The effect of B-cell depletion in the Theiler's model of multiple sclerosis

Journal of the Neurological Sciences 359 (2015) 40–47

Contents lists available at ScienceDirect

Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns

The effect of B-cell depletion in the Theiler's model of multiple sclerosis Francesca Gilli a,⁎, Libin Li b,1, Sandra J. Campbell c, Daniel C. Anthony c, Andrew R. Pachner a,b a b c

Department of Neurology, Geisel School of Medicine at Dartmouth, Hanover, NH, USA Department of Neurology, University of Medicine and Dentistry-New Jersey Medical School, Newark, NJ, USA Department of Pharmacology, University of Oxford, Oxford, United Kingdom

a r t i c l e

i n f o

Article history: Received 3 August 2015 Received in revised form 22 September 2015 Accepted 6 October 2015 Available online 8 October 2015 Keywords: B cell depletion TMEV-IDD Multiple sclerosis Disability progression

a b s t r a c t B cell depletion (BCD) is being considered as a treatment for multiple sclerosis (MS), but there are many uncertainties surrounding the use of this therapy, such as its potential effect in individuals with concurrent viral infections. We sought to discover what effect BCD, induced by an anti-CD20 monoclonal antibody, would have on Theiler's murine encephalomyelitis virus-induced demyelinating disease (TMEV-IDD). Mice were injected with the anti-CD20 monoclonal antibody 5D2, 14 days before or 14 days after infection with TMEV. Efficacy of depletion of B cells was assessed by flow cytometry of CD19+ cells. Mouse disability was measured by Rotarod, viral load was measured by real time PCR for TMEV RNA. Binding and neutralizing antibody levels were determined in sera and CSF by ELISA, and in CNS by real time PCR for IgG RNA. Inflammation, microglial activation, axonal damage and demyelination were assessed using immunohistochemistry. 5D2-induced BCD was confirmed by demonstration of nearly absent CD19+ cells in the blood and lymphoid tissue. Systemic and CNS antibody responses were suppressed during 5D2 treatment. Higher viral loads were detected in 5D2-treated mice than in controls, and the viral levels correlated negatively with IgG production in the brain. Overall, 5D2 caused worsening of the early encephalitis and faster progression of disability, as well as exacerbation of the pathology of TMEV-IDD at the end stage of the disease. These data indicate that BCD in humans might worsen CNS viral infections and might not improve disability accrual in MS. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Multiple sclerosis (MS) is an idiopathic neuroinflammatory disease of the central nervous system (CNS) thought to be due to a viral or autoimmune etiology. Based to a great extent on experience gained in the EAE (experimental autoimmune encephalomyelitis) model [1], emphasis on the immunology of the disease has been predominantly on the role of T cells rather than B cells. Over the past few years, there have been two major changes in basic research into MS. First, the Abbreviations: BCD, B cell depletion; MS, multiple sclerosis; TMEV, Theiler's murine encephalomyelitis virus; TMEV-IDD, Theiler's murine encephalomyelitis virus-induced demyelinating disease; CD, cluster of differentiation; RT-PCR, reverse transcription polymerase chain reaction; RNA, ribonucleic acid; CSF, cerebrospinal fluid; ELISA, EnzymeLinked ImmunoSorbent Assay; CNS, central nervous system; IgG, immunoglobulin G; EAE, experimental autoimmune encephalomyelitis; HHV-6A, human herpes virus 6A; EBV, Epstein Barr virus; VZV, Varicella Zoster Virus; HERVs, human endogenous retroviruses; p.i, post infection; MNCs, mononuclear cells; BAbs, binding antibodies; Nab, neutralizing antibodies; mRNA, messenger ribonucleic acid; CPE, cytopathic effect; FACS, fluorescence activated cell sorting; FITC, fluorescein isothiocyanate; SCH, spinal cord homogenate; BCA, bicinchoninic acid assay; PBS, Phosphate buffered saline; IBA-1, Ionized calcium binding adaptor molecule 1; APP, amyloid precursor protein; SEM, standard error of mean. ⁎ Corresponding author at: Department of Neurology, Geisel School of Medicine at Dartmouth, 1 Medical Center Drive, Lebanon, NH 03756, USA. E-mail address: [email protected] (F. Gilli). 1 Present address: GenScript USA Inc., Piscataway, NJ, USA.

http://dx.doi.org/10.1016/j.jns.2015.10.012 0022-510X/© 2015 Elsevier B.V. All rights reserved.

importance of viral models of demyelination has been increasingly recognized [2,3]. Second, B cells have increasingly moved into the spotlight in MS research [4–6]. In both MS and its animal models, B cells and plasma cells are commonly found in active lesions [7,8] and antibodies have been identified in areas of demyelination [9,10]. In addition, B cell depletion (BCD) using anti-CD20 monoclonal antibodies [11] has resulted in decreased attacks in MS patients [12]. The purpose of this investigation was to elucidate the consequences of anti-CD20 therapy in a mouse viral model of MS. The Theiler's murine encephalomyelitis virus (TMEV)-induced demyelinating disease (TMEV-IDD) model is characterized by progressive weakness and robust systemic and CNS antiviral antibody and cellular responses [13,14], as well as a pathology that closely mimics that of MS [15,16]. We used an anti-mouse CD20 monoclonal antibody, 5D2, to evaluate the effect of BCD in TMEV-IDD. These data represent the first study of BCD in viral models of MS. 2. Methods 2.1. Mice, virus growth, viral quantitation, intracerebral inoculation All animal work utilized protocols reviewed and approved by the University Animal Care and Use Committee at UMDNJ-New Jersey Medical School. SJL mice purchased from Harlan Laboratories (Indianapolis,

F. Gilli et al. / Journal of the Neurological Sciences 359 (2015) 40–47

IN) were 4–8-week old females, and were housed in isolator cages in the Research Resource Facility at New Jersey Medical School. Mice were inoculated intracerebrally (i.c.) with 3 × 108 plaque forming units (PFU) of TMEV, BeAn strain in a final volume of 30 μL of PBS. The virus used was the BeAn strain of TMEV, obtained originally from Steven Miller (Northwestern University), and passaged in a hamster fibroblast line, BHK, as previously described [17]. PFU were determined by a cytopathic effect (CPE) assay. Mice were necropsied at every 10 days post infection (p.i.) for Expt. (− 14) until day 90 p.i. For Expt. (+ 14), biweekly necropsy was performed until day 150 p.i. Techniques used in were performed as previously described [18,19], including anesthesia, i.c. injections, perfusion with PBS, CSF collection by cisternal tap, and the collection of blood and other tissues. 2.2. Anti-CD20 treatment The monoclonal anti-mouse CD20 antibody 5D2 was obtained through a Material Transfer Agreement between Genentech and UMDNJ. Pilot experiments were performed to determine optimal dosing for prolonged depletion of peripheral blood B cells, which was found to be 10 mg/kg intraperitoneally (i.p.) twice a month. In Expt. (− 14), treatment was initiated with 5D2, 15 mice, or control IgG, 15 mice, 14 days prior to infection with TMEV. Expt. (+14) was identical except that treatment was begun at 14 days p.i. 2.3. Enumeration of mononuclear cells (MNCs) in the CNS, total IgG and anti-TMEV binding antibody, and neutralizing antibody (NAb) assay

41

Gaithersburg, MD). cDNA was finally used as a template for the real time RT-PCR analysis based on the 5′ nuclease assay. Transcriptional expression was normalized using the housekeeping gene glyceraldehyde-phosphate-dehydrogenase (GAPDH) as reference. The IgG mRNA level was expressed as the relative expression index of two to the power of −ΔCt, where −ΔCt = −(CtIgG − CtGAPDH). Differently, to determine the viral load in spinal cords, a standard curve was first obtained by plotting the Ct values of the samples against known TMEV concentration; copy number was determined by regression analysis from the standard curve.

2.5. Rotarod testing for progressive disability Progressive disability in mice was assessed as previously described [18]. Rotarod data were expressed as a neurological function index (NFI). The NFI value at any time point was the mean of the last 3 time indices divided by the mean time indices from day 15 to day 45 after infection. Time indices were the time on the Rotarod for that day divided by the mean of the 2 maximum times for that mouse.

2.6. B-cell enumeration FACS analysis for CD19-positive B cells was used to determine adequacy of depletion of B cells, since CD19 and CD20 expression are similar [22]. MNCs from the various tissues or blood underwent FACS after labeling with a FITC-conjugated donkey anti-mouse CD19 antibody (Jackson Immunoresearch, West Grove, PA).

MNCs from CNS tissues were harvested by Percoll density gradient centrifugation (70%/30%) as described previously [18,19]. Total viable cells were then enumerated. Capture ELISA was used to quantitate total IgG in both CSF and serum. In brief, donkey anti-mouse IgG (Jackson ImmunoResearch Laboratories, West Grove, PA) was used as coating antibody, and horseradish peroxidase-conjugated donkey anti-mouse IgG (Jackson) was used as detecting antibody. A purified murine IgG (Jackson ImmunoResearch Laboratories) was serially diluted as a standard positive control on each plate. TMEV-specific BAb level was determined as previously described [20]. Each sample was assigned an arbitrary unit of anti-TMEV binding activity according to a standard curve derived from a known positive control in the anti-TMEV ELISA. TMEV-specific neutralizing antibody (TMEV-NAb) was measured using the modified cytopathic effect assay (CPE) [13]. In brief, 1500 PFU of TMEV was incubated with serial 3-fold dilutions of mouse serum and CSF for 2 h, followed by incubation with BHK cells for 48 h in a 96-well microplate format. Live BHK cells were then exposed to naphthol blue-black dye, and absorbance read on an ELISA reader at 620 nm. Neutralizing titers were expressed as the inverse of that dilution of serum able to block 90% of the cytopathic effect of the virus; i.e. to lower the cytopathic effect of 1500 PFU of virus to 150 PFU of virus under conditions in which the change from 1500 PFU to 150 PFU is in the linear part of the curve of the cytopathic effect dilution curve.

After perfusion of mice with PBS, spinal cords were removed by dissection, and immediately placed in 4% paraformaldehyde at 4 °C. Tissue was processed to paraffin wax and cut in a microtome, 10 μm in the axial plane. The spatial extent of the entire spinal cord was assayed to assess for local effects. For immunohistochemical analysis, the tissue was de-waxed in xylene and rehydrated through alcohols. Primary antibodies against markers for microglial activation (IBA-1, Ab5076), T cell (CD3, Ab5076) and axonal damage (amyloid precursor protein—APP) were obtained from Abcam (Cambridge, MA, USA). Biotinylated secondary antibodies and avidin– biotin complex were from Vector Laboratories (Peterborough, UK). Avidin–horseradish peroxidase was obtained from DAKO (Cambridge, UK). Immunocytochemistry was carried out by the avidin–biotincomplex method with minor modifications depending on the antibody used [23]. For Luxol fast blue staining (Klüver and Barrera method), tissue sections were incubated with 0.1% Luxol fast blue (Sigma, St. Louis, MO, USA) at 56 °C overnight. Then, the slides were soaked in 0.05% lithium carbonate solution, distilled water, and 70% ethanol. Finally, the slides were dehydrated in absolute ethanol, cleared in xylene, and mounted. Lesion severity was evaluated by an examiner who was blinded to the experimental conditions [24].

2.4. Real time reverse transcription polymerase chain reaction (RT-PCR) for in situ IgG production and TMEV viral load

2.8. Statistical analysis

Expression of IgG and TMEV RNA was tested as previously described with specific primers and probes for murine IgG1 and TMEV [19,21]. Total RNA was isolated from fresh, homogenized tissue samples using TRIzol® RNA Isolation Reagents (Life Technologies, Grand Island, NY); total RNA (50 ng/uL) was then reverse transcribed using random hexamer primers with the qScript™ cDNA SuperMix (Quanta Biosciences,

All data are shown as mean ± SEM. The nonparametric Mann– Whitney U-test and Kruskal–Wallis tests were used for the statistical analysis, which was performed using GraphPad Prism version 6.00 for Mac (GraphPad Software, San Diego, California, USA). All reported p values are based on two-tailed statistical tests, with a significance level of 0.05.

2.7. Histology, immunohistochemistry and demyelination (Luxol fast blue)-method of Klüver–Barrera staining

42

F. Gilli et al. / Journal of the Neurological Sciences 359 (2015) 40–47

3. Results 3.1. 5D2 treatment depleted CD19+ cells in peripheral and CNS lymph organs Anti-CD20 antibodies have demonstrated efficacy in depleting CD20+ B cells in human studies and animal models of neuroinflammatory diseases [25–27]. In the present study, we determined the ability of the anti-mouse CD20 antibody, 5D2, to deplete total MNC or B cells in the TMEV-IDD model. No change of total MNC count was found after 5D2 treatment, except in the spleen (Fig. 1A), where total MNCs were lower in BCD mice, consistent with the high percentage of B cells normally found in spleen. CD19+ cell percentages were then determined in MNCs isolated from various organs of mice with 5D2 or control treatment. In all tissues that we tested, 5D2 depleted at least 95% of CD19+ cells, except for the spinal cord, in which there was 78% decrease in CD19+ cells (Fig. 1B). 3.2. BCD abolished antibody production in the CNS and diminished antibody production outside of the CNS Antibody production within the CNS is a prominent feature of TMEV-IDD, as well as in MS [8,28–30]. We characterized both peripheral and CNS antibody responses to evaluate effects of BCD in the TMEV model. The following antibody measures were used in both serum and CSF: total IgG, anti-TMEV BAb by ELISA, and anti-TMEV NAb by CPE assay. In addition, levels of IgG mRNA in the spleen and spinal cord were quantitated by real time RT-PCR from RNA to determine local production of IgG. 3.2.1. Total IgG We measured IgG concentrations in sera and CSF from mice. In infected animals treated with control IgG, total IgG increased in the

CSF from a baseline of about 5 μg/mL prior to infection to approximately 180 μg/mL. This increase was almost completely abolished by BCD. Similarly, the rise in serum IgG concentrations from 500 μg/mL to more than 8000 μg/mL was also prevented by BCD treatment (Fig. 2A and B). Increases in IgG concentrations in the CSF and serum in the control mice in these experiments have also been seen in TMEV-infected controls treated with saline, and in these experiments were thus not related to the very low concentrations of normal mouse IgG injected as negative controls for the 5D2 injections. In Expt. (− 14), IgG levels in BCD mice were 17.3% and 7.2% in sera and CSF respectively to that of control mice. In Expt. (+ 14) mice, these numbers were 25.5% and 12.5% correspondingly. 3.2.2. Anti-TMEV BAbs and NAbs We measured both TMEV-specific BAbs and NAbs. BAb measurements are a general assay for all antigen-specific antibodies, while NAb assays measure antibodies that are of high affinity and bind to neutralizing epitopes of the virus. Both BAbs and NAbs increased in sera and CSF of infected mice during the course of disease. NAb levels in the CSF were very high late in the disease and approached the levels seen in corresponding sera (Fig. 2E and F). BCD markedly lowered BAbs and NAbs in sera and CSF (Fig. 2C–F), an effect more dramatic in CSF than in sera. For example in Expt. (+ 14) mice, BAbs in BCD mice were 11% to that of control mice in sera, but were 4% in CSF. NAbs in BCD mice were 16% to that of control mice in sera, and were 10% in CSF. 3.3. Effect of BCD on IgG mRNA levels in tissues of spleen and spinal cord We measured IgG mRNA levels as a measure of local IgG production in both peripheral and CNS tissues by real time RT-PCR. In the spleen, IgG expression was down-regulated in BCD animals (Fig. 3). IgG production within the parenchyma of the spinal cord is a feature of chronic TMEV-IDD and is readily detected by demonstration of high levels of IgG mRNA in spinal cord; IgG mRNA was decreased significantly after BCD (Fig. 3). 3.4. BCD resulted in greater viral load in the CNS Previous studies have shown that TMEV infection is limited to the CNS in TMEV-IDD. In the first month of infection viral levels in the brain decrease, while levels in the spinal cord rise to a high plateau level in chronic disease [31]. We assessed viral load in the CNS tissues by measuring TMEV RNA levels by real time RT-PCR. In the brain, the usual clearance of virus in the second half of the first month of infection was delayed after BCD treatment in both Expt. (−14) and Expt. (+14) animals. The overall viral levels were relatively higher in the BCD groups than in control animals (Fig. 4). The difference between the BCD and control groups was significant in Expt. (−14), but not in Expt. (+14). 3.5. BCD exacerbated the progressive disability of TMEV-IDD

Fig. 1. Depletion of B cells from blood and tissues by 5D2. MNCs were isolated from peripheral blood, deep cervical lymph nodes (DCLN), superficial cervical lymph nodes (SCLN), other peripheral lymph nodes (OPLN), spleen and spinal cord. The results for blood represent means of serial samples, while the results for tissues are means from necropsy samples obtained every two weeks after infection. A: MNC counts in each organ. B: percentage of CD19+ cells in MNCs. Data are representative of Expt. (+14) mice. (Mean ± SEM, n ≥ 15 in each organ.)

As in previous studies [18,20], 90% of TMEV-infected animals developed severe progressive disability, beginning at about 2 months after infection. BCD exacerbated the disability progression; both Expt. (−14) and Expt. (+14) mice had worse Rotarod performances when treated with 5D2 (Fig. 5A and 5B). The difference between these two groups was in the early phase. In Expt. (−14), a markedly worse disease appeared in the first 2 weeks after infection. This acute phase of the disease has shown characteristics of flaccid paralysis (i.e. poliomyelitis), but the potentiation of the infection was not sufficient to lead to outright paralysis. The animals recovered somewhat in the following 2–3 weeks, and then began the chronic progression of disability; their performance on the Rotarod never caught up with the control animals. In Expt. (+ 14) mice, this worsening did not appear in the early phase, because BCD did not start until 2 weeks after infection. BCD

F. Gilli et al. / Journal of the Neurological Sciences 359 (2015) 40–47

43

Fig. 2. Suppression of antibody responses during the course of TMEV-IDD. Antibody levels in sera (A, C, E) and CSF (B, D, F) were assessed for the following: total IgG (A and B), anti-TMEV BAbs (C and D), and anti-TMEV NAbs (E and F). Measurement is shown as a line figure representing the time course. Data are from Expt. (+14) animals. Mean ± SEM, values are shown for three to four animals at each time point.

Fig. 3. Suppression of local antibody production by 5D2 as measured by real time RT-PCR of RNA from spleen and spinal cord of Expt. (−14) (A) and Expt. (+14) (B) mice. Relative IgG1 expression was expressed as relative IgG mRNA level in each 0.5 μg total RNA from tissue, expressed as 2 to the power of inversed delta Ct (IgG1-GAPDH). Each measurement is shown as a clustered column figure showing mean values of all animal samples. (Mean ± SEM, n ≥ 14 in each subgroup. * p b 0.05.)

44

F. Gilli et al. / Journal of the Neurological Sciences 359 (2015) 40–47

Fig. 4. Increased viral load in mouse CNS tissues of infected mice after treatment assessed by real-time RT-PCR of viral RNA. Brain and spinal cord from Expt. (−14) (A) and Expt. (+14) (B) were analyzed; each is shown as mean values of all animals in the subgroups. Viral load values are expressed as viral RNA copy equivalents referring to a standard of known quantity of in vitro- transcribed BeAn virus RNA. Since the viral load in the brain usually drops to near zero by the third month, only samples from days 20–60 were used for viral load in the brain. Similarly, since spinal cord levels early in infection are very low and then rise to plateau by day 45, only samples from after the first month were used in this analysis for viral load in the spinal cord. Statistical comparisons were between controls and BCD (Mean ± SEM, n ≥ 14 in each subgroup. * p b 0.05).

treated mice began to become weaker relative to control mice around day 50 and this gap between the two groups worsened with time. The worsening of disease correlated with higher viral load in the CNS, consistent with our previous studies and that of others showing correlation between viral levels and disability [18,31]. 3.6. BCD exacerbated the pathology of TMEV-IDD at the end stage of the disease as determined by histology Inflammation, microglial activation, axonal damage and demyelination were assessed in Expt. (+ 14) mice through the spatial extent of the spinal cord using immunohistochemistry for markers for T cells, microglia, and APP and Luxol-fast blue staining for myelin. There was no significant microglial signal in day 4, while some minimal staining was present on day 20 which increased at later time points; however, no obvious difference was observed between the two treatment groups for the first 47 days (20, and 47 days p.i,). However, by day 77 and more prominently by day 111 and 134, the 5D2-treated group displayed significantly more abnormalities than the controls for all four of the assayed markers. This difference is exemplified by Iba1 staining at days 111 and 134 (Fig. 6A–D). The spatial extent of T-cell infiltration, demyelination and axonal damage mirrored the areas of microglial activation and the principal sites affected were the ventral root exit zones throughout the cord. In addition, cresyl violet staining revealed some evidence of neuron loss in the anterior horn, which was more prominent in the 5D2-treated animals at the later time points.

4. Discussion This manuscript is the first describing the effect of BCD in the Theiler's virus model of MS. Chronic TMEV infection in SJL mice is an excellent model for the progressive accrual of disability in MS, since the hallmarks of both diseases are CNS inflammation, progressive demyelination and axonal injury, leading to progressive neurological disability [3]. B lymphocytes are essential contributors to both innate and adaptive immunity. Their depletion by anti-CD20 antibodies in other animal models of disease has had profound effects [25–27,32– 34]; the Theiler's model of MS is no exception. Our work demonstrated that BCD in TMEV-IDD worsened neurological disability, abolished CNS anti-viral antibody responses, increased viral load in the CNS, and resulted in greater neurological disability. We used an anti-CD20 monoclonal antibody to deplete B cells. CD20 is one of the specific B-cell markers, which are expressed at various stages of B cell differentiation; CD20 is expressed on all stages of B cell development except very early cells such as stem cells or early pro-B cells or late, highly differentiated B cells such as plasmablasts and plasma cells [22]. Thus, following anti-CD20 injection, B cells in the peripheral blood are rapidly depleted, but stem cells in the bone marrow are unaffected, allowing the generation of new B cells. The plasma cells in the bone marrow, lymph nodes, and spleen are also unaffected [35]. Since we sought persistent B cell depletion, mimicking the situation in clinical trials of BCD in patients with MS [36], 5D2 was injected every two weeks to ensure that replenishment of the B cell population by stem cells did not occur.

Fig. 5. Worse CNS disability in 5D2-treated mice as determined by Rotarod testing. A: Expt. (−14) mice; B: Expt. (+14) mice. Mean ± SEM, values are shown for 15 mice at each time point. * p b 0.05.

F. Gilli et al. / Journal of the Neurological Sciences 359 (2015) 40–47

45

Fig. 6. Immunohistochemical detection of Iba1 in coronal sections of spinal cord. No qualitative or quantitative differences were observed between the IgG control group (A and C) as compared to the 5D2-treated group (B and D), for the first 77 days. By day 77, and more prominently by day 111 (A and B) and 134 (C and D), a clear separation was observed between the two groups; the structure of the cord was disrupted in the 5D2B-treated group (B and D) because of the extent of the inflammatory pathology whereas the control group (A and C) exhibited significantly less pathology and was essential intact although qualitatively the pattern was similar. Following 5D2-treatment, there were prominent cuffs of activated microglia around the majority of vessels and within the white matter tract. By comparison to the 5D2-treated group, in the IgG controls the microglial activation was more focal compared to the obvious widespread activation observed in the white matter of the 5D2B-treated animals.

Depletion of B cells from peripheral blood is readily achieved with anti-CD20 monoclonal antibodies, but it is not known how much depletion of B cells within other compartments in the body is required for optimal therapeutic effect. In the current study, we used frequent dosing of an anti-mouse CD20 monoclonal antibody, 5D2, which was highly effective in depleting B cells in peripheral blood, lymph nodes, and spleen, but less effective in depleting B cells in the spinal cord. The reason for this relative lower efficiency of depletion is unclear. The optimal dose of anti-CD20 monoclonal antibodies for any disease is a critical issue, since higher doses will likely be more immunosuppressive. Identifying the optimal dose in TMEV-IDD or MS is essential; the trial of the B-cell depleting anti-CD20 antibody, ocrelizumab, in rheumatoid arthritis was halted because of concerns about excess deaths from infections [37]. BCD initiated at day −14 and day +14 relative to infection resulted in abolition of the virus-specific antibody response in the CNS detecting by measuring anti-viral antibodies in the CSF. Following introduction of antigens into the CNS, antigens drain into cervical lymph nodes [38,39]; antibody-secreting cells in the CNS are presumably derived from plasmablasts originating in these cervical lymph nodes, which migrate to the CNS, similar to CNS T cells [39,40]. This sequence of the establishment of an antigen-specific response takes longer in the CNS than in the peripheral system [41]. In support of this, we found in previous studies using ELISpot analysis of IgG secreting cells that ASCs do not appear in significant numbers in the CNS until around 40 days p.i. during TMEV-IDD [19]. Also supporting this delayed establishment of a CNS antibody response is the late appearance of increased levels of IgG in the CSF, about 45 days p.i. in infected mice treated with control IgG in this study. These data indicate that timing of BCD relative to infection is critical since, once established in the CNS, antibodysecreting cells may not be susceptible to BCD, as demonstrated by the inability of BCD to affect CSF measures of CNS antibody production in MS [42].

The issue of timing of BCD relative to infection and subsequent MSlike disease is important for understanding the implications of these studies to clinical situations. Results from post-infection treatment regime experiment are most relevant to patients taking BCD agents who develop a neurotropic virus infection. Our data would predict that these patients would develop worse disease from the viral infection than untreated individuals. Results from the post-infection treatment regime experiment might be also relevant to BCD treatment in very early MS prior to the development of disability or extensive demyelination. Although BCD in MS results in decreased attacks [12], its effect on long-term disability progression has been shown to be only marginal and possibly relapse-reduction-dependent [43]. Accordingly, our data would predict that long-term accrual of disability might be either unaffected or even worsened in MS patients treated with BCD; since the TMEV-IDD model does not include “attacks”, this aspect of MS could not be studied in our experiments. This disconnection between effects of therapies on disability relative to attacks would not be a surprising outcome since many of our current MS agents, which substantially decrease attack rates may have no significant effect on long-term outcomes [44]. The persistence of cytopathic viruses in the CNS has been reported as being predominantly controlled by levels of anti-viral neutralizing antibodies (NAbs) [45]. In our TMEV model, we investigated changes in levels of viral load in the CNS and NAbs during the infection. During the early phase, levels of virus are high in the brain, largely found in neurons in the gray matter; in chronic disease, levels of viral RNA copies in the brain are relatively low, while spinal cord viral load is high, thought to be primarily in macrophages and microglia in the white matter [31]. After B cells were depleted in our experiments, diminished NAb levels in the blood and CSF were associated with significant elevation of viral loads in both brain and spinal cord, consistent with the role of NAbs in controlling virus persistence.

46

F. Gilli et al. / Journal of the Neurological Sciences 359 (2015) 40–47

The above findings on decreased NAbs and elevated CNS viral load are consistent with our findings on histopathology within the CNS. Microglia are a critical cell population in this infection [46] and in our studies were significantly more highly activated in BCD mice than controls. This increased microglial activation, as demonstrated by higher Iba1 expression, was also associated with increased demyelination, axonal injury, and T cell infiltration. Interestingly, a major goal of the study was to determine whether BCD would ameliorate the disability progression of TMEV-IDD, which closely mimics the disability progression of MS [14,47,48]. However, both when mice were B-cell depleted prior to and after infection, BCD exacerbated clinical disease. This was most striking when 5D2 was initiated prior to infection (Expt. − 14). This severe acute worsening of the disease was also seen in the EAE model induced by myelin oligodendrocyte glycoprotein peptide (MOG33–35) [25,34], in which BCD prior to MOG immunization exacerbated EAE. This was thought to be due to heightened inflammation caused by removal of IL-10-producing CD20+ regulatory B10 cells by BCD in EAE. The similar timing and exacerbation of disease symptoms in our studies and those in EAE suggest that a similar B10 mechanism may be operative in our system. The disability in TMEV-IDD was also worsened in post-infection treated mice, but the effect was less dramatic and was delayed until after 50 days p.i. The second month p.i. is a time period in which the antibody production in the CNS in control mice becomes prominent, and the effects seen in the BCD mice may have been due to the reduction of CNS IgG production in these mice. The effects of locally produced IgG in the CNS in TMEV infection are likely complex. The fact that mice had worse neurological function when CNS antibody production was reduced is consistent with a protective role for CNS IgG, in balance. Protection could be accomplished by suppression of viral replication [45,49], clearance of myelin debris [50], promotion of remyelination [51], or other mechanisms. This positive role for some populations of locally produced IgG does not rule out pathological effects for other populations of CNS IgG. These findings may have considerable relevance to the use of B cell depletion in MS and the risks of viral infections in patients receiving BCD for other diseases. Patients with MS receiving BCD therapies may develop highly inflammatory syndromes because of the absence of B regulatory cells, or worsened CNS or systemic infections because of the inability to mount a strong anti-pathogen humoral immune response. In addition, although BCD appears to be effective in suppressing attacks [12] and in decreasing gadolinium-enhancing lesions in MS patients [52], our results do not support an ameliorative effect of BCD on disability progression. Although the relevance in MS of these observations is still speculative at this time, our data suggest that the biology of BCD needs to be better understood prior to its extensive use in human neuroinflammatory diseases. Conflict of interest FG has received research support from Biogen Idec. DCA is in receipt of a research grant from Roche to investigate the role of B cells in demyelinating disease and has acted as a scientific advisor. ARP has received personal compensation for activities with Biogen Idec, EMD Serono, Novartis, Hoffman-LaRoche, Schering AG, NovoNordisk, Biomonitor, and Teva Marion as a consultant. ARP has also received research support from Hoffman-LaRoche, Biogen Idec, and Sanofi-Aventis. LL and SJC declare no competing interests. Acknowledgments This work was supported by a grant from Roche, awarded to the University of Medicine and Dentistry-New Jersey Medical School (UMDNJ). The funding company had no role in the design, collection, analysis, and interpretation of data, nor were they involved in the writing of the

manuscript and the decision to submit it for publication. The authors thank Rob Glanzmann at Roche and Andy Chan at Genentech for their assistance.

References [1] E. Mix, H. Meyer-Rienecker, H.P. Hartung, U.K. Zettl, Animal models of multiple sclerosis—potentials and limitations, Prog. Neurobiol. 92 (2010) 386–404. [2] B. Wootla, M. Eriguchi, M. Rodriguez, Is multiple sclerosis an autoimmune disease? Autoimmune Dis. 2012 (2012) 969657. [3] A.R. Pachner, Experimental models of multiple sclerosis, Curr. Opin. Neurol. 24 (2011) 291–299. [4] A.H. Cross, E. Waubant, MS and the B cell controversy, Biochim. Biophys. Acta 1812 (2011) 231–238. [5] J.L. Browning, B cells move to centre stage: novel opportunities for autoimmune disease treatment, Nat. Rev. Drug Discov. 5 (2006) 564–576. [6] F. Hoffmann, E. Meinl, B cells in multiple sclerosis: good or bad guys?: an article for 28 May 2014 — World MS Day 2014, Eur. J. Immunol. 44 (2014) 1247–1250. [7] J.W. Prineas, F. Connell, The fine structure of chronically active multiple sclerosis plaques, Neurology 28 (1978) 68–75. [8] A.R. Pachner, L. Li, D. Lagunoff, Plasma cells in the central nervous system in the Theiler's virus model of multiple sclerosis, J. Neuroimmunol. 232 (2011) 35–40. [9] C.P. Genain, B. Cannella, S.L. Hauser, C.S. Raine, Identification of autoantibodies associated with myelin damage in multiple sclerosis, Nat. Med. 5 (1999) 170–175. [10] M. Krumbholz, E. Meinl, B cells in MS and NMO: pathogenesis and therapy, Semin. Immunopathol. 36 (2014) 339–350. [11] M.D. Pescovitz, Rituximab, an anti-cd20 monoclonal antibody: history and mechanism of action, Am. J. Transplant. 6 (2006) 859–866. [12] S.L. Hauser, E. Waubant, D.L. Arnold, T. Vollmer, J. Antel, R.J. Fox, et al., B-cell depletion with rituximab in relapsing-remitting multiple sclerosis, N. Engl. J. Med. 358 (2008) 676–688. [13] H.L. Lipton, F. Gonzalez-Scarano, Central nervous system immunity in mice infected with Theiler's virus. I. Local neutralizing antibody response, J. Infect. Dis. 137 (1978) 145–151. [14] A.L. Nelson, A.J. Bieber, M. Rodriguez, Contrasting murine models of MS, Int. MS J. 11 (2004) 95–99. [15] M.C. Dal Canto, H.L. Lipton, Primary demyelination in Theiler's virus infection. An ultrastructural study, Lab. Investig. 33 (1975) 626–637. [16] J. Prineas, Pathology of the early lesion in multiple sclerosis, Hum. Pathol. 6 (1975) 531–554. [17] M. Trottier, P. Kallio, W. Wang, H.L. Lipton, High numbers of viral RNA copies in the central nervous system of mice during persistent infection with Theiler's virus, J. Virol. 75 (2001) 7420–7428. [18] A.R. Pachner, J. Brady, K. Narayan, Antibody-secreting cells in the central nervous system in an animal model of MS: phenotype, association with disability, and in vitro production of antibody, J. Neuroimmunol. 190 (2007) 112–120. [19] A.R. Pachner, L. Li, K. Narayan, Intrathecal antibody production in an animal model of multiple sclerosis, J. Neuroimmunol. 185 (2007) 57–63. [20] L. Li, M. Matsumoto, T. Seabrook, C. Cojean, V. Brinkmann, A.R. Pachner, The Effect of FTY720 in the Theiler's Virus Model of Multiple Sclerosis(submitted for publication) 2011. [21] K. Narayan, D. Dail, L. Li, D. Cadavid, S. Amrute, P. Fitzgerald-Bocarsly, et al., The nervous system as ectopic germinal center: CXCL13 and IgG in lyme neuroborreliosis, Ann. Neurol. 57 (2005) 813–823. [22] T.W. LeBien, T.F. Tedder, B lymphocytes: how they develop and function, Blood 112 (2008) 1570–1580. [23] C. Cunningham, R. Deacon, H. Wells, D. Boche, S. Waters, C.P. Diniz, et al., Synaptic changes characterize early behavioural signs in the ME7 model of murine prion disease, Eur. J. Neurosci. 17 (2003) 2147–2155. [24] S.K. Kim, K.O. Cho, S.Y. Kim, White matter damage and hippocampal neurodegeneration induced by permanent bilateral occlusion of common carotid artery in the rat: comparison between wistar and Sprague–Dawley strain, Korean J. Physiol. Pharmacol. Off. J. Korean Physiol. Soc. Korean Soc. Pharmacol. 12 (2008) 89–94. [25] M.S. Weber, T. Prod'homme, J.C. Patarroyo, N. Molnarfi, T. Karnezis, K. LehmannHorn, et al., B-cell activation influences T-cell polarization and outcome of antiCD20 B-cell depletion in central nervous system autoimmunity, Ann. Neurol. 68 (2010) 369–383. [26] M. Ramos-Casals, I. Sanz, X. Bosch, J.H. Stone, M.A. Khamashta, B-cell-depleting therapy in systemic lupus erythematosus, Am. J. Med. 125 (2012) 327–336. [27] Y.S. Kap, N. van Driel, J.D. Laman, P.P. Tak, B.A. t Hart, CD20+ B cell depletion alters T cell homing, J. Immunol. 192 (2014) 4242–4253. [28] M.S. Freedman, E.J. Thompson, F. Deisenhammer, G. Giovannoni, G. Grimsley, G. Keir, et al., Recommended standard of cerebrospinal fluid analysis in the diagnosis of multiple sclerosis: a consensus statement, Arch. Neurol. 62 (2005) 865–870. [29] S.I. Tschen, S.A. Stohlman, C. Ramakrishna, D.R. Hinton, R.D. Atkinson, C.C. Bergmann, CNS viral infection diverts homing of antibody-secreting cells from lymphoid organs to the CNS, Eur. J. Immunol. 36 (2006) 603–612. [30] J. Bankoti, L. Apeltsin, S.L. Hauser, S. Allen, M.E. Albertolle, H.E. Witkowska, et al., In multiple sclerosis, oligoclonal bands connect to peripheral B-cell responses, Ann. Neurol. 75 (2014) 266–276. [31] M. Trottier, B.P. Schlitt, A.Y. Kung, H.L. Lipton, Transition from acute to persistent Theiler's virus infection requires active viral replication that drives proinflammatory cytokine expression and chronic demyelinating disease, J. Virol. 78 (2004) 12480–12488.

F. Gilli et al. / Journal of the Neurological Sciences 359 (2015) 40–47 [32] K. Yanaba, Y. Hamaguchi, G.M. Venturi, D.A. Steeber, E.W. St Clair, T.F. Tedder, B cell depletion delays collagen-induced arthritis in mice: arthritis induction requires synergy between humoral and cell-mediated immunity, J. Immunol. 179 (2007) 1369–1380. [33] A. Ahuja, J. Shupe, R. Dunn, M. Kashgarian, M.R. Kehry, M.J. Shlomchik, Depletion of B cells in murine lupus: efficacy and resistance, J. Immunol. 179 (2007) 3351–3361. [34] T. Matsushita, K. Yanaba, J.D. Bouaziz, M. Fujimoto, T.F. Tedder, Regulatory B cells inhibit EAE initiation in mice while other B cells promote disease progression, J. Clin. Invest. 118 (2008) 3420–3430. [35] N.L. Monson, The natural history of B cells, Curr. Opin. Neurol. 21 (Suppl. 1) (2008) S3–S8. [36] B. Barun, A. Bar-Or, Treatment of multiple sclerosis with anti-CD20 antibodies, Clin. Immunol. 142 (2012) 31–37. [37] P. Emery, W. Rigby, P.P. Tak, T. Dorner, E. Olech, C. Martin, et al., Safety with ocrelizumab in rheumatoid arthritis: results from the ocrelizumab phase III program, PLoS ONE 9 (2014), e87379. [38] P.M. Knopf, C.J. Harling-Berg, H.F. Cserr, D. Basu, E.J. Sirulnick, S.C. Nolan, et al., Antigen-dependent intrathecal antibody synthesis in the normal rat brain: tissue entry and local retention of antigen-specific B cells, J. Immunol. 161 (1998) 692–701. [39] Navarrete-Talloni MJ, Kalkuhl A, Deschl U, Ulrich R, Kummerfeld M, Rohn K, et al. Transient peripheral immune response and central nervous system leaky compartmentalization in a viral model for multiple sclerosis. Brain Pathol. 20:890–901. [40] Y.V. Mendez-Fernandez, M.J. Hansen, M. Rodriguez, L.R. Pease, Anatomical and cellular requirements for the activation and migration of virus-specific CD8+ T cells to the brain during Theiler's virus infection, J. Virol. 79 (2005) 3063–3070. [41] C.J. Harling-Berg, T.J. Park, P.M. Knopf, Role of the cervical lymphatics in the Th2type hierarchy of CNS immune regulation, J. Neuroimmunol. 101 (1999) 111–127. [42] L. Piccio, R.T. Naismith, K. Trinkaus, R.S. Klein, B.J. Parks, J.A. Lyons, et al., Changes in B- and T-lymphocyte and chemokine levels with rituximab treatment in multiple sclerosis, Arch. Neurol. 67 (2010) 707–714.

47

[43] K. Hawker, P. O'Connor, M.S. Freedman, P.A. Calabresi, J. Antel, J. Simon, et al., Rituximab in patients with primary progressive multiple sclerosis: results of a randomized double-blind placebo-controlled multicenter trial, Ann. Neurol. 66 (2009) 460–471. [44] A. Shirani, Y. Zhao, E. Kingwell, P. Rieckmann, H. Tremlett, Temporal trends of disability progression in multiple sclerosis: findings from British Columbia, Canada (1975–2009), Mult. Scler. (Houndmills, Basingstoke, England) 18 (2012) 442–450. [45] S.A. Stohlman, C. Ramakrishna, S.I. Tschen, D.R. Hinton, C.C. Bergmann, The art of survival during viral persistence, J. Neurovirol. 8 (Suppl. 2) (2002) 53–58. [46] Y.H. Jin, M. Mohindru, M.H. Kang, A.C. Fuller, B. Kang, D. Gallo, et al., Differential virus replication, cytokine production, and antigen-presenting function by microglia from susceptible and resistant mice infected with Theiler's virus, J. Virol. 81 (2007) 11690–11702. [47] D.B. McGavern, L. Zoecklein, K.M. Drescher, M. Rodriguez, Quantitative assessment of neurologic deficits in a chronic progressive murine model of CNS demyelination, Exp. Neurol. 158 (1999) 171–181. [48] H.L. Lipton, M.C. Dal Canto, Chronic neurologic disease in Theiler's virus infection of SJL/J mice, J. Neurol. Sci. 30 (1976) 201–207. [49] T.U. Metcalf, D.E. Griffin, Alphavirus-induced encephalomyelitis: antibody-secreting cells and viral clearance from the nervous system, J. Virol. 85 (2011) 11490–11501. [50] M.E. Vargas, J. Watanabe, S.J. Singh, W.H. Robinson, B.A. Barres, Endogenous antibodies promote rapid myelin clearance and effective axon regeneration after nerve injury, Proc. Natl. Acad. Sci. U. S. A. 107 (2010) 11993–11998. [51] K. Asakura, M. Rodriguez, A unique population of circulating autoantibodies promotes central nervous system remyelination, Mult. Scler. (Houndmills, Basingstoke, England) 4 (1998) 217–221. [52] L. Kappos, D. Li, P.A. Calabresi, P. O'Connor, A. Bar-Or, F. Barkhof, et al., Ocrelizumab in relapsing-remitting multiple sclerosis: a phase 2, randomised, placebo-controlled, multicentre trial, Lancet 378 (2011) 1779–1787.