Nuclear aggregates of polyamines are supramolecular structures that play a crucial role in genomic DNA protection and conformation Luciano D’Agostino 1 , Massimiliano di Pietro 1 and Aldo Di Luccia 2 1 Department of Clinical and Experimental Medicine, ‘Federico II’ University of Naples, Italy 2 Department of Animal Production, University of Bari, Italy Polyamines interact with DNA phosphate groups by means of nonspecific electrostatic bonds [1]. This inter- action has been shown to result in the protection of small DNA molecules from common damaging agents, such as ionizing radiation and reactive oxygen species [2,3]. Polyamines in solution with polynucleotides have also been shown to inhibit the activity of endonucle- ases, including DNase I [4–7]. The protective ability of polyamines is attributable not only to the formation of a steric barrier against DNA-damaging agents, but also to their property to condense the DNA. In fact, polyamines, like other cations, induce DNA condensa- tion as a consequence of the inhibition of > 90% of DNA negative charges [8]. Analogous in vivo experi- ments have demonstrated that spermine and, to a lesser extent, spermidine, prevent DNA fragmenta- tion and the onset of apoptosis. Protection from enzymatic cleavage appears to be the result of a modi- fied chromatin arrangement, rather than inhibition of the endonuclease activity [9]. Condensation of DNA in the presence of poly- amines has also been proposed to be instrumental in genome packaging [10]. This should be regarded of crucial importance if we consider that the total length of cellular DNA is  1 m, whereas the size of the nuc- leus is in the range of several micrometers [11]. How- ever, condensation should not be considered as a static state, as the elasticity is a mechanical property of the DNA, indispensable to cellular processes such as repli- cation and transcription [12,13]. For these reasons the DNA strands in vivo, at the same time, must be pack- aged and protected, but not restrained. The structural impact of polyamines on DNA is also supported by the evidence that these compounds induce, on polynucleotides, a transition from the right- oriented B-form to the left-handed Z-form [14,15]. Such an effect might be important for DNA physiol- ogy, as a tight connection occurs between transcrip- tional activity on DNA and the acquisition of a Z-form [16]. Keywords DNA conformation; DNA protection; molecular aggregates; polyamines; supra- molecular chemistry Correspondence L. D’Agostino, MD, Facolta ` di Medicina ‘Federico II’, Edificio-6, Via S. Pansini, 5, 80131 Naples, Italy Fax: +39 081746 2707 Tel: +39 081746 2707 E-mail: luciano@unina.it (Received 4 April 2005, accepted 19 May 2005) doi:10.1111/j.1742-4658.2005.04782.x In a previous study we showed that natural polyamines interact in the nuclear environment with phosphate groups to form molecular aggregates [nuclear aggregates of polyamines (NAPs)] with estimated molecular mass values of 8000, 4800 and 1000 Da. NAPs were found to interact with genomic DNA, influence its conformation and interfere with the action of nucleases. In the present work, we demonstrated that NAPs protect naked genomic DNA from DNase I, whereas natural polyamines (spermine, sper- midine and putrescine) fail to do so. In the context of DNA protection, NAPs induced noticeable changes in DNA conformation, which were revealed by temperature-dependent modifications of DNA electrophoretic properties. In addition, we presented, for NAPs, a structural model of polyamine aggregation into macropolycyclic compounds. We believe that NAPs are the sole biological forms by which polyamines efficiently protect genomic DNA against DNase I, while maintaining its dynamic structure. Abbreviation NAP, nuclear aggregate of polyamines. FEBS Journal 272 (2005) 3777–3787 ª 2005 FEBS 3777 A body of evidence also indicates a role for poly- amines in the regulation of gene expression [17] and in the cell cycle. A correlation has been shown between increased concentrations of spermine and spemidine in the nucleus and the induction of mitosis [18–20]. Moreover, alterations in the polyamine biosynthetic pathway affect the correct progression of the cell cycle, particularly the S-phase [21]. Temperature is an additional factor capable of affecting DNA conformation. It has been reported that (a) an increase of a few °C is associated with a reduction in the overstretching forces of the DNA strands [22,23], (b) that the DNA melting tempera- ture is directly correlated with the chain length of interacting polyamines [24–27] and (c) that a trans- ition of the DNA chain from a dispersed coil state to a condensed-collapsed state parallels an increase in temperature [28]. These findings draw attention to the relationship existing between temperature and stabilization, aggregation, elasticity and conforma- tional transition of the DNA, all phenomena closely linked to DNA protection and influenced by poly- amines. Recently we described new compounds with mole- cular mass values of  1000, 4800 and 8000 Da, named nuclear aggregates of polyamines (NAPs), whose molecular structure is based on the ionic interaction between polyamines and phosphate groups. These compounds were isolated from the nuclei of several cell types. In vitro aggregation experiments demonstrated that by mixing polyamines (spermine, spermidine and putrescine) in phosphate buffer it is possible to gener- ate compounds with molecular mass identical to the NAPs extracted from cells [29]. This finding suggested that NAPs can form naturally in the nuclear environ- ment, where phosphates are particularly abundant. The positive net charge of NAPs allows interaction with the negatively charged DNA phosphates. We have shown previously that these compounds influence DNA conformation and protect DNA from Exonuc- lease III and DNase I [29]. The compound with a molecular mass of 4800 Da, which was suggested to induce supercoiled DNA forms, is clearly associated with cell replication, being recovered in large quantities in the nuclei of cells in S-phase and absent in non-replicating cells. Experi- mental evidence indicates the NAP with the lowest molecular mass (1000 Da) functions as a precursor for the 4800 Da form of NAP. The concentration of the 8000 Da form of NAP in the nucleus did not vary throughout the phases of the cell cycle. In this study, based on the investigation (by gel elec- trophoresis) of the interaction between polyamines and genomic naked DNA, we compared the protective efficacy of NAPs with that of single polyamines against DNase I. We demonstrated that only NAPs efficiently impede nuclease cleavage, suggesting that they are the biologically effective forms by which poly- amines protect the genomic DNA from endonucleases. We propose a model of molecular organization of polyamines into NAPs extrapolated from our previous and present experimental evidence. NAPs were previously named according to their molecular mass. However, as the molecular mass was only an approximate value, estimated by GPC analysis and therefore different from that calculated by the molecular formulas presented here, we now adopt, in the present study, an alternative nomenclature based on size. Therefore, < 1000 NAP, 4800 NAP and 8000 NAP will subsequently be named s-NAP, m-NAP and l-NAP (small-, medium- and large-sized), respectively. Results and Discussion Single polyamines fail to protect naked genomic DNA from DNase I To assess the protective role of NAPs against endonuc- leases and compare it with that of single polyamines, we carried out electrophoretic assays of genomic DNA treated with DNase I. Single polyamines or NAPs were allowed to interact with genomic DNA before expo- sure to phosphodiesterasic endonuclease. Assuming that the tested compounds, interacting with DNA, pre- vent DNA degradation, the protective effect can be demonstrated by a higher molecular mass of the DNA molecules migrating into the gel. We first tested DNA degradation by DNase I upon incubation with increasing concentrations of single polyamines, starting from a concentration of 1 lm polyamines, which is comparable to the concentration of polyamines forming NAPs in the extractive elution. As shown in Fig. 1A, a 1 lm concentration of poly- amines did not protect from DNase I, and no noticeable increase in DNA protection was observed with 50 or 150 lm single polyamine concentrations. Increasing the concentration of single polyamines up to 600 lm resul- ted in no clear impediment to DNA cleavage (Fig. 1B). At 600 lm spermine, a peculiar effect was observed, namely the paradoxical facilitation of DNase I, result- ing in the complete degradation of DNA (Fig. 1B, lane c). This phenomenon, which probably depends on the electrostatic nature of the polyamines–DNA inter- action, is in accordance with previously published results indicating a biphasic behavior of DNA structure in water solution with increasing concentrations of Polyamine aggregates and DNA L. D’Agostino et al. 3778 FEBS Journal 272 (2005) 3777–3787 ª 2005 FEBS spermine [30]. The authors of this study observed that spermine induces DNA condensation and precipitation as a consequence of the progressive neutralization of the negative charges of DNA. This phenomenon was attributed to the organization of DNA into a liquid- crystalline ordered structure determined by the interac- tion of a single polyamine with several negative DNA charges. However, DNA resolubilization occurred when spermine was supplemented in excess. DNA res- olubilization was attributed to the osmotic stress gen- erated by the high concentration of polyamines, which drives these cations into the nonpolar DNA phase, and to the decreased number of DNA-binding sites per polyamine, making the DNA more hydrophilic. Analogously, in our experimental model, high levels of spermine in solution rendered the genomic DNA extre- mely sensitive to the action of DNase I. We believe that this is a result of the resolubilization of DNA and the massive exposure of the phosphodiester bonds to the active site of the nuclease, probably because of repulsive effects exerted by polyamines present in the nonpolar phase of the DNA on polyamines linked to the backbone phosphates. As single polyamines did protect naked genomic DNA from DNase I, we investigated whether natural aggregates of polyamine, such as NAPs, might possess this ability. As shown in Fig. 2 (left panel), all NAPs protected DNA from degradation, although the migra- tion pattern of DNA preincubated with l-NAP differed from that of DNA preincubated with either m-NAP or s-NAP. As polyamines were not only ineffective in DNA protection, but even detrimental for DNA integrity at higher concentrations, aggregation into NAPs may reflect the need to keep the concentration of intra- nuclear polyamines at low levels and under stringent control. In fact, other studies have demonstrated that an excess of polyamines may result in the perturbation of vital functions dependent on DNA integrity and conformation [31], whereas their drastic decrease under the lower threshold can impede cell mitosis and ⁄ or trigger the mitochondria-mediated apoptotic pathway [32]. In order to accomplish this tight regulation that hampers an excessive increase of polyamines in the A B Fig. 1. Electrophoresis of genomic DNA preincubated with single polyamines and then exposed to DNase I. (A) Electrophoretic migration, at 37 °C, of genomic DNA preincubated with three differ- ent concentrations (1, 50 and 150 l M) of spermine (lanes c, c¢ and c¢¢), spermidine (lanes d, d¢ and d¢¢) or putrescine (lanes e, e¢ and e¢¢) and then exposed to DNase I. Whole genomic DNA (lane a) and DNase I-digested genomic DNA (lane b) were controls. (B) Electrophoretic migration, at 37 °C, of genomic DNA preincubated with 600 l M spermine (lane c), spermidine (lane d) or putrescine (lane e) and exposed to DNase I. Controls were in lanes a (whole genomic DNA) and b (DNA exposed to DNase I). Identical results were obtained at a migration temperature of 40 °C (data not shown). Fig. 2. Nuclear aggregates of polyamines (NAPs) protect genomic DNA from DNase I and, at the same time, influence DNA conforma- tion. The electrophoretic migration at 37 °C (left) and 40 °C (right) of genomic DNA preincubated at 37 °C with small-size NAP (s-NAP; lanes c and c¢), medium-size NAP (m-NAP; lanes d and d¢) or large- size NAP (l-NAP; lanes e and e¢) and exposed to DNase I. Controls were the whole genomic DNA (lane a) and the DNA fully digested by DNase I (lane b). The DNA 1 kb ladder marker was in lane m. L. D’Agostino et al. Polyamine aggregates and DNA FEBS Journal 272 (2005) 3777–3787 ª 2005 FEBS 3779 nucleus, cells are also provided with enzymes that interconvert and catabolize polyamines [33,34]. A key enzyme for polyamine catabolism is diamine oxidase, which is able to bind the DNA and oxidize DNA- bound polyamines [35]. The fact that this enzyme is particularly evident in differentiated enterocytes [36– 38], which are involved in the uptake and distribution of polyamines [39] to the entire organism via the intes- tinal bloodstream, suggests that it probably has the strategic function of coping with the flow of poly- amines potentially harmful to the DNA of intestinal cells. We believe that NAPs are an important part of this physiological scenario. NAPs protect DNA in the context of structural elasticity The data shown in Fig. 2 (left panel) indicated not only that NAPs preserve naked genomic DNA from DNase I-dependent degradation with an efficacy much greater than that of single polyamines, but also that the migration patterns of NAP–DNA complexes differ substantially. Namely, the DNA preincubated with l-NAP showed a diffuse migration pattern, whereas the DNA interacting with s-NAP and m-NAP migrated in a compact form similar to that of naked DNA (lane a), but significantly faster. Hence, we wondered whe- ther such a change in DNA migration properties was caused by a difference in the protection ability of sin- gle NAPs or by conformational changes induced in DNA by the interaction with NAPs. As temperature has been shown to be a variable that dynamically influences DNA conformation in terms of elasticity and condensation status [28], we varied, within physio- logical ranges, the temperature of the electrophoretic run. We believed that this modification of the experi- mental conditions, taking place after inactivation of the endonuclease, could influence migration changes based on DNA conformation, but not on the differen- tial degradation of nucleic acids. By raising the run- ning temperature from 37 °Cto40°C, we observed a mirroring change of the DNA electrophoretic patterns (Fig. 2, right panel). Many aspects of the experiments described in Fig. 2 are worthy of discussion. First, upon incubation with each NAP, DNA, although exposed to the endonuc- lease, showed migration features that, at least in one of the thermal conditions applied, were not dissimilar to that of control DNA. This result undoubtedly dem- onstrates that all NAPs, independently of their molecular mass and net charge, completely protect DNA from DNase I. We believe that the protection occurs as a result of steric hindrance of DNase I access to the DNA phosphodiester bond even though we cannot exclude that modification of the DNA conden- sation status might play an additional role. Theoretic- ally, the possibility exists that the protection of DNA by NAPs depends on modification of the catalytic properties of DNase I, rather than by preventing access of DNase I to the DNA phosphodiester bonds. However, the latter seems to be the likeliest possibility, as the protection of DNA from nucleases is a general property of NAPs, regardless of the type of nuclease tested (NAPs have been shown to prevent exonuclease III – another type of nuclease – degradation of DNA [29]). Furthermore, it has already been suggested that spermine prevents in vivo endonuclease activity as con- sequence of a modified degree of chromatin accessibil- ity to the enzyme [9]. Second, all NAPs increased the electrophoretic speed of genomic DNA, up to induce a diffuse migration pattern, which was determined by each NAP selec- tively at a given temperature (either 37 or 40 °C). As temperature modifications always followed DNase I inactivation, we were able to exclude any interference of this environmental change with the enzymatic activ- ity. Therefore, we concluded that, in the context of constant DNA protection, NAPs interfere with the DNA condensation status in a temperature-dependent manner. We believe that enhancement of the migration speed is attributable to DNA decondensation and strand elongation, which facilitates the penetration of DNA into the gel matrix [40,41]. A third aspect concerns the relationship between DNase I and DNA conformation. Our experiments suggest that the incubation of NAPs–DNA complexes with DNase I was a decisive factor in the above men- tioned conformational effects, as NAPs alone deter- mined only slight modifications in the running properties of DNA [29]. It has already been shown that single polyamines, mainly spermine, can modulate the DNA-binding properties of proteins, either increas- ing or diminishing their affinity for DNA [42]. This ability, to modulate the DNA-binding properties of proteins, was analyzed in relation to the conformational effects of polyamines on DNA and was found to be directly dependent on the degree of their positive charge. Accordingly, we hypothesize that, in the pres- ence of NAPs, DNase I, while prevented from acting as a nuclease as a result of steric inhibition, interacts with NAPs–DNA complexes and, in turn, cooperates with NAPs to modify DNA arrangement. Ionic forces may drive the interaction between DNase I and NAPs–DNA complexes. In fact, under the experi- mental conditions applied (inactivation by EDTA and the electrophoretic run performed at pH 8), the Polyamine aggregates and DNA L. D’Agostino et al. 3780 FEBS Journal 272 (2005) 3777–3787 ª 2005 FEBS enzyme acquires a negative net charge and can there- fore bind the positively charged NAPs. In contrast to polyamines, the effects of NAPs on DNA condensation status do not seem to depend mainly on DNA charge neutralization. By virtue of their net charge, single polyamines have been shown to influence the conformation of high M r DNA [28,41]. Namely, T4 DNA in buffer solution acquired an elon- gated coiled conformation, whereas the progressive neutralization of the global DNA charge by interacting polyamines determined the acquisition of a compact- folded conformation, which showed a slower electro- phoretic mobility [41]. Differently, the modification of the electrostatic properties of DNA induced by NAPs does not appear to be a primary factor driving the change of condensation status, as the different migra- tion patterns were obtained without altering the con- centration of NAPs in the solution. The marked DNA electrophoretic changes indicate, in our opinion, that NAPs efficiently preserve DNA elasticity and can modulate the degree of DNA strand elongation, which is measured by mobility on the gel. These findings sug- gest a possible role for NAPs in the in vivo nuclear environment: NAPs-dependent modification of the DNA condensation status might play a role in the regulation of chromatin complexation onto histones. The elasticity of the DNA was correlated to both its interaction with polyamines [43] and temperature. Melting experiments demonstrated that polyamines stabilize the DNA structure with an ability that is a function of the polyamine chain length [24–26]. More- over, the melting entropy of DNA was determined by measuring the overstretching force of single molecules of DNA [22,23]. This transitional force decreases with the increase of temperature from 11 to 52 °C, thus indicating that the stability of the DNA double helix is a temperature-dependent phenomenon and that DNA melting occurs during the overstretching transition. However, the maintenance of an appropriate func- tional morphology of the DNA seems to require more complex mechanisms. Studies of cation interactions indicate that the size of the DNA grooves depends on the number of charges present on the DNA backbone. In fact, the repulsion of phosphate groups across the minor groove makes it widen, whereas the neutraliza- tion of the phosphate groups reduces the groove width [44,45]. Even though the groove’s flexibility is crucial, its collapse [46] should be considered a detrimental event that can be more efficiently prevented by the interaction with the l-NAPs rather than with the small sized polyamines. Our conviction is strengthened by the fact that the collapse of high M r DNA to toroidal and spheroidal structures has been reported in the presence of multivalent cations, including spermidine and spermine [46]. Recent DNA thermodynamic stud- ies also support this belief, as they indicate that the distension of strands caused by temperature increases widen the grooves [22,23], so permitting the interaction of larger moieties. Therefore, two implications can be inferred from our results (a) the preservation of the DNA integrity is fully assured by NAPs along with the modification of the folding state of the DNA and (b) a few degrees of temperature increase, a normal occurrence in living cells, is able to drive significant conformational chan- ges in the presence of NAPs and DNase I. Both of these events imply that NAPs carry out their defensive function against DNase I having constantly full acces- sibility into the DNA grooves. For all of these reasons, NAPs, compounds with an optimal mass ⁄ charge ratio, represent supramolecular structures able to determine a broader impact on DNA structure and physiology than single polyamines. A model of polyamine aggregation into NAPs As a last step of the present study, we sought to pro- pose a structural model of polyamine organization into NAPs in accordance with the experimental data pro- duced to date about NAPs biochemistry (summarized in Table 1), and the theoretical principles of macro- molecule self-assembly (see the Experimental proce- dures). The NAPs were drawn as macro(poly)cyclic com- pounds on the basis of the following assumptions (a) the attraction of opposite charges of polyamines and phosphate represents the driving force of self- assembly, (b) the intercalation of a phosphate anion between the N-terminal ends of two polyamines per- mits the formation of a cyclical structure character- ized by a minimal repulsive force as a consequence of thermodynamic and kinetic stability and selectiv- ity, (c) ion water solvatation takes part in the supra- molecular aggregation, conferring high flexibility to ionic bonds, (d) amine nitrogen of polyamines, com- pletely protonated at physiological pH, cannot parti- cipate in the formation of hydrogen bonds, whereas phosphate groups are able to form hydrogen bonds, and (e) hydrogen bonds among phosphate groups stabilize adjacent polycyclic units into tridimensional supramolecular structures. The binding of phosphate groups to polyamines to form cyclic supramolecules can overcome the mere phenomenon of attraction of charges to imply the pro- cess of molecular recognition, already described more than 20 years ago [47]. Our previous NMR studies L. D’Agostino et al. Polyamine aggregates and DNA FEBS Journal 272 (2005) 3777–3787 ª 2005 FEBS 3781 showed that phosphate groups which interact with long polyamines (spermidine and spermine), are able to determine molecular rearrangements in their struc- ture [29]. These modifications might include expression of enhanced flexibility on the major axis of poly- amines, which favours their bending, and would be instrumental to the formation of cyclic supra-molecules. A cyclical structure of NAPs has already been sugges- ted in view of their absorbance peak at 280 nm [29], which is compatible with an electron delocalization typical of molecules with p fi p bonds, such as poly- ene systems. Moreover, it has already been postulated that macropoycyclic compounds possess a structure favourable for maximizing and optimizing functional molecular activities. Such molecules are usually large (macro), and may therefore contain central cavities and possess numerous bridges and connections (polycyclic) [47]. Importantly, the formation of polyamine-based macrocyclic compounds has already been described, either as a spontaneous biological event [48], or as a result of in vitro synthetic experiments [49]. The s-NAP, whose molecular mass, extrapolated from the simplest formula, is  1000 Da, is composed of two spermines, one spermidine and one putrescine. As a result, we can predict its structure to be a single tetrameric ring formed by four polyamines linked by four phosphate groups through ionic bonds (Fig. 3A). We have previously shown that, in synchronized Caco- 2 cells stimulated to replicate by gastrin, the diminu- tion of the s-NAP pool was accompanied by increased levels of m-NAP. This observation raised the hypothesis that s-NAPs have the property to aggregate into larger molecules (i.e. m-NAPs) [29]. Further sup- port of this hypothesis came from the detection of compounds with intermediate molecular mass values (ranging from 1000 to 4800 Da) in the above mentioned in vitro aggregation studies [29]. Further- more, previous biochemical studies indicated that m-NAP conserves the same spermine ⁄ spermidine ⁄ putrescine ratio (2 : 1 : 1) of s-NAP. For all of these reasons, and because the molecular mass of m-NAP was estimated to be 4800 Da [29], we proposed its structure to consist of five 4-polyamine monomers con- nected by hydrogen bonds (Fig. 3B). The largest NAP (l-NAP) has a polyamine ratio of 1 : 1 : 1 [29]. In contrast to m-NAP, we did not isolate a pool of monomers for the formation of l-NAP, therefore its structural model results are more specula- tive. However, following the criterion of analogy with the smaller compounds (s-NAP and m-NAP), we pre- dict that its structure may originate from the aggrega- tion, by hydrogen bonds, of five 6-polyamine rings (i.e. two spermines, two spermidines and two putrescines) linked by phosphate groups (Fig. 3C). We predicted 6-polyamine rings in the structure as (a) 3-polyamine rings are likely to require an excessive degree of poly- amine bending and a high energetic level, which would give rise to a less favourable structure, and (b) larger rings (of 9-polyamine units or larger) would be extre- mely large for fitting into the DNA grooves where NAPs are believed to interact. The primum movens in NAPs–DNA aggregation is the charge attraction between DNA phosphates and the amino groups of polyamines. As the amino groups of polyamines are already engaged in ionic bonds with the phosphates of NAPs, secondary amino groups are those available to establish interstrand interaction with the backbone phosphates. In accordance with a recently proposed model of spermine interstrand com- plexation along the major groove, we believe that the interaction of NAPs rings with DNA is then stabilized by intra major groove bonds with DNA bases. They Table 1. Experimental data supporting nuclear aggregates of polyamines (NAPs) modelling. Put, putrescine; Spd, spermidine; Spm, spermine; Ph, phosphate group. s-NAP m-NAP l-NAP Spontaneous aggregation in vitro a Yes Yes Yes Pick of absorbance at 280 nm b Yes Yes Yes Simplest formula c Put-Ph-Spd-Ph-(Spm) 2 Put-Ph-Spd-Ph-(Spm) 2 Put-Ph-Spd-Ph-Spm Estimated molecular mass (Da) d  1000  4800  8000 Calculated molecular mass (Da) e 1035.1 5175.5 9552.15 Number of monomers 1 5 ? (5) g Estimated diameter of each ring f  15 A ˚  15 A ˚  25 A ˚ Compatibility with major groove dimensions Yes Yes Yes Proposed interacting DNA form ? (A-DNA) g Z-DNA ? (B-DNA) g a Gel permeation chromatographic analysis of 25 lM polyamines (Spm, Spd and Put) dissolved in phosphate buffer (pH 7.2). b UV spectro- photometric detection at 280 nm. c Defined on the basis of the molar concentration of polyamines forming NAPs. d Gel permeation chroma- tographic analysis. e On formulae shown in Fig. 3. f Diameters can vary owing to the flexibility of the electrostatic interactions linking polyamines and phosphate groups. g Speculative hypotheses. Polyamine aggregates and DNA L. D’Agostino et al. 3782 FEBS Journal 272 (2005) 3777–3787 ª 2005 FEBS can be either of the hydrophobic type (between the CH 2 group of thymine and the methylene groups of spermine) and ⁄ or the ion-dipole type (between the sec- ondary amino groups of polyamines and purine-N7 or thymine-O4 residues) [1,50]. The insertion of NAP monomers into the DNA grooves forms the basis of the recognition process occurring between the two supramolecular structures (i.e. DNA and NAPs). In our model, the adaptation of both l-NAP and m-NAP to the DNA shape was hypothesized to be favoured by bidirectional movements of the arms of an arch-like structure (Fig. 3B,C). Such an event is made possible by the hydrogen bonds between phosphates belonging to contiguous rings that confer great flexibility to the macropolycyclic structures of NAPs. It could be argued that NAPs, whose structural integrity relies on weak interactions (electrostatic and hydrogen bonds), might disaggregate once in contact with DNA. How- ever the results of electrophoretic experiments allowed us to exclude such a possibility. In fact, the loss of NAPs’ integrity would leave, in solution, single poly- amines (spermine, spermidine, and polyamines), which we showed do not possess any relevant DNA protect- ive activity. Therefore, in order to protect DNA from DNase I, NAPs must maintain their structural integ- rity. Moreover, their ability to influence DNA confor- mation is a further indirect sign that NAPs must be not disaggregated when in contact with DNA. It is also conceivable that NAPs interacting with DNA must stabilize their bond to the double helix by creating a solid structure around it. NAPs would greatly strengthen the alignment along the DNA longi- tudinal axis through the formation of hydrogen bonds between phosphate groups of adjacent molecules (Fig. 4D,E). According to this model, NAPs would thereby form a supra-molecular tunnel capable of enveloping the entire DNA. The final effect would be the formation of an external scaffolding that protects the DNA by masking the sugar-phosphate backbone, as indicated by the evidence of protection against DNase I and exonuclease III [29]. An additional property, namely conformational, might be ascribed to the differences, in terms of size Fig. 3. Structural models of nuclear aggregates of polyamines (NAPs). (A) Small-size NAP (s-NAP). In accordance with the simp- lest formula indicating a spermine (Sm) ⁄ spermidine (Sd) ⁄ putrescine (P) ratio of 2 : 1 : 1, polyamines were represented as terminally linked by phosphate groups to form a single cyclical structure (grey disks, carbon atoms; blue disks, nitrogen atoms; yellow disks, phosphate atoms; red disks, oxygen atoms; white disks, hydrogen atoms). (B) Medium-size NAP (m-NAP). This NAP is represented as a polymer of five s-NAPs linked by hydrogen bonds (green tri- angles). The white arrows indicate the possible opening ⁄ closure movements that allow the adaptation of NAPs to DNA grooves. The closure of the arch (resting state) may occur when the com- pound is in phosphate buffer solution. (C) Large-size NAP (l-NAP). This NAP is represented as a polymer of five 6-polyamine units, linked by hydrogen bonds, according to the simplest formula indica- ting an Sm ⁄ Sd ⁄ P ratio of 1 : 1 : 1. L. D’Agostino et al. Polyamine aggregates and DNA FEBS Journal 272 (2005) 3777–3787 ª 2005 FEBS 3783 and shape, among the single NAPs. Previously, we have shown that m-NAP can enhance the electropho- retic mobility of genomic DNA [29]. Furthermore, we showed by spectrophotometry that m-NAP has the property to increase the absorbance at 260 nm of genomic DNA, whereas other NAPs failed to do so. These results suggest that the m-NAP interacting with DNA might determine structural rearrangements char- acterized by base extrusion, an event occurring in the transition to the left-oriented conformation [16]. For this reason, we predicted a model in which m-NAP favours DNA transition to the Z-form (Fig. 4C), a DNA form characterized by a narrower diameter, more elongated strands and outward exposure of bases. Our previous experiments suggest that the s-NAP works both as a functional NAP, binding the DNA as such, and as a precursor of the m-NAP. Although the aggregation of s-NAPs can occur independently of the DNA structure, the possibility exists that m-NAPs can be built up directly in loco, through the sequential aggregation of three small units, to two s-NAPs already bound to DNA. Thereby, the progressive for- mation of compounds with an increasing number of monomers (up to five) would force, with a type of wedge-like progression (Fig. 4B), the DNA grooves to widen and simultaneously determine the transition towards the Z-form [16], which then proceeds along the two strands in a zip-like manner. Another import- ant consequence of s-NAP complexation into m-NAP is the strong increase of electrostatic forces that the latter compound exerts on DNA. It is known that elec- trostatic forces play a greater role in the A–Z trans- ition than they do in the B–Z transition, because the difference in the linear charges density is greater between the A and the Z forms than between the B and the Z forms [51]. For this reason, we constructed a model in which s-NAP interacts with A-DNA and, among the non-Z fi Z transitional possibilities, we chose the A fi Z possibility (Fig. 4A,C). Additionally, because the A-DNA major groove is narrower than the B-DNA major groove, we hypothesized, by virtue of size compatibility criteria, a preferential interaction of the s-NAP with A-DNA (Fig. 4A). In fact, we eval- uated the diameters of the monomers to be  15 A ˚ for s-NAP and m-NAP and  25 A ˚ for l-NAP. We do not have clear evidence for proposing an interactive model for l-NAP with a specific DNA type. However, as l-NAP is the most widely represented compound in the nuclei of quiescent and replicating cells, and the B-DNA is the most common DNA form [52], a specific role for l-NAP in the protection and conformation of B-DNA can be suggested. However, Fig. 4. Interaction of single nuclear aggre- gates of polyamines (NAPs) with different DNA forms. (A) Small-size NAP (s-NAP) interacting with A-DNA. Grey rings repre- sent the polyamine backbone. Red dots indicate the phosphate groups. To simplify interpretation of the figures, the NAPs phosphates facing the DNA groove were omitted. A-DNA has a groove width more suitable than other DNA forms for interac- tion with this NAP. (B) Progressive forma- tion of medium-size NAP (m-NAP). The addition of s-NAP units to two s-NAPs already bound to DNA (up to five) can allow the formation of m-NAP directly onto the DNA. This may favour the transition to the Z-DNA, through the progressive widening of DNA strands and the exposure of bases. (C) m-NAP interaction with Z-DNA. Z-DNA sta- bilization by the m-NAP arch-like structure was represented as a result of the distan- cing of consecutive A-DNA major grooves. (D) Perspective view of s-NAPs connected by hydrogen bonds. Aggregation of more s-NAPs units can allow the formation of a tunnel-like envelope around the DNA. (E) Perspective view of m-NAPs connected by hydrogen bonds. A 3D m-NAP tunnel structure enveloping the DNA is suggested. Polyamine aggregates and DNA L. D’Agostino et al. 3784 FEBS Journal 272 (2005) 3777–3787 ª 2005 FEBS it should be mentioned that whatever the DNA con- formation, DNA protection is fully assured by each interacting NAP. Studies conducted (based on CD and ⁄ or Raman spectroscopy), to date, on the interaction between polyamines and DNA, analysed DNA molecules much smaller than the genomic DNA used in the present study, testing polyamine ⁄ DNA ratios of 1 : 10 to 1 : 1, which was quite different from those used in our experimental setting (NAP ⁄ DNA ratio of 1 : 5000) [1,50,53]. For this reason, we are convinced that these methodological approaches cannot be easily extrapola- ted to the NAPs setting. Therefore, much work must be carried out to clarify several aspects of our model. However, a recent review from Medina and coworkers stated that ‘despite the great amount of experimental and theoretical works carried out up to now, it is not possible to give an undoubted explanation about how the polyamines bind to DNA’ [54]. In this context, we believe that our work, although not fully exhaustive in all the aspects approached, might shed novel light on the matter. In conclusion, we demonstrated that NAPs are able to preserve the genomic DNA from DNase I-dependent degradation, with an efficacy extraordinarily greater than single polyamines. Furthermore, we showed that DNA, while preserved in its integrity by single NAPs, undergoes temperature-sensitive conformational chan- ges, which are indicative of a preserved DNA elasticity. We believe that NAPs are the sole biologically active forms by which polyamines physiologically interact with and protect genomic DNA, given that these quasi-stable molecular aggregates, natural examples of supramolecu- lar chemistry, are able to reach the maximum effect with the minimum effort. Experimental procedures Human genomic DNA was isolated from peripheral blood leukocytes donated by M. di Pietro. DNA was extracted and purified in phenol ⁄ chloroform and then resuspended in Tris ⁄ EDTA (TE) buffer. NAPs were extracted from the nuclei of preconfluent Caco-2 cells and purified by gel per- meation chromatography, as previously described [29]. Genomic DNA (4 lg per 2.5 lL of phosphate buffer) was incubated for 6 min at 37 °C with 4.5 lL of l-, m- or s-NAP (mean polyamine concentration: 0.25 ngÆlg )1 DNA) or water solutions of single polyamines (putrescine, sper- midine and spermine) at a concentration of 1, 50, 150 or 600 lm. The mixture was then exposed to DNase I (RQ1RNase-free DNase; Promega, Milan, Italy) at a con- centration of 0.025 UÆlg )1 DNA. Briefly, 1 lL of the DNase I solution was added to 1 lL of the reaction buffer solution (400 mm Tris ⁄ HCl, pH 8, 100 mm MgSO 4 and 10 mm CaCl 2 ) and then mixed with NAP–DNA or poly- amine–DNA solutions. The enzyme action was stopped after 30 min of incubation at 37 °C by adding 1 lLof 20 mm EDTA, pH 8. Samples were then loaded onto a 1Æ5% (w ⁄ v) ultrapure DNA grade agarose gel. Electrophor- esis of DNA was carried out for 1 h in an HE 100 supersub (Amersham Pharmacia Biotech, Uppsala, Sweden), at a constant temperature of 37 or 40 °C (controlled by a peri- staltic pump system), by applying an electric field strength of 11.1 VÆcm )1 in Tris ⁄ borate ⁄ EDTA buffer. Each gel was then photographed by using a Polaroid MP-4 L camera. NAPs modelling was carried out, producing molecular structures that were in strict accordance with biochemical data and theoretical rationales. All biochemical data were collected from analytical, elec- trophoretic and NMR studies shown in our previously pub- lished work [29] and in the present study. Theoretical rationales were derived from the universally accepted prin- ciples of the supramolecular chemistry based on the self- assembly by means of weak interactions [47,55–57]. A multistep aggregation plan was developed as follows (a) a cyclic and planar structure representing the single base module, (b) base modules combined in polycyclic planar structures (ladder), (c) the polycyclic modules developed 3D into tunnel-like structures. The compatibility between the NAPs dimensions, evaluated on formulas, and the currently accepted DNA grooves size was the requisite for the inter- action scheme of NAPs with DNA forms. Acknowledgements We are grateful to Dr Alma Contegiacomo and Dr Luigi Gomez-Paloma for helpful suggestions and advice, to Mr Paolo Mastranzo for technical assistance and to Mr Stefano D’Agostino for his contribution in artwork production. This work was supported by a research grant from the Campania Region. References 1 Ouameur AA & Tajmir-Riahi HA (2004) Structural analysis of DNA interactions with biogenic polyamines and cobalt (III) hexamine studied by Fourier transform infrared and capillary electrophoresis. J Biol Chem 279, 42041–42054. 2 Spotheim-Maurizot M, Ruiz S, Sabattier R & Charlier M (1995) Radioprotection of DNA by polyamines. Int J Radiat Biol 68, 571–577. 3 Khan AU, Di Mascio P, Medeiros MH & Wilson T (1992) Spermine and spermidine protection of plasmid DNA against single-strand breaks induced by singlet oxygen. Proc Natl Acad Sci USA 89, 11428–11430. L. D’Agostino et al. Polyamine aggregates and DNA FEBS Journal 272 (2005) 3777–3787 ª 2005 FEBS 3785 4 Pingoud A, Urbanke C, Alves J, Ehbrecht HJ, Zabeau M & Gualerzi C (1984) Effect of polyamines and basic proteins on cleavage of DNA by restriction endonucleases. Biochemistry 23, 5697–5703. 5 Kuosmanen M & Poso H (1985) Inhibition of the activ- ity of restriction endonucleases by spermidine and sper- mine. FEBS Lett 179, 17–20. 6 Baeza I, Gariglio P, Rangel LM, Chavez P, Cervantes L, Arguello C, Wong C & Montanez C (1987) Electron microscopy and biochemical properties of polyamine- compacted DNA. Biochemistry 26, 6387–6392. 7 Oana H, Tsumoto K, Yoshikawa Y & Yoshikawa K (2002) Folding transition of large DNA completely inhi- bits the action of a restriction endonuclease as revealed by single-chain observation. FEBS Lett 530, 143–146. 8 Wilson RW & Bloomfield VA (1979) Counterion- induced condensation of deoxyribonucleic acid. A light- scattering study. Biochemistry 18, 2192–2196. 9 Brune B, Hartzell P, Nicotera P & Orrenius S (1991) Spermine prevents endonuclease activation and apopto- sis in thymocytes. Exp Cell Res 195, 323–329. 10 Saminathan M, Thomas T, Shirahata A, Pillai KS & Thomas TJ (2002) Polyamine structural effects on the induction and stabilization of liquid crystalline DNA, potential applications to DNA packaging, gene therapy and polyamine therapeutics. Nucleic Acids Res 30, 3722–3731. 11 Watson D, Hopkins NH, Roberts JW, Steitz JA & Weiner AM (1987) Molecular Biology of the Gene. Benjamin ⁄ Cummings, Menlo Park. 12 Zhou H, Zhang Y & Ou-Yang Z (2000) Elastic property of single double-stranded DNA molecules: Theoretical study and comparison with experiments. Phys Rev E Stat Phys Plasmas Fluids Relat Interdiscip Topics 62, 1045–1058. 13 Bustamante C, Bryant Z & Smith SB (2003) Ten years of tension: single-molecule DNA mechanics. Nature 421, 423–427. 14 Basu HS, Schwietert HC, Feuerstein BG & Marton LJ (1990) Effects of variation in the structure of spermine on the association with DNA and the induction of DNA conformational changes. Biochem J 269, 329–334. 15 Ruiz-Chica J, Medina MA, Sanchez-Jimenez F & Ramirez FJ (2001) Raman study of the interaction between polyamines and a GC oligonucleotide. Biochem Biophys Res Commun 285, 437–446. 16 Rich A & Zhang S (2003) Timeline: Z-DNA: the long road to biological function. Nat Rev Genet 4, 566–572. 17 Childs AC, Mehta DJ & Gerner EW (2003) Polyamine- dependent gene expression. Cell Mol Life Sci 60, 1394– 1406. 18 Tabor CW & Tabor H (1984) Polyamines. Annu Rev Biochem 53, 749–790. 19 Pegg AE (1988) Polyamine metabolism and its import- ance in neoplastic growth and as a target of chemother- apy. Cancer Res 48, 759–774. 20 Janne J, Alhonen L & Leinonen P (1991) Polyamines: from molecular biology to clinical application. Ann Med 23, 241–259. 21 Oredsson SM (2003) Polyamine dependence of normal cell-cycle progression. Biochem Soc Trans 31, 366–370. 22 Williams MC, Wenner JR, Rouzina I & Bloomfield VA (2001) Entropy and heat capacity of DNA melting from temperature dependence of single molecule stretching. Biophys J 80, 1932–1939. 23 Williams MC, Rouzina I & Bloomfield VA (2002) Ther- modynamics of DNA interactions from single molecule stretching experiments. Acc Chem Res 35, 159–166. 24 Musso M, Thomas T, Shirahata A, Sigal LH, Van Dyke MW & Thomas TJ (1997) Effects of chain length modification and bis (ethyl) substitution of spermine analogs on purine-purine-pyrimidine triplex DNA stabi- lization, aggregation, and conformational transitions. Biochemistry 36, 1441–1449. 25 Antony T, Thomas T, Shirahata A & Thomas TJ (1999) Selectivity of polyamines on the stability of RNA-DNA hybrids containing phosphodiester and phosphorothio- ate oligodeoxyribonucleotides. Biochemistry 38, 10775– 10784. 26 Antony T, Musso M, Hosseini MW, Brand G, Green- field NJ, Thomas T, Van Dyke MW & Thomas TJ (1999) Differential effects of cyclopolyamines on the stability and conformation of triplex DNA. Antisense Nucleic Acids Drug Dev 9, 13–23. 27 Hou MH, Lin SB, Yuann JM, Lin WC, Wang AH & Kan Ls L (2001) Effects of polyamines on the thermal stability and formation kinetics of DNA duplexes with abnormal structure. Nucleic Acids Res 29, 5121–5128. 28 Murayama H & Yoshikawa K (1999) Thermodynamics of the collapsing phase transition in a single duplex DNA molecule. J Phys Chem B 103, 10517–10523. 29 D’Agostino L & Di Luccia A (2002) Polyamines inter- act with DNA as molecular aggregates. Eur J Biochem 269, 4317–4325. 30 Saminathan M, Antony T, Shirahata A, Sigal LH, Thomas T & Thomas TJ (1999) Ionic and structural specificity effects of natural and synthetic polyamines on the aggregation and resolubilization of single-, double- and triple-stranded DNA. Biochemistry 38, 3821–3830. 31 Seiler N, Duranton B, Gosse ` F & Raul F (2000) Sper- mine cytotoxicity to human colon carcinoma-derived cells (CaCo-2). Cell Biol Toxicol 16, 117–130. 32 Nitta T, Igarashi K & Yamamoto N (2002) Polyamine depletion induces apoptosis through mitochondria- mediated pathway. Exp Cell Res 276, 120–128. 33 Seiler N, Delcros JG & Moulinoux JP (1996) Polyamine transport in mammalian cells, an update. Int J Biochem Cell Biol 28, 843–861. 34 Thomas T & Thomas TJ (2001) Polyamines in cell growth and cell death: molecular mechanisms and thera- peutic applications. Cell Mol Life Sci 58, 244–258. Polyamine aggregates and DNA L. D’Agostino et al. 3786 FEBS Journal 272 (2005) 3777–3787 ª 2005 FEBS [...]... D’Agostino et al 35 Bruun L, Hlgdall EV, Vuust J & Houen G (1998) Polyamine-stimulated binding of diamine oxidase to DNA Acta Chem Scand 52, 921–929 36 D’Agostino L, D’Argenio G, Ciacci C, Daniele B, Macchia V & Mazzacca G (1984) Diamine oxidase in rat small bowel, distribution in different segments and cellular location Enzyme 31, 217–220 37 D’Agostino L, Daniele B, Pignata S, Gentile R, Contegiacomo... Contegiacomo A, Tagliaferri P, Silvestro G, Bianco AR & Mazzacca G (1989) Ornythine decarboxylase and diamine oxidase in human colon carcinoma cell line Caco-2 in culture Gastroenterology 97, 888–894 38 Daniele B & D’Agostino L (1994) Proliferation and differentiation of the small intestinal epithelium, from Petri dish to bedside Ital J Gastroenterol 26, 459– 470 39 D’Agostino L, Pignata S, Daniele B, D’Adamo... 2823–2832 42 Panagiotidis CA, Artandi S, Calame K & Silverstein SJ (1995) Polyamines alter sequence-specific DNA protein interactions Nucleic Acids Res 23, 1800–1809 43 Murayama Y, Sakamaki Y & Sano M (2003) Elastic response of single DNA molecules exhibits a reentrant collapsing transition Phys Rev Lett 90, 018102 44 Hamelberg D, Williams LD & Wilson WD (2001) In uence of the dynamic positions of cations... defensive macrocyclic polyamines Science 281, 428–431 Frydman B, Bhattacharya S, Sarkar A, Drandarov K, Chesnov S, Guggisberg A, Popaj K, Sergeyev S, Yurdakul A, Hesse M et al (2004) Macrocyclic polyamines deplete cellular ATP levels and inhibit cell growth in human prostate cancer cells J Med Chem 47, 1051–1059 Ruiz-Chica J, Medina MA, Sanchez-Jimenez F & Ramyrez FJ (2001) Fourier transform Raman study of. .. B, D’Adamo G, Ferraro C, Silvestro G, Tagliaferro P, Contegiacomo A, Gentile R, Bianco AR et al (1990) Polyamine uptake by human colon carcinoma cell line Caco-2 Digestion 46, 325–359 40 Bloomfield VA (1997) DNA condensation by multivalent cations Biopolymers 44, 269–282 41 Yamasaki Y, Teramoto Y & Yoshilawa K (2001) Disappearance of the negative charge in giant DNA with a folding transition Biophys... (2000) Structural basis of polyamine -DNA recognition: spermidine and spermine interactions with genomic B-DNAs of different GC content probed by Raman spectroscopy Nucleic Acids Res 28, 3379–3385 Medina MA, Urdiales JL, Rodriguez-Caso C, Ramirez FJ & Sanchez-Jimenez F (2003) Biogenic amines and polyamines: similar biochemistry for different physiological missions and biomedical applications Crit Rev... (1987) Supramolecular Chemistry – scope and perspectives molecules – supermolecules – and molecular FEBS Journal 272 (2005) 3777–3787 ª 2005 FEBS Polyamine aggregates and DNA 48 49 50 51 52 53 54 55 56 57 devices Nobel lecture, Stockholm, World Scientific Publishing Co., Singapore Schroder FC, Farmer JJ, Attygalle AB, Smedley SR, Eisner T & Meinwald J (1998) Combinatorial chemistry in insects: a library of. .. Sherrington DC & Taskinen KA (2001) Self-assembly in synthetic macromolecular systems via multiple hydrogen bonding interactions Chem Soc Rev 30, 83–93 Greig LM & Philp D (2001) Applying biological principles to the assembly and selection of synthetic superstructure Chem Soc Rev 30, 287–302 Bowes KF, Ferguson G, Lough AJ, Zakaria CM & Glidewell C (2003) Phosphonoacetic acid as a building block in supramolecular. .. cations on the structure of the DNA minor groove, sequence-dependent effects J Am Chem Soc 123, 7745–7755 45 Hamelberg D, Williams LD & Wilson WD (2002) Effect of a neutralized phosphate backbone on the minor groove of B -DNA: molecular dynamics simulation studies Nucleic Acids Res 30, 3615–3623 46 Thomas TJ & Bloomfield VA (1983) Collapse of DNA caused by trivalent cations, pH and ionic specificity effects... structural specificities on the interaction between DNA and biogenic polyamines Biophys J 80, 443–454 Wu HY & Behe MJ (1984) Salt-induced Z -A- Z transition sequence in the mixed ribo-deoxyribo copolymer poly (rG-dC) X poly (rG-dC) Proc Natl Acad Sci USA 81, 7284–7287 Wolfe SL (1993) Molecular and Cellular Biology Wadsworth Publishing Co., Belmont, CA Deng H, Bloomfield VA, Benevides JM & Thomas GJ Jr . Nuclear aggregates of polyamines are supramolecular structures that play a crucial role in genomic DNA protection and conformation Luciano D’Agostino 1 ,. elec- trophoresis) of the interaction between polyamines and genomic naked DNA, we compared the protective efficacy of NAPs with that of single polyamines against DNase