The association between EAE development in mice and the production of autoantibodies and abzymes after immunization of mice with different antigens

Abstract We have previously shown that immunization of C57BL/6 mice, prone to spontaneous development of experimental autoimmune encephalomyelitis (EAE), with three antigens (MOG35‐55, DNA‐histone complex or DNA‐methylated BSA complex), alters the differentiation profiles of bone marrow haematopoietic stem cells (HSCs). These are associated with the production of autoantibodies (auto‐Abs) against these antigens and the formation of abzymes hydrolysing DNA, MOG, myelin basic protein (MBP) and histones. Immunization of mice with antigens accelerates the development of EAE. This work is the first to analyse the ratio of auto‐Abs without and with catalytic activities at different stages of EAE development (onset, acute and remission phases) after immunization of mice with the three specific antigens. Prior to immunization and during spontaneous in‐time development of EAE, the concentration of auto‐Abs against MBP, MOG, histones and DNA and activities of IgG antibodies in the hydrolysis of substrates increased in parallel; correlation coefficients = +0.69‐0.94. After immunization with MOG, DNA‐histone complex or DNA‐met‐BSA complex, both positive (from +0.13 to +0.98) and negative correlations (from −0.09 to −0.69) were found between these values. Our study is the first showing that depending on the antigen, the relative amount of harmful auto‐Abs without and abzymes with low or high catalytic activities may be produced only at onset and in acute or remission phases of EAE. The antigen governs the EAE development rate, whereby the ratio of auto‐Abs without catalytic activity and with enzymatic activities of harmful abzymes hydrolysing MBP, MOG, histones and DNA varies strongly between different disease phases.


| INTRODUC TI ON
Antibodies (Abs) against chemically stable analogs of transition states of chemical reactions and natural antibodies with catalytic activities (abzymes) are well-described in the literature (for review, see refs. [1][2][3][4][5][6]. IgG and/or IgA and IgM (immunoglobulin G, A and M) antibodies that hydrolyse DNA, RNA, adenosine triphosphate (ATP), polysaccharides, peptides and proteins were found in the blood of patients for various autoimmune pathologies (AIPs), including systemic lupus erythematosus (SLE), multiple sclerosis (MS), Hashimoto's thyroiditis, polyarthritis, lymphoproliferative diseases, polyneuritis and malignant tumours, as well as for three viral diseases-viral hepatitis, tick-borne encephalitis and human immunodeficiency (Refs. 1-6 and references therein). In contrast, antibodies from healthy donors or from patients with diseases without significant impairment of the immune status do not show significant catalytic activities. [1][2][3][4][5][6] Autoantibodies (auto-Abs) to various antigens are formed both in healthy donors 7,8 and in patients with AIPs. [1][2][3][4][5][6] In patients with multiple sclerosis (MS), auto-Abs against DNA were found in elevated concentrations in only 17%-18% 9 and abzymes with DNase activity in 90%-95% of patients, 10 compared with healthy donors. For patients with AIPs, the average values of antibody levels in comparison with healthy donors usually increase only in the late or worsening stages of the disease. [1][2][3][4][5][6] Different abzymes with various catalytic activities in healthy donors are usually either absent or show extremely low activities. [1][2][3][4][5][6][11][12][13][14][15] In contrast to enzyme-linked immunosorbent assay (ELISA), the catalysis of different reactions by abzymes is based on a large number of catalyst turns, even if they possess low activity. Therefore, different abzymes can be found even at the onset of AIPs. [1][2][3][4][5][6] The statistically significant appearance of abzymes can be detected in the earliest stages of various pathologies when changes in antibody levels to specific antigens (eg DNA, MBP, thyroglobulin) still correspond to the range of changes seen in healthy donors. [1][2][3][4][5][6] This was confirmed by the ratio of relative auto-antibody concentrations to DNA and abzymes hydrolysing DNA during spontaneous SLE development in autoimmune MRL-lpr/lpr mice [16][17][18] and abzymes hydrolysing proteins during EAE development in C57BL/6 mice. [19][20][21][22] Thus, the appearance or strong increase (dozen-fold or more) of abzymes' relative activities (RAs) in relation to levels seen for healthy individuals can be one of the earliest signs of AIP development. [16][17][18][19][20][21][22] It has been previously suggested that AIP development may be because of defects in haematopoietic stem cells. 23 To investigate this issue, we have in previous studies used two different models of autoimmune prone mice: MRL-lpr-lpr (SLE model) [16][17][18] and C57BL/6 (human MS model). 19,20 We analysed the differentiation profile of bone marrow haematopoietic stem cells (HSCs) during spontaneous and antigen-stimulated development of SLE (DNA immunization) and EAE (immunization with MOG). We found that, compared with the norm, the spontaneous pre-disease phase in SLE and EAE mice led to first changes, and transition to deep pathology led to additional changes in the differentiation profile of HSCs (ie BFU-E, CFU-E, CFU-GM and CFU-GEMM). During initial changes in HSC differentiation profile, abzymes hydrolysing DNA and MBP in the blood of mice showed only relatively low activity, whereas their activity increased greatly during later disease stages. [16][17][18][19][20][21][22] Immunization of SLE mice with DNA accelerates disease development and being associated with an increase in antibody levels against DNA, as well as in the increase in RAs of abzymes hydrolysing DNA, ATP and oligosaccharides. [16][17][18] After treating EAE mice with MOG, DNAhistone complex or DNA-met-BSA complex, the first strong increase in antibody levels against MOG, MBP and DNA and in abzymes that hydrolyse these substrates was already observed after seven days; the maximum increase was observed 18-20 days after immunization. [19][20][21][22] Notably, changes in HSC differentiation profiles were similar for SLE and EAE mice during spontaneous and antigen-induced development of these pathologies. 19 In parallel with the differentiation profile change for HSCs, the relative number of lymphocytes producing auto-Abs and abzymes increased. [19][20][21][22] Interestingly, the RA of abzymes derived from the cerebrospinal fluid of MS patients, hydrolysing MBP, DNA and oligosaccharides depending on their substrate, is about 40-60 times higher than Abs found in the blood of the same patients. [24][25][26] Analysing the development of SLE and EAE showed that the appearance of abzymes is the earliest and also a statistically significant marker for the development of these AIPs. A further significant increase in abzyme activity correlates with additional changes in the differentiation profile of bone marrow stem cells, which associated with increases in urine protein concentration, anti-DNA and anti-protein Abs in the blood and leads to the appearance of a typical pink butterfly in SLE mice and alopecia in EAE mice. [16][17][18][19][20][21][22] Accelerating EAE development in C57BL/6 mice after MOG im- In addition to intranuclear functions, histones can be harmful when they enter the extracellular space. 27 The administration of exogenous histones to animals leads to systemic inflammatory and toxic reactions. Animals treated with various anti-histone preparations are protected from lethal endotoxaemia, sepsis, ischaemia, reperfusion injury, injury, pancreatitis, peritonitis, stroke, coagulation and thrombosis. In addition, elevated concentrations of histones and nucleosomes in the blood affect several pathophysiological processes and the progression of diseases such as AIPs, inflammation and cancer. 27  At the same time, auto-Abs of HIV-infected patients 28 and EAE mice 22 can effectively hydrolyse all five histones. Until now, there are no data on abzymes cross-hydrolysing any proteins against other proteins. [1][2][3][4][5][6]32 Recently, it has been shown that in the blood of HIVinfected patients, Abs against MBP effectively hydrolyse not only MBP but also histones and vice versa. 32 It is believed that virus infection in humans leads to an accumulation of Abs against components of certain viruses that may have structural similarity to components of human blood and cells. [30][31][32] Then, because of the mimicry of certain viral and human proteins, the immune system may fail, and Abs against components of the human body may be accumulated, resulting in the development of AIPs. According to our data, the beginning of AIP development may be associated not only with the mimicry of antigens but also with the homology between certain human proteins. For example, during cell apoptosis, nuclear histones with sequences homologous to MBP and other human proteins enter the blood. It is possible that the existence of cross-catalytic activity of abzymes against histones and MBP is one of the internal causes of MS induction and other AIPs. 32 Many SLE anti-DNA Abs are directed against nucleosomal DNAhistone complexes entering circulation after internucleosomal cleavage during apoptosis. 33 It has been shown that polyclonal abzymes hydrolysing DNA of SLE and MS patients are cytotoxic, enter the nucleus and cause DNA fragmentation, inducing cell death by apoptosis. 34,35 In MS and SLE, anti-MBP abzymes with protease activity can attack MBP of the myelin-proteolipid sheath of axons. 28,29 Thus, these abzymes can play an important negative role in the pathogenesis of SLE and MS pathologies. However, auto-Abs and abzymes hydrolysing other histones could also play a negative role in the pathogenesis of MS. 5 Taking current knowledge into account, we wanted to understand which auto-Abs without catalytic activity and which of the various abzymes demonstrating different levels of catalytic activity are produced at different EAE stages before and after immunization of mice. In this study, we, therefore, compared the levels of anti-MOG, anti-MBP, anti-DNA and anti-histone antibodies in different phases of EAE development (onset, acute and remission phases) before and after immunization of mice with MOG, DNA-histone complex or DNA-met-BSA complex. In parallel, we determined the catalytic activity of abzymes hydrolysing DNA, MOG, MBP and histones and estimated correlations between levels of Abs and their ability to hydrolyse these substrates.

| Reagents and animals
If not stated otherwise, all chemicals, proteins and five histones

| Immunization of mice and analysis of different parameters of EAE development
Immunization of C57BL/6 mice with MOG, 19,20 DNA-met-BSA complex 21 and DNA-histone complex 22 was described previously (for details, see Supplementary methods). Analysis of various parameters during different phases of EAE development, including differentiation profiles of bone marrow HSCs and lymphocyte proliferation in different organs, was described previously [19][20][21][22] (for details, see Supplementary methods). In previous work [19][20][21][22] and in this work, parameter changes before and after immunization were analysed in groups of seven mice each.

| ELISA of anti-DNA, anti-MOG and antiprotein antibodies
In this study, we analysed previously collected samples concerning levels of Abs to different antigens of IgG preparations [19][20][21][22] at different time-points of spontaneous and antigen-induced development of EAE, which were not used before, and obtained additional data from frozen (−70 o C) blood plasma preparations for a more detailed and advanced analysis. [19][20][21][22] The levels of anti-DNA Abs (plasmas were diluted 100-fold) were determined using standard ELISA: plates with immobilized double-stranded DNA, and horseradish peroxidase-conjugated mouse Abs against human IgG of the test system ORGENTEC Diagnostika (Germany) were used as in Ref. 22 according to the manufacturer's instructions .
The relative levels of anti-MBP and anti-MOG IgGs were estimated according to Ref. [19][20][21][22] For analysis of anti-MOG, anti-MBP Abs, plasmas were diluted 50-fold using sodium carbonate buffer, pH 9.6. Diluted plasmas (50 μl) were added to ELISA strips, which were incubated overnight at 23°C. The strips were washed with TBS buffer (20 mmol/L Tris-HCl containing 0.15 mol/L NaCl) supplemented with 0.01% NaN 3 and 0.05% Triton X-100 and three times with the same buffer containing no Triton X-100. To block the strip surfaces, they were treated for 2.5 hours at 30°C using TBS containing 0.2% egg albumin and 0.01% NaN 3 . The strips were washed 8 times with water and then with TBS containing 0.01% NaN 3 . The strips were incubated with 100 μL of TBS containing 1 μg/mL conjugate of monoclonal anti-human IgGs with horseradish peroxidase for 40 minutes at 30°C rewashed 10 times with water. After adding 60 μL citric-phosphate buffer containing 3,3',5,5'-tetramethylbenzidine, and H 2 O 2 , the strips were incubated for 14 minutes at 23°C, and the reaction was stopped by the addition of 6 μL of 50% H 2 SO 4 .
A 450 values were determined using a UNISKAN II Plate Reader

| IgG purification
Electrophoretically homogeneous mouse IgG was purified by two consecutive chromatographies of plasma proteins on Protein G-Sepharose, followed by FPLC gel filtration on Superdex 200 HR 10/30 column in dissociating conditions (pH 2.6), as described previously. [19][20][21][22] Abs were filtered through filter units (Millex; 0.1 μm) and stored in sterilized tubes to protect them from bacterial and viral contamination. [19][20][21][22] Electrophoretic homogeneity of IgG was analysed by SDS-PAGE 4%-15% gradient gels using non-reducing conditions in the absence of DTT (0.1% SDS). IgG was revealed with silver staining gels, as described previously. [19][20][21][22] To exclude possible artefacts because of DNase and protease contaminations, IgG was separated by SDS-PAGE, and its activities were revealed using a gel assay. [19][20][21][22] After electrophoresis, peaks for all activities were revealed only in those bands corresponding to intact IgG, and no other peaks were found for any canonical enzyme activities. More detailed data are given in previous work [19][20][21][22] and Supplementary methods.

| DNA-hydrolysing activity assay
DNase activity of IgG was analysed as described in previous work. [19][20][21][22] The reaction mixture (20 μL) contained 20 mmol/L Tris-HCl (pH 7.5), 20 mmol/L NaCl, 5 mmol/L MgCl 2 , 1 mmol/L ethylenediaminetetraacetic (EDTA), supercoiled (sc) pBluescript DNA (20 μg/mL) and 0.03-0.2 mg/mL of IgG. The mixture was incubated for 1-12 hours at 37°C, and then, DNA hydrolysis was analysed using electrophoresis in 0.8% agarose gel. Ethidium bromide-stained gel was analysed with Gel-Pro Analyzer v9.11. [19][20][21][22] The relative DNase activity was appraised from the percentage of the initial band of intact scDNA and its relaxed form, considering the relative content of DNA between these two bands for scDNA incubation in the absence of Abs. The products were analysed by looking at linear parts of rate dependencies of substrate hydrolysis on IgG concentrations and the time course (15%-40% of DNA hydrolysis). The complete transformation of scDNA to its nicked form was taken as 100% of the activity. The RA (% of the hydrolysis) was finally recalculated to the same standard conditions.

| Protease activity assay
The reaction mixtures (10-30 μL) for analysing protease activities of IgG contained 20 mmol/L Tris-HCl, pH 7.5; 0.7-1.0 mg/mL of either MOG, MBP or equimolar mixture of five histones, and 0.01-0.2 mg/mL of IgG, as described in detail in previous work. [19][20][21][22] They were incubated for 1-21 hours at 37°C. The cleavage products of the proteins were analysed by SDS-PAGE using 4%-15% gradient or 12% gels. Gels were stained with Coomassie R250, scanned and then quantified using Gel-Pro v3.1 software. The RAs of different IgGs were calculated from a decrease in the percentage of original proteins turned to their various hydrolysed forms. The hydrolysis of all proteins incubated without Abs was taken into account. Other measurements similar to the analysis of DNA hydrolysis were taken using pseudo-first-order reaction conditions.

| Statistical analysis
The analysed parameters are given as the mean ± SD over three to four independent experiments, and data for each group are averaged over seven different mice. Correlation coefficients (CCs) between different parameters were estimated using the Pearson parametric method of Microsoft Excel 2010.

| RE SULTS
In several earlier studies, we analysed changes in the differentiation profile of bone marrow HSCs and the level of lymphocyte proliferation after immunization of C57BL/6 mice with MOG, DNA-histone complex or DNA-met-BSA complex. [19][20][21][22] For illustration, previously obtained data are given in the Supplementary results. We evaluated changes in weight and proteinuria levels during spontaneous or antigen-stimulated EAE development ( Figure S1). [19][20][21][22] One of the important indicators for AIP development is an increase in mouse urine protein concentration. 36 Until day 63, after immunization, the proteinuria of mice treated with MOG increased 3.8-fold, whereas for mice treated with a complex of DNA-histone or DNA-met-BSA, it decreased ~1.3-fold ( Figure S1). A decrease in proteinuria may slow down EAE development in mice after treatment with antigens containing DNA. [19][20][21][22] Figure S2). [19][20][21][22] It was previously shown that the production of auto-Abs and abzymes is associated not only with a change in differentiation profiles of HSCs but also with an increase in the level of lymphocyte proliferation (sum of T and B cells). [19][20][21][22] During spontaneous EAE development, we observed a constant increase in the relative content of lymphocytes in mouse bone marrow, spleen and thymus over time; a maximal increase was observed in bone marrow.
Only for lymph nodes, we observed a remarkable decrease in the level of lymphocyte proliferation ( Figure S3). Now, we have, for In this study, we used the following four groups of C57BL/6 mice, which differed in some of the parameters characterizing their EAE development, as described previously [19][20][21][22]  Some data on levels of Abs and their relative catalytic activity were taken from previous work. [19][20][21][22] On top, we obtained some additional data by analysing levels and relative activities of IgG using supplemental blood plasma preparations and IgG isolated from these preparations, according to procedures described previously. [19][20][21][22] New additional IgG preparations were purified from sera of individual mice by chromatography of unfrozen plasma proteins on Protein G-Sepharose, followed by FPLC gel filtration. [19][20][21][22] All IgG preparations were electrophoretically homogeneous and, according to the analysis of the catalytic activities of gel fragment extracts after IgG SDS-PAGE, [19][20][21][22] did not contain canonical DNase and protease enzymes.

| Analysis of different parameters characterizing spontaneous EAE
As shown in previous work, autoimmune C57BL/6 mice reveal slow and spontaneous changes in different parameters, which are also relevant during EAE pathology, with onset as early as three months after birth. [19][20][21][22] Therefore, it is possible to reliably test hydrolysing properties of IgG derived from the blood of three-month-old mice for MOG, MBP, DNA and histones. [19][20][21][22] During the 63-day course of spontaneous EAE development, It is on these proteins or their fragments that the production of auto-Abs with and without catalytic activity can occur.  IgG-dependent hydrolysis of MOG and MBP (+0.63) was positive and high. Interestingly, after immunization of mice with DNA-met-BSA complex, similar to treatment with MOG and DNA-histone complex, there was a distinct interconnection between production of Abs against MOG and MBP and abzymes hydrolysing these substrates.

| Analysis of anti-histone Abs after immunization of mice with different antigens
The main antigen for which Abs against DNA is accumulated is a complex of DNA with histones. 33 Histones can also play an important role in the development of AIPs. 27 ELISA analysis of the relative concentration of Abs against the five histones (H1, H2a, H2b, H3 and H4) was carried out using their equimolar mixture ( Figure 5A).
During spontaneous EAE development, the relative concentration of Abs against five histones increased only 1.5-fold by day 63.
After immunization of mice with MOG, the concentration of Abs first slightly increased and then decreased. The treatment of mice with DNA-met-BSA complex stimulated a constant decrease in the F I G U R E 2 Time-dependent changes in average values of anti-DNA Ab concentration and relative activity of IgG hydrolysing DNA after immunization of mice with MOG (A), DNA-histone complex (B) and DNA-met-BSA complex (C). For comparison, data on parameter changes during spontaneous development of EAE are given concentration. The level of Abs only rose strongly (3.7-fold) when mice were immunized with DNA-histone complex ( Figure 5A).
At all stages of EAE development, the efficiency of hydrolysis of each of the five histones was evaluated separately. 22,31,32 After treating mice with MOG, we observed constant growth of protease activity only for hydrolysis of H1 histones, which increased 2.5-fold by day 63 compared with time-point zero ( Figure 5B). MOG stimulated decrease in efficiency of IgG-dependent hydrolysis of H2a, H2b, H3 and H4 in the period between 7 and 20 days, after which the activity of abzymes hydrolysing these four histones increased ( Figure 5B).
The DNA-histone complex induced the production of abzymes hydrolysing H1 and H3 histones on day 7, and for H2a, H2b and H4 at about 20 days after immunization. Thereafter, the activity of abzymes against all histones decreased significantly.
After immunization of mice with DNA-met-BSA complex, we observed very similar patterns ( Figure 5D). The level of IgG-dependent hydrolysis of all five histones was greatly reduced at 7-14 days, and then sharply increased at about 20 days, with a strong decrease in the activities between 30 and 63 days. Overall, the synthesis of abzymes hydrolysing histones showed the tendency to decrease in the period following the initial 20 days after immunization, corresponding to the period of EAE remission.
It should be noted that the raise in Ab levels against the immunogens used and the hydrolysis activity of the abzymes differ strongly; there are both positive (from + 0.13 to + 0.91) and negative correlations (from − 0.09 to − 0.98).

| D ISCUSS I ON
We have previously demonstrated that changes in the differentiation profile of bone marrow HSCs and in lymphocyte proliferation in different organs lead to EAE development in C57BL/6 mice, associated with the production of auto-Abs to MOG, MBP, DNA and histones and abzymes-harmful for mice-hydrolysing these substrates. 19 All Ab fractions eluted from the sorbent differed in their relative catalytic activities and other properties. Interestingly, depending on the patient, as well as the specific disease and its duration, the repertoire of Abs hydrolysing DNA, RNA, proteins, oligosaccharides and other antigens may be relatively small or very broad. [1][2][3][4][5][6]38 In the later stages of AIPs, the total pool of immunoglobulins they increased between 20 and 63 days, which was associated with the production of IgG with low or without activity ( Figure 3A and Figure 4A). At the same time, mice immunized with the DNAhistone complex in the initial and acute phases (days 6-20) demonstrated a high level of anti-DNA Abs, whereas the activity of the abzymes was relatively low ( Figure 2B). The subsequent ~2.5-fold decrease in the anti-DNA concentration between 20 and 63 days after immunization with the DNA-histone complex is accompanied by a 200-fold increase in abzyme activity ( Figure 2B). More than 20 days after treating mice with the DNA-met-BSA complex, an increase in both levels of anti-DNA Abs and DNase activity of abzymes occurred ( Figure 2C). However, abzyme RA increase (290-fold from day 1 to 63) is 52 times higher than the increase in F I G U R E 5 Time-dependent changes in average values of anti-histone Ab concentration (A) and relative activity of IgG hydrolysing five different histones (H1, H2a, H2b, H3 and H4) after immunization of mice with MOG (B), DNA-histone complex (C) and DNA-met-BSA complex (D). The error in determining the value corresponding to each timepoint does not exceed 10%-12% anti-DNA Ab levels (5.6-fold). Thus, in both cases, immunization of mice with antigens containing DNA led to a strong delay in the production of specific anti-DNA abzymes with high catalytic activity (maximum activity at 63 days) compared with the treatment of mice with MOG (maximum activity at 20 days); there is a 6.5fold difference in DNase activity ( Figure 2) All three immunogens accelerate EAE development in mice.
However, at 63 days, IgG-dependent efficiency of DNA hydrolysis is about 6.5-8.0 times higher after treatment with antigens containing DNA than the maximal DNase activity of IgG after immunization of mice with MOG.

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest.