Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Journal of General Virology (1998), 79, 789–799. Printed in Great Britain ................................................................................................................................................................................................................................................................................... Human polyomavirus JC control region variants in persistently infected CNS and kidney tissue Cornelia Elsner and Kristina Do$ rries Institut fu$ r Virologie und Immunbiologie der Universita$ t Wu$ rzburg, Versbacher Str. 7, D-97078 Wu$ rzburg, Germany The question of a possible role for JC virus (JCV) genomic rearrangements in the pathogenesis of progressive multifocal leukoencephalopathy (PML) was addressed by analysis of the genomic complexity and the transcriptional control region (TCR) of the JCV DNA population in persistently infected CNS and kidney tissue. After cloning of full-length viral DNA, no extensive changes were detected in the coding regions of the JCV genome by restriction analysis suggesting an intact JCV DNA population. For further analysis of the distribution of JCV subtypes, the non-coding region was amplified by PCR. Molecular analysis revealed homogeneous JCV TCR populations in almost 50 % of the individuals. Heterogeneity was found in two CNS samples with three and five different JCV subtypes, respectively, Introduction Infection with the human polyomavirus JC virus (JCV) is usually acquired in the first decade of life with seroconversion rates of more than 90 % in the adult (Arthur & Shah, 1989 ; Kitamura et al., 1990 ; Major et al., 1992). Primary contact leads to lifelong asymptomatic persistence in the kidney (Chesters et al., 1983 ; Do$ rries & Elsner, 1991 ; Flaegstad et al., 1991) and the CNS (Chesters et al., 1983 ; Elsner & Do$ rries, 1992). Additionally, JCV DNA is found in tonsillar cells, leukocytes and the bone marrow (Do$ rries et al., 1994 ; Monaco et al., 1996 ; Myers et al., 1989). Activation of persistent virus infection occurs frequently without causing disease (Coleman et al., 1980). Only in cases of severe impairment of immunocompetence (Schmidbauer et al., 1990 ; Walker & Padgett, 1983) may cytolytic JCV infection of oligodendrocytes cause the fatal CNS disease progressive multifocal leukoencephalopathy (PML). Although the number of cases is steadily increasing due to the AIDS epidemic, the pathomechanisms leading to PML are not yet understood. As well as the role of Author for correspondence : Kristina Do$ rries. Fax 49 931 201 3934. e-mail Doerries!mail.uni-wuerzburg.de and in four kidney specimens with two TCR subtypes. Altogether, seven TCR subtypes were identified. One in each group represented single promoter element TCRs without duplication of sequences. The TCR of the major variant JCV-W1 was comparable in sequence and structure to that of the PML prototype JCV Mad-1 DNA. The identification of dominant PML-derived JCV TCR subtypes in most persistently infected individuals suggests that rearrangements of the JCV TCR can be associated with the persistent state of infection. However, it appears unlikely that PML-associated JCV subtypes are generated anew in each individual host in the course of persistence. The findings rather suggest that a limited number of stable JCV subtypes circulate in different geographical regions of the world. severe impairment of immunological surveillance, genomic heterogeneity of JCV DNA was assumed to be associated with development of the disease. Characterization of JCV genomes in PML tissue revealed that coding DNA segments are conserved with changes at the nucleotide level not necessarily leading to changes in the amino acid composition of proteins. Some of the changes in the carboxy-terminal segments of VP1, T antigen and the intergenic region are conserved in clustered geographical groups. Therefore, they are used as a basis for the so-called VT genotyping scheme (Agostini et al., 1997 ; Ault & Stoner, 1992 ; Frisque et al., 1984 ; Loeber & Do$ rries, 1988). The noncoding DNA segment containing the origin of DNA replication and the transcriptional control region (TCR) is conserved with respect to the nucleotide sequence, but heterogeneous in its structure (Agostini et al., 1997 ; Grinnell et al., 1983 ; Martin et al., 1985 ; Matsuda et al., 1987). Promoter elements in PML are classified into TCR type I and TCR type II DNA. A prerequisite for TCR type I JCV genomes is the fusion of the TATA box and following sequences leading to repetition of elements that include the TATA box. TCR type II JCV DNA has one TATA box proximal to the early mRNA initiation site that is not included in the repeated elements. 0001-5074 # 1998 SGM HIJ Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Mon, 19 Jun 2017 01:04:52 C. Elsner and K. Do$ rries Table 1. Patients and organ material analysed for the presence of JCV DNA in persistently infected organs PCR products Case CNS B2 B14 B17 B20 B23 Kidney K1 K3 K4 K6 K7 K8 K9 K10 Age/sex* T antigen Control region 34}m Not known 74}m 84}m 69}m Multiple sclerosis Liver carcinoma Rectum carcinoma Heart failure Heart failure 59}m 11}m 50}m 58}m 77}f 64}m 77}f 32}m HSV encephalitis Leukaemia Creutzfeldt-Jacob disease Aneurysm Liver cirrhosis Leukaemia Cardio-pulmonal failure Renal transplantation, aneurysm Disease * Age is given in years ; m, male ; f, female. Both subtypes have conserved factor binding sites in common that are framed by highly variable sequences preferentially used as sites of recombination (Agostini et al., 1997 ; Ault & Stoner, 1993 ; Loeber & Do$ rries, 1988). TCRs with unique DNA sequences of single promoter elements were detected in kidney and urine (Flaegstad et al., 1991 ; Loeber & Do$ rries, 1988 ; Tominaga et al., 1992 ; Yogo et al., 1990, 1991 b). In addition to genotype and TCR variants, defective polyomavirus genomes were observed in semipermissive cell cultures and tissue. Assuming that such molecules might be generated in the course of persistent JCV infection, a possible role of a defective DNA population with extensive deletions or substitutions within coding sequences has been repeatedly discussed (Grinnell et al., 1982 ; McCance, 1983 ; Yoshiike et al., 1982). Comparison of JCV DNA populations with a high rate of virus replication in PML brain, and those exhibiting low rates in persistently infected kidney tissue demonstrated JCV DNA molecules of different TCR types but with highly related coding sequences in the respective tissue of one patient (Ault & Stoner, 1994 ; Do$ rries, 1984 ; Loeber & Do$ rries, 1988). Promoter elements of TCR types from cytolytically infected brain tissue included complex rearrangements. In contrast, virus from histopathologically healthy kidney tissue exhibited control regions without extensive sequence duplications. From these findings, the assumption was drawn that the parental rather attenuated JCV archetype may circulate in the human population, invading the CNS during persistence. It appeared HJA to be likely that the PML-type of JCV variants might subsequently be generated anew in each host from the archetype TCR during asymptomatic activation processes (Iida et al., 1993 ; Kato et al., 1994 ; Yogo et al., 1990). The detection of minor heterogeneity in JCV populations from PML brain (Yogo et al., 1994) supported this hypothesis. However, the time-scale of rearrangements in the course of infection and the extent of genomic changes in the CNS prior to disease remained uncertain. In an attempt to elucidate the extent of the genetic heterogeneity in persistently infected organs, we determined the genomic complexity of JCV genomes in CNS and kidney tissue without histopathological changes typical for polyomavirus-induced disease. Analysis of JCV rearrangements in the TCR revealed the presence of one dominant JCV TCR subtype among minor TCR variants in both organs. The TCR was found to be identical in sequence to that of PML-derived JCV TCR types, thus demonstrating that PML-type viruses can similarly invade the CNS and the kidney prior to disease. The characterization of a PML-type TCR as the dominant virus type in peripheral tissue of immunocompetent individuals suggests that rearranged TCR types are circulating in the human population and are not necessarily generated in each host. Methods + Patient material. Autopsy material was analysed from CNS and kidney of 13 patients from Western Europe. Brain samples were examined Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Mon, 19 Jun 2017 01:04:52 JC virus variants in persistently infected tissue from patients with multiple sclerosis, heart failure and carcinomas of the rectum and liver. Kidney material was from eight patients with herpes simplex virus (HSV) encephalitis, Creutzfeldt-Jacob disease, liver cirrhosis, leukaemia, aneurysm and cardio-pulmonal failure. Cellular DNA was prepared by standard procedures (Do$ rries et al., 1994 ; Elsner & Do$ rries, 1992). + Cellular gene libraries from human brain and kidney tissue. JCV genomes were cloned from tissue with the restriction enzyme BamHI in the phage lambda replacement vector NM1151 (Elsner & Do$ rries, 1992). Cellular DNA was cleaved to completion and cloning was performed essentially as described by Sambrook et al. (1989). Between 1¬10& and 6¬10& clones}µg cellular DNA were achieved by in vitro packaging ; 1–3¬10' clones per specimen were analysed by plaque hybridization with radioactive genomic JCV DNA probes (Elsner & Do$ rries, 1992). Characterization of cloned JCV genomes was performed by restriction endonuclease cleavage and followed by Southern blot analysis. + Amplification of JCV-specific DNA by PCR. PCR was performed with primer pairs either covering a DNA fragment of the early coding region (TEP1}PEP2) or the non-coding region flanked by conserved coding sequences (JCV71}72). Cellular DNA (1 µg) was amplified in 2±5 mM MgCl in a reaction volume of 50 µl with 1 U of Taq polymerase. # Cycling was as described previously (Do$ rries et al., 1994 ; Elsner & Do$ rries, 1992). As a precaution against PCR contamination, all reactions were performed in parallel with a complete set of controls. Negative controls included testing of the buffer components in duplicate prior to and after preparation of experimental reactions in laboratories used exclusively for PCR. Characterization of amplification products was performed in another laboratory by electrophoresis and radioactive hybridization of Southern blots with a JCV-specific internal oligonucleotide or a specific ssRNA probe (Do$ rries et al., 1994 ; Elsner & Do$ rries, 1992). + Cloning and typing of JCV control region elements. For cloning, JCV-specific products were amplified by PCR with the primer pair JCV71}72 containing an artificial BamHI and EcoRI restriction site, respectively (Elsner & Do$ rries, 1992), and inserted into the bacterial vector pBluescript SK(). Recombinants were selected by colony hybridization with a radioactive ssRNA probe (Elsner & Do$ rries, 1992). JCV prototypes were cloned by the same strategy (Frisque et al., 1984 ; Loeber & Do$ rries, 1988). The characterization of JCV inserts was performed with the restriction enzymes BamHI}EcoRI and StyI followed by electrophoresis on agarose gels. + Analysis of cloned JCV DNA segments by nucleotide sequencing. DNA sequencing of the complete JCV control region between the start codon of the putative agnogene and that of the T antigen gene was performed from one-tube plasmid mini-preparations (DelSal et al., 1988). Isolated DNA was denatured with 0±2 M NaOH (10 min, 65 °C), neutralized by 0±4 vol. 5 M ammonium acetate (pH 5±5) and ethanol-precipitated. For sequencing virus-specific internal oligonucleotides JCV 69, 87, 60, 53 and pBluescript primers 16 and 31 were used (Do$ rries et al., 1994). Sequencing was performed with the enzyme Sequenase and $&S-dATP (USB, Amersham) essentially as recommended by the supplier. Results Complexity of JCV DNA in persistently infected brain and kidney tissue For characterization of the JCV DNA population in persistently infected tissue, autopsy specimens were selected from brain parenchyma and from the renal medulla of patients with CNS diseases other than PML or with systemic disorders (Table 1). In none of the cases, neurological symptoms or histopathological patterns indicative of PML were observed. Cellular DNA was subjected to amplification of JCV DNA. One primer pair was specific for the T antigen coding region and resulted in a fragment of 199 bp in length. A second primer pair spanning the control region amplified a fragment of about 600 bp in length (Elsner & Do$ rries, 1992). The specificity of electrophoretically separated products was confirmed by Southern blotting and JCV-specific radioactive hybridization. PCR analysis revealed the presence of JCV-specific DNA sequences in at least one of five brain specimens and one of four kidney specimens from each case (Table 1). In order to analyse the complexity of JCV DNA molecules in persistently infected organs, the complete genome was examined by cloning of JCV genomes from CNS tissue in two cases (B20, B23) and from kidney tissue in three cases (K1, K3, K4). The single BamHI restriction site in the early coding region of the JCV genome was used for insertion in phage lambda NM1151. Characterization of cloned JCV molecules was performed by restriction cleavage with the enzyme BamHI and combinations of BamHI with EcoRI, PstI, BglII, XbaI, NsiI, HpaII, HincII or HindIII followed by Southern blot analysis. The analyses revealed a JCV insert of 5 kb in length in all clones from the brain and kidney. Comparison of the DNA fragment pattern with that of the JCV prototypes showed the characteristic JCV pattern with all enzymes used (data not shown). This finding demonstrated that cloned viral DNA molecules in persistently infected tissue were intact without major rearrangements of coding sequences. Comparison of JCV TCR subtype populations in brain and kidney tissue by restriction mapping Analysis of full-length JCV DNA with the enzyme combination BamHI}PvuII focussed on the TCR. Palindromes of the restriction enzyme PvuII are located close to and within the non-coding region of the JCV genome resulting mainly in three DNA fragments spanning conserved coding DNA segments and a variable fragment C including the viral control region. In one case, two out of 16 brain-derived clones exhibited heterogeneous mobility of fragment C. One was not related to the prototype segment C, whereas the second was comparable in length to that of JCV GS}B (991 bp) (Fig. 1 a, lanes 4 and 10). In the remaining clones, fragment C was comparable to that of JCV Mad-1 (1093 bp). In contrast, clones from the second brain specimen and all kidney-derived clones from three cases had an identical fragment pattern comparable in length to that of JCV Mad-1 (Fig. 1 b). To determine the TCR heterogeneity further, JCV DNA was amplified by PCR on brain tissue from three and on kidney tissue from five additional cases (Table 2). Products were directly cloned to avoid loss of less common JCV subtypes involved. Twenty to 40 JCV-specific clones per case were Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Mon, 19 Jun 2017 01:04:52 HJB C. Elsner and K. Do$ rries (a) CNS ( b) Kidney Fig. 1. Analysis of JCV genomes from persistently infected organs ; (a) CNS and (b) kidney. Restriction cleavage was performed with BamHI/PvuII followed by separation on a 1 % agarose gel and Southern blot hybridization with a radioactively labelled JCV-specific RNA probe. JCV genomic DNA was cloned from CNS and kidney tissues. Prototype DNA GS/B, Mad-1 and archetype GS/K DNA are shown on the left. Numbers at the top of each lane represent individual clones. BamHI/PvuII fragments A, B, D represent coding DNA segments ; fragment C includes the non-coding region. isolated, thus giving a representative view on the JCV population present. A total of 91 clones was established from CNS tissue, and 227 clones were achieved from kidney tissue. Restriction cleavage of virus-specific inserts with the cloning enzymes BamHI}EcoRI revealed the presence of heterogeneous JCV TCR fragments in 67 % of the CNS samples and in 80 % of the kidney samples. A detailed analysis of the JCV control region was performed with the restriction enzyme StyI. The enzyme cleaved the insert into 5–6 fragments (Fig. 2 a, b). The fragments A and B spanned sequences upstream of the origin of DNA replication and downstream of late coding sequences. Promoter elements were divided into 2–4 fragments as HJC demonstrated by comparison with those of prototype TCR type I (JCV Mad-1), type II (JCV GS}B) and archetype DNA (JCV GS}K). In CNS and kidney tissue, seven and two different TCR StyI patterns were identified, respectively (Figs 2 and 3 a, b, Table 2). Pattern JCV-W1 was comparable to that of prototype JCV Mad-1. JCV-W6 was similar to that of prototype JCV GS}B, and JCV-W7 was comparable to that of archetype JCV GS}K. Four additional patterns, JCV-W2 to JCV-W5, were detected. The JCV-W2 type exhibited DNA fragments related to those of JCV Mad-1, although the entire insert length was about 100 bp longer ; this was due to a duplicated band. Fragment pattern JCV-W4 was also related to the Mad-1 subtype. The missing 100 bp band was reflected in Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Mon, 19 Jun 2017 01:04:52 JC virus variants in persistently infected tissue Table 2. JCV TCR variants in persistently infected CNS and kidney tissue Tissue Case No. of clones analysed* CNS B20 16a B23 6a B2 17b B14 14b B17 38b 5 K1 91 16a K3 7a K4 6a K6 43b K7 38b K8 42b K9 39b K10 36b Totals Kidney Totals 8 227 JCV TCR type I† JCV TCR type II† W1 W2 W4‡ W6 W5 W3 W7‡ 14}16 (87±5 %) 6}6 (100 %) 17}17 (100 %) – 1}16 (6±3 %) – 1}16 (6±3 %) – – – – – – – – – – – – – – – – – – – – 3}38 (7±9 %) 40}91 16}16 (100 %) 7}7 (100 %) 6}6 (100 %) 40}43 (93 %) 37}38 (97 %) 41}42 (97±6 %) 39}39 (100 %) 27}36 (75 %) 213}227 – – 1}91 – 1}91 – 14}14 (100 %) 12}38 (31±6 %) 26}91 – 1}38 (2±6 %) 1}91 – 1}38 (2±6 %) 1}91 – 21}38 (55±3 %) 21}91 – – – – – – – – – – – – – – 3}43 (7 %) 1}38 (3 %) 1}42 (2±4 %) – – – – – – – – – – – – – – – – – 9}36 (25 %) 14}227 – – – – 0}227 0}227 0}227 0}227 – – – – 0}227 * a, isolated by direct cloning ; b, isolated after PCR amplification. † Type determined by restriction analysis and sequencing. ‡ Related sequence with single elements. the reduced length of the entire insert. JCV-W5 had a pattern with a smaller fragment C that showed similarities to that of the archetype DNA. JCV-W3 clones exhibited a pattern that was not comparable to any of the other JCV TCR types. Thus, in addition to a new as yet unknown TCR pattern, control regions similar to those of PML-derived JCV genomes at both sites of infection were found. Nucleotide sequence of variant control regions from persistently infected tissue The detection of JCV TCR segments of the PML-type in persistently infected tissue was rather unexpected. Therefore, the JCV control region elements were further characterized by DNA sequencing (Fig. 4). To improve the comparison of divergent DNA sequences, the JCV control region was divided into the origin of DNA replication (ori) and seven DNA segments (A–E, F1, F2) with reference to Ault & Stoner (1993). On the early side of all TCR types, ori sequences were highly conserved. The length of TCR type I control regions varied between 726 and 522 bp (Fig. 2). The DNA segment A, a TATA box element, was fused to nucleotides of segment C and segment E. In TCR type II DNA, the control region varied in length between 581 and 662 bp. Segment A was fused to segment B and the following segment C. In TCR type I, segment C was conserved, whereas segment C of TCR type II was variable in length with deletions of 8–21 bp at the 5« end. Segment B separated segments A and C leading to repeated elements without TATA box sequences. These conserved sequence elements were framed by variable sequences leading to differentially constructed repeated elements. Downstream sequences included a 16–18 bp segment E in both TCR types. In TCR type II, these were followed by the 3«-truncated forms of the 66 bp archetype segment D. The sequence of segments F1, F2 and Agno downstream of the initiation codon of the putative agnogene was conserved in six out of seven subtypes identified. Rearranged segments E through to Agno were Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Mon, 19 Jun 2017 01:04:52 HJD C. Elsner and K. Do$ rries (a) CNS (a) (b) (b) Kidney Fig. 2. Restriction endonuclease mapping of the JCV subtype control region. PCR amplification of the control region and flanking early and late coding sequences was carried out with degenerative primers containing restriction sites for BamHI and EcoRI, respectively ; the amplification products were cloned in the bacterial vector pBluescript SK(). (a) Positions of StyI restriction sites in the control region are indicated in the prototype JCV Mad-1, GS/B and archetype GS/K DNA segments. A–E represent StyI restriction fragments ; ORI, origin of DNA replication ; n, nucleotide position ; a, numbering according to Loeber & Do$ rries (1988). Early and late genes show the orientation of the TCR. (b) StyI fragment composition of JCV subtype control regions ; *, values normalized after nucleotide sequence analysis. identified only once (Fig. 5, W6). This TCR subtype was previously observed in PML tissue (Loeber & Do$ rries, 1988). Single point mutations were repeatedly detected. However, comparison of different clones of the same subtype revealed random distribution of the changes. This suggested that single point mutations were probably introduced at least in part by the PCR amplification step. The genotype defining single base change on the late side of ori at the archetype TCR position 133 was represented by the nucleotide adenosine in all cases. This is a common finding in American and European isolates (Yogo et al., 1990). JCV TCR subtype lineages in persistently infected individuals The JCV control regions identified so far were divided into three types. Type I and type II both contain repeated conserved elements that might even include the TATA box or extend to late coding sequences. In contrast, the so-called archetype TCR is formed by unique sequences and a complete segment D that is located downstream of the conserved region C. As shown in Fig. 5, the major JCV subtype W1 carried a type I TCR. The HJE Fig. 3. Characterization of JCV TCR subtypes from CNS (a) and kidney (b) tissue after PCR amplification and subcloning. JCV control regions were amplified by PCR with a degenerative primer pair containing BamHI and EcoRI cloning sites and cloned in the bacterial vector pBluescript SK(). BamHI/EcoRI/StyI restriction fragments were separated and compared with the electrophoretic mobility of the control region fragments of JCV Mad-1, GS/B and GS/K ; the JCV Mad-1 fragment length is indicated in bp. B2, B14 and B17, brain tissue of individual patients ; K7, K6, K8 and K9, kidney tissue ; numbers represent individual clones of one case. DNA nucleotide sequence was almost identical to that of the JCV prototype Mad-1. Additionally, two rearranged TCR type I regions were identified differing only in the number of the first conserved element. In JCV-W2, this element was present in triplicate. JCV-W4 represented a single TCR element identical in sequence to that of the first conserved type I element. The downstream segments E, F1 and F2 were all similarly arranged. Four TCR type II variants were identified by the presence of segment B following the TATA box. JCVW6 was identical in sequence and structure to that of JCV GS}B (Loeber & Do$ rries, 1988). In the early segment, JCV-W5 was highly related to JCV-W6, whereas the late part, including the segments F1 through to Agno, was comparable to that of the archetype. JCV-W5 and JCV-W6 both carried truncated forms of the archetype related segment D suggesting a close Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Mon, 19 Jun 2017 01:04:52 JC virus variants in persistently infected tissue Fig. 4. Alignment of the nucleotide sequence of JCV subtypes isolated from persistently infected individuals. The nucleotide sequences of JCV TCR subtypes W1, W2, W4, W3, W5, W7 from German individuals are aligned from the core origin of DNA replication to the start codon of the agnogene. Nucleotide numbering is according to Loeber & Do$ rries (1988). The DNA strand has the polarity of the late mRNA. The start codon for the agnogene is indicated by an arrow. The sequence is divided in seven boxed segments (A, B, C, D, E, F1, F2) ; shading is identical to that in Fig. 5. ‘ A ’ represents the TATA box and T antigen binding sites are indicated. Deletions are marked by a dash. Fig. 5. Structure of the promoter elements of JCV TCR type I (JCV W1, W2, W4) and type II (JCV W6, W5, W3, W7) isolated from persistently infected individuals. The DNA sequence of the A–F2 segments is identical to that in Fig. 4. The initiation codon for the putative agnogene is indicated by an arrow. Agno represents the putative agnogene coding DNA segment repeated in the JCV-W6 and the GS/B prototype TCR. Segment length is given in bp ; AT represents the TATA box ; ori, origin of DNA replication. *, Nucleotides are numbered according to Loeber & Do$ rries (1988). v, Deletion of 23 bp ; –, conserved sequences. Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Mon, 19 Jun 2017 01:04:52 HJF C. Elsner and K. Do$ rries relationship between these TCR variants. In JCV-W3, segment D is eliminated and a deletion at the 3«-end of segment C and at the 5«-end of the second B element could be observed. JCVW7 represented the archetype, which is thought to be the ancestral type of all differentially arranged TCRs. The TCR classification demonstrates that both lineages of rearranged TCR types and the archetype TCR were present in persistently infected individuals in Western Europe. Distribution of JCV TCR subtypes in infected organs Comparison of the TCR types in brain tissue revealed the presence of major and minor subtypes (Table 2). The rate of minor JCV types varied between 2±6 and 30±8 %. Major types were found at a rate between 53 and 100 %. The PMLassociated subtypes JCV-W1 and JCV-W6 were the dominant types in brain tissue with rates between 87±5 and 100 %. Interestingly, the archetype JCV-W7 was present at a rate of 53 % in one brain specimen. This was higher than the closely related duplicated TCR JCV-W6 (30±8 %) in the same specimen. Although this finding suggested extensive rearrangements in this patient, variant molecules were observed only in two out of five cases. Thus, rearranged JCV TCRs are probably not regularly present in persistently infected brain tissue. In the kidney, JCV TCR populations were similarly homogeneous in about 50 % of all cases (seven out of 13 cases). The dominant TCR was the PML-derived prototype Mad-1. Only one minor variant was detected at rates between 2±4 and 25 % in the four heterogeneous DNA populations. This subtype contained a single element TCR closely related to the Mad-1 subtype. The same subtype was additionally detected in brain tissue ; therefore the TCR type is not tissue-specific. TCR JCV-W1 was the major type identified in 12 of the 13 patients irrespective of the tissue analysed. This revealed that PML-type variants persisted in the CNS and the kidney of individuals without JCV-associated disease. From the distribution of TCR subtypes in kidney and brain tissue, rearrangements in the CNS of individual patients cannot be excluded. However, it appears to be unlikely that changes in the TCR are restricted to the CNS. The identification of the JCV Mad-1 (JCV-W1) prototype as the only variant in the majority of cases rather suggested the presence of conserved stable JCV TCR subtypes circulating in the human population. Discussion The extensive divergence of the JCV control region in PML patients gave rise to the hypothesis that the generation of heterogeneous TCRs might be involved in virulence during the pathogenic process. It was believed that JCV genomes may change from an attenuated persisting subtype with single promoter elements to a virulent virus type with rearranged duplicated promoter elements growing efficiently in the CNS. This hypothesis implicated that the state of disease is preceded by progressive selection of JCV subtypes leading to a heterogeneous JCV population in persistently infected CNS HJG tissue. Besides a hypervariability of the JCV TCR in the noncoding region of the genome, DNA rearrangements of coding sequences were described in brain tissue of a PML patient and as a result of JCV growth in tissue culture (Elsner & Do$ rries, 1992 ; Martin et al., 1982 ; Matsuda et al., 1987 ; Rentier Delrue et al., 1981). Although detected in vivo only in rare cases and in the late stages of disease, it was conceivable that JCV genomes with major changes in the coding region of the genome might become prevalent in the attenuated persistent state of infection. To further evaluate the role of JCV variants in the pathogenesis of PML, we examined the genomic structure of JCV DNA in tissue of patients without histopathological changes indicative for JCV-associated disease. Persistent infection was confirmed by PCR amplification of JCV DNA. However, amplification of JCV DNA fragments alone does not give evidence of the complexity of viral genomes, and full-length genomic JCV PCR is not sensitive enough for JCV detection in persistence (own observations). Therefore, JCV DNA from tissue was cloned in a phage lambda gene library and then analysed. Characterization of JCVspecific inserts revealed exclusively the presence of full-length JCV genomes. Extensive restriction enzyme mapping gave no indication of major changes within coding DNA segments. This confirmed the high conservation rate of the genomic structure (Grinnell et al., 1982 ; Martin et al., 1982 ; Rentier Delrue et al., 1981) and suggested that rearranged coding DNA segments are not regularly generated in the persistent state of infection. Therefore, it appears rather unlikely that defective JCV genomes play a major role in polyomavirus persistence. In contrast to the conserved coding region, restriction analysis of the non-coding region revealed heterogeneity of DNA segments carrying the TCR. Most of the genomes revealed fragment lengths similar to that of the rearranged PML prototype Mad-1 DNA. In view of reports on JCV archetype TCRs with single elements in healthy individuals (Markowitz et al., 1991 ; Yogo et al., 1990, 1991 b), this finding suggested a more extensive involvement of PML-associated promoter elements in persistent infection than thus far assumed (Sundsfjord et al., 1994). To evaluate this finding further, the TCR populations were analysed by PCR of the non-coding region and following molecular characterization. Loss of minor JCV variants was avoided by cloning of the products prior to further analyses. Using this strategy, the clone number per individual sample probably reflected the concentration of JCV DNA, which appeared to be higher in the renal medulla than in the CNS (Kitamura et al., 1994 a, 1994 b ; Stoner et al., 1996 ; Tominaga et al., 1992). Detailed sequence analysis of the JCV control region revealed the presence of seven different TCRs in persistently infected tissue samples from 13 patients. The most prominent types at both sites of infection were identical to those of the prototypes JCV Mad-1 and GS}B isolated from an American and a German PML brain. The archetype structure described Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Mon, 19 Jun 2017 01:04:52 JC virus variants in persistently infected tissue earlier as the dominant TCR type in renal infection was detected as a minor population in brain tissue. This confirmed the results from the cloning experiments above, but is in contrast with the assumed predominance of JCV archetype TCRs in persistently infected individuals (Markowitz et al., 1991 ; Sundsfjord et al., 1994 ; Yogo et al., 1991 b, 1990). In the attempt to find an explanation for this discrepancy, the possible generation of artificial amplification products by PCR-dependent recombination was taken into consideration. Although recombination rates of almost 5 % were suggested, depending on the PCR target DNA used (Meyerhans et al., 1990), in our experiments the major TCR subtypes were not only detected by PCR amplification but also in the cloned fulllength genomes. Therefore, PCR recombination and PCR contamination appeared to be rather unlikely. Besides the extensive control of the PCR experiments, the strongest argument against contamination is the detection of the same rate of JCV Mad-1-like TCRs in the classical cloning approach. Additionally, analysis of the V-T intergenic genotype linked to the described TCRs has been performed. In contrast to the Mad-1 genome isolated from a PML patient in the USA, the JCV-W1 genomes from persistently infected tissue are linked to coding regions of V-T genotype 2. This confirms that the JCV-W1 molecules are unique isolates. Interestingly, V-T genotype 1 sequences were described in urine of individuals from Northern Europe (Yogo et al., 1991 a). In contrast, in the mixed American population almost all genotypes were found, with a predominance of genotypes 1 and 2 in white Caucasians (Agostini et al., 1997 ; Ault & Stoner, 1992). Therefore, the detection of genotype 2 in our group might be representative of the geographical distribution of JCV genotypes in Western Europe. The often propagated hypothesis that each individual carries a unique JCV TCR subtype generated from the circulating archetype (Yogo et al., 1990) is not only contradicted by the data presented here, but also by reports from the USA demonstrating a predominance of the Mad-1 type in persistently infected patients (Myers et al., 1989 ; White et al., 1992). This points to a higher prevalence of the Mad-1 TCR in selected groups in the USA, whereas in others the archetype was found to be the dominant TCR. In Japan, almost exclusive archetype-related TCRs with minor changes were described in persistently infected tissue (Markowitz et al., 1991 ; Yogo et al., 1991 b). These findings support the assumption that a limited number of TCR types dominate the virus population in different geographical regions of the world. The overgrowth of distinct TCRs might be explained by differential stability of TCR structures. Conserved stable TCR segments can be observed in type I and type II TCRs. In case of the Mad-1 isolate, even the complete control region exhibits high stability after a prolonged tissue culture history. This is demonstrated by conservation of the DNA sequence and structure of the direct JCV Mad-1 isolate (Br-1) and the cultured virus (Mad-1) from the same PML brain (Grinnell et al., 1983 ; Martin et al., 1982 ; Rentier Delrue et al., 1981). In contrast to this stable organization, highly heterogeneous TCR populations and defective genomes were found after tissue culture passages of other TCRs (Howley et al., 1976 ; Martin et al., 1979 ; Osborn et al., 1974). Therefore, it is conceivable that TCR structures may have evolved which are more or less susceptible to rearrangement events. Whether this is linked to the biological function of the respective promoter element is not yet determined. Predominantly homogeneous TCRs or subtype populations with limited heterogeneity were detected at both sites of persistent infection. The major TCR subtype identified in healthy CNS was similar to that in lytically infected PML tissue. Therefore, a neurotropic selection of highly active JCV TCRs as a pathogenic event is rather unlikely. This does not exclude occasional rearrangements of TCRs in the course of persistence. The conditions leading to the development of variant TCR types are not yet known. In case of the archetype TCR, sequence rearrangements with only small variations from the archetype were observed in immunosuppressed renal transplant patients (Kitamura et al., 1994 b ; Yogo et al., 1990). This suggested that immunosuppression may support TCR rearrangement. However, in the cases described here a correlation of immune impairment by malignancies and the presence of variant JCV DNA populations could not be found. In a search for other possible factors responsible for TCR rearrangements it became obvious that most studies concentrated on brain and kidney tissue, organs not exhibiting cellular replication activity. In this context, the characterization of lymphoid cells as a site of persistent JCV infection was noticeable (Do$ rries et al., 1994 ; Major et al., 1992). This cell compartment includes actively dividing cell populations and it is conceivable that JCV TCR variants might be generated in the replicating cells. The occurrence of a major PML-derived JCV TCR population in persistently infected tissue of individuals without PML suggests that highly active JCV TCRs circulate in Western Europe. They might already infect the individual host on primary contact and distribute to the target organs of infection. Therefore, it is not likely that active virus growth in the oligodendrocytes depends on the selection of glia cellspecific promoter elements. Activation of infection and induction of disease may rather be controlled by conditions closely related to the immunological competence of the viral host. This study was supported by the Deutsche Forschungsgemeinschaft, Sonderforschungsbereich SFB 165}B3 and by the Wilhelm Sander Stiftung, Grant 91.061. References Agostini, H. T., Ryschkewitsch, C. F., Singer, E. J. & Stoner, G. L. (1997). JC virus regulatory region rearrangements and genotypes in progressive multifocal leukoencephalopathy : two independent aspects of virus variation. Journal of General Virology 78, 659–664. Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Mon, 19 Jun 2017 01:04:52 HJH C. Elsner and K. Do$ rries Arthur, R. R. & Shah, K. V. (1989). Occurrence and significance of papovaviruses BK and JC in the urine. Progress in Medical Virology 36, 42–61. Ault, G. S. & Stoner, G. L. (1992). Two major types of JC virus defined in progressive multifocal leukoencephalopathy brain by early and late coding region DNA sequences. Journal of General Virology 73, 2669–2678. Ault, G. S. & Stoner, G. L. (1993). Human polyomavirus JC promoter} enhancer rearrangement patterns from progressive multifocal leukoencephalopathy brain are unique derivatives of a single archetypal structure. Journal of General Virology 74, 1499–1507. Ault, G. S. & Stoner, G. L. (1994). Brain and kidney of progressive multifocal leukoencephalopathy patients contain identical rearrangements of the JC virus promoter}enhancer. Journal of Medical Virology 44, 298–304. Chesters, P. M., Heritage, J. & McCance, D. J. (1983). Persistence of DNA sequences of BK virus and JC virus in normal human tissues and in diseased tissues. Journal of Infectious Diseases 147, 676–684. Coleman, D. V., Wolfendale, M. R., Daniel, R. A., Dhanjal, N. K., Gardner, S. D., Gibson, P. E. & Field, A. M. (1980). A prospective study of human polyomavirus infection in pregnancy. Journal of Infectious Diseases 142, 1–8. DelSal, G., Manfioletti, G. & Schneider, C. (1988). A one-tube plasmid DNA mini-preparation suitable for sequencing. Nucleic Acids Research 16, 9878. Do$ rries, K. (1984). Progressive multifocal leukoencephalopathy : analysis of JC virus DNA from brain and kidney tissue. Virus Research 1, 25–38. Do$ rries, K. & Elsner, C. (1991). Persistent polyomavirus infection in kidney tissue of patients with diseases other than progressive multifocal leukoencephalopathy. Virology Life Science Advances 10, 51–61. Do$ rries, K., Vogel, E., Gu$ nther, S. & Czub, S. (1994). Infection of human polyomavirus JC and BK in peripheral blood leukocytes from immunocompetent individuals. Virology 198, 59–70. Elsner, C. & Do$ rries, K. (1992). Evidence of human polyomavirus BK and JC infection in normal brain tissue. Virology 198, 59–70. Flaegstad, T., Sundsfjord, A., Arthur, R. R., Pedersen, M., Traavik, T. & Subramani, S. (1991). Amplification and sequencing of the control regions of BK and JC virus from human urine by polymerase chain reaction. Virology 180, 553–560. Frisque, R. J., Bream, G. L. & Cannella, M. T. (1984). Human polyomavirus JC virus genome. Journal of Virology 51, 458–469. Grinnell, B. W., Martin, J. D., Padgett, B. L. & Walker, D. L. (1982). Is progressive multifocal leukoencephalopathy a chronic disease because of defective interfering particles or temperature-sensitive mutants of JC virus? Journal of Virology 43, 1143–1150. Grinnell, B. W., Padgett, B. L. & Walker, D. L. (1983). Comparison of infectious JC virus DNAs cloned from human brain. Journal of Virology 45, 299–308. Howley, P. M., Khoury, G., Takemoto, K. K. & Martin, M. M. (1976). Polynucleotide sequences common to the genomes of simian virus 40 and the human papovaviruses JC and BK. Virology 73, 303–307. Iida, T., Kitamura, T., Guo, J., Taguchi, F., Aso, Y., Nagashima, K. & Yogo, Y. (1993). Origin of JC polyomavirus variants associated with progressive multifocal leukoencephalopathy. Proceedings of the National Academy of Sciences, USA 90, 5062–5065. Kato, K., Guo, J., Taguchi, F., Daimaru, O., Tajima, M., Haibara, H., Matsuda, J., Sumiya, M. & Yogo, Y. (1994). Phylogenetic comparison between archetypal and disease-associated JC virus isolates in Japan. Japanese Journal of Medical Science and Biology 47, 167–178. Kitamura, T., Aso, Y., Kuniyoshi, N., Hara, K. & Yogo, Y. (1990). High HJI incidence of urinary JC virus excretion in nonimmunosuppressed older patients. Journal of Infectious Diseases 161, 1128–1133. Kitamura, T., Kunitake, T., Guo, J., Tominaga, T., Kawabe, K. & Yogo, Y. (1994 a). Transmission of the human polyomavirus JC virus occurs both within the family and outside the family. Journal of Clinical Microbiology 32, 2359–2363. Kitamura, T., Satoh, K., Tominaga, T., Taguchi, F., Tajima, A., Suzuki, K., Aso, Y. & Yogo, Y. (1994 b). Alteration in the JC polyomavirus genome is enhanced in immunosuppressed renal transplant patients. Virology 198, 341–345. Loeber, G. & Do$ rries, K. (1988). DNA rearrangements in organ specific variants of polyomavirus JC strain GS. Journal of Virology 62, 1730–1735. Major, E. O., Amemiya, K., Tornatore, C. S., Houff, S. A. & Berger, J. R. (1992). Pathogenesis and molecular biology of progressive multifocal leukoencephalopathy, the JC virus-induced demyelinating disease of the human brain. Clinical Microbiology Reviews 5, 49–73. McCance, D. J. (1983). Persistence of animal and human papovaviruses in renal and nervous tissues. Progress in Clinical and Biological Research 105, 343–357. Markowitz, R.-B., Eaton, B. A., Kubik, M. F., Latorra, D., McGregor, J. A. & Dynan, W. S. (1991). BK virus and JC virus shed during pregnancy have predominantly archetypal regulatory regions. Journal of Virology 65, 4515–4519. Martin, J. D., Frisque, R. J., Padgett, B. L. & Walker, D. L. (1979). Restriction endonuclease cleavage map of the DNA of JC virus. Journal of Virology 29, 846–855. Martin, J. D., Brackmann, K. H., Grinnell, B. W., Frisque, R. J., Walker, D. L. & Green, M. (1982). Recombinant JC viral DNA : Verification and physical map of prototype. Biochemical and Biophysical Research Communications 109, 70–77. Martin, J. D., King, D. M., Slauch, J. M. & Frisque, R. J. (1985). Differences in regulatory sequences of naturally occurring JC virus variants. Journal of Virology 53, 306–311. Matsuda, M., Jona, M., Yasui, K. & Nagashima, K. (1987). Genetic characterization of JC virus Tokyo-1 strain, a variant oncogenic in rodents. Virus Research 7, 159–168. Meyerhans, A., Vartanian, J.-P. & Wain-Hobson, S. (1990). DNA recombination during PCR. Nucleic Acids Research 18, 1687–1691. Monaco, M. C. G., Atwood, W. J., Gravell, M., Tornatore, C. S. & Major, E. O. (1996). JC virus infection of hematopoietic progenitor cells, primary B lymphocytes, and tonsillar stromal cells : implications for viral latency. Journal of Virology 70, 7004–7012. Myers, C., Frisque, R. J. & Arthur, R. R. (1989). Direct isolation and characterization of JC virus from urine samples of renal and bone marrow transplant patients. Journal of Virology 63, 4445–4449. Osborn, J. E., Robertson, S. M., Padgett, B. L., Zu Rhein, G. M., Walker, D. L. & Weisblum, B. (1974). Comparison of JC and BK human papovaviruses with simian virus 40 : restriction endonuclease digestion and gel electrophoresis of resultant fragments. Journal of Virology 13, 614–622. Rentier Delrue, F., Lubiniecki, A. & Howley, P. M. (1981). Analysis of JC virus DNA purified directly from human progressive multifocal leukoencephalopathy brains. Journal of Virology 38, 761–769. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning : A Laboratory Manual, 2nd edn. Cold Spring Harbor, NY : Cold Spring Harbor Laboratory. Schmidbauer, M., Budka, H. & Shah, K. V. (1990). Progressive multifocal leukoencephalopathy (PML) in AIDS and in the pre-AIDS era. Acta Neuropathologica 80, 375–380. Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Mon, 19 Jun 2017 01:04:52 JC virus variants in persistently infected tissue Stoner, G. L., Agostini, H. T., Ryschkewitsch, C. F., Baumhefner, R. W. & Tourtellotte, W. W. (1996). Characterization of JC virus DNA amplified from urine of chronic progressive multiple sclerosis patients. Multiple Sclerosis 1, 193–199. Sundsfjord, A., Flaegstad, T., Flo, R., Spein, A. R., Pedersen, M., Permin, H., Julsrud, J. & Traavik, T. (1994). BK and JC viruses in human immunodeficiency virus type 1-infected persons : prevalence, excretion, viremia, and viral regulatory regions. Journal of Infectious Diseases 169, 485–490. Tominaga, T., Yogo, Y., Kitamura, T. & Aso, Y. (1992). Persistence of archetypal JC virus DNA in normal renal tissue derived from tumorbearing patients. Virology 186, 736–741. Walker, D. L. & Padgett, B. L. (1983). The epidemiology of human polyomaviruses. Progress in Clinical and Biological Research 105, 99–106. White, F. A., Ishaq, M., Stoner, G. L. & Frisque, R. J. (1992). JC virus DNA is present in many human brain samples from patients without progressive multifocal leukoencephalopathy. Journal of Virology 66, 5726–5734. Yogo, Y., Kitamura, T., Sugimoto, C., Ueki, T., Aso, Y., Hara, K. & Taguchi, F. (1990). Isolation of a possible archetypal JC virus DNA sequence from nonimmunocompromised individuals. Journal of Virology 64, 3139–3143. Yogo, Y., Iida, T., Taguchi, F., Kitamura, T. & Aso, Y. (1991 a). Typing of human polyomavirus JC virus on the basis of restriction fragment length polymorphisms. Journal of Clinical Microbiology 29, 2130–2138. Yogo, Y., Kitamura, T., Sugimoto, C., Hara, K., Iida, T., Taguchi, F., Tajima, A., Kawabe, K. & Aso, Y. (1991 b). Sequence rearrangement in JC virus DNAs molecularly cloned from immunosuppressed renal transplant patients. Journal of Virology 65, 2422–2428. Yogo, Y., Guo, J., Iida, T., Satoh, K., Taguchi, F., Takahashi, H., Hall, W. & Nagashima, K. (1994). Occurrence of multiple JC virus variants with distinctive regulatory sequences in the brain of a single patient with progressive multifocal leukoencephalopathy. Virus Genes 8, 99–105. Yoshiike, K., Miyamura, T., Chan, H. W. & Takemoto, K. K. (1982). Two defective DNAs of human polyomavirus JC adapted to growth in human embryonic kidney cells. Journal of Virology 42, 395–401. Received 9 July 1997 ; Accepted 10 November 1997 Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Mon, 19 Jun 2017 01:04:52 HJJ