Full text of DOI:10.1038/labinvest.3780437

0023-6837/02/8204-443$03.00/0
LABORATORY INVESTIGATION
Vol. 82, No. 4, p. 443, 2002
Copyright © 2002 by The United States and Canadian Academy of Pathology, Inc.
Printed in U.S.A.
Toward Efficient Analysis of Mutations in Single Cells
from Ethanol-Fixed, Paraffin-Embedded, and
Immunohistochemically Stained Tissues
Ernst Heinmöller, Qiang Liu, Yuan Sun, Gudrun Schlake, Kathleen A. Hill,
Larry M. Weiss, and Steve S. Sommer
Department of Molecular Genetics (EH, QL, GS, KAH, SSS), City of Hope National Medical Center/Beckman
Research Center; and Department of Pathology (YS, LMW), City of Hope National Medical Center, Duarte, California
SUMMARY: Only a few studies have demonstrated successful molecular analysis after whole genome amplification using single
cells dissected from paraffin-embedded tissues. The results in these studies were limited by low-amplification efficiency and high
rates of allele dropout. In the present study, the amplification rate using a thoroughly modified primer extension and
preamplification-PCR protocol was improved significantly for single cells microdissected from paraffin-embedded and immu-
nohistochemically stained tissues. Tissue fixation with ethanol (85%) and the addition of 0.2 mmol/L EDTA helped to achieve an
amplification rate between 67% (segments 200 to 400 bp) and 72% (segments Ͻ200 bp). Normal tissue sections were
immunohistochemically double stained for overabundance of p53 protein and proliferating cell nuclear antigen. Microdissection
of single cells was performed with a manual micromanipulator equipped with a Tungsten needle. Sequence analysis of the TP53
gene was performed after improved primer extension preamplification-PCR and multiplex PCR from single microdissected cells.
The rate of allele dropout was at least 68%. These technical advances facilitate routine mutation analysis using a single cell or
a few cells microdissected from routinely processed paraffin-embedded normal and tumor tissues. Allele dropout still represents
a serious problem in single-cell mutation analysis, especially in samples with limited template DNA and prone to DNA damage.
(Lab Invest 2002, 82:443–453).
olecular analysis of single cells has become
routine in reproductive medicine (Hahn et al,
gous mutations from single cells. No protocol has
been identified that eliminates ADO (Hahn et al, 2000).
Of great interest is the development of methodologies
that permit examination of rare heterozygous muta-
tions in single cells from normal tissues with high
sensitivity and accuracy to enable investigation of the
role of mutagenesis and risk of disease.
M
2000). However, few studies have routinely analyzed
rare heterozygous mutations in single cells microdis-
sected from fixed, paraffin-embedded, and immuno-
histochemically stained tissue sections because of
certain limitations (Becker et al, 1996; Roehrl et al,
1997; Schutze and Lahr, 1998; Suarez-Quian et al,
1999). Crosslinking of DNA in tissues fixed with the
commonly used fixative, formalin, limits the efficiency
of PCR amplification and reduces the maximum length
of the PCR amplicon (Karlsen et al, 1994; Williams et
al, 1999). Allele dropout (ADO) occurs when one allele
is preferentially amplified over the other to the extent
that the minor allele is not detected. ADO is particu-
larly problematic when templates consisting of DNA
from less than 10 cells are amplified; it occurs at the
same rate irrespective of whether nested PCR or
whole genome amplification (WGA) is used for initial
amplification (Hahn et al, 1998, 2000). ADO is espe-
cially problematic in the diagnosis of rare heterozy-
Immunohistochemistry is a technique that specifi-
cally provides information about protein expression
patterns at the single-cell level and may indicate
molecular changes relevant to neoplastic develop-
ment. The tumor suppressor gene TP53 is mutated in
50% of human cancers (Greenblatt et al, 1994). In its
mutated form or under certain cellular and/or geno-
toxic stress, the p53 protein becomes stabilized and
its overabundance is readily detectable by immuno-
histochemical staining (May and May, 1999; Wallace-
Brodeur and Lowe, 1999). Overabundance of p53
protein may be found even in normal tissues in the
nuclei of cells with a morphologically normal appear-
ance (Barnes et al, 1992; Huusko et al, 1999; Varley et
al, 1999). Immunohistochemical staining for p53 can
be used to identify cells that are potential candidates
for TP53 gene mutations. Because proliferating cell
nuclear antigen (PCNA) expression is regulated by
p53, double immunostaining for overexpression of
p53 and PCNA proteins was used to enhance detec-
tion of cells with overexpression of p53 caused by a
mutation in the p53 gene of functional relevance.
Received November 8, 2001.
Supported in part by the Deutsche Krebshilfe (to EH). Current address for
Dr. Heinmöller: Institute of Pathology, Klinikum Kassel, Möncheberg-
strasse 41-42, 34125 Kassel, Germany.
Address reprint requests to: Dr. Steve S. Sommer, Chair, Department of
Molecular Genetics, Beckman Research Institute/City of Hope, 1450 East
Duarte Road, Duarte, CA 91010-0269. E-mail: sommerlab@coh.org
Laboratory Investigation • April 2002 • Volume 82 • Number 4 443

Heinmöller et al
In general, for molecular analyses from single cells,
two different PCR techniques can be used: direct
(usually nested) amplification or (nonspecific) pream-
plification of the genetic material by WGA followed by
specific (nested) PCR. A major advantage of WGA is
the possibility of multiple genetic analyses from a
single cell. Of the various WGA techniques available,
primer extension and preamplification (PEP)-PCR and
degenerate oligonucleotide-primed PCR provide the
most complete coverage of the genome (Wells et al,
1999). Improved primer extension and preamplifica-
tion (I-PEP)-PCR is superior to degenerate
oligonucleotide-primed PCR in amplification efficiency
(Dietmaier et al, 1999). PEP-PCR is a technique of
WGA used before specific PCR from minute amounts
of DNA, which enables the molecular analysis of
multiple genes from a single cell or a few cells (Zhang
et al, 1992). PEP-PCR has recently been improved to
facilitate multiple genetic analyses of small cell clus-
ters microdissected from formalin-fixed paraffin-
embedded tissues (Dietmaier et al, 1999). Although
some studies have analyzed DNA from immunohisto-
chemically stained single cells from fresh-frozen tissue
samples (Persson et al, 2000; Ponten et al, 1997), to
our knowledge, no study has conducted WGA before
specific amplification of single-cell genomes from
fixed, paraffin-embedded, and immunohistochemi-
cally stained tissue sections. Herein, DNA integrity,
amplification efficiency, and ADO are evaluated after
an overall enhanced protocol for microdissection of
single cells from ethanol-fixed, paraffin-embedded,
and immunohistochemically stained tissue sections
and amplification with a modified I-PEP-PCR protocol.
In this study major improvements to tissue fixation
and processing and single-cell DNA amplification
were made to enable the analysis of rare heterozygous
mutations in the TP53 gene of single cells microdis-
sected from ethanol-fixed, paraffin-embedded, and
immunohistochemically stained sections of normal
human tissue (Fig. 1).
Results
Tissue Fixation and Immunohistochemistry
To avoid DNA crosslinking by formalin fixation of
tissue, which limits molecular analysis of single cells,
Figure 1.
A flow chart of the methods used for mutation analysis using ethanol-fixed, paraffin-embedded, and immunohistochemically stained single cells or small cell clusters.
444 Laboratory Investigation • April 2002 • Volume 82 • Number 4

Mutations in Microdissected Single Cells
ethanol/EDTA fixation was optimized to maintain high
DNA integrity and provide high-resolution immunohis-
tochemistry. Ethanol has a precipitating effect on
tissues. EDTA inhibits DNases present in surgically
removed tissues for better preservation of DNA. Pro-
teinase K digestion of various tissues fixed with etha-
nol/EDTA (“Materials and Methods”) revealed the
presence of single-stranded, high molecular weight
DNA (~20 Kb) after the final step of immunohisto-
chemical staining in all tissues analyzed (data not
shown). To demonstrate the feasibility of ethanol/
EDTA tissue fixation for immunohistochemistry, dou-
ble staining of p53 protein (red-colored nuclear stain)
together with PCNA (brown-colored nuclear stain) was
used. A mix of 85% ethanol/0.2 mmol/L EDTA was
found to give crisp and clear immunohistochemistry
after a brief antigen retrieval by steam treatment at
95° C for 5 minutes, enabling the clear distinction of
nuclear staining for p53 protein, or PCNA, or both p53
and PCNA in various tissues with physiologically dif-
ferent expression patterns of both proteins (Fig. 2).
Fixation with 95% ethanol/0.1 mmol/L EDTA or 75%
ethanol/1 mmol/L ETDA resulted in less clear staining
(not shown). In addition, cutting of sections was im-
paired when 95% ethanol/0.1 mmol/L EDTA was used
as fixative.
Optimization of I-PEP-PCR: Analysis of Dilutions of
Genomic DNA
Preliminary experiments indicated that the amplifica-
tion success rate of various exons of the TP53 gene
from single cells microdissected from ethanol/EDTA-
fixed tissues was 30% or less when the I-PEP tech-
nique originally described by Dietmaier et al (1999)
was used. Such a low amplification success rate from
single cells would make the routine analysis of genes
like TP53 costly and cumbersome, because exons 5 to
9, which are commonly examined, would be amplified
in only a small number of microdissected single cells.
To optimize I-PEP, genomic DNA was serially diluted
to the single-cell level (6 pg). To ensure optimal
digestion, the concentration of Tween-20 in the diges-
tion buffer was increased from 0.5% to 3%. To
achieve maximum amplification efficiency, various
PCR cycling conditions (annealing temperature, ex-
tension time) and I-PEP-PCR mix ingredients (concen-
tration of primers, MgCl2, dNTP, enzymes) were
tested. The concentrations of MgCl2 and dNTP were
increased stepwise from 2.5 mmol/L and 0.1 mmol/L
to 6.0 mmol/L and 1 mmol/L, respectively. Specific
PCR after I-PEP-PCR was most effective when MgCl2
was 6.0 mmol/L and dNTPs were 1 mmol/L each in
I-PEP-PCR. Of the primer concentrations tested (5
␮mol/L, 10 ␮mol/L, 15 ␮mol/L, 20 ␮mol/L, 25 ␮mol/L,
30 ␮mol/L), 20 ␮mol/L was found to be optimal. Of the
various enzymes tested (Taq Expand High Fidelity
Polymerase with and without Bst DNA polymerase
[Roche, Basel, Switzerland], Pfu Turbo [Stratagene, La
Jolla, California], and AdvanTaq [Clontech Laborato-
ries, Palo Alto, California]), Taq Expand High Fidelity
Polymerase without Bst DNA polymerase was found
Figure 2.
Immunohistochemical staining for p53 and PCNA protein (overabundance) in
nucleus and cytoplasm, respectively, for normal human tissues that have been
fixed in ethanol and embedded in paraffin. A, Normal colon tissue with crypt
epithelial cells staining positive for PCNA (brown cytoplasmic stain) and a
single cell staining positive (red nuclear stain) for p53 overabundance (arrow)
(magnification ϫ100). B, Normal lung tissue with bronchial epithelial cells. A
cell staining for PCNA (brown cytoplasmic stain, white arrow) can be seen in
addition to a cell staining for p53 overabundance (red nuclear stain, black
arrow) (magnification ϫ40). C, Benign proliferative lesion of the mammary
gland. A cell staining red for p53 protein overabundance (white arrow) is seen
together with a cell staining mixed red-brown for both p53 and PCNA (black
arrow) proteins and many other cells staining brown for PCNA overabundance
(magnification ϫ250).
to be superior. No amplification was seen when Pfu
Turbo was used in I-PEP (not shown). Lowering the
annealing temperature from 37° C to 28° C, reducing
the annealing time from 4 minutes to 2 minutes, and
prolonging the extension time from 30 seconds to 3
minutes enabled the amplification of exons 5 to 9 of
Laboratory Investigation • April 2002 • Volume 82 • Number 4 445

Heinmöller et al
the TP53 gene after I-PEP-PCR. A nested amplifica-
tion of ~2 Kb from high molecular weight genomic
DNA was possible even to a dilution of 13.2 pg of
template DNA, which is equivalent to four copies of
mammalian genome.
Measurement of ADO After Amplification from Single
Human Fibroblasts
Amplification efficiency using single fibroblasts (HF57
human fibroblasts) was assessed using amplification
of various segment lengths within exons 5 to 9 of the
TP53 gene in a nested PCR configuration after I-PEP.
Unfixed fibroblasts were air dried for 2 hours and used
for analysis of amplification from single cells and for
measurement of ADO. Segments of ~500 bp (eg,
spanning exons 5 and 6) were amplified with a suc-
cess rate of 68% (33 of 48 single cells); segments of
300 bp or less (eg, exon 6) were amplified with a
success rate of 87% (42 of 48 single cells; not shown).
A ~2 Kb segment (spanning exons 5 to 9) could be
amplified in 4 (8%) of 50 single cells after I-PEP.
HF57 human fibroblasts are heterozygous for a
polymorphism (GϾA bp 13494) in intron 6 of the TP53
gene, which makes them useful for the study of ADO
after I-PEP-PCR from single cells. Exon 6 was ampli-
fied after I-PEP-PCR from 37 air-dried, single HF57
fibroblast cells. ADO was determined after DNA se-
quencing. The ADO rate was 68% (25 of 37 single
cells). Twelve single cells (32%) demonstrated het-
erozygosity at bp 13494. Of the single cells with ADO,
32% (12 single cells) showed the mutant allele se-
quence GϾA bp 13494, and 35% (13 single cells)
exhibited the wild-type sequence (Fig. 3). ADO was
not observed when five or more cells were used for
I-PEP-PCR.
Figure 3.
Evaluation of allele dropout (ADO) after improved primer extension and
preamplification (I-PEP)-PCR. Sequence analysis of a heterozygote sequence
change intron 6 bp 13494 (GϾA) of TP53 gene in single HF57 fibroblasts.
Lanes 1 and 3: single cell, heterozygosity; lanes 2, 4, and 5: single cell,
wild-type, allele dropout; lane 6: single cell, GϾA, ADO.
Amplification from Single Cells Microdissected from
Ethanol-Fixed, Paraffin-Embedded, and
Immunohistochemically Stained Tissue Sections
36% from pancreatic tissue. The amplification rates of
at least two exons were considerably higher in colon
tissues (Table 3). High multiplex amplification success
rate was achieved using tagged primers (see “Materi-
als and Method”) in first-round PCR in addition to an
antibody that ensures a hot start PCR by neutralizing
Taq polymerase.
Amplification efficiency using single microdissected
cells from ethanol-fixed, paraffin-embedded, and im-
munohistochemically stained tissue sections was de-
termined by amplification of various exons of the TP53
gene after I-PEP-PCR and a nested PCR configura-
tion. First-round amplification was performed in a
multiplex PCR using primers covering exons 5 to 9 of
the TP53 gene. Segments between 200 and 400 bp
(single exons of TP53 gene) were amplified in up to
67% of cells; segments smaller than 200 bp were
amplified in up to 72% of cells (Fig. 4). Amplification
efficiency was dependent on the type of tissue ana-
lyzed. Pancreatic tissue has a high degree of enzy-
matic DNA degradation starting early during surgical
manipulation. DNA amplification from pancreatic tis-
sue was compared with colon tissue, in which DNA is
usually not affected by enzymatic damage during
surgery. As expected, amplification success was
greater using single cells from colon than from pan-
creatic tissue. In a series of amplification experiments,
all five exons (5 to 9) of the TP53 gene were amplified
in 50% of single cells harvested from colon versus
Identification of Rare Heterozygous Mutations in Single
Cells of Normal Colon Identified To Have an
Overabundance of P53 Protein
Normal colon tissue from five patients who were
treated for colon carcinoma was fixed in ethanol/
EDTA, embedded in paraffin, and immunohistochemi-
cally stained for identification of p53 protein overex-
pression. Normal tissues from two of these patients
did not stain positive for p53 protein. Normal tissues
from the remaining three patients exhibited a positive
nuclear staining for p53 protein in up to nine epithelial
cells/slide; the total number of epithelial cells per slide
was estimated to be between 40,000 and 60,000 cells.
A total of 23 single cells and a cell cluster (40 cells)
446 Laboratory Investigation • April 2002 • Volume 82 • Number 4

Mutations in Microdissected Single Cells
Figure 4.
Representative examples of the amplification rate of various segments of exons 5 to 9 from colon cells after I-PEP-PCR. Upper: Nested PCR exon 9 (first-round
segment 378 bp); middle: different samples with nested PCR exon 5 (first-round segment 383 bp); lower: different samples with nested PCR exon 5 (first-round
segment 181 bp). Lane 1: 30 cells; lanes 2 to 16: single cell; lanes 17 and 18: negative control, no cells.
with positive p53 protein immunohistochemistry were
microdissected and amplified. Sequence analysis re-
vealed five sequence changes (Fig. 5).
for molecular analyses is usually much lower in com-
parison to paraffin-embedded tissues. To circumvent
the disadvantages associated with formalin fixation
and frozen tissues, tissues were fixed in an ethanol/
EDTA solution before paraffin embedding.
Discussion
Formalin Fixation Limits Molecular Analysis, Especially
Using Single Cells
Ethanol/EDTA Fixation Improves Molecular Analyses
Using Single Cells
There is no better place to study the underlying
molecular genetic dynamics of normal and pathologic
processes than in vivo. Histologic tissue sections are
one way to gain access to a fingerprint of these
processes within a defined morphologic compartment
and at the level of a single cell (Emmert-Buck et al,
1996; Schutze and Lahr, 1998). However, technical
limitations have hindered widespread genomic analy-
ses especially at the single-cell level in large part
because of the routine formalin fixation of tissues,
which causes DNA crosslinking, thereby reducing
PCR efficiency. This is one of the reasons why the
achieved amplification rates reported from single cells
In the present study, a mixture of 85% ethanol/0.2
mmol/L EDTA gave crisp and clear immunohisto-
chemical staining for two proteins (p53 and PCNA)
simultaneously in various tissues, providing the basis
for coupled morphologic and molecular analyses in
paraffin-embedded tissues at a single cell-level. It has
been demonstrated that ethanol fixation is the method
of choice for immunohistochemical staining of a vari-
ety of tissue antigens, especially the p53 protein
(Arnold et al, 1996; Bassarova and Popov, 1998).
Furthermore, addition of EDTA has been shown to
inhibit DNA degradation in formalin-fixed tissues to a
certain degree (Yagi et al, 1996). The specificity of the
immunohistochemical double staining for both pro-
teins was demonstrated by frequent expression of
both proteins in highly dividing tissues like colon and a
proliferative lesion of the mammary gland, by moder-
ate expression of both proteins in moderately dividing
tissues like bronchial epithelium, and by rare expres-
sion of one of the proteins in slowly dividing tissues
like liver tissue.
microdissected
from
formalin-fixed
paraffin-
embedded tissue slides did not exceed 35% in short
segments up to 246 bp (Becker et al, 1996; Roehrl et
al, 1997). Considerably higher amplification rates (up
to 70%) have been reported from single cells dis-
sected from frozen, immunohistochemically stained
tissues (Persson et al, 2000). However, major limita-
tions in the analysis of frozen tissue sections include
the usually considerably lower quality of immunohis-
tochemical staining, especially if double staining is
performed. In addition, the availability of frozen tissues
Laboratory Investigation • April 2002 • Volume 82 • Number 4 447

Heinmöller et al
Modified I-PEP Shows Improved Amplification Success
Herein, a recently published I-PEP-PCR technique
(Dietmaier et al, 1999) was thoroughly optimized. The
revised protocol demonstrated amplification of ~2 Kb
segments after I-PEP-PCR when high molecular
weight DNA was used as the template, even using
dilutions down to 12 pg, ie, four copies (or two cells) of
mammalian genome. Successful amplification of 2 Kb
segments after PEP has been described previously
but only after using 5000 cells as initial template
(Duddy et al, 1998). Using air-dried single fibroblasts,
a 2 Kb segment could be amplified but only with a
small number of cells. This may be a result of the use
of unfixed cells that were air dried for at least 2 hours,
when oxygen-related damage may have already led to
DNA degradation. When smaller segments were am-
plified after I-PEP-PCR, amplification efficiency was
up to 87%. Most importantly, the amplification rate of
segments covering single exons of the TP53 gene was
increased from 30% to up to 72% from single ethanol-
fixed, paraffin-embedded, and immunohistochemi-
cally stained single cells after the present modifica-
tions of I-PEP-PCR. As a result of this improvement,
serial molecular analyses of single cells stained for
overabundance of proteins like p53 or others can be
done in a reasonable number of PCR experiments.
Exons 5 to 9 of the TP53 gene were amplified in 50%
of single cells dissected from colon samples. The
amplification success rate in single pancreatic cells
was lower, most likely because of the much higher
degree of pancreatic protease-related DNA damage
(autodigestion) beginning early during surgical manip-
ulation of this tissue, in addition to the long surgical
procedure if performed as Whipple procedure. But
even under these most unfavorable conditions, multi-
plex PCR after WGA offers the opportunity of multiple
genetic analyses of various genes even if multiple
exons are to be analyzed. In addition, if mutations are
found by sequence analysis, reamplification for con-
firmation of a particular base change can be per-
formed from the original template, which is not possi-
ble when single cells are amplified directly by specific
PCR without prior preamplification by any WGA
method. Of interest is the rate of polymerase errors
during the amplification from single cells. Herein, over
10,000 nucleotides were sequenced in the analysis of
single, air-dried fibroblasts and not a single artificial
mutation was observed.
ADO After Modified I-PEP Amplification from Single Cells
Is Still High
ADO or preferential amplification of one allele over the
other during PCR amplification is a well-known phe-
with a heterozygous mutation: C to T transition (arrow) exon 7 bp 14,049
(pro/val). B, Single cell with deletion of four bases (ATGG) and insertion of
three bases (“TAA” arrows) exon 5 (bp 13,157 to 13,161). C, Single cell with
a silent G to A transition (arrow) in exon 8 (bp 14,506). D, Single cell with a
C to T transition (arrow) in exon 5 (bp 13,234, asp/val). E, Single cell with a
heterozygous sequence change in intron 5 (G to A transition, bp 13,292).
Figure 5.
Identification of mutations in exons 5 to 9 of the TP53 gene from single cells
that were immunohistochemically stained positive for p53 protein overabun-
dance. Please note that arrows are used to mark the mutant sequence. The
single cells were from paraffin-embedded, normal colon tissue. A, Forty cells
448 Laboratory Investigation • April 2002 • Volume 82 • Number 4

Mutations in Microdissected Single Cells
Table 1. Protocols for Cell Digestion and PEP-PCR Mixes
A. This study
B. Dietmaier et al. (Am J Pathol 1999;154:83–95)
Y
Y
Y
Y
Y
Y
Y
Y
Tween-20: 3%
Tween-20: 0.5%
Proteinase K: 0.5 mg/ml
PEP-primer: 20 ␮mol/L
MgCl2: 6 mmol/L
Gelatin: 0.05 mg/ml
dNTPs: 4 mmol/L
DMSO: 5%
Proteinase K: 4 mg/ml
Pep-primer: 16 ␮mol/L
MgCl2: 2.5 mmol/L
Gelatin: 0.05 mg/ml
dNTPs: 0.1 mmol/L
DMSO: 5%
Taq Expand HiFi: 2.5 U
Taq Expand HiFi: 1.8 U
Final concentrations are given.
nomenon limiting detection of heterozygote base
changes (Hahn et al, 1998, 2000). ADO rates between
0% and 83% have been reported and are dependent
on the source and quality of the template DNA ampli-
fied (Findlay et al, 1995, 1998; Garvin et al, 1998; Hahn
et al, 1998; Rechitsky et al, 1996; Snabes et al, 1994;
Wells et al, 1999). As a result of ADO, mutation
analysis may fail to detect heterozygous mutations in
a significant number of cases. The distribution of
alleles dropping out during PCR amplification occurs
at random. The occurrence of ADO is influenced by
multiple factors, eg, cell lysis, conditions of PCR mix,
thermal cycling conditions, guanine/cytosine content
of the DNA segment amplified, or degradation of the
DNA template (El-Hashemite and Delhanty, 1997; Ray
and Handyside, 1996; Walsh et al, 1992). Furthermore,
ADO occurs in either direct, specific (nested) amplifi-
cation or in PEP-PCR before specific PCR from single
cells (Hahn et al, 1998). This ADO estimate is specifi-
cally for the amplification method and is a minimal
estimate because it is expected that ADO is much
higher for analysis of single cells microdissected from
paraffin-embedded tissues. No method described to
date has been able to eliminate ADO.
Forty-one cell culture human fibroblasts harboring a
heterozygous polymorphism in intron 6 of the TP53
gene were used for the analysis of ADO using the
modified I-PEP-PCR protocol. Fibroblasts were
spread on slides, air dried, and stained with methylene
blue. Because fibroblasts were not prepared with a
fixative like ethanol or acetone to ensure at least a
limited protection, the fibroblasts were not protected
from oxygen radical–induced DNA damage. Because
DNA degradation starts immediately after cells are
exposed to oxygen, this situation closely resembles
the exposure of immunohistochemically stained cells
to DNA damaging oxygen radicals while sitting as a
single-cell layer on slides in the course of tissue
Table 2. PCR Primers
PCR primers
5Ј to 3Ј Sequence
First round
E5-9PD2:
E5PU1X:
E6PD1X:
E6PU1X:
E7PD1X:
E7PU2X:
E8PD2X:
E8PU1X:
E9PD1X:
E8-9SU1X:
Second round
E5-26D:
E5-291U:
E6SD2:
E6SU2:
E7SD2:
E7SU1:
E8PD3:
E8PU2:
E9PD2:
E9PU1:
5Ј gcggtcccaaaagggtcagtctcgctagtgggttgcagga 3Ј
5Ј gcggtcccaaaagggtcagtaagagcaatcagtgaggaa tc 3Ј
5Ј gcggtcccaaaagggtcagtagctggggctggagagac 3Ј
5Ј gcggtcccaaaagggtcagtcaaataagcagcaggagaaag 3Ј
5Ј gcggtcccaaaagggtcagtccctgcttgccacaggtc 3Ј
5Ј gcggtcccaaaagggtcagtaagaaatcggtaagaggtggg 3Ј
5Ј gcggtcccaaaagggtcagtttgggagtagatggagcctg 3Ј
5Ј gcggtcccaaaagggtcagtgaaagaggcaaggaaaggtg 3Ј
5Ј gcggtcccaaaagggtcagtaagcaggacaagaagcggtg 3Ј
5Ј gcggtcccaaaagggtcagtggagccattgtctttgaggc 3Ј
5Ј ttcacttgtgccctgactt 3Ј
5Ј ccgttggtcgggacagc 3Ј
5Ј acagggctggttgcccag 3Ј
5Ј gccactgacaaccaccctt 3Ј
5Ј ccaaggcgcactggcctc 3Ј
5Ј tggggcacagcaggccag 3Ј
5Ј ttccttactgcctcttgcttc 3Ј
5Ј agtgaatctgaggcataactg 3Ј
5Ј gagaccaagggtgcagttatg 3Ј
5Ј ggcaaatgccccaattgcag 3Ј
U ϭ Upstream; D ϭ Downstream; X ϭ (5Ј tagged 20-mer universal sequence); P ϭ oligo designed for PCR amplification; S ϭ oligo was designed for DNA
sequencing.
Laboratory Investigation • April 2002 • Volume 82 • Number 4 449

Heinmöller et al
Table 3. Amplification Efficiency of Exons 5 to 9 of the
Conclusion
TP53 Genea
In summary, we present a reproducible and reliable
fixation protocol for paraffin-embedded tissues that
allows clear and crisp immunohistochemical staining
of at least two proteins simultaneously and preserves
DNA integrity to a degree that allows the molecular
analysis of multiple genes at the single-cell level.
Single-cell microdissection was achieved by a fast
and easy-to-handle simple manual microdissection
device that is affordable to many laboratories, being
about one tenth the price of a laser microdissection
device. Furthermore, the success rate of PCR ampli-
fication from DNA obtained from a single cell dis-
sected from paraffin-embedded tissue is increased to
a level where serial analyses of mutations in a single
cell or a few cells from normal or tumor tissue can be
performed in a reasonable number of experiments.
This could be the basis for further studies to determine
an individual TP53 mutation load in normal tissues in
the context of morphology, enabling an estimation of
individual cancer risk well in advance of the develop-
ment of malignant disease.
Number of exons
Amplification
%
Colon tissue
2 (e.g. exons 6ϩ8)b
17/18
16/18
12/18
9/18
94
89
66
50
3
4
5 (exons 5–9)
Pancreatic tissue
2 (e.g. exons 6ϩ8)
3
17/28
15/28
12/28
10/28
61
54
43
36
4
5 (exons 5–9)
a Eighteen microdissected paraffin-embedded single cells from normal
colon tissue and 28 single cells from normal pancreatic tissue.
b At least two exons.
processing and microdissection. Herein the ADO rate
was 68% and affected both alleles randomly. If a
heterozygous mutation is present, according to the
data herein, the chance of the correct diagnosis of
heterozygosity will be 22%.
Material and Methods
Cell and Tissue Samples
Mutation Analysis in Single Cells from Ethanol-Fixed,
Paraffin-Embedded, and Immunohistochemically Stained
Normal Tissues
Tissue samples from patients in surgery for colorectal
cancer or pancreatic cancer were obtained from the
Department of Pathology from the City of Hope Na-
tional Medical Center (Duarte, California). Suspen-
sions of cultured human fibroblasts (HF57) were a
generous gift from Steven Bates (Department of Biol-
ogy, City of Hope).
This is the first report of specific molecular analyses
after WGA from immunohistochemically stained single
cells microdissected from paraffin-embedded tissues.
In the analysis of a limited number of microdissected
single cells from normal colon tissue staining positive
for p53 protein, sequence changes in the TP53 gene
were found in 4 of 23 single cells and in 1 cell cluster
of 40 cells. The sequence changes were not published
mutational hotspots of the TP53 gene. The relevance
of the heterozygote mutations detected in intron 5 of
one single cell and the heterozygous base change
seen in the cell cluster in regard to inactivation of the
TP53 gene is not clear. However, heterozygous muta-
tions inactivating the remaining wild-type allele have
been described (Forrester et al, 1995; Hachiya et al,
1994; Milner and Medcalf, 1991; Villadsen et al, 2000).
Possible reasons besides ADO for detecting only
wild-type sequence in a subset of single cells with
positive p53 immunohistochemistry dissected from
tissue slides may be contamination of the PCR or loss
of genetic material containing the allele harboring the
base change during tissue cutting. The chance of
losing genetic material from a single cell nucleus in the
course of tissue sectioning can be estimated to be
between 10% and 20% if sections of 6 ␮m are
processed. It should be noted that switching from the
analysis of single paraffin-embedded cells to single
cells from fresh-frozen tissues might not be helpful in
solving the ADO problem. A recently published article
reported an ADO rate of 50% from single cells dis-
sected from immunohistochemically stained fresh-
frozen tissue sections (Persson et al, 2000).
Tissue Processing and Immunohistochemistry
After tissues were handled for pathologic-anatomical
diagnosis, samples were fixed immediately in 85%
ethanol supplemented with 0.2 mmol/L EDTA for 24
hours and embedded in paraffin thereafter. Six-micron
serial sections of ethanol (EDTA)-fixed, paraffin-
embedded tissues were mounted on Probe-On slides
(Ventana Medical Systems, Tucson, Arizona) and
baked at 56° C for 1 hour. Slides were deparaffinized
by incubating the slides in xylene for 2 ϫ 10 minutes
and rehydrating in 100% ethanol for 2 ϫ 5 minutes, in
96% ethanol for 2 ϫ 5 minutes, and in 70% ethanol for
2 ϫ 5 minutes. Antigen retrieval was performed by
placing the slides into steam heat for 5 minutes.
Antibodies used for double staining were PCNA (Ab-1)
(Oncogene Research Products, Cambridge, Massa-
chusetts) and NCL-p53-DO7 (Novocastra, Newcastle-
upon-Tyne, United Kingdom). Staining was performed
using a Biotech Techmate 1000 Immunostainer (Ven-
tana Medical Systems). Double staining of slides was
performed using a modified ABC-Method (Ventana
Medical Systems). Human PCNA was stained brown
using the chromogen 3'3-diaminobenzidine, and the
p53 protein was stained red using the chromogen
Bio-Red. Appropriate positive tissue controls and neg-
ative controls for primary antibodies were included in
450 Laboratory Investigation • April 2002 • Volume 82 • Number 4

Mutations in Microdissected Single Cells
every experiment. Tissue sections from a lung adeno-
carcinoma exhibiting p53 protein overexpression of
tumor cells served as positive control. Staining without
primary antibody was performed on the lung cancer
tissue sections for demonstration of specificity of
staining. Stained tissue sections were covered with
aqueous mounting medium (Aquatex; Merck, Darm-
stadt, Germany) and a glass coverslip. Samples were
stored frozen at Ϫ20° C until examined.
For determination of optimal tissue fixation and
immunohistochemistry, fresh tissue samples were cut
in three pieces and fixed in 95% ethanol/0.01 mmol/L
EDTA, 85% ethanol/0.2 mmol/L EDTA, and 70% eth-
anol/1 mmol/L EDTA, respectively. After 24 hours,
tissues were paraffin embedded, and immunohisto-
chemical staining was performed as described above.
The quality and reproducibility of staining was as-
sessed independently by two pathologists (YS and
LMW). DNA integrity was routinely assessed after
Proteinase K digestion (2 mg/ml final concentration for
16 hours followed by a 10-minute inactivation step at
90° C) of cells from a variety of tissues scraped from
serial sections taken at different stages of tissue
processing and immunohistochemical staining. Eight-
microliter aliquots of digested cell samples were elec-
trophoresed through a 1% agarose gel for 30 minutes
and stained with ethidium bromide.
from tissue sections) at 50° C followed by a 10-minute
inactivation at 90° C.
PCR
A recently described protocol for improved primer
extension preamplification (I-PEP)-PCR (Dietmaier et
al, 1999) was used with major modifications (Table 1)
in a Perkin Elmer 9600 thermocycler (Foster City,
California). I-PEP-PCR was set up by adding 30 ␮l of
I-PEP mix [final concentration: 0.05 mg/ml gelatin, 20
␮mol/L (N)15 random primer, 1 mmol/L each dNTP, 3
U of Taq Expand High Fidelity polymerase, 6 mmol/L
MgCl2, in 1ϫ PCR buffer No. 3 from Roche] to 10 ␮l of
lysis buffer containing the lysed cell. PCR was run for
50 cycles as follows: Step 1: 95° C for 2 minutes; Step
2: 95° C for 30 seconds; Step 3: 28° C for 90 seconds;
Step 4: ramp 0.1° C per second to 55° C; Step 5:
55° C for 2 minutes; Step 6: 68° C for 3 minutes; Step
7: go to Step 2, 49 times; Step 8: 68° C for 15 minutes;
and Step 9: 4° C. Exons 5 to 9 of the TP53 gene were
amplified in a nested PCR approach: first-round PCR
was performed as a multiplex PCR using upstream
and downstream primers specific for each exon
tagged with an unrelated chimeric 20-nucleotide se-
quence at the 5' end of all primers (Shuber et al, 1995)
(Table 2). Second round PCR was performed in a
nested or hemi-nested condition using primers given
in Table 2. Five microliters of I-PEP-PCR product was
aliquoted to 25 ␮l of the specific PCR first-round mix:
0.2 mmol/L each dNTP; 1.5 mmol/L MgCl2; 0.4 ␮mol/L
primers exons 5, 7, and 9; 0.6 ␮mol/L primers exons 6
and 8; 0.5 U of Taq Expand High Fidelity Polymerase;
and 0.5 U of Anti-Taq Antibody (Gibco, Eggenstein,
Germany). Each exon was amplified separately in a
second-round PCR. Three microliters of first-round
PCR product was aliquoted to 22 ␮l of mix containing
1.5 mmol/L MgCl2; 0.2 mmol/L each dNTP; 0.4
mmol/L primers exons 5 and 7; 0.6 mmol/L primers
exons 6, 8, and 9; and 0.3 U of Taq Expand High
Fidelity Polymerase. Cycling conditions were identical
for first-round and second-round PCR: Step 1: 95° C
for 2 minutes; Step 2: 95° C for 30 seconds; Step 3:
55° C for 45 seconds; Step 4: 72° C for 1 minute; Step
5: go to step 2, 29 times; Step 6: 72° C for 10 minutes;
and Step 7: 4° C. Negative control reactions with
water instead of DNA were included in each experi-
ment. The presence and relative quantity of PCR
product was ascertained by resolution on a 2% aga-
rose gel. The primers of the nested PCR reaction were
used for DNA sequencing.
Preparation of Single Cells from Cell Cultures and from
Paraffin-Embedded, Immunohistochemically Stained
Tissue Sections
Fibroblasts resuspended in PBS were spread onto
glass slides and were air dried for 2 hours. After
staining with methylene blue, single cells were picked
with sterile needles (Microlance; Becton Dickinson,
Franklin Lakes, New Jersey). Single cells or small cell
clusters on tissue slides staining positive for p53
protein overabundance were microdissected using a
manual micromanipulator (MP285; Sutter Instruments,
Novato, California) equipped with a Tungsten needle
(FHC Inc., Bowdoinham, Massachusetts). An inverted
microscope (Nikon TMS, Melville, New York) was
used. Microdissection was performed as follows: with
a fine surgical needle and at lower magnification, the
target cell was chosen and the area around it was
freed from surrounding tissue by scraping a free space
as close as possible to the target. After changing to a
magnification of ϫ40, the cell was then pushed into
the free space with the joystick-operated micromanip-
ulator, picked up with a new surgical needle, and
transferred into the PCR tube. Approximately 5 to 10
minutes are necessary to microdissect a single cell
from normal colonic epithelium. Fibroblasts in a
single-cell layer were dissected in less than 2 minutes
per cell and could be dissected manually without the
micromanipulator. Cells were transferred into 10 ␮l of
1ϫ Taq PCR buffer No. 3 without MgCL2 (Roche)
containing 400 ␮g/ml (fibroblasts) or 1 mg/ml (single
cells from tissue sections) of Proteinase K and 3%
Tween-20 (Roche). Cell lysis was performed by incu-
bation for 3 hours (fibroblasts) or 16 hours (single cells
DNA Sequencing
PCR products were purified using Microcon-100 Mi-
croconcentrators (Amicon Inc., Beverly, Massachusetts)
and sequenced using the Big Dye Terminator Sequenc-
ing Kit (Applied Biosystems, Foster City, California) ac-
cording to the manufacturers instructions. Sequence
changes were confirmed by reamplifying from the origi-
nal I-PEP-PCR and sequencing in the opposite direction.
Repeated amplifications of the exons from another PEP
Laboratory Investigation • April 2002 • Volume 82 • Number 4 451

Heinmöller et al
Garvin AM, Holzgreve W, and Hahn S (1998). Highly accurate
analysis of heterozygous loci by single cell PCR. Nucleic
Acids Res 26:3468–3472.
aliquot of the same cell were performed to show that the
mutation could be reproducibly identified in independent
aliquots of the PEP. Sample chromatograms were ana-
lyzed in a blinded manner by two individuals. Seq Ed (PE
Biosystems Foster City, California) and Sequencher soft-
ware (Gene Codes, Inc., Ann Arbor, Michigan) were used
to generate two or multiple sequence alignments, re-
spectively. Differences between sequences were com-
pared with the published wild-type sequence using the
IARC TP53 mutation database (http://www.iarc.fr/p53/).
Greenblatt MS, Bennett WP, Hollstein M, and Harris CC
(1994). Mutations in the p53 tumor suppressor gene: Clues to
cancer etiology and molecular pathogenesis. Cancer Res
54:4855–4878.
Hachiya M, Chumakov A, Miller CW, Akashi M, Said J, and
Koeffler HP (1994). Mutant p53 proteins behave in a domi-
nant, negative fashion in vivo. Anticancer Res 14:1853–1859.
Hahn S, Garvin AM, Di Naro E, and Holzgreve W (1998). Allele
drop-out can occur in alleles differing by a single nucleotide
and is not alleviated by preamplification or minor template
increments. Genet Test 2:351–355.
Acknowledgements
We thank Dr. Wolfgang Dietmaier for helpful discus-
sions, Mrs. Wenyan Li, Mrs. Martha Magallanes, Mrs.
Xuemin Li and Mrs. Sofia Loera for expert technical
assistance, and Dr. S. Bates for the cells.
Hahn S, Zhong XY, Troeger C, Burgemeister R, Gloning K,
and Holzgreve W (2000). Current applications of single-cell
PCR. Cell Mol Life Sci 57:96–105.
Huusko P, Castren K, Launonen V, Soini Y, Paakkonen K,
Leisti J, Vahakangas K, and Winqvist R (1999). Germ-line
TP53 mutations in Finnish cancer families exhibiting features
of the Li-Fraumeni syndrome and negative for BRCA1 and
BRCA2. Cancer Genet Cytogenet 112:9–14.
References
Arnold MM, Srivastava S, Fredenburgh J, Stockard CR,
Myers RB, and Grizzle WE (1996). Effects of fixation and
tissue processing on immunohistochemical demonstration of
specific antigens. Biotech Histochem 71:224–230.
Karlsen F, Kalantari M, Chitemerere M, Johansson B, and
Hagmar B (1994). Modifications of human and viral deoxyri-
bonucleic acid by formaldehyde fixation. Lab Invest 71:604–
611.
Barnes DM, Hanby AM, Gillett CE, Mohammed S, Hodgson
S, Bobrow LG, Leigh IM, Purkis T, MacGeoch C, and Spurr
NK (1992). Abnormal expression of wild type p53 protein in
normal cells of a cancer family patient. Lancet 340:259–263.
May P and May E (1999). Twenty years of p53 research:
Structural and functional aspects of the p53 protein [pub-
lished erratum appears in Oncogene 2000 Mar 23;19(13):
1734]. Oncogene 18:7621–7636.
Bassarova AV and Popov AA (1998). Immunohistochemical
detection of p53: Effect of fixation and methods of antigen
retrieval. Folia Histochem Cytobiol 36:127–132.
Milner J and Medcalf EA (1991). Cotranslation of activated
mutant p53 with wild type drives the wild-type p53 protein
into the mutant conformation. Cell 65:765–774.
Becker I, Becker KF, Rohrl MH, Minkus G, Schutze K, and
Hofler H (1996). Single-cell mutation analysis of tumors from
stained histologic slides. Lab Invest 75:801–807.
Persson AE, Ling G, Williams C, Backvall H, Ponten J, Ponten
F, and Lundeberg J (2000). Analysis of p53 mutations in
single cells obtained from histological tissue sections. Anal
Biochem 287:25–31.
Dietmaier W, Hartmann A, Wallinger S, Heinmoller E, Kerner
T, Endl E, Jauch KW, Hofstadter F, and Ruschoff J (1999).
Multiple mutation analyses in single tumor cells with im-
proved whole genome amplification. Am J Pathol 154:83–95.
Ponten F, Williams C, Ling G, Ahmadian A, Nister M, Lunde-
berg J, Ponten J, and Uhlen M (1997). Genomic analysis of
single cells from human basal cell cancer using laser-
assisted capture microscopy. Mutat Res 382:45–55.
Duddy SK, Gorospe S, and Bleavins MR (1998). Genetic
analysis of multiple loci in microsamples of fixed paraffin-
embedded tissue. Toxicol Sci 46:317–323.
El-Hashemite N and Delhanty JD (1997). A technique for
eliminating allele specific amplification failure during DNA
amplification of heterozygous cells for preimplantation diag-
nosis. Mol Hum Reprod 3:975–978.
Ray PF and Handyside AH (1996). Increasing the denatur-
ation temperature during the first cycles of amplification
reduces allele dropout from single cells for preimplantation
genetic diagnosis. Mol Hum Reprod 2:213–218.
Emmert-Buck MR, Bonner RF, Smith PD, Chuaqui RF,
Zhuang Z, Goldstein SR, Weiss RA, and Liotta LA (1996).
Laser capture microdissection [see comments]. Science 274:
998–1001.
Rechitsky S, Freidine M, Verlinsky Y, and Strom CM (1996).
Allele dropout in sequential PCR and FISH analysis of single
cells (cell recycling). J Assist Reprod Genet 13:115–124.
Roehrl MH, Becker KF, Becker I, and Hofler H (1997).
Efficiency of single-cell polymerase chain reaction from
stained histologic slides and integrity of DNA in archival
tissue. Diagn Mol Pathol 6:292–297.
Findlay I, Matthews P, and Quirke P (1998). Multiple genetic
diagnoses from single cells using multiplex PCR: Reliability
and allele dropout. Prenat Diagn 18:1413–1421.
Findlay I, Ray P, Quirke P, Rutherford A, and Lilford R (1995).
Allelic drop-out and preferential amplification in single cells
and human blastomeres: Implications for preimplantation
diagnosis of sex and cystic fibrosis. Hum Reprod 10:1609–
1618.
Schutze K and Lahr G (1998). Identification of expressed
genes by laser-mediated manipulation of single cells. Nat
Biotechnol 16:737–742.
Shuber AP, Grondin VJ, and Klinger KW (1995). A simplified
procedure for developing multiplex PCRs. Genome Res
5:488–493.
Forrester K, Lupold SE, Ott VL, Chay CH, Band V, Wang XW,
and Harris CC (1995). Effects of p53 mutants on wild-type
p53-mediated transactivation are cell type dependent. On-
cogene 10:2103–2111.
Snabes MC, Chong SS, Subramanian SB, Kristjansson K,
DiSepio D, and Hughes MR (1994). Preimplantation single-
452 Laboratory Investigation • April 2002 • Volume 82 • Number 4

Mutations in Microdissected Single Cells
cell analysis of multiple genetic loci by whole-genome am-
plification. Proc Natl Acad Sci USA 91:6181–6185.
Wells D, Sherlock JK, Handyside AH, and Delhanty JD
(1999). Detailed chromosomal and molecular genetic analy-
sis of single cells by whole genome amplification and com-
parative genomic hybridisation. Nucleic Acids Res 27:1214–
1218.
Suarez-Quian CA, Goldstein SR, Pohida T, Smith PD, Peter-
son JI, Wellner E, Ghany M, and Bonner RF (1999). Laser
capture microdissection of single cells from complex tissues.
BioTechniques 26:328–335.
Williams C, Ponten F, Moberg C, Soderkvist P, Uhlen M,
Ponten J, Sitbon G, and Lundeberg J (1999). A high fre-
quency of sequence alterations is due to formalin fixation of
archival specimens. Am J Pathol 155:1467–1471.
Varley JM, McGown G, Thorncroft M, James LA, Margison
GP, Forster G, Evans DG, Harris M, Kelsey AM, and Birch JM
(1999). Are there low-penetrance TP53 alleles? Evidence
from childhood adrenocortical tumors. Am J Hum Genet
65:995–1006.
Yagi N, Satonaka K, Horio M, Shimogaki H, Tokuda Y, and
Maeda S (1996). The role of DNase and EDTA on DNA
degradation in formaldehyde fixed tissues. Biotech Histo-
chem 71:123–129.
Villadsen R, Nielsen KV, Bolund L, and Briand P (2000).
Complete loss of wild-type TP53 in a nontransformed human
epithelial cell line is preceded by a phase during which a
heterozygous TP53 mutant effectively outgrows the homozy-
gous wild-type cells. Cancer Genet Cytogenet 116:28–34.
Zhang L, Cui X, Schmitt K, Hubert R, Navidi W, and Arnheim
N (1992). Whole genome amplification from a single cell:
Implications for genetic analysis. Proc Natl Acad Sci USA
89:5847–5851.
Wallace-Brodeur RR and Lowe SW (1999). Clinical implica-
tions of p53 mutations. Cell Mol Life Sci 55:64–75.
Walsh PS, Erlich HA, and Higuchi R (1992). Preferential PCR
amplification of alleles: Mechanisms and solutions. PCR
Methods Appl 1:241–250.
Laboratory Investigation • April 2002 • Volume 82 • Number 4 453