www.elsevier.com/locate/ycimm

Cellular Immunology 247 (2007) 12–17

Immunogenicity of mitochondrial DNA modified by hydroxyl radical

Khurshid Alam *, Moinuddin, Suraya Jabeen

Department of Biochemistry, Faculty of Medicine, J.N. Medical College, A.M.U., Aligarh 202 002, India

Received 9 April 2007; accepted 28 June 2007Available online 22 August 2007

Abstract

Mitochondria consume about 90 percent of oxygen used by the body, and are a particularly rich source of reactive oxygen species(ROS). In this research communication mitochondrial DNA (mtDNA) was isolated from fresh goat liver and modified in vitro by hydro-xyl radical generated from UV irradiation (254 nm) of hydrogen peroxide. As a consequence of hydroxyl radical modification, mtDNAshowed hyperchromicity and sensitivity to nuclease S1 digestion as compared to control mtDNA. Animals immunized with mtDNA andROS-modified mtDNA induced antibodies as detected by direct binding and competition ELISA. The data suggest that immunogenicityof mtDNA got augmented after treatment with hydroxyl radical. IgG isolated from immune sera showed specificity for respective immu-nogen and cross-reaction with other nucleic acids. Binding of induced antibodies with array of antigens clearly indicates their polyspecificnature. Moreover, the polyspecificity exhibited by induced antibodies is unique in view of similar multiple antigen binding properties ofnaturally occurring anti-DNA antibodies derived from SLE patients.� 2007 Elsevier Inc. All rights reserved.

Keywords: mtDNA; Hydroxyl radical; Immunogenicity; Antibodies; SLE

1. Introduction

Human mitochondrial DNA is a double-stranded circu-lar molecule of 16,568 base pairs that codes for thirteenessential genes of oxidative phoshorylation [1–3]. Eachhuman cell has hundreds of mitochondria and multiplecopies of mtDNA and is the only extra chromosomalDNA in human cells [4]. Lack of a chromatin structure,histones protection, and inefficient repair system makesmtDNA susceptible to oxidative damage. ROS-induceddamage to mtDNA has resulted in accumulation of 8-oxodGuo (damage marker) in UV-irradiated hepatoma cellline [5]. Approximately 1–5% of the oxygen consumed bymitochondria is converted to ROS under physiologicalconditions [6] and thus ROS production appears to beessentially a function of oxygen consumption. Further-more, increased mitochondrial activity per se can be an oxi-

0008-8749/$ - see front matter � 2007 Elsevier Inc. All rights reserved.

doi:10.1016/j.cellimm.2007.06.007

* Corresponding author. Fax: +91 571 2702758.E-mail address: kalam786@rediffmail.com (K. Alam).

dative stress to cell [7]. Electron transport chain of theorganelle is a major source of ROS and mtDNA is there-fore exposed to high levels of damaging oxygen species.Indeed, oxidative DNA-base damage measured as 8-hydroxydeoxyguanosine (oxodGuo) has been detected inmtDNA at steady-state levels several fold higher than innuclear DNA [8]. This apparent difference in damage couldbe due to proximity of mtDNA to ROS generated duringelectron transport. Mitochondrial dysfunction is importantcause of certain human diseases [9,10] and cumulativemtDNA damage is implicated in the aging process and inthe progression of such common diseases as diabetes, can-cer, and heart failure [1,7]. Mitochondrial DNA damage, ifnot repaired, leads to disruption of electron transport chainand production of more ROS. This vicious cycle of ROSproduction and mtDNA damage ultimately leads to energydepletion in the cell and apoptosis [11,12]. Activation ofapoptotic pathway, in turn, inhibits the rate of mitochon-drial translation. The damaged mtDNA could accumulatein senescent cells. The cells loaded with perturbed DNAundergo apoptosis and the contents are eliminated.

K. Alam et al. / Cellular Immunology 247 (2007) 12–17 13

However, some molecules of in vivo ROS-modifiedmtDNA might escape elimination and present itself tothe immunoregulatory network as neoantigen or migrateto nucleus and integrate into the nuclear genome [13]. Incertain pathology increased apoptosis and delayed removalof apoptotic cell debris may also favor persistence of ROS-modified mtDNA in circulation inter alia exposure toimmunoregulatory network and thus making the moleculeimmunogenic.

Double stranded DNA (B-conformation) against whichmost of the antibodies are detected in human lupus is nolonger regarded as an antigen initiating the disease becauseimmunization with B-form of DNA does not induce clini-cal features typical of SLE [14]. Involvements of majororgans like heart, lungs, kidneys and central nervous sys-tem have been shown to be the cause of mortality resultingfrom SLE [15]. The anti-dsDNA autoantibodies, markerantibody for the confirmed diagnosis of SLE, show diverseantigen binding [16,17] which include components ofDNA, its different conformations and chemically modifiedstructures [16,18].

In this communication immunology of mtDNA and itshydroxyl radical modified counterpart has been studied.The binding properties of experimentally induced antibod-ies were compared with SLE derived pathogenic anti-DNAantibodies.

2. Materials and methods

2.1. Isolation and purification of mitochondrial DNA

Mitochondrial DNA was isolated as per the proceduresdescribed elsewhere [19]. Briefly, fresh goat liver washomogenized in ice-cold buffer of following composition:0.3 M sucrose, 25 mM Tris, 10 mM EDTA (pH 7.0). Thehomogenate was centrifuged at 4000 rpm (0 �C) for10 min and supernatant was preserved. The pellet was sus-pended in one-tenth volume of above-mentioned buffer andre-centrifuged. The mitochondrial pellet, thus obtained,was suspended in 20 ml of 10 mM Tris, 0.1 M NaCl,0.1 M EDTA (pH 8.0). SDS was added to a final concen-tration of 2% and the mixture was incubated for 60 s at37 �C followed by 9 min incubation at room temperatureto complete the lysis. The contaminant high molecularweight nuclear DNA (if any) was precipitated by one-thirdvolume of 4 M NaCl (final concentration, 1 M). The con-tents were gently mixed by inverting the tube and mixturewas placed at 0 �C for 4–16 h. The material was centrifugedat 27,000g for 15 min at 0 �C and supernatant was deprote-inized thrice using chloroform/isoamylalcohol (24:1) mix-ture. The aqueous layer rich in mitochondrial nucleicacids was concentrated by precipitation with two volumesof ethanol at �20 �C. The alcohol precipitate was collectedby centrifugation and dissolved in minimum volume ofTris–EDTA containing 0.1 M NaCl. The material waspassed through Sepharose CL 4B column to obtain puremtDNA.

2.2. Modification of mitochondrial DNA

Mitochondrial DNA was modified by hydroxyl radicalas described earlier [20]. Aqueous solution of mitochon-drial DNA (0.15 mM base pair) in PBS was irradiatedunder 254 nm UV light for 30 min at room temperaturein presence of hydrogen peroxide (15.1 mM). Excess hydro-gen peroxide was removed by extensive dialysis againstPBS (pH 7.4).

2.3. SLE patients

Informed consent was obtained from patients (allfemales) before taking blood samples. All patients fulfilledat least four criteria of American College of Rheumatologyfor the classification of SLE.

2.4. Immunization scheme

Mitochondrial DNA and its hydroxyl-modified counter-part (25 lg) were separately complexed with equal amountof methylated BSA and emulsified with complete Freund’sadjuvant. The complex was injected intramuscularly in thehind leg muscles of rabbits (three animals in each group).Subsequent injections were given in incomplete Freund’sadjuvant. Weekly injections of antigen (mtDNA/hydro-xyl-modified mtDNA) were given for seven weeks and thuseach animal received a total of 175 lg antigen during thecourse of immunization. One week after the last dose ofimmunogen, blood was collected; serum separated anddecomplemented by heating at 56 �C for 30 min. Preim-mune blood was collected prior to immunization. Thematerial was stored in small aliquots at �80 �C withsodium azide (0.1%) as preservative.

2.5. Isolation of IgG by protein-A-Sepharose

Protein-A-Sepharose CL 4B was used to purify serumIgG. Serum (0.5 ml) diluted with equal volume of PBS(pH 7.4), was applied to the column (0.9 · 15 cm) pre-equilibrated with above buffer. The wash through was recy-cled 2–3 times. Unbound IgG was removed by extensivewashing with PBS (pH 7.4). The bound IgG was elutedwith 0.58% acetic acid in 0.85% sodium chloride [21]. Threemilliliter fractions were collected in a measuring cylinderalready containing 1 ml of 1 M Tris–HCl, pH 8.5, andabsorbance was recorded. The IgG concentration wasdetermined considering 1.4 OD280 = 1.0 mg mammalianIgG/ml. The isolated IgG was then dialyzed against PBS(pH 7.4) and stored at �80 �C.

2.6. Enzyme linked immunosorbent assay (ELISA)

Serum antibodies were detected by ELISA on flat bot-tom microtiter wells [22]. Test wells were filled with100 ll DNA antigens (2.5 lg/ml in TBS) and incubatedfor 2 h at 37 �C and overnight at 4 �C. The wells were

14 K. Alam et al. / Cellular Immunology 247 (2007) 12–17

washed thrice with TBS-T to remove unbound antigen.Unoccupied sites were blocked with 150 ll of 1.5% BSA(in TBS) for 5 h at room temperature. The plates werewashed once with TBS-T and preimmune/immune sera(100 ll per well of 1:100 dilution), or IgG isolated fromsera, were absorbed for 2 h at room temperature and over-night at 4 �C. The unbound antibody was washed off fourtimes using TBS-T. An appropriate anti-immunoglobulinalkaline phosphate conjugate (100 ll of 1:2000 dilution)was added to each well. In case of inhibition ELISA,immune complex was added instead of serum/IgG. Theimmune complex was formed by mixing varying concentra-tions of antigen (inhibitor) with fixed amount of antibody.The mixture was incubated for 2 h at room temperatureand overnight at 4 �C and complex thus formed was coatedin place of serum/IgG. After incubation at 37 �C for 2 h,the plates were washed four times with TBS-T and threetimes with distilled water and developed using p-nitro-phenyl phosphate. Absorbance was read at 410 nm.

3. Results

3.1. Characterization of mtDNA and its modification by

hydroxyl radical

Incubation of isolated mtDNA with single strand spe-cific nuclease S1 enzyme did not affect its co-migration withmtDNA alone in agarose gel and the material was consid-ered to be pure (Fig. 1a). The circular mtDNA was modi-fied with hydroxyl radical generated by UV irradiation(254 nm) of hydrogen peroxide. The UV absorption profileof hydroxyl-modified mtDNA showed hyperchromicitycompared to control (Fig. 1b). Furthermore, at 260 nmwavelength the hyperchromicity was approximately 35 per-cent. The observed hyperchromicity in mtDNA suggeststructural changes in the molecule as a consequence ofhydroxyl radical attack.

Fig. 1. (a) Agarose gel electrophoretogram of mtDNA (lane 1) treatedwith nuclease S1 (lane 2). (b) Ultraviolet absorption spectra of mtDNA(- - -) and hydroxyl-modified mtDNA (—).

3.2. Physico-chemical characterization of ROS-modified

mtDNA

Heat induced melting of mtDNA and hydroxyl-modi-fied mtDNA was monitored at 260 nm from 30 to 95 �Cat the rate of 1.5 �C/min [19]. While mtDNA did not showheat-induced denaturation under our experimental condi-tions, the melting temperature (Tm) of hydroxyl-mtDNAwas observed to be 78 �C. Table 1 summarizes the UVand thermal denaturation properties of mtDNA and itsmodified form.

3.3. Nuclease S1 treatment of hydroxyl-modified mtDNA

Mitochondrial DNA and its modified form were incu-bated with nuclease S1 (20 IU/lg DNA) for 30 min at37 �C. EDTA terminated the reaction and samples wereelectrophoresed in agarose gel. The enzyme digested hydro-xyl-modified mtDNA (figure not shown), which is indica-tive of structural changes in mtDNA as a consequence ofhydroxyl radical attack. Mitochondrial DNA, used as con-trol, did not show any sensitivity towards nuclease S1.

3.4. Immunogenicity of mtDNA and hydroxyl-modified

mtDNA

Rabbits were immunized with mtDNA and hydroxyl-modified mtDNA. At the end of immunization blood waswithdrawn and serum separated. Direct binding ELISArevealed that hydroxyl-modified mtDNA is more immuno-genic than mtDNA (Fig. 2). Preimmune sera, used ascontrol, did not show appreciable binding with eitherimmunogen. Specificity of the induced antibodies wasevaluated by competitive binding assays [22,23]. Specificityof induced antibodies towards respective immunogen wasevident from inhibition in antibody activity usingimmunogen as inhibitor (Fig. 3).

3.5. Cross-reactivity of induced antibodies

IgG was isolated from anti-mtDNA and anti-hydroxyl-modified mtDNA sera and subjected to binding studies onsolid phase coated with nucleic acid antigens. Althoughinduced antibodies were highly specific for their immuno-gen, they did cross-react with nucleic acids of varying size

Table 1UV and thermal denaturation characteristics of mtDNA and hydroxyl-modified mtDNA

Parameter mtDNA Hydroxyl-modifiedmtDNA

Physical shape Circular LinearAbsorbance ratio (A260/A280) 1.84 1.47Percent hyperchromicity at 95 �C Not observed 20Melting temperature (Tm), �C Not observed 78Onset of melting, �C Not observed 65

- log serum dilution

Abs

orba

nce

at 4

10 n

m

2.0 2.3 2.6 2.9 3.2 3.5 3.80.0

0.5

1.0

1.5

Fig. 2. Direct binding ELISA of anti-mtDNA antibodies (n) and anti-hydroxyl-modified mtDNA antibodies (d) with their respective immuno-gen. Corresponding unfilled symbols (h, s) represent binding ofpreimmune serum with immunogen.

Inhibitor concentration, μg/ml

Per

cent

inhi

bitio

n

0.1 1.0 10.0 100.00

25

50

75

100

Fig. 3. Inhibition ELISA of anti-mtDNA antibodies (s) and anti-hydroxyl-modified mtDNA antibodies (d) with their respectiveimmunogen.

Table 2Competitive inhibition data of anti-ROS-mtDNA IgG

Inhibitor Maximum percentinhibition at 20 lg/ml

Concentrationrequired for 50%inhibition (lg/ml)

ROS-mtDNA 80.5 6mtDNA 63 9ssDNA 58 17DNA 48 —ROS-DNA 65 13Superoxide-modified DNA 27 —Z-DNA 8 —RNA 40 —ROS-RNA 75 8Poly (G) 25 —ROS-poly(G) 29 —

The microtiter plates were coated with ROS-mtDNA (2.5 lg/ml).

Table 3Competitive ELISA of anti-mtDNA IgG

Inhibitor Maximum percentinhibition at 20 lg/ml

Concentrationrequired for 50%inhibition (lg/ml)

mtDNA 72 8ROS-mtDNA 41 —ssDNA 31DNA 59 11ROS-DNA 47 —Superoxide-modified DNA 32 —Z-DNA No inhibition —RNA 35 —ROS-RNA 28Poly (G) 22 —ROS-poly(G) 33 —

The microtiter plates were coated with mtDNA (2.5 lg/ml).

Table 4Inhibition of SLE anti-DNA antibodies by native DNA, mtDNA, andOH-mtDNA

SLE serum Maximum percent inhibition at 20 lg/ml

nDNA mtDNA OH-mtDNA

1 49 45 512 66 51 623 22 35 414 62 46 565 68 48 596 69 52 647 37 45 568 59 54 65

Means ± SEM 54 ± 5.96 47 ± 2.08 56.75 ± 2.78aNHS (n = 5) 15 ± 6.17 12 ± 8.06 18 ± 4.2

The microtiter wells were filled with nDNA (2.5 lg/ml).a Means ± SEM values of serum autoantibodies of normal healthy

subjects binding with respective inhibitors.

K. Alam et al. / Cellular Immunology 247 (2007) 12–17 15

and structure (Tables 2 and 3). The cross-reactive proper-ties of induced antibodies appear to be due to theirpolyspecific nature, a characteristic feature of most of theSLE derived pathogenic anti-DNA autoantibodies.

3.6. Binding of human SLE autoantibodies with DNA,

mtDNA and hydroxyl-modified mtDNA

Anti-dsDNA positive SLE samples were subjected tocompetitive binding assay with DNA, mtDNA and hydro-xyl-modified mtDNA. The human autoantibodies werewell recognized by both mtDNA and hydroxy-modifiedmtDNA (Table 4). Serum autoantibodies of normalhealthy subjects (NHS) showed little binding withinhibitors.

4. Discussion

The mtDNA is not protected by histones, as is thenuclear DNA, and it lies in close proximity to the free

16 K. Alam et al. / Cellular Immunology 247 (2007) 12–17

radical producing process of oxidative phosphorylation.Approximately 1–5% of the oxygen consumed by mito-chondria is converted to reactive oxygen species underphysiological conditions. The production of ROS is essen-tially a function of O2 consumption. Hence, increasedmitochondrial activity per se can be an oxidative stress tocell. Mammalian cells have numerous enzymatic andnon-enzymatic antioxidant defenses to guard the level ofpotentially dangerous reactive oxygen species. Cumulativedamage of mtDNA is implicated in the progression ofmany human diseases.

The objective of this study was twofold. The first aimwas to evaluate the immunogenicity of mtDNA in experi-mental animals and secondly the mtDNA was subjectedto in vitro modification by hydroxyl radical and experimen-tal animals were challenged with this neo-antigen. Nativeform of mammalian DNA (B-conformation) is non-immu-nogenic. In contrast, bacterial DNA is a strong immuno-gen [24]. The immunogenicity of bacterial DNA has beenattributed to the presence of nucleotide hexamers contain-ing unmethylated CpG motifs. Similarly, the role ofunmethylated CpG dinucleotide sequences in the immuno-genicity of plasmid DNA is well recognized [25–27]. Othermodified forms of DNA and polynucleotides [28,29],modified self determinants [30] and self proteins [31] havebeen reported to be immunogenic and the antibodies thusgenerated are cross reactive with native DNA. MammalianDNA complexed with synthetic peptide Fus-1 elicitedanti-dsDNA response in mice [32].

Mitochondrial DNA modified with ROS exhibitedhyperchromicity at 260 nm and showed single strandbreaks. These changes in mtDNA are attributed to semiloss of secondary structure as a consequence of hydroxylradical insult. Mitochondrial DNA and its ROS form weresubjected to nuclease S1 digestion. The enzyme selectivelychopped off the single stranded regions in ROS-modifiedmtDNA. Unmodified mtDNA was not affected by nucleasetreatment.

Both mtDNA and its ROS counterpart were immuno-logically active and induced antibodies in experimental ani-mals. Comparative studies revealed that ROS-modifiedmtDNA is a better immunogen. Antibodies against ROS-modified mtDNA were highly specific for its immunogenand showed cross-reactivity with mtDNA. Cross-reactionsof antibodies (experimentally induced or naturally occur-ring) with closely related antigens are attributed to epitopessharing between/among antigens. The preferential bindingof anti-ROS-mtDNA IgG with ROS forms of DNA, RNAand poly (G) is a quasi evidence of active role being playedby activated oxygen species in modifying antigenic struc-ture of macromolecules.

Raised respiration of mitochondrial cells will increaseROS production to levels beyond detoxifying defenses,which may damage mtDNA. Subsequent in vitro studieson ROS and mtDNA revealed base modification andstrand breaks in mtDNA [33]. Furthermore, based on thefinding that mtDNA is fragmented by ROS, Richter [34]

suggested that mtDNA fragments carrying modified basesescape from mitochondria and accumulate in a time-depen-dent manner in nuclear DNA. In our view there are fairchances that body’s immune cells might consider this frag-mented and modified mtDNA as a foreign body and elicitantibody response typical of anti-DNA autoantibodies.Systemic lupus erythematosus (SLE) is an autoimmune dis-order characterized by high titer polyspecific anti-dsDNAantibodies in patients’ sera. What causes SLE is reallynot known and all attempts to induce antibodies againstnDNA in experimental animals have failed. Anti-DNApositive SLE sera were subjected to competitive bindingassay using mtDNA and ROS-modified-mtDNA as inhib-itors (Table 4). Analysis of data in Table 4 clearly indicatesthat mtDNA and its modified form were fairly recognizedby anti-nDNA positive SLE samples. In view of the non-immunogenic nature of nDNA and the immunogenicityexhibited by mtDNA and hydroxyl-modified mtDNA inour experimental conditions, it is possible that mtDNAand its modified from might act as triggering antigen in asubgroup of lupus patients.

Acknowledgment

This study was supported by a research grant [37(980)/98/EMR-II] to K.A. from Council of Scientific and Indus-trial Research, New Delhi.

References

[1] S. Ohta, A multifunctional organelle mitochondria is involved incell death, proliferation and disease, Curr. Med. Chem. 10 (2003)2485–2494.

[2] T. Stevnsner, T. Thorslund, N.C. Souza-Pinto, V.A. Bohr, Mito-chondrial repair of 8-oxoguanine and changes with aging, Exp.Gerontol. 37 (2002) 1189–1196.

[3] J.M. Shoffner, D.C. Wallace, Oxidative phosphorylation diseases.Disorders of two genomes, Adv. Hum. Genet. 19 (1990) 267–330.

[4] T.M. Wardell, E. Ferguson, P.F. Chinnery, G.M. Borthwick,R.W. Taylor, G. Jackson, A. Craft, R.N. Lightowlers, N. Howell,D.M. Turnbull, Changes in the human mitochondrial genomeafter treatment of malignant disease, Mutat. Res. 525 (2003)19–27.

[5] D. Takai, S.H. Park, Y. Takada, S. Ichinose, M. Kitagawa, M.Akashi, UV-irradiation induces oxidative damage to mitochondrialDNA primarily through hydrogen peroxide: analysis of 8-oxodGuoby HPLC, Free Radic. Res. 40 (2006) 1138–1148.

[6] S. Papa, Mitochondrial oxidative phosphorylation changes in the lifespan. Molecular aspects and physiopathological implications, Bio-chim. Biophys. Acta 1276 (1996) 87–105.

[7] D. Kang, N. Hamasaki, Mitochondrial oxidative stress and mito-chondria, Clin. Chem. Lab. Med. 41 (2003) 1281–1288.

[8] C. Richter, Reactive oxygen and DNA damage in mitochondria,Mutat. Res. 275 (1992) 249–255.

[9] R. Luft, The development of mitochondrial medicine, Proc. Natl.Acad. Sci. USA 91 (1994) 8731–8738.

[10] D.C. Wallace, Mitochondrial diseases in man and mouse, Science 283(1999) 1482–1488.

[11] M. Mirabella, S. DiGiovanni, G. Silvestri, P. Tonali, S. Servidei,Apoptosis in mitochondrial encephalomyopathies with mitochondrialDNA mutations: a potential pathogenic mechanism, Brain 123 (2000)93–104.

K. Alam et al. / Cellular Immunology 247 (2007) 12–17 17

[12] B.S. Mandavilli, J.H. Santos, B. VanHouten, Mitochondrial DNArepair and aging, Mutat. Res. 30 (2002) 127–151.

[13] G. Kim, H. Sikder, K.K. Singh, A colony color method identifies thevulnerability of mitochondria to oxidative damage, Mutagenesis 17(2002) 375–381.

[14] D.S. Pisetsky, J.P. Grudier, G.S. Gilkeson, A role for the immuno-genic DNA in the pathogenesis of systemic lupus erythematosus,Arth. Rheum. 33 (1990) 159–164.

[15] J. Nossent, N. Cikes, E. Kiss, A. Marchesoni, V. Nassonova, M.Mosca, M. Olesinska, G. Pokorny, B. Rozman, M. Schneider, P.G.Vlachoyiannopoulos, A. Swaak, Current causes of death in systemiclupus erythematosus in Europe, 2000–2004: relation to diseaseactivity and damage accrual, Lupus 16 (2007) 309–317.

[16] K. Alam, R. Ali, Human autoantibody binding to multiple confor-mations of DNA, Biochem. Int. 26 (1992) 597–605.

[17] J. Ara, R. Ali, Reactive oxygen species modified DNA fragments ofvarying size are the preferred antigen for human anti-DNA autoan-tibodies, Immunol. Lett. 34 (1992) 195–200.

[18] K. Alam, N. Islam, R. Hasan, A. Ali, R. Ali, Naturally occurringSLE anti-DNA antibodies recognize unique conformation on DNA-lysine photoadduct, Microbiol. Immunol. 36 (1992) 1003–1007.

[19] B. Chatterjee, G.R.K. Rao, A simple method for purification ofmtDNA, Indian J. Biochem. Biophys. 21 (1984) 378–380.

[20] K. Alam, A. Ali, R. Ali, The effect of hydroxyl radical on theantigenicity of native DNA, FEBS Lett. 319 (1993) 66–70.

[21] J.W. Goding, Use of staphylococcal protein-A as an immunologicalreagent, Immuol. Methods 20 (1976) 241–253.

[22] R. Ali, K. Alam, Evaluation of antibodies against oxygen free radical-modified DNA by ELISA, first ed., in: D. Armstrong (Ed.), Methodsin molecular biology: oxidative stress biomarkers and antioxidantprotocols, vol. 186, Humana Press, New Jersey, 2002, pp. 171–181.

[23] G. Waris, K. Alam, Immunogenicity of superoxide radical modified-DNA: studies on induced antibodies and SLE anti-DNA autoanti-bodies, Life Sci. 75 (2004) 2633–2642.

[24] D.S. Pisetsky, The immunologic properties of DNA, J. Immunol. 156(1996) 421–423.

[25] A.M. Kreig, A. Yi, S. Matson, T.J. Waldschmidt, G.A. Bishop, R.Teasdale, G.A. Koretzky, D.M. Klinman, CpG motifs in bacterialDNA triggers B-cell activation, Nature (London) 374 (1995)546–549.

[26] Y. Sato, M. Roman, H. Tighs, D. Lee, M. Corr, M. Nguyen, G.J.Silverman, M. Lotz, D.A. Carson, E. Raz, Immunostimulatory DNAsequences necessary for effective intra-dermal gene immunization,Science 273 (1996) 352–354.

[27] R. Grifantini, O. Finco, E. Bartolini, M. Draghi, G.D. Giudice, C.Kochen, A. Thomas, S. Abrignain, G. Grandi, Multi-plasmid DNAvaccination avoids antigenic competition and enhances immunoge-nicity of a poorly immunogenic plasmid, Eur. J. Immunol. 28 (1998)1225–1232.

[28] Z. Arif, R. Ali, Antigenicity of poly(dA–dT).poly(dA–dT) photo-crosslinked with 8-methoxypsoralen, Arch. Biochem. Biophys. 329(1996) 191–198.

[29] J. Ahmad, B.T. Ashok, R. Ali, Reactive oxygen species modifiedthymine and poly(dT) represent unique epitope for human anti-DNAautoantibodies, Immunol. Lett. 58 (1997) 69–74.

[30] M.J. Mamula, Lupus autoimmunity: from peptides to particles,Immunol. Rev. 144 (1995) 301–314.

[31] D.J. Zack, K. Yamamoto, A.L. Wong, M. Stempniak, C. French,R.H. Weisbart, DNA mimics a self-protein that may be a target forsome anti-DNA autoantibodies in systemic lupus erythematosus, J.Immunol. 154 (1994) 1987–1994.

[32] D.D. Desai, M.P. Krishnan, J.T. Swindle, T.N. Marion, Antigen-specific induction of antibodies against native mammalian DNA innon-autoimmune mice, J. Immunol. 151 (1993) 1614–1626.

[33] C. Richter, Oxidative damage to mitochondrial DNA and itsrelationship to ageing, Int. J. Biochem. Cell. Biol. 27 (1995) 647–653.

[34] C. Richter, Do mitochondrial DNA fragments promote cancer andageing? FEBS Lett. 241 (1988) 1–5.