DOI: 10.1002/chem.200902029Role of Environmental Factors on the Structure and Spectroscopic Responseof 5’-DNA–Porphyrin Co...
IntroductionDNA is a very attractive scaffold for precise arrangement ofchromophores and fluorophores on the nanoscale.[1–...
signate profile of the CD spectra in the porphyrin Soretregion, together with the fact that the intercept with the xaxis (...
at 422.7 nm (+67.8mÀ1cmÀ1). Variable temperature CD dataare presented in Figure 4. The bisignate exciton couplet col-lapse...
to-center inter-porphyrin distance ranges 32–36 Š, while thedihedral angle (q=15,5,5’,15’), relevant for porphyrin–por-phy...
porphyrin transitions is the most effective. Within this ap-proximation, the CD in the Soret region may be thought tobe we...
sembly of the end-to-end stacked duplex (see Figures S1–S16).The complexity of the CD signal observed at À58C in theporphy...
this case an increase in ionic strength initiates partial por-phyrin–porphyrin inter-duplex aggregation as a water mole-cu...
The observed experimental outcomes may be rationalizedupon assuming the formation of a rod-like p–p stackedZnPor-8mer inte...
This approach results in quali-tative agreement with more so-phisticated methods, even forp-stacked porphyrindimers.[67,88...
substantial energy separation between the two (Bx andBy) components.[88,89]This is, however, not observed inthe case of te...
region (trisignate or tetrasignate CD curves). The origin ofthe observed multisignate CD profile is the strong electronici...
[15] S. Bahatyrova, R. N. Frese, C. A. Siebert, J. D. Olsen, K. O. van derWerf, R. van Grondelle, R. A. Niederman, P. A. B...
[83] G. Bringmann, D. C. G. Gçtz, T. A. M. Gulder, T. H. Gehrke, T.Bruhn, T. Kupfer, K. Radacki, H. Braunschweig, A. Heckm...
Upcoming SlideShare
Loading in …5
×

Role of Environmental Factors on the Structure and Spectroscopic Response

262 views

Published on

We have explored the utility,
strength, and limitation of throughspace
exciton-coupled circular dichroism
in determination of the secondary
structure of optically active chromophoric
nanoarrays using the example of
end-capped porphyrin– and metalloporphyrin–
oligodeoxynucleotide conjugates.
We put special emphasis on the
explanation of the origin and significance
of the distinctive multiple bands
in the CD spectra (trisignate and tetrasignate
CD bands). Such CD profiles
are often observed in chiral aggregates
or multichromophoric arrays but have
never before been studied in detail. We
found that variation of temperature
and ionic strength has a profound
effect on the geometry of the porphyrin–
DNA conjugates and thus the
nature of electronic interactions. At
lower temperatures and in the absence
of NaCl all three 5’-DNA–porphyrin
conjugates display negative bisignate
CD exciton couplets of variable intensity
in the Soret region resulting from
through-space interaction between the
electric transition dipole moments of
the two end-capped porphyrins. As the
temperature is raised these exciton
couplets are transformed into single
positive bands originating from the
porphyrin–single-strand DNA interactions.
At higher ionic strengths and low
temperatures, multisignate CD bands
are observed in the porphyrin Soret
region. These CD signature bands originate
from a combination of intermolecular,
end-to-end porphyrin–porphyrin
stacking between duplexes and porphyrin–
DNA interactions. The intermolecular
aggregation was confirmed
by fluorescence and absorption spectroscopy
and resonance light scattering.
DeVoe theoretical CD calculations, in
conjunction with molecular dynamics
simulations and Monte Carlo conformational
searches, were used to mimic
the observed bisignate exciton-coupled
CD spectra as well as multiple CD
bands. Calculations correctly predicted
the sign and shape of the experimentally
observed CD spectra. These studies
reveal that the exciton-coupled circular
dichroism is a very useful technique for
the determination of the structure of
optically active arrays.

Published in: Technology
0 Comments
1 Like
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
262
On SlideShare
0
From Embeds
0
Number of Embeds
2
Actions
Shares
0
Downloads
2
Comments
0
Likes
1
Embeds 0
No embeds

No notes for slide

Role of Environmental Factors on the Structure and Spectroscopic Response

  1. 1. DOI: 10.1002/chem.200902029Role of Environmental Factors on the Structure and Spectroscopic Responseof 5’-DNA–Porphyrin Conjugates Caused by Changes in the Porphyrin–Porphyrin InteractionsAngela Mammana,[a]Gennaro Pescitelli,[c]Tomohiro Asakawa,[a]Steffen Jockusch,[a]Ana G. Petrovic,[a]Regina R. Monaco,[a]Roberto Purrello,[d]Nicholas J. Turro,[a]Koji Nakanishi,[a]George A. Ellestad,*[a]Milan Balaz,*[b]and Nina Berova*[a]Abstract: We have explored the utility,strength, and limitation of through-space exciton-coupled circular dichro-ism in determination of the secondarystructure of optically active chromo-phoric nanoarrays using the example ofend-capped porphyrin– and metallo-porphyrin–oligodeoxynucleotide conju-gates. We put special emphasis on theexplanation of the origin and signifi-cance of the distinctive multiple bandsin the CD spectra (trisignate and tetra-signate CD bands). Such CD profilesare often observed in chiral aggregatesor multichromophoric arrays but havenever before been studied in detail. Wefound that variation of temperatureand ionic strength has a profoundeffect on the geometry of the porphy-rin–DNA conjugates and thus thenature of electronic interactions. Atlower temperatures and in the absenceof NaCl all three 5’-DNA–porphyrinconjugates display negative bisignateCD exciton couplets of variable inten-sity in the Soret region resulting fromthrough-space interaction between theelectric transition dipole moments ofthe two end-capped porphyrins. As thetemperature is raised these excitoncouplets are transformed into singlepositive bands originating from theporphyrin–single-strand DNA interac-tions. At higher ionic strengths and lowtemperatures, multisignate CD bandsare observed in the porphyrin Soretregion. These CD signature bands orig-inate from a combination of intermo-lecular, end-to-end porphyrin–porphy-rin stacking between duplexes and por-phyrin–DNA interactions. The inter-molecular aggregation was confirmedby fluorescence and absorption spec-troscopy and resonance light scattering.DeVoe theoretical CD calculations, inconjunction with molecular dynamicssimulations and Monte Carlo confor-mational searches, were used to mimicthe observed bisignate exciton-coupledCD spectra as well as multiple CDbands. Calculations correctly predictedthe sign and shape of the experimental-ly observed CD spectra. These studiesreveal that the exciton-coupled circulardichroism is a very useful technique forthe determination of the structure ofoptically active arrays.Keywords: circular dichroism ·DNA aggregation · helical struc-tures · molecular modeling ·porphyrinoids[a] Dr. A. Mammana, Dr. T. Asakawa, Dr. S. Jockusch,Dr. A. G. Petrovic, Dr. R. R. Monaco, Prof. N. J. Turro,Prof. K. Nakanishi, Dr. G. A. Ellestad, Prof. N. BerovaDepartment of Chemistry, Columbia University3000 Broadway, New York, NY 10027 (USA)Fax: (+1)212-932-8273E-mail: ndb1@columbia.edugae2104@columbia.edu[b] Prof. M. BalazDepartment of Chemistry, University of Wyoming1000 E. University Avenue, Laramie, WY 82071 (USA)Fax: (+1)307-766-2807E-mail: mbalaz@uwyo.edu[c] Dr. G. PescitelliDipartimento di Chimica e Chimica IndustrialeUniversita degli Studi di Pisa, 56126 Pisa (Italy)[d] Prof. R. PurrelloDepartment of Chemical Sciences, University of CataniaViale Andrea Doria 6, 95125 Catania (Italy)Supporting information for this article is available on the WWWunder http://dx.doi.org/10.1002/chem.200902029.Chem. Eur. J. 2009, 15, 11853 – 11866 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim 11853FULL PAPER
  2. 2. IntroductionDNA is a very attractive scaffold for precise arrangement ofchromophores and fluorophores on the nanoscale.[1–7]Chro-mophore and fluorophore arrays often exhibit differentphysical and chemical properties from monomers fromwhich they are assembled.[1–4,8,9]When chromophores comein close proximity of each other their transition dipole mo-ments may interact upon photoexcitation of one of them,and the excitation energy becomes delocalized over thechromophores forming an exciton.[10–13]Nature takes advant-age of this phenomenon in numerous systems, the most im-portant being photosystems I and II.[14–16]When the interact-ing chromophores are oriented to form a chiral array, theexciton transitions can be detected by circular dichroism(CD) spectroscopy, with spectra typically displaying a char-acteristic bisignate signal (overlap of positive and negativeCotton effects).[17]The sign of this bisignate CD curve canbe correlated with the absolute orientation of the interactingchromophores, according to the exciton chirality method.[18]Exciton coupling has found widespread use as a diagnostictool in the field of chiral supramolecular assemblies[1–7,11,19–25]as well for the determination of absolute configuration ofchiral organic molecules.[26–29]Exciton coupled circular dichroism associated with bispor-phyrin arrays makes porphyrins ideal chromophores for CDstructural studies.[30]A porphyrin exciton couplet occurs inthe visible spectral region around 420 nm,[31,32]it has mini-mal overlap with CD signals of other chromophores and isobservable over a large distance (up to 66 Š).[33–38]Chiralsupramolecular assemblies of porphyrins, mediated by pstacking or ligand-to-metal binding, often display strong anddiagnostic CD Soret signals.[39–59]In our present work we ex-plore the utility, strengths, and limitations of the through-space exciton-coupled circular dichroism for the elucidationof the secondary structure of optically active chromophoricnanoarrays. Previously, we studied end-capped porphyrin–DNA conjugates joined through 3’-amide and more flexible5’-phosphate linkers.[35–38,60,61]In an extension of these stud-ies we have now examined structural and spectroscopicproperties of 5’-amide end-capped free base, copper andzinc porphyrin–oligonucleotide conjugates so as not to dis-turb the conventional 3’–5’ DNA synthesis. In this study weput special emphasis on the origin and interpretation of thedistinctive multiple bands in the CD spectra (trisignate andtetrasignate CD curves) of porphyrin– and metalloporphyr-in–DNA conjugates induced by an increase in ionic strength.Such CD profiles have sometimes been observed in chiralmultichromophoric nanoassemblies[39,62–65]but have neverbefore been studied in detail. The use of porphyrin end-capped DNA sequences permitted the exploration of both,a) the through-space CD exciton coupling between two por-phyrins separated by a DNA duplex, as well as b) the elec-tronic coupling between two p-stacked porphyrins from ag-gregated DNA duplexes by simply varying the ionic strengthand the metallation of the porphyrin unit (free base, zinc orcopper). We rationalize the influence and the contributionof both types of porphyrin–porphyrin couplings (intra- andinter-duplex) on the experimentally observed CD spectra.Molecular mechanics (MM), molecular dynamics (MD) sim-ulations and theoretical DeVoe CD calculations, along withfluorescence spectroscopy, and resonance light scattering(RLS) have been used to support and explain the unusualchanges in the CD profile within the Soret region in re-sponse to the environmentally induced structural changes.Results and DiscussionThe structures of porphyrin–DNA conjugates 2HPor-8mer,CuPor-8mer and ZnPor-8mer studied in this manuscript aredepicted in Figure 1. Their synthesis has been reported pre-viously.[66]Their UV/Vis absorption characteristics are sum-marized in Table 1.Free base porphyrin–DNA conjugate (2HPor-8mer) withoutNaCl: The circular dichroism spectrum below 300 nm of2HPor-8mer recorded in 50 mm potassium phosphate buffer(pH 7.0) at À58C in the absence of NaCl revealed spectralfeatures of right handed B-DNA (see Figure S1, SupportingInformation). The CD spectrum in the Soret band absorp-tion region at À58C (Figure 2a, blue line) displayed a strongnegative exciton couplet with a positive band at 415.5 nm(De=++32.0mÀ1cmÀ1) and negative band at 430.1 nm (De=À19.0mÀ1cmÀ1). In the case of porphyrins, the Soret band isdue to two mutually orthogonal and quasi-degenerate transi-tions (Bx and By).[31]However, if unrestricted libration of theporphyrin ring is possible around a preferential axis, corre-sponding to the bond linking the porphyrin to the rest ofmolecule, an effective transition dipole moment directedalong such axis, for example, 5–15 (Figure 1), can efficientlydescribe the coupling between two B transitions.[67]The bi-Table 1. UV/Vis absorption characteristics of 2HPor–8mer, ZnPor–8mer,and CuPor–8mer.Por–DNA Soret band [nm] (e [mÀ1cmÀ1])[a]Q bands [nm]2HPor–8mer 422.0 (230600) 516, 554, 585, 639ZnPor–8mer 425.0 (351000) 557, 594CuPor–8mer 418.0 (251500) 541[a] e = extinction coefficients.Figure 1. Sequence and schematic representation of 5’-DNA–porphyrinand 5’-DNA–metalloporphyrin conjugates.www.chemeurj.org 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim Chem. Eur. J. 2009, 15, 11853 – 1186611854
  3. 3. signate profile of the CD spectra in the porphyrin Soretregion, together with the fact that the intercept with the xaxis (424 nm) coincides with the maximum in the UV spec-trum (422 nm), reveals that the observed CD signal is due tolong distance through-space, electric dipole–dipole, excitoncoupling of the two porphyrin chromophores. A negative ex-citon CD couplet originates from the negative twist betweeninteracting effective transition moments of two porphyr-ins.[17]Thermal DNA melting experiments were carried out tostudy the influence of the DNA strand separation on theporphyrin–porphyrin through-space interactions. The melt-ing of the duplex 2HPor-8mer (as evidenced by the decreaseof the negative CD band at 255 nm, Tm =48.38C, Figure S2,Supporting Information), revealed a radical change in theshape and intensities of the CD bands in the Soret region(Figure 2a). Upon heating from À5 to 408C the bisignateCD signal observed at À58C is transformed into a positiveband (green line, 420 nm, De=++27.3mÀ1cmÀ1). At tempera-tures higher than Tm (508C) a very small positive CDband (425 nm) is the only feature that can be observed inthe Soret band region of the 2HPor-8mer. Induced positiveCD bands in the Soret region have been previously linkedto a porphyrin–single-stranded DNA interaction.[36]The CDmelting curve recorded at 415 nm shows a sigmoidal profile(Figure 2b) with a melting temperature of 45.38C. The nega-tive CD Cotton effect of the exciton couplet (430 nm) ob-served at À58C disappeared completely above 458C, a tem-perature very close to the Tm of 2HPor-8mer.Although a direct comparison with our previously pub-lished results[35–38,60,61]on spectroscopic properties of freebase porphyrin–DNA conjugates is not possible due to theuse of a combination of a different linker (phosphate), a dif-ferent attachment position (3’), and a different porphyrin(bisphenyl–bispyridyl caboxylate vs phenyl–trispyridyl car-boxylate) the following conclusions can be drawn: a) por-phyrins attached to a short oligonucleotide via a 5’-phos-phate linker give rise to a weak positive exciton couplet CDsignal; b) porphyrins attached to the short oligonucleotidevia a 5’- or 3’-amide linker give rise to a medium to strongexciton couplet CD signal; c) porphyrins attached at the 5’-position give rise to a positive CD band at temperatureshigher than Tm for both, amide and phosphate linker; d)porphyrins attached via a phosphate linker tend to destabi-lize DNA sequences while porphyrins attached via amidelinker stabilize DNA sequences.Copper porphyrin–DNA conjugate without NaCl: The CDsignal in the Soret region displayed a non-conservative bisig-nate curve (Figure 3a) with a strong positive Cotton effectat 413 nm (De=++32.8mÀ1cmÀ1) and a weak negative Cottoneffect at 425 nm (De=À9.7mÀ1cmÀ1) similar to that of the2HPor-8mer. Heating the sample from À5 to 408C causedthe transformation of the bisignate CD signal into a singlepositive band (416.7 nm, De=++43.3mÀ1cmÀ1). Further heat-ing to 808C was accompanied with a significant decrease ofthe CD intensity resulting in a small positive CD band at420 nm. The CD melting curve recorded at 414 nm showeda “hill-like” profile with maximum at 158C (Figure 3b). Theweak negative CD band of the exciton couplet (423 nm) ob-served at À58C disappeared above 108C.Zinc porphyrin–DNA conjugate without NaCl: Zinc por-phyrins have been shown to have a water molecule axiallycoordinated to the central atom in aqueous solutions. Thiscoordination causes the zinc to deviate from planarity (0.19–0.22 Š above the N4 plane), increases its hydrophilicity anddefines its binding modes with DNA.[69]Axial ligandshamper intercalation into the DNA and the outside minorgroove binding is the major DNA binding mode. The CDspectrum of the ZnPor-8mer in the absence of NaCl isshown in Figure 4. At À58C the through-space exciton cou-pling between the two porphyrin chromophores appendedat the DNA termini is the origin of the very intense longrange (43 Š, see Figure 8) exciton CD couplet with a nega-tive band at 431.9 nm (À78.0mÀ1cmÀ1) and a positive bandFigure 2. Variable temperature CD spectra of 2HPor-8mer (c=2.5 mm)from À5 to 808C.[68]Figure 3. Variable temperature CD spectra of CuPor-8mer (c=3.6 mm)from À5 to 808C.[68]Chem. Eur. J. 2009, 15, 11853 – 11866 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim www.chemeurj.org 11855FULL PAPERDNA–Porphyrin Conjugates
  4. 4. at 422.7 nm (+67.8mÀ1cmÀ1). Variable temperature CD dataare presented in Figure 4. The bisignate exciton couplet col-lapsed completely above 608C. A sigmoidal melting profileat 422 and 432 nm (Figure 4b) is evidence for strong cooper-ativity of the melting, most likely related to changes in theporphyrin-porphyrin through-space interaction.Influence of the porphyrin metallation on the CD profile(no NaCl): The exciton couplet of the ZnPor-8mer has thesame sign as the CuPor-8mer and 2HPor-8mer confirmingtheir common origin but it is remarkably stronger in intensi-ty (Figure 5). It is well known that the sign and intensity ofthe exciton couplet are strongly dependent on the molecularextinction coefficients emax, interchromophoric distance R12as well as chiral twist between the interacting chromo-phores.[17]As all three porphyrins are attached at the sameposition in the DNA sequence using the same linker we canapproximately assume a similar inter-porphyrinic distance.From Table 1 we know that the extinction coefficient of Zn–porphyrin is 40 and 52% stronger than the Cu–porphyrinand 2H-porphyrin, respectively. Even if couplet amplitudes(A=DemaxÀDemin) are proportional to emax2, the differentvalues of emax can not entirely account for the difference ob-served between the CD couplets of the three conjugates. Itis likely that conformations with a more favorable twist be-tween the two Zn–porphyrins contribute to the stronger ex-citon couplet as also suggested by the conformational studydescribed below.Molecular dynamics and conformational analysis: theoreti-cal explanation of observed exciton-coupled CD: Our effortin finding the prevailing conformation(s) of 5’-porphyrinend-capped DNA duplexes has involved the combination oftwo computational approaches: molecular dynamics (MD)simulations and Monte Carlo conformational search (MC).These simulations were essential due to our inability toobtain any useful NMR structural data for these conjugatesbecause of the aggregation at the necessary NMR concen-trations. More specifically, we were interested in exploringthe rotational freedom of the porphyrins appended at the 5’-DNA termini. All simulations were carried out using theSchrçdinger suite of programs:[70]MD was carried out usingthe Desmond module,[71,72]and MC conformational searchwith the MacroModel v9.5 module. Dynamics simulationswere run for 2 ns. MC conformational search was run underthe “enhanced search” setting, with the maximum numberof steps set to 10000. For both MD and MC initial structureshave been constructed with B-form duplex DNA. This formhas been preserved during the simulations by implementingharmonic constraints on the GC base pairs. For further de-tails on MD and MC calculations see Supporting Informa-tion.MD simulations on the 2HPor-8mer duplex showed thatone of the porphyrins remains capped (overlapping the ter-minal A–T plane) throughout the 2 ns simulation, while theother porphyrin is more mobile. More precisely, when theinitial conformation of the conjugate is fully capped (FigureS13A), within ~800 ps of the start of the simulation one ofthe porphyrins in the conjugate clearly uncaps (does notshow any significant paralleling with the plane of the A–Tpair) and remains more mobile than the other. The secondporphyrin displays, relatively, less mobility and remainsmainly stacked over the A–T base pair at a distance of ap-proximately 4 to 5 Š. MD has sampled 200 conformationsalong the 2 ns dynamics trajectory at a constant interval of100 ps. The most energetically stable snapshot conformation,labeled 2HPor-8mer-MD1, is shown in Figure 6a. A repre-sentative semicapped structure 2HPor-8mer-MD2 is dis-played in Figure 6b (semicapped means one porphyrin iscapped, while the other one uncapped). It was determinedby first calculating the average structure of all 200 confor-mations obtained from the molecular dynamics trajectory.Next, the resulting average structure was compared with thepool of 200 conformers and one of them with the lowestroot-mean-square deviation (RMSD) from the calculatedaverage chosen as the representative structure. Both 2HPor-8mer-MD1 and 2HPor-8mer-MD2 display preserved A–Thydrogen bonding at one of the DNA termini. The center-Figure 4. Variable temperature CD spectra of ZnPor-8mer (c=3.9 mm)from À5 to 808C.[68]Figure 5. Soret region CD spectra of 2HPor-8mer (red dotted curve),CuPor-8mer (green dashed curve), and ZnPor-8mer (blue curve) at a) 0and b) 408C.[68]www.chemeurj.org 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim Chem. Eur. J. 2009, 15, 11853 – 1186611856M. Balaz, N. Berova, G. A. Ellestad et al.
  5. 5. to-center inter-porphyrin distance ranges 32–36 Š, while thedihedral angle (q=15,5,5’,15’), relevant for porphyrin–por-phyrin coupling[67]is positive (q=++388 and +568 for confor-mations MD1 and MD2, respectively). It is important tonote that while a fully capped starting structure of 2HPor-8mer (Figure S13A) preserves porphyrin capping at least atone end of the conjugate, starting from a fully uncappedstructure (Figure S13C) leads mostly to uncapped conforma-tions. Nevertheless, these fully uncapped structures are~280 kJmolÀ1energetically less stable than those involvingcapping.The MC conformational search on fully-capped and fully-uncapped starting structures of the 2HPor-8mer supportsthe MD data since it results in a preferred semicapped con-formation. In other words, neither MD nor MC runs provideevidence for a fully-uncapped conformation for 2HPor-8mer. However, MD simulations on the ZnPor-8mer duplexdisplay significantly different dynamic behavior. Both of theporphyrins are clearly equally mobile over the course of a2 ns simulation. The initially fully-capped conformation(Figure S13B) evolves throughout this simulation such thatfirst one of the porphyrins uncaps within ~200 ps. Withinthe next ~500 ps the second one also uncaps and is seen tobe as equally mobile as the first porphyrin. Throughout thebalance of this simulation, all conformations show both por-phyrins remain uncapped. The initially fully-uncapped con-formation (Figure S13D) remains uncapped. Importantly,unlike the non-metallated conjugate, where only one por-phyrin is predominantly mobile, in the zinc case both termi-nal porphyrins are equally mobile.The MC conformational search on ZnPor-8mer supportsthe MD based observations, as the entire lowest energy(most stable) pool of 200 conformations shows that bothtethered porphyrins in the conjugate remain uncapped withrespect to the terminal A–T base pairs. The representativelow energy structure from the MC conformational search isdisplayed in Figure 7, labeled as ZnPor-8mer-MC. A higherdegree of conformational freedom of porphyrins leads todisruptions of the terminal A–T hydrogen bonding. This ob-served A–T disruption, however, is much smaller than in thecase of non-metallated conjugate. Additional conformation-al details of ZnPor-8mer-MC (Figure 8) are the interchomo-phoric distance of 43 Š and a negative dihedral angle (q=À1298). This preferred negative interporphyrinic twistindeed is in agreement with the experimentally observednegative CD exciton couplet (Figure 4a).DeVoe CD calculations of MD and MC/CS structures: CDcalculations within the coupled oscillator framework(DeVoe method[73,74]were run on some representative struc-tures obtained by molecular mechanics calculations de-scribed above. The aim of the present CD calculations wasnot to reproduce the whole CD spectra of porphyrin–DNAconjugates. Rather, we restricted our analysis to the Soretregion only, and considered only the strongest and red-shift-ed transitions from the DNA bases, whose coupling withFigure 6. Most energetically stable capped 2HPor-8mer-MD1 (a) and therepresentative semicapped 2HPor-8mer-MD2 (b) conformations of2HPor-8mer conjugate resulting from 2 ns MD run. Distances betweenporphyrin and A–T planes are indicated.Figure 7. Lowest energy structure (ZnPor-8mer-MC) from the conforma-tional search of the ZnPor-8mer conjugate.Figure 8. MC lowest energy structure (ZnPor-8mer-MC) showing inter-chromophoric (Zn-Zn) distance and dihedral angle associated with theporphyrin chromophores.Chem. Eur. J. 2009, 15, 11853 – 11866 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim www.chemeurj.org 11857FULL PAPERDNA–Porphyrin Conjugates
  6. 6. porphyrin transitions is the most effective. Within this ap-proximation, the CD in the Soret region may be thought tobe well predicted, at least qualitatively, while that in theDNA region is clearly not, as it will require a full descrip-tion of all relevant transitions.[75–78]Free-base and zinc-metallated porphyrin DNA conjugates,2HPor-8mer-MD2 and ZnPor-8mer-MC, were considered asrepresentative of porphyrin–DNA duplexes (Figures 6 and7). From the viewpoint of exciton CD calculations, the dif-ferent capping versus uncapping between the two structuresis especially relevant. For the uncapped ZnPor-8mer-MCstructure, calculated CD spectra show symmetric CD cou-plets in the Soret region (Figure 9a) due to degenerate exci-ton coupling between the two circular oscillators describingthe porphyrin chromophores. The calculated CD couplet isquite intense (A % À120mÀ1cmÀ1) despite the very largedistance (R12 % 43 Š). The non-degenerate exciton couplingbetween the porphyrins and the DNA bases is instead quiteweak. The calculated CD spectrum reproduces very well themain features (shape, sign and magnitude) of the experi-mental one for ZnPor-8mer (Figure 4a, blue line) and is typ-ical of a B-DNA form with two uncapped porphyrins. Yet, amore precise simulation of the CD spectrum of porphyrinDNA duplexes would necessarily involve a thorough confor-mational sampling and CD calculations on several differentstructures, which is beyond the scope of this study. As forthe semicapped 2HPor-8mer-MD2 structure, a weak sym-metrical CD couplet is again obtained when only the degen-erate porphyrin-porphyrin coupling is allowed in the calcula-tions. However, the non-degenerate coupling with the DNAbases is much stronger and leads to an overall monosignateCD band (Figure 9b). The strongest non-degenerate cou-plings are those between each porphyrin and the two closestA–T base pairs, onto which the former is stacked. With thecautions stressed above, it may be noticed that the calculat-ed CD reproduces the dominant positive CD band at ca.420 nm of 2HPor-8mer at intermediate temperatures (with-out NaCl, Figure 2a, green line).In both cases just discussed, the inter-duplex interactionsare not taken into account, and the calculated CD in theSoret region of porphyrin DNA duplexes appear as a moreor less symmetric monosignate band or couplet. More com-plicated features, multiple bands or couplets with sidebands,were not observed in our calculations. Moreover, theyreveal that the representative fully uncapped structureZnPor-8mer-MC allows for a much stronger porphyrin-por-phyrin exciton coupling than the semicapped structure2HPor-8mer-MD2, and that the contribution of porphyrin-base coupling is much stronger for a capped structure thanfor an uncapped structure.Porphyrin-porphyrin aggregation promoted by increasedionic strengthFree base 2HPor-8mer with 150mm NaCl: To promote por-phyrin–DNA and porphyrin–porphyrin interduplex interac-tions and to study the effect of increased ionic strength onthe structure and spectroscopic response of 2HPor-8mer, werepeated the variable temperature experiments in the pres-ence of 150 mm NaCl. We anticipated that an increase inionic strength would reinforce the hydrophobic interactionsbetween nucleobases, stimulate tighter interactions betweencovalently linked porphyrin and neighboring base pairs, butalso promote porphyrin–porphyrin interduplex interactions.Figure 10b shows the Soret region CD spectra of a meltingexperiment of 2HPor-8mer with 150 mm NaCl. A dramaticchange, in comparison to the spectrum without NaCl, is de-tected in the porphyrin Soret band region upon heating. Amultisignate (three Cotton effects, +/À/+ from the red,440, 427 and 415 nm) CD curve is clearly observed at À58C(Figure 10b, blue line). We have ascribed the origin of thesemultiple bands to interduplex interactions between the end-capped porphyrins coming from two different DNA duplex-es (for more detailed discussion see below). The melting ofthe duplex 2HPor-8mer in the presence of 150 mm NaCl re-veals a transition, which despite the obvious complexity ofits CD curve, appears similar to that observed in the ab-sence of NaCl at lower temperatures. Upon heating fromÀ5 to 408C the multisignate CD collapses to a single posi-tive band. The CD melting curve recorded at 440 nm (posi-tive CD band) showed a sigmoidal profile (melting tempera-ture of 26.08C), indicative of cooperativity during the disas-Figure 9. DeVoe-calculated CD spectra on structures a) ZnPor-8mer-MCand b) 2HPor-8mer-MD2.Figure 10. Variable temperature CD spectra of 2HPor-8mer (c=2.5 mm)from À5 to 808C in a) the absence of NaCl (see Figure 2a) and b) thepresence of 150 mm NaCl.www.chemeurj.org 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim Chem. Eur. J. 2009, 15, 11853 – 1186611858M. Balaz, N. Berova, G. A. Ellestad et al.
  7. 7. sembly of the end-to-end stacked duplex (see Figures S1–S16).The complexity of the CD signal observed at À58C in theporphyrin Soret band region cannot be explained accordingto a simple exciton coupling between two transition dipoles.The multisignate (+/À/+) CD curve is strongly indicativeof more complex phenomena than the simple through-spacelong-range exciton coupling between two porphyrin chromo-phores. We have previously reported that a bisignate profilewith distinctive side-bands (“tetrasignate” CD signal) is theresult of p–p interactions between two closely placed por-phyrins appended to a chiral scaffold with restricted confor-mational flexibility.[67]In such a case the common applica-tion of the exciton chirality method as an interaction be-tween two transition dipoles (in classical terms, linear oscil-lators) is no longer justified, and the porphyrin Soret transi-tion must be depicted through a circular oscillator, that is, acombination two mutually orthogonal degenerate or quasi-degenerate dipoles.[67]Our results clearly indicate that underconditions of increased ionic strength close interduplexhead-to-tail interactions between 5’-appended free base por-phyrins take place. This results in restricted rotation be-tween interacting porphyrins giving rise to a multisignateCD curve in the Soret region. A similar chromophore-pro-moted DNA aggregation has been recently reported by Wa-genknecht et al. in a perylene-end-capped DNA duplex (noCD studies)[79]and by Lewis et al. in perylenediimide-linkedDNA hairpins.[80]Copper porphyrin–DNA conjugate with 150mm NaCl: Avery distinctive multisignate (+/À/+/À) CD signal is ob-served in the porphyrin Soret band region of the CuPor-8mer at À58C with two positive Cotton effects at 438 and412 nm and two negative at 425 and 400 nm (Figure 11b).The observed CD profile supports also, in this case, thechiral head-to-tail inter-duplex porphyrin–porphyrin interac-tion. The variable temperature CD experiments of CuPor-8mer in the presence of 150 mm of NaCl revealed a dramaticchange in the shape and intensity of the Soret band CD pro-file as shown in Figure 11b. The multisignate CD profile ob-served at À58C (blue curve) collapsed to a single positiveband at 418 nm at 408C (green line). The sigmoidal shape ofthe melting curves within the Soret region points to the co-operativity of the disassembly of the porphyrin stacked ag-gregates (Figure S6), similar to the 2HPor-8mer.In order to determine the effect of concentration on ag-gregation, we carried out a series of dilution experimentsusing CuPor-8mer as an example where changes in the CDin the Soret region were monitored as a function of concen-tration (73 to 0.69 mm), constant ionic strength (150 mmNaCl) and constant absorbance by adjusting the path lengthof the cuvette. As expected the aggregation is clearly con-centration-dependent (see Figure S5) with a multisignateprofile observed at 73 mm and a monosignate positive signalat 0.69 mm.Zinc porphyrin-DNA conjugate with 150mm NaCl: In-creased ionic strength has also a clear effect on the structureand stability of the ZnPor-8mer as evidenced again by amultisignate (+/À/+) CD curve in the porphyrin Soretband region where an additional positive band to the origi-nal negative couplet is clearly detectable (Figure 12b). InFigure 11. Variable temperature CD spectra of CuPor-8mer (c=3.6 mm)from À5 to 808C in a) the absence of NaCl (see Figure 3a) and b) thepresence of 150 mm NaCl.Figure 12. Variable temperature CD spectra of ZnPor-8mer from À5 to 808C in a) the absence of NaCl (see Figure 4a) and b) the presence of 150 mmNaCl, and c) 450 mm NaCl.Chem. Eur. J. 2009, 15, 11853 – 11866 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim www.chemeurj.org 11859FULL PAPERDNA–Porphyrin Conjugates
  8. 8. this case an increase in ionic strength initiates partial por-phyrin–porphyrin inter-duplex aggregation as a water mole-cule axially bound to the zinc atom hinders the porphyrininter-duplex interaction. As seen in Figure S7, the meltingprofile of the porphyrin–DNA conjugate ZnPor-8mer in thepresence of 150 mm of NaCl revealed that the extra bandobserved at low temperature disappeared above 308C andonly a clear bisignate CD remained visible due to porphy-rin–porphyrin intra-duplex exciton coupling. We anticipatedthat the higher ionic strength of 450 mm NaCl would pro-mote even stronger inter-DNA duplex aggregation due toporphyrin p–p stacking. As expected a very characteristicmultisignate CD curve (+/À/+, Figure 12c, blue line) wasobserved within the porphyrin Soret band region at À58C.Fluorescence and absorptionspectroscopy: Fluorescenceand absorption spectroscopyprovided additional supportfor the presence of aggregatesat high salt concentrations.The absorbance spectra ofZnPor-8mer in buffer solutionshow a strong dependence onthe ionic strength. At low ionicstrength ([NaCl]=0 to150 mm) the expected absorb-ance spectrum of the Soretband of ZnPor is observed(Figure 13b, red and blue spec-tra). At a high ionic strength([NaCl]=450 mm) and 108Cthe Soret band at 425 nm is de-creased and a new hypso-chromically shifted band ap-pears at 413 nm (Figure 13a,red curve) indicates H-type ag-gregation.[81]Upon denaturingat elevated temperatures(708C) single stranded ZnPor-8mer is generated, whichshows the typical Soret bandabsorption of ZnPor, inde-pendent of the ionic strength (blue curve).The strong dependence of the absorbance under highionic strength is also observed by fluorescence spectroscopy.Excitation into the Soret band of non-stacked ZnPor(424 nm) generates a fluorescence with maxima at 604 and657 nm (Figure 14). In the absence of NaCl, only minorchanges in fluorescence intensities are observed upon dena-turing of ZnPor-8mer (Figure 14, dashed spectra). At highionic strength ([NaCl]=450 mm) and 108C the fluorescenceintensity decreased by a factor of ~5 (red solid line). Thefluorescence excitation spectrum of this emission revealedthat this remaining fluorescence originated from excitationof non-stacked ZnPor (Figure S9). Further support for thisassignment is provided by fluorescence lifetime studies.Fluorescence decay kinetics after pulsed excitation at424 nm are almost identical at high and low ionic strength(Figure S10). At high ionic strength ([NaCl]=450 mm) anadditional minor fluorescence component with a lifetime of0.1 ns was observed which contributed less than 10% tothe overall fluorescence signal. This short-lived minor com-ponent is probably caused by fluorescence of p–p stackedZn-Por. The deconvoluted fluorescence spectrum of theweakly fluorescing p–p stacked ZnPor is shown in FigureS11, which is strongest at an excitation wavelength of413 nm. In the case of a single porphyrin per duplex such asnon-self-complementary ZnPor-T(GC)3C and its comple-mentary 5’-G(CG)3A sequence, the aggregation is much lesspronounced (Figure S12).Figure 13. Absorbance spectra of ZnPor-8mer in K-phosphate buffer(50 mm) with a) 450 mm NaCl and b) 150 mm NaCl at 108C (red) and708C (blue).Figure 14. Fluorescence spectra (lex =424 nm) of ZnPor-8mer (4 mm) in K-phosphate buffer (50 mm), withoutNaCl (dashed lines) and 450 mm NaCl (solid lines) at 10 (red) and 708C (blue). a) single stranded ZnPor-8mer, b) ZnPor-8mer duplex, c) representation of the ZnPor-8mer aggregate. A non-stacked Zn-Por is repre-sented as a purple box, p–p stacked as a cyan box.www.chemeurj.org 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim Chem. Eur. J. 2009, 15, 11853 – 1186611860M. Balaz, N. Berova, G. A. Ellestad et al.
  9. 9. The observed experimental outcomes may be rationalizedupon assuming the formation of a rod-like p–p stackedZnPor-8mer interduplex aggregate containing two non-stacked ZnPor at each end (Figure 14c). Because the ob-served weak fluorescence at high ionic strength ([NaCl]=450 mm) after excitation at 425 nm is dominated by fluores-cence of the peripheral non-stacked ZnPor (Figure 14c,solid red line), this fluorescence intensity can be used to es-timate the size of the p–p stacked ZnPor-8mer assembly.An increase of fluorescence intensity by a factor of ~5 is ob-served upon denaturation of the duplex at elevated temper-atures (708C) (fluorescence of single stranded ZnPor-8mer,Figure 14a). This fluorescence increase is consistent with anaggregate of approximately five p–p stacked ZnPor-8merduplexes reaching the size of approximately 150 Š and con-taining ten ZnPor-8mer units as represented in Figure 14c.The effect of temperature and oligonucleotide hybridizationon the fluorescence intensity is negligible as demonstratedby the comparison of the fluorescence intensities in the ab-sence of NaCl (Figure 14a, dashed lines). We postulate thatthe p–p stacked ZnPor-8mer assemblies show a rod-likestructure and not closed loops. The formation of closedloops, which would also result in a significant fluorescencereduction (because all ZnPors will be stacked), is entropical-ly disfavored (higher degree of organization). It is also en-thalpically unlikely (disruption of the continuous p stack ofnucleobases and porphyrins and distortion in the DNA skel-eton and thus distortion of the CD profile below 300 nm).Experimentally, no such change in the CD profile was ob-served in the UV spectral region under conditions favoringaggregation.Aggregation of ZnPor-8mer: Resonance light scattering(RLS): Resonance light scattering is a useful technique tostudy electronically coupled chromophore arrays where en-hanced light scattering of several orders of magnitudes canbe observed at the absorption bands of the aggregates.[82]RLS experiments can be conveniently performed in a stan-dard steady-state fluorescence spectrometer with right angleconfiguration. Figure 15 shows the RLS spectra of ZnPor-8mer at different NaCl concentrations. At low ionic strength([NaCl]=150 mm) and in the absence of NaCl, only negligi-ble RLS was observed in the Soret band spectral region. In-terestingly, at 150 mm NaCl no evidence of aggregation wasobserved in the fluorescence spectra (see Figures S9 andS11). In contrast, CD did provide evidence of aggregation asshown by the appearance of multisignate bands (Fig-ure 12b). However, at high ionic strength ([NaCl]=450 mm)a strong RLS signal was observed (Figure 15, red solid line).This enhanced light scattering at high ionic strength is con-sistent with the formation of p–p stacked ZnPor-8mer ag-gregates as shown in Figure 14c. Upon denaturing at elevat-ed temperatures (708C) single stranded ZnPor-8mer aregenerated, which show no RLS (dashed spectra, Figure 15).Theoretical circular dichroism analysis: A series of CD cal-culations with the coupled oscillator model was run to inter-pret and explain the origin of the distinctive multiple bandsin the CD spectra of aggregates of porphyrin–DNA conju-gates promoted by the addition of salt. Similar distinctivefeatures of multiple CD bands in the Soret region had beenobserved in the case of covalent bis-porphyrin derivativeswhere strongly interacting porphyrins undergo energeticallyfavored p–p stacking,[67]or where the porphyrins are directlylinked.[83–85]Similarly, for porphyrin aggregates in which theporphyrins are tethered to a chiral template,[36]or monomer-ic porphyrins are aggregated on a chiral template,[86,87]anomalous CD profiles in the Soret region were observedunder conditions favoring strong porphyrin aggregationmediated by p–p stacking. In the current case of 5’-porphy-rin–DNA conjugates, multiple CD Soret bands are again ob-served as a consequence of DNA inter-duplex aggregation,formed through porphyrin p–p stacking, as proved by vari-ous spectral evidences such as concentration-dependent CD,UV/Vis absorption, fluorescence and RLS. The collection ofpresent and literature data let us conclude that the occur-rence of strongly interacting porphyrins is a necessary butnot sufficient condition for the appearance of multiple CDSoret bands. Chiral porphyrin stacking does not necessarilylead to multiple CD Soret bands, however, the latter havealways been associated with very strong p–p stacking (orother short-distance interactions). A theoretical demonstra-tion of the above statement is offered below.DeVoe CD calculations on porphyrin dimers: CD calcula-tions on aggregates between porphyrin–DNA conjugate du-plexes such as those sketched in Figure 14 would be impossi-ble due to the system complexity and lack of safe structuraldata. Therefore, suitable models needed to be chosen. Ourcalculation strategy consisted of three sets of DeVoe calcula-tions involving various kinds of p-stacked porphyrin dimers.Despite the known limitations of the coupled-dipole approx-imation (especially at short interchromophoric distances)used in the DeVoe-type calculations employed here, this ap-proach is known to reproduce with sufficient accuracy theexciton-coupled CD spectra of bis-porphyrinic compounds.Figure 15. Resonance light scattering of ZnPor-8mer (4 mm) in K-phos-phate buffer (50 mm), without NaCl (black), 150 mm NaCl (blue) and450 mm NaCl (red) at 108C (solid line) and 708C (dashed line).Chem. Eur. J. 2009, 15, 11853 – 11866 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim www.chemeurj.org 11861FULL PAPERDNA–Porphyrin Conjugates
  10. 10. This approach results in quali-tative agreement with more so-phisticated methods, even forp-stacked porphyrindimers.[67,88,89]DeVoe calcula-tions may directly take into ac-count the circular oscillatorpicture,[30,34,67,90,91]where thetwo mutually orthogonal com-ponents (Bx and By) of theSoret transition are consideredexactly degenerate. Details onDeVoe calculations are givenin the Supporting Information.The simplest possible situa-tion of a dimer composed oftwo ideal, D4h-symmetric andperfectly parallel porphyrins,described as circular oscilla-tors, defines a non-chiral struc-ture devoid of CD.[92]There-fore, we envisaged three possible limiting situations of CD-active p–p stacked porphyrin dimers: a) the quasi-parallelstacking between two porphyrins distorted from planarity;b) the non-parallel stacking between two perfectly planarporphyrins; c) the parallel stacking between two perfectlyplanar porphyrins in the presence of additional (extrinsic)transition dipoles close to the porphyrins.a) Distorted porphyrin dimers: A set of plausible geome-tries for p-stacked porphyrin dimers was obtained byusing MMFF geometry optimizations on dimers ofcopper(II) 5-phenyl-10,15,20-tris(2-pyridyl)-porphyrin(CuPor). This compound was chosen as a model of theporphyrin moieties found in the CuPor-8mer conjugate.Thanks to the presence of regions with complementarycharge (positive on the metal core, negative on the pyri-dine nitrogens, Figure S14), face-to-face stacking be-tween two CuPor nuclei is strongly favored. MMFF-opti-mized dimers have geometries in keeping with thosecommonly observed for p–p stacked porphyrins,[93,94]with 4.1 Š mean plane-to-plane distance and 4.3 Š later-al shift (horizontal offset). Although the optimized struc-ture for the monomer CuPor is achiral and its porphyrincore almost planar, the opposite is true for the fourdimers, where the porphyrin cores are heavily distortedfrom planarity. This leads to structures endowed withvariable chirality reflected in the heterogeneity of thecalculated CD couplets. Four distinguishable and non-en-antiomeric possibilities exist (indicated as N–N, N–S, N–E, and N–W, Figure S15), relative to the positions of thetwo phenyl rings; the optimized structure for the “N–N”dimer is shown in Figure 16. DeVoe calculations on thefour structures led to multiple CD signals in the Soretregion (Figure 16) with variable intensity. With the ex-ception of the N–S dimer, four bands of alternating signwere predicted in all other cases, which can be describedas a central couplet with pronounced side bands.b) Tilted porphyrin dimers: A second set of DeVoe calcula-tions was run on a system of two ideal, perfectly planarporphyrins, stacked on each other, with a relative orien-tation (plane-to-plane distance and x offset) similar tothat discussed above. In this case a tilt angle q was intro-duced to generate chirality, which describes the deviationfrom parallelism along the y direction (Figure 17). Pa-rameters relative to CuPor were employed. CalculatedCD spectra in the Soret region for the idealized dimersshow again a central couplet with strong sidebands, evenfor very small values of q (Figure 17).From the comparison of the last two sets of calculations,it is clear that upon the formation of a p–p stacked por-phyrin dimer, comparable CD effects may be obtainedeither as a consequence of distortion from planarity, orof deviation from perfect parallelism. It must be stressedthat multiple absorption and CD Soret bands can also beobtained for various dimer geometries by introducing aFigure 16. DeVoe-calculated CD spectra on four MMFF-optimized structures of [Cu–PhPyr3P]2 dimers. Thestructure of the “N–N” dimer is shown on the right, in two views (from the top and from one side); pyridylrings in red, phenyl rings in yellow (in the “northern” quadrant), arrows depicting interporphyrin N!Cu coor-dinations. For other structures with phenyl rings in different quadrants, see Figure S15.Figure 17. DeVoe-calculated CD spectra on ideal quasi-parallel porphyrindimers with geometries schematized on the right, as a function of tiltangle q.www.chemeurj.org 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim Chem. Eur. J. 2009, 15, 11853 – 1186611862M. Balaz, N. Berova, G. A. Ellestad et al.
  11. 11. substantial energy separation between the two (Bx andBy) components.[88,89]This is, however, not observed inthe case of tetraarylporphyrins such as 2HPor and itsmetallated derivatives ZnPor and CuPor, whose Soretabsorption bands are not detectably split and relativelynarrow.[31,32,67,20]c) Porphyrin dimers perturbed by external dipoles: A wayto generate chirality in the otherwise achiral assembly oftwo perfectly parallel circular oscillators consists in theintroduction of external dissymmetry elements. In theporphyrin-appended DNAs, the porphyrins experiencethe presence of the chiral environment provided by thepolynucleotide, which can be described by means of p–p* transition dipoles. Thus, we constructed an input forDeVoe calculations constituted by two ideal and parallelcircular oscillators (geometry as in Figure 17 above, q=08), plus four linear oscillators describing the main tran-sitions of the base pairs (A–T) closest to the porphyrins.The position and orientation of the dipoles (Figure 18,right) reproduced the geometry of the A–T base pairsfound in the MD-generated semicapped structure2HPor-8mer-MD2, and parameters relative to 2HPorwere used. This input aimed at modeling the perturba-tion exerted by DNA bases on the porphyrin-porphyrincoupling in the aggregates of DNA duplexes. Figure 18shows the DeVoe-calculated CD spectra. When only thedegenerate coupling between the two parallel circularoscillators is allowed, the calculated CD is zero as ex-pected. When only the non-degenerate couplings be-tween the two circular and the four linear oscillators areallowed, a monosignate Soret CD band appears, similarto the spectrum described above for the 2HPor-8mer-MD2 structure (Figure 9B). Very interestingly, however,when both the degenerate and the non-degenerate typesof couplings are allowed at the same time, the calculatedCD spectrum shows a complex pattern of multiple bandsin the Soret region, although weak. This phenomenon isdue to so-called high-order exciton couplings.[28,67,95]Thus, when interactions with extrinsic linear dipoles areinvolved, the coupling between the two circular oscilla-tors, which is negligible at the first order, becomes al-lowed at higher orders, and contributes strongly and in arather complicated way to the observed CD. As a result,multiple CD Soret bands may be generated due to theperturbation exerted by extrinsic dipoles on an otherwiseachiral p–p stacked porphyrin dimer.ConclusionsThe CD spectra of the three porphyrin-DNA conjugates2HPor-8mer, CuPor-8mer and ZnPor-8mer were investigat-ed in different conditions. At negligible ionic strength, allthree conjugates exhibit negative exciton CD couplets in theSoret region but with different intensities. The double-stranded conformation of the zinc porphyrin–DNA conju-gate with apparently more favorable twist between two in-teracting porphyrins is reflected in a very strong conserva-tive exciton couplet CD signal with an amplitude A of aboutÀ160mÀ1cmÀ1at an interchromophoric distance of about43 Š. The negative exciton chirality between the two inter-acting porphyrins is corroborated by MD calculations. Atthe shorter interchromophoric distance of about 32–36 Š,the copper and free base porphyrins show weaker long-range porphyrin–porphyrin interactions in the Soret regionthan the zinc conjugate. The porphyrins of these two conju-gates tend to interact more strongly with adjacent base pairsdue to their planarity (no axial ligand) and hydrophobicity.The axial ligand together with increased hydrophilicity pre-vents effective p–p interactions of zinc porphyrin with anadjacent terminal adenine–thymine base pair. Therefore, itis conceivable that the axial zinc–water coordination whichprevents a close porphyrin interaction with adjacent basepairs (capping) is the determining factor for this significantlymore intense, through-space CD exciton coupling observedin the case of 5’-end-appended zinc porphyrin DNA duplex-es. It follows that a p–p interaction of porphyrin chromo-phore with adjacent nucleobases has a disturbing effect onthe intensity and the shape of the bisignate CD curve associ-ated with an interporphyrin coupling.Interestingly, the large difference in CD intensities of thethree conjugates is counter-intuitive. The two uncapped Zn-porphyrins, endowed with larger conformational mobility,seem able to attain a more effective chiral interporphyrintwist (Figure 8) for exciton coupling. It is noteworthy thatDeVoe calculations on representative fully uncappedZnPor-8mer and semicapped 2HPor-8merMD2 structuressubstantiate the experimental findings. In particular, theydemonstrate that the stacking of a porphyrin on the adjacentbases forces it into a poorly effective arrangement for theexciton coupling with respect to the second porphyrin at theend of the duplex.At increased ionic strength (150 mm NaCl) porphyrin mo-lecular caps promote the formation of porphyrin–DNAnanoaggregates with very characteristic although unusualmultisignate CD Cotton effects within the porphyrin SoretFigure 18. DeVoe-calculated CD spectra on an ideal parallel porphyrindimer perturbed by vicinal A–T base pairs with a geometry (right) de-rived from 2HPor–8mer-MD2 structure.Chem. Eur. J. 2009, 15, 11853 – 11866 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim www.chemeurj.org 11863FULL PAPERDNA–Porphyrin Conjugates
  12. 12. region (trisignate or tetrasignate CD curves). The origin ofthe observed multisignate CD profile is the strong electronicinteraction between stacked porphyrins, which arise fromNaCl-promoted head-to-tail interduplex aggregates. Such astrong coupling overcomes the through-space exciton cou-pling between the two end-capped porphyrins within theduplex. Porphyrins play the role of DNA “molecular glue”in formation of the observed aggregates. This explanation isalso supported by MD and MC conformational analysis.Fluorescence and resonance light scattering experiments notonly confirmed the formation of porphyrin–DNA nanoag-gregates but shed light on their size and structure. An in-crease of fluorescence intensity by a factor of ~5 is observedupon denaturation of ZnPor-8mer at 450 mm NaCl which isconsistent with interduplex rod-like aggregates containingapproximately ten ZnPor-8mer units.MD and MC calculations were carried out to determinethe preferred conformation(s) of the porphyrins around theDNA duplexes. These simulations revealed a significant dif-ference in the dynamic behavior of the conjugates depend-ing on the type of porphyrin (presence and nature of metalin the porphyrin center). The 2HPor-8mer conjugate dis-played a reduced structural flexibility at one end of thiscomplex, with one of the porphyrins clearly showing lessmobility while retaining some degree of hydrophobic over-lap with the terminal A–T base pair to which it is covalentlytethered. In contrast, the ZnPor-8mer conjugate showedboth tethered porphyrins to be equally mobile, uncappingfully from the terminal A–T base pairs and moving freely inthe solvent, and in some cases, rotating fully, transiently cap-ping the A–T base pairs before uncapping again. In bothcases, the dynamic movement of the porphyrins is striking.It is clear that no single structure can capture the nature ofthis complex. It is found to be a highly dynamic complex,and the access of the porphyrins to the solvent and possiblyformation of other transient aggregates between distinctporphyrin DNA duplexes cannot be disregarded.DeVoe CD calculations demonstrated that multiple CDbands in the Soret region (including couplets with strongside bands) may be obtained only as a consequence of anadditional strong interaction between two (or more) por-phyrins, for example, that observed upon face-to-face pstacking. Multiple CD Soret bands may arise by several ide-alized conditions: 1) exciton coupling between stacked por-phyrins with sizeable deviations from planarity; 2) excitoncoupling between quasi-planar, non-parallel stacked por-phyrins; 3) exciton coupling between quasi-planar, quasi-parallel stacked porphyrins, perturbed by surrounding chro-mophores. These three possibilities do not exclude eachother. On the contrary, for supramolecular aggregates ofporphyrin–DNA conjugate duplexes, they are likely to existall at the same time. Therefore, the observed CD arisesmost probably from a complex combination of the three fac-tors, and an accurate prediction of the CD spectra of theseaggregates would require one to define many subtle struc-tural parameters at a high level of precision. Qualitativelyspeaking, however, comparison of Figures 16 and 17 withFigure 11b reveals a surprising similarity between the tetra-signate CD observed for aggregates of CuPor-8mer andsome of our simulated CD spectra. The aggregates of2HPor-8mer and ZnPor-8mer have less symmetrical, trisig-nate CD spectra in the Soret region, which may be thoughtto possibly arise from a superposition of a tetrasignate-likespectrum with other types of features. The most importantconclusion is that CD spectroscopy has shown to be able tosensitively detect the occurrence of aggregation betweenend-capped porphyrin–DNA duplexes promoted by certainconditions of concentration, temperature and ionic strength.In summary, we have shown that CD spectroscopy, and inparticular the exciton chirality method, is a very useful ex-perimental technique for the determination of the structureof optically active multi-chromophoric assemblies. Asshown, however, using the example of the porphyrin- andmetalloporphyrin–DNA conjugates, a close interaction be-tween spatially-fixed chromophores may prevent the con-ventional application of this technique based on through-space interchromophoric dipole–dipole interactions. Yet,even in such more complicated cases, when supported byother techniques and calculations, CD spectroscopy provesto be vital for detection of conformational changes andsupramolecular events.AcknowledgementsA.M. thanks the PhD Program in Scienze Biochimiche e Biomolecolari,University of Catania, Italy for financial support. S.J. and N.J.T. thank theNational Science Foundation for financial support through grant NSF-CHE-07-17518. M.B. thanks the University of Wyoming Start-up grantfor financial support. We thank Columbia undergraduate summer-re-search students, Amanda Wolfe and Warren McGee for contributions atthe initial stages of the CD and computational studies, respectively.A.G.P. thanks Istvan Kolossvary from DeShaw Co for technical sup-port in MD simulations. G.P. thanks MIUR for financial support(PRIN 2007PBWN44)[1] L. A. Fendt, I. Bouamaied, S. Thoni, N. Amiot, E. Stulz, J. Am.Chem. Soc. 2007, 129, 15319–15329.[2] V. L. Malinovskii, F. Samain, R. Haner, Angew. Chem. 2007, 119,4548–4551; Angew. Chem. Int. Ed. 2007, 46, 4464–4467.[3] E. Mayer-Enthart, H. A. Wagenknecht, Angew. Chem. 2006, 118,3451–3453; Angew. Chem. Int. Ed. 2006, 45, 3372–3375.[4] M. Endo, M. Fujitsuka, T. Majima, J. Org. Chem. 2008, 73, 1106–1112.[5] K. C. Hannah, B. A. Armitage, Acc. Chem. Res. 2004, 37, 845–853.[6] F. D. Lewis, X. Liu, Y. Wu, X. Zuo, J. Am. Chem. Soc. 2003, 125,12729–12731.[7] F. D. Lewis, L. Zhang, X. Liu, X. Zuo, D. M. Tiede, H. Long, G. C.Schatz, J. Am. Chem. Soc. 2005, 127, 14445–14453.[8] Z. S. Yoon, M. C. Yoon, D. Kim, J. Photochem. Photobiol. C 2005, 6,249–263.[9] B. Bouvier, T. Gustavsson, D. Markovitsi, P. Millie, Chem. Phys.2002, 275, 75–92.[10] G. D. Scholes, G. Rumbles, Nat. Mater. 2006, 5, 683–696.[11] M. Simonyi, Z. Bikadi, F. Zsila, J. Deli, Chirality 2003, 15, 680–698.[12] A. Huijser, T. J. Savenije, S. C. J. Meskers, M. J. W. Vermeulen,L. D. A. Siebbeles, J. Am. Chem. Soc. 2008, 130, 12496–12500.[13] G. D. Scholes, ACS Nano 2008, 2, 523–537.[14] W. Kuhlbrandt, D. N. Wang, Nature 1991, 350, 130–134.www.chemeurj.org 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim Chem. Eur. J. 2009, 15, 11853 – 1186611864M. Balaz, N. Berova, G. A. Ellestad et al.
  13. 13. [15] S. Bahatyrova, R. N. Frese, C. A. Siebert, J. D. Olsen, K. O. van derWerf, R. van Grondelle, R. A. Niederman, P. A. Bullough, C. Otto,C. N. Hunter, Nature 2004, 430, 1058–1062.[16] G. S. Beddard, J. Phys. Chem. B 1998, 102, 10966–10973.[17] N. Berova, K. Nakanishi in Exciton Chirality Method: Principles andApplications (Eds.: N. Berova, K. Nakanishi, R. W. Woody), Wiley-VCH, Weinheim, 2000, pp. 337–382.[18] N. Harada, S. L. Chen, K. Nakanishi, J. Am. Chem. Soc. 1975, 97,5345–5352.[19] P. G. A. Janssen, J. Vandenbergh, J. L. J. van Dongen, E. W. Meijer,A. Schenning, J. Am. Chem. Soc. 2007, 129, 6078–6079.[20] R. Iwaura, F. J. M. Hoeben, M. Masuda, A. Schenning, E. W. Meijer,T. Shimizu, J. Am. Chem. Soc. 2006, 128, 13298–13304.[21] N. Sakai, P. Talukdar, S. Matile, Chirality 2006, 18, 91–94.[22] G. A. Ellestad, Chirality 2006, 18, 134–144.[23] F. D. Lewis, Y. Wu, L. Zhang, X. Zuo, R. T. Hayes, M. R. Wasielew-ski, J. Am. Chem. Soc. 2004, 126, 8206–8215.[24] S. Werder, V. L. Malinovskii, R. Haner, Org. Lett. 2008, 10, 2011–2014.[25] R. Häner, F. Samain, V. L. Malinovskii, Chem. Eur. J. 2009, 15,5701–5708.[26] N. Berova, L. Di Bari, G. Pescitelli, Chem. Soc. Rev. 2007, 36, 914–931.[27] S. Superchi, R. Bisaccia, D. Casarini, A. Laurita, C. Rosini, J. Am.Chem. Soc. 2006, 128, 6893–6902.[28] S. Superchi, E. Giorgio, C. Rosini, Chirality 2004, 16, 422–451.[29] S. G. Allenmark, Nat. Prod. Rep. 2000, 17, 145–155.[30] S. Matile, N. Berova, K. Nakanishi, J. Fleischhauer, R. W. Woody, J.Am. Chem. Soc. 1996, 118, 5198–5206.[31] A. Ghosh in Modern Quantum Chemical Studies of Porphyrins andMetalloporphyrins, Vol. 7 (Eds.: K. M. Kadish, K. M. Smith, R. Gui-lard), Academic Press, San Diego, 2000, pp. 1–38.[32] M. Gouterman in Optical Spectra and Electronic Structure of Por-phyrins and Related Rings, Vol. III (Ed.: D. Dolphin), AcademicPress, New York, 1978, pp. 1–165.[33] K. Tsubaki, K. Takaishi, H. Tanaka, M. Miura, T. Kawabata, Org.Lett. 2006, 8, 2587–2590.[34] S. Matile, N. Berova, K. Nakanishi, Chem. Biol. 1996, 3, 379–392.[35] M. Balaz, A. E. Holmes, M. Benedetti, P. C. Rodriguez, N. Berova,K. Nakanishi, G. Proni, J. Am. Chem. Soc. 2005, 127, 4172–4173.[36] M. Balaz, J. D. Steinkruger, G. A. Ellestad, N. Berova, Org. Lett.2005, 7, 5613–5616.[37] M. Balaz, B. C. Li, J. D. Steinkguger, G. A. Ellestad, K. Nakanishi,N. Berova, Org. Biomol. Chem. 2006, 4, 1865–1867.[38] M. Balaz, B. C. Li, G. A. Ellestad, N. Berova, Angew. Chem. 2006,118, 3610–3613; Angew. Chem. Int. Ed. 2006, 45, 3530–3533.[39] F. J. M. Hoeben, M. Wolffs, J. Zhang, S. De Feyter, P. Leclere, A.Schenning, E. W. Meijer, J. Am. Chem. Soc. 2007, 129, 9819–9828.[40] R. F. Pasternack, A. Giannetto, P. Pagano, E. J. Gibbs, J. Am. Chem.Soc. 1991, 113, 7799–7800.[41] R. F. Pasternack, C. Bustamante, P. J. Collings, A. Giannetto, E. J.Gibbs, J. Am. Chem. Soc. 1993, 115, 5393–5399.[42] E. Bellacchio, R. Lauceri, S. Gurrieri, L. M. Scolaro, A. Romeo, R.Purrello, J. Am. Chem. Soc. 1998, 120, 12353–12354.[43] R. Purrello, E. Bellacchio, S. Gurrieri, R. Lauceri, A. Raudino,L. M. Scolaro, A. M. Santoro, J. Phys. Chem. B 1998, 102, 8852–8857.[44] R. Purrello, L. Monsu Scolaro, E. Bellacchio, S. Gurrieri, A.Romeo, Inorg. Chem. 1998, 37, 3647–3648.[45] R. Lauceri, A. Raudino, L. M. Scolaro, N. Micali, R. Purrello, J.Am. Chem. Soc. 2002, 124, 894–895.[46] A. Mammana, M. De Napoli, R. Lauceri, R. Purrello, Bioorg. Med.Chem. 2005, 13, 5159–5163.[47] N. E. Mukundan, G. Pethç, D. W. Dixon, M. S. Kim, L. G. Marzilli,Inorg. Chem. 1994, 33, 4676–4687.[48] S. Arimori, M. Takeuchi, S. Shinkai, J. Am. Chem. Soc. 1996, 118,245–246.[49] T. S. Balaban, A. D. Bhise, M. Fischer, M. Linke-Schaetzel, C. Rous-sel, N. Vanthuyne, Angew. Chem. 2003, 115, 2190–2194; Angew.Chem. Int. Ed. 2003, 42, 2140–2144.[50] A. S. R. Koti, N. Periasamy, Chem. Mater. 2003, 15, 369–371.[51] T. S. Balaban, M. Linke-Schaetzel, A. D. Bhise, N. Vanthuyne, C.Roussel, Eur. J. Org. Chem. 2004, 3919–3930.[52] P. Prochµzkovµ, Z. Zelinger, K. Lang, P. Kubµt, J. Phys. Org. Chem.2004, 17, 890–897.[53] T. Park, J. S. Shin, S. W. Han, J.-K. Son, S. K. Kim, J. Phys. Chem. B2004, 108, 17106–17111.[54] H. Onouchi, T. Miyagawa, K. Morino, E. Yashima, Angew. Chem.2006, 118, 2441–2444; Angew. Chem. Int. Ed. 2006, 45, 2381–2384.[55] D. Monti, M. Venanzi, M. Stefanelli, A. Sorrenti, G. Mancini, C. Di-Natale, R. Paolesse, J. Am. Chem. Soc. 2007, 129, 6688–6689.[56] P. Stepanek, M. Dukh, D. Saman, J. Moravcova, L. Kniezo, D.Monti, M. Venanzi, G. Mancini, P. Drasar, Org. Biomol. Chem.2007, 5, 960–970.[57] A. Tsuda, M. A. Alam, T. Harada, T. Yamaguchi, N. Ishii, T. Aida,Angew. Chem. 2007, 119, 8346–8350; Angew. Chem. Int. Ed. 2007,46, 8198–8202.[58] R. van Hameren, A. M. vanBuul, M. A. Castriciano, V. Villari, N.Micali, P. Schon, S. Speller, L. Monsu Scolaro, A. E. Rowan,J. A. A. W. Elemans, R. J. M. Nolte, Nano Lett. 2008, 8, 253–259.[59] J. H. Fuhrhop, C. Demoulin, C. Boettcher, J. Koening, U. Siggel, J.Am. Chem. Soc. 1992, 114, 4159–4165.[60] M. Balaz, K. Bitsch-Jensen, A. Mammana, G. A. Ellestad, K. Naka-nishi, N. Berova, Pure Appl. Chem. 2007, 79, 801–809.[61] M. Balaz, M. De Napoli, A. E. Holmes, A. Mammana, K. Nakanishi,N. Berova, R. Purrello, Angew. Chem. 2005, 117, 4074–4077;Angew. Chem. Int. Ed. 2005, 44, 4006–4009.[62] K. Toyofuku, M. A. Alam, A. Tsuda, N. Fujita, S. Sakamoto, K. Ya-maguchi, T. Aida, Angew. Chem. 2007, 119, 6590–6660; Angew.Chem. Int. Ed. 2007, 46, 6476–6480.[63] Z. Tomovic, J. Van Dongen, S. J. George, H. Xu, W. Pisula, P. Le-clere, M. M. J. Smulders, S. De Feyter, E. W. Meijer, A. Schenning,J. Am. Chem. Soc. 2007, 129, 16190–16196.[64] Y. Q. Zhang, P. L. Chen, L. Jiang, W. P. Hu, M. H. Liu, J. Am.Chem. Soc. 2009, 131, 2756–2757.[65] J. Tabei, M. Shiotsuki, F. Sanda, T. Masuda, Macromolecules 2005,38, 9448–9454.[66] A. Mammana, T. Asakawa, K. Bitsch-Jensen, A. Wolfe, S. Chaturan-tabut, Y. Otani, X. X. Li, Z. M. Li, K. Nakanishi, M. Balaz, G. A. El-lestad, N. Berova, Bioorg. Med. Chem. 2008, 16, 6544–6551.[67] G. Pescitelli, S. Gabriel, Y. K. Wang, J. Fleischhauer, R. W. Woody,N. Berova, J. Am. Chem. Soc. 2003, 125, 7613–7628.[68] 0m NaCl, 50 mm K-phosphate buffer, RT, pH 7.0.[69] M. Nakash, Z. Clyde-Watson, N. Feeder, J. E. Davies, S. J. Teat,J. K. M. Sanders, J. Am. Chem. Soc. 2000, 122, 5286–5293.[70] Schrçdinger-Inc. Portland, 2007–2008. http://www.schrodinger.com.[71] Desmond developed by D. E. Shaw Inc., New York NY 10036(USA), 2008.[72] P. Maragakis, K. Lindorff-Larsen, M. P. Eastwood, R. O. Dror, J. L.Klepeis, I. T. Arkin, M. O. Jensen, H. F. Xu, N. Trbovic, R. A. Fries-ner, A. G. Palmer, D. E. Shaw, J. Phys. Chem. B 2008, 112, 6155–6158.[73] H. Devoe, J. Chem. Phys. 1964, 41, 393–400.[74] H. Devoe, J. Chem. Phys. 1965, 43, 3199–3208.[75] C. L. Cech, W. Hug, I. Tinoco, Jr., Biopolymers 1976, 15, 131–152.[76] C. L. Cech, I. Tinoco, Jr., Biopolymers 1977, 16, 43–65.[77] W. C. Johnson, Jr., I. Tinoco, Jr., Biopolymers 1969, 7, 727–749.[78] V. Rizzo, J. A. Schellman, Biopolymers 1984, 23, 435–470.[79] D. Baumstark, H. A. Wagenknecht, Angew. Chem. 2008, 120, 2652–2654; Angew. Chem. Int. Ed. 2008, 47, 2612–2614.[80] M. Hariharan, Y. Zheng, H. Long, T. A. Zeidan, G. C. Schatz, J.Vura-Weis, M. R. Wasielewski, X. Zuo, D. M. Tiede, F. D. Lewis, J.Am. Chem. Soc. 2009, 131, 5920–5929.[81] M. Kasha, Radiat. Res. 1963, 20, 55–70.[82] R. F. Pasternack, P. J. Collings, Science 1995, 269, 935–939.Chem. Eur. J. 2009, 15, 11853 – 11866 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim www.chemeurj.org 11865FULL PAPERDNA–Porphyrin Conjugates
  14. 14. [83] G. Bringmann, D. C. G. Gçtz, T. A. M. Gulder, T. H. Gehrke, T.Bruhn, T. Kupfer, K. Radacki, H. Braunschweig, A. Heckmann, L.Christoph, J. Am. Chem. Soc. 2008, 130, 17812–17825.[84] M. Ikeda, S. Shinkai, A. Osuka, Chem. Commun. 2000, 1047–1048.[85] M. Takeuchi, T. Imada, S. Shinkai, Bull. Chem. Soc. Jpn. 1998, 71,1117–1123.[86] A. H. Shelton, A. Rodger, D. R. McMillin, Biochemistry 2007, 46,9143–9154.[87] L. Palivec, M. Urbanova, K. Volka, J. Pept. Sci. 2005, 11, 536–545.[88] J. Sˇebek, P. Bourˇ, J. Phys. Chem. A 2008, 112, 2920–2929.[89] T. Yamamura, T. Mori, Y. Tsuda, T. Taguchi, N. Josha, J. Phys.Chem. A 2007, 111, 2128–2138.[90] M. C. Hsu, R. W. Woody, J. Am. Chem. Soc. 1971, 93, 3515–3525.[91] M.-C. Hsu, R. W. Woody, J. Am. Chem. Soc. 1969, 91, 3679–3681.[92] C. Rosini, R. Ruzziconi, S. Superchi, F. Fringuelli, O. Piermatti, Tet-rahedron: Asymmetry 1998, 9, 55–62.[93] C. A. Hunter, J. K. M. Sanders, J. Am. Chem. Soc. 1990, 112, 5525–5534.[94] W. R. Scheidt, Y. J. Lee in Recent Advances in the Stereochemistry ofMetallotetrapyrroles, Vol. 64 (Ed.: J. W. Buchler), Springer, Berlin,1987, pp. 1–70.[95] C. Rosini, M. Zandomeneghi, P. Salvadori, Tetrahedron: Asymmetry1993, 4, 545–554.Received: July 22, 2009Published online: October 20, 2009www.chemeurj.org 2009 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim Chem. Eur. J. 2009, 15, 11853 – 1186611866M. Balaz, N. Berova, G. A. Ellestad et al.

×