mesothelioma cancer

Simian virus (SV40) was discovered as one of the viruses capable of infecting
Macacus rhesus as well as Macacus cynomolgus monkey kidney cells (1). It also
had the possibility to infect and to transform human cells grown in vitro.
Simian virus 40 is a DNA tumor virus that not only induces tumors in rodents but
also is capable of immortaliz¬ing human mesothelial cells in vitro. Except for a
report demonstrat¬ing SV40 in one metastatic melanoma, SV40 was not considered
to be oncogenic in humans (2). Simian virus 40 DNA has been found in several
human tumors such as choroid plexus tumors, osteosarcomas, malignant
mesotheliomas, and lymphoproliferative diseases such as non-Hodgkin’s lymphomas
(3–5). The role of the SV40 virus in human tumors has been extensively discussed
in several excellent reviews (4,6,7).

Simian virus 40 is a double-stranded DNA virus whose genome encodes two tumor
(T)-antigens known as large T-antigen and small t-antigen. Replication of the
double-stranded DNA genome occurs in the nucleus of the host cell. Transcription
of the genome is carried out by host cell RNA polymerase II, and large T-antigen
plays a major role in regulating transcription of the viral genome by binding to
the origin region of the viral genome. Protein–protein interactions between
T-antigen and DNA polymerase alpha directly stimulate replication of the viral
genome. Small t-antigen is not essential for virus replication but allows viral
DNA to accumulate in the nucleus. Both proteins contain nuclear localization
signals, which results in their accumula¬tion in the nucleus, where they migrate
after being synthesized in the cytoplasm. After infection, early messenger RNAs
(mRNAs) are expressed from the early promoter, which contains a strong
transcrip-

Methodology for Detection of SV40 DNA

Experimental work suggested that SV40 could cause mesotheliomas. Syrian hamsters
injected with SV40 wild-type virus developed malig¬nant tumors such as true
histiocytic lymphomas, sarcomas, and meso-theliomas according to the site of
injection (15). The SV40 mutant virus that lacked the gene sequence for tag was
not able to induce malig¬nant mesothelioma but initiated only the growth of true
histiocytic lymphomas or sarcomas (14). In vitro experiments suggest that human
mesothelial cells are unusually susceptible to SV40-mediated transfor¬mation and
asbestos co-carcinogenicity (16). The experimental findings in hamsters prompted
Carbone et al (17) to investigate human mesothelial tumor tissue for the
presence of SV40. The presence of SV40 DNA in 29 of 48 mesotheliomas was found
using polymerase chain reaction (PCR) with several primer sets PYV.for/PYV.rev
and SV.for3/ SV.rev. These findings were confirmed by sequencing and blotting

experiments. Carbone et al were able to prove that the viral DNA was actively
transcribed and translated to viral Tag. Immunohistochemistry carried out on
frozen sections as well as immunoprecipitation assays with the mouse monoclonal
antibody Pab419 confirmed the presence of viral Tag.

Most studies dealing with the detection of SV40 DNA now describe the use of PCR
or a PCR-based assay as the gold standard. The PCR technique described by
Bergsagel and coworkers (3) for detection of the SV40 DNA in ependymomas and
choroid plexus has become the method used by most research groups. The DNA of
the tumor sample is extracted, purified, and amplified with specific beta-globin
primers AG1 and AG-2 for checking the integrity and amplifiability of the
extracted DNA. All specimens from which adequate DNA is extracted are examined
for viral sequences with the primers PYV.for and PYV.rev. These primers amplify
a conserved region of the Tag that is common to the papovaviruses SV40 virus,
BK, and Jamestown Canyon (JC) virus resulting in an amplicon of 172bp. This
region codes for the Rb, p107, and the Rb2/p130 binding domain of Tag. When
using this primer set one only detects papovavirus, but since this primer set
does not dis¬tinguish among BK, JC, and SV40, additional techniques such as
South¬ern blotting have to be performed for demonstrating with certainty the
SV40 origin of the amplified PCR.

Probes specific for BK, JC, SV40 were used in a Southern blot tech¬nique to
determine whether SV40 was present or whether the ampli-con was derived from BK
or JC virus. However, under some conditions BK- and JC-specific probes were
found to cross-react with SV40. To resolve ambiguities regarding the
cross-reactivities of specimens with BK and JC DNA probes, one has to sequence
the PCR amplicons. The sequences of BK and JC virus contain an additional 9-bp
insert in contrast to SV40, which lacks this 9-bp insert. If fresh tissue or
frozen tissue is available for analysis, one can use the primer set SV.for2 and
SV.rev amplifying a 574-bp region of the SV40 Tag gene.

The primers SV.for3 and SV.rev can be used to amplify a 105-bp frag¬ment of the
SV40 that only partially overlaps the fragment amplified with PYV.for and
PYV.rev. The former set can be used easily for analy¬sis of paraffin-embedded,
formalin-fixed tissues where the DNA is par¬tially degraded due to formalin
fixation and subsequent processing. Other primer sets such as SV5-SV6 result in
amplicons of 169bp, while TA1-TA2 amplify a segment of 441bp (Table 19.1).

Different positivity rates for SV40 DNA have been found in malig¬nant
mesothelioma when using different primer sets. The wide vari¬ability in the
detection rates may be partly explained by the relatively small numbers of cases
analyzed in each study. However, one must take into account the differences in
tissue quality, formalin-fixed tissue versus fresh frozen samples, the
differences in DNA extraction method, the variability in PCR amplification and
detection methods, as well as the geographic differences.

It remains unclear why different primer sets give different results. Mutations
and deletions occur in most viruses and also in SV40 virus. Mutations at the 3¢
site where the primer has to bind can lead to mis-priming with no resulting Taq
polymerase activity and no production of specific SV40 amplicon. However,
mutations are unlikely to occur in Rb-pocket binding domain as this site is
highly relevant for Tag-

mediated cell transformation. According to this hypothesis, the primer set
specific for the Rb-pocket, SV3.for/SV.rev would be expected to give the highest
positive results (18,19).

There is little known about the state and nature of the SV40 genome in tumoral
cells. The infective viral state of the SV40 genome has so far been isolated
from only one choroid plexus tumor (20). In the majority of investigated tumors
in this series the SV40 DNA was predominantly present in its full-length
episomal state. The same was true of the viral DNA associated with osteosarcomas
(21). In the brain tumor study, each tumor appeared to be associated with a
single homogeneous viral DNA species defined by its unique sequence related to
the variable domain (the last 87 amino acids, 622 to 708) of the C-terminus of
Tag. The pos¬sible integrated state of the virus has been suggested in three of
69 human papillary thyroid carcinomas and in five of 10 osteosarcomas (22,23).
Integration of the SV40 genome into the DNA of the host results in opening of
the circular SV40 genome and leads to a linearized SV40 genome. This can lead to
disruption of certain genes not only from the host but also of viral genes. It
may end in the total loss of viral genes. This phenomenon has also been
described for cervical carcinoma– associated oncogenic human papillomavirus
(HPV) 16 where integra¬tion leads to changes or even loss of the L1 region,
resulting in negative PCR results for primers detecting sequences in the L1
region. In contrast, the E6 and E7 regions coding for the oncogenic proteins E6
and E7 remain highly conserved throughout the complete multistep carcinogenesis
from dysplasia to frankly malignant and invasive cer¬vical carcinoma (24). It is
also noteworthy that different HPV primer sets also give different rates of
positivity for a given sample (25). Those primers that amplify smaller segments
are found to give a higher pos¬itive yield in contrast to those amplifying
larger segments. Primer sets detecting sequences in highly conserved regions of
HPV such as the E6 and E7 genes also give higher yields than those situated in
other regions such as the L1 region known to be less conserved during the
carcinogenesis proces. It is very remarkable and maybe also similar to SV40 that
in very early lesions the HPV genome is present in the epi-somal form and that
when integration of the genome occurs the lesion starts further to evolve into
high-grade dysplasia and invasive frankly malignancy. When applying these
findings on the SV40 situation, one can readily appreciate why some SV40 primer
sets give different pos-itivity rates as some regions, especially the Rb-binding
pocket, remain highly conserved while others may be lost during the multistep
car-cinogenesis process. It is also understandable that primer sets that amplify
smaller segments will yield a higher positivity rate than those that amplify
larger segments. One has to bear in mind that some primers can amplify also Tag
sequences from other papovaviruses, and even Southern blot techniques with
specific probes may not always be sufficient to discern the SV40 Rb-binding
pocket from other papova Tags. A likely candidate would be the Tag of the BK
virus as this virus has been found to be very ubiquitous in the human
population.

The SV40 and BK tags are very similar, but a 9-bp insert facilitates
distinction. For this reason, sequencing is the preferred method when

working with PYV.for-PYV.rev and SV3.for-SV.rev primer sets. One cannot rule out
with absolute certainty that the sequences detected do not belong to a
recombinant SV40-BK virus or another unknown virus.

In one study the presence of SV40 DNA was confirmed in 14 of 25 malignant
mesotheliomas using the novel Primed in situ (PRINS) method on paraffin-embedded
formalin-fixed tissue specimens (26). The PRINS labeling method is based on a
primer-mediated DNA syn¬thesis starting with the annealing of an oligonucleotide
DNA primer adjacent to the DNA region of interest. This oligonucleotide serves
as a primer for the DNA polymerase Taq polymerase incorporating the four
nucleotides dATP, dGTP, dUTP, and dCTP, of which the dUTP is labeled with
digoxigenin. The label can be visualized immunohisto-chemically using alkaline
phosphatase-bound antibody or peroxidase-bound antibody. The application of the
PRINS methodology has been focused primarily on cytogenetics (27). This method
can be applied on paraffin-embedded formalin-fixed tissue sections (28).

The PRINS reactions were carried out with three separate primers, PYV.rev,
SV2.for and SV.rev, detecting a DNA sequence coding for the viral tag of SV40.
The latter is detecting a region (4425–4402) coding for Tag, which is common to
BK, JC, and SV40. The former two primers are SV40 specific. The SV.rev primer is
homologous to the 4399 to 4372 region and the SV2.for primer is identical to the
region 4920 to 4945. Both are located in the DNA sequence coding for the Tag but
outside the Rb-p107-Rb2/p130 binding domain.

Nuclear staining was found in 14 of the 25 investigated mesothe-lioma tissue
specimens using the PYV.rev primer in the PRINS method, indicating the presence
of SV40 or SV40-like DNA, as this primer also recognizes a sequence identical to
SV40, JC, and BK virus. A separate PRINS reaction with the Sv2.for primer and
Sv.rev primer, considered to be specific for the SV40 virus, revealed a nuclear
signal in all those cases that tested positive with the PYV.rev primer. The
percentage of positive cells varied considerably among different tumors. The
percentage of positive neoplastic cells ranged from 2% in an epithelial
mesothelioma to nearly 100% in a mesenchymal mesothelioma (Figs. 19.1 and 19.2).
Stromal elements such as blood vessels, adipocytes, and inflammatory cells were
found not to be positive. In all the cases that tested positive for SV40 or
SV40-like DNA, cytoplasmic and nuclear immunoreactivity was found for tag with
the monoclonal antibody pAb280. The amount of immunoreactive cells was
comparable to those with nuclear signal for SV40-like DNA in a given sample. No
nuclear signal indicative for the presence of SV40 or SV40-like DNA or
immunoreactivity for tag with the monoclonal antibody Pab 280 was found in those
cases with nonneoplastic mesothelium or pleural carci¬noma metastasis.

Polymerase chain reaction contamination could result from the SV40 DNA itself or
from plasmids containing various segments of the SV40 genome. It is very
unlikely to assume that all reported positive results would be the result of
contamination by one or another source. Some laboratories reporting positive
results had never before worked with SV40 or known laboratory SV40 strains and
some cases displayed several different mutations or deletions. One would not
expect such data when dealing with contamination, where always the same bands or
band would be found. Tag protein and mRNA were found in mesothelioma, thereby
making the possibility of PCR contamination by exogenous SV40 DNA very unlikely.

In some cases, the sequences are from SV40 itself, as the virus could be rescued
or extensive sequencing confirmed the SV40 origin. For the majority of cases,
where extensive sequencing has not been carried out, it is better to use “SV40
like” instead of SV40. The possibility of a recombinant SV40 cannot be ruled out
with certainty, nor can the pos¬sibility of a human virus related to SV40 or
even a cellular homologue of Tag.

Specificity of PCR for SV40 Detection

The question of specificity of SV40 detection has been challenged on two
grounds. The failure of Strickler et al (29) to detect SV40 DNA in their
paraffin-embedded samples has been used as an argument to consider SV40
positivity in mesothelioma specimens as a result of intra- or interlaboratory
contamination. The possibility of cross-contamination by exogenous SV40 to be
high due to the long and widely distributed usage of SV40 early region base
vectors and gene constructs was hypothesized (19). This criticism has been
addressed by the experi¬ments of Testa and coworkers (30). They reported the
conclusions of a multilaboratory study they had directed following the
recommendation of the 1997 National Institutes of Health (NIH) conference and
upon the request of the International Mesothelioma Interest Group (IMIG). This
study included the laboratory of Michele Carbone, who had reported positive
findings for SV40 in mesotheliomas in the United States, and that of Kaija
Linnainmaa, who had reported negative findings in mesotheliomas in Finland. The
laboratory of Kalily, which specialized in JC virus, was also involved in this
study to confirm that the isolated DNAs did not belong to other papovaviruses
such as BK or JC virus. The mesothelioma material was distributed to four
independent labo¬ratories, three of which never had worked with SV40. The DNA
was extracted in one single independent center not previously exposed to SV40.
This study confirmed the presence of SV40 DNA and proteins in 10 of 12 (83%)
mesotheliomas in the United States. The positive and neg¬ative controls yielded
consistent results in all centers (30). High reliabil¬ity for SV40 detection was
demonstrated by revealing SV40 positivity in

U.S. mesothelioma specimens and absence of SV40 DNA in all investi¬gated Finnish
specimens using a double-blind analysis approach (31). To verify these results,
Testa et al organized a second study where blinded DNAs of 11 mesothelioma
samples in the U.S. together with nine mesothelioma specimens in Turkey were
investigated. They found four of 11 U.S. mesotheliomas to be positive for SV40
in contrast to Turkish mesotheliomas, which were all negative for SV40. This
discrep¬ancy in results between U.S. mesotheliomas on one side and Turkish as
well as Finnish mesotheliomas on the other side, suggested that SV40 prevalence
can display regional differences. It is noteworthy that in con¬trast to the
United States, Turkey started to administer polio vaccines in the 1970s that are
considered to be cleared of SV40. Whether these dif¬ferences can be solely
attributed to SV40-contaminated polio vaccines or other hitherto unknown factors
remains to be determined.

The validity of the contamination theory was further investigated by screening
coded DNA aliquots of 32 matched blood samples and prostate biopsies for SV40
DNA by PCR (32). The SV40 DNAwas found in nine of 64 specimens and in 31 of 32
patients the prostate and blood samples were completely concordant. The low
level of SV40 DNA and the nearly completely concordance is inconsistent with the
possibility of contamination. The found positivity was attributed to circulating
mononuclear cells in the blood rather than to malignant cells. This underscores
the importance of confirming positive PCR results by other techniques, such as
immunohistochemistry, in situ hybridization, or microdissection before
attributing SV40 to a certain tumor type. Investigation of microdissected
mesothelioma cells and nearby stromal cells from the same paraffin section using
PCR followed by sequenc¬ing displayed SV40 DNA in 57% of the mesothelioma
samples. The SV40 sequences were present in both the preinvasive and invasive
com¬ponent of the tumor cells, which indicated that SV40 is present in the early
stages of the process. Microdissected cells from matched adjacent lung tissue
were negative for SV40 sequences (33).

One research group conducted its expriments with new mesothe-lioma specimens in
a separate laboratory in a new building where experiments with viruses or
plasmids had never been carried out (34). The group was able to confirm its
former results by demonstrating SV40 regulatory sequences in 10 of 18
mesothelioma samples and SV40 Tag sequences in eight of 10 mesothelioma samples
(34,35).

Another study confirmed the presence of SV40 DNA in eight of 12 malignant
mesothelioma specimens but detected SV40 DNA only in 11 of 49
lymphoproliferative disorders (5). DNAextractions and PCR reac¬tions were
carried out at the same time by the same investigators, thereby ruling out the
possibilities that either the technical procedure used was not sufficiently
sensitive to detect SV40 or the high percent¬age of SV40 DNA–positive
mesotheliomas might be related to PCR contamination (5).

Reproducibility of the PCR Reaction

The reproducibility of SV40 DNA detection by the PCR technique has been
challenged by Strickler and Mulatero. Studies conducted by Shah

upon the request of Strickler and Goedert, failed to detect SV40 DNA in any of
the investigated mesothelioma specimens. However, this study was completely
based on paraffin-embedded material (Strickler et al, 1997). A number of
possibilities were formulated to account for the possible negative data such as
limited sensitivity and different tech¬nical approach. But it remains unclear
why Shah’s analyses appear to be at odds with those published by other groups.
The author did not use Southern blotting or filter hybridization techniques as
confirmatory technique after PCR amplification. In the study of Strickler, only
four cases were found positive after ethidium bromide staining of agarose gels
compared to 26 of 26 after filter hybridization in Carbone et al’s study (17).
In a new nine-laboratory multicenter investigation, none of the selected normal
human lung tissues and none of the 25 pleural mesothelioma samples obtained from
archival samples were repro-ducibly positive for SV40 DNA (36). Eight
laboratories used a PCR assay and one laboratory used Southern blotting without
prior PCR amplification. Since several negative control samples gave positive
results for SV40 DNA in eight of nine laboratories, one might seriously question
the global outcome of this study. The negative results reported by Shah’s team
may have been caused by the limited sensitivity of the used methodology.
Apparently, Shah acknowledged under oath sensi¬tivity problems that raise
concerns about the validity of the reported findings (37).

Different methods used for DNA extraction may account for the variable detection
rate since the spooling technique used for high-quality DNA extraction from
fresh tissue is not applicable to paraffin-embedded, formalin-fixed tissue
material. Pass and co-workers stressed the advantage of DNA centrifugation at
high speed after phenol-chloroform extraction for efficient isolation of small
molecular weight DNA including episomal SV40 DNA. One must take into account
that some extraction techniques other than phenol-chloroform extraction may
result in less efficient recovery of viral DNA as described for adenovirus DNA
and cytomegalovirus DNA (38,39). Additional differences, such as shorter
protease digestion time for cul¬tured cells in contrast to cells in tissue
blocks, intrinsic differences between cells in culture and cells in tissue, may
influence very strongly the performance of the PCR assay for detecting SV40 DNA.

It is interesting to note that Strickler had used minute amounts of DNA
extracted from individual paraffin wax sections scraped off glass slides.
Estimating an average of 6pg of genomic DNA/diploid cell and isolating 5000
cells out of one 5-mm-thick paraffin section, one can assume that one can
collect 30ng DNA. This means that approximately 20 to 25 sections are needed to
have an optimal amount of DNA.

In contrast, Mulatero et al (40) had relied on antemortem diagnostic biopsy
material from which 200ng of extracted DNA was used per PCR reaction instead of
the more optimal amounts of 500 to 1000pg. The findings of Strickler and
Mulatero are in contradiction with the results reported by Carbone and
coworkers, who found SV40 DNA in 29 of 48 (60%) mesothelioma tumor samples.
However, Carbone and co-workers found their PCR assay was able to detect 1 to 10
genome

copies per PCR reaction in contrast to Strickler, who found a lower sen¬sitivity
level of only 10 to 100 SV40 copies with his technique. Carbone et al used the
same primer sets as Bergsagel, amplifying the Rb pocket binding domain of Tag,
which is the region that binds pRb, p107, and p130/Rb2. Tag expression was
detected by immunohistochemistry and Western blot using the anti-Tag pAb 419,
which is specific for SV40 Tag and does not recognize Tags from BK and JC virus.
Patients’ sera were found to contain Tag antibodies. The finding of Tag-protein
expression in some human mesotheliomas further decreased the possibility of PCR
contamination by commonly used vectors containing SV40 sequences. More than 20
independent research teams have confirmed world¬wide the presence of SV40 or
SV40-like DNA in pleural malignant mseothelioma (41).

initiation of viral DNA replication during lytic growth. In addition, Tag binds
DNA polymerase and the transcription factor AP-2 and forms a specific complex
with the p105 product of the retinoblastoma suppres¬sor gene.

Pab419 (Ab-1) is a mouse monoclonal antibody with specificity for an antigen
localized to the amino terminal domain of the 94-kd SV40 Tag. The sequences
recognized by this reagent are also present on 21-kd SV40 tag. Therefore, the
antibody detects Tag as well as tag and will stain SV40-infected cells. The use
of the antibody is validated for immunoprecipitation and immunofluorescence
(46).

Pab416 (Ab-2) is a mouse monoclonal antibody with specificity for antigenic
determinants unique to the SV40 Tag and nonreactive with SV40 tag. Clone Pab416
is derived by immunization with purified Tag and fusion of mouse spleen cells
with NS-1 mouse myeloma cells. This antibody is reactive to an epitope situated
in the N-terminal region. The antibody is reactive by immunoprecipitation with
the 94-kd SV40 Tag and stains SV40-infected cells (46). Cross-reactivity has
been noted with BK virus Tag. The use of the antibody has been described for
Western blotting, immunoprecipitation, and immunofluorecence. Diffuse nuclear
staining is a characteristic staining pattern when using this antibody in an
immunohistochemical or immunofluorescent technique on frozen tissue sections.

Pab101 is a mouse monoclonal antibody that is recognizing an epitope situated on
the SV40 Tag but does not cross-react with the 21-kd SV40 tag. The antibody
reacts with a denaturation-sensitive determinant on Tag. It reacts with higher
affinity than Pab100 but does recognize a subclass of Tag that is also
recognized by sera from mice bearing SV40-induced tumors. The antibody
precipitates Tag but not tag (nor the smaller SV80 tag) from extracts of SV80
cells and SV40-infected TC-7 cells (47,48). Punctuate nuclear staining is
observed when immunostaining SV40-infected cells by immunofluorescence but not
with BK-infected human cells.

Pab280 (Ab-3) is a mouse monoclonal antibody with specificity for antigenic
determinants unique to the SV40 tag and nonreactive with SV40 Tag. Its binding
site within the unique region of tag was localized by studying its reaction
within SV40 mutants, other papovaviruses, and bacterial expression vectors
coding for fragments of tag. The anti¬body was used to define the cellular
localization of tag by immuno-histochemistry and by immunoprecipitation of
subcellular extracts of

Most investigators have used monoclonal antibodies directed against Tag. Using
the mouse monoclonal antibody pAb419 on frozen sections of human pleural
malignant mesothelioma biopsies Carbone et al (9,17) demonstrated nuclear
expression of the viral Tag in neoplastic mesothelial cells in two separate
series. The immunoreactivity was restricted to the nucleus. Cytoplasmic
immunoreactivity was not noticed. Immunoreactivity in nonneoplastic elements
such as stromal cells or leukocytes was not found. In one of these series the
immuno-histochemical findings were confirmed by specific mRNA in situ
hybridization for Tag. A unique nuclear staining pattern with the monoclonal
antibody was found by Orengo et al (51) in two SV40 DNA–positive newly
established mesothelioma cell lines. The major¬ity of neoplastic cells (90%)
were immunoreactive for Tag without cytoplasmic immunoreactivity. In contrast,
Dhaene et al (52) found only cytoplasmic immunoreactivity in their SV40
DNA–positive fresh frozen mesothelioma tissue biopsies and no nuclear
immunoreactivity with the Pab419.

Investigations with the Pab101 did not reveal any immunoreactivity in the same
mesothelioma cases. These negative results are not so surprising as this
antibody is reactive to an antigenic site located at Tag that is very sensitive
to denaturation (47,48). Similar results were reported by Galateau-Salle et al
(53), working on paraffin-embedded, formalin-fixed mesothelioma tissue biopsies
with the same Pab419 monoclonal antibody. One study describes the use of
monoclonal antibodies Pab416 and Pab101 for detection of SV40 Tag in
paraffin-embedded, formalin-fixed cell block sections from pleural effusions of
32 malignant mesotheliomas. No immunoreactivity was found despite strong
staining of positive controls. The authors stated that the small sample size in
the cytology block sections, the low viral copy number in infected cells, and
the effects of formalin fixation were the reason for their negative
immunohistochemical investigation (54). Only cytoplas-mic immunoreactivity for
SV40 Tag was found in six mesothelioma cell lines with the monoclonal antibodies
Pab419 and Pab101 (55). The authors hypothesized the presence of a contaminating
90-kd protein in the commercially available anti-Tag monoclonal antibodies
Pab419 and Pab101, resulting in false-positive Western blotting results and
immunohistochemistry results. This 90-kd Tag–like protein appeared to be an
artifact caused by the incomplete separation of the heavy

When discussing the puzzling findings of immunohistochemistry for SV40 antigens
in malignant mesothelioma, most authors do not take into account the specificity
of the used monoclonal antibody. The widely used Pab419 monoclonal antibody is
not only reactive to epi-topes situated at Tag but also detects tag. This latter
protein is found not only in the nucleus but also in the cytoplasm of infected
and trans¬formed cells, thereby explaining the observed cytoplasmic
immuno-reactivity with this antibody (44,50).

Another point of consideration is that the use of this antibody has been
validated for immunohistochemistry only on frozen sections but not for
application on paraffin-embedded, formalin-fixed tissues.

Formalin as a cross-linking agent can alter epitopes by rendering them less
accessible to the antibody, thereby necessitating the use of antigen retrieval
methods. Failure to do this may result in false-negative reports. When
addressing the issue of immunoreactivity, one has to take into account that some
epitopes are not resistant to forma¬lin fixation, leading to prevention of
specific binding to its antibody. Application of the antibody on these fixed
tissues can result in false-negative findings. A similar situation has been
found when addressing the presence of the nuclear antigen p53 in nonneoplastic
mesothelial cells. Formalin fixed mesothelial cells do not display any
immunore-activity for p53 (58). In contrast, the same acetone fixed
nonneoplastic mesothelial cells display slight nuclear immunoreactivity with
anti¬bodies directed against p53 (59). Application of antigen retrieval

methods, such as pressure cooking in a citrate buffer or a microwave
pretreatment in the same buffer, is mandatory when investigating the presence of
nuclear antigens such as p53 or hormone receptors for estrogen and progesterone
in paraffin-embedded, formalin-fixed tissues. None of the described studies
reporting negative findings for SV40 protein in paraffin-embedded mesothelioma
tissues addressed this issue properly, thereby invalidating their discussion.

Another effect of prolonged formalin fixation is diffuse cytoplasmic
immunoreactivity for nuclear antigens. Cytoplasmic immunoreactivity has also
been described for other classic nuclear antigens such as c-myc, the cellular
homologue of the v-myc antigen. Fixation times that are longer than 60 to 120
minutes seem to influence the cellular local¬ization of nuclear proteins such as
c-myc and result in cytoplasmic immunoreactivity (60). This phenomenon has also
been described for c-fos and c-myc in paraffin-embedded, formalin-fixed tissue
of pleural malignant mesothelioma (61). It is very remarkable that the studies
by Carbone and Orego reporting solely nuclear immunoreactivity in neo-plastic
mesothelial cells use the same fixation protocol, e.g., short acetone fixation
followed by air drying. The studies reporting solely cytoplasmic staining or no
immunostaining describe another fixation protocol, e.g., air drying for several
hours to overnight followed by acetone fixation. This difference in fixation
protocol might be an expla¬nation for the conflicting and at first sight
contradictory results reported.

Not only are the antibody and the type of material used critical for obtaining
maximal sensitivity and specificity, but also the detection system for
revelation of the bound anti–large T-antibody is important. Most research groups
describe the use of a classic horseradish per-oxidase linked streptavidin-biotin
method (ABC method) (17,51,53). However, some groups report the use of more
sensitive systems such as dextran polymers coated with alkaline phosphatase,
resulting in signal amplification, and the alkaline phosphatase–antialkaline
phos-phatase (APAAP) method or tyramine amplified immunohistochem-istry
(26,52,55). These facts also have to be kept in mind when interpreting the
results of the various immunohistochemical studies reporting the presence of
SV40 antigens in malignant mesothelioma.

Another important point is the use of adequate positive controls when conducting
state-of-the-art immunohistochemical surveys. Most studies use SV40-infected or
transformed cell lines such as SVEC4-10 or WI-26VA4 or mesothelial cells
transfected with SV40. This approach is, in my opinion, scientifically correct,
but the positive control cells or tissues have to be processed in the same way
as the tissues and cells that are under investigation. In my opinion
acetone-fixed, SV40-posi-tive control cells deposited on a glass slide are not
adequate controls for an immunohistochemical survey investigating
paraffin-embedded, formalin-fixed tissues. Appropriate negative controls have to
be included for validating the immunohistochemical procedure. Most studies
describe the omission of the first antibody and replacement by buffer solution.
It is advisable to include a second negative control per case where the primary
antibody has been replaced by another

2. Soriano F, Shelburne E, Gokcen M. Simian virus 40 in a human cancer. Nature
1974;249:421–424.

3. Bergsagel D, Finegold M, Butel J, Kujsky W, Garcea R. DNA sequences similar
to those of SV40 in ependymomas and choroid plexus tumours of childhood. N Engl
J Med 1992;326:988–993.

4. Carbone M, Rizzo P, Pass HI. Simian virus 40, poliovaccines and human cancer:
a review of recent developments. Oncogene 1997;15:1877–1888.

5. Rizzo P, Carbone M, Fisher S, et al. Simian virus 40 is present in most
United States human mesotheliomas but is rarely present in non-Hodgkin
lymphomas. Chest 1999;116:470S-473S.

6. Ferber D. Monkey virus link to cancer grows stronger. Science 2002;296:
1012–1015.

7. Gazdar A, Butel J, Carbone M. SV40 and human tumours: myth, associa¬tion or
causality? Nature Cancer Rev 2002;2:957–964.

8. Morwood K, Price TNC, Mayne LV. Mutation of p53 is not a prerequisite for
immortalization of human fibroblasts by SV40 T antigen. Exp Cell Res
1996;223:308–313.

9. Carbone M, Rizzo P, Grimpley P, et al. Simian virus 40 large T-antigen binds
p53 in human mesotheliomas. Nat Med 1997;3:908–912.

10. De Luca A, Baldi A, Esposito V, Howard C, Bagella L, Rizzo P. The
retino-blastoma gene family pRb/p105, p1087,pRb2/p130 and simian virus 40 T
antigen in human mesotheliomas. Nat Med 1997;8:913–916.

11. Fanning E, Knippers R. Structure and function of simian virus 40 large
tumour antigen. Annu Rev Biochem 1992;61:55–85.

12. Bikel I, Montano X, Agha M, et al. SV40 small T-antigen enhances the
trans¬formation activity of limiting concentrations of SV40 large T-antigen.
Cell 1987;48:321–330.

13. Cicala C, Avantaggiati M, Graessman A, Rundell K, Levine A, Carbone M.
Simian virus 40 small t-antigen stimulates viral DNA replication in per¬missive
monkey cells. J Virol 1994;68:3138–3144.

14. Cicala C, Pompetti F, Nguyen P, Dixon K, Levine A, Carbone M. SV40 small t
deletion mutants preferentially transform mononuclear phagocytes and
B-lymphocytes in vivo. Virology 1992;190:475–479.

15. Cicala C, Pompetti F, Carbone M. SV40 induces mesotheliomas in hamsters. Am
J Pathol 1993;142:1524–1533.

16. Bochetta M, Di Resta I, Powers A, et al. Human mesothelial cells are
unusu¬ally susceptible to simian virus 40 mediated transformation and asbestos
cocarcinogenity. Proc Natl Acad Sci USA 2000;97:10214–10219.

17. Carbone M, Pass H, Rizzo P, et al. Simian virus 40-like DNA sequences in
human pleural mesothelioma. Oncogene 1994;9:1781–1790.

18. Pass H, Rizzo P, Donington J, Wu P, Carbone M. Further validation of SV40
like DNA in human pleural mesotheliomas. Dev Biol Stand 1998;94:143– 145.

19. Griffith D, Nicholson A, Weiss R. Detection of SV40 sequences in human
mesotheliomas. Dev Biol Stand 1998;94:127–136.

20. Lednicky J, Garcea R, Bergsagel D, Butel J. Natural simian virus 40 strains
are present in human choroid plexus and ependymoma tumors. Virology
1995;212:710–717.

21. Lednicky J, Stewart A, Jenkins J, Finegold M, Butel J. SV40 DNA in human
osteosarcomas shows sequence variation among T-antigen genes. Int J Cancer
1997;72:791–800.

22. Mendoza S, Konishi T, Miller C. Integration of SV40 in human osteosar-coma.
Oncogene 1998;17:2457–2462.

23. Pacini F, Vivaldi A, Santoro M, et al. Simian virus 40-like DNA in human
papillary thyroid carcinomas. Oncogene 1998;16:665–669.

24. zur Hausen H. Papillomavirus infections—a major cause of human cancers.
Biochim Biophys Acta 1996;1288:F55–F78.

25. Kleter B, van Doorn J, ter Schegge J. Novel short fragment PCR assay for
highly sensitive broad spectrum detection of anogenital human papillo-maviruses.
Am J Pathol 1998;53:1731–1739.

26. Ramael M, Nagels J, Heylen H, De Schepper S, Paulussen J, Van Haesendonck C.
Detection of SV40 like viral DNA and viral antigens in malignant pleural
mesothelioma. Eur Res J 1999;14:1381–1386.

27. Koch J, Kolvraa S, Peetersen P, Gregersen N, Boland L. Oligonucleotide
priming methods for the chromosome specific labeling of alpha satellite DNA in
situ. Chromosoma (Berlin) 1989;98:259–265.

28. Ramael M, Van Steelandt H, Stuyven G, Van Steenkiste M, Degroote J.
Application of the primed in situ labeling (PRINS) method for detection of
numerical chromosomal aberrations in paraffin embedded formalin fixed tissue of
molar and non-molar pregnancies. Biochemica 1997;2:18– 20.

29. Strickler H, Goedert JJ, Fleming M, et al. Simian virus 40 and pleural
mesothelioma in humans. Cancer Epidemiol Biomarkers Prev 1996;5:473– 475.

30. Testa J, Carbone M, Hirvonen A, et al. A multi-institutional study confirms
the presence and expression of simian virus 40 in human malignant mesotheliomas.
Cancer Res 1998;58:4505–4509.

31. Hirvonen A, Mattson K, Karjalainen A, et al. Simian virus 40 (SV40) like DNA
sequences not detectable in Finnish mesothelioma patients not exposed to
SV40-contaminated polio vaccines. Mol Carcinogen 1999;26:93– 99.

32. Jasani B, Cristaudo A, Emri S, et al. Association of SV40 with human
tumours. Semin Cancer Biol 2001;11:49–61.

33. Shivapurkar N, Wiethege T, Wishuba II, et al. Presence of simian virus 40
sequences in malignant mesotheliomas and mesothelial cell proliferations. J Cell
Biochem 1999;76:181–188.

34. Cristaudo A, Powers A, et al. SV40 can be reproducibly detected in paraf¬fin
embedded mesothelioma samples. Anticancer Res 2000;20:895–898.

35. Cristaudo A, Vivaldi A, Sensales G, et al. Molecular biology studies on
mesothelioma tumor samples: preliminary data on H-ras, p21, and SV40. J Environ
Pathol Toxicol Oncol 1995;14:29–34.

36. Strickler H, et al. A multicenter evaluation of assays for detection of SV40
DNA and results in masked mesothelioma specimens. Cancer Epidemiol Biomarkers
Prev 2001;10:523–532.

37. MacLachlan D. SV40 in human tumours: new documents shed light on the
apparent controversy. Anticancer Res 2002;22:957–964.

38. Fahle G, Fisher S. Comparison of six commercial DNA extraction kits for
recovery of cytomegalovirus DNA from spiked human specimens. J Clin Microbiol
2000;38:3860–3863.

39. Henderson Y, Liu T, Clayman G. A simple and sensitive method for detect¬ing
adenovirus in serum and urine. J Virol Methods 1998;71:51–56.

40. Mulatero C, Surentheran T, Breuer J, Rudd R. Simian virus 40 and human
pleural mesotheliomas. Thorax 1999;54:60–61.

41. Klein G, Powers A, Croce C. Association of SV40 with human tumors. Oncogene
2002;21:1141–1149.

42. McLaren B, Haenel T, Stevenson S, Mukerjee S, Robinson B, Lake R. SV40 like
sequences in cell lines and tumour biopsies from Australian malignant
mesotheliomas. Aust NZ J Med 2000;30:450–456.

43. Greer C, Peterson S, Kiviat N, Manos M. PCR amplification of paraffin
embedded tissues: effects of fixative and fixation time. Am J Clin Pathol
1991;95:117–124.

44. Ellman M, Bikel I, Figge J, Roberts I, Schlossman R, Livinston D.
Localiza¬tion of the simian virus 40 small t-antigen in the nucleus and
cytoplasm of monkey and mouse cells. J Virol 1984;50:623–627.

45. Tieman F, Zerrahn J, Deppert W. Cooperation of SV40 large and small
T-antigens in metabolic stabilization of tumor suppressor p53 during cel¬lular
transformation. J Virol 1995;69:6115–6121.

46. Harlow E, Crawford L, Pim D, Williamson N. Monoclonal antibodies spe¬cific
for simian virus 40 tumour antigens. J Virol 1981;39:861–869.

47. Deppert W. Monoclonal antibodies against SV40 tumour antigens: analy¬sis of
antigenic binding sites, using adenovirus 2-simian virus 40 hybrid viruses. J
Virol 1981;37:478–482.

48. Gurney E. Monoclonal antibodies against simian virus 40 T-antigens:
evi¬dence for distinct subclasses of large T-antigen and for similarities among
nonviral T-antigens. J Virol 1980;34:752–763.

49. Montano X, Lane DP. Monoclonal antibody analysis of simian virus 40 small
t-antigen. J Virol 1984;51:760–767.

50. Montano X, Lane DP. Monoclonal antibody analysis of simian virus 40 small
t-antigen in infected and transformed cells. J Virol 1989;63:3128– 3134.

51. Orengo A, Spoletini L, Procopio A, et al. Establishment of four new
mesothelioma cell lines: characterization by ultrastructural and
immuno-phenotypic analysis. Eur Respir J 1999;13:527–534.

52. Dhaene K, Verhulst A, Van Marck E. SV40 large T-antigen and human pleural
mesothelioma. Screening by tyramine amplified immunohisto-chemistry. Virchows
Arch 1999;435:1–7.

53. Galateau-Salle F, Bidet Ph, Iwatsubo Y, et al. SV40-like DNA sequences in
pleural mesothelioma, bronchopulmonary carcinoma and non-malignant pulmonary
disease. J Pathol 1998;184:252–257.

54. Simsir A, Fetsch P, Bedrossian C, Ioffe O, Abati A. Absence of SV40 large
T-antigen in malignant mesothelioma effusions: an immunocytochemical study.
Diagn Cytopathol 2001;25:203–207.

55. Pilatte Y, Vivo C, Renier A, Kheuang L, Greffard A, Jaurand MC. Absence of
SV40 large T-antigen expression in human mesothelioma cell lines. Am J Respir
Cell Mol Biol 2000;23:788–793.

56. Wynnford-Thomas D. P53 in tumour pathology: can we trust
immunohis-tochemistry? J Pathol 1992;166:329–330.

57. Crawford L, Leppard K, Lane D, Harlow E. Cellular proteins reactive with
monoclonal antibodies directed against simian virus 40 T-antigen. J Virol
1982;42:612–620.

58. Ramael M, Eerdekens C, Lemmens G, Deblier I, Jacobs W, Van Marck E.
Immunoreactivity for p53 protein in neoplastic and non-neoplastic mesothelial
tissues. J Pathol 1992;168:371–375.

59. Dowell S, Derias N, Wilson P, Lane D, Hall P. Expression of p53 in reactive
mesothelium. Histopathology 1993;22:96–97.

60. Loke S, Neckers L, Schwab G, Jaffe S. c-Myc protein in normal tissue:
effects of fixation on its apparent subcellular distribution. Am J Pathol
1988;131: 29–37.