Patchy supramolecules as versatile tools to probe hydrophobicity in nanoglobular systems

Article abstract of DOI:10.1021/ja401373h

We describe precise supramolecules that enable the evaluation of the effective hydrophobicity of amphiphilic or patchy nanoglobular systems. These supramolecules exhibit the lower critical solution temperature phenomenon, which provides a quantitative measure of their effective hydrophobicity. Specifically, two isomeric 8-aryl-2′-deoxyguanosine derivatives with a transposed pair of methylene groups self-assemble into hexadecameric nanoglobular supramolecular G-quadruplexes (SGQs) that show large differences in their transition temperatures as determined by turbidity and differential scanning calorimetry studies. Molecular modeling studies suggested that differential clustering of the hydrophobic patches on the surface is responsible for the striking differences between the two isomeric supramolecules.

Full text of DOI:10.1021/ja401373h

Communication
pubs.acs.org/JACS
Patchy Supramolecules as Versatile Tools To Probe Hydrophobicity
in Nanoglobular Systems
Luis M. Negron, Yazmary Melendez-Contes, and Jose M. Rivera*
́
́
́
́
Department of Chemistry, University of Puerto Rico, Río Piedras Campus, Río Piedras, Puerto Rico 00931
S
* Supporting Information
ABSTRACT: We describe precise supramolecules that
enable the evaluation of the effective hydrophobicity of
amphiphilic or “patchy” nanoglobular systems. These
supramolecules exhibit the lower critical solution temper-
ature phenomenon, which provides a quantitative measure
of their effective hydrophobicity. Specifically, two isomeric
8-aryl-2′-deoxyguanosine derivatives with a transposed pair
of methylene groups self-assemble into hexadecameric
nanoglobular supramolecular G-quadruplexes (SGQs) that
show large differences in their transition temperatures as
determined by turbidity and differential scanning calorim-
etry studies. Molecular modeling studies suggested that
differential clustering of the hydrophobic patches on the
surface is responsible for the striking differences between
the two isomeric supramolecules.
he juxtaposition of hydrophobic and hydrophilic regions
T(patches) on the surface of proteins plays a pivotal role in
both health and disease.1 For example, hydrophobic patches on
the surface of proteins mediate protein−protein interactions
and have provided a means for the evolution of multimeric
systems.2 Conversely, the emergence of a hydrophobic patch
́
Figure 1. Kekule structures for 1, 2, and 3. Addition of two methylene
on a mutant hemoglobin results in detrimental polymerization
leading to sickle cell anemia.3 However, despite its long and
interesting history4 in addition to recent seminal findings,5
critical details regarding the hydrophobic effect have yet to be
elucidated, such as in the context of the rough amphiphilic
(patchy) surfaces on precise nanoglobular systems, including
many soluble proteins.6
groups (red) to 2 (control) produces the structural isomers 1 with two
methylenes in the aryl group and 3 with one methylene in each side
chain attached to the ribose. Addition of KI in aqueous solution
generates self-assembled nanoglobules (116, 216, and 316) with patchy
surfaces. The models on the right represent the solvent-accessible
surface areas of the SGQs with the methylenes highlighted in red.
They were generated using Maestro after minimizations in Macro-
Model with the AMBER 94 force field.10
The development of model systems with patchy or
amphiphilic surfaces has the potential to clarify some of those
details and also to enable technological applications.7 A number
of model systems such as polymers or micelles could be
envisioned to address this challenge. However, they have
multiple limitations such as their polydispersity and the
difficulty of precisely controlling their composition and
structure. Here we describe a family of hexadecameric
nanoglobular supramolecular G-quadruplexes (SGQs) with
precise structures and amphiphilic “patchy” surfaces that offer
a complement to polymeric systems (Figure 1).8 These systems
are thermoresponsive [i.e., they show the lower critical solution
temperature (LCST) phenomenon], which gives us a
quantitative measure of their effective hydrophobicity.9 We
demonstrate that the distribution of the patches can in fact be
used to modulate the transition temperature for the onset of
the LCST phenomenon and thus their hydrophobicity.
Compounds 1, 2, and 3 were prepared using the method-
ology described previously for this family of compounds
(Figure S1 in the Supporting Information).11−13 In aqueous
phosphate-buffered solutions (pH 7.4, 2 M KI), all three
1
compounds (10 mM) showed signature peaks in the H NMR
spectra confirming the formation of the corresponding
hexadecamers 116, 216, and 316 (Figures S18−S20).13 Turbidity
experiments (transmittance at 500 nm) confirmed the
thermoresponsive behavior of these SGQs and provided the
first indication that the different patchy surfaces indeed impact
their effective hydrophobicities (Figure 2). The cloud-point
temperature (Tcp) for 116 was 10 °C, while Tcp = 32 °C for its
Received: February 6, 2013
Published: February 25, 2013
© 2013 American Chemical Society
3815
dx.doi.org/10.1021/ja401373h | J. Am. Chem. Soc. 2013, 135, 3815−3817

Journal of the American Chemical Society
Communication
Table 1. Thermodynamic Parameters for Supramolecules
116, 216, and 316 Obtained from Turbidity and DSC
measurements
a
b
c
SGQ Tcp (°C)
Tt (°C)
Tm (°C)
ΔHt (kcal/mol)
116
216
316
10
60
32
6.35 0.03
55.1 0.4
59.50 0.03
68.4 0.1
0.75 0.04
12.2 0.5
12.4 0.1
56.19 0.06
69.98 0.05
a
Cloud-point temperature as determined by turbidity measurements.
Transition temperature as determined by DSC. Melting temperature
b
c
as determined by DSC.
(0.75 kcal/mol) compared with the corresponding values for
216 and 316 (∼12 kcal/mol) (Figure 2 and Table 1). The
apparent large discrepancy between the onset of turbidity (Tcp)
and the maximum of the DSC peak (Tt) for 316 results from an
athermal association process between the nanoglobular
assemblies that is likely mediated by interactions between the
peripheral hydroxyl groups and water molecules. This
phenomenon is similar to the one reported by Winnik and
co-workers17 using telechelic poly(N-isopropylacrylamide)
(PNIPAM) derivatives that form micellar assemblies. Accord-
ing to Winnik, this water-mediated association (of micelles, in
their case) leads to globules that are large enough to scatter
light (thus the onset of turbidity), but since it does not involve
partitioning of the water molecules of hydration to the bulk, it
seems to be below the detection limit of the calorimeter. The
fact that this occurs exclusively with 316 (since for the other two
assemblies the onset of turbidity occurs slightly above Tt) will
be the subject of future molecular dynamics studies aimed at
illuminating this particular phenomenon.
Figure 2. (top) Turbidity curves (measured at 500 nm) and (bottom)
DSC endotherms for 116 (blue), 216 (green), and 316 (red).12 The
measurements were performed in aqueous phosphate-buffered
solutions (10 mM in 1−3, pH 7.4, 2 M KI).
Molecular modeling studies of these SGQs provided vital
clues to help explain why 116 shows a significantly higher
hydrophobicity than its isomer 316 (Figure 1). The former
shows clusters of patches concentrated around the core’s
surface, while the latter presents a more scattered distribution
of such patches. These smaller and more distributed patches on
the surface of 316 can be accommodated in the tetrahedral
water network with minimal disruption of the hydrogen bonds.
Interestingly, this remodeling of the patches on the surface of
316 has an effect very similar to the actual removal of the
hydrophobic patches (216). While the distribution of patches
enables the formation of clusters of 316 at lower temperatures
(i.e., lower Tcp) relative to 216, the former shows Tt and ΔHt
values very similar to those of the latter, although the sharper
endotherm transition for 316 (Figure 2) hints at a more
cooperative dehydration process. These results underscore the
importance of having systems that are modular and molecularly
precise, which could enable future molecular dynamics studies
to illuminate these phenomena further.
The hydrophobic effect never ceases to provide surprising
twists in its manifestations. Although much progress has been
made since its first descriptions in the early 20th century,4 it
remains a fertile ground for new discoveries and applications.
We have demonstrated an additional use for SGQs in the
elaboration of useful tools for studying the hydrophobic effect
in a precise nanoglobular system. We have shown how the
three-dimensional distribution of hydrophobic elements leads
to a profound impact on the hydrophobicity of the assembly.
We expect these results to enable further studies on the
dynamics of water molecules in contact with patchy surfaces
having different sizes and shapes, which should complement
studies performed with polymers. For example, related systems
isomer 316 and 60 °C for 216.14 These results suggest that 116 is
substantially more hydrophobic than 316 (ΔTcp ≈ 22 °C), while
216 appears to be the least hydrophobic of all.15 The turbidity
above Tcp resulted from the formation of a colloidal suspension
of microglobules, which were shown by dynamic light scattering
(DLS) to have average hydrodynamic diameters (DH) in the
range between 4.5 and 21.7 μm (Figures S25 and S26).11,16
Differential scanning calorimetry (DSC) experiments pro-
vided further support of the turbidity measurements while
revealing additional data associated with the thermodynamic
parameters of these SGQs (Figures S23−S25).13 After data
deconvolution, the DSC endotherms revealed two different
processes: the transition temperature (Tt) corresponding to the
LCST and, at higher temperatures, the melting temperature
(Tm) of the SGQs within the microglobules (Figures S23−
S25).13 The endotherms for 116 and 316 showed maxima
corresponding to Tt values of 6.4 and 56.2 °C, respectively
(ΔTt ≈ 50 °C), while that for 216 (with two fewer methylene
groups per subunit) showed a Tt of ∼55.0 °C (Table 1).
Furthermore, these measurements revealed that these systems
have Tm values in the range 59−70 °C, with 116 being the least
thermally stable of the three assemblies. This is likely the result
of its larger steric bulk imposed by the butoxy group near the
core of the assembly (vs ethoxy in 2 and 3; Figure 1).
Higher Tt values relate to the energy required to expel the
associated water molecules, while the corresponding enthalpy
(ΔHt) indicates the number of water molecules associated
within the hydration shells. For example, the fact that there are
fewer water molecules of hydration associated with 116 is
reflected in the lower magnitude of the enthalpy for the process
3816
dx.doi.org/10.1021/ja401373h | J. Am. Chem. Soc. 2013, 135, 3815−3817

Journal of the American Chemical Society
Communication
(11) (a) Betancourt, J. E.; Rivera, J. M. J. Am. Chem. Soc. 2009, 131,
16666−16668. (b) Betancourt, J. E.; Subramani, C.; Serrano-Velez, J.
L.; Rosa-Molinar, E.; Rotello, V. M.; Rivera, J. M. Chem. Commun.
2010, 46, 8537−8539.
(12) Monomers 1 and 2 (Figure 1) were synthesized as previously
reported for 3 (see ref 11; Figure S1). All of the data for 3, including
its self-assembly and LCST properties, were previously reported in ref
11b.
should enable the evaluation of how the specific nature of the
patches (e.g., saturated versus unsaturated) in conjunction with
their relative sizes and distributions affect the hydrophobicity.
We believe that the modularity of the system will aid both
experimental and theoretical approaches aimed at answering
these and related questions.
ASSOCIATED CONTENT
* Supporting Information
(13) See the Supporting Information.
■
S
(14) The values of Tt for these SGQs showed no significant changes
as a function of concentration over the range 5−10 mM.
(15) Cooling all of the SGQs below their respective Tt (116 to 0 °C;
216 and 316 to 25 °C) reestablished the initial state of solubility with
fairly transparent solutions. Furthermore, the process could be
repeated at least seven times for each sample with little sign of fatigue
(Figures S21 and S22).
(16) Below Tt, the DH values were 3.6−4.0 nm for 116, 4.0−4.6 nm
for 216, and 5 nm for 316.
(17) (a) Obeid, R.; Tanaka, F.; Winnik, F. M. Macromolecules 2009,
42, 5818−5828. (b) Kujawa, P.; Segui, F.; Shaban, S.; Diab, C.; Okada,
Y.; Tanaka, F.; Winnik, F. M. Macromolecules 2006, 39, 341−348.
Detailed synthetic procedures, characterization data for all new
compounds, experimental protocols, and NMR data. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
riveralab.upr@gmail.com
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank NIH-SCORE (Grant 5SC1GM093994) and the
RISE Program (Grant 5R25GM061151) for financial support.
L.M.N. thanks the Alfred P. Sloan Foundation RISE Program
for graduate fellowships. We are also grateful to Dr. Irving E.
Vega and Edwin Vazquez-Rosa for assistance with mass
spectrometry measurements.
REFERENCES
■
(1) (a) Ball, P. Nature 2011, 478, 467−468. (b) Ball, P. Chem. Rev.
2008, 108, 74−108.
(2) (a) Levy, E. D.; Boeri Erba, E.; Robinson, C. V.; Teichmann, S. A.
Nature 2008, 453, 1262−1265. (b) Tsai, C. J.; Nussinov, R. Protein Sci.
1997, 6, 1426−1437.
(3) Galkin, O.; Vekilov, P. G. J. Mol. Biol. 2004, 336, 43−59.
(4) Tanford, C. Protein Sci. 1997, 6, 1358−1366.
(5) For selected examples related to the hydrophobic effect, see:
(a) Li, I. T. S.; Walker, G. C. Acc. Chem. Res. 2012, 45, 2011−2021.
(b) Patel, A. J.; Varilly, P.; Jamadagni, S. N.; Acharya, H.; Garde, S.;
Chandler, D. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 17678−17683.
(c) Garde, S.; Patel, A. J. Proc. Natl. Acad. Sci. U.S.A. 2011, 108,
16491−16492. (d) Li, I. T. S.; Walker, G. C. Proc. Natl. Acad. Sci.
U.S.A. 2011, 108, 16527−16532. (e) Jamadagni, S. N.; Godawat, R.;
Dordick, J. S.; Garde, S. J. Phys. Chem. B 2009, 113, 4093−4101.
(f) Giovambattista, N.; Lopez, C. F.; Rossky, P. J.; Debenedetti, P. G.
Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 2274−2279. (g) Giovambattista,
N.; Debenedetti, P. G.; Rossky, P. J. J. Phys. Chem. C 2007, 111, 1323−
1332. (h) Chandler, D. Nature 2005, 437, 640−647. (i) Lum, K.;
Chandler, D.; Weeks, J. D. J. Phys. Chem. B 1999, 103, 4570−4577.
(6) Granick, S.; Bae, S. Science 2008, 322, 1477−1478.
(7) (a) Glotzer, S. C.; Solomon, M. J. Nat. Mater. 2007, 6, 557−562.
(b) Zhang, Z.; Glotzer, S. C. Nano Lett. 2004, 4, 1407−1413.
(8) For examples of supramolecular systems in aqueous environ-
ments, see: (a) Gibb, B. C. Nat. Chem. 2010, 2, 512−513.
(b) Uhlenheuer, D. A.; Petkau, K.; Brunsveld, L. Chem. Soc. Rev.
2010, 39, 2817−2826. (c) Zayed, J. M.; Nouvel, N.; Rauwald, U.;
Scherman, O. A. Chem. Soc. Rev. 2010, 39, 2806−2816. (d) Oshovsky,
G. V.; Reinhoudt, D. N.; Verboom, W. Angew. Chem., Int. Ed. 2007, 46,
2366−2393.
(9) (a) Urry, D. W. J. Phys. Chem. B 1997, 101, 11007−11028.
(b) Urry, D. W.; Gowda, D. C.; Parker, T. M.; Luan, C. H.; Reid, M.
C.; Harris, C. M.; Pattanaik, A.; Harris, R. D. Biopolymers 1992, 32,
1243−1250.
(10) (a) MacroModel, version 9.5; Schrodinger, LLC: New York,
̈
2007. (b) Maestro, version 8.0.315; Schrodinger, LLC: New York,
̈
2007.
3817
dx.doi.org/10.1021/ja401373h | J. Am. Chem. Soc. 2013, 135, 3815−3817