Dendrimers functionalized with a single fluorescent dansyl group attached 'off center': Synthesis and photophysical studies

Article abstract of DOI:10.1021/ja000949l

A series of three new fluorescent dendrimers containing a single, focally located dansyl group and 3 (1), 9 (2), and 27 (3) carboxylic acid groups in their peripheries were synthesized and characterized. The photophysical properties of these dendrimers were investigated in aqueous solution. The host-guest interactions of the dendrimers through their dansyl subunits with β-cyclodextrin and polyclonal anti-dansyl antibodies were also investigated by various methods. Photophysical measurements on the dendrimers demonstrate that the dansyl residue is progressively shielded from the solvent as the dendrimer generation increases, resulting in marked changes in spectral features, fluorescence quantum yields, excited-state fluorescence lifetimes, radiative and nonradiative decay rates, and rotational reorientation times. The excited-state intensity decay kinetics for 1-3 are well described by a single exponential. Contrary to the popularly held belief that lower generation dendrimers are 'floppy' species in solution, the molecular motions of 1-3 are described by a single rotational reorientation time. Access to the dansyl moiety is impeded with increasing dendrimer size as the dendrimer mass affords a significant degree of protection from binding by nonselective (β-cyclodextrin (βCD)) and selective (anti-dansyl antibody) hosts for the dansyl residue. The equilibrium constant for β-CD binding of the dansyl residue in 1 is ~2.5-fold lower than that for binding to dansylamine (DA). Dendrimers 2 and 3 do not associate with β-CD at all. Anti-dansyl antibodies can bind to the dansyl residue in dendrimers 1-3 with remarkably large binding affinities. The equilibrium constant for the antibody complex decreases systematically from 5.0 x 10⁷ M^-1 for DA to 1.5 x 10⁶ M^-1 for 3.

Full text of DOI:10.1021/ja000949l

VOLUME 122, NUMBER 26
J ULY 5, 2000
© Copyright 2000 by the
American Chemical Society
Dendrimers Functionalized with a Single Fluorescent Dansyl Group
Attached “Off Center”: Synthesis and Photophysical Studies
Claudia M. Cardona,† Julio Alvarez,† Angel E. Kaifer,*,† Tracy Donovan McCarley,‡
Siddharth Pandey,§ Gary A. Baker,§ Neil J. Bonzagni,§ and Frank V. Bright*,§
Contribution from the Center for Supramolecular Science and Department of Chemistry, UniVersity of
Miami, Coral Gables, Florida 33124-0431, Choppin Laboratories of Chemistry, Louisiana State
UniVersity, Baton Rouge, Louisiana 70803-1804, and Department of Chemistry, Natural Sciences Complex,
UniVersity of Buffalo, State UniVersity of New York, Buffalo, New York 14260-3000
ReceiVed March 16, 2000
Abstract: A series of three new fluorescent dendrimers containing a single, focally located dansyl group and
3 (1), 9 (2), and 27 (3) carboxylic acid groups in their peripheries were synthesized and characterized. The
photophysical properties of these dendrimers were investigated in aqueous solution. The host-guest interactions
of the dendrimers through their dansyl subunits with â-cyclodextrin and polyclonal anti-dansyl antibodies
were also investigated by various methods. Photophysical measurements on the dendrimers demonstrate that
the dansyl residue is progressively shielded from the solvent as the dendrimer generation increases, resulting
in marked changes in spectral features, fluorescence quantum yields, excited-state fluorescence lifetimes, radiative
and nonradiative decay rates, and rotational reorientation times. The excited-state intensity decay kinetics for
1-3 are well described by a single exponential. Contrary to the popularly held belief that lower generation
dendrimers are “floppy” species in solution, the molecular motions of 1-3 are described by a single rotational
reorientation time. Access to the dansyl moiety is impeded with increasing dendrimer size as the dendrimer
mass affords a significant degree of protection from binding by nonselective (â-cyclodextrin (â-CD)) and
selective (anti-dansyl antibody) hosts for the dansyl residue. The equilibrium constant for â-CD binding of the
dansyl residue in 1 is ∼2.5-fold lower than that for binding to dansylamine (DA). Dendrimers 2 and 3 do not
associate with â-CD at all. Anti-dansyl antibodies can bind to the dansyl residue in dendrimers 1-3 with
remarkably large binding affinities. The equilibrium constant for the antibody complex decreases systematically
from 5.0 × 107 M-1 for DA to 1.5 × 106 M-1 for 3.
Introduction
these systems offers interesting possibilities concerning light
harvesting, energy storage, and conversion.2 As a recent
example, Fre´chet and co-workers have synthesized dendritic
macromolecules containing multiple light-harvesting chro-
mophores in their periphery that, upon photoexcitation, can
Dendrimer functionalization is a very active area of research
in modern chemistry.1 In this regard, functionalization of
dendritic structures with luminescent groups has attracted
considerable attention because the molecular architecture of
(1) For recent reviews on dendrimer research, see: (a) Bosman, A. H.;
Janssen, H. M.; Meijer, E. W. Chem. ReV. 1999, 99, 1665. (b) Newkome,
G. R.; He, E.; Moorefield, C. N. Chem. ReV. 1999, 99, 1689. (c) Fischer,
M.; Vo¨gtle, F. Angew. Chem., Int. Ed. 1999, 38, 884.
† University of Miami.
‡ Louisiana State University.
§ State University of New York.
10.1021/ja000949l CCC: $19.00 © 2000 American Chemical Society
Published on Web 06/16/2000

6140 J. Am. Chem. Soc., Vol. 122, No. 26, 2000
Cardona et al.
Scheme 1. Synthesis of Second Generation Dansyl
transfer energy efficiently to a single central core.3 The excited
core, in turn, releases the excitation energy by fluorescing.
While synthetic progress with functionalized dendrimers
continues at a fast pace, detailed physicochemical studies on
their properties lag somewhat behind. In this work, we report
the synthesis of a series of unsymmetric dendrimers (first,
second, and third generations, see structures below) containing
a single dansyl group4 covalently attached “off center”, that is,
at a focal point in these macromolecules. Taking advantage of
the well-known fluorescent properties of the dansyl group,5 we
have carried out a detailed photophysical investigation to
ascertain the effect of the dendrimer structures on the microen-
vironment that surrounds the dansyl residues and on the
dynamics of these macromolecular systems. We have also
investigated the effects of the dendrimer structure on the binding
interactions of the dansyl residue (guest) with â-cyclodextrin6
(â-CD) and polyclonal anti-dansyl antibodies (hosts). All of
these data are of considerable interest to improve our funda-
mental understanding of dendrimer properties. Furthermore, the
structure of our dendrimers is well-suited to investigate site
isolation properties in dendrimers and related macromolecules.
The type of dendritic matrix used in this work, first developed
by Newkome and co-workers,7 has been less frequently
investigated than the more commonly utilized Fre´chet-type
dendrimers.3,8
Dendrimer 2
with the corresponding amine building blocks to yield lipophilic
forms of the dendrimers having tert-butyl groups on their
periphery. Hydrolysis of these lipophilic dendrimers with formic
acid leads to the more hydrophilic macromolecules 1, 2, and 3,
having 3, 9, and 27 surface carboxylic acid groups, respectively.
This procedure is illustrated by the synthesis of dendrimer 2,
shown in Scheme 1. The overall dendrimer yields range from
44 to 67%. MALDI-TOF mass spectra were extremely important
for the characterization of dendrimers 1-3, providing verifica-
tion of the purity of the precursor amine building blocks as well
as of the completeness of the final hydrolysis process. We also
prepared nitrotriacid 4 and nitrononaacid 5 for control experi-
ments regarding cyclodextrin binding studies.
Results and Discussion
Synthesis. The synthetic procedures for the preparation of
the dansyl dendrimers follow those previously reported by us
for the synthesis of unsymmetric ferrocene-containing dendrim-
ers.9 Very briefly, dansyl chloride was reacted in dry acetonitrile
(2) (a) Balzani, V.; Campagna, S.; Denti, G.; Juris, A.; Serroni, S.;
Venturi, M. Acc. Chem. Res. 1998, 31, 26. (b) Jiang, D.-L.; Aida, T. J.
Am. Chem. Soc. 1998, 120, 10895. (c) Devadoss, C.; Bharathi, P.; Moore,
J. S. J. Am. Chem. Soc. 1996, 118, 9635.
(3) Adronov, A.; Gilat, S. L.; Fre´chet, J. M. J.; Ohta, K.; Neuwahl, F.
V. R.; Fleming, G. R. J. Am. Chem. Soc. 2000, 122, 1175.
(4) Multiple dansylation of poly(propylene imine) dendrimers has been
recently reported: Archut, A.; Gestermann, S.; Hesse, R.; Kauffmann, C.;
Vo¨gtle, F. Synlett 1998, 546.
(5) (a) Hoenes, G.; Hauser, M.; Pfleiderer, G. Photochem. Photobiol.
1986, 43, 133. (b) Li, Y.-H.; Chan, L.-M.; Tyer, L.; Moody, R. T.; Himel,
C. M.; Hercules, D. M. J. Am. Chem. Soc. 1975, 97, 3118.
(6) For a series of recent reviews on the properties, uses and applications
of cyclodextrin hosts, see: Chem. ReV. 1998, 98, 8, issue 5.
(7) Newkome, G. R.; Behera, R. K.; Moorefield, C. N.; Baker, G. R. J.
Org. Chem. 1991, 26, 4376.
(8) Hawker, C. J.; Wooley, K. L.; Fre´chet, J. M. J. J. Am. Chem. Soc.
1993, 115, 4375.
(9) Cardona, C. M.; McCarley, T. D.; Kaifer, A. E. J. Org. Chem. 2000,
65, 1857.
Steady-State Fluorescence Measurements. Absorbance and
fluorescence maxima for dansylamide (DA) and dendrimers 1-3
(1 µM) dissolved in aqueous buffer are presented in Table 1.
The hypsochromic shifts of the fluorescence spectra and the
decrease in the Stokes shift suggest that the dansyl residue is
systematically encountering a less polar microenvironment as
the dendrimer generation increases from 1 to 2 to 3. Although

Synthesis and Photophysical Studies of Dendrimers
J. Am. Chem. Soc., Vol. 122, No. 26, 2000 6141
Table 1. Measured and Calculated Photophysical Parameters for Dansylamine (DA) and Dansyl-Labeled Dendrimers 1, 2, and 3 Dissolved in
Aqueous Buffer (pH ) 7)
solute
λAbsmax (nm)a
λFlumax (nm)b
Φf
τf (ns)
kr (107 s-1
)
knr (108 s-1
)
φ (ns)
V (nm3)c
DA
1
2
324
329
329
324
548
547
542
535
0.028 ( 0.002
0.044 ( 0.003
0.068 ( 0.003
0.086 ( 0.003
3.17 ( 0.02d
4.90 ( 0.05
6.32 ( 0.05
7.94 ( 0.06
0.88 ( 0.06
0.90 ( 0.06
1.08 ( 0.05
1.08 ( 0.04
3.07 ( 0.02
1.95 ( 0.02
1.47 ( 0.01
1.15 ( 0.01
0.07 ( 0.01
0.19 ( 0.02
0.59 ( 0.03
1.86 ( 0.05
0.29 ( 0.04
0.77 ( 0.08
2.40 ( 0.12
7.57 ( 0.20
3
a Absorption maxima for repeat measurements are reproducible to within (1 nm. b Fluorescence maxima have a (2 nm imprecision. c Hydrodynamic
volumes (V) are calculated from the Debye-Stokes-Einstein expression assuming viscosity η ) 1 mPa‚s at 22 °C. d The actual intensity decay for
DA is best described by a double exponential decay law; the average value is reported here. The double exponential decay arises from the simultaneous
emission from the locally excited and twisted intramolecular charge transfer states.
increasing the dendrimer generation might enhance the local
polarity surrounding the dansyl residue because of the introduc-
tion of additional amide or carboxylate groups in proximity to
the dansyl moiety, our data indicate that the enhanced hydro-
carbon character predominates. Thus, an increase in dendrimer
generation results in a less polar microenvironment surrounding
the dansyl residue relative to DA dissolved in aqueous buffer.
Fluorescence Quantum Yields. The fluorescence quantum
yields,Φf, of DA and 1-3 were determined (Table 1) by using
the Parker and Rees method10 with quinine sulfate dissolved in
0.1 N H2SO4 serving as the reference standard (Φs ) 0.577).
The results of this experiment show that the Φf value for the
dansyl residue increases in the order DA < 1 < 2 < 3 and the
increase parallels the hypsochromic shifts in the fluorescence
spectra. It is important to mention here that the Φf value
determined in this work for DA (0.028 ( 0.002) is statistically
equivalent to the value reported for ꢀ-dansyl-lysine (0.026-
0.029) dissolved in water.11 To provide the reader with a better
feel for the increases in effective hydrophobicity surrounding
the dansyl residue on going from DA to 3, we note that 1
and 3 dissolved in aqueous buffer exhibit a Φf value similar
to that measured for ꢀ-dansyl-lysine dissolved in 90:10 and
80:20 (v:v) H2O:dioxane mixtures, respectively.11
Figure 1. Typical multifrequency phase-modulation data for dendrim-
ers 1 (9), 2 (2), and 3 (b) dissolved in aqueous buffer (pH ) 7.0).
The solid curves represent the best fit between each data set and a
single-exponential decay law.
Excited-State Fluorescence Intensity Decay Kinetics. Fig-
ure 1 presents typical multifrequency phase-modulation12-16 data
sets for 1-3 dissolved in aqueous buffer. The symbols cor-
respond to dendrimer 1 (9), 2 (2), and 3 (b). The solid curves
that pass through the data points are the best fits between the
experimental data and a single-exponential decay law. In all
cases, the time-resolved fluorescence intensity decay for 1-3
dissolved in buffer was well described by a single fluorescence
lifetime, τf (ø2 e 1.07). Table 1 collects the recovered τf values.
These results show that the fluorescence lifetime for the dansyl
residue increases significantly as we proceed from DA to 3.
Table 1 also collects the radiative, kr, and nonradiative rates,
knr,17,18 for DA and dendrimers 1-3 dissolved in aqueous buffer.
These results show that the reason for the increase in τf on going
from DA to 3 is a consequence of two factors: (1) a slight
(∼20%) increase in kr and (2) an almost 3-fold decrease in knr.
Figure 2. Typical multifrequency polarized modulation ratio and
differential polarized phase angle data for DA (b), 1 (9), 2 (2), and
3 (1) dissolved in aqueous buffer (pH ) 7.0). The solid curves represent
the best fit between each data set and a isotropic rotor model with one
rotational reorientation time.
(10) Parker, C. A.; Rees, W. T. Analyst 1962, 87, 83.
(11) Parker, C. W.; Yoo, T. J.; Johnson, M. C.; Godt, S. M. Biochemistry
1967, 6, 3408.
(12) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 2nd ed.;
Kluwer Academic/Plenum Publishers: New York, 1999.
(13) Bright, F. V.; Betts, T. A.; Litwiler, K. S. Crit. ReV. Anal. Chem.
1990, 21, 389.
These data indicate that successive dendrimer generations result
in a shutting down of nonradiative pathways (e.g., twisted
intramolecular charge transfer from the excited manifolds)5 of
the dansyl residue with very little change in radiative path-
ways.17,18
Figure 2 presents typical multifrequency polarized modulation
ratio and differential polarized phase angle12-16 data for DA
and dendrimers 1-3 dissolved in aqueous buffer. The symbols
correspond to DA (b) and dendrimers 1 (9), 2 (2), and 3 (1).
The solid curves that pass through the data points are the best
fits between the experimental data and an isotropic rotor model
with a single rotational reorientation time, φ. In all cases, the
time-resolved anisotropy decay for DA and 1-3 dissolved in
(14) Gratton, E.; Jameson, D. M.; Hall, R. D. Annu. ReV. Biophys.
Biophys. Bioeng. 1984, 13, 105.
(15) Jameson, D. M.; Gratton, E.; Hall, R. D. Appl. Spectrosc. ReV. 1984,
20, 55.
(16) Lakowicz, J. R.; Lackzo, G.; Gryczynski, I.; Szmacinski, H.; Wiczk,
W. J. Photochem. Photobiol. B: Biol. 1988, 2, 295.
(17) Turro, N. J. Modern Molecular Photochemistry; University Science
Books: Sausalito, CA, 1991.
(18) Birks, J. B. Photophysics of Aromatic Molecules; Wiley-Inter-
science: London, 1970.

6142 J. Am. Chem. Soc., Vol. 122, No. 26, 2000
Cardona et al.
Figure 3. Chemical shift for the protons of 1 (1 mM) in D2O buffered
at pD 7.0. The values in parentheses indicate the shifts induced by the
addition of 4 mM â-CD.
Figure 4. Chemical shift for the protons of 2 (1 mM) in D2O buffered
at pD 7.0. The values in parentheses indicate the shifts induced by the
addition of 4 mM â-CD.
buffer was well-described by a single φ (ø2 e 1.12). Table 1
collects the recovered φ values. These results show that there
is no evidence for the dansyl residue moving independently from
the dendrimer; the dansyl unit moves in registry with the
dendrimer’s molecular framework.
The Debye-Stokes-Einstein expression12 provides a con-
venient way to estimate the hydrodynamic volume, V, of the
reorienting species. Table 1 collects the recovered V values for
DA and 1-3. Examination of Table 1 reveals that V increases
systematically from DA to 1 to 2 to 3.
Binding with Cyclodextrin Hosts. Dansyl derivatives are
well-known guests for the formation of inclusion complexes
with CD hosts.19-21 Our interest in the investigation of host-
guest interactions between these dansyl-containing dendrimers
and cyclodextrins derives from previous results obtained with
ferrocene-containing dendrimers.9 Our electrochemical data
suggested that the stepwise growth in the dendritic matrix (from
first to second to third generation) effectively decreases the
equilibrium association constant (Ka) between the dendrimer-
attached ferrocene group and â-CD. This effect is presumably
due to the increasing steric difficulty that the CD host
encounterssas the dendritic matrix growssto approach and
engulf the ferrocene residue. In fact, cyclodextrin binding studies
may constitute an additional method to probe how deeply buried
is the functional group of choice into the dendritic matrix and
provide insight into the extent of enVironmental isolation that
the dendrimer affords to the guest unit (ferrocene, dansyl, etc.).
Initially we investigated binding of dendrimers 1 and 2 with
â-CD using 1H NMR spectroscopy. Figure 3 shows the changes
in the observed chemical shifts for the proton resonances of
dendrimer 1 induced by the addition of the â-CD host. Clearly,
the aromatic protons in the dansyl unit, as well as the methylene
protons in the three branches of the dendrimer unit, are sensitive
to the presence of the CD. Regression analysis of the chemical
shifts measured as a function of the concentration of added â-CD
leads to a value of 110 M-1 for Ka. Similar experiments with
dendrimer 2 yielded smaller â-CD-induced chemical shift
changes (see Figure 4). Regression analysis of these data led to
much smaller values of the corresponding Ka, but accurate values
could not be determined with a reasonable confidence level
because the binding is very weak. Notice that â-CD-induced
chemical shifts in 2 were only observed for the protons of the
first dendrimer layer, that is, the layer closer to the dansyl group.
No interaction between the CD host and the outside, carboxylate-
Figure 5. Effect of added â-CD on the steady-state fluorescence of
for DA (b), 1 (9), 2 (2), and 3 (1) in aqueous buffer (pH ) 7.0). The
solid curves represent the regression analysis.
terminated layer of the dendrimer was detected from these NMR
experiments. To further verify that the carboxylate surface of
the dendrimer does not interact with â-CD, we performed NMR
control experiments with dendrimers 4 and 5 in which the dansyl
residue is replaced by a nitro group. Neither one of these
compounds exhibited detectable â-CD-induced chemical shift
changes, unequivocally confirming that the CD binding site in
1 and 2 is the dansyl residue. Our NMR data also indicate that
the stability of the â-CD‚2 complex is much lower than that of
the â-CD‚1 complex. We did not run similar NMR experiments
with dendrimer 3 due to the well-documented problems
encountered for the observation of core proton resonances in
third and larger generation dendrimers.
We also used steady-state fluorescence intensity and anisot-
ropy methods21 to determine the Ka values for binding of DA
and 1-3 to â-CD. Figure 5 presents the integrated total dansyl
fluorescence intensity as a function of added â-CD for DA (b)
and dendrimers 1 (9), 2 (2), and 3 (1) dissolved in aqueous
buffer. Examination of these data reveals that the addition of
â-CD to the solution causes a marked increase in the dansyl
group fluorescence intensity only for DA and 1. The increase
in fluorescence for 2 is nominal, and no statistical change in
fluorescence intensity is observed for 3 upon addition of â-CD.
According to these intensity profiles, dendrimer 3 is not bound
by â-CD. To confirm the absence/presence of binding we
measured the steady-state fluorescence anisotropy of DA and
(19) Dunbar, R. A.; Bright, F. A. Supramol. Chem. 1994, 3, 93.
(20) Ikeda, H.; Nakamura, M.; Ise, N.; Oguma, N.; Nakamura, A.; Ikeda,
T.; Toda, F.; Ueno, A. J. Am. Chem. Soc. 1996, 118, 10980.
(21) Catena, G. C.; Bright, F. V. Anal. Chem. 1989, 61, 905.

Synthesis and Photophysical Studies of Dendrimers
J. Am. Chem. Soc., Vol. 122, No. 26, 2000 6143
Table 2. Recovered Equilibrium Association Constants, Ka, for
Binding of DA and Dansyl-Labeled Dendrimers 1, 2, and 3 with
â-CD in Aqueous Buffer (pH ) 7.0) at 22 °C
complex
Ka (M-1
)
r2
DA + â-CD
1 + â-CD
307 ( 35
0.999
0.999
136 ( 23
110a
2 + â-CD
3 + â-CD
<1
0.957
∼0 (no association
observed)
a from NMR measurements.
dendrimers 1, 2, and 3 as a function of added â-CD (data not
shown). The anisotropy data confirm that 3 does not bind to
â-CD, while association of 2 with â-CD is extremely weak.
Table 2 collects the recovered Ka values for the complexation
between DA and 1-3 with the â-CD host.
Figure 6. Effect of added anti-dansyl antibodies on the steady-state
fluorescence anisotropy of for DA (b), 1 (9), 2 (2), and 3 (1) in
aqueous buffer (pH ) 7.0). The solid curves represent the regression
analysis.
Inspection of the results presented in Table 2 shows several
important points. The Ka recovered by fluorescence for 1 with
1
â-CD is in good agreement with the value obtained using H
NMR spectroscopy (vide supra). However, although 1 can bind
â-CD, its Ka is significantly lower than that for the â-CD‚DA
complex (307 vs 136 M-1). Finally, the â-CD access to the
dansyl residue in dendrimers 2 and 3 is totally impeded by the
dendrimer mass.
Table 3. Recovered Equilibrium Association Constants, Ka, for
Binding of DA and Dansyl-Labeled Dendrimers 1, 2, and 3 with
Polyclonal Anti-Dansyl Antibodies in Aqueous Buffer (pH ) 7.0) at
22 °C
from steady-state anisotropy from fluorescence intensity
Anti-Dansyl Antibody Binding. The binding between a
hapten and an antibody can be a remarkably selective event
that is often characterized by a high affinity (Ka > 106 M-1).
In contrast to other anti-fluorophore antibodies that quench the
fluorescence of the fluorophore upon association, the association
between polyclonal anti-dansyl antibodies and dansyl leads to
a significant increase in the dansyl fluorescence and a hypso-
chromic spectral shift in the dansyl emission spectrum.22,23 To
assess the anti-dansyl/dendrimer Ka, we measured both the
steady-state anisotropy and fluorescence intensity of the dansyl
residue in DA and dendrimers 1, 2, and 3 as a function of added
anti-dansyl antibody. This approach is superior to an intensity
measurement alone because binding may not result in a
measurable spectral shift and/or intensity change, but binding
must be associated with a change in the steady-state fluorescence
anisotropy and/or intensity.21 Plots of steady-state fluorescence
vs anti-dansyl concentration (not shown) indicated binding only
for DA and 1 to anti-dansyl. While this is fully consistent with
the â-CD results (Table 2), the intensity-based measurements
are not conclusive evidence against binding between dendrimers
complex
Ka (M-1
)
r2
Ka (M-1
)
r2
DA + Ab 4.97 ( 0.23 × 107 0.999 1.40 ( 0.51 × 108 0.987
1 + Ab
2 + Ab
4.07 ( 0.20 × 106 0.997 3.89 ( 0.94 × 106 0.910
1.91 ( 0.12 × 106 0.992 No association
observed
3 + Ab
1.46 ( 0.10 × 106 0.958 No association
observed
appears to move in concert with the entire dendrimer; no local
motion was detectable. Association between a nonselective guest
(â-CD) and the dansyl residue within the larger dendrimers can
be impeded entirely. A selective host (anti-dansyl antibodies)
can complex with each dendrimer to some degree, and the
stability of the complexes was only decreased 30-fold when
one compared the free guest to a sterically crowded dendrimer
like 3. This latter result is somewhat surprising given that â-CD
and anti-dansyl differ by ∼100-fold in mass (â-CD, 1135 g/mol;
anti-dansyl, 150 000 g/mole), suggesting that the dendrimer
conformation/structure can be altered to allow access to the
dansyl residue if the binding affinity is great enough.
21
2 and 3 and anti-dansyl antibodies.
Figure 6 presents the steady-state fluorescence anisotropy DA
(b), 1 (9), 2 (2), and 3 (1) as a function of added antibody.
These data clearly indicate that DA and dendrimers 1, 2, and 3
bind to polyclonal anti-dansyl antibodies. The recovered Ka
values are compiled in Table 3. These results illustrate that:
(1) there is remarkably strong binding between the dansyl
residue within each dendrimer and the anti-dansyl antibody and
(2) the Ka decreases only about 30-fold when we compare the
binding affinity of DA (∼5 × 107 M-1) to that of dendrimer 3
(1.5 × 106 M-1).
Experimental Section
Reagents for Synthesis. All chemicals were reagent grade (Aldrich)
and were used as received. â-Cyclodextrin (â-CD) was a gift from
Cerestar and used without any further purification. Solvents, such as
CH3CN, were freshly distilled. Column chromatography was performed
with Scientific Adsorbents silica gel (63-200 µm).
3 Cascade:Dansyl[1]:(3-Oxo-2-azapropylidyne):Propanoic (1).
Dansyl chloride (0.520 g, 1.93 mmol), Behera’s amine (0.503 g, 1.21
mmol), and K2CO3 anhydrous (0.561 g, 4.01 mmol) were stirred in 10
mL of dry CH3CN under N2 at 23 °C. The progress of the reaction
was followed by TLC (CH2Cl2). The reaction mixture was filtered, and
the solvent was removed at reduced pressure. The residue was purified
by column chromatography (SiO2, CH2Cl2 followed by 5:1 CH2Cl2/
EtOAc) to furnish the ester product (0.550 g, 70%) as a pale green
foam. This compound was hydrolyzed by stirring the foam in 96%
formic acid for 2 days to produce a white powder (95%) after complete
removal of the formic acid under vacuum. FT-IR (KBr): 3300-2500
(ν(OH)) cm-1; 3463 (ν(NH)) cm-1; 1723 (ν(CdO)) cm-1; 1569
Conclusions
The results of these experiments illustrate several points. As
the dendrimers grow, the dansyl residue encounters a signifi-
cantly less polar microenvironment. The dansyl reporter group
(22) Parker, C. W. In Handbook of Experimental Immunology; Weir, D.
M., Ed.; Blackwell: Oxford, 1973; Vol. 1, Chapter 14.
(23) Hanson, D. C.; Yguerabide, J.; Schumaker, V. N. Biochemistry 1981,
20, 6842.
(ν(AMIDE II, N-H)) cm-1; 1317, 1142 (ν(SdO)) cm-1
(DMSO-d6): 1.60-1.93 (m, CH2CH2, 12H); 2.81 (s, CH3, 6H); 7.57
.
1H NMR

6144 J. Am. Chem. Soc., Vol. 122, No. 26, 2000
Cardona et al.
(sb, NH, 1H); 7.2-8.4 (m, C10H6, 6H); 12.0 (bs, OH, 3H). 13C NMR
(DMSO-d6): 27.59, 30.33 (CH2CH2); 45.01 (CH3); 60.45 (C(CH2-
CH2)3); 114.85, 119.06, 123.50, 127.63, 127.839, 128.77, 128.96,
129.35, 138.63, 151.34 (C10H6); 173.85 (COOH). UV-vis [λmax, nm
(ꢀ, M-1 cm-1), phosphate buffer pH 7]: 3295 (4.33 × 103). MALDI-
TOF MS: 479.926 (M+, calcd 480.53).
EtOAc). Hydrolysis of the obtained white foam in 96% formic acid
procured a white powder in 95% yield. FT-IR (KBr): 3300-2500
(ν(OH)) cm-1; 3339 (ν(N-H)) cm-1; 1721 (ν(CdO)) cm-1; 1657
(ν(AMIDE I, CdO)) cm-1; 1539 (ν(AMIDE I, N-H)) cm-1; 1295
1
(ν(CdO)) cm-1; 1201 (ν(O-H)) cm-1. H NMR (DMSO-d6): 2.03
(m, CH2CH2, 12H); 1.75-2.12 (m, CH2CH2, 36H); 7.30 (bs, NH, 3H);
12.0 (bs, OH, 9H). 13C NMR (DMSO-d6): 28.00, 29.04 (CH2CH2);
30.06, 31.10 (CH2CH2); 56.39 (C(CH2CH2)3); 93.27 (O2NC); 170.37
(CONH); 174.32 (COOH). MALDI-TOF MS: 966.137 (M+, calcd
964.93); 989.251 (M + Na+); 1005.63 (M + K+); 915.921 (M - NO2).
Reagents for Photophysical Studies. DA (99%, Aldrich Chemical
Co.), Na2HPO4‚7H2O and NaH2PO4‚H2O (J. T. Baker, Inc.), 1,4-bis-
(4-methyl-5-phenyl-2-oxazolyl)benzene (Me2POPOP, Aldrich Chemical
Co.), ethanol (dehydrated, 200 proof, Pharmco), â-CD (Sigma Chemical
Co.), and polyclonal anti-dansyl antibody (rabbit IgG fraction, Molecular
Probes, Inc.) were used as received. All aqueous buffer solutions were
prepared from doubly distilled-deionized water (Millipore) and stock
solutions were refrigerated in the dark at 4 ( 2 °C.
9 Cascade:Dansyl[1]:(3-Oxo-2-azapropylidyne):(3-Oxo-2-azap-
entylidyne):Propanoic (2). As described above, dansyl chloride (0.266
g, 0.99 mmol) was stirred with the second generation amine building
block (1.00 g, 0.69 mmol) in dry acetonitrile over K2CO3 anhydrous
(0.26 g, 1.9 mmol) under N2. After filtration and removal of the solvent,
the residue was purified by column chromatography, eluting with CH2-
Cl2, followed by 5:1 CH2Cl2/EtOAc and finally 2:1 CH2Cl2/EtOAc.
The recovered product (0.530 g, 46%), a pale green foam, was
hydrolyzed as described above to procure a pale yellow powder in 95%
yield after complete removal of the formic acid. FT-IR (KBr): 3300-
2500 (ν(OH)) cm-1; 3348 (ν(NH)) cm-1; 1715 (ν(CdO)) cm-1; 1643
(ν(AMIDE I, CdO)) cm-1; 1549 (ν(AMIDE II, N-H)) cm-1; 1300
1
(ν(CdO)) cm-1; 1208 (ν(H-O)) cm-1. H NMR (DMSO-d6): 1.62-
Instrumentation. 1H and 13C NMR spectra were recorded on a
Varian VXR-400 spectrometer; all the chemical shift (δ) values are
reported in ppm. Electronic absorption spectra were obtained in either
a Shimadzu model 2101 or a Milton-Roy model 1201, equipped with
a temperature-controlled cell holder that was maintained at 25 °C. FAB-
MS data were recorded with a VG-Trio mass spectrometer. MALDI-
TOF MS spectra were recorded at the LSU Mass Spectroscopy Facility.
Fluorescence experiments were performed with an SLM-AMINCO
model 48000 MHF. For all steady-state fluorescence measurements, a
450 W xenon arc lamp was used as the excitation source with single
grating monochromators serving as the wavelength selection devices.
The excitation and emission spectral band-passes were kept at 8 nm,
and all emission spectra were background-corrected using appropriate
blanks. The time-resolved fluorescence anisotropy and intensity decay
kinetics were measured in the frequency domain12-16 using the same
instrument. For these particular experiments, a CW argon-ion laser
(Coherent, Innova 90-6) operating at 351.1 nm was the excitation
source. An interference filter (Oriel) was placed in the excitation beam
path to minimize extraneous plasma tube superradiance from reaching
the detection system. The sample fluorescence was monitored in the
typical L-format after passing through a 420 nm long-pass filter and a
Glan-Thompson calcite polarizer. The Pockels cell modulator was
operated at a 5 MHz base repetition rate. Typically, data were acquired
for 60 s between 5 and 200 MHz (40 total frequencies), and at least 10
discrete multifrequency data sets were acquired for each sample. For
the excited-state intensity decay measurements, we used a dilute solution
of Me2POPOP dissolved in ethanol as the reference lifetime standard;
its lifetime was assigned a value of 1.45 ns.12 Magic angle polarization
conditions were used for all excited-state intensity decay measurements
to eliminate bias arising from fluorophore rotational reorientation. The
excited-state fluorescence lifetimes and rotational reorientation times
were recovered from the frequency-domain data by using a com-
mercially available nonlinear least-squares software package (Globals
Unlimited). The actual imprecision in each datum was used as a
weighting factor. Unless otherwise specified, all spectroscopic measure-
ments were carried out at 22 ( 2 °C in 0.1 M pH 7.0 phosphate buffer.
For the dynamic experiments, the samples were subjected to four
freeze-pump-thaw cycles to remove residual oxygen.
1.92 (m, CH2CH2, 12H) gen 1; 1.82-2.12 (m, CH2CH2, 36H) gen 2;
2.86 (s, CH3, 6H); 7.11 (sb, NH, 3H) gen 2; 7.73 (sb, NH, 1H) gen 1;
7.2-8.4 (m, C10H6, 6H); 12.2 (bs, OH, 9H). 13C NMR (DMSO-d6):
28.16, 29.13 (CH2CH2) gen 2; 30.32, 32.42 (CH2CH2) gen 1; 45.03
(CH3); 56.22 (C(CH2CH2)3) gen 2; 60.95 (C(CH2CH2)3) gen 1; 114.81,
119.40, 123.46, 127.30, 127.68, 128.84, 128.03, 129.13, 139.43, 151.31
(C10H6); 171.71 (CONH); 174.53 (COOH). UV-vis [λmax, nm (ꢀ, M-1
cm-1), phosphate buffer pH 7]: 329 (3.80 × 103). MALDI-TOF MS:
1168.02 (M+, calcd 1168.23); 920.311 (M - H3N+C(CH2CH2COOH)3).
(27 Cascade:Dansyl[1]:(3-Oxo-2-azapropylidyne):(3-Oxo-2-aza-
pentylidyne):(3-Oxo-2-azapentylidyne):Propanoic (3). Following the
procedure described above, dansyl chloride (0.033 g, 0.12 mmol) and
the third generation amine building block (0.50 g, 0.11 mmol) were
stirred in a suspension of K2CO3 anhydrous in dry CH3CN under N2.
The residue obtained after filtration and solvent evaporation was purified
by column chromatography (SiO2, 5:1 CH2Cl2/EtOAc followed by
EtOAC). A very pale green, almost white foam was recovered (0.263
g, 50%), which upon hydrolysis gave a white-green powder (95%).
FT-IR (KBr): 3300-2500 (ν(OH)) cm-1; 3339 (ν(NH)) cm-1; 1717
(ν(CdO)) cm-1; 1652 (ν(AMIDE I, CdO)) cm-1; 1542 (ν(AMIDE II,
1
N-H)) cm-1; 1291 (ν(CdO)) cm-1; 1203 (ν(O-H)) cm-1. H NMR
(DMSO-d6): 1.7-2.2 (m, CH2CH2, 156H) gen 1, 2, and 3; 2.80 (s,
CH3, 6H); 7.24 (sb, NH, 9H) gen 3; 6.9-8.5 (m, C10H6, 6H & NH
gen1 and 2, 4H); 12.1 (bs, OH, 27H). 13C NMR (DMSO-d6): 28.10,
29.07 (CH2CH2) gen 3; 30.26, 30.90 (CH2CH2) gen 2; 44.99 (CH3);
56.31 (C(CH2CH2)3) gen 3; 172.26 (CONH) gen 3; 174.48 (COOH).
UV-vis [λmax, nm (ꢀ, M-1 cm-1), phosphate buffer pH 7]: 324 (3.31
× 103). MALDI-TOF MS: 3230.69 (M+, calcd 3231.33); 2993.08 (M
- DansylH+).
3 Cascade:Nitro[1]:(3-Oxo-2-azapropylidyne):Propanoic (4). The
nitrotriester (4.55 g, 10.2 mmol), precursor of Behera’s amine, was
hydrolyzed by stirring it in 10 mL of 96% formic acid for 15 h at 23
°C. After complete removal of the formic acid at reduced pressure and
100 °C, a white powder was obtained (2.75 g, 97%). FT-IR (KBr):
3300-2500 (ν(OH)) cm-1; 1715 (ν(CdO)) cm-1; 1532 (ν(NdO))
cm-1; 1295 (ν(CdO)) cm-1. 1H NMR (DMSO-d6): 2.1-2.2 (m, CH2-
CH2, 12H); 12.3 (bs, OH, 3H). 13C NMR (DMSO-d6): 28.16, 29.73
(CH2CH2); 92.77 (O2NC); 173.04 (COOH). FAB+ MS: 277 (M+, calcd
277.23); 278 (M + 1).
Acknowledgment. The authors are grateful to the NSF for
the support of this research work (to A.E.K.,
CHE-9982014; to F.V.B., CHE-9626636). F.V.B. is also grateful
for support from the DOE (DEFG0290ER14143) and ONR
(N000149610501 and N0001497100773). C.M.C. acknowledges
a minority predoctoral fellowship from NIH.
9 Cascade:Nitro[1]:(3-Oxo-2-azapropylidyne):(3-Oxo-2-azapen-
tylidyne):Propanoic (5). The triacid made as described above (2.5 g,
9.0 mmol) was coupled to Behera’s amine (11.2 g, 27.0 mmol) in the
presence of DCC (5.6 g, 27.1 mmol) and HOBT (1.23 g, 9.1 mmol).
The mixture was stirred in 100 mL of dry THF for 6 days at 23 °C
under N2. The product was filtered, and after evaporation of the solvent,
the residue was purified by column chromatography (2:1 CH2Cl2/
JA000949L