Re: Lack of serologic evidence for prevalent simian virus 40 infection in humans.

Full text

Page 1

Re: Lack of Serologic Evidence
for Prevalent Simian Virus 40
Infection in Humans
We read with interest the recent arti-
cle by Carter et al. (1), in which human
sera were screened for antibodies to sim-
ian virus 40 (SV40). Sera were screened
for SV40 antibodies by an enzyme-
linkedimmunosorbent
SV40 VP1 virus-like particles (VLPs) as
antigens. Although
SV40-positive human sera (6.6%) in the
Carter et al. study (1) is in agreement
with the results obtained in other inves-
tigations (2–6), Carter et al. report that
SV40 reactivity in human sera disap-
peared after serum preadsorption with
VLPs of BK virus (BKV) and JC virus
assayusing
the fractionof
Journal of the National Cancer Institute, Vol. 96, No. 10, May 19, 2004CORRESPONDENCE 803
at University of Portland on May 24, 2011
jnci.oxfordjournals.org
Downloaded from

Page 2

(JCV), two ubiquitous human polyoma-
viruses that are related to SV40. As a
consequence, Carter et al. concluded
that the SV40 antibodies detected in hu-
man sera were not authentic SV40 anti-
bodies but were BKV and/or JCV anti-
bodies that cross-react with SV40.
The article by Carter et al. (1) de-
serves some comments. First, due to the
sequence homology of the VP1 struc-
tural proteins (i.e., more than 80% iden-
tical) of the three polyomaviruses, BKV,
JCV, and SV40, it is not surprising that
the vast majority of the human antibody
repertoire against the VP1 protein of the
three polyomaviruses can largely over-
lap. This cross-reactivity, in turn, in-
creases the probability that immuno-
dominantepitopes
within the family of VP1 antigenic de-
terminants common to all three poly-
omaviruses.Coexistence
nodominance and cross-reactivity has
been largely documented in the litera-
ture (7). Thus, both qualitative and
quantitative differences in the antibody
response to the epitopes present only in
the SV40 VP1 structural protein may be
difficult to determine, particularly if an
assay based on serum preadsorption
with BKV and JCV VLPs is used as the
sole means of detection.
Second, there is variability in the an-
tibody response to SV40 among individ-
uals. For example, in a study (6) of
human sera from children under 5 years
of age in Morocco who were affected by
poliomyelitis, 100% of the serum sam-
ples contained antibodies to SV40.
However, in the same study, human sera
from SV40-unaffected individuals in
Morocco, Zaire, Sierra Leone, and Po-
land contained antibodies to SV40 in
only 0.4%–5.3% of the samples, a per-
centage that is comparable to the results
of other studies (1–5). This observation
suggests that, under particular circum-
stances, humans can display a strong
and specific antibody response to SV40.
Third, because of the ubiquity of the
two human polyomaviruses, BKV and
JCV, and because of the antigenic cross-
reactivity in the structural proteins of the
three polyomaviruses, BKV, JCV, and
SV40, it is very difficult to find human
sera that are positive for antibodies to
SV40 and negative for antibodies to
BKV and JCV. However, a human se-
rum sample with antibodies only to
SV40 has been detected in one study (6),
arerepresented
ofimmu-
suggesting that specific SV40 antibodies
can be produced by humans.
Fourth, although seroconversion to
BKV and JCV is age dependent (5),
there is no detectable age-dependent se-
roconversion to SV40 (1,5,6), suggest-
ing that most of the SV40 antibodies
detected in human sera are not generated
by infection with BKV or JCV.
Fifth, in human sera from two immu-
nosuppressed renal transplant patients
that were examined sequentially for an-
tibodies to BKV, JCV, and SV40 over a
period of 82 and 51 weeks, respectively,
a substantial rise in SV40 antibody titers
was detected after the kidney transplant
(6), suggesting that a latent SV40 infec-
tion, similar to a latent BKV and JCV
infection, can be reactivated in humans
by immunosuppression. Moreover, dur-
ing the follow-up, the curve of antibody
titers to SV40 was different from that of
antibody titers to either BKV or JCV
(6), suggesting a specific immunologic
response to SV40 in these two patients.
In conclusion, although antibody
cross-reaction can occur during the im-
munologic response of humans to poly-
omavirus infection,
specific antibodies can be detected in
human sera. SV40-specific antigens,
such as peptides that include SV40
structural epitopes that do not cross-
react with BKV and JCV capsid anti-
gens, should be used to detect the
antibody-specific response to SV40 in-
fection in the human population.
unique SV40-
GIUSEPPE BARBANTI-BRODANO
ALFREDO CORALLINI
ROBERTO S. ACCOLLA
FERNANDA MARTINI
MAURO TOGNON
REFERENCES
(1) Carter JJ, Madeleine MM, Wipf GC, Garcea
RL, Pipkin PA, Minor PD, et al. Lack of
serologic evidence for prevalent simian virus
40 infection in humans. J Natl Cancer Inst
2003;95:1522–30.
(2) Shah KV, Ozer HL, Pond HS, Palma LD,
Murphy GP. SV40 neutralizing antibodies in
sera of US residents without history of polio
immunization. Nature 1971;231:448–9.
(3) Jafar S, Rodriguez-Barradas M, Graham DY,
Butel JS. Serological evidence of SV40 infec-
tions in HIV-infected and HIV-negative
adults. J Med Virol 1998;54:276–84.
(4) De Sanjose S, Shah KV, Domingo-Domenech
E, Engels EA, Fernandez de Sevilla AF, Al-
varo T, et al. Lack of serological evidence for
an association between simian virus 40 and
lymphoma. Int J Cancer 2003;104:522–4.
(5) Knowles WA, Pipkin P, Andrews N, Vyse A,
Minor P, Brown DW, et al. Population-based
study of antibody to the human polyomavi-
ruses BKV and JCV and the simian polyoma-
virus SV40. J Med Virol 2003;71:115–23.
(6) Minor P, Pipkin P, Jarzebek Z, Knowles WA.
Studies of neutralising antibodies to SV40 in
human sera. J Med Virol 2003;70:490–5.
(7) Berzofsky JA, Berkover IJ. Immunogenicity
and antigen structure. In: Paul WE, editor.
Fundamental immunology, 4th edition. Phila-
delphia (PA): Lippincott Raven; 1999. p.
651–99.
NOTES
Affiliations of authors: Department of Experi-
mental and Diagnostic Medicine and Center of
Biotechnology (GBB, AC), Center of Biotechnol-
ogy and Department of Morphology and Embry-
ology (FM, MT), University of Ferrara, Ferrara,
Italy; Department of Clinical and Biological Sci-
ences (RSA), University of Insubria, Varese, Italy.
Correspondence to: Mauro Tognon, PhD, De-
partment of Morphology and Embryology, Chair
of Applied Biology, School of Medicine, Univer-
sity of Ferrara, Via Fossato di Mortara 64/B,
44100 Ferrara, Italy (e-mail: tgm@unife.it).
DOI: 10.1093/jnci/djh151
RESPONSE
We thank Barbanti-Brodano et al. for
their interest in our recent manuscript
(1). We set out to examine the hypoth-
esis that antibodies to simian virus 40
(SV40) would be prevalent in the popu-
lation; given that so many tumor sites
have been reported to harbor SV40
DNA, even in individuals too young to
have been exposed to SV40, one predic-
tion would be that SV40 is a commonly
circulating virus in humans. There ap-
pears to be general agreement that,
whereas antibodies to BK virus (BKV)
and JC virus (JCV) are frequent, i.e.,
present in more than 50% of the popu-
lation (1–3), only a small percentage of
various populations examined have an-
tibodies to SV40, and in these popula-
tions seropositivity to SV40 is highly
correlated with seropositivity to BKV
and JCV.
We disagree with Barbanti-Brodano
et al.’s suggestion that this correlation is
to be expected because of the high de-
gree of sequence homology among the
polyomavirus capsid proteins (VP1s).
First, cross-reactivity between BKV and
JCV has not been observed in several
studies; in fact, very different patterns of
seroreactivity to BKV and JCV have
been observed that are related to age
(1–3). Second, most human sera that are
804 CORRESPONDENCEJournal of the National Cancer Institute, Vol. 96, No. 10, May 19, 2004
at University of Portland on May 24, 2011
jnci.oxfordjournals.org
Downloaded from

Page 3

positive for JCV or BKV antibodies do
not cross-react with SV40; conversely,
the majority of macaque sera that are
seropositive for SV40 do not cross-react
with BKV or JCV. Third, human papil-
lomaviruses (HPVs) have a T?7 icosa-
hedral virion that is similar to that of the
polyomaviruses. The capsid proteins
(L1s) of HPVs are closely related, yet
HPVs elicit type-specific serologic re-
sponses (4,5). Structural studies of the
HPV capsid (or virus-like particles
[VLPs]) have shown that conserved res-
idues of L1 form the internal core of the
capsid, whereas residues that vary
among HPV types are exposed on the
surface (6). These type-specific residues
form the immunoreactive epitopes that
are seen in natural infections with HPV
and in experimental infections with
VLPs. Broadly cross-reactive antibodies
to HPV can be generated only in re-
sponse to denatured L1 protein. Thus,
the fact that polyomavirus VP1 proteins
share extensive homology reflects their
conserved virion structure, not their
shared antigenicity.
The authors cite one study (7) that
examined more than 2000 sera from
several geographic areas and found neu-
tralizing antibodies to SV40 in 3–5% of
the sera, although two small sets of sera
showed greatervariation
100%). The meaning of finding 100%
seropositivity to SV40 in Moroccan
children with poliomyelitis is unclear.
Given the generally negative findings of
SV40 seropositivity, this result would
need to be confirmed before concluding
that the sera had antibodies specific for
SV40. We agree with the authors that
there is no apparent relationship be-
tween SV40 seropositivity and age, un-
like the distinct patterns seen for JCV
and BKV (1,3), but we question what
that result means. If the SV40 antibodies
are the response to infection with SV40,
when does the infection occur?
At this time, serological testing for
antibodies to polyomaviruses is a young
field. It is not clear why some people
have antibodies that cross-react between
SV40 and human polyomaviruses. How-
ever, certain findings about SV40 anti-
bodies are consistent: 1) anti-SV40 an-
tibodies in humans are infrequent and
are of low titer, whereas BKV and JCV
antibodies are frequent and of high titer,
like anti-SV40 antibodies in macaques
(1–3); 2) SV40 antibodies are not asso-
ciated with poliovirus vaccination his-
(0% and
tory or age; 3) SV40 antibodies are not
associated with osteosarcoma in chil-
dren in the United States (1) nor with
non-Hodgkinlymphoma
from Spain (2). It has been suggested (3)
that SV40 cross-reactive antibodies oc-
cur in individuals who have very high
titers of BKV or JCV infection. Perhaps
cross-reactive epitopes are seen in re-
sponse to high viral loads or certain
types of infection or occur in individuals
with a particular immunogenetic back-
ground. Or perhaps they occur because a
small number of people have been in-
fected by SV40. All of these possibili-
ties await further research.
inpatients
DENISE A. GALLOWAY
JOSEPH J. CARTER
REFERENCES
(1) Carter JJ, Madeleine MM, Wipf GC, Garcea
RL, Pipkin PA, Minor PD, et al. Lack of
serologic evidence for prevalent simian virus
40 infection in humans. J Natl Cancer Inst
2003;95:1522–30.
(2) de Sanjose S, Shah KV, Domingo-Domenech
E, Engels EA, Fernandez de Sevilla A, Garcia-
Villanueva M, et al. Lack of serological evi-
dence for an association between simian virus
40 and lymphoma. Int J Cancer 2003;104:
522–4.
(3) Knowles WA, Pipkin P, Andrews N, Vyse A,
Minor P, Brown DW, et al. Population-based
study of antibody to the human polyomavi-
ruses BKV and JCV and the simian polyoma-
virus SV40. J Med Virol 2003;71:115–23.
(4) Carter JJ, Koutsky LA, Hughes JP, Lee SK,
Kuypers J, Kiviat N, et al. Comparison of
human papillomavirus types 16, 18, and 6
capsid antibody responses following incident
infection. J Infect Dis 2000;181:1911–9.
(5) Roden RB, Hubbert NL, Kirnbauer R, Chris-
tensen ND, Lowy DR, Schiller JT. Assess-
ment of the serological relatedness of genital
human papillomaviruses by hemagglutination
inhibition. J Virol 1996;70:3298–301.
(6) Chen XS, Garcea RL, Goldberg I, Casini G,
Harrison SC. Structure of small virus like par-
ticles assembled from the L1 protein of human
papillomavirus 16. Mol Cell 2000;5:557–67.
(7) Minor P, Pipkin P, Jarzebek Z, Knowles W.
Studies of neutralising antibodies to SV40 in
human sera. J Med Virol 2003;70:490–5.
NOTES
Affiliation of authors: Fred Hutchinson Cancer
Research Center, Seattle, WA (DAG, JJC).
Correspondence to: Denise A. Galloway, PhD,
Program Head of Cancer Biology, Fred Hutchin-
son Cancer Research Center, 1100 Fairview Ave.
N., Rm. C-105, Seattle, WA 98109-1024 (e-mail:
dgallowa@fhcrc.org).
DOI: 10.1093/jnci/djh152
Journal of the National Cancer Institute, Vol. 96, No. 10, May 19, 2004CORRESPONDENCE 805
at University of Portland on May 24, 2011
jnci.oxfordjournals.org
Downloaded from