Multiple Sclerosis (MS) is one of the most frequent neurological disorders causing disability in young adults and affects approximately 2.5 million people worldwide [1],[2]. An interplay between both susceptibility genes and still unknown environmental factors is considered to be causative of the disease. Due to acute inflammatory demyelination and axonal loss with partly structural repair and recovery of function, most patients suffer from a relapsing-remitting course.

In order to develop new therapeutic strategies and a better understanding of this autoimmune disorder of the central nervous system (CNS), intensive research efforts dealing with the underlying disease pathomechanisms have been undertaken. For a long time, mainly T cells were considered as the initiator and perpetuator of the disease. However, during the last two decades, the important role of B cells as antigen presenting cells and producers of autoantibodies in the pathogenesis of MS has increasingly been appreciated [3]-[5]. In particular, the role of B cells in MS is hardly understood. This is surprising because intrathecal antibody synthesis and oligoclonal IgG of yet unknown specificity are a diagnostic hallmark of MS [6]. Clonally expanded B cells persist in the CNS of MS patients [7] and antibody deposition with concomitant complement activation represents the most frequently observed pattern of demyelination in MS brain lesions [8]. Consistently, plasmapheresis can be beneficial in exacerbations in relapsing forms of MS [9]. Although for decades no MS-specific autoantibody has been identified, the discovery of antibodies against the potassium channel KIR4.1 in a substantial proportion of MS patients has revived interest in antibody-mediated autoimmunity in MS [10]. Work performed in MS-like preclinical models suggests a role for B cells in initiating inflammatory responses in the CNS [11] and treatment of relapsing-remitting MS (RRMS) patients with the B cell depleting monoclonal antibody rituximab rapidly and markedly reduced active CNS inflammation [12]. A similar effectiveness was shown for alemtuzumab [13] and ofatumumab [14]. Nonetheless, today there is still no first-line treatment option in MS that specifically targets B cells and B cell subsets.

The multiple lines of evidence for a contribution of B cells to the disease pathogenesis raise the question whether a sub-typing of patients according to their B cell response in the peripheral blood is not only possible, but may also permit the identification of B cell-dependent MS, thus paving the way for a target-oriented and individualized therapy.

To this end, we have recently introduced an assay based upon the enzyme-linked immunospot technique (ELISPOT) for the detection of CNS antigen-specific B cells in the blood of patients with MS. These B cells only occurred in MS patients and were absent in healthy donors and in patients with other inflammatory and non-inflammatory neurological diseases as well as other autoimmune disorders [15]. Our previous analyses focused on measurements of the brain antigen-specific B cell response after 96 h of polyclonal stimulation. Here we extend our findings by introducing a direct ex vivo assay for patients with clinical manifestations of an acute MS relapse. This assay allowed us to visualize acute ongoing B cell immune responses to antigens prominent in the CNS in a subgroup of patients and to correlate this response to clinical relapse parameters.

After binding of a specific antigen to the B cell receptor and its presentation to a corresponding effector T cell, B cell proliferation and differentiation into plasma cell precursors and memory B cells occur. Whereas antibody producing plasma cells are predominantly located in the bone marrow after emigration from the lymphatic follicles, resting B lymphocytes recirculate in the body and can be converted into antibody-producing plasma cells with the help of polyclonal stimulation in vitro. Only in the context of a relapse and at the stage of the emigration from the lymphatic follicles to the bone marrow, plasma cells become detectable in the blood and can be directly analyzed for CNS specificity.

PBMC from 30 patients with MS experiencing an acute relapse were analyzed for their response to brain antigen using the ELISPOT technique. In order to differentiate between an acute and a memory B cell response, we introduced two different assay types. In a direct ex vivo approach PBMC were incubated on the plates for 24 h immediately after separation from the blood sample without any prestimulation. Alternatively, PBMC were tested after 96 h of polyclonal stimulation with R-848, IL-2 and β-mercaptoethanol, which is an established method for the activation of resting memory B cells [19]. Secreted antibodies were captured on the ELISPOT plates and visualized as spots that corresponded to the numbers of brain antigen-specific B cells.

Distinct size morphology of B cell spots produced directly ex vivo and after polyclonal stimulation in ELISPOT assays

In a total of nine patients brain antigen-specific B cells could be detected in the blood directly ex vivo and after polyclonal stimulation (Table 2). We compared the morphology of spots produced directly ex vivo and after polyclonal stimulation using a specialized ImmunoSpot® image analysis software. Spots that were produced directly ex vivo were much more distinct and significantly smaller than spots produced after 96 h of polyclonal stimulation (compare a mean spot size of 0.008282 mm² ± 0.00852 mm² to 0.033343 mm² ± 0.052619 mm²; p < 0.001) (Figure 1). These results underline the notion that spots detected directly ex vivo correspond to antibodies produced by recently activated and still recirculating plasma cells, while spots produced after polyclonal stimulation are derived from resting memory B cells that produce significantly larger spots following in vitro reactivation.

Variable	Pattern I	Pattern II	Pattern III
No. of patients	9	6	15
Age – yrs
Mean	27.9 ± 7.4	41.8 ± 13.7	42.4 ± 12.4
Range	18-42	22-59	18-61
Female sex – no. (%)	9 (100)	5 (83.3)	10 (66.7)
EDSS in remission†
Mean	3.1 ± 1.1	3.5 ± 1.7‡	3.5 ± 2.0‡
Range	1.5-5.0	2.0-5.5	1.0-7.0
MS severity¶
Mean	73.9 ± 16.0	81.8 ± 10.1‡	71.4 ± 20.2‡
Range	51-91	72-93	37-94
RRMS/SPMS	8/1	4/2	12/3

EDSS = Expanded Disability Status Scale; RRMS = relapsing-remitting MS; SPMS = secondary progressive MS; SD = standard deviation.

^*Plus-minus values are means ± SD.

^†Scores on the EDSS range from 0 to 10, with higher scores indicating a greater degree of disability.

^‡One patient was lost to follow-up.

^¶MS severity refers to the percentile rank of each individual study patient compared to a matched MS cohort of the MSBase Registry. Values were determined using MS Curves[18].

Ongoing brain-specific B cell activity is detectable in the blood in a subset of patients with an acute disease relapse

In addition to the nine patients that showed a positive B cell response both directly ex vivo and after polyclonal stimulation, we identified six of 30 patients that did not show any B cell activity associated with the relapse, but were classified as B cell positive donors only after polyclonal stimulation. The remaining 15 of 30 patients displayed brain antigen-specific B cells neither in direct ex vivo assays nor after 96 h of stimulation. These data suggest that the MS patients tested in our study fall into three different categories, depending on the presence of an actively ongoing and/or CNS antigen-specific memory B cell response in the blood. These categories are summarized as “patterns” in Tables 1 and 2 and Figure 2. No significant differences were found between the patterns in relation to relapse or disease severity (Table 2). However, patients who were classified as “pattern I” with a positive B cell response both directly ex vivo and after polyclonal stimulation were significantly younger than the cohort of patients represented by both pattern II and III (p = 0.003). In order to delineate that the CNS antigen-specific B cell response was not a transient phenomenon, but a characteristic feature of a disease subtype, we retested 21 MS patients in clinical remission three to nine months after relapse. In 17 of 21 patients the CNS antigen-specific B cell response detected after polyclonal stimulation was comparable to the results obtained during relapse. Four patients who had been tested negative initially, now showed a positive response. Loss of a brain antigen-specific B cell response in previously positive patients was not observed. Importantly, the formerly evident B cell activity in nine of 30 patients as measured directly ex vivo was absent in remission. Brain antigen-specific B cells were also absent directly ex vivo in healthy donors (n = 17) and patients with other neurological or other inflammatory neurological diseases (n = 22).

Association between the development of brain-specific B cell responses in the blood and acute disease reactivation

One interesting finding pertained to a patient who originally did not display any CNS antigen-specific B cell responses directly ex vivo and after polyclonal stimulation. In this patient clinical examination three months after relapse revealed a recovery of function and no signs for a renewed clinical disease exacerbation despite the persistence of paresthesia. However, a positive brain antigen-specific B cell response was evident in direct ex vivo ELISPOT testing indicating recent immune reactivation. Ten days later the patient was admitted to the hospital with symptoms of a clinical relapse including deficits in visual, cerebellar, sensory as well as motor functions. ELISPOT analysis confirmed the presence of plasma cells actively secreting brain antigen-reactive antibodies. The patient was examined at three subsequent time points during which we were able to demonstrate the disappearance of brain antigen-specific plasma cells from the blood. At the same time, a brain antigen-specific memory B cell response became detectable (Figure 3). The dramatic drop in the number of B cell spots measured directly ex vivo and after polyclonal stimulation within four days between the third and fourth measurement may be explained either by the high dose intravenous glucocorticoid therapy that was administered at the time point of relapse for three consecutive days [20] and/or by the migration of the autoreactive B cells into the target tissue, that is the CNS.

A positive brain-specific B cell response in the blood is predictive of a subsequent relapse

The association between a positive brain-reactive B cell response in the blood in direct ex vivo assays and the development of a consecutive relapse as shown for one case in Figure 3 suggests that the presence of brain-reactive B cells in the blood could be linked to a higher relapse rate. To test this hypothesis, we followed n = 8 patients that displayed a brain-reactive B cell response in the blood directly ex vivo and after polyclonal stimulation (pattern I), n = 5 patients that displayed brain-reactive B cells only after polyclonal stimulation (pattern II) and n = 14 patients with a negative B cell response in both assays (pattern III). The patients were followed for a period of ten months and the time to the next relapse after initial testing was recorded. Data are summarized in Table 4. The Kaplan-Meier plot shows that the relapse-free interval in patients that were tested positive in direct ex vivo ELISPOT assays was significantly shorter than in patients that were classified as pattern II or III (p = 0.0005) (Figure 4). There was no significant difference between patients with pattern II versus pattern III (p = 0.348). The hazard ratio for the development of a consecutive relapse in the setting of a positive direct ex vivo brain-reactive B cell response in the blood (pattern I) compared to patients with pattern II or III was 6.08 (95% confidence interval 1.87-19.77).

Time [months]	Pattern I	Pattern II	Pattern III
0	8	5	14
1	7	5	14
2	7	5	14
3	5	5	14
4	4	3	14
5	3	3	12
6	2	3	12
7	2	3	12
8	1	3	12
9	1	3	11
10	1	3	11

In the pathogenesis of MS, multiple lines of evidence indicate an important role of B cells and autoantibodies [21]-[24]. For decades, intensive research efforts were made to identify the target antigen in MS, but antibodies against myelin components or other structures prominent in the CNS were also frequently found in other neurological diseases and/or healthy individuals [4],[25]. The assumption that antibodies are pathogenic in the development of MS is mainly supported by histopathological findings that provided evidence for the frequent deposition of immunoglobulins and complement factors in MS brain lesions [8]. In the past an association between the neuropathological pattern II defined by Lucchinetti et al. and a benefit from plasma exchange was suggested [26] and treatment with the B cell-specific monoclonal antibodies rituximab, alemtuzumab or ofatumumab were effective in ameliorating disease severity by means of reduction in the number of total gadolinium-enhancing magnetic resonance imaging (MRI) lesions as well as lower annualized relapse rates [12]-[14]. It has initially been suggested that in particular antibodies against myelin oligodendrocyte glycoprotein (MOG) and myelin basic protein (MBP) were predictive of a conversion from a clinically-isolated syndrome (CIS) to clinically definite MS [27]. However, a subsequent corroboration of these results failed [28]. Here we show that the presence of a brain-specific memory B cell response in the blood as measured by ELISPOT was associated with an increased risk of the development of a MS relapse.

In our previous work we have demonstrated that the detection of brain antigen-specific B cells in the PBMC population permitted the identification of a B cell-dependent subtype of MS [15]. These data were in line with earlier reports that showed the presence of proteolipid protein (PLP)- and MOG-specific B cells in the blood and cerebrospinal fluid (CSF) of patient with MS using the ELISPOT approach [29],[30]. The data presented here extend these findings not only by presenting the predictive value of this test, but also by the introduction of a direct ex vivo assay that indicated actively ongoing disease and was associated with clinical disease reactivation in a fraction of patients. It is tempting to speculate that the presence of brain-reactive B cell responses in the blood can be used to subdivide MS patients into different categories as suggested in the current study. A categorization of patients has already been done following different patterns of demyelination in brain lesions and the most frequently observed pattern was characterized by the deposition of antibodies. Interindividual heterogeneity in the patterns of demyelinative pathology has been suggested [8]. Our data support this concept, but imply that there might also be intraindividual heterogeneity in regard to the brain-specific B cell response over the course of the disease.

Acute clinical relapses of MS often lead to the deterioration of clinical symptoms and the failure of functional CNS systems. As yet, the evidence-based standard treatment for MS relapses is a high-dose intravenous glucocorticosteroid pulse therapy. Since relapse treatment can only be initiated when new clinical symptoms of MS are evident and persist, there is always a risk of the development of irreversible deficits. The ability to detect relapses before they become clinically evident and their consecutive early treatment would provide an option to prevent the accumulation of CNS damage. A subtyping of blood immune responses as suggested here might be one possible option along these lines. Even if the CSF is also relatively easy to access and it might be argued that immune responses in the CSF reflect the pathogenic processes in the CNS more closely, the risks of side effects and the ethical problems with exposing patients to repeated spinal taps emphasize the clinical and practical advantage of a test that can be performed on peripheral blood.

The data presented here strengthen the central role of B cells in the immune pathogenesis of MS. It is conceivable that our results will help to identify patients with B cell-/antibody-dependent MS and relapses, thereby guiding the development and use of B cell-directed therapeutic strategies. It remains to be elucidated if the detection of recirculating B cells that produce CNS-specific antibodies ex vivo will allow the diagnosis of MS reactivation even before the occurrence of clinically evident symptoms, which would help to facilitate the initiation of early treatment that could potentially include plasmapheresis [26]. Finally, the detection of antibody-producing B cells in MS patients corroborates the autoimmune hypothesis of the disease and its association with clinical disease parameters.