Progressive multifocal leukoencephalopathy (PML) is a rare, fatal demyelinating disease caused by the human polyomavirus JC (JCV), which induces lytic infection of oligodendrocytes. JCV is an ubiquitous virus that infects at least 70% of the human population and could persist lifelong in humans . A small number of immunocompromised patients, including 3-4% of patients with AIDS, develop PML [1-4]. This neurological pathology is rapidly fatal and consequently contributes significantly to the mortality of HIV-infected patients .
Since the neurological symptoms of PML result from the viral destruction of myelin-producing oligodendrocytes, JCV has long been classed as a neurotrophic virus . Many studies, however, have demonstrated that JCV can infect various other cell types including immune system cells . Using in situ hybridization and immunocytochemistry, JCV DNA and viral capsid protein were detected in mononuclear cells located within the central nervous system (CNS) and the bone marrow from one patient with PML . These mononuclear cells were shown to be B lymphocytes (CD45+R) by dual labelling techniques, demonstrating that JCV was a lymphotrophic virus . Since this original description, JCV DNA has been detected by polymerase chain reaction (PCR) and in situ hybridization in mononuclear cells present in bone marrow, in CNS perivascular spaces and in peripheral blood from patients with PML, suggesting a haematogenous viral spread to the CNS .
Recently, Dubois et al. [9,10] have shown that JCV DNA was detectable by PCR in the peripheral blood lymphocytes from HIV-1-infected patients even in the early stage of infection (28.9% of 135), in HIV-negative immunocompromised patients (16.4% of 61) and also, less frequently (8% of 50), in peripheral blood lymphocytes from healthy blood donors. These recent data suggest that peripheral blood mononuclear cells (PBMC) could be a site of viral latency following primary infection in HIV-seropositive patients and in healthy subjects. However, the role of these peripheral blood cells in CNS invasion before the onset of PML remains controversial [6,7,9,11].
JCV latency or reactivation status in PBMC was examined to determine whether JCV productively infects PBMC in HIV-1-positive immunocompromised and immunocompetent patients before the onset of PML. When JCV DNA detection was positive in PBMC, the presence of JCV mRNA early transcripts, T and t genes, was investigated ; active JCV multiplication was ascertained by the detection of JCV late transcripts for a capsid-coding gene (VP1).
Material and methods
The study of HIV-1-positive patients included 69 immunocompetent patients, 82 immunocompromised patients who did not have PML and 10 patients with AIDS and PML.
Blood specimens were collected from 151 HIV-1-infected patients without signs of PML after clinical consultation at the Tourcoing and Valenciennes General Hospitals (France) from February 1996 to January 1997. The patients formed two groups: 69 immunocompetent patients (mean age 38.41 years) and 82 immunocompromised patients (mean age 36.80 years). Within the immunocompetent group, 57 had CD4 lymphocyte levels of 200-499×106cells/l and 12 had >500×106cells/l. The immunocompromised patients had <200×106cells/l. Blood and cerebrospinal fluid (CSF) specimens were obtained from eight immunocompromised HIV-1-infected patients from the Pellegrin and Haût-Levêque hospitals (Bordeaux) and four immunocompromised HIV-1-infected patients from Tourcoing hospital who presented with clinical and radiological signs of PML .
Total leukocytes were separated by sedimentation and centrifugation of 5-7ml whole blood collected from each patient in EDTA-containing tubes. PBMC were isolated from EDTA-treated whole blood using a classical Ficoll extraction procedure . CSF was stored in portions at -80ºC until DNA extraction. For PBMC or CSF samples, DNA was extracted using proteinase K digestion (40μg per sample; Boehringer Mannheim, Mannheim, Germany) and a classical phenol- chloroform procedure .
Polymerase chain reaction
Each PCR assay was performed in an MJ research thermocycler (Watertown, MA, USA) in a total volume of 50μl, using 1μg DNA extracted from total PBMC, or 0.5μg DNA extracted from CSF, in the presence of 200μM deoxynucleotide triphosphates, 1.5mM MgCl2, 2.5IU Taq polymerase (Gold, Perkin Elmer, Branchburg, NJ, USA) and 50pmol of each primer [JB3 (position 4179-4198): 5′-GTATACACA GCAAAAGAAGC-3′; JB4 (position 4790-4809): 5′- GCTCATCAGCCTGATTTTGG-3′], which recognize nucleotidic sequences localized in JCV T-antigen-encoding early gene. In a second PCR run, 1μl of the amplified products was added to 49μl of the previously described PCR mixture including the primers PEP1 (position 4255-4274; 5′-AGTCTTTAGGGTCTTCT ACC-3′) and JB3. Amplification cycling was performed as described by Vago et al. . The determination of the semi-nested PCR products was performed by agarose gel electrophoresis (2%) with ethidium bromide staining. DNA-free negative controls containing only the PCR reaction mix and positive controls obtained with brain tissues from a deceased HIV-1-infected patient with PML were included. A pBr322 plasmid containing the complete genome of JCV mad-1 strain was used as a positive control in order to test the sensitivity of the amplification and detection procedures .
RNA extraction and analysis
RNA was extracted from the DNA-positive PBMC by using guanidium thiocyanate, phenol-chloroform procedure as described by Chomczynski and Sacchi . RNA was precipitated from the aqueous phase in isopropanol, washed in 70% ethanol and dissolved in 50μl distilled water until used in reverse transcriptase (RT) PCR assays.
The primer JB3 was used to synthesize a part of the large tumor (T) and small tumor (t) antigen-encoding cDNA by using a Superscript II transcriptase reverse according to the manufacturer‚s protocole (Gibco BRL, Cergy Pontoise, France) before a semi-nested PCR within the JCV T-encoding early gene as described above.
The universal oligonucleotide dT15 primer was used to synthesize single-stranded cDNA by reverse transcription, before a nested PCR within VP1 using two external primers, VPP5 (position 1280-1308; 5′-ATGATGCAGACAGCATTGAAGAAGTTACC-3′) and VPM9 (position 2421-2399: 5′-TCCATGCCAT ACATAGGCTGCCC-3′), and two internal primers, VPP13 (position 1725-1748: 5′-TTCCACTACCC AATCTAAATGAGG-3′) and VPM5 (position 2263-2240: 5′-GTTTGTAAACATGCCACAGAC ACT-3′), as previously described by Dubois et al. .
In order to check the quality and the efficiency of RNA extraction, glucose 6-phosphate dehydrogenase mRNA was amplified for each sample analysed [15,16]. JCV RNA samples extracted from brain tissues of subjects with PML were used as positive controls for each serial of RT-PCR assays.
Southern blot analysis
Intron-differential RNA PCR in T and t genes was performed by Southern blotting. The filters were pre-hybridized at 60ºC for 2h into the hybridization solution (×5 saline sodium citrate buffer (SSC), 0.1% N-lauroylsarcosine, 0.02% sodium dodecyl sulphate, 5% blocking solution; provided by Boehringer Mannheim) . In order to probe the large T and small t cDNAs amplified by RT-PCR, the biotin-labelled oligonucleotide JEP1 (5′-bCTTTTTA GGTGGGGTAGAGTGTTGGGATCCTGTGT TTTCA-3′) was then used at concentration of 15pmol/ml. Hybridization was performed overnight at 55ºC. The biotin-labelled hybrids were detected using streptavidin-conjugated to alkaline phosphatase (Boehringer Mannheim). The filters were then covered with chemiluminescent substrate (CSPD, provided by Boehringer Mannheim) and exposed on X-ray films (Amersham, Les Ullis, France) for 2h.
The clinical and biological results were analysed by the chi-square test, by the chi-square Fischer exact test or by the chi-square Mantel-Haenszel test (Epi Info 6/CDC-OMS statistical analysis software).
Detection of JCV DNA in CSF from patients with clinical and/or radiological signs of PML
The sensitivity of JCV PCR within the T early- encoding gene was assessed by limit detection of the signal in serial ten dilutions of a full-length JCV mad-1 strain cloned into a pBR322 vector. Our nested-PCR procedure was able to detect 60 copies of JCV mad-1 strain genome after agarose gel electrophoresis . Using this JCV PCR procedure, we examined CSF samples from immunocompromised HIV-1 patients who presented clinical and radiological criteria of PML at time of inclusion. Of the 12 CSF samples tested, 10 were positive by JCV PCR, demonstrating an ongoing viral neurological infection (data not shown). The two HIV-1-positive immunocompromised patients who presented a neurological pathology associated with a negative JCV PCR in CSF were exluded from this study.
Detection of JCV DNA in peripheral blood mononuclear cells
To assess whether JCV is harboured in peripheral blood cells from HIV-1 patients before the onset of PML, JCV DNA was detected by PCR in PBMC from HIV-1-infected patients with or without PML (Table 1). Within the group of HIV-1-positive immunocompromised patients with PML, JCV PCR was positive in six of 10, demonstrating that an active JCV neurological infection could be developed without viral infection of peripheral blood leukocytes. In the group of 151 HIV-1-infected patients without PML, 31 (20.53%) were positive by JCV PCR in PBMC (data not shown).
Correlation with clinical and biological status of the patients
The distibution of the PBMC positive for JCV DNA according to the clinical status [Centers for Disease Control and Prevention (CDC) stage of infection] and biological status (degree of immunosuppression determined by CD4 lymphocyte counts) of patients is presented in Table 2. JCV DNA was detected by PCR in the blood of HIV-1-infected patients who did not have PML independently of the clinical severity of the HIV-1-induced disease (25.9, 14.5 and 25.4% in CDC stages A, B and C, respectively; P=0.24, chi-square test). We found no significant relationship between the existence of a JCV infection of PBMC and the degree of lymphopenia (16.6, 17.6 and 23.2% for HIV-1-infected-patients who did not have PML with CD4 cell counts >500×106, 200-499×106 and <200×106cells/l, respectively; P>0.34, chi-square test). A statistical analysis stratified according the CDC clinical stage confirmed the absence of association between a positive JCV DNA detection in PBMC and the CD4 lymphocyte counts (17.4 and 23.2% for HIV-1 patients who did not have PML with CD4 lymphocyte counts above and below 200×106cells/l, respectively; P=0.44, chi-square Mantel-Haenszel test) (Table 2). Moreover, the presence or absence of JCV DNA in PBMC did not differ significantly with the age of HIV-1 patients in those who did not have PML (mean ages 37.55 and 37.90 years, respectively; P=0.830, chi-square test) (Table 1).
Detection of JCV early and late mRNA in PBMC
In order to assess JCV latency or reactivation status in peripheral blood leukocytes from HIV-1 patients before the onset of PML, the expression of JCV early and late mRNA was investigated in PBMC samples positive for JCV DNA detection (Table 1).
An intron-differential RT-PCR was developed for the detection of mRNA in PBMC for the large T and small t antigens . A specific reverse transcription of the alternatively spliced mRNA into cDNA was carried out with the JB3 primer, before semi-nested PCR with the JB3, JB4 and PEP1 primers (Fig. 1). The amplified large T and small t cDNA were hybridized for identification with an oligonucleotidic probe (JEP1) in a classical Southern blotting procedure (Fig. 1). The amplification of the small t cDNA yielded a fragment of 487 base pairs (bp) that was clearly differentiated from the 211bp large T cDNA fragment and from the 555bp genomic DNA fragment. Figure 2 shows the bands amplified from large T and small t cDNA from brain tissues after agarose gel electrophoresis and Southern blot analysis for brain tissues with PML (lane 4) and contaminating viral genomic DNA amplified products (555bp) for some PBMC samples (lanes 8, 9, 14 and 15).
The presence of JCV late mRNA was indicated by detection of VP1 using RT-PCR . Brain tissues with known PML were used as positive controls for each serial of RT-PCR within VP1 (data not shown).
Among HIV-1-positive patients without PML, 12 immunocompetent and 19 immunocompromised patients had detectable JCV DNA in PBMC; none presented with JCV early and late mRNA in blood, demonstrating that JCV does not productively infect the peripheral blood leukocytes from HIV-1-infected patients without PML with or without leukopenia (Table 1). Interestingly, among six patients with PML and AIDS who were positive for JCV DNA in leukocytes, none was positive for mRNA for JCV early or late genes, suggesting the absence of JCV reactivation in peripheral blood cells at the onset of neurological disease (Table 1).
PML is a fatal demyelinating disease that affects up to 4% of HIV-1-infected patients. The presence of JCV DNA in kidneys, lymphoid nodes, spleen and liver suggests a widespread haematogenous distribution in patients with PML [5,9]. The presence of JCV DNA in PBMC has been detected in groups of HIV-1-infected patients, indicating that peripheral lymphocytes can harbour JCV [18,19]. The hypothesis conferring a role to certain leukocyte subpopulations for the reactivation and transport of virus across the blood-brain barrier is attractive. This concept is supported by the observation that PML initial lesions are located around terminal arterial branches and fits with the multifocal character of these lesions [20-23]. Neverthless, few studies have focused on the role of lymphocytes before the onset of PML. Houff et al.  were first to demonstrate the presence of JCV-infected leukocytes in PML primary lesions. Recently, Dubois et al. found JCV DNA in peripheral blood lymphocytes in 35 (40.3%) of 87 HIV-1-positive patients without PML and in five of seven (71%) HIV-1-positive patients with PML . Althogether, these data suggest that peripheral blood leukocytes may play a central part in the onset of PML, their precise role in JCV latency and reactivation remains hypothetical [6,8-10,24].
In our study, JCV DNA was amplified from PBMC from 20.53% of 151 HIV-1-seropositive patients without PML. This frequency of JCV infection of peripheral leukocytes is not significantly lower than those previously published by Dubois et al. (28.5% of 157) and by Tornatore et al. (38% of 28), who used similar methods for JCV DNA detection [9,19]. Dubois et al.  have recently demonstrated that JCV DNA was intermittently amplified from peripheral blood leukocytes in 40.4% of 29 HIV-1 patients without PML, thus supporting the hypothesis of viral reservoirs in areas such as lymphoid organs. The intermittent appearance of JCV-infected peripheral mononuclear cells could well explain the low rate of JCV DNA detection observed in our study and in other studies [6,7,10].
Surprisingly, among the HIV-1-positive patients without PML included in our study, the clinical status and severity of lymphopenia, in particular a low CD4 count, did not seem to correlate with detectable JCV DNA in peripheral blood cells (Table 2). These results corroborate those initially published by others and suggest that CD4 lymphocyte depletion does not enhance the risk of infection of peripheral lymphocytes in these patients [6,9]. Moreover, independent of biological and clinical status, our data indicate that increasing age may not be a decisive factor accounting for the presence of JCV DNA in the blood leukocytes of HIV-1-positive patients without PML. Taken together, our results demonstrate that JCV-containing lymphocytes can be detected by PCR in HIV-1-seropositive patients without PML irrespective of their immunocompetence and age.
Within the group of patients with a detectable JCV DNA in CSF, only six of 10 had JCV infection of PBMC at the onset of PML. These results suggest that JCV DNA detection in PBMC is not an adequate diagnostic for PML since it may remain negative in the presence of the disease [6,9,19]. This has also been observed by other authors and could be a consequence of an intermittent appearance of JCV in peripheral blood . However, in our patients with PML, 60% were positive for JCV DNA detection in PBMC compared with 17.4% and 23.2% positive for immunocompetent and immunocompromised patients, respectively, in the HIV-1-positive group without PML (Table 1). These results could be explained by active replication and infection of mononuclear precursor cells in lymphoid tissue and bone marrow at the onset of PML, increasing the reservoir of infected PBMC. Alternatively, it may result from virus replication in only a small percentage of the subset of lymphocyte population, for example B cells of PBMC, which could be too small for detection even with our semi-nested PCR assay. This hypothesis is supported by recent results showing in vitro infection of haematopoietic precursor cells and a small percentage of CD19 B cells by JCV . However, the selective presence of JCV within peripheral B cells has only been demonstrated in PBMC from two patients with AIDS, and larger studies are necessary to define the lymphotrophic specificity of JCV further .
In our study, JCV early and late mRNA was not detectable in the PBMC of HIV-1-positive patients with or without PML in spite of demonstrable JCV DNA (Table 1). RNA for early genes such as T and t may be present in the absence of viral particle production because it encodes for regulatory factors in early and late transcription and in DNA replication [25-27].The absence of the JCV early mRNA for T and t suggests that JCV was mainly latent in blood. These results were confirmed by our results concerning VP1 mRNA, the presence of which is considered to be specific for complete and productive JCV replication . Indeed, the absence of mRNA for VP1 supported the failure to detect early mRNA for T viral protein, which is implicated in the activation of JCV late transcription [12,25].
Overall, our molecular results indicate that PBMC harbour latent JCV in immunocompetent or immunocompromised HIV-1-positive patients whether or not they were affected by PML. These results are in agreement with recent data on the JCV mRNA detection in leukocyte compartments and suggest that JCV reactivation does not take place in PBMC before the onset of PML [6,10]. However, our failure to detect JCV RNA with RT-PCR in HIV-1-infected patients with and without PML cannot exclude the possibility of active replication in a small number of JCV-infected B cells. Indeed, a very low number of peripheral B cells, as CD19 cells, harbouring replicative JCV could remain undetectable in a PBMC sample using classic or nested RT-PCR procedures (11). Therefore, it should be interesting in further studies to recover only the peripheral lymphotrophic JCV cells and to analyse them for the expression of both viral early and late gene mRNA in PBMC before the onset of PML.
In conclusion, our results agree with the theory of haematogenous distribution of virus and do not eliminate the possibility of transport and activation of virus across the blood-brain barrier [6,7]. However, our data do not indicate any direct link between peripheral virus and dissemination in the central nervous system at times of immunodepression. Moreover, our RT-PCR results suggest that active replication of JCV in PBMC appears to be absent or at least a very rare event in HIV-1-positive immunocompromised patients with and without PML. Further studies using flow cytometry followed by RNA PCR assays are needed to define precisely the capacity of JCV replication in subpopulations of lymphocytes.
Pr Didier Ingrand (laboratoire de Virologie, Centre Hospitalier Universitaire de Reims, France) kindly provided pmad-1. We thank Dr Adel (laboratoire d‚Anatomo-pathologie de l‚hôpital Henri Mondor, Paris) for providing brain tissues with PML. We are indebted to the clinicians from the Hospitals of Tourcoing and Valenciennes for their active collaboration in the inclusion of the HIV-1-infected patients during this study. We are grateful to Dr Pascal Vincent for skilful assistance in the statistical analysis.
1. Major EO, Amemiya K, Tornatore C, Houff SA, Berger J. Pathogenesis and molecular biology of progressive multifocal leukoencephalopathy, the JC virus-induced demyelinatined disease of the human brain. Clin Microbiol Rev 1992, 2:49-73.
2. Moret H, Guichard M, Matheron S, et al. Virological diagnosis of progressive multifocal leukoencephalopathy: detection of JC virus in cerebrospinal fluid and brain tissue of AIDS patients. J Clin Microbiol 1993, 31:3310-3313.
3. Henson J, Rosenblum M, Armstrong D, Furneaux H. Amplification of JC virus DNA from brain and cerebrospinal fluid of patients with progressive multifocal leukoencephalopathy. J Infect Dis 1991, 162:858-861.
4. Weber TR, Turner W, Frye S, et al. Specific diagnosis of progressive multifocal leukoencephalopathy by polymerase chain reaction. J Infect Dis 1994, 169:1138-1141.
5. Berger JR, Concha M. Progressive multifocal leukoencephalopathy: the evolution of a disease once considered rare: J NeuroVirol 1995, 1:5-18.
6. Gallia GL, Houff SA, Major EO, Khalili K. JC virus infection of lymphocytes-revisited. J Infect Dis 1997, 176:1603-1609.
7. Houff SA, Major EO, Katz DA. Involvement of JC virus infected mononuclear cells from the bone marrow and spleen in the pathogenesis of progressive multifocal leukoencephatopathy. N Engl J Med 1988, 318:301-305.
8. Major EO, Amemiya K, Elder G, Houff SA. Glial cells of the human developing brain and B cells of the immune system share a common DNA binding factor for recognition of the regulatory sequence of the human polyomavirus, JCV. J Neurosci Res 1990, 27:461-471.
9. Dubois V, Lafon ME, Ragnaud JM, et al. Detection of JC virus DNA in the peripheral blood leukocytes of HIV-infected patients. AIDS 1996, 10:353-358.
10. Dubois V, Dutronc H, Lafon ME, et al. Latency and reactivation of JC virus in peripheral blood of human immunodeficiency virus type-1 infected patients. J Clin Microbiol 1997, 35:2288-2292.
11. Monaco MCG, Atwood WJ, Gravell M, Tornatore CS, Major EO. JC virus infection of hematopoietic progenitor cells, primary B cells lymphocytes, and tonsillar stromal cells: implications for viral latency. J Virol 1996, 70:7004-7012.
12. Ishaq M, Stoner GL. Differential expression of mRNAs for JC virus large and small tumor antigens in brain tissus from progressive multifocal leukoencephalopathy patients with and without AIDS. Proc Natl Acad Sci USA 1994, 91:8283-8287.
13. Vago L, Cinque P, Sala E, et al. JCV DNA and BKV DNA in the CNS tissue and CSF of AIDS patients and normal subjects. Study of 41 cases and review of the literature. J Acquir Immune Defic Syndr 1992, 12:139-146.
14. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987, 162:156-159.
15. Taylor JJ, Heasman PA. Control gene for reverse transcription/polymerase chain reaction. Br J Haematol 1994, 86:444-447.
16. Andréoletti L, Hober D, Decoene C, et al. Detection by PCR of enteroviral RNA in endomyocardial tissue of patients with chronic cardiac diseases. J Med Virol 1996, 48:53-59.
17. Leparc I, Fuchs F, Kopecka H, Aymard M. Use of the polymerase chain reaction with a murine model of picornavirus-induced myocarditis. J Clin Microbiol 1993, 31:2890-2894.
18. Atwood WJ, Amemiya K, Traub R, Harms J, Major EO. Interaction of the human polyomavirus, JCV, with human B-lymphocytes. J Virol 1992, 190:716-723.
19. Tornatore C, Berger JR, Houff SA, et al. Detection of JC virus DNA in peripheral lymphocytes from patients with and without progressive multifocal leukoencephalopathy. Ann Neurol 1992, 31:454-462.
20. Berger JR, Kaszovitz B, Post MJ, Dickinson G. Progressive multifocal leuko-encephalopathy associated with human immunodeficiency virus infection. A review of the literature with a report of sixteen cases. Ann Intern Med 1987, 107:78-87.
21. Lang W, Miklossy J, Deruaz JP, et al. Neuropathology of the acquired immune deficiency syndrome (AIDS): a report of 135 consecutive autopsy cases from Switzerland. Acta Neuropathol 1989, 77:379-390.
22. Whiteman ML, Post MJ, Berger JR, et al. Progressive multifocal leukoencephalopathy in 47 HIV-seropositive patients: neuroimaging with clinical and pathological correlation. Radiology 1993, 187:233-240.
23. Weber TR, Major EO. Progressive multifocal leukoencephalopathy: molecular biology, pathogenesis and clinical impact. Intervirology 1994, 40:98-111.
24. Khalili K, Rappaport J, Khoury G. Nuclear factors in human brain cells bind specifically to the JCV regulatory region. EMBO J 1988, 7:1205-1210.
25. Cole CN. Polyomavirinae: the viruses and their replication. In Virology, 3th edn, Vol. 2. Edited by Fields BN, Knipe DM, Howley PM, et al., Philadelphia: Lippincott-Raven; 1996:1997-2025.
26. Frisque RJ, Bream GL, Cannelle MT. Human polyomavirus JC virus genome. J Virol 1984, 51:458-469.
27. Daniel AM, Frisque RJ. Transcription initiation sites of prototypes and variants JC virus early and late messenger RNAs. Virology 1993, 194:97-109
© 1999 Lippincott Williams & Wilkins, Inc.