- Email: [email protected]
JC virus in human glial-derived tumors
Case Studies JC VIRUS IN HUMAN GLIAI.-DERIVED TUMORS RITA CALDARELLI-STEFANO,RENZO BOLDORINI, MD, GUIDO MONGA, MD, EMANUELAMERAVIGLIA, ELISABETTAOMODEO ZORINI, AND PASQUALEFERRANTE, MD
To investigate the presence and the role of polyomaviruses JC (JCT), BK (BKV),and the simian polyomavirus (SV40) in human brain tumors, samples from 25 glial-derived tumors (10 astrocytomas, 5 ependymomas, 5 oligodendrogliomas, and 5 glioblastomas) were examined by means of molecular biology and immunohistochemistry. Nested PCR of the large T (LT) region and its sequence analysis showed JCV in 6 cases (4 astrocytomas, 1 oligodendroglioma, and 1 ependymoma), while the transcriptional control region (TCR) was amplified ordy in 1 astrocytoma, the oligodendrogiioma, and the ependymoma, one of which (astrocytoma) also stained positively by immunohistochemistry (ICV LT). TCR sequence analysis of the
oligodendroglioma showed a JCV rearranged structure not related to a known viral strain, while the astrocytoma and the ependymoma disclosed a JCV Mad.4 strain that is known to induce brain tumors in animals. We suggest that JCV could have played a role in the pathogenesis of these brain tumors. HUM PATHOL 31:394-395. Copyright © 2000 by W.B. Saunders Company Key words:brain tumors, polyomavirus, molecular biology. Abbreviations: LT, large T; TCR, transcriptional control region; JCV, JC polyomavirus; BKV, BK polyomavirus; SV40, simian polyomavirus; n-PCR, nested polymerase chain reaction; bp, base pair.
Human polyomaviruses JC (]CV), BK (BKV), and the simian polyomavirus (SV40) are known to induce brain tumors after intracerebral inoculation in experimental animals. ] Moreover, SV40 and BKV DNA has been found in various types of human lung, bone, and brain tumors, a n d J C virus genome has also been recently isolated from human glial tumors. 24 On the other hand, there is evidence that different JCV strains, identified by means of sequence analysis of their transcriptional control region (TCR), have different pathogenic potential; Mad-1 has been isolated from both normal brain and brain with progressive multifocal leukoencephalopathy, whereas Mad-4 has been shown to be associated with human and animal glial tumors. 1
sought in all of the positive samples, and sequence analysis was also performed. DNA belonging to the JCV TCR was positive in only the following 3 tumors: the oligodendroglioma, the ependymoma, and 1 of the astrocytomas. Nucleotide sequence analysis of the TCR region isolated from the astrocytoma and ependymoma showed the following identical rearrangements (Fig 1): a first TATA-box (block A), a sequence of 55 bp (block C) followed by an IS-base pair (bp) fragment (block E) repeated in tandem, separated by the sequence "5"-GGAAGG"3." These findings indicated the presence of a Mad-4 JCV strain. The TCR sequence amplified from the oligodendroglioma had a rearranged structure in comparison with the Mad-1 and archetype strains (Fig 1): one 21-bp insertion (block B) and 2 others of 10 bp (blocks D) in addition to the A, C, and E blocks. Paraffin-embedded sections of all the 25 tumor samples were immunohistochemically examined with an antibody that recognizes the large T antigen of the JC virus. Positively stained nuclei were only found in a few of the neoplastic ceils of the astrocytoma whose TCR region was identified by n-PCR (Fig 2).
METHODS AND RESULTS To investigate the presence of polyomaviruses and their possible association with human brain tumors, we retrospectively examined surgical samples from glial-derived tumors of 25 patients (13 women and 12 men; mean age, 52 years; range, 1%76) with no sign of immunosuppression: 10 astrocytomas, 5 ependymomas, 5 glioblastomas, and 5 oligodendrogliomas, all of which were histologically diagnosed. DNA was extracted from paraffin-embedded tissue, and nested polymerase chain reaction (n-PCR) for the large T (LT) antigen region showed LT polyomavirus sequences in 4 astrocytomas, 1 ependymoma, and 1 oligodendroglioma. All of these samples were proved to be JCV-positive by means of direct nucleotide sequencing. In order to differentiate the JCV strains, the TCR was
From Fondazione Don C. Gnocchi, IRCCS, Cattedra di Virologia, Universit~ di Milano, Milan, Italy; and Servizio di Anatomia Patologica Ospedale Maggiore della Carit~tand Dipartimento di Scienze Mediche, Universit~ del Piemonte Orientale "Amedeo Avogadro," Novara, Italy. Address correspondence and reprint requests to Renzo Boldorini MD, Servizio di Anatomia Patologica, Ospedale Maggiore della Carit~t, Corso Mazzini 18, 28100 Novara, Italy. Copyright © 2000 by W.B. Saunders Company 0046-8177/00/3103-0020510.00/0 doi: 10.1053/hp.2000.5231
394
DISCUSSION The link between the presence of polyomavirus and brain tumor development is still controversial. The brain is considered to be one of the sites of JCV latency but, as genomic sequences of the Mad-1 viral strain can also be found in normal brains, the detection ofJCV genome is not enough to support viral involvement in the development of brain tumors. 5 However, it is known that Mad-1 can undergo viral rearrangements in the TCR region and therefore generate mutated strains, one of which (Mad-4) has been found to have oncogenic properties in experimental animals. 1 Two of our cases (1 astrocytoma and 1 ependymoma) showed rearranged sequences typical of the Mad-4 strain in the viral TCR region, and 1 (the astrocytoma) stained positively with antibody againstJC virus T antigen. AlthoughJCV was detected in only a small number of our cases, it can be speculated that the virus did play a role in the pathogenesis of these 2 brain tumors (2 of 25, 8%). As far as the oligodendroglioma is concerned, the
CASE STUDIES
ori
A
.......
Archetype FIGURE 1. Genomic organization of the regulatory region (TCR) amplified from human brain tumor tissues in comparison with the archetype and Mad- 1 prototypes. The TCR amplified from the astrocytoma and ependymama have an identical organization that is similar to that of Mad- 1 but very different from the almost unique organization found in the oligodendroglioma.
B
ori
A
C ~~.~..~ ,,~. . . .
,
C
Mad-1
D ,
E
E
..~
.~
A
F1
~
~
C
::
E
F1
® ori
A
C
E
A
C
E
F1
ori
A
C
E
A
C
E
F1
ori
A
Astrocytoma Ependymoma Oiigodendroglioma
B
C
D
E
C
D
E
F1
L
FIGURE2. Immunohistochemical staining with a polyclonal antibody against JC virus T antigen shows nuclear positivity only in few astrocytes (arrowheads) of a fibrillary astrocytoma. (Haematoxylin counterstain, original magnification x250.)
2. Khalili K, Krynska B, del Valle L, et al: Medulloblastoma a n d the h u m a n neurotropic polyomavirusJC virus. Lancet 353:1152-1153, 1999 3. Boldorini R, Caldarelli-Stefano R, Monga G, et al: PCR detection of JC virus DNA in the brain tissue of a 9-year-old child with pleomorphic xanthoastrocytoma.J Neurovirol 4:2424-245, 1998 4. Rencic A, G o r d o n J, Otte J, et al: Detection of JC virus DNA sequence a n d expression of the viral oncoprotein, t u m o r antigen, in brain o f i m m u n o c o m petent patient with oligoastrocytoma. Proc Natl Acad Sci U S A 93:7352-7357, 1996 5. Gallia GL, Gordon J, Khalfli K: T u m o r pathogenesis of h u m a n neurotropicJC virus in the CNS.J Neurovirol 4:175-181, 1998
rearranged ]GV sequence (which is not strictly related to a viral strain) has not yet been associated with the presence of neoplasia, but its oncogenic potential cannot be excluded "a priori" and is worthy of further investigation. REFERENCES 1. Padgett BL, Walker DL, ZuRhein GM, et al: Differential neurooncogenicity of strains of JC virus, a h u m a n polyoma virus, in newborn syrian hamsters. Cancer Res 37:718-720, 1997
395
To investigate the presence and the role of polyomaviruses JC (JCT), BK (BKV),and the simian polyomavirus (SV40) in human brain tumors, samples from 25 glial-derived tumors (10 astrocytomas, 5 ependymomas, 5 oligodendrogliomas, and 5 glioblastomas) were examined by means of molecular biology and immunohistochemistry. Nested PCR of the large T (LT) region and its sequence analysis showed JCV in 6 cases (4 astrocytomas, 1 oligodendroglioma, and 1 ependymoma), while the transcriptional control region (TCR) was amplified ordy in 1 astrocytoma, the oligodendrogiioma, and the ependymoma, one of which (astrocytoma) also stained positively by immunohistochemistry (ICV LT). TCR sequence analysis of the
oligodendroglioma showed a JCV rearranged structure not related to a known viral strain, while the astrocytoma and the ependymoma disclosed a JCV Mad.4 strain that is known to induce brain tumors in animals. We suggest that JCV could have played a role in the pathogenesis of these brain tumors. HUM PATHOL 31:394-395. Copyright © 2000 by W.B. Saunders Company Key words:brain tumors, polyomavirus, molecular biology. Abbreviations: LT, large T; TCR, transcriptional control region; JCV, JC polyomavirus; BKV, BK polyomavirus; SV40, simian polyomavirus; n-PCR, nested polymerase chain reaction; bp, base pair.
Human polyomaviruses JC (]CV), BK (BKV), and the simian polyomavirus (SV40) are known to induce brain tumors after intracerebral inoculation in experimental animals. ] Moreover, SV40 and BKV DNA has been found in various types of human lung, bone, and brain tumors, a n d J C virus genome has also been recently isolated from human glial tumors. 24 On the other hand, there is evidence that different JCV strains, identified by means of sequence analysis of their transcriptional control region (TCR), have different pathogenic potential; Mad-1 has been isolated from both normal brain and brain with progressive multifocal leukoencephalopathy, whereas Mad-4 has been shown to be associated with human and animal glial tumors. 1
sought in all of the positive samples, and sequence analysis was also performed. DNA belonging to the JCV TCR was positive in only the following 3 tumors: the oligodendroglioma, the ependymoma, and 1 of the astrocytomas. Nucleotide sequence analysis of the TCR region isolated from the astrocytoma and ependymoma showed the following identical rearrangements (Fig 1): a first TATA-box (block A), a sequence of 55 bp (block C) followed by an IS-base pair (bp) fragment (block E) repeated in tandem, separated by the sequence "5"-GGAAGG"3." These findings indicated the presence of a Mad-4 JCV strain. The TCR sequence amplified from the oligodendroglioma had a rearranged structure in comparison with the Mad-1 and archetype strains (Fig 1): one 21-bp insertion (block B) and 2 others of 10 bp (blocks D) in addition to the A, C, and E blocks. Paraffin-embedded sections of all the 25 tumor samples were immunohistochemically examined with an antibody that recognizes the large T antigen of the JC virus. Positively stained nuclei were only found in a few of the neoplastic ceils of the astrocytoma whose TCR region was identified by n-PCR (Fig 2).
METHODS AND RESULTS To investigate the presence of polyomaviruses and their possible association with human brain tumors, we retrospectively examined surgical samples from glial-derived tumors of 25 patients (13 women and 12 men; mean age, 52 years; range, 1%76) with no sign of immunosuppression: 10 astrocytomas, 5 ependymomas, 5 glioblastomas, and 5 oligodendrogliomas, all of which were histologically diagnosed. DNA was extracted from paraffin-embedded tissue, and nested polymerase chain reaction (n-PCR) for the large T (LT) antigen region showed LT polyomavirus sequences in 4 astrocytomas, 1 ependymoma, and 1 oligodendroglioma. All of these samples were proved to be JCV-positive by means of direct nucleotide sequencing. In order to differentiate the JCV strains, the TCR was
From Fondazione Don C. Gnocchi, IRCCS, Cattedra di Virologia, Universit~ di Milano, Milan, Italy; and Servizio di Anatomia Patologica Ospedale Maggiore della Carit~tand Dipartimento di Scienze Mediche, Universit~ del Piemonte Orientale "Amedeo Avogadro," Novara, Italy. Address correspondence and reprint requests to Renzo Boldorini MD, Servizio di Anatomia Patologica, Ospedale Maggiore della Carit~t, Corso Mazzini 18, 28100 Novara, Italy. Copyright © 2000 by W.B. Saunders Company 0046-8177/00/3103-0020510.00/0 doi: 10.1053/hp.2000.5231
394
DISCUSSION The link between the presence of polyomavirus and brain tumor development is still controversial. The brain is considered to be one of the sites of JCV latency but, as genomic sequences of the Mad-1 viral strain can also be found in normal brains, the detection ofJCV genome is not enough to support viral involvement in the development of brain tumors. 5 However, it is known that Mad-1 can undergo viral rearrangements in the TCR region and therefore generate mutated strains, one of which (Mad-4) has been found to have oncogenic properties in experimental animals. 1 Two of our cases (1 astrocytoma and 1 ependymoma) showed rearranged sequences typical of the Mad-4 strain in the viral TCR region, and 1 (the astrocytoma) stained positively with antibody againstJC virus T antigen. AlthoughJCV was detected in only a small number of our cases, it can be speculated that the virus did play a role in the pathogenesis of these 2 brain tumors (2 of 25, 8%). As far as the oligodendroglioma is concerned, the
CASE STUDIES
ori
A
.......
Archetype FIGURE 1. Genomic organization of the regulatory region (TCR) amplified from human brain tumor tissues in comparison with the archetype and Mad- 1 prototypes. The TCR amplified from the astrocytoma and ependymama have an identical organization that is similar to that of Mad- 1 but very different from the almost unique organization found in the oligodendroglioma.
B
ori
A
C ~~.~..~ ,,~. . . .
,
C
Mad-1
D ,
E
E
..~
.~
A
F1
~
~
C
::
E
F1
® ori
A
C
E
A
C
E
F1
ori
A
C
E
A
C
E
F1
ori
A
Astrocytoma Ependymoma Oiigodendroglioma
B
C
D
E
C
D
E
F1
L
FIGURE2. Immunohistochemical staining with a polyclonal antibody against JC virus T antigen shows nuclear positivity only in few astrocytes (arrowheads) of a fibrillary astrocytoma. (Haematoxylin counterstain, original magnification x250.)
2. Khalili K, Krynska B, del Valle L, et al: Medulloblastoma a n d the h u m a n neurotropic polyomavirusJC virus. Lancet 353:1152-1153, 1999 3. Boldorini R, Caldarelli-Stefano R, Monga G, et al: PCR detection of JC virus DNA in the brain tissue of a 9-year-old child with pleomorphic xanthoastrocytoma.J Neurovirol 4:2424-245, 1998 4. Rencic A, G o r d o n J, Otte J, et al: Detection of JC virus DNA sequence a n d expression of the viral oncoprotein, t u m o r antigen, in brain o f i m m u n o c o m petent patient with oligoastrocytoma. Proc Natl Acad Sci U S A 93:7352-7357, 1996 5. Gallia GL, Gordon J, Khalfli K: T u m o r pathogenesis of h u m a n neurotropicJC virus in the CNS.J Neurovirol 4:175-181, 1998
rearranged ]GV sequence (which is not strictly related to a viral strain) has not yet been associated with the presence of neoplasia, but its oncogenic potential cannot be excluded "a priori" and is worthy of further investigation. REFERENCES 1. Padgett BL, Walker DL, ZuRhein GM, et al: Differential neurooncogenicity of strains of JC virus, a h u m a n polyoma virus, in newborn syrian hamsters. Cancer Res 37:718-720, 1997
395