ORIGINAL ARTICLE

Bharathi Æ Kosagi S. Jagannatha Rao Æ Reuven Stein

First evidence on induced topological changes in supercoiled DNAby an aluminium D-aspartate complex

Received: 31 March 2003 / Accepted: 10 July 2003 / Published online: 27 September 2003� SBIC 2003

Abstract Aluminium is a debatable and suspected etio-logical factor in neurodegenerative disorders. Alumin-ium–amino acid complexes also play an important rolein the complex biology of the metal. Recent reportsindicate the presence of D-aspartate and D-glutamate inaging brain, human breast tumors, core amyloid plaquesand neurofibrillary tangles of Alzheimer’s brain. Thisstereoinversion from the L- to the D-enantiomer isenhanced by Al. Further, the observation that Al islocalized in the chromatin region encouraged the presentstudy of the interaction of Al–amino acid complexeswith DNA. This study used circular dichroism ofsupercoiled DNA and showed that Al–D-Asp caused anative B-DNA to C-DNA conformational change, whileAl–L-Asp, Al–L-Glu and Al–D-Glu did not alter thenative B-DNA conformation. This differential DNAbinding property of Al–amino acid complexes isassigned to the stereoisomerism and chirality of thecomplexes. Interestingly, polyamines like spermine fur-ther induced an asymmetric condensation of the ‘‘limitC-motif’’ induced by Al–D-Asp to a Y-DNA. The resultswere supported by computer modeling, gel studies andethidium bromide binding. We also propose a mecha-nism of Al–D-Asp binding and its ability to modulateDNA topology.

Keywords Aluminium Æ C-DNA Æ DNA topology ÆRacemization Æ Spermine

Introduction

Aluminum has never been demonstrated to have defi-nite biological function, suggesting that Al possessesproperties which are incompatible with fundamentallife processes. It is usually excluded from normal bio-chemical and metabolic processes because of its lowsolubility and chemical forms, namely silicates, phos-phates and oxides, which render the Al physiologicallyunavailable. The potential role of Al in the pathogen-esis of human neurodegenerative disorders remainscontroversial. Recent studies in experimental animalsas well as circumstantial evidence provided clues thatAl might be one of the suspected and also highlydebatable factors causing neuronal cell death in neu-rodegenerative brain [1]. Al is the only trivalent metalion that appears capable of inducing progressiveencephalopathy in the brain [2]. In biological systems,proteins and free amino acid pools predominantlycontain the L-forms of amino acids. Similarly in normalhuman brain, vast number of amino acids exists in theL-configuration while only a few are in the D-configu-ration. However, in neurodegenerative brain theamount of racemized forms of amino acids, in partic-ular D-aspartate (D-Asp) and D-glutamate (D-Glu), arerelatively in large proportions in core amyloid plaquesand neurofibrillary tangles (NFTs) [3], and D-Asppromotes aggregation and fibril formation of the Abpeptide [4, 5]. Biologically uncommon D-Asp residueshave also been reported in proteins of tooth enamel,the eye lens, the aorta and the brain (Ab and tau) inelderly humans [6, 7, 8, 9]. The reason for amino acidracemization under in vivo conditions is not known.Further, the chelation of amino acids by positivelycharged metal ions stabilizes the intermediate carban-ion succinimide by loss of alpha hydrogen, thusincreasing the racemization rate from the L-form to theD-form [10, 11]. Recently, Latha et al. [12] reportedthat Al favors racemization from L-Asp and L-Glu toD-Asp and D-Glu, respectively, in the rabbit brain. The

J Biol Inorg Chem (2003) 8: 823–830DOI 10.1007/s00775-003-0484-1

Bharathi Æ K. S. Jagannatha Rao (&)Department of Biochemistry and Nutrition,Central Food Technological Research Institute,570013 Mysore, IndiaE-mail: kjr4n@yahoo.comFax: +91-821-2517233

R. SteinDepartment of Neurobiochemistry,Tel-Aviv University, Tel-Aviv, Israel

study also revealed that Al preferentially binds to theL-forms of Asp and Glu than to the D-forms. Alsorecently, Deloncle et al. [11] reported that Al in com-plexation with L-glutamate and L-transferrin was ableto cross the blood/brain barrier and deposit in thebrain. Interestingly, Al is also reported [13, 14] to belocalized in the chromatin region of nuclei. In thisperspective, it is important to study whether bothL- and D-forms of Al–Asp and Al–Glu complexeswould interact with DNA and contribute to DNAconformational change. It is quite pertinent to choosesupercoiled DNA (scDNA) as a model system for thisstudy, since a vast array of supercoiled pockets ofDNA exist in viruses, animals and human cells, whichare known to be involved in gene expression [15]. Thesesupercoiled pockets are comparable to the pUC-19DNA supercoiling with non-histone binding sites [16].Hence the results can be implicated to eukaryoticchromosomal DNA to provide an insight modelexplaining the possible role of Al–amino acid com-plexes in binding to DNA. To the best of our knowl-edge, this is the first report presenting comprehensivedata on interaction of Al–L-Asp, Al–D-Asp, Al–L-Gluand Al–D-Glu with supercoiled DNA, by monitoringchanges in the conformational pattern of DNA.

Materials and methods

Materials

pUC-19 supercoiled DNA (cesium chloride purified) was procuredfrom Bangalore Genei (India). Spermine, desferoximine, EDTA,agarose, chloroquine, ethidium bromide and HEPES were fromSigma (USA).

Preparation of aluminium–amino acid complexes

AlCl3 (10)2 M) was dissolved in MilliQ water and precipitated atpH 7.0 as Al(OH)3 by NaOH. Al(OH)3 was isolated by centrifu-gation, washed with MilliQ water five times and then redissolved in10)2 M of L-aspartic, D-aspartic, L-glutamic and D-glutamic acidsseparately. Solutions of 10)3 M Al–amino acid complexes wereobtained, which were preserved in the dark after sterilization. Alcomposition in the Al–amino acid complexes was verified by elec-trothermal atomic absorption spectroscopy and the amino acidcontent by HPLC. The pH values of Al–L-Asp, Al–D-Asp, Al–L-Glu and Al–D-Glu were 4.32, 4.22, 4.32 and 4.28, respectively.The quality of complexation was verified by proton NMR [12].

Aluminium maltolate synthesis

Al maltolate was prepared in our laboratory from AlCl3.6H2O andmaltol (3-hydroxy-2-methyl-4H-pyran-4-one) following the methodof Finneagan et al. [17]. Al maltolate was used in the present studyto understand Al–DNA interactions. This Al complex is hydro-lytically stable from pH 2.0 to 12.0. It enhances free Al existence by60–70% at neutral pH compared to any other inorganic or organicAl complex. Al speciation chemistry is a complex phenomenon,and to overcome this observation Al maltolate was used in thepresent investigation. Al maltolate increases the soluble Al con-centration from 40% to 60% compared to other inorganic salts.[17].

Circular dichroism studies

Supercoiled DNA was titrated against different concentrations ofthe Al–amino acid complexes, Al–L-Asp, Al–D-Asp, Al–L-Glu andAl–D-Glu, as well as Al maltolate (1·10)6 M to 5·10)4 M). TheAl–D-Asp–DNA complexes were titrated against spermine(5·10)6 M to 2·10)4 M), EDTA (4·10)5 M) and desferoximine(4·10)5 M). Circular dichroism (CD) spectra were recorded foreach concentration in 0.1 mM HEPES buffer at pH 7.0. The CDspectra were recorded on a JASCO-715 spectropolarimeter using apath length of 2 mm at 1 nm intervals between 190 and 320 nm.Repetitive scans were used to improve the signal-to-noise ratio andan average of four scans was recorded for each concentration.DNA conformations were deduced by the guidelines of Gray et al.[18] and Hanlon et al. [19].

Agarose gel studies

Integrity of the scDNA was studied using 1% agarose. The samplesscDNA (control), scDNA+Al maltolate, scDNA+Al–L-Asp,scDNA+Al–D-Asp, scDNA+Al–L-Glu and scDNA+Al–D-Gluwere loaded onto the gel and electrophoresed at 5 V/cm at roomtemperature. The DNA concentrations in all lanes were maintainedat 1 lg. The sensitivity of the above complexes to topoisomer sep-aration was carried out using chloroquine, a drug which mimics theactivity of topoisomerase-I. All samples prior to loading onto thegel were incubated with chloroquine for 45 min at room tempera-ture. The gels were stained with ethidium bromide (0.1 lg/mL).

Fluorescence studies

Fluorescence emission studies were carried out using equimolarconcentrations of DNA and EtBr (1:1). The EtBr binding patternsof scDNA, scDNA+Al maltolate, scDNA+Al–L-Asp,scDNA+Al–D-Asp, scDNA+Al–L-Glu, scDNA+Al–D-Glu,scDNA+Al–D-Asp+spermine (2·10)4 M), scDNA+Al–D-As-p+EDTA (4·10)5 M) and scDNA+Al–D-Asp+desferoximine(4·10)5 M) complexes were analyzed. DNA/EtBr solutions wereexcited at 540 nm and emission was recorded at 590 nm using aHitachi F-2000 spectrofluorimeter. The amount of EtBr bound andthe average number of base pairs bound per EtBr molecule in theDNA–Al amino acid complexes were calculated using Scatchard’sequation [20, 21].

Desktop molecular modeling studies

Computer modeling studies were carried out for better under-standing of complexation of the Al–amino acid complexes and alsoto know the spatial arrangements of the molecules when aminoacids complex with Al. The x, y, z atomic coordinates of the L- andD-forms of aspartate and glutamate were transferred from theProtein Databank on to the Desktop Molecular Modeler and thevan der Waals forces were checked. The energy minimization of theL- and D-forms of aspartate and glutamate was done and variousparameters, namely x, y, z coordinates, bond lengths, bond anglesand torsion angles, were obtained. The above parameters wereverified according to Pauling et al. [22]. Similarly, these coordinateswere also derived for the L- and D-forms of complexed Al aspartateand Al glutamate and energy minimization was carried out for thesame. The computer modeling was done using the DesktopMolecular Modeler package, version 1.2 [23].

Results

We have carried out the following biophysical,biochemical and computer modeling studies to

824

understand the role of Al–amino acid complexes inmodulating the helicity of scDNA.

CD studies

It is hoped to understand the role of Al–amino acidcomplexes in modulating DNA topology and we haveeffectively used a CD spectroscopic study to map theDNA helicity changes. The CD spectra of scDNAshowed the B-DNA conformation with a characteristicpositive peak at 275 nm and a negative peak at 245 nm.On addition of increasing concentrations (1·10)6 M to5·10)4 M) of the Al–amino acid complexes Al–L-Asp(Fig. 1), Al–L-Glu (Fig. 2) and Al–D-Glu (Fig. 3) to thescDNA, there was a decrease in the magnitude of boththe positive peak at 275 nm and the negative peak at245 nm. Al maltolate (Fig. 4) did not affect the CDspectra of the scDNA. This indicates a strong binding toDNA without altering the B-DNA conformation.Interestingly in the case of Al–D-Asp (Fig. 5A), the

positive peak at 275 nm decreased with no concomitantchange in the magnitude of the negative peak at 245 nm.This characteristic decrease in the positive peak at275 nm without any change in the negative peak at245 nm reveals the C-DNA conformation. The effect ofAl–D-Asp on the CD spectra shows no clear-cut isos-bestic point, unlike the other complexes. The magni-tudes of the change in the positive peak at 275 nm andnegative peak at 245 nm are plotted with molar ellip-ticity values versus the concentration of the Al–aminoacid complexes, to determine the percent change at bothwavelengths for each concentration. The results indi-cated a consistent decrease in the molar ellipticity of thepositive peak at 275 nm, with no change in the negativepeak at 245 nm for Al–D-Asp and a feature whichreflects the C-DNA conformation (Fig. 5B and Fig. 5C).Maestre and Wang [24] have reported that lowering thesuperhelical density of the scDNA is possibly responsi-ble for decreasing the positive peak at 275 nm with nochange in negative peak at 245 nm.

Fig. 1 Effect of Al–L-Asp on the CD spectra of pUC-19 scDNA in10)4 M HEPES buffer (pH 7.0): a, 1·10)6 M; b, 5·10)5 M; c,3·10)4 M; d, 4·10)4 M; e, 5·10)4 M

Fig. 2 Effect of Al–L-Glu on the CD spectra of scDNA in 10)4 MHEPES (pH 7.0): a, 1·10)6 M; b, 5·10)5 M; c, 3·10)4 M; d,4·10)4 M; e, 5·10)4 M

Fig. 3 Effect of Al–D-Glu on the CD spectra of scDNA in 10)4 MHEPES (pH 7.0): a, 1·10)6 M; b, 5·10)5 M; c, 3·10)4 M; d,4·10)4 M; e, 5·10)4 M

Fig. 4 Effect of Al–maltol on the CD spectra of scDNA in 10)4 MHEPES (pH 7.0): a, 1·10)6 M; b, 5·10)5 M; c, 3·10)4 M; d,4·10)4 M; e, 5·10)4 M

825

In an attempt to understand the relevance of theabove concept, we have studied the role of supercoils bytreating scDNA with chloroquine, which uncoils thescDNA like topoisomerase-I. In the CD spectra there

was no change in the positive or negative peak onuncoiling of scDNA with chloroquine (data not shown).Hence it is clear that the relaxation of supercoils isprobably not associated with the decrease in the positivepeak at 275 nm. Thereby the spectral changes observedin the C-form of DNA are not because of simpleuncoiling but a conformational change from B- toC-DNA caused by Al–D-aspartate.

To understand the role of spermine in modulating theC-DNA conformation, we have added spermine(5·10)6 M to 2·10)4 M) to Al–D-Asp–DNA complexes(C conformation). The CD changes observed were apositive peak ranging from 207 nm to 209 nm and anincrease in magnitude of the negative peak at 275 nm.This is a characteristic feature of the Y-DNA confor-mation (Fig. 6A). Spermine (2·10)4 M) interaction withscDNA did not alter the native B-DNA conformation(Fig. 6B), in accordance with Gosule and Schellman [25,

Fig. 5 A Effect of Al–D-Asp on the circular dichroism spectra ofpUC-19 scDNA in 10)4 M HEPES buffer (pH 7.0). a, 1·10)6 M;b, 5·10)5 M; c, 3·10)4 M; d, 4·10)4 M; e, 5·10)4 M. B Magni-tudinal change in CD spectra of Al–amino acid complexes at275 nm. C Magnitudinal change in CD spectra of Al–amino acidcomplexes at 245 nm

Fig. 6 A Effect of spermine on the CD spectra of scDNA+Al–D-

Asp complex in 10)4 M HEPES (pH 7.0): a, 1·10)6 M; b,5·10)5 M; c, 1·10)4 M; d, 2·10)4 M. B Effect of spermine onthe CD spectra of scDNA in 10)4 M HEPES (pH 7.0): a,5·10)5 M; b, 1·10)4 M; c, 2·10)4 M

826

26]. Further, we examined the role of the metal chelatorsEDTA and desferoximine in reversing the C to B con-formation induced by Al–D-Asp. Treatment with EDTA(4·10)5 M) caused an increase in the positive peak at275 nm but no change in the negative peak at 245 nm,resembling the B-DNA conformation (Fig. 7a) andsuggesting that EDTA reversed the DNA conformationfrom the C to the B form (Fig. 7b). However, withdesferoximine (4·10)5 M) there was no change in thepositive peak at 275 nm and negative peak at 245 nmand a similar spectra to C-DNA (Fig. 7d) was observed,indicating that desferoximine could not reverse the C toB conformation (Fig. 7c).

Agarose gel studies

Gel studies were conducted to monitor the uncoilingprocess of scDNA and topoisomer separation in thepresence of different Al–amino acid complexes. scDNA(1 lg) with Al–amino acid complexes and Al maltolate(5·10)4 M) were subjected to agarose gel electrophoresisto study the DNA uncoiling pattern. The gel profilesshowed that the complexes did not affect the supercoil-ing of DNA (Fig. 8). The sensitivity of the above com-plexes for topoisomer separation was tested usingchloroquine (1 lg/mL), a drug that mimics topoisom-erase-I, and these data provide an idea on the stability ofthese complexes. The electrophoretic band patternclearly showed that the topoisomer separation in the Al–D-Asp–DNA complex is lower (Fig. 9D) compared toother DNA–Al–amino acid complexes (Fig. 9C, E, F).

Fluorescence studies

The EtBr binding pattern to DNA was monitored in thepresence and absence of Al–amino acid complexes. This

study provides information on the EtBr binding withrelevance to DNA conformation. Further, we alsostudied EtBr titration and used Scatchard’s plot tocalculate the amount of EtBr bound per base pair. EtBrfluorescence intensities were recorded for scDNA andthe Al–amino acid complexes (5·10)4 M). The dataclearly indicate that Al–D-Asp significantly decreased thefluorescence intensity (Fig. 10C), whereas the othercomplexes enhanced EtBr fluorescence with DNA(Fig. 10B, D, E). In case of Al–D-Asp, the EtBr bindingto DNA is apparently low due to the C-DNA confor-mation. It has been reported by Hanlon et al. [19] thatthe grooves are shallower in C-DNA because the basesare off-axis, rather narrower compared to the B-formwhere the base stacking is significantly high. Hence theEtBr binding towards C-DNA is apparently low as the

Fig. 7 Effect of EDTA and desferoximine on the CD spectra ofscDNA+Al–D-Asp complex in 10)4 M HEPES (pH 7.0): a,scDNA alone (control); b, 4·10)5 M EDTA; c, 5·10)4 M Al–D-Asp+scDNA; d, 4·10)5 M desferoximine

Fig. 8 Effect of Al–maltol and Al–amino acid complexes on theintegrity of scDNA: A, scDNA (control); B, 5·10)4 M Al–maltol;C, 5·10)4 M Al–L-Asp; D, 5·10)4 M Al–D-Asp; E, 5·10)4 M Al–L-Glu; F, 5·10)4 M Al–D-Glu

Fig. 9 Mapping the sensitivity of Al–amino acid complexes tochloroquine-induced topoisomer separation: A, 1·10)6 g scDNA(control); B, 1·10)6 g scDNA+5·10)4 M Al; C, 1·10)6 gscDNA+5·10)4 M Al–L-Asp; D, 1·10)6 g scDNA+5·10)4 MAl–D-Asp; E, 1·10)6 g scDNA+5·10)4 M Al–L-Glu; F, 1·10)6 gscDNA+5·10)4 M Al–D-Glu

827

EtBr cannot intercalate between the base pairs. How-ever, EtBr binding to other Al–amino acid–DNA com-plexes is greater and may be due to enhanced basestacking. In the case of Al–D-Asp–spermine (2·10)4 M),EtBr fluorescence was apparently very low compared tothe Al–D-Asp–DNA complex, since the DNA hadundergone condensation. The effect on the EtBr fluo-rescence intensity of EDTA (4·10)5 M) to Al–D-Asp–DNA was enhanced, suggesting the reversibility of C- toB-DNA, while desferoximine (4·10)5 M) could not alterthe fluorescence intensity, indicating no reversibility(Fig. 10B). From the Scatchard plot it is clear that theamount of EtBr bound and the average number of basepairs bound per EtBr molecule showed the trend given inTable 1. The results indicate that the average numberof base pairs bound per EtBr molecule is low for the Al–D-Asp–DNA and Al–D-Asp–DNA–spermine complexes.

Desktop molecular modeling studies

Computer modeling studies employing DeskTopMolecular Modeler software, version 1.2 [23], werecarried out to understand the complexation (2:3) of Alto the L- and D-forms of the Al–amino acid complexesand also to map the Al exposure in the complex. Alpreferentially binds to the COO– group of the L- and D-forms of aspartate and glutamate. The findings fromthese modeling studies indicate that Al in the case of theL-forms is embedded inside the complex (Fig. 11A andC). However, in the case of the D-forms the Al is moreexposed on the surface. In addition, with respect to Aland D-Glu, because of the large molecular size of glu-tamate, Al is less exposed and buried inside the complexstructure compared to D-Asp (Fig. 11D). The distancebetween the Al and NH is very close in case of Al–D-Asp,

when compared to other Al–amino acid complexes.Hence there is efficient interaction of Al ion with DNAin case of Al–D-Asp (Fig. 11B).

Discussion

Aluminium has a coordination number of six, so itspreferred geometries are octahedral and Al forms com-plexes with oxygen donor ligands, especially with theCOO) of aspartate and glutamate irrespective of theL- or D-form. When Al forms complexes with aspartateand glutamate, the overall charge of the complexes willbe neutralized. The net neutral charge of Al–Asp andAl–Glu at pH 7.0 is of special interest because it pro-vides a means by which Al may pass through mem-branes, especially the blood/brain barrier [27]. In thepresent study, aspartic acid was used for Al competitivecomplexation, since it contains one carbon less thanglutamic acid and is more acidic (pK1=2.0, pK2=3.9 foraspartic acid and pK1=2.2, pK2=4.3 for glutamic acid).Complexation constants (AH)2 for aspartic acid andglutamic acids taking aluminium hydrolysis into accountare 6.4·1011 and 7.3·1011, respectively [28]. In neuro-degenerative brain, racemized forms of amino acids, inparticular D-Asp and D-Glu, are in relatively largeproportions in core amyloid plaques and neurofibrillarytangles [5]. Latha et al. [12] reported that Al favorsracemization from L-Asp and L-Glu to D-Asp and D-Glu,respectively, in aged rabbit brain. Recent studies showedthat Al in complexation with L-glutamate and L-trans-ferrin were able to cross the blood/brain barrier andaccumulate in selective regions of the brain [11]. Further,it was observed that the Al–amino acid complexes Al–L-Glu and Al–D-Asp could induce conformationalchange from an a-helix to a b-pleated sheet in the Abpeptide. Notably, Al–D-Asp induced greater loss of he-licity and induced more random coil content than free Alin binding with the Ab peptide, suggesting that D-Aspyields more stable Al complexes than the L-form [28].

Based on our results, we propose a mechanism(Fig. 12) that the potential reactive sites for Al in theAl–D-Asp complex found with the DNA molecule are

Fig. 10 EtBr binding pattern of Al–amino acid–DNA complexes:A, 1·10)6 g scDNA (control); B, 1·10)6 g scDNA+5·10)4 M Al–L-Asp; C, 1·10)6 g scDNA+5·10)4 M Al–D-Asp; D, 1·10)6 gscDNA+5·10)4 M Al–L-Glu; E, 1·10)6 g scDNA+5·10)4 M Al–D-Glu; F, 1·10)6 g scDNA+5·10)4 M Al–maltol; G, 1·10)6 gscDNA+4·10)5 M spermine (control); H, 4·10)5 M sper-mine+5·10)4 M Al–D-Asp; I, 4·10)5 M desferoxi-mine+5·10)4 M Al–D-Asp; J, 4·10)5 M EDTA+5·10)4 M Al–D-Asp

Table 1 Data on EtBr intensity on binding to DNA and the effectof Al–amino acid complexes in altering the EtBr fluorescence

Al–amino acid complexes Fluorescencemeasurements

Average no. of basepairs bound perEtBr molecule

Supercoiled DNA alone 0.409 347.7Al–L-aspartate+scDNA 0.567 790.8Al–D-aspartate+scDNA 0.270 140.2Al–L-glutamate+scDNA 0.518 653.4Al–D-glutamate+scDNA 0.495 588.9Al–maltolate+scDNA 0.540 715.1Al–D-aspartate+spermine 0.228 127.4Al–D-aspartate+EDTA 0.494 138.9Al–D-aspartate+desferoximine 0.256 633.8Spermine+DNA 0.396 299.5

828

heterocyclic nitrogen atoms, exocyclic carbonyls on pu-rines and pyrimidine bases and the phosphate backbone.Al in the form of Al–D-Asp binds to DNA, neutralizesrepulsive interactions between the negatively chargedsugar–phosphate backbone and stabilizes interactionbetween the base pairs. This further stabilizes thehydrogen bonding that pairs the two strands. Thephosphate–phosphate repulsion along the backbone andbetween the strands and also the repulsive interactionbetween the phosphate groups of the backbone and thenegative sites on the bases are neutralized by Al and thefree NH3 group of the Al–D-Asp complex. Al interactswith bases and possibly prevents the strands fromcoming close to each other; in the case of C-DNA thebases tilt in such a way that they move further apartfrom the helix axis. In the case of B-DNA the intra- andinterstrand distances between the phosphate groups

is maximal, but when Al–D-Asp interacts with DNA, Albinds to the N7 position of purines and the exocyclic C2

of pyrimidines, which drives the base pairs towards thecenter of the helix in order to maximize the distancebetween the N and O sites on the bases and the phos-phate backbone [19]. Also, binding of the free NH3

group of aspartate to the phosphate backbone leads to asmaller winding angle with greater base pair overlap anda more perpendicular arrangement of base pairs relativeto the helix axis. The net effect is the generation of ahelical structure which conforms to the characteristicsand geometry similar to the B conformation, but wherethere is a significant reduction in electrostatic free en-ergy: this is a characteristic feature of the C-form.Hence, conversion from the native B-form to a slightlycondensed C-form of DNA occurs with Al–D-Asp as thestrain is on both the bases and the phosphate backbone.

Fig. 11 DeskTop MolecularModeller computer modeling ofAl–amino acid complexes. Theasterisks indicates aluminiumexposure in Al–amino acidcomplexes

Fig. 12 A proposed mechanismof binding of Al–D-Asp to theDNA double helical structureand its influence on the DNAconformational pattern from Bto C

829

Further, it was reported that the major grooves andminor grooves of C-DNA are shallower, as the basepairs tilt in such a way that they are situated off the axis[21, 29]. Further, EtBr fluorescence studies support theabove results observed in the case of Al–D-Asp. Inaddition to intercalation, EtBr also binds to phosphategroups. Fluorescence studies on Al–amino acidcomplexes–DNA interactions indicated that Al–D-Aspreduced the fluorescence intensity significantly in com-parison with Al–L-Asp, Al–L-Glu and Al–D-Glu.According to LePecq and Paoletti [30], EtBr at low saltconcentrations shows a second mode of binding whichappears to be the electrostatic binding of the cationic dyeto the negatively charged phosphate groups. In the caseof Al–D-Asp, the displacement of EtBr from the boundPO3) sites by the NH3

+ groups of the amino acids andAl to the N7 position of purines, leading to base stack-ing, might be the reason for the reduction in the EtBrfluorescence emission.

Based on our findings and the literature, we proposea hypothesis on the relevance of Al–D-Asp in modulatingDNA topology. Recent studies in our laboratory evi-dence the presence of left-handed Z-DNA in severelyaffected Alzheimer’s disease brains [31]. This findingleads to a hypothesis that etiological factors like Al andthe Ab peptide are likely to play a pivotal role in the B toZ conformational change in DNA [31]. In the presentstudy we showed that only Al–D-Asp is able to modulatethe conformation of DNA from B to C, while sperminewas able to convert C-DNA to the Y-DNA conforma-tion; Y-DNA is structurally closer to Z-DNA [32]. It isreported that Y-DNA is an ordered, twisted, tightpacking arrangement of the double helix which isstructurally and immunologically related to the Z-DNAfamily [32, 33]. It is left handed in conformation likeZ-DNA and binds strongly to polyclonal anti-Z-anti-body. Thus Al in the form of Al–amino acid complexesmay play a key role in contributing to the complexhelical transitions B–C–Y of DNA and these complexDNA conformations show reductions in electrostaticfree energy and hence they are likely to assume theZ-DNA conformation [31].

Acknowledgements The authors profoundly thank Dr. V. Prakash,Director, Central Food Technological Research Institute, Mysore,for all his support and encouragement. This work was supported bya grant from the Department of Biotechnology, India, and theMinistry of Science, Culture and Sports, Israel.

References

1. Rao KSJ, Katsetos CD, Herman MM, Savory J (1998) ClinLab Med 18:687–698

2. Lukiw WJ, Yasui M, Strong MJ, Ota K, Verity MA (1997)Mineral and metal neurotoxicology. CRC Press, Boca Raton,FL, USA, pp 113–125

3. Fisher GH, Payan IL, Shou-Jian-Chou, Man EH, CerewanskiS (1991) Brain Res Bull 28:127–131

4. Vyas SB, Duffy LK (1995) Biochem Biophys Res Commun206:718–723

5. Shapira R, Austin GE, Mirra SS (1988) J Neurochem 50:649–54

6. Helfman PM, Bada JL (1976) Nature 262:279–2817. Masters PM, Bada JL, Zigler JS Jr (1977) Nature 268:71–738. Roher AE, Lowenson JD, Clarke C, Wolkow C, Wang R

(1993) J Biol Chem 268:3072–30839. Man EH, Sandhouse ME, Burg J, Fisher GH (1980) Science

220:1407–140810. Anderson KK, Perecz GL, Fisher GH, Man EH (1989) Neu-

rosci Res Commun 6:45–4911. Deloncle R, Guillard O, Clanet F (1990) Biol Trace Elem Res

15:1239–124512. Latha KS, Anitha S, Rao KSJ, Bali G, Easwaran KRK (2001)

Alz Rep 4:197–20413. Crapper DR, Quittkat S, Krishnan SS, Dalton AJ, De Boni U

(1980) Acta Neuropathol 50:19–2414. Perl DP, Brody AR (1980) Science 208:297–29915. Bauer WR, Crick FHC, White XX (1980) Sci Am XX:243–24516. Serban D, Benevides JM, Thomas GJ Jr (2002) Biochemistry

41:847–85317. Finneagan MM, Retigg SJ, OrvigC (1986) J Am Chem Soc

108:5033–503518. Gray DM, Taylor TN, Lang D (1978) Biopolymers 17:1–4519. Hanlon S, Chan A, Berman S (1978) Biochim Biophys Acta

519:526–53620. Chatterjee B, Koteswara Rao (1994) Ind J Biochem Biophys

31:77–7921. Suh D, Sheardy RD, Chaires JB (1991) Biochemistry 30:8722–

872622. Pauling L, Corey RB, Branson HR (1951) Proc Natl Acad Sci

USA 37:205–21123. Crabbe MJC, Appleyard JR (1990) DeskTop Molecular

Modeller software, version 1.2. Oxford University Press,Oxford

24. Maestre MF, Wang JC (1971) Biopolymers 10:1021–XXXX25. Gosule LC, Schellman JA (1978) J Mol Biol 121:311–32626. Gosule LC, Schellman JA (1976) Nature 259:333–33527. Martin RB (1986) Clin Chem 32:1797–180628. Deloncle R, Fauconneau B, Piriou A, Huguet F, Guillard O

(2002) Brain Res 946:247–25229. Tunis Schneider MJB, Maestre MF (1970) J Mol Biol 52:521–

54130. LePecq JB, Paoletti C (1967) J Mol Biol 27:87–9231. Anitha S, Latha KS, Rao KSJ, Viswamitra MA (2002) Neuro

Mol Med 2:287–29532. Thomas TJ, Thomas T (1990) J Biomol Struct Dyn 6:1221–

123533. Shin Y A, Eichhorn GL (1984) Biopolymers 23:325–335

830