Rapid release of entrapped contents from multi-functionalizable, surface cross-linked micelles upon different stimulation

Article abstract of DOI:10.1021/ja103391k

Hydrophobic guests such as pyrene could be readily trapped inside the micelles of an alkynylated surfactant in the presence of an azide-functionalized cross-linker using the click reaction. The cross-linker was designed to contain cleavable bonds such as geminal diol, disulfide, and acetal. The resulting pyrene-containing water-soluble nanoparticle was under electrostatic stress when diluted below the CMC of the surfactant. Extremely rapid (<1 min) release of the hydrophobic content was observed when the cross-linker was cleaved. This method combines the ease of physical entrapment and the precision of chemical ligation, and potentially is highly useful in the delivery and controlled release of pharmaceutical agents.

Full text of DOI:10.1021/ja103391k

Published on Web 07/16/2010
Rapid Release of Entrapped Contents from Multi-Functionalizable, Surface
Cross-Linked Micelles upon Different Stimulation
Shiyong Zhang and Yan Zhao*
Department of Chemistry, Iowa State UniVersity, Ames, Iowa 50011-3111
Received April 21, 2010; E-mail: zhaoy@iastate.edu
Scheme 1. Preparation of the Pyrene-Containing SCM
Abstract: Hydrophobic guests such as pyrene could be readily
trapped inside the micelles of an alkynylated surfactant in the
presence of an azide-functionalized cross-linker using the click
reaction. The cross-linker was designed to contain cleavable
bonds such as geminal diol, disulfide, and acetal. The resulting
pyrene-containing water-soluble nanoparticle was under electro-
static stress when diluted below the CMC of the surfactant.
Extremely rapid (<1 min) release of the hydrophobic content was
observed when the cross-linker was cleaved. This method
combines the ease of physical entrapment and the precision of
chemical ligation, and potentially is highly useful in the delivery
(<1 min) upon cleavage of the cross-linkages. Because of the
outstanding tolerance of the click reaction to functional groups,
we can prepare SCMs with a variety of cross-linkers and, as a result,
employ different environmental stimuli to trigger the release.
We chose pyrene as a mock hydrophobic drug because of its
environmentally sensitive fluorescence.9 Its five vibronic bands
respond to environmental polarity differently. The intensity ratio
between the third (∼384 nm) and the first band (∼372 nm) is
particularly sensitive to environmental changes. As shown by Figure
1a, the emission spectrum of pyrene was dependent upon the
concentration of 1. According to I3/I1, pyrene was in a more
hydrophobic microenvironment in the 1 mM aqueous solution of
1 than in 0.05 mM. The CMC of the surfactant was determined to
be ∼1.5 × 10-4 M by monitoring I3/I1 at different concentrations
of the surfactant (Figure 1b). The number compares favorably with
the 1.4 × 10-4 M obtained by surface-tension measurement.6
Pyrene-containing SCMs were prepared according to Scheme
1. An aqueous solution containing 74 µM pyrene and 10 mM 1
was combined with a cross-linker (2-4) and CuCl2/sodium ascor-
bate. Because the solubility limit of pyrene in water is 0.67 µM,10
the majority of the dissolved pyrene resided within the surfactant
micelles.11 The reaction was generally allowed to continue for 24 h
at room temperature.
and controlled release of pharmaceutical agents.
Approximately half of potential drug candidates identified in high
throughput screening have poor solubility in water and thus are
often denied further chance of development.1 Although surfactant
micelles can solubilize hydrophobic agents in water, their usage in
drug delivery is hampered by the high critical micelle concentration
(CMC), low thermodynamic stability, and the exceedingly dynamic
nature of the assembly.
Polymeric micelles represent significant improvements over
surfactant micelles because macromolecular amphiphiles tend to
aggregate at concentrations orders of magnitude lower than their
small molecule counterparts and produce micelles with greater
thermodynamic stability.2 A hydrophobic drug may be physically
trapped inside the hydrophobic core of a polymeric micelle3 or
covalently attached to the polymer.4 The latter approach enables
the controlled release of drugs by specific stimuli and is more
effective at preventing premature drug release than physical
entrapment—features of particular importance in the delivery of
drugs with high cytotoxicity. It has been reported, for example,
that physically entrapped anticancer drugs display as high cyto-
toxicity as the small molecule versions.5 Nonetheless, covalent
linking between the drug and the delivery vehicle puts severe
constraints on the structure of both components and adds consider-
able complexity to the production and formulation of the therapeutic
package.
Entrapment of pyrene was confirmed by fluorescence spectros-
copy. When the pyrene-containing SCMs were diluted in water so
that the concentration of (the cross-linked) 1 was below its CMC,
We recently reported a simple method to capture the micelles
of alkynylated surfactants such as 1 by covalent cross-linking
(Scheme 1).6 Cross-linking was readily achieved by the highly
efficient alkyne-azide click reaction7 in the presence of 1 equiv
of 2 and a catalytic amount of Cu(I).8 The resulting surface-cross-
linked micelles (SCMs), 8-10 nm in diameter, had numerous
residual alkynes on the surface. Multivalent postmodification was
conveniently accomplished via the same click reaction by adding
desired azide-functionalized polymers or ligands after the cross-
linking.
Figure 1. (a) Normalized emission spectra of pyrene in the presence of
surfactant 1 in water. (b) Pyrene I3/I1 ratio as a function of [1]. [pyrene] )
0.1 µM.
In this communication, we report the surprising discovery that
these SCMs can release entrapped contents extremely rapidly
9
10642 J. AM. CHEM. SOC. 2010, 132, 10642–10644
10.1021/ja103391k  2010 American Chemical Society

C O M M U N I C A T I O N S
I3/I1 of pyrene remained unchanged at 0.84-0.85, the same as that
above the CMC of uncross-linked 1 (Figure 1b). Importantly, the
pyrene emission showed no change over at least a period of six
months, suggesting that the hydrophobic guest was physically
trapped inside the nanoparticle and could not escape.
these nanoparticles are properly described as under “electrostatic
stress” below the CMC. It is reasonable to expect, as soon as the
covalent constraint is removed, the nanoparticle would “explode”
like an “electrostatic bomb”. Of course, not all cross-linkages have
to be cleaved and “partial explosion” might be sufficient to release
the pyrene. It is also likely that the electrostatic stress of the SCMs
may accelerate the cleaving reaction. After all, any stress in the
starting materials of a reaction, whether steric, conformational, or,
in this case, electrostatic, should raise the ground-state energy of
the system and lower the activation energy.
We also encapsulated pyrene within SCMs using acetal-contain-
ing 4 as the cross-linker, with the intention that the guest would be
released under acidic conditions. Acid-triggered release is important
in many delivery applications. Endosomes and lysosomes are more
acidic than cytosols.17 Successful delivery via endocytosis thus often
requires acid-triggered release. Cancerous and inflammatory tissues
are also known to be more acidic than normal tissues.18
To our dismay, the pyrene-containing SCMs prepared with 4
only showed a slight change in I3/I1 from 0.81 to 0.77 at pH ) 5
over a period of 96 h, although treatment with 0.01 M HCl did
decrease the number to 0.7 (data not shown). The result was initially
very puzzling to us as the SCMs were surrounded by acidic water
and the p-methoxybenzyl acetal group in 4 was highly prone to
hydrolysis. The compound, for example, underwent partial hy-
drolysis during aqueous workup in our hands.
As soon as periodic acid (HIO4) was added to the mixture to
cleave the 1,2-diol group in the cross-linker, however, I3/I1
dropped quickly (Figure 2a).12 To our amazement, release of
pyrene was so rapid that, by the time HIO4 was added and the
solution was mixed by gentle vortexing (<1 min), the change in
I3/I1 was complete. The end I3/I1 value was somewhat dependent
on the amount of HIO4 added. 1 equiv of the cleaving agent (20
µM) reduced the I3/I1 to 0.74-0.75, higher than the 0.70
observed in the uncross-linked surfactant below the CMC (Figure
1b). Quite likely, not all the 1,2-diol groups in the SCMs were
cleaved when an equivalent amount of the cleaving agent was
added. Release seemed to be complete in the presence of 10 or
100 equiv of HIO4, as the final I3/I1 was similar or even slightly
lower than 0.70 in the uncross-linked micelles below the CMC.13
The outstanding functional group compatibility of the click
reaction enabled us to incorporate cross-linkers sensitive to different
stimuli. Diazide 3, for example, contains a disulfide bond cleavable
under reducing conditions, e.g., upon addition of 5. Release of
pyrene once again was found to occur extremely rapidly (Figure
2b). Excess 5 was no longer needed to reduce the I3/I1 ratio to 0.70.
It seems that the entrapped pyrene was completely released even
with just 1 equiv or 20 µM of 5.
The specificity of the release was demonstrated by the control
experiments.14 Considering the low concentration of the pyrene-
containing SCMs15 and the cleaving agent (HIO4 or 5), the release
was remarkably efficient. Cleavage of disulfide bonds in cross-
linked polymers was reported to take hours to days to complete
and often requires millimolar concentrations of reducing thiols.16
Why did the SCMs expel pyrene so rapidly? Cationic surfactants,
such as 1, form micelles under two opposing forces. Hydrophobic
interactions among the hydrocarbon tails favor micellization, and
electrostatic repulsion among the headgroups disfavors it. Once
captured by covalent cross-linkages, the SCMs cannot disassemble.
Because the stability of the SCMs is maintained by covalent bonds,
Considering that pyrene fluorescence did not directly monitor
what happened to the nanoparticles, we turned to dynamic light
scattering (DLS), which correlates the scattered light with the
diffusion coefficient of a particle in solution. Figure 3 shows the
change in the scattering intensity for various pyrene-containing
SCMs. Disintegration of the nanoparticles was once again found
to be extremely rapid for SCMs prepared with 2 or 3 as the cross-
linker (4 and 0). The scattering intensity dropped to 15 and 24%
of the original value, respectively, when 1 equiv of HIO4 or 5 was
added to the corresponding SCMs. On the other hand, the change
in scattering intensity was clearly slower in the acid-triggered
disassembly. The nanoparticles prepared with 4 as the cross-linker
disintegrated gradually over a period of 60 min at 37 °C, although
the majority of the change occurred in the first 20 min.19 In contrast,
the nanoparticles at pH ) 7 displayed negligible changes in
scattered light over the same period of time (×).
Why did the acid-triggered disassembly have a different reaction
profile? These SCMs are positively charged on the surface. Anionic
periodate should be attracted electrostatically to the surface of the
nanoparticles. Dithiol 5 at least would not be repelled. For the acetal
cross-linkers to be hydrolyzed, however, not only is the reaction
catalyzed by proton but the intermediate is also a carbocation. Neither
prefers to stay in a positively charged microenvironment. Moreover,
Figure 3. Relative intensity of scattered light for the pyrene-containing
SCMs upon different stimulation: 1 equiv of HIO4 for SCMs cross-linked
with 2 (4), 1 equiv of 5 for SCMs cross-linked with 3 (0), and pH 5 (])
and 7 (×) acetate buffer at 37 °C for SCMs cross-linked with 4. [1] ) 20
µM.
Figure 2. Change of pyrene I3/I1 ratio after addition of 0 (4), 1 (0), 10
(]), and 100 equiv (×) of cleaving agent to the pyrene-containing SCMs
in deionized water at ambient temperature. (a) Cross-linker ) 2, cleaving
agent ) HIO4. (b) Cross-linker ) 3, cleaving agent ) 5. [1] ) 20 µM.
9
J. AM. CHEM. SOC. VOL. 132, NO. 31, 2010 10643

C O M M U N I C A T I O N S
(2) Kwon, G. S. Colloids Surf., B 1994, 2, 429–434.
given its hydrophobicity, the acetal may be located in a relatively
hydrophobic region of the SCM and thus is unlikely to be fully solvated
by water. All the above factors would slow down the hydrolysis.20
What could be the reason for the small change of pyrene emission
in the acid-sensitive SCMs? According to Figure 3, the SCMs were
only partially disintegrated in our experiments, as uncross-linked
surfactants are too small to scatter light at 20 µM. The I3/I1 value
for the SCMs prepared with 2 and 3 was 0.84-0.85 (Figure 2),
the same as that for the uncross-linked micelles above the CMC,
but only 0.81 for those prepared with 4. Thus, the former had higher
cross-linking density than the latter.21 As soon as some surfactants
become free and escape from the SCM, pyrene would not be
protected by the remaining surfactants if their hydrocarbon tails
cannot adjust themselves around the guest; this is most likely to
be the case when the initial SCM is highly cross-linked. If, however,
the initial cross-linking density is low, the nanoparticles would be
less rigid. As some surfactants are removed, the remaining structure
can reconfigure itself easily and might still be able to protect its
hydrophobic guest. As long as the binding affinity is sufficiently
high between the partially hydrolyzed SCM and pyrene, the guest
would want to stay inside the hydrophobic particle.
(3) Kabanov, A. V.; Chekhonin, V. P.; Alakhov, V. A.; Batrakova, E. V.;
Lebedev, A. S.; Melik-Nubarov, N. S.; Arzhakov, S. A.; Levashov, A. V.;
Morozov, G. V.; Severin, E. S.; Kabanov, V. A. FEBS Lett. 1989, 258,
343–345.
(4) Yokoyama, M.; Kwon, G. S.; Okano, T.; Sakurai, Y.; Seto, T.; Kataoka,
K. Bioconjugate Chem. 1992, 3, 295–301.
(5) Bae, Y.; Kataoka, K. AdV. Drug DeliVery ReV. 2009, 61, 768–784.
(6) Zhang, S.; Zhao, Y. Macromolecules 2010, 43, 4020–4022.
(7) (a) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew.
Chem., Int. Ed. 2002, 41, 2596–2599. (b) Kolb, H. C.; Finn, M. G.;
Sharpless, K. B. Angew. Chem., Int. Ed. 2001, 40, 2004–2021. (c) Finn,
M. G.; Kolb, H. C.; Fokin, V. V.; Sharpless, K. B. Prog. Chem. 2008, 20,
1–4.
(8) The popular CuSO4-sodium ascorbate combination produced a precipitate
from a micellar solution of 1, possibly because the divalent sulfate anion
affects the solubility of the cationic surfactant. CuCl2 (5 mol %) and sodium
ascorbate (25 mol %) were found to work well in the cross-linking.
(9) Kalyanasundaram, K.; Thomas, J. K. J. Am. Chem. Soc. 1977, 99, 2039–
2044.
(10) (a) Abraham, M. H.; Le, J. J. Pharm. Sci. 1999, 88, 868–880. (b) Futoma,
D. J. Polycyclic aromatic hydrocarbons in water systems; CRC Press: Boca
Raton, FL, 1981; pp 14-22.
(11) Approximately one-third of the micelles contained pyrene in our experiment.
The ratio of [1]/[pyrene] was ∼130 in our experiments. Our previous work
(ref 6) estimated the micelle aggregation number of 1 to be 40-50. Thus,
on average, every three micelles contained one pyrene molecule. The
emission spectra of pyrene showed the absence of pyrene excimer,
indicating that few (if any) micelles contained more than one pyrene.
(12) The releasing experiments were performed in triplicates. The relative error
in I3/I1 was <5%.
As demonstrated by our previous research,6 multivalent surface
functionalization was extremely simple in the alkynyl-terminated
SCMs. If a combination of several cross-linkers is used, the stability
of these nanoparticles should be tailored readily for specific
applications. Our finding that the SCMs can eject entrapped
hydrophobic content rapidly is significant. The hydrophobic guest
is apparently trapped in a high energy state when the surface is
heavily cross-linked. During cleavage, if the hydrocarbon tails
cannot rearrange to accommodate the loss of surfactants, the
remaining structure could not bind the guest very well. The result
is much higher sensitivity toward stimulation; even partial cleavage
can cause complete ejection of the guest. Overall, our entrapment-
release strategy combines the ease of physical entrapment and the
precision of chemical ligation and, as a result, requires no covalent
modification of the entrapped agents. With additional benefits of
multivalent surface modification, outstanding tolerance of the click
reaction for functional groups, and tunable surface charge (e.g., by
using anionic, nonionic, or zwitterionic surfactants with multiple
alkynyl or azide groups), the SCMs may become very useful in
the delivery and controlled release of pharmaceutical agents.
(13) Periodic acid at high concentration (2 × 10-3 M) was found to quench the
fluorescence of pyrene significantly. This could be the reason why the I3/I1
ratio was slightly lower that the 0.70 observed for pyrene below the CMC
of 1. Possibly, not all the vibronic bands were quenched to the same extent.
(14) SCMs prepared with 2 only released pyrene in the presence of periodic
acid, but not with dithiol 5 or at pH ) 5; those prepared with 3 released
pyrene only in the presence of 5, not with periodic acid or at pH ) 5.
SCMs prepared with an uncleavable cross-linker, p-xylyl diazide, on the
other hand, showed no release in the presence of 1 equiv of periodic acid,
5, or at pH ) 5. The corresponding figures are given in the Supporting
Information (Figures 1S-3S).
(15) The concentration of the cross-linked 1 and the cleaving agent was 20 µM.
The aggregation number of 1 was estimated to be 40-50 by DLS. The
concentration of the SCM was thus 0.4-0.5 µM.
(16) (a) Sun, K. H.; Sohn, Y. S.; Jeong, B. Biomacromolecules 2006, 7, 2871–
2877. (b) Koo, A. N.; Lee, H. J.; Kim, S. E.; Chang, J. H.; Park, C.; Kim,
C.; Park, J. H.; Lee, S. C. Chem. Commun. 2008, 6570–6572. (c) Zhang,
L.; Liu, W.; Lin, L.; Chen, D.; Stenzel, M. H. Biomacromolecules 2008,
9, 3321–3331.
(17) Mellman, I.; Fuchs, R.; Helenius, A. Annu. ReV. Biochem. 1986, 55, 663–
700.
(18) (a) Helmlinger, G.; Sckell, A.; Dellian, M.; Forbes, N. S.; Jain, R. K. Clin.
Cancer Res. 2002, 8, 1284–1291. (b) Trevani, A. S.; Andonegui, G.;
Giordano, M.; Lo´pez, D. H.; Gamberale, R.; Minucci, F.; Geffner, J. R.
J. Immunol. 1999, 162, 4849–4857. (c) Engin, K.; Leeper, D. B.; Cater,
J. R.; Thistlethwaite, A. J.; Tupchong, L.; Mcfarlane, J. D. Int. J. Hyper-
thermia 1995, 11, 211–216.
(19) The scattering intensity dropped to 15% of the original value after 48 h.
(20) Overall, the reaction was still quite fast. Hydrolyses of similar benzaldehyde
acetals at comparable pH were reported to take hours to days to complete.
See: Jensen, J. L.; Herold, L. R.; Lenz, P. A.; Trusty, S.; Sergi, V.; Bell,
K.; Rogers, P. J. Am. Chem. Soc. 1979, 101, 4672–4677.
Acknowledgment. This paper is dedicated to Prof. Joseph B.
Lambert for his retirement from Northwestern University. We thank
the NSF (CHE-0748616) for supporting the research.
(21) It is difficult to know exactly how many cross-links an SCM contained.
Digestion of the diol-cross-linked SCMs in our previous work (ref 6)
indicated that the cross-linking density was in line with the stoichiometry
of the reagents when a water-soluble cross-linker (i.e., 2) was employed.
For water-insoluble cross-linkers, such as 3 and 4, the SCMs were expected
to have a lower cross-linking density, as the cross-linkers were not
completely consumed at the end of the cross-linking reaction. SCMs
prepared with 4 had lower cross-linking density probably because the cross-
linker underwent hydrolysis during cross-linking.
Supporting Information Available: Experimental details for the
synthesis and additional figures. This material is available free of charge
via the Internet at http://pubs.acs.org.
References
(1) (a) Lipinski, C. A. J. Pharmacol. Toxicol. Methods 2000, 44, 235–249. (b)
Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J. AdV. Drug
DeliVery ReV. 2001, 46, 3–26.
JA103391K
9
10644 J. AM. CHEM. SOC. VOL. 132, NO. 31, 2010