WO2011124652A1 - Immunodominant peptide of polyomavirus jc and uses thereof - Google Patents

Abstract

The present invention relates to a peptide comprising the sequence selected from the group of: GMFTNRSGSQ (SEQ ID No.12), GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ (SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24) and its use in the diagnostic and/or prognostic of the several pathologies including progressive multifocal leukoencephalopathy.

Description

Immunodominant peptide of polyomavirus JC and uses thereof

FIELD OF THE INVENTION

The present invention relates to an immunodominant peptide of polyomavirus and its use in an assay for the assessment of the JCV-specific T cell responses.

In particular such assay can be used for predicting the risk of developing or for monitoring progressive multifocal leukoencephalopathy as well as other pathologies involving the JC virus. BACKGROUND OF THE INVENTION

Progressive multifocal leukoencephalopathy (PML) is a severe demyelinating infectious disease of the central nervous system (CNS) that affects person with immune dysfunction (Richardson 1983). It is caused by the reactivation of the virus JC (JCV), a human polyomavirus that has also been claimed to be responsible of other rare conditions, the cerebellar granule cell neuronopathy, JCV encephalopathy (Tan 2010) and the JCV- induced nephropathy in kidney transplanted patients. There is no established specific therapy for PML, and reversion of the immune suppression, when feasible, remains the only approach with proven efficacy to management of PML, pointing to the importance of JCV-specific host immunity in the control of this disease.

PML usually develops in the context of immune disorders, such as HIV infection, hematological malignancies or transplantation (Richardson 1983; Garcia-Suarez 2005; Cinque 2009). Interest in PML has increased in recent years because of several cases developing in patients receving immunomodulant treatments with monoclonal antibodies, such as the alpha-4 integrin antagonist natalizumab, used in multiple sclerosis (Kappos 2007; Hartung 2009), the CD 11a antagonist efalizumab, used for treatment of psoriasis (Molloy 2009) and, possibly, the anti-B-cell antigen CD20 rituximab rituximab, used in the treatment of lymphoroliferative and autoimmune disorders (Carson 2009).

PML is in most of the cases progressively fatal within a few months from onset. There is no established specific therapy, and reversion of the immune suppression, when feasible, remains the only approach with proven efficacy to management of PML (Cinque 2009). This involves initiation or augmentation of antiretro viral treatment (ART) in HIV-related PML, or reduction or elimination of immune suppressive treatments in PML cases associated with other underlying conditions. These observations point to the importance of JCV-specific host immunity in the control of PML and suggest that assessment of JCV- specific immune responses can be useful to monitor disease outcome. Previous work in patients with HIV-related PML has indeed shown that disease remission following initiation of ART (i.e., in the so-called "PML survivor" patients) is associated with restoration of JCV-specific CD4 and CD8 T-cell responses (Gasnault 2003; Du Pasquier 2004; Khanna 2009).

JCV-specific T-cell immunity has so far been assessed in PML using several approaches, including T-cell lymphoproliferation (studying the CD4 cell responses, Gasnault 2003), cytotoxic T cells (CTL) chromium release or tetramer assays (measuring the CD8 cell activity, Du Pasquier 2004) and the functional cytokine secretion and Elispot assays, measuring both CD4 and CD8 responses (Khanna 2009). Koralnik et al. have identified two main A*02-restricted JCV 9-aa epitopes of the JCV capsid protein VPl, corresponding to the VPl-p36 and VPl-pl00 peptides. VPl-p36 and VPl-plOO-specific CTLs were detected by tetramer binding in respectively 71% and 89% of PML survivors and none of PML progressors (Koralnik 2001; Koralnik 2002). Preliminary work from Hirsh's groups using the IFN-gamma Elispot with a pool of 15-mer peptides (overlapping by 11) also showed that JCV-specific T-cell responses were selectively impaired in non-survivors and that PML survivors had significant increases in JCV-specific T-cells (Khanna 2009). The use of Elispot with JCV-specific peptides appears to be a practical tool for the identification of the JCV-specific T cell immunity. However, the Elispot assay coupled with p36 and pi 00, seems to display a low sensitivity and required a two-week incubation to allow PBMC expansion and detect significant responses (Chen 2009). In vitro cell expansion is not only cumbersome, but may also introduce additional variation between experiments. In addition, T lymphocytes recognizing JCV VPl -pi 00 have been shown to cross-react with the homologous BK virus peptide, suggesting that the same epitope can be recognized in response to either BK or JC virus infection (Krymskaya 2005; Li 2006; Sharma 2006). On the other hand, using large peptide pools (Khanna 2009; Chen 2009) may also be disadvantageous because, within the pool, individual peptides may be displaced by competing peptides for MHC binding and thus responses to individual peptides be missed. This approach also does not provide information on the mechanisms involved in immune responses, e.g., role of CD4+ versus CD8+ cells and specifically involved epitopes. In JCV infection, the presence of specific T-cell responses at some times during the disease, has been associated with better outcome of PML, indicating that measuring JCV- specific T-cell responses could be used for prognostic purposes. The outcome of PML is usually unpredictable at the time of diagnosis, although there are several measurements that might help real-time monitoring of the disease after its diagnosis. These include clinical neurological assessment, brain magnetic resonance imaging and CSF examination for JCV DNA copy numbers. None of these assessments, however, provides an optimal tool for patient management: clinical assessment is often subjective and the clinical picture may even worsen in cases with virological improvement accompanied by inflammatory reaction, such as in the immune recosntitution inflammatory syndrome; PML lesions at MRI often continue to increase in volume beyond clinical and virological improvement; quantification of JCV DNA in CSF requires the lumbar puncture, which, although repeatable, is considered an invasive procedure. In this context, measuring the T-cell response through an easy and rapid procedure, such as the Elispot assay using blood PBMC, would represent an additional objective and non invasive approach that enables real-time monitoring of PML patients and help provide prognostic information in real-time. In addition, most of the PML patients show low JCV-specific CTL responses at the time of symptoms onset, thus measuring these responses could also be useful to predict patients at risk for PML. This would be particularly useful in the context of using treatments associated with PML onset, such as natalizumab or rituximab, both before starting and during treatment (Chen 2009). Indeed, criteria for risk stratification have been proposed for patients starting natalizumab treatment or already being treated with this drug. Presence of JCV- VP 1 IgG and previous treatment with immunosuppressive drugs are associated with a greater risk to develop PML. In addition, the risk seems to increase over time, with most of the cases of PML occurring after 24 months of natalizumab therapy. Therefore, the risk appears to be higher in natalizumab treated patients with positive anti-JCV IgG and who have previously been treated with immunosuppressive drugs (Gorelik 2010). In general, the study of T-cell responses during the course of JCV infection, from chronic persistent infection to widespread CNS active infection, would be key to understand the host- pathogen relationships and better define the immunological background for PML to occur. Current approaches to measurement of T-cell responses in PML have included T-cell lymphoproliferation (studying the CD4 cell responses), CTL chromium release and CD8+ tetramer identification, as well as the functional cytokine secretion and IFN-gamma Elispot assays, measuring both CD4 and CD8 responses. Recently, Elispot has been used to measure IFN-gamma release by JCV-specific T-cells after PBMC stimulation with overlapping peptide pools from both VP1 and Large T-antigen (Chen 2009; Khanna 2009; Jilek 2010). Elispot offers a relatively easy, fast and quantitative measure of the immune responses against specific antigens. The approach of using a large pool of 15mer overlapping peptides was also useful to disclose different responses between patients that could be useful for patient monitoring. However, peptides in a pool can affect each other binding, the procedure is rather costly, and, importantly for pathogenesis studies, it does not enable to assess which epitopes are involved in CTL recognition and responses.

Therefore there is the need to identify immunodominant peptides that can be used to measure JCV-specific T-cell response through an easy and rapid procedure.

The international application WO2010/090757 describes methods and compositions for determining whether a subject is at risk for PML, including subjects being treated with immunosuppressants by determining whether the subject harbors a JCV variant with reduced binding for sialic acid relative to a normal JCV.

WO2010/100182 describes an immunological method for detecting active JCV infection by screening for the presence of activated T lymphocytes against JC virus in a blood sample of a patient, by exposing the blood sample to epitopes of the virus for up to 48 hours.

WO2009/105212 discloses methods and kits for testing for the presence or absence of a polyomavirus, such as BKV, in a sample. The methods and kits are useful for quantifying BKV and differentiating BKV from JCV.

WO2009/038684 relates to compositions, methods, and kits for treating subjects infected by or at risk of infection with a DNA virus (e.g., a JC Virus or a BK virus). Aspects of the invention are useful to prevent or treat DNA virus associated conditions (e.g., PML) in subjects that are immunocompromised. Compositions are provided that inhibit intracellular replication of DNA viruses.

In WO1992/019774 methods for detecting the propensity for an individual to be affected by a polyomavirus are disclosed. The methods include an assay wherein a biological specimen from a female is contacted with at least one probe capable of determining whether the female has been exposed to a polyomavirus. A method for prophylactically treating the female is also described.

US2009/0099335 relates to HLA-A*02-restricted cellular epitopes within the VP1 polypeptide of human polyomaviruses, which are useful as diagnostic reagents for virus infection. Preferred peptides correspond to amino acids residues 107-116, 108-116 and 44- 52 of BKV VP1, and are processed in vivo in natural infection with BKV. Effector T cell populations stimulated by the peptides represent functional CTLs as assessed by cytotoxicity and cytokine production, and are reactive against cells presenting both the BKV peptides above and the JC virus homo log sequences.

The authors' approach to the study of JCV-VP1 specific responses consisted in the screening of a library of seventy 10-mer peptides spanning the entire coding sequence of the major viral capside protein- 1 (VP1, total length 354 aa), in search for immunodominant peptides.

The present invention relates to an assay for assessing the JCV-specific T-cell responses and identifies immunogenic JCV-specific VP1 epitopes.

SUMMARY OF THE INVENTION

To select immunogenic JCV-specific epitopes, the authors synthesized a library of seventy 10-mer peptides spanning the entire coding sequence of the major viral capside protein- 1 (VP1, total length 354 aa) and tested these peptides in a matrix combination for IFN- gamma Elispot from peripheral blood mononuclear cells (PBMC). The authors identified a novel major immunogenic target peptide, named VPl-p261. Responses above a cut-off of 25 SFU/10^A6 cells (established based on cut-off values used in other Elispot assays according to the authors' experience) was observed to this peptide in 12/14 (86%) JCV- seropositive healthy donors, 5/7 (71%) who developed PML and survived ("PML survivors"), 9/14 (64%>) patients with active PML, but in 0/6 JCV-seronegative healthy controls and 2/9 (20%) HIV+ non-PML controls. By comparing the sensitivity of the VP1- p261 -Elispot assay to that of Elispot using the only peptides so far described in literature, the HLA-A*02-restricted VPl-p36 and VPl-pl00 peptides, responses against VPl-p261 were observed in 10/11 (91%) HLA-A*02 PML survivors or healthy subjects and in 4/5 (80%) HLA-A*02 patients with active PML. In contrast, only a minority of HLA-A*02 individuals responded to VPl-p36 or VPl-pl00. The authors have recently showed that JCV strains present in the cerebrospinal fluid (CSF) of PML patients carry one of several VP 1 -specific substitutions at critical sites for sialic acid binding, with S269F mutation being the most commonly observed and, together with mutations at positions 265 and 267 (N265D, S267F/L/Y) accounting for 42% of all observed VP1 mutations (Gorelik, 2011). Of note, the VPl-p261 peptide of the invention is located in the proximity of the VP1 binding pocket with the sialic acid cell receptor and includes aa at position 265, 267 and 269 in its core region. Indeed, the insertion of these mutations in the VPl-p261 peptide (VPl-p261/S267L/F/Y; VPl-p261/269F), abolished or significantly reduced the Elispot response in nearly all of the healthy subjects and PML survivors.

Thus, the authors have identified a novel VP1 immunodominant peptide, named VP1- p261, which is of relevance because:

1) In combination with an assay, in particular an IFN-gamma Elispot assay, it provides a sensitive tool for assessment of JCV-VP1 specific T cell responses without requiring cell expansion; and

2) Following substitution of core aminoacids (S267F/L/Y, S269F), corresponding to common PML-associated mutations in vivo, it abrogates T-cell response, suggesting that immune escape may be a strategy by which the virus evades host immunity and establishes its progressively fatal course.

It is therefore an object of the present invention a peptide comprising the sequence selected from the group of: GMFTNRSGSQ (SEQ ID No.12), GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ (SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24) for use in the diagnostic and/or prognostic of the following pathologies: progressive multifocal leukoencephalopathy (PML), cerebellar granule cell neuronopathy and/or JCV nephropathy.

The pathologies progressive multifocal leukoencephalopathy (PML), cerebellar granule cell neuronopathy and JCV nephropathy share common features in that they are all mediated by JCV and result in JCV reactivation.

Preferably the peptide consists of the sequence selected from the group of: GMFTNRSGSQ (SEQ ID No.12), GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ (SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24).

Still preferably the peptide consists of the sequence GMFTNRSGSQ (SEQ ID No.12). In a preferred embodiment the diagnostic and/or prognostic of said pathologies is performed by measuring JCV T cell response.

Preferably the JCV T cell response is measured by means of an Elispot assay.

Still preferably the Elispot assay is an interferon-gamma Elispot assay. JCV T cell response may also be measured by methods known by the skilled person in the art such as intracellular staining methods, cytofluorimetric methods, measurement of lymphocytes proliferation.

It is another object of the invention a peptide comprising the sequence selected from the group of: GMFTNRSGSQ (SEQ ID No.12), GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ (SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24) for predicting the risk of developing progressive multifocal leukoencephalopathy (PML) in a subject being anti-JCV IgG positive.

Preferably the peptide consists of the sequence selected from the group of: GMFTNRSGSQ (SEQ ID No.12), GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ (SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24).

Still preferably the peptide consists of the sequence GMFTNRSGSQ (SEQ ID No.12). In a preferred embodiment the subject has been previously exposed to an immunomodulant treatment.

Preferably the immunomodulant treatment is natalizumab.

Still preferably the prediction of risk of developing progressive multifocal leukoencephalopathy (PML) in a subject being anti-JCV IgG positive is performed by measuring JCV T cell response.

Yet preferably the JCV T cell response is measured by means of an Elispot assay.

Preferably the Elispot assay is an interferon-gamma Elispot assay.

It is a further object of the invention a method for monitoring a JCV T cell response in a sample obtained from a subject comprising the steps of:

a) isolating PBMCs or mononucleic cells from the subject;

b) incubating the isolated PBMCs or mononucleic cells with a peptide comprising the sequence GMFTNRSGSQ (SEQ ID No.12);

c) measuring the JCV T cell response;

d) repeating steps a), b) and c) at regular time intervals;

e) comparing the measured JCV T cell response to appropriate control response.

It is a further object of the invention a method for predicting the risk of developing PML in subject being anti-JCV IgG positive or to predict the outcome of PML and/or to monitor the course of PML in a subject affect by PML comprising the steps of: a) isolating a sample of PBMCs or mononucleic cells from the subject;

b) incubating the isolated PBMCs or mononucleic cells with a peptide comprising the sequence GMFTNRSGSQ (SEQ ID No.12);

c) measuring the JCV T cell response;

d) comparing the measured JCV T cell response to appropriate control response.

Preferably step b) comprises incubating the isolated PBMCs or mononucleic cells with the peptide comprising the sequence GMFTNRSGSQ (SEQ ID No.12) and at least one further peptide comprising a sequence selected from GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ (SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24).

Preferably the peptide comprising the sequence GMFTNRSGSQ (SEQ ID No.12) consists of SEQ ID No. 12.

Still prefererably the peptide comprising a sequence selected from GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ (SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24) is a peptide consisting of SEQ ID No. 20, SEQ ID No. 21, SEQ ID No. 22, SEQ ID No. 23 or SEQ ID No. 24.

In a preferred embodiment the mononucleic cells are isolated from cerebrospinal fluid.

In a still preferred embodiment the JCV T cell response is measured by means of an Elispot assay.

Yet preferably the Elispot assay is an interferon-gamma Elispot assay.

In the present invention, the methods may comprise obtaining PBMCs from fresh blood or utilizing cryopreserved PBMCs, similarly mononucleic cells may be obtained freshly or may be used after cryopreservation. PBMCs or mononucleic cells may be then incubated with the peptide as defined above on microwell plates coated with anti human interferon-γ (IFNy) antibody, subsequently there may be provided the addition of biotinylated detection-antibody against human IFNy, of streptavidin-AP conjugate and of the substrate to develop the reaction. The method may include counting the spot-forming units (SFU) by an ELISpot reader. Values may be expressed as the mean of duplicate experiments.

JCV T cell response may also be measured by methods known by the skilled person in the art such as intracellular staining methods, cytofluorimetric methods, measurement of lymphocytes proliferation. JCV T cell response obtained in PBMCs or mononucleic cells may be analysed at regular intervals (either weekly, or every 2^nd, 3^rd or 4^th week or at longer and variable intervals) during the course of a disease such as PML or during the course of a treatment. This approach is of particular significance when immunesuppression is reduced by pharmacological or other interventions, i.e., initiation of HAA T in HIV-related PML, or, in other patients, interruption or reduction of immunosuppressive treatments, or plasma exchange to eliminate circulating immunomodulant molecules.

The analysis of the responses at the different time points may be compared to first and previous examinations to identify relative changes in responses. The appropriate control response may be a response before the start of a treatment, the response in a non JCV IgG positive subject, the response in a healthy subject, a response at various time points during the course of a disease or a treatment.

JCV T cell response may be assesses by counting SFU in specific contexts (e.g., treatment with immunomodulant or immunosuppressive drugs) and analysis of subject PBMCs or mononucleic cell may be performed before the initiation of treatment or during treatment. Depending on the results, one could then consider to treat only subjects who show responses against the peptide or the combination of peptides (peptide p261 and its mutated froms). Based on previous experience with other Elispot assays and on the responses observed against VPl-p261 in the JCV IgG negative subject, a value of about 25 SFU x 10^A6 cells may be chosen as a cut-off to distinguish responses from non-responses. Also, depending on the results, one could consider to withdraw a pharmacological treatment in a subject who shows a significant decrease in the response against the peptide above indicated, e.g., below the cut-off value, during the treatment itself.

The present invention refers to definitions as indicated in Cinque 2003.

In the present invention a PML survivor is a patient in whom PML remitted months to years before current evaluation. A PML progressor is a patient in whom PML progressed until to death. The term "Good outcome" means a patient in whom PML would remit (clinical stabilization or improvement), the term "bad outcome" means a patient in whom PML would progress (clinical progression and death). A patient with active PML is a patient with signs of active disease (often at the beginning of the disease and/or close to time of diagnosis) in whom the outcome is still unpredictable, "good" and "bad outcome" are here used to classify patients with active disease at the time of evaluation according to subsequent good or bad disease evolution. In the methods of the present invention the peptide comprising or consisting of SEQ ID No. 12, SEQ ID No. 20, SEQ ID No. 21 , SEQ ID No. 22, SEQ ID No. 23 or SEQ ID No. 24 may be used at concentration in a range of 2-4 μg/mL (per well), preferably at 2 μg/mL. The incubation time may range between 6 and 36 hours. Longer incubation times of 7 to 10 days can be used if T cell need be expanded in vitro .

In the methods of the present invention, the number of cells per well may be between 100,000 to 500,000 cells. Preferably 200,000 cells per well are used.

The invention will now be illustrated by means of non-limiting examples referring to the following figures.

Figure 1. Configuration of the 17 10-mer peptide pools created by a matrix method and identified by letters A to S. The number in the table is the number code of each peptide. Figure 2. IFN-gamma Elispot T-cell responses in 7 JCV-antibody positive healthy donors and 9 HIV-positive "PML survivors" to the 17 10-mer peptide pools created by a matrix method and identified by letters A to S.

Figure 3. IFN-gamma T-cell responses in healthy donors and PML patients against VP1- p261 peptide. Ab neg HD: JCV-VP1 IgG negative healthy donors; Ab neg HIV+ pt: JCV- VP1 IgG negative HIV positive control patients; Ab pos HD: JCV-VP1 IgG positive healthy donors; Ab pos HIV+ pt: JCV-VP1 IgG positive HIV positive control patients; PML survivors: Patients in whom PML remitted months to years before current evaluation; Active PML: Patients with signs of active disease (analysis performed at the beginning of the disease and/or close to time of diagnosis) in whom the outcome is still unpredictable. Figure 4. Prolmmune Class I REVEAL rate assay, on and off rates of VPl-p261 binding to HLA-A*02 molecule. A, on-rate of VPl-p261 binding; B, on-rate of intermediate positive control (known T-cell epitope) binding; C, on-rate of positive control (known T-cell epitope) binding; D, off-rate of VPl-p261 binding; E, off-rate of intermediate control binding; F, off-rate of positive control binding.

Figure 5. IFN-gamma T-cell responses in healthy donors and PML patients against VP1- p261 and other peptides as defined in Table 3.

Figure 6. Elispot responses in CD4-depleted and CD8-depleted blood cell populations. Figure 7. IFN-gamma T-cell responses in HLA-A*02 healthy donors and HLA-A*02 PML patients against the two previously described HLA-A*02 restricted peptides VPl-p36 and VP 1 -p 100 and the novel peptide VP 1 -p261. Figure 8. IFN-gamma T-cell responses against VPl-p261 in patients with active PML at the time of diagnosis according to subsequent disease evolution. Good outcome: patients in whom PML would remit; bad outcome: patients in whom PML would progress.

Figure 9. IFN-gamma T-cell responses against VPl-p261 in sequential samples collected from PML patients at subsequent times following diagnosis Panels A, B, C show the responses in three patients with favorable outcome following initiation of highly active antiretroviral therapy (HAART) in a patient with HIV-related PML (A), following plasma exchange (Plex) in a patient with natalizumab-related PML (B) or following interruption of immunosuppressive (IS) drugs (C); panels D, E, F show the responses in three patients with progressively fatal disease, with non-Hodgkin lymphoma (NHL)(D), or HIV-related PML, despite initiation (E) or continuation (F) of HAART.

Figure 10. IFN-gamma T-cell responses in controls and PML patients against peptide VP1- p261 and against its "mutated" peptides (as indicated in Table 3). DETAILED DESCRIPTION OF THE INVENTION MATERIALS AND METHODS

Patients and Samples

Twenty healthy controls and twenty-one PML patients were examined. Characteristics of patients and controls are shown in Table 1.

Table 1: Patients characteristics (n.a = not applicable)

Healthy subjects PML

Number 20 21

Female :male 14:6 7: 14

HIV status n.a.

HIV+ 17

HIV- 4

Outcome n.a.

Survivors 15

Progressors 6

Fresh blood samples were obtained from study subjects. PBMC were isolated by Ficoll- Hypaque density gradient centrifugation. After washing with RPMI-1640, the PBMC were resuspended and cultured in RPMI-1640 culture medium (SIGMA, Italy) supplemented with 20% fetal bovine serum (FBS, SIGMA), 2 mM of L-glutamine and 1% PenStrep (SIGMA) at concentration of 2* 10^A6/ml. Cerebrospinal Fluid was collected from PML patients by lumbar puncture as part of the diagnostic work-up, Patients gave their written informed consent to store CSF aliquots and to analyse CSF for the purposes of the present study Peptides

Seventy 10-mer peptides, overlapping by 5 amino acids, spanning the whole 354 aa long JCV VP1 protein

MAPTKRKGERKDPVQVPKLLIRGGVEVLEVKTGVDS ITEVECFL PEMGDPDEHLR GFSKSI SI SD FESDSPNRDMLPCYSVARI PLPNLNEDLTCG ILMWEAVTLKTEV IGVTSLMNVHSNGQATHDNGAGKPVQGTSFHFFSVGGEALELQGVLFNYRTKYPDG TIFPKNATVQSQVMNTEHKAYLDKNKAYPVECWVPDPTRNENTRYFGTLTGGENVP PVLHITNTATTVLLDEFGVGPLCKGDNLYLSAVDVCGMFTNRSGSQQWRGLSRYFK VQLRKRRVKNPYPI SFLLTDLINRR PRVDGQPMYGMDAQVEEVRVFEGTEELPGD PDMMRYVDKYGQLQTKML (SEQ ID No. 41); GenBank: AAA82101.1; Frisque 1984) were initially synthesized, to be pooled and analysed according to the matrix method. Subsequently, additional 7-mer, 8-mer, 9-mer and 10-mer peptides, including "mutated" peptides were also synthesized (Primm srl, Milan, Italy). All the peptides had been purified by HPLC, the peptide purity >70%, and were further analyzed by mass spectrometry. The lyophilized peptides were solved in 100% DMSO, tipically 16.5 mg/ml depending on solubility, further diluted with PBS to a final concentration of 4 mg/ml and frozen as aliquots at -20°C. Peptides were pooled (as indicated in Fig. 1) and also tested individually (those indicated in Table 4), at a final concentration of 2 ug/ml.

ELISpot assay

96-well PVDF Multiscreen-HA plates (Millipore, Italy) were coated with 100 ul of anti- human interferon-γ (IFNy) antibody (5 ug/ml; clone 2G1; Pierce Biotechnology, Rockford, USA) in PBS. The plate was incubated overnight at 4°C. The content of the plate was thoroughly washed with 200 ul/well of PBS to remove unbound antibody, and this procedure was repeated for a total of three washes. After washing, the plate was turned upside down and blotted onto paper towels. The coated wells were blocked with 200 ul media (RPMI-1640 supplemented with 20% FBS, 2 mM L-glutamine and 10% PenStrep). After two hours at 37°C, the blocking medium was discarded. Then 2* 10^A5 cells were plated in each well and pulsed with antigens (2 ug/ml/peptide) in a final volume of 200 ul/well. Plates were incubated overnight (18h) at 37°C in a 5% C0₂ incubator. Cells were removed by six washings with PBS and three with PBS/0.05% Tween-20 and 100 ul of biotinylated detection-antibody against human ΙΚΝγ (0.5 ug/ml; clone B133.5; Pierce Biotechnology) in PBS/4%> BSA were added per well. After two hours at 37°C the plates were rinsed six times with PBS, then 100 ul of streptavidin-AP conjugate (Amersham, GE Healthcare) were added to each well at a dilution of 1/1000 in PBS/0.5% BSA and incubated for 45 minutes at room temperature. Unbound complexes were removed by six washings with PBS. The authors added 100 ul of NBT/BCIP substrate solution to each well, the color reaction was developed at room temperature, in the dark, for 5-10 minutes. The content of the plate was discarded and it was rinsed with ultra pure water to stop the reaction. The plate was turned upside down, allowing the membrane to completely dry prior to analysis. The spot- forming units (SFU) were counted by an ELISpot reader. Values were expressed as mean and standard deviation for parametric variables.

In the present invention, the peptide concentration of 2 μg/mL was used as the standard, however, it may be used in the range of 2-4 μg/mL (per well). Similarly, the incubation time may range from 6 to 36 hours. Longer incubation times can also be used (7 to 10 days) if T cell need be expanded in vitro. The number of cells per well is typically 200,000 cells per well, however the number of cells may range may from 100,000 to 500,000 cells Human Leukocyte Antigen Typing

The authors characterized all the donors to verify the specificity of HLA class I binding peptides and to evaluate the strength of the peptide/HLA complex. The authors determined the major histocompatibility complex class I allele A using Sequence Specific Primer (SSP) PCR with HLA EZ-TYPE low resolution kit (GTI Diagnostic ITALIA srl). DNA was extracted from fresh or frozen blood with QIAamp DNA Blood Mini Kit, QIAGEN. Results were analyzed by using the interpretative software EZ-Type.

JCV VPl amplification and sequencing

DNA was extracted from 200 of CSF using the QIAamp Blood Kit (Qiagen) and eluted in a final volume of 50 μΐ_^. VPl was amplified either using primers flanking the whole VPl gene (full VPl PCR), or, when amplification with this method was not successful, by a semi-nested PCR assay that amplified separately shorter VPl regions (short fragment VPl PCR). The full VPl PCR consisted of a nested assay that used the outer primers VPl- LF and VP1-LR, amplifying a 2027 bp fragment; and the inner primers VP1-SF and VP1- SR, amplifying a 1233 bp long fragment (Table 2).

The PCR reaction mixture consisted of 5 μΐ. of 10X PCR buffer, 4 mM of each dNTP, 0.7 μΜ of primers VP1-LF and VP1-LR in the first round and primers VP1-SF and VP1-SR in the second round, 1.25 unit of Platinum Taq HF (Invitrogen) and 1 of extracted DNA (first round) or 2.5 μΐ of amplified product (second round) in a total volume of 50 μΐ,. Cycling parametres were (for both first and second round) 30 cycles at 94 °C for 20 sec, at 58 °C for 30 sec and at 68 °C for 90 sec in an automated thermal cycler (Applied Biosystems). The short fragment PCR consisted of a semi-nested PCR that used primers VPl-1 and VPl-4a in the first round, amplifying a 797 bp fragment, followed by two semi- nested assays with primers VPl-1 and VPl-2a, amplifying a 481 bp-long fragment, or with primers VPl-1.5 and VPl-4a, amplifying a 490 bp-long fragment (Zheng 2005, Table 2).

Table 2: Sequences of primers used for amplifiction of JCV-VP1 from cerebrospinal fluid of PML patients.

Name Sequence

VP1-LF GCAGCCAGCTATGGCTTTAC (SEQ ID No. 1)

VP1-LR GCTGCCATTCATGAGAGGAT (SEQ ID No. 2)

VP1-SF CCTCAATGGATGTTGCCTTT (SEQ ID No. 3)

VP1-SR AAAACCAAAGACCCCT (SEQ ID No. 4)

VP 1 - 1 TTGACTCAATTACAGAGGTAGAAT (SEQ ID No. 5)

VPl-4a AGAAATTGGGTAGGGGTTTTTAAC (SEQ ID No. 6)

VPl-2a AGGTACGCCTTGTGCTCTGTGTTC (SEQ ID No. 7)

VP 1-1.5 GTGCAGGGCACCAGCTTTCATT (SEQ ID No. 8)

The PCR reaction mixture consisted of 2.5 μΕ 10X PCR buffer, 200 μΜ of each dNTP, 1.5 mM MgC12, 0.5 μΜ primers and 1.25 unit of AmpliTaq Gold DNA Polymerase (Applied Biosystems). In the first round 4 μΕ of extracted DNA were added in a total volume of 25 μΐ, and the cycling parameters were 20 cycles at 94 °C for 20 sec, at 55 °C for 30 sec and at 68 °C for 90 sec in an automated thermal cycler (Applied Biosystems). This first step PCR product was purified with ExoSAP-IT PCR Clean-up Kit, the protocol consists of a single pipetting step (enzyme mixture addition), a 30-min incubation at 37 °C followed by enzyme inactivation at 80 °C for a further 15 min. In the second round of semi-nested PCR 4 μΕ of cleaned PCR product were added in a total volume of 25 μΐ, and the cycling parameters were 40 cycles at 94 °C for 20 sec, at 62 °C for 30 sec and at 68 °C for 90 sec in an automated thermal cycler (Applied Biosystems). With both assays, following amplification with the inner primers, 10 μΐ of the amplified product from the second mixture was electrophoresed on a 2% agarose gel containing 0.5 μg/ml ethidium bromide. The results were photographed under U.V. illumination and regarded as positive when a band corresponding to the expected bp long DNA fragment was present. The amplification product was purified by the Qiagen purification kit. A's were added to the ends of the cleaned up PCR product by Taq polymerase (A-overhang reaction) and cloning was carried out by the TOPO TA cloning kit (Invitrogen). Mini-prep DNA was prepared (Qiagen) from colonies containing the cloned VP1 PCR product. Two to 48 clones were sequenced for each sample sample (median 23). Following translation of the VP1 sequences, amino acid mutations were marked by comparison to the large selction of VP1 sequences from PML and non-PML cases (Sunyaev 2009). Only mutations present in more than one clone for sample were considered.

EXAMPLES

Example 1. Screening of a VP1 peptide library by INF-gamma Elispot

To analyse the immunogenicity of VP1 epitopes the authors synthesized a library of seventy 10-mer peptides spanning the entire coding sequence of the major viral capside protein- 1 (VP1, total length 699 aa) and tested these peptides in a matrix combination by IFN-gamma Elispot from peripheral blood mononuclear cells (PBMC). To identify immunogenic epitopes the authors grouped peptides into 17 different pools, identified by letters A to S, where each peptide was contained in two pools only (Anthony 2003, Precopio 2008, Figure 1).

The authors initially analysed PBMC from 7 JCV-antibody positive healthy donors and 9 HIV-positive "PML survivors", i.e., patients with no longer active disease at the time of present sampling - who had developed PML months to years ago, and in whom disease progression had halted both clinically, radio logically and virologically (Cinque 2003). Because PML remission is associated with the immune reconstitution induced by ART (Du Pasquier 2004) and these patients are still treated with antiretro viral agents, it is likely that they have developed and maintained an efficient immune response against JCV. The authors found the most frequent and highest responses against three pools, D, Q and R (Figure 2). By the matrix method, two major immunogenic target peptides were identified, corresponding to VPl-p231 (from pools D and Q, AAs 231-240, progressive number 47 in Figure 1) and VPl-p261 (from pools D and R, AAs 261-270, progressive number 53 in Figure 1, Table 3).

Table 3: Sequences of the JCV-VPl peptides used in the gamma-IFN Elispot assay

To assess the possible cross-reaction between JCV and BKV peptides, The authors compared VPl-p231 and VPl-p261 sequences to those of the analogue peptides in BK virus, (BKV), the other human polyomavirus that has 75% sequence homology with JCV and causes nephropathy in kidney transplant recipients (Table 4). The previously described VP 1 -pi 00 and VPl-p36 peptides have been shown to cross-react with BK virus, indicating that the same population of CD8+ T cell could be functionally active against both viruses (Krymskaya 2005, Li 2006, Sharma 2006, Tagaram 2008). The authors found a 100% identity for the VPl-p231, but of 70% (7/10 aa homology between the two viruses) for VPl-p261. Of note, the sequence homology of VPl-p36 was of 89%> (8/9 aa homology) and of 78% for VPl-pl00 (7/9 aa homology).

Example 2. Elispot responses against selected peptides

To assess the T cell responses against VPl-p261 the authors tested the peptide individually in PML patients and controls. Responses above a cut-off of 25 SFU/10^A6 cells (established based on cut-off values used in other Elispot assas in our laboratory) was observed to this peptide in 12/14 (86%>) JCV-seropositive healthy donors, 5/7 (71%) who developed PML and survived ("PML survivors"), 9/14 (64%) patients with active PML, but in 0/6 JCV- seronegative healthy controls and 2/9 (20%) HIV+ non-PML controls. These results show the immunogenicity of the novel peptide, and the specificity of the T-cell response elicited by VPl-p261 (Figure 3).

In general, 10-mer peptides, like those included in the JCV-VP1 peptide library, are more likely to be recognized by MHC Class I antigens and thus mediate CD8+ T cell responses. However, MHC Class II antigen mediated CD4+ T cell response may also be possible, depending on the features of the peptide. Indeed, the authors believe that the response against the authors' peptide VPl-p261 is predominantly mediated by MHC class II antigens for the following reasons:

a) using the SYFPEITHI Epitope Prediction method (http://www.syfpeithi.de/home.htm), it was predicted that the HLA antigens with better binding to VPl-p261 were HLA- A* 02 and A*l l, which are also the two most frequent alleles of the Caucasian population. However, the predicting binding score was low and, in all patient groups, the authors observed similar responses between HLA- A* 02 and non-HLA-A*02 patients, suggesting that it is unlikely for this peptide to be MHC Class I, HLA-A*02 restricted.

b) The authors assessed, by the PROIMMUNE, REVEAL™ assays, the on- and off-rates for VPl-p261 binding to MHC class I molecule HLA-A*0201. Results indicated fast binding but also fast release of JCV-VP1 from MHC, confirming that the complex HLA A*0201/VPl-p261 is not stable and thus unlikely to efficiently present the antigen to the T-cell (Figure 4). c) The authors also assessed the IFNy Elispot responses against additional shorter (including 9- and 8-mers) peptides involving aa from VPl-p261 or adjacent positions. Overall, responses against VPl-p261 remained the most frequently observed and the highest compared to those against all of its related peptides. In particular, the authors did not observed significant responses against the 9-mer peptides containing 1 aa deletion at either the N- or C-terminal of VPl-p261 (peptides VPl-p261(9) and VPl-p262(9), as indicated in Table 4), suggesting that the aVPl-p261 peptide is unlikely to include the epitope(s) involved in CD8+ T cell response (Figure 5).

d) Finally, the authors assessed, in 3 healthy donors, the response to VPl-p261 of selected PBMC populations obtained by CD4+ or CD8+ T-cell depletion using magnetic beads.

The authors observed in all case studies that the response against VPl-p261 was stronger in CD8-deleted than in the CD4-deleted T-cell population (Figure 6), supporting the previous observation in favour of a prominent MHC Class II-mediated response.

To test the sensitivity of the VPl-p261 Elispot, the authors compared the T cell responses to VPl-p261 to the responses elicited against VPl-p36 and VPl-plOO, which have previously been shown to be immunogenic and to correspond to HLA-A*02-restricted JCV epitopes (Koralnik 2001, Koralnik 2002). Significant responses against VPl-p261 were observed in 10/11 (91%) HLA-A*02 PML survivors or healthy subjects, in 4/5 (80%) HLA-A*02 patients with active PML, and in 10/20 (50%) non HLA-A*02 subjects. By contrast, only 4 subjects elicited a response against either p36 or pi 00 (Figure 7).

To assess the potential of VPl-p261 Elispot to predict PML evolution in patients with PML, the authors tested the T cell responses against this peptide both at the time of diagnosis and during the course of the disease. Among 14 patients with active PML, the authors observed a response above 25 SFU/10^A6 cells against VPl-p261 in a total of 9 patients, including 6/8 (75%) PML patients who would progress ("bad outcome") and 3/6 (50%)) with later PML remission ("good outcome", Figure 8). Thus,the presence of an increasing response to VPl-p261 is indicative of a good outcome, as also reported in Figure 9. Indeed, the longitudinal assessment through the course of PML in 6 patients with multiple samples indicated increasing T-cell responses in all the patients showing disease remission, but only in 2 out of 3 patients showing disease progression (Figure 9). Because good PML outcome may be associated with increasing T cell responses against VPl-p261 over time, measuring such responses after PML onset provides a prognostic marker of PML outcome. Example 3. Assessment of Elispot responses against wild-type and mutant peptides

It has recently been showed that JCV mutant virus containing aminoacid substitutions at several specific JCV VPl positions, including, among the others, aminoacids 55, 60, 267 and 269, are positively selected during PML from pre-existent wild-type virus within the patient (Sunyaev 2009, Gorelik, in press). Notably, these amino acids are located within the outer JCV VPl loops at critical sites for cell binding with the sialic acid residues of cell receptors (Gee 2004). By the analysis of CSF and plasma sequences from 40 patients with PML, the authors have very recently shown that JCV strains present in the cerebrospinal fluid (CSF) and plasma of PML patients carry one of several VPl -specific substitutions likely involved in cell binding, with L55F and S269F mutations being the most commonly observed (in 25% of PML patients each).

The authors noted that the serine at position 269 is in the core region in the VPl-p261 peptide. The substitution of serine with phenilalanine at position 269, because of the large difference between these two amino acids, might indeed induce steric clashes between the peptide and the MHC molecule with the consequent decrease in the affinity of their interaction (Ruppert 1993, Sunyaev 2009). In addition, mutations at positions 265 and 267 (N265D, S267F/L/Y) are also observed in vivo, and, together with the S269F mutation, account for 42% of all observed VPl mutations in PML patients (Gorelik, in press). Similarly to the S269F substitution, the substitution of serine in position 267 with leucine, phenylalanine or tyrosine, and the substitution of asparagine in position 265 with aspartic acid, could also decrease petpide-MHC interaction affinity

To test the hypothesis that VPl mutations could affect the formation of the MHC-peptide complex and thus the host immune response in PML patients carrying the mutated virus, the authors synthesized mutant VPl-p261 peptides containing the mutations at position 265, 267 and 269, namely VPl -p261/265D; VPl-p261/267L/F/Y; VPl-p261/269S.

In both controls and PML survivors, T-cell responses were significantly reduced, and in some cases completely abrogated, after PBMC stimulation with the mutated peptides compared to the wild-type VPl-p261 peptide (p<0.0001, Mann- Whitney test for all 267 and the 269 mutants, p=n.s, Figure 10; for the 265 mutant similar results were obtained, data not shown). These results suggest that VPl acquired adaptive changes may affect the host immune response against JCV and that immune escape may be a strategy by which the virus evades host immunity and establishes progressive infection. Example 4. VP1 amino acids substitution in vivo and T-cell response against peptides containing PML-specific mutation

For 12 PML patients (4 survivors, and 8 progressors, including 1 with slowly progressing "relapsing/remitting" disease) the authors were able to compare the presence of JCV VP1 mutations in CSF with the Elispot responses (Table 4).

Table 4: IFN-gamma T-cell responses in PML patients with known PML-associated CSF JCV-VP1 mutations.

Note: values indicate the number of Spot Forming Units-SFU per 10^A6 PBMC

n.d.: Not Done, Active PML-Progr.: The analysis was performed in patients with signs of active disease (at the beginning of the disease and/or close to time of diagnosis), who would later progress (bad outcome), Active PML-Survivor: The analysis was performed in patients with signs of active disease (at the beginning of the disease and/or close to time of diagnosis), who would later remit (good outcome); PML-Survivor: patients in whom PML remitted months to years before current analysis.

The authors hypothesized that PML progressors are either unable to mount any response against JCV, because of profound immunosuppression (e.g., as observed in HIV-related PML in the pre-ART era); or that their response, although adequate against wild-type virus, may not be efficient against mutated epitopes (e.g., as observed in HIV-related PML in patients efficiently treated with HAART). Thus, a PML patient whose virus carries a mutation at position 265, 267 or 269 might be unable to clear the infection if, because of his HLA repertoire, T cell response requires recognition of this epitope and no efficient responses can be elicited against different epitopes.

Only three patients (1 survivor and 2 progressors) had the S269F or the S267Y mutation in their CSF virus. However, the survivor and one of the progressors, who carried the S269F mutation, did not show a response against the wild-type VPl-p261 (response below the cut off of 25 SFU). The second progressor patient, who carried the S267Y mutation, showed an abrogated response to the mutated peptide p261/269F. The p261/267Y/L/F peptides were not tested. These results suggest that disease outcome or progression in patient carrying a mutated peptide could be monitored using both the wild type and the mutated peptide.

DISCUSSION

The authors identified a novel JCV-VP1 immunodominant peptide, which provides a sensitive tool for assessment of JCV-VP1 specific T cell responses. In addition, the authors demonstrated that substitutions of its core region (S267F/L/Y, S269F), corresponding to common PML-associated mutations in vivo, abrogate or significantly reduce T-cell responses, suggesting that immune escape may be a strategy by which the virus evades host immunity and establishes its progressively fatal course.

In JCV infection, the presence of specific T-cell responses at some times during the disease, has been associated with better outcome of PML, indicating that measuring JCV- specific T-cell responses could be used for prognostic purposes. The outcome of PML is usually unpredictable at the time of diagnosis, although there are several measurements that might help in real-time monitoring of the disease after its diagnosis. These include clinical neurological assessment, brain magnetic resonance imaging and CSF examination for JCV DNA copy numbers. None of these assessments, however, provides an optimal tool for patient management: clinical assessment is often subjective and the clinical picture may even worsen in cases with virological improvement accompanied by inflammatory reaction, such as in the immune reconstitution inflammatory syndrome; PML lesions at MRI often continue to increase in volume beyond clinical and virological improvement; quantification of JCV DNA in CSF requires the lumbar puncture, which, although repeatable, is considered an invasive procedure. In this context, measuring the T-cell response through an easy and rapid procedure, such as the Elispot assay using blood PBMC, would represent an additional objective and non- invasive approach that enables real-time monitoring of PML patients and help provide prognostic information in real-time. In addition, most of the PML patients show low JCV-specific CTL responses at the time of symptoms onset, thus measuring these responses could also be useful to predict patients at risk for PML. This would be particularly useful in the context of using treatments associated with PML onset, such natalizumab or rituximab, both before starting and during treatment (Chen 2009). Indeed, criteria for risk stratification have been proposed for patients starting natalizumab treatment or already being treated with this drug. Presence of JCV-VP1 IgG and previous treatment with immunosuppressive drugs are associated with a greater risk to develop PML. In addition, the risk seems to increase over time, with most of the cases of PML occurring after 24 months of natalizumab therapy. Therefore, the risk appears to be higher in natalizumab treated patients with positive anti-JCV IgG and who have previously been treated with immunosuppressive drugs (Gorelik 2010). In this context, the absence of an efficient T cell response against JCV or its antigen could represent an additional risk factor and help select patients who are more eligible for treatment, or, during treatment, who are at lower risk of developing PML. More in general, the study of T-cell responses during the course of JCV infection, from chronic persistent infection to widespread CNS active infection, would be key to understand the host- pathogen relationships and better define the immunological background for PML to occur. Current approaches to measurement of T-cell responses in PML have included T-cell lymphoproliferation (studying the CD4 cell responses), CTL chromium release and CD8+ tetramer identification, as well as the functional cytokine secretion and IFN-gamma Elispot assays, measuring both CD4 and CD8 responses. Recently, Elispot has been used to measure IFN-gamma release by JCV-specific T-cells after PBMC stimulation with overlapping peptide pools from both VP1 and Large T-antigen (Chen 2009; Khanna 2009; Jilek 2010). Elispot offers a relatively easy, fast and quantitative measure of the immune responses against specific antigens. The approach of using a large pool of 15mer overlapping peptides was also useful to disclose different responses between patients that could be useful for patient monitoring. However, peptides in a pool can affect each other binding, the procedure is rather costly, and, importantly for pathogenesis studies, it does not enable to assess which epitopes are involved in CTL recognition and responses.

The authors' approach to the study of JCV- VP 1 specific responses consisted in the screening of a library of seventy 10-mer peptides spanning the entire coding sequence of the major viral capside protein- 1 (VP1, total length 354 aa), in search for immunodominant peptides.

In this way the authors identified a new immunodominant 10-mer JCV-VP1 epitope (VP1- p261). This peptide was recognized by 86% of JCV-seropositive healthy donors, 71% PML 'survivors', 64% of patients with active PML, but none of JCV-seronegative healthy controls and 20% HIV+ non-PML controls. These findings were in sharp contrast to the responses to the two previously described HLA-A*02 restricted JCV-VP1 immunogenic peptides VPl-p36 or VP 1 -pi 00, which each elicited a response only in a minority of subjects.

Another peptide (VPl-p231) was also identified through the matrix method of the present inevntion, and found to elicit T-cell responses in most of the subjects examined. However, the sequence of VP 1-231 was identical to the homo log sequence of BKV, clearly indicating that these responses were not JCV-specific. On the contrary, the VPl-p261 peptide shows 70% identity with the BKV homo log peptide, lower than that of the VP1- p36 and VP 1 -pi 00, which is of 89%> and 78%, respectively. Indeed these two latter peptides were shown to cross-react with BK virus (Krymskaya 2005, Li 2006, Sharma 2006, Tagaram 2008).

While the authors' work was initially planned to look for epitopes involved in CTL responses, and for this reasons the authors screened a library of 10-mer VP1 peptides, the most immunogenic peptide, i.e., VPl-p261, is likely to be principally recognized by MHC Class II molecule. Although this is not definitely demonstrated, there are several clues that make VPl-p261 unlikely to be recognized by MHC Class I. By the use of an epitope prediction algorythm, the HLA Class I antigens with better binding with VPl-p261 were HLA-A*02 and A* 11, which are also the two most frequent alleles of the Caucasian population; however, the binding score was of 12, thus low (www.syfpeithi.de). Low affinity between VPl-p261 and HLA-A*02 was confirmed by binding experiments, which showed fast association between VPl-p261 and HLA-A*02 molecule, but also rapid dissociation of the complex, suggesting that the peptide-MHC antigen complex has low stability and unlikely to work appropriately. To better characterize the epitope(s) possibly responsible for T-cell responses in the patients studied, including whether it involved CD4+ or CD8+ T cells, the authors also synthesised several peptides of different length, all including the residue at position 267 and 269. None of them, including the 9-mer peptides deleted of 1 aa at either peptide terminal, was able to elicit responses against VP1 higher than or similar to those observed against the non-mutated original 10-mer VPl-p261 peptide. Finally, by testing recognition of the VPl-p261 peptide by CD4+ or CD8+ deleted blood cell population, better responses were observed with the latter, providing an additional clue that the peptide VPl-p261 could be primarily recognized by MHC Class II molecules.

In the present invention, an increasing response to VPl-p261 over time was observed in patients with favourable outcome, as opposed to patients with progressive disease. Thus, VPl-p261 Elispot strongly helps in the prognosis of the subsequent phases of PML and thus in the clinical management of the disease. In this perspective, the test might be performed periodically (e.g., monthly or every second week) in a patient with PML: when using an Elispot assay, a SFU value above the cut-off level or an increased SFU value compared to previous responses may predict a favourable PML outcome (remission). Any other method to monitor T cell response may be used.

The authors and others recently showed that JCV VPl mutants are selected in PML patients that carry one of several aminoacid substitutions that confer a change in binding with their cell receptor. The most frequent substitution, observed in 25% of PML patients, involves aminoacid position 269, with substitution of serine with phenylalanine. This aminoacid does indeed involve the core region of the immunodominant VPl-p261 peptide. Furthermore, two other commonly observed PML mutations are located within the authors' VPl l-p261 peptide: the 265D and the S267F/L/Y mutations. The authors thus hypothesised that a mutation at one of these positions does not only confer a change in binding affinity to cell receptor, but also affect the immune response directed against VPl . By comparing the Elispot responses against VPl-p261 to those against the peptide containing the S267L/FLY or S269F substitutions, the authors observed that the latter were abolished compared to the wild-type mutant. Thus, the authors propose that selection of the mutant in those sites is a mechanism of JCV immune evasion. There are a number of examples of viral evasion. In most of the cases, however, they involve envelope molecules of R A viruses, which are characterized by high intrinsic variability or DNA viruses, such as HBV virus, which is also characterized by high variability due to its way if replication through a reverse transcriptase enzyme. JCV is a DNA virus that is by nature less subjected than RNA viruses to random mutations in the viral genome. On the other hand, adaptive VPl mutations have been shown to occur in PML, as well as complex rearrangements of the non coding control region (NCCR). Indeed, and different from other "true" latent viruses, primarily herpes viruses, asymptomatic JCV persistence in the host following primary infection is characterized by an intense replicative activity in the urinary tract. This lifelong high-level replication may indeed favor the development of viral mutants, which, in the presence of favourable conditions, may expand and become the prevalent population. The emergence of a viral variant involving an immunodominant epitope is certainly one of the most unfavourable situation for the host, because it will impair most of the responses against the virus.

One of the most puzzling observations following the introduction of HAART, was that in approximately half of PML patients disease progression was not stopped, despite achievement of a viro-immuno logical response to HAART. This was quite in contrast to what is observed for other opportunistic infections, for which survival increased dramatically due to HAART-induced immune -reconstitution. In fact PML is the opportunistic infection with the worst prognosis and currently the second most frequent cause of HIV-related mortality. This phenomenon has been interpreted as a possible result of the inflammatory changes concomitant to immune reconstitution, which may indeed worsen the clinical picture and even lead to death. However, most of the fatal cases of PML in patients receiving HAART were not indeed characterized by an inflammatory picture, suggesting that there might be other causes to explain PML progression in these cases. The authors hypothesized here that such PML progressors are unable to clear CNS JCV infection because their T-cell response, although adequate against wild-type virus, is not efficient against mutated epitopes. Thus, a PML patient whose virus carries, e.g., the S269F mutation might be unable to clear the infection if, also related to his HLA repertoire, T cell response requires recognition of this epitope.

In conclusion, the present invention provides a new tool (VPl-p261 and related assay, in particular an Elispot based assay) that is useful for real-time monitoring of JCV T-cell specific responses in PML patients or other patients, such as patient with multiple sclerosis or HIV+ patients. VPl-p261 is highly immunogenic, i.e., it is recognised by more than two thirds of healthy controls, thus it would be an essential component of peptide pools that are to be applied to any individual to test T-cell responses for prognostic purposes.

In addition, because PML occurs in patients with impaired JCV-specific immunity, measuring JCV-specific T-cell responses may help predict patients who are more at risk of developing PML. Thus, such approach may provide useful in patients who are going to initiate or who are receiving treatments with immunomodulatory molecules that have been associated with development of PML, such as natalizumab. In the case of natalizumab treatment, low or absent T-cell response would represent an additional variable to be used for PML risk stratification of patients (together with anti-JCV IgG, previous history of immune suppressive treatments and, in those already on natalizumab, duration of treatment, Gorelik 2010). In addition, the authors disclosed a new possible mechanism of JCV evasion of the CTL response by selection of VPl mutations. This mechanism may provide an explanation to the frequently observed phenomenon by which JCV may propagate and lead to progressively fatal disease. More in general, these findings may help pave the way for new immune-based treatment approaches.

REFERENCES

Richardson EP, Jr. N Engl J Med 1961; 265: 815-23.

Tan CS, Koralnik IJ. Lancet Neurol. 2010 Apr;9(4):425-37

Garcia-Suarez J, et al., Am J Hematol 2005; 80: 271-81.

Cinque P, et al, Lancet Infect Dis. 2009 Oct;9(10):625-3

Kappos L, et al. Lancet Neurol 2007; 6: 431-41

Hartung HP. Lancet Neurol 2009; 8: 28-31.

Molloy ES, Calabrese LH. Nat Rev Rheumatol 2009; 5: 418-19.

Carson KR, et al. Lancet Oncol 2009; 10: 816-24

Gasnault J, et al, AIDS 2003; 17: 1443-9.

Du Pasquier RA, et al, Brain. 2004 Sep;127(Pt 9): 1970-8

Khanna N, et al, Clin Infect Dis 2009; 48: 1459-66.

Koralnik IJ, et al, J Immunol. 2002 Jan l;168(l):499-504

Koralnik IJ, et al, J Virol. 2001 Apr;75(7):3483-7.

Chen Y, et al, N Engl J Med. 2009 Sep 10;361(11): 1067-74

Krymskaya L, et al, J Virol. 2005 Sep;79(17): l 1170-8.

Li J, et al, J Gen Virol. 2006 Oct;87(Pt 10):2951-60.

Sharma MC, et al, Virol. 2006 Jun 20;350(1): 128-36.

Gorelik L et al. Ann Neurol. 2010 Sep;68(3):295-303.

Jilek S et al, Lancet Neurol. 2010 Mar;9(3):264-72.

Gorelik L et al, J Infect Dis., in press

Cinque P, et al, J Neurovirol 2003; 9 Suppl 1 : 88-92.

Frisque R et al, J Virol 1984 Aug;51(2):458-69. H-Y Zheng, et al, J. of General Virol. 2005, 2035-45.

S.R. Sunyaev, et al., PLoS Genetics 2009

Anthony DD, Lehmann PV. Methods 2003 Mar;29(3):260-9. MX. Precopio, et al, Cytometry A 2008 Nov;73(l l): 1071-8. Tagaram HR, et al, Virology 2008; 376: 183-90.

G.V. Gee, et al, J. of Biological Chemistry 2004, 49172-76. Ruppert, J. Sidney, et al, Cell 74: 929-937.

Claims

1- A peptide comprising the sequence selected from the group of: GMFTNRSGSQ (SEQ ID No.12), GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ (SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24) for use in the diagnostic and/or prognostic of the following pathologies: progressive multifocal leukoencephalopathy (PML), cerebellar granule cell neuronopathy and/or JCV nephropathy.

2- The peptide according to claim 1 consisting of the sequence selected from the group of: GMFTNRSGSQ (SEQ ID No.12), GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ

(SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24).

3- The peptide according to claim 2 consisting of the sequence GMFTNRSGSQ (SEQ ID No.12).

4- The peptide according to any one of claim 1 to 3 wherein the diagnostic and/or prognostic of said pathologies is performed by measuring JCV T cell response.

5- The peptide according to claim 4 wherein the JCV T cell response is measured by means of an Elispot assay.

6- The peptide according to claim 5 wherein the Elispot assay is an interferon-gamma Elispot assay.

7- A peptide comprising the sequence selected from the group of: GMFTNRSGSQ (SEQ ID No.12), GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ (SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24) for predicting the risk of developing progressive multifocal leukoencephalopathy (PML) in a subject being anti-JCV IgG positive.

8- The peptide according to claim 7 consisting of the sequence selected from the group of: GMFTNRSGSQ (SEQ ID No.12), GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ (SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24).

9- The peptide according to claim 8 consisting of the sequence GMFTNRSGSQ (SEQ ID No.12).

10-The peptide according to any one of claim 7 to 9 wherein the subject has been previously exposed to an immunomodulant treatment. 11- The peptide according to claim 10 wherein the immunomodulant treatment is natalizumab.

12- The peptide according to any one of claim 7 to 11 wherein the prediction of risk of developing progressive multifocal leukoencephalopathy (PML) in a subject being anti-JCV IgG positive is performed by measuring JCV T cell response.

13- The peptide according to claim 12 wherein the JCV T cell response is measured by means of an Elispot assay.

14- The peptide according to claim 13 wherein the Elispot assay is an interferon-gamma Elispot assay.

15- A method for monitoring a JCV T cell response in a sample obtained from a subject comprising the steps of:

a) isolating PBMCs or mononucleic cells from the subject;

b) incubating the isolated PBMCs or mononucleic cells with a peptide comprising the sequence GMFTNRSGSQ (SEQ ID No.12);

c) measuring the JCV T cell response;

d) repeating steps a), b) and c) at regular time intervals;

e) comparing the measured JCV T cell response to appropriate control response.

16- A method for predicting the risk of developing PML in subject being anti-JCV IgG positive or to predict the outcome of PML and/or to monitor the course of PML in a subject affect by PML comprising the steps of:

a) isolating a sample of PBMCs or mononucleic cells from the subject;

b) incubating the isolated PBMCs or mononucleic cells with a peptide comprising the sequence GMFTNRSGSQ (SEQ ID No.12);

c) measuring the JCV T cell response;

d) comparing the measured JCV T cell response to appropriate control response.

17- The method according to any one of claim 15 or 16 wherein step b) comprises incubating the isolated PBMCs or mononucleic cells with the peptide comprising the sequence GMFTNRSGSQ (SEQ ID No.12) and at least one further peptide comprising a sequence selected from GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ (SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24).

18- The method according to claim 17 wherein the peptide comprising the sequence GMFTNRSGSQ (SEQ ID No.12) consists of SEQ ID No. 12. 19- The method according to claim 17 or 18 wherein the peptide comprising a sequence selected from GMFTDRSGSQ (SEQ ID No. 20), GMFTNRLGSQ (SEQ ID No. 21), GMFTNRFGSQ (SEQ ID No. 22), GMFTNRYGSQ (SEQ ID No. 23), or GMFTNRSGFQ (SEQ ID No. 24) is a peptide consisting of SEQ ID No. 20, SEQ ID No. 21, SEQ ID No. 22, SEQ ID No. 23 or SEQ ID No. 24.

20- The method according to any one of claim 15 to 19 wherein the mononucleic cells are isolated from cerebrospinal fluid.

21- The method according to any one of claim 15 to 20 wherein the JCV T cell response is measured by means of an Elispot assay.

22- The method according to claim 21 wherein the Elispot assay is an interferon-gamma Elispot assay.