Exploring the binding ability of phenanthroline-based polyammonium receptors for anions: Hints for design of selective chemosensors for nucleotides

Article abstract of DOI:10.1021/jo901423m

(Chemical Equation Presented) The synthesis of receptor 2,6,10,14,18-pentaaza[20]-21,34-phenanthrolinophane (L1), containing a pentaamine chain linking the 2,9 positions of a phenanthroline unit, is reported. The protonation features of L1 and of receptor

Full text of DOI:10.1021/jo901423m

pubs.acs.org/joc
Exploring the Binding Ability of Phenanthroline-Based Polyammonium
Receptors for Anions: Hints for Design of Selective Chemosensors for
Nucleotides
Carla Bazzicalupi, Andrea Bencini,* Silvia Biagini, Enrico Faggi, Stefano Meini,
Claudia Giorgi, Alessio Spepi, and Barbara Valtancoli
ꢀ
Dipartimento di Chimica, Universita di Firenze, Via della Lastruccia, 3, 50019 Sesto Fiorentino, Firenze,
Italy
andrea.bencini@unifi.it
Received July 4, 2009
The synthesis of receptor 2,6,10,14,18-pentaaza[20]-21,34-phenanthrolinophane (L1), containing a
pentaamine chain linking the 2,9 positions of a phenanthroline unit, is reported. The protonation
features of L1 and of receptor 2,6,10,14,18,22-hexaaza[23]-24,37-phenanthrolinophane (L2) have
1
been studied by means of potentiometric, H NMR, and spectrofluorimetric measurements; this
study points out that the fluorescent emission of both receptors depends on the protonation state of
the polyamine chain. In fact, the receptors are emissive only at neutral or acidic pH values, where all
the aliphatic amine groups are protonated. Potentiometric titrations show that L2 is able to bind
selectively ATP over TTP, CTP, and GTP. This selectivity is lost in the case of L1. 1H and 31P NMR
measurements and molecular mechanics calculations show that the phosphate chains of nucleotides
give strong electrostatic and hydrogen-bonding interactions with the ammonium groups of the
protonated receptors, while the nucleobases interact either via π-stacking with phenanthroline or via
hydrogen bonding with the ammonium groups. Of note, MM calculations suggest that all nucleo-
tides interact in an inclusive fashion. In fact, in all adducts the phosphate chain is enclosed within
the receptor cavities. This structural feature is confirmed by the crystal structure of the
[(H6L2)2(TTP)2(H2O)2]4+ adduct. Fluorescence emission measurements at different pH values show
that L2 is also able to ratiometrically sense ATP in a narrow pH range, thanks to emission quenching
due to a photoinduced electron transfer (PET) process from an amine group of the receptor to the
excited phenanthroline.
Introduction
substrates within clefts or pockets with the appropriate
dimensions and disposition of the binding sites to optimally
host a selected nucleotide. It is now accepted that the
binding process is regulated by different noncovalent inter-
actions, such as charge-charge and charge-dipole inter-
actions, hydrogen bonding, hydrophobic effects, and
stacking interactions, that work cooperatively to stabilize
the adducts.
Nucleotides are ubiquitously present in biological sys-
tems and play crucial roles in many cellular functions, such
as transport across membranes, DNA synthesis, cell signal-
ing, and energy- or electron-transfer processes.1 These
functions are generally regulated by recognition processes
involving proteins able to selectively bind the appropriate
nucleotide anion, thanks to the encapsulation of the anionic
DOI: 10.1021/jo901423m
r
Published on Web 09/10/2009
J. Org. Chem. 2009, 74, 7349–7363 7349
2009 American Chemical Society

Bazzicalupi et al.
JOCArticle
In this context, the design of synthetic receptors able to
bind phosphate anions in aqueous solution represents one of
the approaches to the analysis of the weak forces which
regulate the recognition processes in biological systems.2
Actually, several examples of nucleotide anion binding by
synthetic receptors, mostly of polyammonium type, have
been recently reported.3-29 Similarly to natural systems, the
formation of stable host-guest adducts requires the incor-
poration in the receptor of sites for multiple interactions with
substrates. In fact, to achieve a better recognition of nucleo-
tide anions, the receptor needs to contain not only positively
charged ammonium groups able to interact with the anionic
phosphate moiety but also binding sites able to interact via
hydrogen bonding or π-stacking with the nucleobases or the
sugar moieties.2-11,13,19-21 Finally, the binding sites of the
receptor need to be opportunely preorganized to optimally
interact with the anionic substrate. From this point of view,
encapsulation of the nucleotide, or of a portion of the
nucleotide, in clefts or cavities of the receptor may strengthen
the overall host-guest interaction affording particularly
stable adducts or favoring recognition of a selected nucleo-
tide anion. Actually, encapsulation of a determined subunits
of nucleotides, e.g., the phosphate chain or the nucleobase,
within the cavity of cyclic polyammonium receptors has
been often proposed on the basis of molecular modeling
results.4,5
(1) (a) Hirsch, A. K. H.; Fischer, F. R.; Diederich, F. Angew. Chem., Int.
Ed. 2007, 46, 338–352. (b) Dugas, H. Bioorganic Chemistry: a Chemical
Approach to Enzyme Action; Springer: New York, 1996. (c) Davidson, A. M. L.;
Dassa, E.; Orelle, C.; Chen, J. Microbiol. Mol. Biol. Rev. 2008, 72, 317–364.
(d) Alton, G. R.; Lunney, E. A. Expert Opin. Drug Discovery 2008, 3, 595–605.
(e) Lewis, J. A.; Lebois, E. P.; Lindsley, C. W. Curr. Opin. Chem. Biol. 2008, 12,
269–280. (f) Matte, A.; Delbaere, L. T. J. Handbook Proteins 2007, 1, 114–118.
(g) Turk, B. E. Curr. Opin. Chem. Biol. 2008, 12, 4–10. (h) Hollenstein, K.;
Dawson, R. J. P.; Locher, K. P. Curr. Opin. Struct. Biol. 2007, 17, 412–418.
(i) Morrow, J. R.; Amyes, T. L.; Richard, J. P. Acc. Chem. Res. 2008, 41, 539–
548. (l) Rye, C. S.; Baell, J. B. Curr. Med. Chem. 2005, 12, 3127–3141.
(2) Selected reviews: (a) Atwood, J. L.; Holman, K. T.; Steed, J. W. Chem.
Commun. 1996, 1401–1407. (b) Fabbrizzi, L.; Licchelli, M.; Rabaioli, G.;
Taglietti, A. Coord. Chem. Rev. 2000, 205, 85–108. (c) Amendola, V;
Bonizzoni, M.; Esteban-Gomez, D.; Fabbrizzi, L.; Licchelli, M.; Sancenon,
~
F.; Taglietti, A. Coord. Chem. Rev. 2006, 250, 1451–1470. (d) Garcia Espana,
E.; Diaz, P.; Llinares, J. M.; Bianchi, A. Coord. Chem. Rev. 2006, 250, 2952–
~
2980. (e) Bianchi, A., Garcia-Espana, E., Bowman-James K., Eds. Supramole-
cular Chemistry of Anions; Wiley-VCH: New York, 1997. (f) Tamaru, S.;
Hamachi, I. Struct. Bonding 2008, 129, 95–125. (g) Katayev, E. A.; Ustynyuk,
Y. A.; Sessler, J. L. Coord. Chem. Rev. 2006, 250, 3004–3037. (h) Caltagirone,
C.; Gale, P. A. Chem. Soc. Rev. 2009, 38, 520–563. (i) Kim, S. K.; Lee, D. H.;
Hong, J.; Yoon, J. Acc. Chem. Res. 2009, 42, 23–31. (j) Kang, S. O.; Hossain,
M. A.; Bowman-James, K. Coord. Chem. Rev. 2006, 250, 3038–3052.
(k) Gunnlaugsson, T.; Glynn, M.; Tocci, G. M.; Kruger, P. E.; Pfeffer, F. M.
Coord. Chem. Rev. 2006, 250, 3094–3117. (l) Lankshear, M. D.; Beer, P. D.
Coord. Chem. Rev. 2006, 250, 3142–3160. (m) Gimeno, N.; Vilar, R. Coord.
Chem. Rev. 2006, 250, 3161–3189. (n) Gale, P. A.; Quesada, R. Coord. Chem.
Rev. 2006, 250, 3219–3244. (o) Amendola, V.; Fabbrizzi, L. Chem. Commun.
2009, 513–531. (p) Kubik, S. Chem. Soc. Rev. 2009, 38, 585–605. (q) Chmie-
lewski, M. J.; Davis, J. J.; Beer, P. D. Org. Biomol. Chem. 2009, 7, 415–424.
(r) Parola, A. J.; Lima, J. C.; Lodeiro, C.; Pina, F. Springer Ser. Fluoresc. 2008, 4,
117–149. (s) Gamez, P.; Mooibroek, T. J.; Teat, S. J.; Reedijk, J. Acc. Chem. Res.
2007, 40, 435–444. (t) Kang, S. O.; Begum, R. A.; Bowman-James, K. Ang.
Chem. Int.Ed. 2006, 45, 7882–7894. (u) Anzenbacher, P.; Nishiyabu, R.;
Palacios, M. A. Coord. Chem. Rev. 2006, 250, 2929–2938. (v) Schmidtchen,
F. P. Coord. Chem. Rev. 2006, 250, 2918–2928. (w) Schmidtchen, F. P.; Berger,
M. Chem. Rev. 1997, 97, 1609–1646. (y) de Silva, A. P.; Gunaratne, H. Q. N.;
Gunnlaugsson, T.; Huxley, A. J. M.; McCoy, C. P.; Rademacher, J. T.; Rice, T. E.
Chem. Rev. 1997, 97, 1515–1566.
The insertion of heteroaromatic subunits with fluoro-
genic characteristics within the receptor structure may
represent an important “added value” to this class of
receptors. In fact, in this case, the aromatic moiety can be
used not only to bind the substrates via π-stacking and or
hydrophobic interactions but also to signal their presence
in solution, thanks to quantifiable changes of its emission
properties.2,12-29
(17) Casasus, R.; Climent, E.; Marcos, M. D.; Martinez-Manez, R.;
Sancenon, F.; Soto, J.; Amoros, P.; Cano, J.; Ruiz, E. J. Am. Chem. Soc.
2008, 130, 1903–1917.
(18) Ambrosi, G.; Formica, M.; Fusi, V.; Giorgi, L.; Guerri, A.; Macedi,
E.; Micheloni, M.; Paoli, P.; Pontellini, R.; Rossi, P. Inorg. Chem. 2009, 48,
5901–5912.
(19) (a) Sakamoto, T.; Ojida, A.; Hamachi, H. I. Chem. Commun. 2009,
141–152. (b) Ojida, A.; Nanaka, H.; Miyahara, Y.; Tamaru, S.; Sada, K.;
Hamachi, I. Angew. Chem., Int. Ed. 2006, 45, 5518–5521. (c) Ojida, A.;
Takashima, I.; Kohira, T.; Nonaka, H.; Hamachi, I. J. Am. Chem. Soc. 2008,
130, 12095–12101.
(3) (a) Hosseini, M. W.; Blaker, A. J.; Lehn, J. M. J. Am. Chem. Soc. 1990,
112, 3896–3904. (b) Dhaenens, M.; Lehn, J. M.; Vigneron, J. P. J. Chem.
Soc., Perkin Trans. 2 1993, 1379–1381.
€
(4) (a) Schneider, H. J.; Blatter, T.; Palm, B.; Pfingstag, U.; Rudiger, V.;
Theis, I. J. Am. Chem. Soc. 1992, 114, 7704–7708. (b) Eliseev, A. V.;
Schneider, H. J. J. Am. Chem. Soc. 1994, 116, 6081–6088.
~
(5) (a) Aguilar, J. A.; Garcia-Espana, E.; Guerrero, J. A.; Luis, S. V.;
Llinares, J. M.; Miravet, J. F.; Ramirez, J. A.; Soriano, C. J. Chem. Soc.,
Chem. Commun. 1995, 2237–2238. (b) Aguilar, J. A; Celda, B.; Fusi, V.;
(20) Casasus, R.; Climent, E.; Marcos, M. D.; Martinez-Manez, R.;
Sancenon, F.; Soto, J.; Amoros, P.; Cano, J.; Ruiz, E. J. Am. Chem. Soc.
2008, 130, 1903–1917.
~
Garcia-Espana, E.; Luis, S. V.; Martinez, M. C.; Ramirez, J. A.; Soriano, C.;
Tejero, R. J. Chem. Soc., Perkin Trans. 2 2000, 1323–1328.
(6) Furuta, H.; Magda, D.; Sessler, J. L. J. Am. Chem. Soc . 1991, 113,
978–985.
(7) Menger, F. M.; Catlin, K. K. Angew. Chem., Int. Ed. Engl. 1995, 34,
2147–2150.
(8) (a) Bazzicalupi, C.; Bencini, A.; Bianchi, A.; Faggi, E.; Giorgi, C.;
Santarelli, S.; Valtancoli, B. J. Am. Chem. Soc. 2008, 130, 2440–2441.
(b) Arturoni, E.; Bazzicalupi, C.; Bencini, A.; Caltagirone, C.; Danesi, A.;
Giorgi, C.; Garau, A.; Lippolis, V.; Valtancoli, B. Inorg. Chem. 2008, 47,
6651–6563.
(21) Zapata, F.; Caballero, A.; Espinosa, A.; Tarraga, A.; Molina, P.
J. Org. Chem. 2008, 73, 4034–40.
(22) Zyryanov, G. V.; Palacios, M. A.; Anzenbacher, P. Angew. Chem.,
Int. Ed. 2007, 46, 7849–7852.
(23) Jose, D. A.; Mishra, S.; Ghosh, A.; Shrivastav, A.; Mishra, S. K.;
Das, A. Org. Lett. 2007, 9, 1979–1982.
(24) Khatua, S.; Choi, S. H.; Lee, J.; Kim, K.; Do, Y.; Churchill, D. G.
Inorg. Chem. 2009, 48, 2993–2999.
(25) Sun, Y.; Zhong, C.; Gong, R.; Fu, E. Org. Biomol. Chem. 2008, 6,
3044–3047.
~
(9) Padilla-Tosta, M. E.; Lloris, J. M.; Martinez-Manez, R.; Pardo, T.;
(26) Ghosh, A.; Shrivastav, A.; Jose, D. A.; Mishra, S. K.; Chandrakanth,
C. K.; Mishra, S.; Das, A. Anal. Chem. 2008, 80, 5312–5319.
(27) Wang, H.; Chan, W. H. Org. Biomol. Chem. 2008, 6, 162–168.
(28) Lee, H. N.; Xu, Z.; Kim, S. K.; Swamy, K. M. K; Kim, Y.; Kim, S. J.;
Yoon, J. J. Am. Chem. Soc. 2007, 129, 3828–3829.
Soto, J.; Benito, A.; Marcos, M. D. Inorg. Chem. Commun. 2000, 3, 45–48.
(10) Vickers, M. S.; Martindale, K. S.; Beer, P. D. J. Mat. Chem. 2005, 15,
2784–2790.
(11) Delepine, A. S.; Tripier, R.; Handel, H. Org. Biomol. Chem. 2008, 6,
1743–1750.
(12) Kwon, J. Y.; Singh, N. J.; Kim, H. N.; Kim, S. K.; Kim, K. S.; Yoon,
J. J. Am. Chem. Soc. 2004, 126, 8892–8893.
(13) Huston, M. E.; Akkaya, E. U.; Czarnik, A. W. J. Am. Chem. Soc.
1989, 111, 8735–8736.
(29) Lee, D. H.; Kim, S. Y.; Hong, J. I. Angew. Chem., Int. Ed. 2004, 43,
4777–4780.
(30) (a) Bazzicalupi, C.; Bencini, A.; Biagini, S.; Bianchi, A.; Faggi, E.;
Giorgi, C.; Marchetta, M.; Totti, F.; Valtancoli, B. Chem.;Eur. J. 2009, 15,
8049–8063. (b) Bazzicalupi, C.; Bencini, A.; Bussotti, L.; Berni, E.; Biagini,
S.; Faggi, E.; Foggi, P.; Giorgi, C.; Lapini, A.; Marcelli, A.; Valtancoli, B.
Chem. Commun. 2007, 1230–1232. (c) Bazzicalupi, C.; Bencini, A.; Bianchi,
A.; Borsari, L.; Danesi, A.; Giorgi, C.; Lodeiro, C.; Mariani, P.; Pina, F.;
Santarelli, S.; Tamayo, A.; Valtancoli, B. Dalton Trans. 2006, 33, 4000–4010.
(d) Lodeiro, C.; Parola, A. J.; Pina, F.; Bazzicalupi, C.; Bencini, A.; Bianchi,
A.; Giorgi, C.; Masotti, A.; Valtancoli, B. Inorg. Chem. 2001, 40, 2968–2975.
(e) Bencini, A.; Bernardo, M. A.; Bianchi, A.; Fusi, V.; Giorgi, C.; Pina, F.;
Valtancoli, B. Eur. J. Inorg. Chem. 1999, 1911–1918.
(14) Kejik, Z.; Zaruba, K.; Michalik, D.; Sebek, J.; Dian, J.; Pataridis, S.;
Volka, K.; Kral, V. Chem. Commun. 2006, 1533–1535.
(15) Atilgan, S.; Akkaya, E. U. Tetrahedron Lett. 2004, 45, 9269–9271.
~
(16) (a) Albenda, M. T.; Bernardo, M. A.; Garcia-Espana, E.; Godino-
Salido, M. L.; Luis, S. V.; Melo, M. J.; Pina, F.; Soriano, C. J. Chem. Soc.,
Perkin Trans. 2 1999, 2545–2549. (b) Aucejo, R.; Diaz, P.; Garcia-Espana, E.;
Alarcon, J.; Delgado-Pinar, E.; Torres, F.; Soriano, C.; Guillem, C. M.
New J. Chem. 2007, 31, 44–51.
7350 J. Org. Chem. Vol. 74, No. 19, 2009

Bazzicalupi et al.
JOCArticle
SCHEME 1. Drawings of Ligands and Nucleotides with Atom Labeling
In the course of our study on polyamine ligands contain-
ing heteroaromatic units as fluorescent receptors for both
metal cations and anions,8,30 we have synthesized receptors
L1 and L2, which contain respectively a pentaamine and a
hexaamine chain linking the 2,9 positions of 1,10-phenan-
throline (Scheme 1). In principle, both ligands may give
highly charged polyammonium cations in aqueous solut-
ions, due to the high number of protonable amine groups
gathered on the receptors. At the same time, the protonated
species of these receptors are characterized by large and
positively charged cavities, where the anionic phosphate
chain of nucleotides could be enclosed. Finally, phenan-
throline is characterized by an intense fluorescence emis-
sion at ca. 370 nm, which could be affected by substrate
binding.
Therefore, we decided to analyze the ability of L1 and L2
to bind and sense nucleotide triphosphate anions, e.g., ATP,
GTP, TTP, and CTP. A preliminary investigation on L2
showed that this receptor is able to bind and selectively sense
at pH 6 ATP over GTP, TTP, and CTP,31 although the four
nucleotides present similar structures and possess similar
binding sites. This result suggests that recognition processes
can be controlled by slightly different structural parameters.
At the same time, it cannot be neglected the fact that
nucleotide recognition by polyamines in aqueous solutions
can be strongly influenced by the protonation states of the
host and guest species, i.e., by the pH of the medium.
With this in mind, we have now compared the binding
features of L1 and L2, which possess an overall similar
structure, toward nucleotide triphosphate, with the pur-
pose to evaluate the structural parameters which determine
nucleotide binding and sensing. At the same time, we have
also carried out a study on the binding ability of the receptors
toward nucleotides in a wide pH range, with the aim to draw
useful indications for design of selective chemosensors for
anions in aqueous solutions.
(32) In the case of L2, the values of the first and second protonation
constants, as well as those of the fifth and sixth ones, are very similar. This
behavior has been already found in other cyclic and acyclic polyamine
receptors, and in some cases, the constant for the addition of an acidic
protons to a [HxL]xþ species, e.g., for the equilibrium [HxL]xþ þ Hþ
[Hxþ1L](xþ1)þ, was found to be equal or even higher than the constant for the
addition of a proton to the less protonated species [Hx-1L](x-1)þ, e.g.,
relative to the equilibrium [Hx-1L](x-1) þ Hþ=[HxL]xþ (see, for instance:
(a) Micheloni, M.; Sabatini, A.; Paoletti, P. J. Chem. Soc., Perkin Trans. 2
1978, 828–830. (b) Micheloni, M.; Paoletti, P.; Vacca, A. J. Chem. Soc.,
Perkin Trans. 2 1978, 1978–1980. (c) Hosseini, M. W.; Lehn, J.-M. J. Am.
Chem. Soc. 1982, 104, 3525–3526. (d) Hosseini, M. W.; Lehn, J.-M. Helv.
Chim. Acta 1986, 69, 587-603 and (e) Dietrich, B.; Hosseini, M. W.; Lehn, J.-M.;
Session, R. B. Helv. Chim. Acta 1983, 66, 1262–1278. Other examples can be
found in ref 33). This unexpected behavior is generally due either to the formation
of a stabilizing intramolecular hydrogen-bonding network in the [Hxþ1L](xþ1)þ
species upon protonation of the [HxL]xþ one, or to energetically expensive
conformational changes of the receptor upon protonation of [Hx-1L](x-1) to
=
give [HxL]xþ. Both these effects can energetically favor the [HxL]xþ þ Hþ
=
[Hxþ1L](xþ1)þ process with respect to the [Hx-1L](x-1) þ Hþ=[HxL]xþ one, leading
to similar values for the protonation constants of the [Hx-1L](x-1) and [HxL]xþ
species (see ref 33 for a more detailed discussion).
(31) Bazzicalupi, C.; Biagini, S.; Bencini, A.; Faggi, E.; Giorgi, C.;
Matera, I.; Valtancoli, B. Chem. Commun. 2006, 4087–4089.
J. Org. Chem. Vol. 74, No. 19, 2009 7351

Bazzicalupi et al.
JOCArticle
TABLE 1. Protonation Constants of L1 and L2 (0.1 M NMe4Cl, 298.
1 K)
log K
reaction
L = L1
L = L2a
L + H+ = [HL]+
10.24(1)
9.87(1)
8.02(1)
7.24(1)
6.51(2)
10.14(1)
10.10(1)
8.59(1)
7.79(1)
6.98(1)
6.90(1)
[HL]+ + H+ = [H2L]2+
[H2L]2+ + H+ = [H3L]3+
[H3L]3+ + H+ = [H4L]4+
[H4L]4+ + H+ = [H5L]5+
[H5L]5+ + H+ = [H6L]6+
aFrom ref 31.
FIGURE 1. pH dependence of the 1H NMR chemical shifts of the
benzylic protons of L1 (b) and distribution diagrams of its proto-
nated forms (298 K, 0.1 M NMe4Cl).
Results and Discussion
Protonation of the Receptors. Since the binding properties
of polyammonium receptors are generally determined by the
formation of protonated species in aqueous solutions and by
the localization of the ammonium groups within the receptor
structure, we preliminarily carried out a study on the basicity
properties of L1 and L2 by coupling potentiometric, 1H
NMR, spectrophotometric, and fluorimetric measurements
in aqueous solutions. The protonation constants of L2 have
been previously communicated,31 while the protonation
equilibria of L1 have been potentiometrically determined
in this study. The basicity constants of both ligands are
reported in Table 1.
Receptors L1 and L2 can bind up to five and six acidic
protons in the pH range investigated (2.5-10.5), respect-
ively.32 Their protonation constants are higher than those
observed for the corresponding protonation equilibria in the
phenanthroline-containing macrocycle 2,5,8,11,14-pentaa-
aza[15]-16,29-phenanthrolinophane,30e which contains five
amine groups separated by ethylenic chains, in agreement
with the larger +I inductive effect exerted on amine groups
by propylenic chains and with the higher flexibility of L1 and
L2 which allow a better minimization of the electrostatic
repulsion between positive charges in the polyprotonated
forms of the receptors.33 All protonation constants of L1 and
L2 are also higher than those reported for 1,10-phenanthro-
line (log K = 4.96),34 suggesting that the heteroaromatic
nitrogens are not directly involved in the process of proton
binding in aqueous solutions, at least in the pH range
FIGURE 2. Emission spectra of L1 at selected pH values (a) and
pH dependence of its emission intensity at 366 nm (298 K, λexc
270 nm, 0.1 M NMe4Cl) (b).
1
investigated. The analysis of the pH dependence of the H
A similar behavior is also found in the case of L2, where a
marked upfield shift of this signal is observed upon the
formation of the [H5L2]5+ and [H6L2]6+protonated spe-
cies (Supporting Information, Figure S2). The first three (L1)
or four protonation steps (L2) take place on the central
amine groups of the aliphatic chain. The lower proton
affinity of the Na nitrogen atoms, in comparison with the
Nb and Nc ones, can be reasonably ascribed to the electron-
withdrawing effect of the heteroaromatic unit on the adja-
cent Na amine groups. Therefore, L1 and L2 present a
preferential binding zone for protons, i. e., the central amine
groups Nb and Nc of the aliphatic chain, located far from the
heteroaromatic units. Of note, in both L1 and L2, the
resonances of the heteroaromatic protons show only minor
shifts in the pH range 2-12, supporting the hypothesis that
phenanthroline is not directly involved in the process of
proton binding.
NMR signals often give useful information on the localiza-
tion of the acidic protons in polyammonium cations.33
However, in the case of L1 and L2, the resonances of the
hydrogen atoms of the propylenic chains cannot be safely
attributed at all pH values, due to their similar chemical
shifts. Conversely, the sharp singlet of the benzylic methy-
lene group (4 in Scheme 1) at 4.69 (L1, pH 1.9) or 4.72 ppm
(L2, pH 2), adjacent to the phenanthroline unit, is easily
recognizable at all pH values in both receptors.
As shown in Figure 1 for L1, the chemical shift is not
affected by pH in the pH range 11-8. A remarkable down-
field shift is observed below pH 8, i.e., with the formation of
the [H4L1]4+ and [H5L1]5+ species in solution, suggesting
that the last two protonation steps of the receptor occur on
the benzylic nitrogen atoms adjacent to phenanthroline Na.
~
(33) Bencini, A.; Bianchi, A.; Garcia-Espana, E.; Micheloni, M.; Ramirez,
J. A. Coord. Chem. Rev. 1999, 188, 97–156.
(34) Smith, R. M.; Martell, A. E. NIST Stability Constants Database,
version 4.0; National Institute of Standards and Technology: Washington,
DC, 1997.
This fact is substantially confirmed by the analysis of the UV
spectra of L1 and L2 recorded in aqueous solutions at different
pH values. In fact, the UV band of phenanthroline at 270 nm is
7352 J. Org. Chem. Vol. 74, No. 19, 2009

Bazzicalupi et al.
JOCArticle
TABLE 2. Stepwise Stability Constants of the Adducts Formed by L1 and L2 with Nucleotides (0.1 M NMe4Cl, 298.1 K)
log K
reaction
S
4- = ATP
S4- = CTP
S4- = TTP
S4- = GTP
[HL1]+ + S4- = [HL1S]3-
2.61(3)
3.12(3)
4.11(3)
5.66(3)
5.67(4)
5.73(4)
3.57(1)
3.80(1)
4.27(2)
5.70(2)
5.64(2)
6.09(2)
[H2L1]2+ + S4- = [H2L1S]2-
[H3L1]3+ + S4- = [H3L1S]-
[H4L1]4+ + S4- = [H4L1S]
3.24(2)
4.07(2)
5.59(3)
5.30(3)
5.93(3)
3.01(4)
4.64(5)
[H3L1]3+ + [HS]3- = [H4L1S]
[H4L1]4+ + [HS]3- = [H5L1S]+
[H5L1]5+ + S4- = [H5L1S]+
[H5L1]5+ + [HS]3- = [H6L1S]2+
[H5L1]5+ +[H2S]2- = [H7L1A]3+
[H5L1]5+ + [H3S]- = [H8L1S]4+
6.01(5)
6.13(5)
5.29(5)
5.83(5)
6.05(4)
5.30(4)
5.64(4)
5.31(2)
4.29(2)
5.48(3)
4.60(3)
[H2L2]2+ + S4- = [H2L2S]2-
[H3L2]3+ + S4- = [H3L2S]-
[H4L2]4+ + S4- = [H4L2S]
4.29(2)a
5.41(2)
7.08(2)
8.86(3)
8.89(3)
9.68(3)
9.91(3)
9.55(3)
9.17(3)
3.16(2)a
4.38(2)
5.98(2)
7.84(2)
7.27(2)
8.95(2)
8.62(2)
8.58(4)
8.06(4)
3.03(3)a
4.15(3)
5.24(3)
7.11(3)
7.21(3)
7.67(4)
7.98(3)
8.06(3)
3.57(1)
2.87(3)a
4.47(3)
5.75(3)
6.60(3)
6.93(3)
7.87(3)
8.40(3)
7.16(3)
[H4L2]4+ + [HS]3- = [H5L2S]+
[H5L2]5+ + S4- = [H5L2S]+
[H5L2]5+ + [HS]3- = [H6L2S]2+
[H6L2]6+ + S4- = [H6L2S]2+
[H6L2]6+ + [HS]3- = [H7L2S]3+
[H6L2]6+ + [H2S]2- = [H8L2S]4+
aFrom ref 31.
affected only by a small redshift (ca. 4 nm) accompanied by a
slight increase of the absorbance below pH 9.
These spectral changes can be attributed to hydrogen
bonding with the ammonium groups of aliphatic chains rather
than to direct involvement of the phenanthroline in proton
binding. In fact, protonation of 1,10-phenanthroline35 is
accompanied by a far higher redshift (ca. 20 nm) of its
absorption band. Conversely, the emission spectra of L1
and L2 are strongly pH-dependent. As shown in Figure 2
for L1, the receptor is not emissive in alkaline aqueous
solutions and shows the typical emission of phenanthroline
only at acidic pH values, i.e., upon protonation of the
[H4L1]4+ species to give the [H5L1]5+ one. This effect can
beattributed tothe presence, in[H4L1]4+, of anunprotonated
benzylic amine group (Na in Scheme 1). In fact, as already
observed in other phenanthroline-containing polyamine li-
gands,30d,e the lone pair of the benzylic nitrogens, located close
to phenanthroline, can efficiently quench the fluorescence
emission of the fluorophore through an electron transfer
process. Protonation of both the benzylic amine groups below
pH 7 makes their lone pairs no longer available for the PET
process and gives rise to a consequent renewal of the fluores-
cence emission of phenanthroline. A similar behavior is found
for L2, where only the hexaprototonated [H6L2]6+ species,
which is prevalent is solution at in the acidic pH region, is
emissive (see the Supporting Information, Figure S3).
Binding of Nucleotide Triphosphates in Aqueous Solutions.
Since the binding properties of polyamine receptors toward
anionic species in aqueous solution are often dependent on
their protonation state and, therefore, on the pH of the
solutions, wedecidedtoinvestigate firsttheirabilitytointeract
FIGURE 3. Distribution of the ATP adducts with L1 (a) and L2
(b) ([L1]=[L2]=[ATP]=1 10-3 M, 298 K, 0.1 M NMe4Cl).
3
with nucleotide triphosphate anions by means of potentio-
metric measurements. Both receptors form stable 1:1 adducts
in aqueous solutions (Table 2). Although the formation of
both 1:1 and 2:1 anion/receptor adducts has been observed in
previously reported cases,36 data analysis with the program
(35) (a) Armaroli, N.; De Cola, L.; Balzani, V.; Sauvage, J.-P.; Dietrich-
Buchecker, C. O.; Kern, J. M. J. Chem. Soc., Faraday Trans. 1992, 88, 553–
556. (b) Kern, J. M.; Sauvage, J.-P.; Weidmann, J. L.; Armaroli, N.;
Flamigni, L.; Ceroni, P.; Balzani, V. Inorg. Chem. 1997, 36, 5329–5338.
(c) Armaroli, N.; De Cola, L.; Balzani, V.; Sauvage, J.-P.; Dietrich- Buchecker,
C. O.; Kern, J. M.; Bailal, A. J. Chem. Soc., Dalton Trans. 1993, 3241–
3247. (d) Armaroli, N.; Ceroni, P.; Balzani, V.; Kern, J. M.; Sauvage,
J.-P.; Weidmann, J. L. J. Chem. Soc., Faraday Trans. 1997, 93, 4145–
4150.
(36) Dietrich, B.; Hosseini, M. W.; Lehn, J.-M.; Session, R. B. J. Am.
Chem. Soc. 1981, 103, 1282–1283.
(37) Gans, P.; Sabatini, A.; Vacca, A. Talanta 1996, 43, 1739–1753.
J. Org. Chem. Vol. 74, No. 19, 2009 7353

Bazzicalupi et al.
JOCArticle
FIGURE 5. pH dependence of the 31P NMR chemical shifts of the
ATP signals in the absence and in the presence of 1 equiv of L1 and
L2 ([L1]=[L2]=[ATP]=5 ꢀ 10-3 M).
equilibria in the same pH range and by the values of the
protonation constants of the substrates, which are quite
similar, in some cases, to the protonation constants of
receptors. In these cases, in fact, different equilibria can be
proposed for the formation of the same supramolecular
adduct. For instance, the [H6L2(ATP)]2+ adduct could be
formed either through the equilibrium [H5L2]5+ +[HATP]3-=
[H6L2(ATP)]2+ or [H6L2]6+ + ATP4- = [H6L2(ATP)]2+
,
FIGURE 4. Plot of log Keff as a function of pH for the L1 (a) and L2
with calculated equilibrium constants of 9.68 and 9.91,
respectively. The values of the stability constants of the
adducts, calculated by using this approach are given in
Table 2.
To overcome the presence of different equilibria present in
solution at the same pH value, it can be useful to calculate
effective stability constants.38 For a given pH value, if the
total amount of free substrate (ΣH(h-i)S), free receptor
(ΣHiS) and adduct formed (ΣHhLS) are known, one can
define an effective stability constant by using the following
equation:
(b) adducts with nucleotides (Keff=ΣHhLS/(ΣH(h-i)S ΣHiL), with
L=L1 or L2 and S=ATP, CTP, TTP or GTP).
3
HYPERQUAD37 under our experimental conditions reveals
only 1:1 stoichiometries for all species detected in our
systems. The eventual formation of oligomeric adducts (with
2:2 stoichiometries, for instance) was also ruled out by
performing potentiometric titrations in a wide range of con-
centrations of receptors and substrates (see the Experimental
Section).
Analysis of the titration curves allows one to determine the
species formed in solution and their overall formation con-
stants βHLS (S=ATP, CTP, TTP or GTP, L=L1 or L2),
Keff ¼ ΣHhLS=ðΣH
S ΣHiLÞ
3
ðh -iÞ
Plots of the pH dependence of the logarithms of the
effective constants (Figure 4) point out that in all cases the
stability of the adducts is strongly influenced by pH. In fact, a
marked increase of the log Keff values is generally observed
from alkaline to slightly acidic pH values, with a maximum
between pH 7 and 4; then the stability generally shows a
slight decrease at lower pH values. In the alkaline and
slightly acidic pH region, the substrates are in their less
protonated and highly charged forms (S4- or HS3-), and
therefore the observed increasing stability from pH 10 to
slightly acidic pH values can be reasonably attributed to the
increasing number of protonated ammonium functions gath-
ered on the receptor, which enhances the receptor ability to
give electrostatic and hydrogen-bonding interactions with
the anionic substrates. Conversely, the decrease of the log
Keff values at more acidic pH values can be reasonably
attributed to the progressive formation in solution of highly
relative to equilibria of the type L + S4- + nH+
=
[HnLS](n-4)+ (see the Supporting Information, Table S1).
Plots of the distribution curves of the species formed by L1
and L2 with the nucleotides (see Figure 3 for the ATP
adducts with both receptors and the Supporting Informa-
tion, Figure S4-S6, for the CTP, GTP, and TTP adducts)
outline that complexation occurs in a wide pH range (for
instance ATP is almost completely bound by both polyam-
monium receptors for pH values lower than 9), with the
formation of a large number of 1:1 adducts with different
protonation degrees.
The determination of the stepwise formation constants for
the adducts between the protonated receptors and the anio-
nic substrates, relative to equilibria of the type Hh-iL +
HiS=[HhLS], implies the knowledge of the localization of
the acidic protons on receptor and substrates in the [HhLS]
host-guest species. This task is made difficult by the pre-
sence in aqueous solution of a large number of overlapping
(38) Bianchi, A.; Garcia-Espan~a, E. J. Chem. Educ. 1999, 76, 1727–1732.
7354 J. Org. Chem. Vol. 74, No. 19, 2009

Bazzicalupi et al.
JOCArticle
TABLE 3. 31P NMR Shifts (δ, ppm) of Nucleotides in Their Adducts with L1 and L2 and Corresponding Complexation-Induced 31P NMR Chemical
Shifts (CIS, ppm), Measured in D2O Solution at pH 6, 298 K
ATP
CTP
TTP
GTP
PR
Pβ
Pγ
PR
Pβ
Pγ
PR
Pβ
Pγ
PR
Pβ
Pγ
Receptor L1
δ (ppm)
CIS
-10.7
0.2
-21.8
2.2
-6.0
3.0
-10.7
0.2
-22.0
0.7
-7.7
2.9
-10.8
0.0
-22.6
0.5
-7.9
2.6
-11.0
0.1
-21.9
0.5
-8.6
2.0
Receptor L2a
δ (ppm)
CIS
-9.9
1.0
-20.3
2.7
-5.2
4.8
-10.5
0.4
-21.5
1.2
-6.5
3.8
-10.8
0.1
-22.2
0.9
-7.6
2.9
-10.9
0.2
-21.3
1.1
-7.4
3.2
aFrom ref 31.
protonated and less negatively charged forms of the sub-
strates (H2S2- and H3S-) and to the consequent reduced
electrostatic and hydrogen-bonding interactions between the
two partners.
If the pH dependence of the conditional constants is
similar for both receptors, significant differences are instead
observed in the binding ability at a given pH value of the two
receptors. In fact, L2 shows a remarkable higher binding
affinity for each substrate than L1 and a marked selectivity
for ATP over GTP, TTP, and CTP (Figure 4b) all over the
pH range investigated. Conversely, a similar binding affinity
for all four nucleotides features L1, which displays only a
slight preference for ATP binding in the pH range 4-6.
While the higher affinity of L2 toward nucleotide anions
could be simply attributed to the higher positive charge
gathered on the receptor at a given pH (for instance, at pH
6, L1 and L2 are present in solution as penta- and hexapro-
tonated species, [H5L1]5+ and [H6L2]6+), the selectivity
displayed by L2 for ATP cannot be interpreted only in terms
of charge-charge and hydrogen-bonding interactions be-
tween receptors and substrates. Therefore, we also analyzed
1
the different host-guest systems by means of H and 31P
NMR spectra recorded at various pH values.
Figure 5 reports the chemical shifts of the 31P NMR signals
of ATP in the presence and in the absence of L1 or L2 at
different pH values. Similar plots have been also obtained for
CTP, GTP, and TTP and are supplied within the Supporting
Information (Figure S7), while Table 3 lists the CIS
(complexation-induced shifts) of the different signals meas-
ured at pH 6, where the most stable adducts are generally
formed. Binding of all substrates to L1 or L2 gives noticeable
changes in the chemical shifts of the 31P NMR signals of
the triphosphate chain. In fact, marked upfield shifts are
observed for the signals of the terminal Pγ phosphate group
and, to a lesser extent, of the central Pβ group. The signal of
the PR phosphate is generally only slightly affected by the
presence of the receptors, and therefore, it is probably
weakly involved in the stabilization of the adducts.
Considering nucleotide complexation by L2, Table 3 and
Figures 5 and S7 show that the upfield shifts of the 31P signals
in the presence of the receptor decrease in the order ATP >
CTP > TTP ≈ GTP, which is almost the same sequence
found for the stability constants of the adducts determined
by potentiometric measurements (Figure 4). At the same
time, for a given nucleotide, the adducts with L2 display
higher complexation induced upfield shifts than those observed
for L1. At a first glance, these results seem to suggest that
the stability of the adducts is mainly determined
FIGURE 6. pH dependence of the 1H NMR signals of the adenine
protons H2 and H8 and of the ribose proton H10 of ATP in the
presence of 1 equiv of L1 and L2.
by charge-charge and hydrogen-bonding interactions
involving the phosphate chain. On the other hand, anal-
ysis of the 1H NMR spectra recorded at different pH
values on D2O solutions containing the nucleotides and
L1 or L2 provides unambiguous evidence of the involve-
ment of the nucleobases in the recognition processes. In
fact, all nucleotides, in the presence of L1 and L2, display
downfield shifts for the signals of the aromatic protons of
nucleobases and of the anomeric proton H10 of the sugar
moiety.
Plots of the chemical shifts of these signals of nucleotides
in the presence and in the absence of L1 or L2 vs pH (see
Figure 6 for ATP and Supporting Information, Figures S8,
for the other nucleotides) show that the complexation-in-
duced chemical shifts are higher at slightly acidic pH values
and strongly decrease in the alkaline pH region. However,
the observed complexation-induced upfield shifts of the
heteroaromatic protons of phenanthroline and nucleobases
are less affected by pH than the downfield shifts observed for
J. Org. Chem. Vol. 74, No. 19, 2009 7355

Bazzicalupi et al.
JOCArticle
TABLE 4. 1H NMR Chemical Shifts for Protons of Phenanthroline, for Aromatic Protons of Nucleobases, and for the Anomeric H10 Proton of Ribose in
the Adducts of L1 and L2 with ATP, CTP, GTP, and TTP and Corresponding Complexation-Induced Chemical Shifts (CIS, ppm), Measured in D2O
Solution at pH 6, 298 K
H2
H1
H3
H8
H2
H10
L1
δ
CIS
δ
CIS
8.31
-0.09
8.27
7.78
-0.11
7.63
7.76
-0.07
7.65
ATP
ATP
8.22
-0.33
7.91
7.95
-0.34
7.49
5.83
-0.38
5.25
L2a
-0.26
-0.32
-0.14
-0.60
-0.76
-0.96
H2
H1
H3
H6
H5
H10
L1
δ
CIS
δ
CIS
8.32
-0.08
8.39
-0.14
H2
7.82
-0.07
7.73
-0.22
H1
7.65
-0.08
7.72
-0.08
H3
CTP
CTP
7.75
-0.20
7.67
-0.28
H6
6.11
-0.11
6.02
5.88
-0.14
5.79
-0.24
H10
L2a
-0.20
L1
δa
CIS
δ
CIS
8.34
-0.06
8.35
7.83
-0.06
7.75
7.67
-0.06
7.71
TTP
TTP
7.68
-0.10
7.43
6.13
-0.19
5.90
L2a
-0.18
-0.20
-0.09
-0.35
-0.42
H2
H1
H3
H8
H10
L1
δa
CIS
δ
CIS
8.28
-0.11
8.35
7.78
-0.11
7.67
7.63
-0.10
7.70
GTP
GTP
8.08
-0.15
7.79
5.75
-0.10
5.46
L2a
-0.23
-0.28
-0.10
-0.44
-0.39
aFrom ref 31.
the 31P signals of the phosphate chain. Whereas the latter are
strongly reduced or, in the same cases, almost absent below
pH 4-5 and at alkaline pH values, the former show only
minor changes in a wide pH range, generally from pH 2 to 8,
indicating that the interactions between phenanthroline and
nucleobases are less sensible to the pH of the medium than
the charge-charge interactions involving the phosphate
moieties of nucleotides.
In the case of the larger receptor L2, the observed downfield
shifts induced by complexation with the receptor (Table 4)
decreases in the order ATP . GTP > TTP > CTP account-
ing for a stronger interaction of adenine of ATP with the
receptor. Of note, the 1H NMR signals of the phenanthroline
unit of L2 also exhibit relevant downfield shifts (see Table 4
and Figure 7b for ATP and Supporting Information, Figures
S9-S11) in the presence of all nucleotides. The downfield
shiftsofthephenanthrolineprotons of L2 are larger in thecase
ofATP and decrease in the order ATP . GTP > TTP ≈ CTP,
a sequence similar to that observed for the protons of nucleo-
bases in the presence of L2. These results may indicate that the
phenanthroline unit of L2 and the nucleobases are coupled by
of π-stacking interactions, which decrease on passing from
ATP to CTP.
In the case of the smaller receptor L1, the formation of the
host-guest adducts with all nucleotides induces smaller
shifts of the 1H resonances of both nucleobases and phenan-
throline with respect to complexation with the larger host L2
(Table 4, Figures 6 and 7 and Supporting Information,
Figures S8-S11). In particular, the shift of the phenanthro-
line signals of L1 in the presence of nucleotides are only
slightly affected by complexation, suggesting that the hetero-
aromatic unit of this receptor is weakly involved in the
stabilization of the adducts.
sphate chains of nucleotides, but also by the interactions
between the nucleobases and the two polyammonium hosts.
Overall, the ATP adducts with L2, the most stable among
all the systems investigated, display the strongest char-
ge-charge interactions involving the phosphate chain and
robust π-stacking interactions between adenine and phenan-
throline. Conversely, the phenanthroline unit of the smaller
receptor L1 is likely to play a minor role in the stabilization of
the adducts, accounting for the lower stability of the adducts
with this receptor.
To corroborate these results, we carried out a MM analy-
sis on the adducts with L1 and L2, and we tried to obtain
crystals suitable for X-ray analysis. Our attempts to crystal-
lize the complexes were unfruitful, with the only exception of
the TTP-L2 adduct, which was structurally characterized.
Molecular Modeling and Structure Analysis. The hos-
t-guest interactions between the two receptors and nucleo-
tides were analyzed by means of molecular modeling studies
by using an implicit simulation of aqueous solution
(preliminary results on ATP and CTP complexation with
L2 have been previously presented).31 In particular, we
focused our attention on the interaction between the
[H5L1]5+ and [H6L2]6+protonated receptors, which are the
main species present in aqueous solution at neutral and
slightly acidic pH values, and the tetraanionic forms of
nucleotides. The most interesting finding is the encapsulation
of the triphosphate chain of nucleotides within the macro-
cyclic cavity of L1 and L2. This structural feature is common
to all sampled conformers of the four different anionic
substrates. In all adducts, the two protonated receptors
assume a similar conformation, folded along the axis con-
necting the two benzylic amine groups, while the nucleotides
are characterized by a bent conformation which allows the
simultaneous encapsulation of the phosphate chain within
the cavity of the protonated guest and the interaction of the
nucleobase either with the phenanthroline unit or with the
polyammonium chain.
These NMR results suggest that the stability sequences
observed for the host-guest adducts is not only determined
by electrostatic interactions and hydrogen-bonding contacts
between the ammonium groups of L1 or L2 and the tripho-
7356 J. Org. Chem. Vol. 74, No. 19, 2009

Bazzicalupi et al.
JOCArticle
FIGURE 7. pH dependence of the signals of the phenanthroline
protons of L1 (a) and L2 (b) in the absence and in the presence of 1
equiv of ATP.
In the case of the adducts with L2, the sampled conformers
can be grouped in two different families, which differ in the
interaction mode with the receptor. In the first family
(Figure 8, A family) the nucleobases interact essentially via
face to face π-stacking with the heteroaromatic unit of the
receptor. In the case of ATP and GTP, the purine bases are
stacked above the phenanthroline unit and lie on planes
almost parallel to that of phenanthroline. In the adducts with
CTP and TTP, the pyrimidine bases lie on planes slightly
bent (24.6° and 7.6° for CTP and TTP, respectively) with
respect to that phenanthroline and are less “overlapped”
with this heteroaromatic unit.
The A family is the most populated in the case of ATP
(70% of sampled conformers) and the less populated in the
adducts with CTP (35%), TTP (30%) and GTP (40%),
accounting for a higher tendency of ATP to form π-stacked
adducts. In the second family, the most populated for CTP,
TTP, and GTP, the nucleobases are still located close to the
phenanthroline unit, but the interaction takes place essent-
ially via hydrogen bonding with the protonated nitrogens of
FIGURE 8. Lowest energy conformers of the adducts between
[H6L2]6+ and ATP (a), GTP (b), CTP (c), and TTP (d) in the
A families of conformers.
the receptor (Figure 9). In particular, the oxygen group O8 of
the carbonyl function of GTP and TTP or the nitrogen atoms
N7 and N3 of ATP and CTP are hydrogen bonded with one
ammonium group of the aliphatic chain (see the Supporting
Information, Figures S12-S14 for labeling).
In both families, encapsulation of the triphosphate unit
within the receptor cavity enables the formation of a number
of hydrogen-bonding contacts with the charged polyammo-
nium chain (see the Supporting Information for a complete
J. Org. Chem. Vol. 74, No. 19, 2009 7357

Bazzicalupi et al.
JOCArticle
1
observed for the 31P and H NMR signals of ATP in the
presence of the receptor and can account for the higher
stability found for the ATP adducts with L2. It seems likely
that the phosphate chain and the adenine unit of ATP play a
more pronounced cooperative role in the binding process
than the corresponding triphosphate units and nucleobases
of the other nucleotides. If this can be ascribed, in the case of
CTP and TTP, to the poorer tendency of cytosine and
thymine to give π-stacking and hydrophobic interactions
than adenine, due to their less extended aromatic structure,
this consideration cannot be applied to GTP, since guanine
and adenine generally display similar ability to give
π-stacked complexes. On the other hand, it is known that
steric effects can influence the stability of host-guest
adducts. From this point of view, a tentative explanation
for the higher stability displayed by the ATP adducts with
respect to the GTP ones could reside in the mutual disposit-
ion of the nucleobase and the aliphatic chain of L2. In fact, to
achieve simultaneously inclusion of the phosphate chain and
face-to-face π-stacking, the receptor assumes a folded con-
formation, with the polyamine chain almost perpendicular
to the phenanthroline plain. Therefore, to “overlap” phe-
nanthroline, guanidine and adenine need to assume a posit-
ion very close and almost perpendicular to the aliphatic
chain of L2. The guanine structure is somewhat more
hindered than adenine, due to the simultaneous presence of
the two bulky CdO and -NH2 groups. Differently from the
-NH2 group of ATP, in the guanine adduct with L2 these
two groups are also oriented toward the aliphatic chain of L2
(Figure 8b). Steric hindrance between guanine and the
aliphatic portion of L2 could justify the observed less
tendency of GTP to give π-stacked adducts.
In the L1 adducts, the receptor and nucleotides assume
bent conformations similar to those found in the case of L2,
which enable the encapsulation of the triphosphate unit
within the cavity of the receptor (Supporting Information,
Figure S12) and the formation of a hydrogen bonding net-
work with the ammonium groups of the receptor. The most
significant difference with respect to the adducts with L2 can
be observed in the interaction mode of the nucleobases with
the receptor. In fact, most of the sampled conformers display
a mutual disposition of the nucleobases and phenanthroline
similar to that observed in the B families of the L2 adducts,
featured by hydrogen bonding contacts between the nucleo-
bases and the ammonium groups of the receptor. Actually,
for all nucleotides, only a limited number of adducts
(<15%) display face-to-face π-stacking pairing, suggesting
that in L1 this kind of interaction plays a minor role in the
stabilization of the adducts and, overall, that phenanthroline
is weakly involved in the recognition process. This result,
once again, is in agreement with the small complexation
induced shift observed for the 1H NMR signals of the
phenanthroline unit of L1.
FIGURE 9. Lowest energy conformers of the adducts between
[H6L2]6+ and ATP (a), GTP (b), CTP (c), and TTP (d) in the
B families of conformers.
list of the hydrogen bonds, Tables S3-S5). The terminal and
central phosphate groups are the most involved in the
resulting hydrogen-bonding network, in keeping with the
31P NMR results. Of note, among the different nucleotides,
the triphosphate unit of ATP gives a robust hydrogen-
bonding network with the polyammonium chain and, at
the same time, exhibits the most pronounced π-stacking
interaction between the nucleobase and phenanthroline.
These results are in good agreement with the largest shifts
The different interaction mode of the nucleobases in the
L1 adducts can be related to the smaller dimension of this
receptor. In fact, in the case of the larger receptor L2, the
nucleotides assume a bent conformation, which allows the
simultaneous encapsulation of the phosphate chain within
the receptor cavity and the formation of π-stacking interact-
ions with phenanthroline. In the case of L1, the triphosphate
chain is still fitted within the receptor cavity, in a fashion
similar to that observed in the adducts with L2. If encapsulat-
7358 J. Org. Chem. Vol. 74, No. 19, 2009

Bazzicalupi et al.
JOCArticle
network of charge-assisted hydrogen-bonding interactions
with the ammonium groups, while its pyrimidine base is
sandwiched between the phenanthroline unit of the second
receptor and the thymine group of the other TTP anion. This
leads to the formation of a face-to-face π-stacked array of
two thymidine and two phenanthroline units, with interpla-
˚
nar distances of 3.3 A between phenanthroline and thymine
˚
planes and of 3.6 A between the two thymine units.
Of note, each phosphate chain is fitted inside a macro-
cyclic cavity affording hydrogen-bonding interactions bet-
ween all phosphate groups and the protonated amine
groups. Indirect hydrogen bonds contacts, i.e., mediated by a
bridging water molecule, are also observed between the Pβ
phosphate and both phenanthroline nitrogens (N4a or N4b)
and a single benzylic ammonium group (N10a or N10b).
Finally, the thymine carbonyl group O2a or O2b interacts via
hydrogen bonding with a protonated nitrogen of the receptor,
while the thymine nitrogen atom N2a or N2b is hydrogen
bonded to the Pβ phosphate of a different TTP unit. The
interaction modes observed in this crystal structure, such
as the inclusion of the phosphate chain within the cavity
of L2 and the π-pairing between phenanthroline and
thymine, resemble those observed in the calculated con-
formers for the nucleotides/L2 adducts.
Conversely, the formation of a dimeric assembly, not
found in aqueous solution, may reflect the scarce tendency
of thymine to give π-stacking interaction with the phenan-
throline of the receptor in the 1:1 adducts. As a consequence,
in the solid state the phosphate chain of each TTP unit is
enclosed in the cavity of one of the polyammonium receptor,
while the corresponding thymine moiety prefers to interact
via π-stacking with the phenanthroline of a second proto-
nated receptor and via hydrogen bonding and π-stacking
with a second TTP anion.
FIGURE 10. Crystal structure of the [(H6L2)2TTP2(H2O)2]4+ cation.
˚
Hydrogen bond distances (A): N2a O9b 2.75(1), N2b O10a
3 3 3
3 3 3
2.76(1), N6a O13b 2.63(2), N9b O9a 2.78 (3), N8a O7b
2.77(2), N8a O14b 2.73 (2), N5a O2a 2.96(1), N5a O9b
3 3 3 3 3 3 3 3 3
3 3 3 3 3 3 3 3 3
2.78(2), N5a O12b 2.82(2), N7a O6b 2.72(3), N9a O10b
3 3 3 3 3 3 3 3 3
2.61(3), N5b O2b 3.04(2), N5b O10a 2.69 (2), N5b O14a
3 3 3 3 3 3 3 3 3
2.75(3), N6b O13a 2.70(2), N10a Ow5 2.79(3), N8b O12a
3 3 3 3 3 3 3 3 3
2.83(3), N8b O6a 2.86(2), N7b O7a 2.72(3), N10b Ow6
2.76(3), N4a Ow5 2.89(2), Ow5 O10b 2.94(2), N4b Ow6
2.94(2), Ow6 O9a 2.87(2).
3 3 3
3 3 3
3 3 3
3 3 3
3 3 3
3 3 3
3 3 3
ion of the phosphate chains enables the formation of a
stabilizing network of hydrogen-bonding interactions, the
smaller dimension of L1 does not allow to the nucleobases to
assume simultaneously a spatial disposition “overlapped” to
the phenanthroline unit of the receptor, preventing the
formation of face-to-face π-stacking interactions.
To obtain further information on the structural features of
the adducts, we attempted to get crystals suitable for X-ray
analysis from aqueous solutions containing receptors
and nucleotides in 1:1 molar ratio. In the case of the system
TTP-L2 we obtained crystals of compound [(H6L2)2(TTP)2-
Although nucleotide complexation by synthetic polyam-
monium receptors has been widely investigated, only a single
crystal structure of this type of adducts has been previously
reported.8a At the same time, no crystal structure displaying
an inclusive binding mode of the anionic moiety of nucleo-
tides has been previously shown, although this binding
feature has been often proposed to justify the formation of
stable adducts between nucleotides and cyclic polyammo-
nium receptors.4,5
The resolution of this crystal structure allowed us to
perform a molecular modeling analysis of the adduct bet-
ween [H6L2]6+ and TTP in its tetraanionic form, using as
starting coordinates those obtained from the X-ray analysis.
In [(H6L2)2TTP2(H2O)2]4+, each polyammonium receptor
interacts simultaneously with one TTP anion via inclusion
within its cavity of the triphosphate chain and with a second
TTP anion essentially via π-stacking. Therefore, we per-
formed two different MD simulations by using either the
coordinates of the polyammonium receptor and of the TTP
anion interacting together in the inclusive fashion (indicated
as “macrocyclic unit 1” and “TTP anion 1” in Figure 10), or
the coordinates of the macrocycle and of TTP interacting
together mainly via π-stacking (“macrocyclic unit 1” and
“TTP anion 2” in Figure 10). In both cases, simulated
annealing leads to two different families of conformers with
structural characteristics almost equal to those previously
achieved using starting coordinates obtained by manual
docking (see Supporting Information, Tables S6 and S7
(H2O)2]Br4 7H2O, which was structurally characterized.
3
The crystal structure is constituted by a dimeric adduct
[(H6L2)2(TTP)2(H2O)2]4+ (Figure 10), four bromide anions
and seven crystallization water molecules. The [(H6L2)2-
(TTP)2(H2O)2]4+ adduct, featured by a non crystallographic
inversion symmetry, is formed by two protonated L2 recep-
tors, two TTP anions, and two water molecules, held
together by several hydrogen-bonding contacts and π-stack-
ing interactions. As shown in Figure 10, each protonated
receptor is characterized by a conformation folded along the
axis linking the benzylic nitrogen atoms, in a manner similar
to that observed in the calculated conformers of the
[(H6L2)(TTP)]2+ adduct. The two receptors assume a
head-to-tail cyclic disposition and define a three-dimens-
ional cavity, where two TTP anions are hosted. The tripho-
sphate chain of each TTP anion is encapsulated within the
macrocyclic cavity of one protonated receptor to give a
J. Org. Chem. Vol. 74, No. 19, 2009 7359

Bazzicalupi et al.
JOCArticle
to selectively sense ATP in a narrow pH range (pH 4.5-7),
thanks to a complexation induced quenching of fluores-
cence.
In principle, the quenching of the fluorescent emission of
L2 in the presence of ATP in the pH range 4.5-7 could occur
through an electron-transfer (PET) process involving either
the adenine moiety or a not protonated amine group of the
receptor. In fact, the electron-rich adenine unit may give rise
to a PET process to excited phenanthroline, a typical aro-
matic electron-rich moiety. This mechanism has been
already proposed for similar quenching effects upon ATP
binding to fluorogenic polyammonium cations.3b This hy-
1
pothesis is supported by MM calculations and H NMR
measurements, which clearly show that, among the four
nucleotides, adenine of ATP gives the strongest π-stacking
interactions with phenanthroline of L2, and assumes a
spatial position very close to that of phenanthroline, which
could favor the PET process. On the other hand, Figure 6
shows that the interaction between adenine and phenanthro-
line is not much affected by pH and it is present also at pH
values lower than 4.5, i.e., in the pH region where the ATP
adducts are emissive.
However, among the adducts with L2, ATP gives not only
the more robust π-stacked pairing, but also the strongest
charge-charge interactions between its triphosphate unit
and the polyammonium chain of the receptor. As already
observed in the absence of nucleotides at pH > 7, the
quenched status of phenanthroline is due to the presence of
a not protonated amine group. L2 becomes emissive below
pH 7, where all amine groups are protonated, and therefore,
their lone pairs are not available to quench the excited
fluorophore through a PET process. The fact that L2 in
the presence of ATP is not emissive in the pH range 4.5-7
may be related to the hydrogen-bonding and charge inter-
actions between the anionic phosphate moiety and the poly-
ammonium chain of the protonated receptor, which could
lead to an intracomplex proton transfer from one ammo-
nium group to a phosphate, making an amine lone pair
available for the quenching process in this pH range. This
mechanism has been already proposed to explain the
quenching of the emission of other polyammonium receptors
in the presence of phosphate anions.13 Of note, Figure 5
shows that the downfield shift observed for the signals of the
phosphate chain in the presence of ATP is maximum at
slightly acidic pH values, where the L2-ATP adduct is
quenched, but decreases remarkably below pH 4.5, where
protonation of the triphosphate moiety of the nucleotide
reduces its charge-charge interactions with the polyammo-
nium groups. This may prevent the process of proton
transfer from the polyammonium chain, renewing the fluor-
escence emission of phenanthroline. This observation sug-
gests that the quenching of the emission at slightly acidic pH
vales is probably due to a PET process involving the lone pair
of an amine group, which deprotonates upon a proton-
transfer process induced by the strong electrostatic and
hydrogen-bonding interactions present in this pH range.
L2 is also able to ratiometrically detect ATP.31 In fact,
addition of increasing amounts of ATP to a solution of L2 in
the pH range 5-7 leads to linear decrease of the fluorescence
emission of the receptor and the fluorescence of L2 is
completely quenched in the presence of 1 equiv of ATP.
The other triphosphate nucleotides produce only small
FIGURE 11. Emission spectra of L2 at selected pH values (a) and
pH-dependence of its emission intensity at 366 nm in the absence
(-9-) and in the presence of 1 equiv of ATP (-0-), CTP (-4-), GTP
(-O-) and TTP (-)-) ([L2]=[ATP]=[CTP]=[GTP]=[TTP]=2.5 ꢀ
10-5 M, 298 K, λexc 270 nm, 0.1 M NMe4Cl) (b).
and Figure S16). In fact, both families display the tripho-
sphate anion enclosed within the macrocyclic cavity, while
the nucleobase interacts either via π-stacking (A family) with
phenanthroline or via hydrogen bonding between one car-
bonyl oxygen thymine and ammonium group of the receptor
(B family). This second interaction mode is the preferred one
by TTP (the 68% of sampled conformers belong to the B
families), as already found by means of the manual docking
procedure. Although these results do not add new informat-
ion on binding mode of the receptor, they give confidence to
the overall minimization procedure used in the absence of
crystallographic data as starting coordinates.
Fluorescence Sensing of ATP. As discussed above, both
receptors exhibit an ON-OFF pH dependence of their fluor-
escence emission. In fact, they are not emissive in the alkaline
pH region and show the typical emission band of phenan-
throline only below pH 7-8 (Figure 2 and Supporting
Information, Figure S3). Analogous pH profiles are also
observed for the adducts of L1 with the four nucleotides.
Conversely, ATP gives significant changes in the pH depen-
dence of the emission spectra of the larger receptor L2
(Figure 11), while the pH dependence of the spectra is only
slightly affected by the presence of the other nucleotides. In
fact, the phenanthroline unit of L2 becomes emissive at
remarkably more acidic pH values in the presence of ATP
(below pH 4.5), while only minor displacements of the pH
profile toward acidic pH values are observed in the case of
CTP, GTP, and TTP.
In consequence, in the pH range 4.5-7 L2 is emissive in the
presence of CTP, TTP and GTP, while is quenched in the
presence of ATP (Figure 11b). Therefore, L2 is not only able
of selectively bind ATP over the other nucleotides, but also
7360 J. Org. Chem. Vol. 74, No. 19, 2009

Bazzicalupi et al.
JOCArticle
SCHEME 2. Synthetic Procedure To Achieve L1
FIGURE 12. Fluorescence intensity of L2 at 366 nm in the presence
of increasing amounts of ATP in the absence (-9-) and in
the presence of 1 eqiv of TTP (-4-), GTP (-O-) and CTP (-1-).
(λexc 270 nm, NMe4Cl 0.1 M, 298.1 K, pH=6, [L]=2.5 ꢀ 10-5).
decrease of the fluorescence emission (at maximum 18% for
CTP). Of note, ratiometrically sensing of ATP is scarcely
affected by the presence in solution of 1 equiv of the other
nucleotides. In fact, a competition experiment, carried out by
adding increasing amounts of ATP to a solution buffered at
pH 6 containing 1 equiv of TTP or GTP, shows a linear
decrease of the fluorescence emission of the fluorophore very
similar to that observed in the absence of these nucleotides.
In the presence of 1 equiv of CTP, the most efficient
competing agent for ATP, the fluorescence emission of L2
linearly decreases up to a 1:0.8 L2/ATP molar ratio and
achieves a constant value for L2/ATP molar ratios greater
than 1.4 (Figure 12).
These results indicate not only that the sensing ability for
ATP is scarcely affected by the presence of the other nucleot-
ides, but also confirms, at least qualitatively, that the binding
ability of L2 toward nucleotides decreases in the order ATP
. CTP > TTP = GTP, in keeping with the results obtained
by means of potentiometric measurements.
strong interactions with the guest species. Despite the similar
structure of nucleotide triphosphates, ATP gives the most stable
adducts with L2, thanks to a pronounced cooperative role
played in the binding process by the phosphate chain and the
adenine unit, which give rise respectively to a robust hydrogen-
bonding network with the polyammonium chain of L2 and to
strong π-stacking interactions with phenanthroline. This bind-
ing mode justifies not only the higher stability of the ATP-L2
adducts, but also the quenching of emission of phenanthroline,
which takes place via a PET process from the receptor to the
excited phenanthroline. Analysis of pH-dependence of the
fluorescence emission shows that quenching of phenanthroline
is induced by ATP binding only in a narrow pH range, between
pH 4.5 and 7, that is the same pH region where the phosphate
moiety of ATP gives the strongest interactions with the ammo-
nium chain of L2. This suggests that the quenching effect is
likely to be related to a proton-transfer process from an
ammonium group to the anionic phosphate chain.
Concluding Remarks
The comparison of the binding properties of L1 and L2
shows that slight differences in the receptor and/or nucleo-
tide structures can determine the binding mode as well as the
fluorescence emission of the adducts. The host-guest
adducts formed by the highly protonated forms of L1 and
L2 present similar inclusive coordination modes of the
phosphate chain within the receptor cavity, which allow
the formation of an hydrogen-bonding network between
the ammonium groups of L1 and L2 and the anionic
phosphate groups of nucleotides. In both the adducts with
L1 and L2, all nucleotides assume folded conformations,
which allow the simultaneous encapsulation of their tripho-
sphate chain and the interaction of the nucleobases with the
receptor. However, only receptor L2 is larger enough to
allow the π-pairing of nucleobases with phenanthroline. In
fact, despite the bent conformation assumed by nucleotides,
the nucleobases cannot achieve an appropriate disposition to
give π-stacking with the phenanthroline of L1, due to the
somewhat smaller dimension of this receptor. The adducts
with L2 are consequently more stable than those with L1.
To achieve selective recognition of a given substrate, the
receptor binding sites need to be simultaneously involved in
Experimental Details
Synthesis of Receptors. The procedure for the synthesis of
receptors L1 is depicted in Scheme 2. L2,31 2,9-bis-
(bromomethyl)-1,10-phenanthroline,39 and tosylated amine
1,5,9-tritosyl-1,5,9-tetraazanonane (1)40 were prepared as pre-
viously described. The reaction of pertosylated polyamine
4 with 2,9-bis(bromomethyl)-1,10-phenanthroline in dry acet-
onitrile at reflux and in presence of K2CO3 afforded macrocycle
4 in moderate yield. Deprotection of the nitrogen atoms was
achieved by treatment of 4 with HBr in acetic acid and in
(39) Chandler, C. J.; Deady, L. W.; Reiss, J. A. J. Hetercycl. Chem. 1981,
18, 599–601.
~
(40) (a) Bencini, A.; Burguete, M. I.; Garcia-Espana, E.; Luis, S. V.;
Miravet, J. F.; Soriano, C. J. Org. Chem. 1993, 58, 4749–4753. (b) Aguilar, J.;
ꢀ
~
Diaz, P.; Escarti, F.; Garcia-Espana, E.; Gil, L.; Soriano, C.; Verdejo, B.
Inorg. Chim. Acta 2002, 339, 307–316.
J. Org. Chem. Vol. 74, No. 19, 2009 7361

Bazzicalupi et al.
JOCArticle
presence of phenol. Deprotected macrocyclic receptor L1 was
obtained in good yield as its hydrobromide salt.
48.1, 46.7, 46.4, 46.3, 28.3, 28.1, 28.0, 21.2. Anal. Calcd for
C61H69N7O10S5: C, 60.03; H, 5.70; N, 8.03. Found: C, 59.9; H,
5.9; N, 7.9.
1,17-Phthalimido-5,9,13-tritosyl-5,9,13-triazaheptadecane (2).
Tosylated amine 1 (15.8 g, 26.6 mmol) and K2CO3 (29.7 g, 215 mmol)
were suspended in refluxing dry acetonitrile (150 mL). To this mixture
was added dropwise a solution of N-(3-bromopropyl)phthalimide
(14.7 g, 55 mmol) in dry acetonitrile (150 mL). After the addition was
complete, the suspension was refluxed for 48 h and then filtered off.
The solvent was evaporated, and the residue suspended in refluxing
ethanol to give 2 as a white solid. Yield: 23.2 g, (90%). Mp: 265-
267 °C. 1H NMR (CDCl3, 300 MHz) δ (ppm): 7.83-7.81 (m, 4H),
7.72-7.64 (m, 10H), 7.34-7.28 (m, 6H), 3.69 (t, J=7.1 Hz, 4H),
3.23-3.13 (m, 12H), 2.43 (s, 9H), 1.97-1.87 (m, 8H). 13C NMR
(CDCl3, 75.4 MHz) δ (ppm): 168.1, 143.2, 136.0, 133.9, 132.0, 129.7,
127.1, 123.1, 46.8, 46.6, 35.6, 28.6, 27.8, 21.5. Anal. Calcd for
C49H53N5O10S3: C, 60.79; H, 5.52; N, 7.23. Found: C, 60.9; H, 5.6;
N, 7.3.
5,9,13-Tritosyl-1,5,9,13,17-pentaazaheptadecane (3). A mixt-
ure of 2 (3.48 g, 3.6 mmol) and hydrazine hydrate (85% in water,
8 mL) in THF (400 mL) was refluxed overnight and then cooled
to room temperature and the resulting solid filtered off. The
filtrate was concentrated to dryness, affording the deprotected
amine 3 as a white solid. Yield: 2.51 g (98%). Mp: 150-152 °C.
1H NMR (CDCl3, 300 MHz) δ (ppm): 7.67-7.53 (m, 6H), 7.23
(m, 6H), 3.12-3.03 (m, 12H), 2.64 (t, J = 6.6 Hz, 4H), 2.34
(s, 9H), 1.83-1.75 (m, 4H), 1.62-1.54 (m, 4H). 13C NMR
(CDCl3, 75.4 MHz) δ (ppm): 143.3, 143.1, 135.9, 135.7, 129.6,
129.5, 127.0, 126.7, 46.8, 46.5, 38.9, 32.1, 28.7, 21.4. Anal. Calcd
for C33H49N5O6S3: C, 55.98; H, 6.98; N, 9.89. Found: C, 56.1;
H, 6.9; N, 9.7.
1,5,9,13,17-Pentatosyl-1,5,9,13,17-pentaazaheptadecane (4).
To a stirred solution of tosyl chloride (6.67 g, 35 mmol) in
pyridine (55 mL) at 0 °C was added dropwise a solution of the
amine 3 (8.5 g, 12 mmol) in 50 mL of pyridine. After the addition
was complete, the red solution was stirred at room temperature
overnight and then poured in a water-ice mixture: an orange,
gummy solid was obtained; it was ground and stirred in water
until it turned into a pale pink powder, which was filtered and
dried in vacuo at 40 °C overnight. The product 4 was recrys-
tallized from hot CHCl3 by addition of cold EtOH and obtained
as a white solid (10.3 g, 85% yield). Mp: 203-205 °C. 1H NMR
(CDCl3, 300 MHz) δ (ppm): 7.84-7.73 (m, 12H), 7.43-7.35
(m, 12H), 3.24-3.12 (m, 16H), 3.09-3.06 (m, 4H), 2.53-2.51
(m, 18H), 2.02-1.88 (m, 8H). 13C NMR (CDCl3, 75.4 MHz)
δ (ppm): 143.5, 143.4, 143.2, 136.8, 135.4, 129.8, 129.6, 127.1,
127.0, 47.5, 47.2, 46.7, 40.2, 29.4, 29.1, 21.5. Anal. Calcd for
C47H61N5O10S5: C, 55.54; H, 6.05; N, 6.89. Found: C, 55.7; H,
6.2; N, 6.6.
2,6,10,14,18-Pentaaza[19](2,9)cyclo-1,10-phenanthrolino-
phane Hexahydrobromide (L1 6HBr). The tosylated macrocycle
3
5 (1.23 g, 1.0 mmol) and phenol (9.4 g, 100 mmol) were dissolved
in HBr/AcOH 33% (100 mL); the reaction mixture was stirred
and refluxed for 20 h until a white solid formed. The reaction
mixture was cooled to room temperature, 100 mL of dichlor-
omethane was added to complete the precipitation, and the
mixture was stirred for an additional 1 h. The solid was filtered
and washed with dichloromethane to remove the residual
phenol. Receptor L1, as hydrobromide salt, was then recrystal-
lized from a mixture of EtOH/water 3:1 (v/v), filtered/ and dried
at 40 °C overnight in the presence of KOH (559 mg, 55% yield).
Mp: 200 °C dec. 1H NMR (D2O, 300 MHz, pH=1.9) δ (ppm):
8.46 (d, J=8.4 Hz, 2H), 7.92 (s, 2H), 7.75 (d, J=8.1 Hz, 2H), 4.68
(s, 4H), 3.29 (m, 4H), 3.15 (m, 12H), 2.19 (m, 4H), 2.03 (t, J=7.0 Hz,
4H). 13C NMR (D2O, 75.4 MHz, pH = 1.9) δ (ppm): 151.4,
144.6, 139.4, 129.2, 127.1, 123.6, 52.5, 45.2, 45.95, 44.86, 42.6,
23.0, 22.4. ESI-MS (m/z) 450.33 (100) ([M + H]+), 226.2 (58)
(M + 2H)2+. Anal. Calcd for C26H39N7 6HBr: C, 33.39;
3
H, 4.85; N, 10.49. Found: C, 33.1; H, 4.8; N, 10.2.
Synthesis of the [(H6L2)2(TTP)2(H2O)2]Br4 7H2O Adduct.
L2 6HBr (10 mg, 0.01 mmol) and thymidine triphosphate
3
3
monosodium salt (4.8 mg, 0.01 mmol) were dissolved in water,
and the pH of the solution was raised to 6 with 0.1 M NaOH.
The solution was then kept under an acetone atmosphere until
the formation of colorless crystals suitable for X-ray analysis.
Yield: 7.0 mg (56.9%) Anal. Calcd for C78H148Br4N20O37P6:
C, 38.13; H, 6.06; N, 11.37. Found: C, 38.3; H, 6.2; N, 11.2.
Potentiometric Measurements. Equilibrium constants for
protonation and complexation reactions with L were deter-
mined by pH-metric measurements at 298.1 ( 0.1 K in 0.1 M
NMe4Cl, by using equipment and procedures which have been
already described.8b,31 In the experiments to determine the
stability of the adducts with L1 and L2, the receptor concentra-
tions were varied from 5 ꢀ 10-4 to 5 ꢀ 10-3 M, while the
concentration of nucleotides was varied in the range 4 ꢀ 10-4
-
9 ꢀ 10-3 M. At least three measurements (about 100 data points
each one) were performed for each system in the pH range
2.5-10.5, and the relevant emf data were treated by means of
the computer program HYPERQUAD.37 Full details are given
within the Supporting Information.
Spectrophotometric and Spectrofluorimetric Measurements.
Absorption and fluorescence emission spectra were recorded
as previously described.30 In the measurements carried out at
different pH values, HCl and NaOH were used to adjust the pH
values which were measured on a pH meter. TRIS buffer (1 mm)
was used in the titrations performed at pH 6.0. In the competi-
tion experiments, successive readings of the emission intensity
were carried out after each addition of CTP, TTP, or GTP to the
ATP solution to ensure that the equilibrium was reached.
NMR Spectroscopy. 1H (300 MHz) and 13C (75 MHz) in
CDCl3 and D2O solutions at different pH values were recorded
at 298 K on a 300 MHz spectrometer. To adjust the pD, small
amounts of 0.01 m NaOD and DCl were added to the solutions.
The pH was calculated by the measured pD values by using the
following formula: pH=pD - 0.40.41 2D correlation experi-
2,6,10,14,18-Pentatosyl-2,6,10,14,18-pentaaza[19](2,9)cyclo-
1,10-phenanthrolinophane (5). To a refluxing, stirred suspension
of the tosylated polyamine 4 (2.54 g, 2.5 mmol) and K2CO3 (3.45 g,
25 mmol) in dry acetonitrile (250 mL) and under nitrogen atmo-
sphere was added a suspension of 2,9-bis-(bromomethyl)-1,10-
phenanthroline (1.0 g, 2.75 mmol) in dry acetonitrile (250 mL)
during 4 h. At the end of the addition, the mixture was stirred and
refluxed for additional 2 h and then cooled to room temperature.
The suspension was filtered on Celite and washed with acetonitrile,
and the filtrate was evaporated to obtain a solid, which was
purified by column chromatography on neutral aluminum oxide,
using a mixture of PE/AcOEt 1:1.5 as the eluant. The fractions
containing 6 (Rf=0.5) were evaporated to dryness; 5 was obtained
1
ments were performed to assign the H NMR signals. Com-
plexation-induced 31P and 1H NMR chemical shifts (CIS, ppm)
were measured as the difference δOBS - δnucleotide where δOBS is
the chemical shift of a signal measured in D2O solutions contain-
ing receptor and substrate in a 1:1 molar ratio (both 5 ꢀ 10-3 M,
in these conditions all adducts are completely formed in solution
1
as a pale brown solid (1.08 g, 35% yield). Mp: 195-197 °C. H
NMR (CDCl3, 300 MHz) δ (ppm): 8.29 (d, J=8.4 Hz, 2H), 8.18
(d, J = 8.1 Hz, 2H), 7.70-7.64 (m, 6H), 7.62-7.56 (m, 8H),
7.34-7.26 (m, 8H), 4.78 (s, 4H), 3.24 (t, J= 6.9 Hz, 4H), 2.99
(m, 4H), 2.66 (m, 4H), 2.55 (s, 9H), 2.41 (s, 6H), 1.84 (m, 4H), 165
(m, 4H). 13C NMR (CDCl3, 75.4 MHz) δ (ppm): 157.7, 143.2,
143.4, 143.0, 137.5, 135.9, 135.4, 129.6, 127.9, 127.1, 77.5, 55.3,
(41) Covington, A. K.; Paabo, M.; Robinson, R. A.; Bates, R. G. Anal.
Chem. 1968, 40, 700–709.
7362 J. Org. Chem. Vol. 74, No. 19, 2009

Bazzicalupi et al.
JOCArticle
and no amounts of uncomplexed receptor or substrate is present
in solution), and δnucleotide is the chemical shift of the corre-
sponding signal of the not complexed receptor or substrate.
Crystal Structure Analyses. Data for the X-ray structural
Computational Details. MD calculations were performed by
using the AMBER3 forcefield, as implemented in the Hyperch-
em 7.51 package,46 by using the procedure and method which
has been already described.8b,31 Starting conformations of the
analysis of [(H6L2)2TTP2(H2O)2]Br4 7H2O were collected on
an Oxford Diffraction Xcalibur3 diffractometer equipped with
[H6L2(S)]2þ or [H5L1(S)]þ adducts (S=ATP4-, GTP4-, CTP4-
,
3
TTP4--), obtained by manual docking of the minimized con-
former of the receptor to the substrate (minimum distance
CCD area detector and graphite-monochromated Mo KR
˚
˚
radiation (λ = 0.71069 A, T = 150 K). Data collection was
between the atoms of receptor and substrate >5 A) and, in
performed using ω scan with the CrysAlis CCD program.42
Data reduction was carried out with the CrysAlis Red pro-
gram,43 and an empirical absorption correction was applied
using spherical harmonics, implemented in SCALE3 AB-
SPACK scaling algorithm.43
the case of the [H6L2(TTP)]2þ complex, derived from the X-ray
crystal structure, were freely minimized. Full details are given
within the Supporting Information.
Acknowledgment. Financial support from MIUR (PRIN
2007) is gratefully acknowledged
Crystals of the compound belong to the triclinic family
˚
(C78H148Br4N20O37P6, Mw =2463.62, P1, a=13.1824(8) A,
Supporting Information Available: Full details for potentio-
metric measurements and molecular mechanics calculations;
overall stability constants of the adducts; list of files containing
the atomic coordinates of the minimized conformers; hydrogen-
bonding distances in the low energy conformers of the adducts
of [H5L1]5þ and [H6L1]6þ with nucleotides; titration data and fit
curves for potentiometric titrations performed to determine the
protonation constants of L2; pH dependence of the chemical
shifts of the benzylic protons and of the fluorescence emission at
365 nm of L2; distribution diagrams of the adducts of L1 and L2
with CTP, TTP, and GTP; pH dependence of the 31P NMR and
1H NMR chemical shifts of the phosphate chain and of the
nucleobases of CTP, TTP, and GTP in the absence and in the
presence of L1 and L2; pH dependence of the 1H NMR chemical
shifts of the phenanthroline protons of L1 and L2 in the absence
and in the presence of CTP, TTP, and GTP; labeled lowest
energy conformers of the adducts of [H5L1]5þ and [H6L1]6þ
with nucleotides; ORTEP drawing of the [(H6L2)2-
(TTP)2(H2O)2]4þ cation; 1H and 13C NMR spectra of com-
pounds 2-5 and of L1; atomic coordinates of the calculated low
energy conformers in MDL_Mol format. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
˚
˚
b=14.2832(9) A, c=17.092(1) A, R=91.782(5)°, β=109.531(5)°,
γ =115.968(6)°; V=2666.3(3) A ; Z=1; Fcalc=1.534 mg/m3; μ=
3
˚
1.689 mm-1; F(000) = 1282; θmax = 24.71). Structures were
solved by direct methods (SIR2004)44 and refined against F2
by using SHELXL-97,45 with non-hydrogen non-carbon atoms
anisotropic. Carbon atoms were refined isotropically in order to
avoid a too low data/parameters ratio. Hydrogen atoms linked
to carbon or nitrogen atoms were introduced in calculated
position and refined in riding mode. Hydrogen atoms belonging
to the solvent water molecules were not localized in the final ΔF
map and not introduced in the calculation. All six-membered
aromatic rings were restrained to idealized hexagons. Due to low
intensity of data at high resolution, only data up to senθ/λ=0.59
were used during refinement. Agreement factors at the end of
refinement were R1=0.0777, wR2=0.1686 (4847 reflections
with I > 2σ(I) - 846 parameters refined); R1=0.1769, wR2=
0.2180 (all data). The Flack parameter at the end of refinement
was -0.019(16).
(42) CrysAlis CCD, Oxford Diffraction Ltd., Version 1.171.32.5, 2008.
(43) CrysAlis RED, Oxford Diffraction Ltd., Version 1.171.32.5, 2008.
(44) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.; Giacov-
azzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna, R.
J. Appl. Crystallogr. 1999, 32, 115–119.
(46) Hyperchem β1 release 7.51 for Windows MM System, Hypercube,
Inc., Gainesville, FL, 2002.
€
(45) Sheldrick, G. M., SHELX-97, Gottingen, 1997.
J. Org. Chem. Vol. 74, No. 19, 2009 7363