Changes in cell differentiation and proliferation lead to production of abzymes in EAE mice treated with DNA–Histone complexes

Abstract Experimental autoimmune encephalomyelitis (EAE)‐prone C57BL/6 mice are used as a model of human multiple sclerosis. We immunize mice with myelin oligodendrocyte glycoprotein (MOG), DNA–histone and DNA‐methylated bovine serum albumin (met‐BSA) complexes to reveal different characteristics of EAE development including bone marrow lymphocyte proliferation and differentiation profiles of hematopoietic stem cells. Immunization of C57BL/6 mice with MOG35‐55 results in the acceleration of EAE development. Anti‐DNA antibodies are usually directed against DNA–histone complexes resulting from cell apoptosis. During the acute EAE phase (7‐20 days after immunization), catalytic antibodies efficiently hydrolysing myelin basic protein (MBP), MOG and DNA are produced with parallel suppression of antibodies hydrolysing histones. We could show that in contrast to MOG, immunization with histone‐DNA results in a reduction of proteinuria, a significant increase in anti‐DNA, anti‐MBP and anti‐MOG antibody titres, as well as an increase in their catalytic activities for antigen hydrolysis, but slightly changes the concentration of cytokines. Contrary to MOG, DNA–histone and DNA‐met‐BSA only stimulated the formation of anti‐DNA antibodies hydrolysing DNA with a long delay (15‐20 days after immunization). Our data indicate that for C57BL/6 mice immunization with DNA‐met‐BSA and DNA–histone complexes may have opposing effects compared to MOG. DNA–histone stimulates the appearance of histone‐hydrolysing abzymes in the acute EAE phase, while abzymes with DNase activity appear at significantly later time‐points. We conclude that MOG, DNA–histone and DNA‐met‐BSA have different effects on numerous bone marrow, cellular, immunological and biochemical parameters of immunized mice, but all antigens finally significantly stimulate the development of the EAE.


| INTRODUCTION
Multiple sclerosis (MS) is the inflammatory and demyelinating disease of the central nervous system (CNS) with perivascular infiltrates composed largely of macrophages and T lymphocytes. Numerous studies support the role of autoimmune mechanisms in the destruction of myelin, while the precise cause of MS remains unknown. 1 Data indicate that activated myelin-reactive CD4 + T cells may be principal mediators of MS. 1 Several recent findings also specify an important role of B cells and autoantibodies (auto-Abs) against myelin autoantigens in MS pathogenesis. [1][2][3] The appearance of oligoclonal immunoglobulin G (IgG) and increased levels of antibodies (Abs) in the cerebrospinal fluid (CSF), the accumulation of clonal B cells in the CSF, and lesions of MS patients provide key evidence for the involvement of a humoural response in demyelination. 4 Evidence from recent clinical studies in humans and animal models demonstrates that auto-Abs against myelin components (involved in antibody-mediated demyelination 3 ) and auto-Abs against oligodendrocyte progenitor cell protein, which may interfere with remyelination by eliminating or impeding these cells, 5 seem to play a crucial role in MS immunopathogenesis. In addition to MS-specific mechanisms, there seem to be some common pathways in the development of different autoimmune diseases (AIDs). For example, the level of lymphocyte apoptosis is usually increased for all AIDs.
In addition to carrying out various intranuclear cell functions, histones act as damage-molecules in the extracellular space. 6 Treating animals with exogenous histones leads to systemic inflammation and toxic reactions through activation of Toll-like receptors and inflammasome pathways. The blood-level increase in nucleosome fragments and histones is associated with multiple adverse pathophysiological processes including progression in various AIDs and inflammatory diseases. 6 It has been shown that multifunctional cytokines can influence the human immune system leading to its dysregulation and resulting in the development of different AIDs. 7 In addition, an important and harmful role in the development of MS 8 and other AIDs can be attributed to auto-Abs with catalytic activities (abzymes; ABZs). [9][10][11][12][13] Artificial abzymes against chemically stable analogues of transition states of different chemical reactions are well described. 9,[14][15][16] In the blood of healthy human beings and animals, it is possible to detect auto-Abs against different proteins, peptides, DNA, RNA, oligosaccharides and various other components; their titres vary significantly, but these auto-Abs are usually catalytically inactive. [8][9][10][11][12][13] Researchers conducted during the last 30 years have revealed different enzymatic activities for auto-Abs. Auto-Ab elaboration is a specific feature of AID patients. [8][9][10][11][12][13] Similar to artificial abzymes, natural ABZs are antibodies against enzyme substrates acting as protein haptens and mimicking transition states of chemical reactions. Antiidiotypic ABZs against catalytic enzyme centres also possess catalytic activities. [8][9][10][11][12][13] Abzymes with many different enzymatic activities were revealed as the earliest and most significant markers of onset and development of various AIDs in humans and mammals [8][9][10][11][12][13][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32] including systemic lupus erythematosus (SLE), [27][28][29][30][31][32]21 in experimental mice. Enzymatic activities of ABZs are actually detectable at the pre-disease stage prior to appearance of obvious AID markers and changes in proteinuria. At that time, antibody titres for various anti-antigens correspond to a range of indices typical for healthy humans and mice. A detectable level of abzymes can indicate the disease onset or pre-disease stage, while an increase in catalytic activities follows the development of clear pathological symptoms. [11][12][13][17][18][19][20][21] To understand the disease, it is therefore essential to identify all possible parallel and complementary mechanisms of MS development.
Current data demonstrate the appearance of myelin basic protein (MBP)-and DNA-hydrolysing abzymes in the blood of patients with MS, [22][23][24][25][26] and several other AIDs. [10][11][12][13] IgGs from the CSF of MS patients, which hydrolyse MBP, DNA and polysaccharides, are on average 30-to 60-fold higher than those taken from the sera of the same patients. [30][31][32] In AIDs, abzymes hydrolysing MBP can attack the protein of the myelin-proteolipid sheath of axons and can play a harmful role in MS, SLE and other disease pathogenesis. [10][11][12][13] Abzymes with DNase activity are harmful: they are cytotoxic and can penetrate into the nucleus of the cell; they hydrolyse nuclear DNA inducing cell death by apoptosis, which in turn stimulates the development of AIDs. [33][34][35] Anti-DNA Abs and ABZs in SLE, MS and other AIDs are usually directed against nucleosomal histone-DNA complexes, which appear during apoptosis and result in internucleosomal cleavage. 36 All autoimmune diseases have initially been suggested to originate from defects in hematopoietic stem cells (HSCs). 37 Later, it was shown that the spontaneous and antigen-induced development of profound experimental autoimmune encephalomyelitis (EAE) 20,21 and SLE [17][18][19] pathologies in autoimmune (AI) prone mice was associated with a specific reorganization of the immune system, including specific changes in the profile of differentiation of bone marrow HSCs in combination with the production of abzymes hydrolysing DNA, ATP, polysaccharides and proteins.
There are several different EAE models, each mimicking a specific facet of MS (for review see 38 ). C57BL/6 mice were recently used to study possible mechanisms of spontaneous, myelin oligodendrocyte glycoprotein (MOG)-and DNA-accelerated development of EAE. 20,21 We confirmed that immunization with MOG and DNA led to an acceleration of EAE development but was associated with various changes in HSC differentiation profiles, lymphocyte proliferation and apoptosis in different organs of mice. The data showed that DNA and MOG may have to some extent antagonistic effects on these parameters and revealed a noticeable delay of EAE development after immunization with DNA. At the same time, as mentioned above, the main immunogen stimulating the development of various AIDs is not DNA but its complex with histones. 6 Moreover, it is believed that an important role in AID development is played by cytokines. 7 Similar to SLE pathology, high-affinity anti-DNA Abs have recently been identified as a major component of intrathecal IgGs in the brain and CSF cells of MS patients. 39 AULOVA ET AL.

| 5817
Taking all above considerations into account, here we carried out the first extended analysis of the DNA-histone complex effect on the changes of differentiation profiles, lymphocyte proliferation and apoptosis of bone marrow HSCs, relative titres of MBP, DNA, histones and relative activities of abzymes hydrolysing these substrates, as well as on the concentration of cytokines.

| Reagents
If not stated otherwise reagents, including chemicals, proteins, five histones (equimolar mixture of H4, H3, H2a, H2b and H1), polymeric bovine DNA, the Superdex 200 HR 10/30 column and Protein G-Sepharose were purchased from Sigma-Aldrich (Munich, Germany) or GE Healthcare. We used purified human MBP containing 18.5 kD from RCMDT (Moscow, Russia) and MOG  from EZBiolab. These preparations were free from lipids, oligosaccharides, nucleic acids and other possible contaminants. The ELISA test system for the analysis of cytokines was obtained from Biolegend (USA).
To prepare a conjugate, a solution of 23.6 mg five histones (in 11.8 mL of water) was mixed with 23 mg calf thymus DNA (in 3 mL of water), adding 80 μL of 3 mol/L NaOH (pH 10). After complete dissolution, the mixture was titrated with 1 mol/L hydrochloric acid to pH 7.0 and diluted to 18.8 mL with a physiological solution containing 0.235 mol/L NaCl. As final step a mixture of 18.8 mL of antigen solution, 101.5 μg Pertussis toxin (Mycobacterium tuberculosis) in 20 μL of water, and 18.8 mL complete Freund's adjuvant solution was composed. This mixture was repeatedly passed through a syringe needle to form a homogeneous gel. One hundred microlitres of the gel was injected subcutaneously per mouse; 50 μL into each paw pad. The second immunization of each mouse was performed (after 2 days) in the same manner using a 150-μL mixture of incomplete Freund's adjuvant containing 0.4 μg of Pertussis toxin. Immunization of mice with MOG and Pertussis toxin, and immunization with DNA without toxin was carried out as described in previously published protocol. 37 The relative weight of mice and proteinuria (relative concentration of total protein in the urine, mg/mL) was analysed as before. [17][18][19][20][21] Protein concentration in urine was measured using the Bradford assay with a bovine serum albumin standard. For further experiments including the purification of Abs and analysis of their enzymatic activity, 0.5-0.8 mL of blood was collected after decapitation using standard approaches. 20,21

| ELISA of anti-DNA and anti-protein antibodies
Anti-MBP and anti-MOG (for both plasma was diluted 50-fold), and anti-DNA (plasma was diluted 100-fold) antibody concentrations were analysed using ELISA in accordance with previous work. 20,21 After a consecutive treatment of blood plasma samples with rabbitspecific anti-mouse Abs conjugated with horseradish peroxidase, the reaction mixtures were incubated with H 2 O 2 and tetramethyl benzidine. The reaction was stopped with H 2 SO 4 , and the optical density (A 450 ) of the solutions was determined using a Uniskan II plate reader (MTX Lab Systems, USA).
The relative level of auto-Abs against histones (a mixture of all five histones) was estimated with ELISA test system from EUROIM-MUN (Germany) using purified homogeneous IgGs (0.02 mg/mL for one analysis) and adhering to the manufacturer's instructions, as described previously. 40 The relative concentrations of cytokines (IFNγ, IL-1β and TNFα) were estimated with ELISA test system from Vector-best (Novosi-

| Statistical analysis
The results are reported as the mean ± SD of at least three to four independent experiments for each mouse, averaged over seven different mice. Differences between the samples and mouse groups were analysed using Student's t test; P ≤ 0.05 was considered to be statistically significant.  initiates the production of abzymes efficiently hydrolysing not only MBP and MOG but also DNA. 20,21 When immunizing mice with MOG, the significant acceleration of EAE development strongly depends on the use of Pertussis toxin (PT). 41 PT was shown to affect the innate immune response; it can cross the blood-brain-barrier by increasing its permeability. 42 Thus, PT can stimulate the development of EAE in mice immunized with MOG because of increasing the blood-brain barrier permeability for MOG.
Previously we have demonstrated that DNA itself is a weak immunogen, but its immunogenic efficacy dramatically increases in different DNA-protein complexes. 43 Immunization of autoimmune MRL-lpr/lpr mice with a complex of DNA and methylated BSA, without Pertussis toxin, led to a strong acceleration of the development of mouse SLE, associated with the development of antibodies against DNA and against the abzymes efficiently hydrolysing DNA, ATP and oligosaccharides. [17][18][19] Comparable to the role of anti-MBP in SLE, high-affinity anti-DNA intrathecal IgGs have been identified as the major components of the cerebrospinal fluid cells and brains of MS patients. 39 DNase abzymes were revealed in~80%-90% of MS patients. 25,26 Similar to SLE, 34 DNase abzymes of MS patients 35 are cytotoxic and induce apoptosis; they therefore can play an important role not only in SLE but also in MS pathogenesis. As SLE and MS patients demonstrate some overlap in medical, biochemical, immunological and psychiatric indexes, 44 abzymes with DNase activity to some extent play a similar role in the development of both pathologies. Therefore, analogous to the treatment of MRL-lpr/lpr mice, [17][18][19] we first immunized C57BL/6 mice with DNAmet-BSA without toxin. 21 As already shown for the treatment with MOG, the immunization of C57BL/6 mice with DNA without Pertussis toxin leads to the development of EAE (similar to the development of SLE in MRL-lpr/lpr mice), but with a very long delay (approximately 16-20 days after immunization) until the manifestation of typical disease indicators. 21 In the absence of toxin, this delay could be because of a reduced blood-brain barrier permeability for the DNA-met-BSA complex. Anti-DNA Abs in AIDs are usually directed against nucleosomal DNA-histone complexes resulting in cell apoptosis. 36 As these complexes in cooperation with intrinsic human MBP could have an important role in the development of MS, we chose to immunize C57BL/6 mice with a DNA-histone complex combined with Pertussis toxin. In addition to various intranuclear functions, five histones act as damage molecules in the extracellular space. 6 Taking this into account, we analysed the development of Abs targeting histones and abzymes hydrolysing these proteins.

DNA Abs
The blood of healthy humans and animals usually contains auto-Abs against DNA and different proteins in low concentration. 10 Figure 1C). Surprisingly, immunization of

| The analysis of the criteria directly associated with antibody activities
IgGs were purified from individual mice by affinity chromatography of plasma proteins on Protein G-Sepharose using special conditions to remove non-specifically bound proteins as described before. [13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29] IgGs were then additionally purified using an FPLC gel filtration. For Previous studies showed that IgGs from control non-autoimmune BALB and CBA mice (at 3-7 months of age) did not possess detectable DNase and protease activities, while IgGs from untreated, MOG + toxin treated and DNA-met-BSA treated EAE mice demonstrated detectable or high activity in hydrolysis of DNA, MOG and MBP. 20,21 In our study, we confirm these data. Within the usual range of data variability, the estimations and average RAs in this study are very well comparable with those obtained previously. 20,21 After treatment of C57BL/6 mice with DNA-histones + toxin, IgGs are also active in the hydrolysis of all these substrates.

| Time-dependent changes in antibodies catalytic activities
Comparable to DNA-met-BSA, the activity began to grow very rapidly starting at 25-30 days and at day 63 it was 202-fold higher than at the onset of the experiment and 8-fold higher than maximal activity after mice immunization with MOG + toxin ( Figure 3A).
Immunization of mice with MOG + toxin led to a statistically sig-
In untreated control mice, the average numbers of BFU-E and CFU-E were nearly the same during the first 20-25 days, and increased 2.1-and 4.1-fold, respectively, by day 63 (Figure 5A,B).
Treating mice with MOG + toxin led to changes in BFU-E and CFU-E colonies going in opposite directions. The number of BFU-E colonies decreased 1.9-fold in the acute phase of EAE (at 6-20 days), and continuously increased until day 63 by a factor of 1.8 compared to time zero of the experiment ( Figure 5A). There was a very sharp 6.5-fold increase of the relative number of CFU-E colonies at 6-10 days, with a slight~1.1-fold decrease in their number during days 20-63 ( Figure 5B). In the case of only mice immunized with the DNA-histone + toxin, there was a significant continuous increase in the number of BFU-E colonies, 1.5-fold on average from day 7-63.
The mice immunized with DNA-met-BSA showed an increase in the number of colonies after 25 days and at day 63 the numbers were approximately 1.4-fold higher than that at time zero ( Figure 5A).
Both antigens (DNA-histone + toxin and DNA-met-BSA without toxin) resulted in very small changes in the number of CFU-E colonies during 0-63 days ( Figure 5B).
To some extent, a comparable increase in the number of CFU-GM colonies was observed in untreated control mice and animals immunized with DNA-histone + toxin and DNA-met-BSA without toxin ( Figure 5C). Only for mice immunized with MOG + toxin, a sharp 1.6-fold decrease in the CFU-GM colonies was observed up to day 10, followed by a slow growth in their number until day 63. can only minimally penetrate the blood-brain barrier.
The treatment of mice with DNA-histone + toxin resulted in a dramatic increase in the lymphocyte proliferation especially during days 14-20 (1.7-1.8-fold) but then significantly decreased ( Figure 6A). This is most likely because of the presence of a toxin which helps DNA-histone + toxin to penetrate the blood-brain barrier and stimulate the synthesis of lymphocyte producing antibodies against MOG, MBP, DNA and histones.
The relative level of spleen lymphocyte proliferation during spontaneous EAE rises slowly and increases 1.9-fold starting 20 days after immunization until day 63 ( Figure 6B). Treating mice with DNA-met-BSA without toxin suppresses the proliferation of the spleen lymphocytes, and until day 63 proliferation rates increase only 1.3fold. Treating mice with MOG + toxin and DNA-histone + toxin leads to comparable profiles and levels of changes in lymphocyte proliferation in the spleen, with two well-defined peaks at 14 (2.6-and 2.3-fold, respectively) and 40 days (1.6-and 2.3-fold, respectively).
We cannot exclude that a significant difference in the profiles of spleen lymphocyte proliferation by DNA-met-BSA without toxin and DNA-histone + toxin may occur because of the effect of the toxin on the proliferation. Moreover, we cannot rule out that to some extent a stronger proliferation of lymphocytes after immunization with DNA-histone + toxin is associated with the increased immunogenicity of the DNA complex with histones. In our experiments, mice treated with DNA-histone + toxin exhibit an increase in the concentration of TNFα, which is involved in systemic inflammation and the acute phase reactions, 48,49 and IL1β, which is an important mediator of the inflammatory response and activator of lymphocytes 49 in the acute phase of EAE ( Figure 1F).

| Cell apoptosis assay
Interestingly, immunization of mice with MOG + toxin leads to the production of Abs and abzymes not only for this protein but also targeting DNA, while the treatment of animals with DNA-histone + toxin induces the development of IgGs hydrolysing histones, DNA, MOG and MBP. It is an important finding that these activities are significantly higher after immunization of mice with antigens used together with the toxin (Figure 3), as the latter increases the penetration of antigens through the blood-brain barrier.
As shown earlier [17][18][19][20][21] (Figure 4). In the latter cases, the level of antibody activity with histone hydrolysis is much lower than the activity seen for non-immunized mice (Figure 4). This finding speaks in favour of the competition of histones and mice MBP for interaction with lymphocytes and direction of their following differentiation. As indicated above, anti-DNA Abs are usually directed against histone-DNA complexes, resulting in internucleosomal cleavage during apoptosis. 36 Taking this into account, one would expect that after mice immunization with DNA-histone + toxin, the production of Abs against DNA and histones can occur in parallel approximately at the same time. However, antibodies efficiently hydrolysing histones are detected during the acute AULOVA ET AL.
| 5829 phase of EAE (7-20 days after immunization) (Figure 4), while the initial stage of production of DNase antibodies is observed only after 20-25 days ( Figure 3A). Interestingly, a strong delay (at 18-20 days) of the initial stage of production of DNA-hydrolysing antibodies is observed for both antigens containing DNA (DNA-met-BSA and histone-DNA + toxin) ( Figure 3A). It should be told that in this situation, that is in the earliest stage of human AIDs, there is usually a production of abzymes against only one specific antigen, for example SLE (anti-DNA), MS (anti-MBP) or autoimmune Hashimoto thyroiditis (anti-thyroglobulin). [8][9][10][11][12][13] In the later stages of autoimmune pathologies, there is usually a very strong expansion of the abzymes reper-  52 In addition, SLE and MS patients demonstrate some similarity in the development of medical, biochemical and immunological indexes, including a comparable change in the profiles of brain stem cell differentiation and lymphocyte proliferation. [11][12][13]20,21,53 The similarity of some immunological indexes between MS and SLE supports the assumption that anti-MBP with proteolytic activity may occur in SLE patients. Interestingly, DNase Abs are on averagẽ 57-fold more active in SLE than MS, while MBP-hydrolysing activity in MS is~7-to 10-fold higher than in SLE patients. 27 Taken together, the development of EAE in mice is associated with changes in the profiles of the differentiation of bone marrow stem cells, changes in the levels of lymphocyte proliferation in various organs, significant levels of cell apoptosis, the production of harmful auto-Abs to DNA, MBP and histones as well as abzymes with various activities and increased concentrations of cytokines.
It is believed that in individual patients with AIDs, the development of initial normal immune reactions can be stimulated by foreign antigens of different bacterial as well as viral infections. [54][55][56][57] Molecular homology and mimicry between human and viral agents such as Epstein-Barr, measles, hepatitis B, herpes simplex, influenza and papilloma viruses may be involved in the autoimmune pathogenesis of MS. 58 Perhaps the antigens of such viruses (and/or some bacteria) are able to penetrate the blood-brain barrier and induce a change in the profile of differentiation and the level of proliferation of bone marrow stem cells in case of long periods of illness and complications. As the abzymes in the bone marrow of MS patients are more active than those in the blood of the same MS patients, [30][31][32] we can assume that the bone marrow lymphocytes are subjected by viral antigens to a special type of differentiation, leading them to synthesize harmful autoantibodies and abzymes in the bone marrow and in different organs.
Here we show for the first time that rapid development of EAE in mice can be stimulated not only by their immunization with MOG + toxin, but also with DNA-met-BSA, and DNA-histone + toxin. We cannot exclude that the initial stages of MS development in humans may be associated with an immunization with a variety of antigens.