Journal of Neurology

, Volume 256, Issue 8, pp 1228–1235

Development of ALS-like disease in SOD-1 mice deficient of B lymphocytes


  • Shulamit Naor
    • Department of Immunology, Bruce Rappaport Faculty of MedicineTechnion-Israel Institute of Technology
  • Zohar Keren
    • Department of Immunology, Bruce Rappaport Faculty of MedicineTechnion-Israel Institute of Technology
  • Tomer Bronshtein
    • Faculty of Food Engineering and BiotechnologyTechnion-Israel Institute of Technology
  • Efrat Goren
    • Faculty of Food Engineering and BiotechnologyTechnion-Israel Institute of Technology
  • Marcelle Machluf
    • Faculty of Food Engineering and BiotechnologyTechnion-Israel Institute of Technology
    • Department of Immunology, Bruce Rappaport Faculty of MedicineTechnion-Israel Institute of Technology
    • Bruce Rappaport Faculty of Medicine and Rappaport Family Institute for Research in the Medical SciencesTechnion-Israel Institute of Technology
Original Communication

DOI: 10.1007/s00415-009-5097-3

Cite this article as:
Naor, S., Keren, Z., Bronshtein, T. et al. J Neurol (2009) 256: 1228. doi:10.1007/s00415-009-5097-3


Several recent studies proposed a role for innate immunity and inflammation in the pathogenesis of amyotrophic lateral sclerosis (ALS). However, possible links, if any, between disease and adaptive immunity are poorly understood. The present study probed for the role of B cells in ALS disease using the G93A-SOD-1 transgenic mouse model. In agreement with other studies, we show here that autoantibodies are detectable in SOD-1 mice. However, SOD-1 B cells did not express any altered phenotype and exhibited indistinguishable responsiveness to immunogenic stimuli relative to wild-type B cells. This was obtained for B cells isolated before, during and after the onset of ALS-like disease. Finally, to obtain an in vivo conclusion, we generated SOD-1 mice that are deficient of B cells, by crossing SOD-1 mice with Igμ-deficient mice (μMT), where B cell development is blocked at the proB stage. The meteoric assays performed on a rota-rod clearly showed the development of ALS-like disease in SOD-1 mice that are deficient of B cells not differently than in control SOD-1 mice. Our results propose that B lymphocytes do not have a major role in the pathogenesis of ALS-like disease in SOD-1 mice.


Amyotrophic lateral sclerosis (ALS)B lymphocytesAntibodiesAutoimmunity


Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease characterized by a loss of brain and spinal cord motor neurons with clinical manifestation of progressive muscular weakness, paralysis and death of patients within 3–5 years [27, 35]. The pathogenesis of ALS is incompletely understood, although different hypotheses have been suggested, including mitochondria dysfunction [45], mutation in the superoxide dismutase gene [9], defects in neuronal glutamate transports [13], and autoimmunity [4]. Despite of all this, no effective remedy is available today for these patients. In about 20% of the patients missense mutation (G93A) in the Cu/Zn superoxide dismutase (SOD1) were identified as the primary cause for this fatal disease (familial ALS [2, 3, 14]). These findings have led to the development of a transgenic mouse model, which express the mutant human SOD-1 protein and develop symptoms of familial ALS-like disease. However, most cases of ALS are sporadic and their etiology is unknown [43].

Inflammatory mechanisms and immune reactivity are hypothesized to play a role in the pathogeneses of central nervous system (CNS) diseases [26, 33]. Numerous studies have shown that autoreactive T and B lymphocytes are responsible for the onset of the disease, and these cells are the target of many drug therapies [15]. Studies now suggest that dysregulation in the immune system may also contribute to the pathogenesis of ALS [6, 41]. The observations proposing a role of the immune system include inflammatory signals [1, 25], activated microglia [36], infiltration of lymphocytes into the spinal cord [42], depositions of IgG antibodies [11], and dysregulation of cytokine expression in the spinal cord of ALS patients [18]. Recently, increased amounts of complement components have been shown in serum, cerebrospinal fluid and neuronal tissue of diseased ALS individuals and in SOD-1 animal models [46]. The findings of autoantibodies to neuronal antigens in patients with lower motor neuron syndromes (LMNs) [30] may also support a potential involvement of the immune system in ALS. Amyotrophic lateral sclerosis occurs both as a sporadic and a familial disease with a similar pathology of neuronal damage [7]. Thus, although most studies utilize the SOD-1 model for familial ALS, it is hoped that therapies effective in mutant SOD1 models will translate to sporadic ALS.

B lymphocytes are key regulators in the autoimmune process. In addition to antibody secretion, other B cells functions include antigen presentation, T cell regulation and cytokine production [16, 23, 47]. Targeted B cell therapy appears to be very efficient treatment to many autoimmune diseases including rheumatoid arthritis, lupus erythematosus and multiple sclerosis [10, 17]. There are several studies proposing a role of B lymphocytes in the pathogenesis of ALS by showing that autoantibodies to neural antigens are produced in ALS patients. Thus, anti-neural antibodies (GM1-gangliosides, AGM1-gangliosides and sulfatides) were found in serum and cerebrospinal fluid of ALS patients [29], and antibodies isolated from ALS patients inhibit dopamine release mediated by L-type calcium channels [31]. The present study was undertaken to determine whether or not B lymphocytes contributes to the pathogenesis of ALS-like disease in SOD-1 mouse. We show here that SOD-1 B cells are indistinguishable from wild-type (wt) B cells by both phenotypic marker expression and responsiveness to immunogenic stimulations and that lack of B cells does not prevent, inhibit or reduce ALS-like disease in SOD-1 mice. These results argue against an essential role of B cells in the pathogenesis of ALS.

Materials and methods

Experimental mice

The mice used for the experiments were 60–130 days old normal C57Bl6 or deficient of B cells (μMT) [21]. Mice transgenic for the mutant G93A form of human superoxide dismutase-1 gene (SOD-1) were purchased from Jackson Laboratory, Bar Harbor, USA. Typing for SOD-1 G93A transgene was carried out by PCR reaction using a tail DNA template. Primer used for SOD-1 are oIMR0113 (5′-CATCAGCCCTAATCCATCTGA-3′) and oIMR0114 (5′-CGCGACTAACAATCAAAGTGA-3′) [34]. The PCR for the wild type IL-2 gene was carried out for control using primers oIMR0042 (5′-CTAGGCCACAGAATTGAAAGATCT-3′) and oIMR0043 (5′-GTAGGTGGA AATTCTAGCATCATCC-3′). The PCR conditions were 90 s at 94°C followed by 37 cycles as follows: 45 s at 94°C, 60 s at 59°C and 45 s at 72°C. Final elongation at 72°C lasted 10 min. In some experiments, SOD-1 Tg and μMT mice were crossed to generate SOD-1 mice that are deficient of B cells. The mice were housed and bred at the animal facility of the Faculty of Biotechnology and Food Engineering, Technion, and all studies were approved by the committee for the supervision of animal experiments. In our mouse colony, the onset of the ALS-like disease occurs at around 90 days after birth and the sick mice tend to die between days 140–180. In these experiments, the principles of laboratory animal care (NIH publication No. 86-23, revised 1985) were followed, as well as specific national laws as approved by the institutional ethic committee.

Cell culture and stimulation

Mature B cells were purified untouched from spleens after negative selection by magnetic beads (B cell isolation kit, Miltenyi Biotec, Bergisch Gladbach, Germany). Cells were cultured in supplemented DMEM tissue culture medium (Biological Industries, Kibbutz Beit Haemek, Israel). For stimulation, we used the synthetic ODNs 1826 (CpG-ODN) [44] (InvivoGen, San Diego, USA) at 100 nM, or with LPS (50 μg/ml) (Sigma). Cells were stimulated for 96 h and proliferation was measured by the MTT assay [20]. Supernatants of stimulated cells were collected and analyzed for antibodies by ELISA (see below).

Flow cytometry and cell sorting

Single cell suspensions from spleens were stained for surface markers using FITC-, PE-, APC and biotin-conjugated antibodies, followed with streptavidin PerCP. Abs used were as follows: B220, RA3-6B2; CD19, 1D3; CD5, 53-7.3 (Pharmingen, Becton Dickinson, San Diego, USA); CD45, 13/2.3; CD69, H1.2F3 [Southern Biotechnology Associates Inc. (SBA), Birmingham, USA]; rat anti mouse IgM (Zymed Laboratories, San Francisco, CA, USA); IgD JA12.5. Data for four-color analysis were collected on a FACSCalibur™ (BD Biosciences, Immunocytometry Systems, Mountain View, CA, USA).and analyzed using FlowJo software.

Tissue array

Tissue array analysis was carried out as described [38]. Briefly, normal mouse tissue samples from healthy mice were homogenized in lysis buffer and after centrifugation, cleared tissue lysates were separated SDS-PAGE gel and transferred to nitrocellulose membrane blots. After blocking, membranes were incubated with sera collected from SOD-1 or from wt mice (diluted 1:500–1:1,500). HRP conjugated goat anti-mouse IgG (Jackson) was used for detection. Visualization of specific bands was performed by ECL reaction.

Analysis of antibodies

Serum samples or supernatants of stimulated cells were analyzed for total amounts of IgM and IgG by sandwich ELISA using specific goat anti-mouse IgM or IgG polyclonal reagents (SBA) as described [37]. Antibody concentrations were calculated using a reference standard curve.

Motor analysis—rota-rod test

Neuromuscular function of SOD1 G93A transgenic mice was assessed using the rota-rod test on an ENV-575 five station mouse USB rota-rod treadmill, MED Associates Inc., St. Albans, USA. Following a 1-week pre-training session on the rotating rod, 70-day-old mice were subjected to gradually increasing rotation speeds from 2.5 to 25 RPM over a 5 min time course and were then subjected to the maximal rotation speed (25 RPM) for an additional 3 min. The time to fall (seconds) and maximal speed achieved (RPM) were recorded twice a week. In each test session the protocol was repeated for 3–5 times and the average value for each time point was assessed. Performance of the mice on the rota-rod was measured as the time the animal holds (seconds)/age(days), or as the maximum speed reached MAX–V(RPM)/age(days). The slopes (dV/dT) of the linear fits were calculated and their analysis of variance was determined. The standard deviation of the slopes was predicted from the upper and lower borders comprising 95% of the data results.

Statistical analysis

The statistical significance of differences between groups was determined using unpaired Student’s t test, with differences considered significant at P < 0.05.


Tissue array analysis of SOD-1 serum antibodies

To determine a possible role of B cells in ALS-like disease in SOD-1 mice, we analyzed serum samples from SOD-1 mice for the presence of autoantibodies, using a mouse tissue array. To do so, serum samples were collected from SOD-1 mice before and after the onset of the diseases (60, 110 and 130 days old), and from aged-matched wt control mice. Serum samples were analyzed for autoreactive antibodies using tissue array membranes that were loaded with tissue homogenates collected from healthy mice. Most of the normal serum samples that were tested in this tissue array (in dilutions of 1:500 up to 1:1,500) provided a similar pattern of bands, probably corresponding to Ig-heavy and Ig-light chains found in these tissues as well as general cross-reactivity (Fig. 1 top blot). However, serum samples collected from 50% of the SOD-1 mice (total of eight) revealed additional bands in lysates prepared from brain and muscle (Fig. 1 middle and bottom blots). These bands were reproducibly observed for the individual SOD-1 mice at all time points that were analyzed. These data suggest that tissue reactive autoantibodies are produced in many SOD-1 mice.
Fig. 1

Tissue reactivity of antibodies in serum of SOD-1 mice. Serum samples from SOD-1 mice were tested in a tissue array for tissue reactivity as described in the methods section. In these arrays, the membranes were loaded with tissue homogenates that were prepared from healthy mice. Shown are representative blot arrays of one control wild-type mouse (out of five analyzed, top) and two SOD-1 mice (out of eight analyzed, middle and bottom). Serum samples were diluted 1:3,000 and adjusted to be approximately 1 μg/ml of total IgG. Tissue specific bands identified by SOD-1 sera are pointed with an arrow. Tissues order is: 1 kidney, 2 brain, 3 spinal cord, 4 lungs, 5 thymus, 6 skin, 7 muscle, 8 spleen, 9 heart

Analysis of splenic B cells and serum immunoglobulin

Next, we analyzed splenic B cells for the expression of B-lineage and surface activation markers and determined total IgM and IgG antibodies in serum. To do so spleen cells and serum samples were prepared from sick SOD-1 mice (130 days) and age-matched control. The spleen cells were stained for CD19, IgM, IgD, CD45 and CD69, and analyzed by flow cytometry (Fig. 2a). Serum samples from the mice were analyzed for the total amounts of IgM and IgG by quantitative ELISA (Fig. 2b). The results reveal no differences between SOD-1 and control B cells in surface marker expression, and no differences in the levels of serum IgM and IgG. Similar results were obtained for the analysis conducted using mice that are 60 or 110 days old (not shown). There were also no differences in the total number of spleen and LN B cells between the mice (not shown).
Fig. 2

Phenotypic analysis of splenic B cells and serum immunoglobulin. a Spleen cells from SOD-1 and from wt control mice were stained for the indicated surface markers and analyzed by flow cytometry. The plots are representatives of four individual mice in each group. The experiment was conducted for mice aged 60, 110 and 130 days. b Serum samples from SOD-1 and from wt control mice were analyzed for total IgM and IgG using standard ELISA. Quantitative measurements were calculated using a reference standard curve. Results are expressed as mean ± SD of five mice in each group

Responsiveness to immunogenic stimulation

To test the immune competence of SOD-1 B cells, we purified splenic B cells from sick SOD-1 (130 days old) and aged-matched control mice. The purified B cells were stimulated for 96 hours with LPS or with CpG DNA. Proliferation of the cells was measured by the MTT assay, and the amounts of secreted IgM and IgG were determined in supernatants by ELISA (Fig. 3). We found that immune responsiveness of SOD-1 B cells to LPS and to CpG DNA is not different than that of the control B cells as measured by proliferation (Fig. 3a), production of IgM (Fig. 3b) and production of IgG (Fig. 3c). Similarly, we did not find any differences in the responsiveness of B cell isolated from mice that are 60 and 110 days old (not shown).
Fig. 3

Responsiveness of splenic B cells to immune stimuli. Splenic B cells from SOD-1 and wt control mice were purified untouched by magnetic beads. Cells were cultured for 96 h unstimulated or stimulated with LPS or with CpG DNA. Proliferation of cells was measured by the MTT assay (a). Supernatants were collected and amounts of secreted IgM (b) and IgG (c) were determined by ELISA using standard curve for reference. Results are expressed as mean ± SD of five mice in each group

Development of ALS-like disease in SOD-1 mice deficient of B cells

Finally, we have directly assessed the role of B cells in the initiation of ALS-like disease in SOD-1 mice in vivo. To do so, we crossed SOD-1 mice with μMT mice that are deficient of B cells due to the insertion of a stop codon in the transmembrane tail exon of the μH chain [21]. After two generations we obtained μMT homozygous mice that carry the SOD-1 transgene on a C57/Bl6 genetic background (μMT/SOD-1), and are completely deficient of B cells (Fig. 4a). All mice were subjected to meteoric assays performed on a rota-rod. The mice were subjected to different velocities (5–22 rpm) and their ability to hold on the rota-rod was measured. Performance of the mice on the rota-rod was measured as the time the animal holds/age, or as the maximum speed (MAX)/age (Fig. 4). Rota-rod performance rate of wild type and μMT transgenic mice was found to be stable at maximal rotating duration and speed for the whole period of time tested. No muscle weakness or motor neuron dysfunction was visible (not shown). Starting from the age of 70 days, all mice over expressing the mutated SOD1 transgene with or without the μMT mutation gradually developed the disease symptoms: motor impairment, hind and fore-limb gradual paralysis and massive weight loss. Rota-rod performance was progressively affected to yield shorter rod duration and lower maximal velocities. Comparative results of the rota-rod performance of the SOD1 and the μMT SOD1 mice are presented in Fig. 4. The analysis revealed that μMT/SOD-1 mice have a greater reduction in the rota-rod performance time (4.4 ± 0.7 s/day relative to 3.5 ± 0.3 s/day in SOD-1 mice, P < 0.05) (Fig. 4b). Similarly, the μMT/SOD-1 mice had a greater reduction in the rota-rod max velocity (0.32 ± 0.05 RPM per day relative to 0.26 ± 0.02 RPM per day in SOD-1 mice, P < 0.05) (Fig. 4c). There were no differences in mortality rates between SOD-1 and μMT/SOD-1 mice (Fig. 4d). Hence, the lack of B cells did not prevent or inhibit the development of ALS-like disease in SOD-1 mice.
Fig. 4

Comparison of the neurodegenerative process development in SOD1G93A and B cell deficient SOD1G93A mice. a Flow cytometry analysis of blood sample mononuclear cells collected from the indicated mice and stained for CD5 and CD19. b The rota-rod performance time of SOD-1 (N = 10) and μMT/SOD-1 mice (N = 10) versus age was measured from day 70 until death. A significant linear correlation was established between mice age and performance time. c The rota-rod max velocity of SOD-1 and μMT/SOD-1 mice versus age was measured from day 70 until death for the same mice. A significant linear correlation was established between mice age and max velocity. d Survival curve for the indicated groups of mice


The rational for this study stems from several publications proposing that ALS disease is associated with dysregulation in the immune system. In here, we specifically addressed the contribution of B lymphocytes to the pathogenesis of ALS-like disease in SOD-1 mice. The conclusions drawn by this study exclude an essential role of B cells in this disease. However, it is important to mention that the mutated form of SOD-1 gene (G93A) in the mouse model is linked with only 20% of ALS cases and that over 140 mutations are now identified [6].

While dysfunction of innate immunity has been shown as a pathogenic feature of ALS [6, 19, 22], there are very few studies proposing a pathogenic role of the adaptive immunity [5, 11, 29, 31, 39]. A functional link between the adaptive immune system and neurodegenerative diseases has been shown for Alzheimer’s disease [40], multiple sclerosis (MS) [26] and Parkinson’s disease [24], proposing the existence of a destructive autoimmune process. Recently, the central role of the B lineage in autoimmunity becomes more appreciated as B cell targeting therapies, such as rituximab [10], yield very promising results even in autoimmune diseases that are traditionally considered as T cell regulated such as MS [17]. The finding that SOD-1 mice that are deficient of B cells develop ALS-like diseases (Fig. 4) appears to exclude a central role of B cells in this disease. However, similar studies using B cell deficient mice proposed that B cells do not play a vital role in experimental autoimmune encephalomyelitis (EAE) [8], a conclusion that is now reversed in light of the successful use of B cell depletion therapy in MS patients [17]. In addition, the SOD-1 model may only represent 20% of familial ALS cases, providing a rational for more clinical studies using other ALS animal models and patients with a different form of ALS before excluding the role of B cells in the disease.

Despite the fact that autoantibodies are detected in ALS patients and in SOD-1 mice (here and in [29, 31]), it is also possible that B lymphocytes may have a smaller contribution to the pathogenesis of ALS than initially proposed. In autoimmune diseases such as lupus, peripheral B cells are hyperreactive [32] and express higher levels of activation markers [12], thereby contributing to the pathogenesis of the diseases. We show here that SOD-1 B cells are not expressing activated phenotype and are not different than wt B cells in responsiveness to LPS and to CpG DNA. Thus, ALS-like disease in the SOD-1 mouse model does not involve polyclonal activation of B lymphocytes or an inherent defect in B cell responsiveness. However, the fact that autoantibodies are produced argues that few, self-reactive B cell clones are activated by specific neuronal tissue. The contribution of these autoantibodies to the pathogenesis of ALS is still questionable although a correlation with the severity of disease course has been shown [28]. In many organ-specific autoimmune diseases autoantibodies can be detected a long time before the clinical onset and during the course of the diseases. In these cases, autoantibodies can be used as predictive markers of an ongoing disease (in healthy individuals) and of disease activity and severity (in patients). So far no biological marker has been described for ALS and extensive research is conducted to identify such a marker for early diagnosis. Several autoantibodies have been identified in cerebrospinal fluid (CSF) and serum of ALS patients including anti-GM-1 ganglioside [29], neurofilament 68 [28]. However, because not all ALS patients develop autoantibodies and that no pathogenic role has been described for these autoantibodies, it is conceivable that autoantibody production is a secondary immunological consequence of neuronal death. The fact that ALS-like disease develops in B cell-deficient SOD-1 mice strongly supports this hypothesis. Nevertheless, it is possible that the autoantibodies may accelerate the course of neuronal degeneration, as a correlation with the severity of disease course has been shown [28].


This work was supported by grants provided by IsrA.L.S.—The Association for ALS Research in Israel, and by the Elias Fund for Medical Research.

Copyright information

© Springer-Verlag 2009