 1
Primer pairs and PCR Conditions for p53 Sequencing
Exon

50 Primer

30 Primer

Product size (BP)

Annealing temp (8C)

23
4
56
7
89
10
11

ATCCCCACTTTTCCTCTTGC
CGTTCTGGTAAGGACAAGGG
TCTTTGCTGCCGTCTTCC
TGCTTGCCACAGGTCTCC
GGACAGGTAGGACCTGATTTCC
TTACTTCTCCCCCTCCTCTGTTG
TCATCTCTCCTCCCTGCTTCTGTC

GAAAAGAGCAGTCAGAGGAC
ACAAAAGAAATGCAGGGGG
AGGGCCACTGACAACCAC
GTCAGAGGCAAGCAGAGGC
AAACAGTCAAGAAGAAAACGGC
GCTTTCCAACCTAGGAAGGCAG
TGCTTCTGACGCACACCTATTG

360
412
517
235
441
193
224

55
55
55
55
55
55
55

BP indicates base pair; temp, temperature.

>90% tumor) and minimal necrosis (<10%) were

selected for inclusion. The paraffin-embedded tumor
blocks were retrieved from immediately adjacent tumor sections and also verified by histology. For all
selected soft tissue sarcomas, frozen tumor, autologous normal tissue, and paraffin-embedded tumor
samples were available. Formalin-fixed, paraffinembedded, tissue sections were cut to 5 lm for
immunohistochemistry studies. Genomic DNA was
extracted from frozen tumor and autologous normal
tissue by using a QIAamp DNA Mini Kit (Qiagen
Sciences, Valencia, Calif) according to manufacturers
instructions. The integrity of the extracted DNA (usually much higher from frozen tissue than from paraffin-embedded tissue) was assessed by running the
genomic DNA on a 1% agarose gel and confirming
the presence of high molecular weight genomic DNA
by ethidium bromide before using the DNA for polymerase chain reaction (PCR).

Direct p53 Sequencing

DNA sequencing was performed as described by
Stockton et al.25 Primers were designed to intron
sequences flanking exons of the p53 gene (exons 2
through 11; Table 1). In brief, 100 ng of genomic
DNA was used as a template for PCR amplification of
exon and intronic boundaries. Each PCR reaction
included 1.5 mM MgCl2, 50 mM KCl, 10 mM Tris
(pH 8.3), 0.015% gelatin, 200 lm each of dATP, dTTP,
dCTP, and dGTP, 10% DMSO, and 5 units of Taq polymerase (Promega, Madison, Wis) in a final volume of
50 lL. PCR cycling was carried out on a Perkin Elmer
(Waltham, Mass) PE 9600 thermocycler using a denaturation cycle at 958C for 10 minutes followed by 40
cycles of denaturation at 948C for 1 minute, annealing at 558C for 1 minute and extension at 728C for 2
minutes, with a final extension step at 728C for 10
minutes. Direct sequencing of DNA amplicons was
performed on an Applied Biosystems (Foster City,
Calif) 373 automated DNA sequencer. Sequences

were analyzed using Chromas-Pro software (Technelysium; Tewantin, Queensland, Australia) and verified
against the sequence deposited in GenBank (USA
National Center for Biotechnology Information
[NCBI]).

Immunohistochemistry
Immunohistochemistry was performed as previously
described.10 Briefly, 5 lm formalin-fixed, paraffinimbedded, tissue sections were dewaxed and rehydrated through step-wise immersion in xylene and
standard ethanol dilutions to water. Sections were
microwaved in a 10 mM sodium citrate solution with
a pH of 6.0 for antigen retrieval. Endogenous peroxidase activity was quenched with 1% hydrogen peroxide in Tris-buffered saline (TBS)/calcium solution at
room temperature for 30 minutes before blocking
with 5% horse serum in 1% bovine serum albumin
(BSA), or Dako protein block, in phosphate-buffered
saline (PBS) overnight at 48C. Anti-p53: DO1 monoclonal (Santa Cruz Biotechnology Inc., Calif) was
applied in 1% BSA-PBS according to manufacturers
instructions and detected by using Dako Envision kit
(DakoCytomation, Carpinteria, Calif) with diamino
benzidine (DAB) development before counterstaining
with hematoxylin and performing step-wise dehydration. Simultaneously, the specimens were subjected
to p53 immunohistochemistry in the Pathology Clinical Laboratory of the University of Texas M. D.
Anderson Cancer Center according to protocol by
using the clone DO7 anti-p53 antibody (DakoCytomation, Carpinteria, Calif).
The DO1 monoclonal antibody maps to amino
acid residues 2025 in the N-terminus, whereas the
DO7 monoclonal antibody maps to an epitope
between amino acids 19 and 26, and, thus, these
2 antibodies have overlapping specificity. Antigen
retrieval is important for reliable results with DO7.
p53 immunohistochemistry expression was evaluated by 2 independent reviewers. The samples were

2326

CANCER

June 1, 2007 / Volume 109 / Number 11

evaluated for both the percentage of positive tumor

cells in a selected area of at least 5 high-power microscopic fields (identified as exhibiting the largest
proportion of positive-stained cells) and for nuclear
staining intensity; a score was assigned to each sample based on these 2 parameters.

STATISTICAL ANALYSIS
Statistical analysis was performed by using the Fisher
exact test; significance was set as P < .05. Cumulative
incidence was estimated by using the method of
Gooley et al26 with death as competing risk. The estimates of cumulative incidences by various risk factors were calculated by using a publicly available SAS
(SAS Institute; Cary, NC) macro created by Erik Bergstralh of the Mayo Clinic (Rochester, Minn). The correlation between p53 expression (as determined by
immunohistochemistry) and p53 mutational status
(determined by sequencing) was analyzed for sensitivity (probability of positive p53 immunohistochemistry staining when a p53 mutation was identified
by sequencing), specificity (probability of negative
p53 immunohistochemistry staining when p53 was
found to be wild-type by sequencing), and concordance (probability of exact agreement between the
2 methods).

RESULTS
Thirty-one stage III soft tissue sarcoma specimens
were included in the study (Table 2). Specimens
were collected from 20 men and 11 women with a
median age of 56 years. Median follow-up time was
78 months, during which 23 (74%) patients died of
their disease because of metastasis, unresectable
local recurrence, or both.

p53 Mutational Spectra in Soft Tissue Sarcoma; a High

Frequency of Exon 4 Mutations
Previously published reports have identified p53
mutations in 20% to 60% of soft tissue sarcomas.27,28
In this study, we sought to confirm this high incidence by using our archived specimens while further
expanding knowledge of p53 mutational spectra in
soft tissue sarcoma. Although previous studies have
evaluated the presence of mutations in the p53 corebinding domain, we chose to expand the search by
examining for the presence of p53 mutations
throughout the entire transcribed regions of the p53
gene. Consequently, we sequenced exons 2 to 11 of
the p53 gene encoding for the entire p53 protein.
p53 mutations were detected in 10 of 31 (32%) of the
soft tissue sarcoma specimens (Table 3; Fig 1). Addi-

TABLE 2
Patient and Tumor Characteristics
Characteristics
Sex
Males
Females
Median age, y (range)
Men
Women
Histological subtype
Synovial sarcoma(SS)
Liposarcoma(LS)
Malignant fibrous histiocytoma (MFH)
Rhabdomyosarcoma(RMS)
Leiomyosarcoma (LMS)
Fibrosarcoma (FS)
Tumor site
Upper extremity
Lower extremity
Head & neck
Visceral
Retroperitoneal
Pelvic
Superficial trunk
Tumor status
Primary (P)
Recurrent (R)

No.

20
11
56 (1876)
56 (2485)
7
3
11
4
5
1
7
10
1
3
4
5
1
15
16

tional sequencing of corresponding autologous normal tissues for these 10 identified p53-mutated soft
tissue sarcoma samples verified that these were true
mutations rather than polymorphisms.
Although the incidence of identified p53 mutations was concordant with previous reports, only 3 of
the mutations were in the classic exon 5 to 8 p53
core-binding domain. Instead, the mutations were
spread throughout the p53 gene, and it is of special
interest that 40% of the mutations were found in
residues 6396, corresponding to exon 4 of the p53
gene. Although p53 is, perhaps, the most extensively
investigated tumor-suppressor gene, the structure
and function of p53 exon 4 still remains obscure. To
the best of our knowledge, this is the first report to
identify p53 exon 4 mutations in soft tissue sarcoma.
Seven (70%) of these mutations were missense synonymous mutations, whereas 3 mutations were
either insertions or frameshift mutations resulting in
early truncation proteins. A search of literature,
established p53 databases,22 and the Universal Mutation Database23 confirmed that the point mutation at
codon 250 we observed in 2 specimens has been previously observed in soft tissue sarcoma.29 The point
mutation in codon 96, although not previously
reported in soft tissue sarcoma, has been previously
observed in bladder cancer.30 In contrast, the other 7
p53 mutations are apparently novel and have not

p53 Mutations in Soft Tissue Sarcoma/Das et al.

2327

TABLE 3
Mutation Type and Location in STS Samples. IHC Status Corresponding to Each Mutation Is Shown in Comparison
Pat
ID

Tumor histological
subtype

Tumor
status

Exon

Codon

Nucleotide

Change

Amino acid
substitution

IHC

561
121
344
390
474
583
537
500
244
508

SS
SS
SS
SS
SS
FS
LS
MFH
MFH
RMS

P
P
P
R
P
P
R
R
P
R

2
7
7
4
4
4
11
4
6
11

8
250
250
96
63
75
377
77
213
393

144
868
868
407
307
344
1249
350
759
1297

C in at 144*
CCC >ACCy
CCC>ACCy
TCT>TGT{
GCT>CCT*
CCT>CAT*
T insertion at nt.1249*
CCA>CAA*
CGA>CGG(Insertion)*
GAC>AAC*

Truncated protein at codon 9

P250T
P250T
S96C
A63P
P75H
T377Y and truncated at 381 codon
P77Q
Frameshift at 213 and truncation at 215
D393N

Negative
Negative
Weak
Negative
Moderate
Weak
Weak
Negative
Strong
Moderate

Pat ID indicates patient identification code; IHC, immunohistochemistry; SS, synovial sarcoma; FS, fibrosarcoma; LS, liposarcoma; MFH, malignant fibrous histiocytoma; RMS, rhabdomyosarcoma; P, primary;
R, recurrent.
* Novel mutations not reported in p53 mutation databases.
y
Reference 30.
{ Reference 31.

FIGURE 1. (A) Chromatographs demonstrating p53 mutations were identified by direct sequencing in 3 selected soft tissue sarcoma samples. (B) p53 mutation spectra in soft tissue sarcoma. The entire transcribed p53 gene is depicted (exons 211). Thin arrows above represent the mutations found in our study
compared with common hotspot p53 mutations represented by arrows below. The functional protein domains of transactivation, DNA binding, and tetramerization are depicted in blue, red, and green, respectively. Codon boundary numbers for the exons are directly below the figure.

2328

CANCER

June 1, 2007 / Volume 109 / Number 11

TABLE 4
Tumor Characteristics and Patient Outcome Stratification According
to p53 Mutational Status
p53 status by sequencing

Histology
SS (n7)
MFH (n11)
RMS (n4)
LS (n3)
LMS (n5)
FS (n1)
Site
Extremity (n17)
Nonextremity (n14)
Tumor status
Primary (n15)
Recurrent (n16)
Survival status
Dead
Alive
Median survival, mo

wt (n = 21)

Mut (n = 10)

2
9
3
2
4
0

5
2
1
1
1
1

11
10

6
4

9
12

6
4

15
6
52

8
2
33

wt indicates wild type; Mut, mutated; SS, synovial sarcoma; MFH, malignant fibrous histiocytoma;
RMS, rhabdomyosarcoma; LS, liposarcoma; LMS, leiomyosarcoma; FS, fibrosarcoma; mo, months.

been previously reported in soft tissue sarcoma or

any other malignancy.
Patient age and sex did not correlate with the
probability of harboring a p53 mutation, neither did
the soft tissue sarcoma site nor whether the soft tissue sarcoma was primary or recurrent (Table 4). A
high rate of mutation was seen in synovial sarcomas
(71%) compared with all other histological subtypes
(25%; P < .05). The median overall survival for patients
with and without p53 mutation (33 vs 52 months)
did not reach statistical significance (P .1), possibly
because of the small patient sample size. Similarly, a
trend toward a shorter time interval to local recurrence as well as increased incidence of lung metastases were observed in soft tissue sarcoma patients
whose tumors bore mutp53. No difference in survival
could be identified for patients harboring a p53 corebinding domain mutation compared with other
mutations. Similarly, patients harboring a p53 exon 4
mutation exhibited a trend toward decreased survival
compared with patients having wtp53 tumors.

Positive Staining for p53 Correlates With Survival in Soft

Tissue Sarcoma
Immunostaining for p53 was performed both in the
research laboratory as described by using DO1 antip53 antibody and also in the University of Texas M.
D. Anderson Cancer Center clinical laboratory core

facility by using DO7 anti-p53 antibody. Four distinct

staining patterns were identified for both antibodies
(Fig. 2A-D) including, 1) negative: no positive staining or <5% positive staining in nuclei of tumor cells;
2) weak: nuclear staining in <20% but >5% of tumor
cells; 3) moderate: variable intensity in >20% but
<50% of the tumor cells; 4) strong: intense nuclear
staining in the majority of cells. Staining with both
antibodies yielded comparable results: 13 (42%) specimens were negative for p53 protein expression,
whereas 18 (58%) of samples were positive (Table 5).
Of the specimens exhibiting p53 immunopositivity, 5
(16%) were strong, 6 (19%) were moderate, and 7
(23%) were scored as weak expressers. No correlation
could be found between p53 immunopositivity and
patient age, sex, site of tumor, or whether the tumor
was primary versus recurrent. The malignant fibrous
histiocytoma histological subtype exhibited the highest level of p53 immunopositivity when compared
with all other histologic subtypes combined. Interestingly, in biphasic synovial sarcoma, p53 staining
could be observed only in the more epithelial or
glandular component of the tumor (Fig. 2E). Despite
these caveats, overall, a clear inverse correlation
could be demonstrated between p53 immunopositivity and survival. All 8 patients still alive at the end of
the follow-up period exhibited negative p53 staining;
in contrast, none of the patients bearing soft tissue
sarcoma expressing p53 were alive at the end of the
study period (Kaplan-Meier, P < .001; Table 3). The
estimated median survival for soft tissue sarcoma
patients that were p53 immunohistochemistry negative was 84 months compared with 22 months for
soft tissue sarcoma patients who were p53 immunohistochemistry strong expressers (P < .05). Time to
either local recurrence or metastasis development
according to cumulative incidence analysis was
found to be shorter for patients who were immunohistochemistry p53 strong expressers (Fig. 3).

p53 Staining Does Not Correlate With p53 Mutational

Status in Soft Tissue Sarcoma
Next we wanted to evaluate the accuracy of soft tissue sarcoma immunohistochemistry p53 staining in
predicting the presence of a p53 mutation identified
by direct sequencing (Table 6). Strong immunohistochemistry p53 staining identified only 1 of 10 sequencing-identified mutated specimens (sensitivity,
10%). If any level of immunohistochemical p53 staining was defined as positive, the sensitivity increased
to 60%, in that 6 of the 10 mutated specimens were
so identified. When only immunohistochemically
strong staining was considered positive, the specificity

p53 Mutations in Soft Tissue Sarcoma/Das et al.

2329

FIGURE 2. p53 immunohistochemistry staining score. Stained sections representing the 4 distinct staining patterns identified, (A) negative, (B) weak, (C) moderate, (D) strong. (E) A biphasic synovial sarcoma demonstrates high expression of p53 in the epithelioid components of the tumor (magnification, 3200).

2330

CANCER

June 1, 2007 / Volume 109 / Number 11

TABLE 5
Tumor Characteristics and Patient Outcome Stratification According
to p53 IHC Staining Score
IHC p53 status

Histology
SS (n7)
MFH (n11)
RMS (n4)
LS (n3)
LMS (n5)
FS (n1)
Site
Extremity (n17)
Nonextremity (n14)
Tumor status
Primary (n15)
Recurrent (n16)
Survival status
Dead
Alive
Median survival, mo

Negative
(n = 13)

Weak
(n = 7)

Moderate
(n = 6)

Strong
(n = 5)

3
7
1
1
1
0

1
0
2
1
2
1

3
0
1
1
1
0

0
4
0
0
1
0

7
6

4
3

3
3

3
2

7
6

2
5

3
3

3
2

5
8
84

7
0
38

6
0
26

5
0
22*

IHC indicates immunohistochemistry; SS, synovial sarcoma; MFH, malignant fibrous histiocytoma;
RMS, rhabdomyosarcoma; LS, liposarcoma; LMS, leiomyosarcoma; FS, fibrosarcoma; mo, months.
* P < .05

of the assay was 81%; ie, 17 of the 21 sequenced wtp53

were identified as negative. However when any immunohistochemical staining was considered positive,
the specificity decreased to 42%, identifying only 9 of
the 21 wtp53 sequenced specimens. Of the 5 specimens strongly stained by immunohistochemistry for
p53, 4 harbored a wtp53 according to sequencing,
and of the 18 specimens exhibiting any level of immunohistochemical p53 staining, 12 harbored wtp53
according to sequencing. Of the 13 samples immunohistochemically stained negative for p53, 4 harbored a p53 mutation. Thus, a significant level of
discordance could be demonstrated between the
2 assays: there were 18 concordant cases when only
strong staining was considered positive staining
(58%), and only 15 (48%) when any level of staining
was considered positive.

FIGURE 3. Increased p53 protein expression in soft tissue sarcoma inversely correlates with time to local recurrence (A) and time to metastasis (B)
as demonstrated by cumulative incidence curves.

TABLE 6
IHC Staining Score Correlation With p53 Mutational Status According
to Direct Sequencing
IHC*

Negative
(n = 13)

Weak
(n = 7)

Moderate
(n = 6)

Strong
(n = 5)

Sequencing
wtp53 (n 21)
Mutp53 (n 10)

9
4

4
3

4
2

4
1

DISCUSSION
A search of PubMed (USA National Library of Medicine searchable journal database) using the keyword
p53 revealed more than 40,000 publications referring to this gene and its protein product that have
been published since its first description just 3 decades ago, rendering p53 the most studied molecule
in the context of cancer, yet it is still not fully understood. More than 10,000 p53 mutations have been

IHC indicates immunohistochemistry; wtp53, wild type p53; Mutp53, mutated p53; n, number.
* IHC sensitivity 60%; IHC specificity 42%; IHC to sequencing concordance rate 48% (when
any degree of staining is considered positive).

recorded in various tumor types including soft tissue

sarcoma, where p53 mutations are thought to occur
in 20% to 60% of cases and may impact on patient
outcome. Taking into account this relatively high

p53 Mutations in Soft Tissue Sarcoma/Das et al.

incidence of p53 mutations in soft tissue sarcoma, it

is pertinent to further analyze the mutational load,
spectra, and types within soft tissue sarcoma in the
hope that this could lead to better understanding of
the molecular determinants driving this disease.
Direct sequencing of p53 is a valuable tool in
gauging the mutational load in a given tumor sample.31 The advantages of direct sequencing include
the ability to detect the exact type and location of
mutation, to identify the presence of more than 1
mutation, and to determine the existence of single
nucleotide polymorphisms (SNP). In sequencing 31
stage III soft tissue sarcoma specimens, we identified
10 (32%) that harbored p53 mutations, consistent
with the incidence of p53 mutation in soft tissue sarcoma as described by others.8,32 Moreover, none of
these have been previously defined as a hotspot p53
mutation. Although 2 of the 10 mutations have been
previously identified, 1 in liposarcoma29 and the
other in bladder cancer,30 the other 7 mutations were
novel, have not been previously recorded in established p53 mutation databases, and thus merit further investigation for possible relevance in soft tissue
sarcoma tumorigenesis and progression.
Unlike other soft tissue sarcoma studies, we
sequenced the entire protein-encoding portions of
the gene; only 3 of the mutations were located in the
classic core-binding domain of p53. Of special interest was the high rate of mutations in exon 4 of the
p53 gene. Study of the p53 protein has focused
mainly on 4 domains;3335 1) a transcriptional activation domain that contacts TAF components of TFIID
(residues 1 to 40); 2) a sequence-specific DNA binding domain (residues 120 to 290); 3) a tetramerization domain (residues 310 to 360); and 4) a domain
that recognizes and binds to damaged DNA nonspecifically (residues 364 to 390). A fifth functional
domain, localized between residues 61 to 96, containing a putative proline-rich signaling domain, has
not been the focus of intense study. Deletion of this
domain from the p53 protein leaves a normal p53
protein regarding transcriptional transactivation
capacity.36 Therefore, despite its physical juxtaposition between the transcriptional activation and DNA
binding domains, the proline-rich region of p53 is
not essential to the function of these domains. However, this domain has been found likely to participate
in the transmission of p53 nontranscriptional, antiproliferative, proapoptotic activities.36 The high rate
of exon 4 p53 mutations found in our study suggest
that this may be a mutation-prone region in soft tissue sarcoma, especially in synovial sarcoma, an observation that needs validation in a larger cohort of
tumors. If larger-scale studies confirm this hypothe-

2331

sis, then more emphasis could justifiably be placed

on investigating the structure and function of this
region.
Under normal conditions, wtp53 is normally a
very short-lived protein, with a half-life of less than
30 minutes; consequently, it can barely be detected
in normal cells.16 However, under conditions of DNA
damage and other stresses, the wtp53 level increases
rapidly because of stabilization rather than increased
synthesis. In contrast, the majority of mutated p53
protein product in tumor cells typically has a much
longer half-life (more than 24 hours) and can be
detected in cells despite the presence of environmental stimuli.37 Using p53 immunohistochemistry as a
tool to predict the presence of p53 mutation in lieu
of submitting the specimen for expensive singlestrand conformational polymorphism testing or
direct sequencing is based on the above observations
and assumes that only stable (therefore mutated)
protein will be expressed and subsequently observed
upon staining. However, there is no agreement in
published reports on whether or not both methods
(immunohistochemistry and direct sequencing) actually yield similar results. Whereas some publications
report a high concordance between immunohistochemistry and sequencing,38 others, reminiscent of
the present study, identify a high rate of discordance.39,40 In the present study, we used 2 different
anti-p53 N-terminal antibodies (clone DO1 and DO7)
for p53 immunohistochemistry, conducted by 2 different operators, yielding comparable results, thereby
further reducing the possibility that technical discrepancies were the root cause for the discordance
between immunohistochemical and sequencing
results.
Four of our samples harboring a p53 mutation
were found to be negative upon immunostaining.
This was expected in 1 case (ID 561) that harbored
an inactivating mutation leading to a very short truncated protein that was undetectable by the antibodies used.41 However, the other 3 cases harbored
missense mutations, and it could be anticipated that
the protein would have been detectable by immunohistochemistry. It is possible that, in these latter
cases, the mutated protein had not been stabilized42
or that the mutant protein acquired an irreversible
conformational change, resulting in failure to expose
the antigen residue site for recognition by the antibody and, thereby, rendering sequencing as a more
sensitive method for identifying the existence of
mutations, if that information is of importance. Conversely, in our study, 12 soft tissue sarcoma specimens harboring wtp53 stained positive to various
degrees. One explanation is that this overexpression

2332

CANCER

June 1, 2007 / Volume 109 / Number 11

signifies increased environmental stress signals leading to stabilization of wtp53. However, in that situation, we would probably expect to observe increased
tumor apoptosis and a relatively less aggressive clinical behavior, which was not the case. A second
option is that down-stream alterations in the p53
pathway could lead to overexpression of p53 without
mutation. In such cases, sequencing will not be of
much identifying help, whereas immunohistochemistry may serve as a better indicator of a functionally
abnormal pathway leading to a more aggressively
dysregulated tumor.
Even though all these hypotheses are possible,
there is another explanation for the discrepancy
between sequencing and p53 immunohistochemistry
that may be unique to soft tissue sarcoma; soft tissue
sarcoma can commonly attain large size (as was the
case in all the cases evaluated in the present study)
and are known to be highly heterogeneous; it is possible that results of the 2 assays represent 2 different
fields of the tumor. Thus, the part of the tumor used
for DNA collection harbors a p53 mutation while the
segment used for immunohistochemistry does not,
and vice versa. Although it is impossible to assay an
entire soft tissue sarcoma specimen, analyzing p53
alterations by using both sequencing and immunohistochemistry may reduce the chance for error. In
addition, different sections of the tumor should be
studied.
Whereas our data cannot support replacing p53
sequencing by immunohistochemistry for the identification of p53 mutations in soft tissue sarcoma, it
does suggest the potential benefit of using p53
immunohistochemistry as a prognostic marker in
stage III soft tissue sarcoma. p53 immunopositivity
has been previously shown to correlate with soft tissue sarcoma outcome.7,8 Current soft tissue sarcoma
prognostic staging systems are based on crude clinical and tumor variables such as size, grade, and the
presence or absence of overt regional nodal or distant visceral metastasis. There is a growing awareness that more sensitive prognostication and
therapeutic decision-making algorithms will need to
incorporate relevant molecular determinants. In light
of the findings reported above, it may be particularly
useful to prospectively investigate the role of p53
expression as potentially integral to the soft tissue
sarcoma staging system.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.
17.

18.

REFERENCES
1.

Pisters PW, Leung DH, Woodruff J, Shi W, Brennan MF.

Analysis of prognostic factors in 1,041 patients with localized soft tissue sarcomas of the extremities. J Clin Oncol.
1996;14:16791689.

19.

Weitz J, Antonescu CR, Brennan MF. Localized extremity

soft tissue sarcoma: improved knowledge with unchanged
survival over time. J Clin Oncol. 2003;21:27192725.
Billingsley KG, Burt ME, Jara E, et al. Pulmonary metastases from soft tissue sarcoma: analysis of patterns of diseases and postmetastasis survival. Ann Surg. 1999;229:602
610; discussion 610612.
Billingsley KG, Lewis JJ, Leung DH, Casper ES, Woodruff
JM, Brennan MF. Multifactorial analysis of the survival of
patients with distant metastasis arising from primary extremity sarcoma. Cancer. 1999;85:389395.
Latres E, Drobnjak M, Pollack D, et al. Chromosome 17
abnormalities and TP53 mutations in adult soft tissue sarcomas. Am J Pathol. 1994;145:345355.
Taubert H, Meye A, Wurl P. Prognosis is correlated with
p53 mutation type for soft tissue sarcoma patients. Cancer
Res. 1996;56:41344136.
Kawai A, Noguchi M, Beppu Y, et al. Nuclear immunoreaction of p53 protein in soft tissue sarcomas. A possible
prognostic factor. Cancer. 1994;73:24992505.
Schneider-Stock R, Onnasch D, Haeckel C, Mellin W,
Franke DS, Roessner A. Prognostic significance of p53 gene
mutations and p53 protein expression in synovial sarcomas. Virchows Arch. 1999;435:407412.
Pollock RE, Lang A, Luo J, El-Naggar AK, Yu D. Soft tissue
sarcoma metastasis from clonal expansion of p53 mutated
tumor cells. Oncogene. 1996;12:20352039.
Milas M, Yu D, Lang A, et al. Adenovirus-mediated p53
gene therapy inhibits human sarcoma tumorigenicity.
Cancer Gene Ther. 2000;7:422429.
Pollock R, Lang A, Ge T, Sun D, Tan M, Yu D. Wild-type
p53 and a p53 temperature-sensitive mutant suppress
human soft tissue sarcoma by enhancing cell cycle control.
Clin Cancer Res. 1998;4:19851994.
Zhang L, Yu D, Hu M, et al. Wild-type p53 suppresses
angiogenesis in human leiomyosarcoma and synovial sarcoma by transcriptional suppression of vascular endothelial growth factor expression. Cancer Res. 2000;60:3655
3661.
Liu J, Zhan M, Hannay JA, et al. Wild-type p53 inhibits nuclear factor-kappaB-induced matrix metalloproteinase-9
promoter activation: implications for soft tissue sarcoma
growth and metastasis. Mol Cancer Res. 2006;4:803810.
Zhan M, Yu D, Lang A, Li L, Pollock RE. Wild type p53 sensitizes soft tissue sarcoma cells to doxorubicin by downregulating multidrug resistance-1 expression. Cancer. 2001;
92:15561566.
Zhan M, Yu D, Liu J, Glazer RI, Hannay J, Pollock RE.
Transcriptional repression of protein kinase Calpha via Sp1
by wild type p53 is involved in inhibition of multidrug resistance 1 P-glycoprotein phosphorylation. J Biol Chem.
2005;280:48254833. Erratum in: J Biol Chem. 2005;280:
22555.
Levine AJ. p53, the cellular gatekeeper for growth and division. Cell. 1997;88:323331.
Martin AC, Facchiano AM, Cuff AL, et al. Integrating mutation data and structural analysis of the TP53 tumor-suppressor protein. Hum Mutat. 2002;19:149164.
Di Agostino S, Strano S, Emiliozzi V, et al. Gain of function
of mutant p53: the mutant p53/NF-Y protein complex
reveals an aberrant transcriptional mechanism of cell cycle
regulation. Cancer Cell. 2006;10:191202.
Peart MJ, Prives C. Mutant p53 gain of function: the NF-Y
connection. Cancer Cell. 2006;10:173174.

p53 Mutations in Soft Tissue Sarcoma/Das et al.

20. Sigal A, Rotter V. Oncogenic mutations of the p53 tumor
suppressor: The demons of the guardian of the genome.
Cancer Res. 2000;60:67886793.
21. Beroud C, Soussi T. p53 gene mutation: Software and database. Nucleic Acids Res. 1998;26:200204.
22. Cariello NF, Beroud C, Soussi T. Database and software for
the analysis of mutations at the human p53 gene. Nucleic
Acids Res. 1994;22:35493550.
23. Hainaut P, Hernandez T, Robinson A, et al. IARC database
of p53 gene mutations in human tumors and cell lines:
updated compilation, revised formats and new visualisation tools. Nucleic Acids Res. 1998;26:205213.
24. Hamroun D, Kato S, Ishioka C, Claustres M, Beroud C,
Soussi T. The UMD TP53 database and website: update
and revisions. Hum Mutat. 2006;27:1420.
25. Stockton DW, Das P, Goldenberg M, DSouza RN, Patel PI.
Mutation of PAX9 is associated with oligodontia. Nat Genet.
2000;24:1819.
26. Gooley TA, Leisenring W, Crowley J, Storer BE. Estimation
of failure probabilities in the presence of competing risks:
new representations of old estimators. Stat Med. 1999;18:
695706.
27. Andreassen A, Oyjord T, Hovig E, et al. p53 abnormalities
in different subtypes of human sarcomas. Cancer Res.
1993;53:468471.
28. Castresana JS, Rubio MP, Gomez L, Kreicbergs A, Zetterberg A, Barrios C. Detection of TP53 gene mutations in
human sarcomas. Eur J Cancer. 1995;31A:735738.
29. Schneider-Stock R, Haeckel C, Radig K, Pein CD, Roessner
A. Improved detection of p53 mutations in soft tissue
tumors using new gel composition for automated nonradioactive analysis of single-strand conformation polymorphism. Electrophoresis. 1997;18:28492851.
30. Prescott JL, Montie J, Pugh TW, McHugh T, Veltri RW. Clinical sensitivity of p53 mutation detection in matched bladder tumor, bladder wash, and voided urine specimens.
Cancer. 2001;91:21272135.
31. Casey G, Lopez ME, Ramos JC, et al. DNA sequence analysis of exons 2 through 11 and immunohistochemical stain-

32.

33.

34.

35.

36.

37.
38.

39.

40.

41.

42.

2333

ing are required to detect all known p53 alterations in

human malignancies. Oncogene. 1996;13:19711981.
Yoo J, Lee HK, Kang CS, Park WS, Lee JY, Shim SI. p53
gene mutations and p53 protein expression in human soft
tissue sarcomas. Arch Pathol Lab Med. 1997;121:395399.
Kitayner M, Rozenberg H, Kessler N, et al. Structural basis
of DNA recognition by p53 tetramers. Mol Cell. 2006;22:
741753.
Okorokov AL, Sherman MB, Plisson C, et al. The structure
of p53 tumour suppressor protein reveals the basis for its
functional plasticity. EMBO J. 2006;25:51915200.
Wang P, Reed M, Wang Y, et al. p53 domains: structure, oligomerization, and transformation. Mol Cell Biol. 1994;14:
51825191.
Walker KK, Levine AJ. Identification of a novel p53
functional domain that is necessary for efficient growth
suppression. Proc Natl Acad Sci USA. 1996;93:15335
15340.
Vogelstein B, Kinzler KW. p53 function and dysfunction.
Cell. 1992;70:523526.
Kinoshita Y, Shiratsuchi H, Tamiya S, et al. Mutations of
the p53 gene in malignant rhabdoid tumors of soft tissue
and the kidney: immunohistochemical and DNA direct
sequencing analysis. J Cancer Res Clin Oncol. 2001;127:
351358.
Schneider-Stock R, Radig K, Oda Y, et al. p53 gene mutations in soft-tissue sarcomascorrelations with p53 immunohistochemistry and DNA ploidy. J Cancer Res Clin Oncol.
1997;123:211218.
Mousses S, McAuley L, Bell RS, Kandel R, Andrulis IL. Molecular and immunohistochemical identification of p53
alterations in bone and soft tissue sarcomas. Mod Pathol.
1996;9:16.
Stephen CW, Helminen P, Lane DP. Characterisation of epitopes on human p53 using phage-displayed peptide
libraries: Insights into antibody-peptide interactions. J Mol
Biol. 1995;248:5878.
Wynford-Thomas D. P53 in tumour pathology: can we trust
immunocytochemistry? J Pathol. 1992;166:329330.