Molecular design and synthesis of self-assembling camptothecin drug amphiphiles

Article abstract of DOI:10.1038/aps.2016.151

The conjugation of small molecular hydrophobic anticancer drugs onto a short peptide with overall hydrophilicity to create self-assembling drug amphiphiles offers a new prodrug strategy, producing well-defined, discrete nanostructures with a high and quantitative drug loading. Here we show the detailed synthesis procedure and how the molecular structure can influence the synthesis of the self-assembling prodrugs and the physicochemical properties of their assemblies. A series of camptothecin-based drug amphiphiles were synthesized via combined solid- and solution-phase synthetic techniques, and the physicochemical properties of their self-assembled nanostructures were probed using a number of imaging and spectroscopic techniques. We found that the number of incorporated drug molecules strongly influences the rate at which the drug amphiphiles are formed, exerting a steric hindrance toward any additional drugs to be conjugated and necessitating extended reaction time. The choice of peptide sequence was found to affect the solubility of the conjugates and, by extension, the critical aggregation concentration and contour length of the filamentous nanostructures formed. In the design of self-assembling drug amphiphiles, the number of conjugated drug molecules and the choice of peptide sequence have significant effects on the nanostructures formed. These observations may allow the fine-tuning of the physicochemical properties for specific drug delivery applications, ie systemic vs local delivery.

Full text of DOI:10.1038/aps.2016.151

Acta Pharmacologica Sinica 2017: 1–11
© 2017 CPS and SIMM All rights reserved 1671-4083/17
www.nature.com/aps
Original Article
Molecular design and synthesis of self-assembling
camptothecin drug amphiphiles
Andrew G CHEETHAM1, 2, Yi-an LIN1, 2, Ran LIN1, 2, Honggang CUI1, 2, 3, 4,
*
1Department of Chemical and Biomolecular Chemistry and 2Institute for NanoBioTechnology (INBT), Johns Hopkins University,
Baltimore, MD 21211, USA; 3Department of Oncology and Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University
School of Medicine, Baltimore, MD 21205, USA; 4Center for Nanomedicine, The Wilmer Eye Institute, Johns Hopkins University School
of Medicine, Baltimore, MD 21231, USA
Abstract
The conjugation of small molecular hydrophobic anticancer drugs onto a short peptide with overall hydrophilicity to create self-
assembling drug amphiphiles offers a new prodrug strategy, producing well-defined, discrete nanostructures with a high and
quantitative drug loading. Here we show the detailed synthesis procedure and how the molecular structure can influence the synthesis
of the self-assembling prodrugs and the physicochemical properties of their assemblies. A series of camptothecin-based drug
amphiphiles were synthesized via combined solid- and solution-phase synthetic techniques, and the physicochemical properties of
their self-assembled nanostructures were probed using a number of imaging and spectroscopic techniques. We found that the number
of incorporated drug molecules strongly influences the rate at which the drug amphiphiles are formed, exerting a steric hindrance
toward any additional drugs to be conjugated and necessitating extended reaction times. The choice of peptide sequence was found
to affect the solubility of the conjugates and, by extension, the critical aggregation concentration and contour length of the filamentous
nanostructures formed. In the design of self-assembling drug amphiphiles, the number of conjugated drug molecules and the choice
of peptide sequence have significant effects on the nanostructures formed. These observations may allow the fine-tuning of the
physicochemical properties for specific drug delivery applications, ie systemic vs local delivery.
Keywords: camptothecin; chemotherapy; peptides; self-assembly; drug amphiphiles; prodrugs; nanomedicine
Acta Pharmacologica Sinica advance online publication, 6 Mar 2017; doi: 10.1038/aps.2016.151
been introduced into clinical use[31–33], and a number of purely
Introduction
The use of peptides in therapeutic applications, either as drugs peptide-based drugs are on the market[34], peptide–drug con-
themselves[1–3] or as a delivery vector for another bioactive jugates for cancer therapy are lagging behind with regards to
agent[4–12], is a widely explored area within the drug delivery FDA approval and are still undergoing assessment in clinical
community, especially now that their synthesis has become trials. The reasons for this may be that peptide–drug conju-
more facile and cost effective[13, 14]. The versatility in peptide gates can still suffer from the same pharmacokinetic issues
design that the twenty naturally occurring amino acids offer, that afflict both peptide and small molecule drugs[35–37], that
coupled with the potential for high affinity associations toward of rapid excretion, hepatic metabolism and non-specific, pre-
target receptors, make their usage highly attractive, especially mature degradation in plasma during circulation and as such
in the creation of bioactive supramolecular nanomaterials[15–30]
.
no clear benefit has been seen. Polymer-drug conjugates[38–42]
,
In particular, the attachment of anticancer drugs to small in which many drug molecules are covalently attached to
peptides is expected to offer a number of advantages, such a hydrophilic polymer, offer an alternative approach with
as greater aqueous solubility and improved biodistribution improved control over the pharmacokinetic profile as the poly-
through targeting and modified pharmacokinetics. However, mer properties (eg, radius of gyration) can be adjusted accord-
while their larger brethren, antibody–drug conjugates, have ingly to increase circulation time and reduce drug excretion.
A key issue with this approach, however, concerns polydis-
persity both in the polymer size and the number of drug mol-
ecules successfully conjugated per polymer chain. The inevi-
table batch-to-batch variability makes it difficult to accurately
*
To whom correspondence should be addressed.
E-mail hcui6@jhu.edu
Received 2016-10-01 Accepted 2016-11-14

www.nature.com/aps
2
Cheetham AG et al
characterize the individual properties of each polymer-drug by conventional reversed-phase HPLC methods, allowing the
conjugate as only averaged properties can be obtained. As a drug content to be precisely controlled through the molecular
means to address the shortcomings of both approaches, the design.
use of small molecule prodrugs that can assemble into larger
aggregate structures is garnering increasing attention[43–53]
.
Materials and methods
These small molecule prodrugs are designed to assemble into Materials
larger discrete nanostructures, combining the small molecular Fmoc amino acids (unless otherwise stated) and coupling
nature with the controlled pharmacokinetics potential offered reagents (HBTU or HATU) were purchased from Advanced
by larger structures such as polymer-drug conjugates or Automated Peptide Protein Technologies (AAPPTEC, Lou-
nanoparticle-based drug delivery systems[54–56]
.
isville, KY, USA). Rink Amide MBHA resin and Fmoc-Lys
Herein, we outline the design choices involved in the syn- (Fmoc)-OH were obtained from Novabiochem (San Diego, CA,
thesis of peptide-based drug amphiphiles and demonstrate USA). Camptothecin was purchased from AvaChem Scientific
how changes to this design can be used to influence the physi- (San Antonio, TX, USA) and all other reagents were sourced
cochemical properties and self-assembled morphology. The from Sigma-Aldrich (St Louis, MO, USA) or VWR (Radnor,
efficient production of oligo-peptides is best accomplished PA, USA).
using solid-phase synthesis techniques, but the conditions
required for their post-synthesis isolation (strong acid and Instruments and methods
scavengers) can be incompatible with many chemotherapeu- RP-HPLC was performed on a Varian ProStar Model 325
tics and therefore requires that these be introduced using HPLC (Agilent Technologies, Santa Clara, CA, USA) equipped
solution-phase chemistry. This dual approach methodology with a fraction collector. Preparative separations utilized a
is therefore necessary if we are to develop a generic modular Varian PLRP-S column (100 Å, 10 µm, 150 mm×25 mm), whilst
platform that can be altered to incorporate any peptide or analytical HPLC used a Varian Pursuit XRs C18 column (5 µm,
anticancer drug of our choosing, allowing the conjugate to be 150 mm×4.6 mm). Water and acetonitrile containing 0.1%
tailored to a particular therapeutic application. Our proof- v/v TFA were used as the mobile phase. Purified molecules
of-concept design conjugated the hydrophobic drug, camp- were lyophilized using a FreeZone -105°C 4.5 L freeze dryer
tothecin[57, 58], onto a β-sheet forming peptide via a reduction- (Labconco, Kansas City, MO, USA). Mass spectrometric data
sensitive disulfide linker that, under physiological conditions, for characterization was acquired using either a Finnigan
assembles into filamentous structures (Figure 1). We will use LCQ ion trap mass spectrometer (Thermo-Finnigan, Waltham,
these and other conjugates to illustrate how the design choices MA, USA) for ESI-MS or Autoflex III Smartbeam (Bruker,
influence the resulting nanodrugs. The great advantage of this Billerica, MA, USA) for MALDI-Tof MS using α-cyano-4-
approach is that their small molecule nature, in comparison to hydroxycinnamic acid as the matrix. LC-MS analysis was
macromolecules, means they can be purified to homogeneity performed using a Thermo-Finnigan Surveyor LC-MS system
equipped with a PDA spectroscopic detector and an LCQ Fleet
Ion Trap Mass spectrometer. Chromatographic separation
was carried out using an Agilent Eclipse Plus C18 column (50
mm×2.1 mm, 1.8 μm), eluting with a water-acetonitrile gra-
dient containing 0.1% v/v formic acid. Data was processed
using Thermo XCaliber software. Bruker Avance 300 or 400
MHz FT-NMR spectrometers were used for the acquisition of
1H and 13C NMR spectra.
Peptide synthesis
The peptides Ac-CGVQIVYKK-NH2 (mCys-Tau), (Ac-C)2-
KGVQIVYKK-NH2 (dCys-Tau), and [(Ac-C)2K]2KGVQIVYKK-
NH2 (qCys-Tau) were synthesized using a combination of
automated and manual Fmoc-solid phase synthesis techniques
(Scheme 1). A Focus XC automated peptide synthesizer
(AAPPTEC, Louisville, KY, USA) was used for the synthesis
of Fmoc-GVQIVYKK-Rink, using 20% 4-methylpiperidine in
DMF for Fmoc deprotections and amino acid/HBTU/DIEA
(4:4:6 relative to the resin) in DMF for couplings (with 2 min
activation time and 1 h reaction time), further modifications
Figure 1. Illustration of the drug amphiphile concept. A hydrophilic
peptide is conjugated to a hydrophobic drug via a degradable linker.
were then carried out manually. For dCys-Tau and qCys-Tau,
the branching lysines were introduced using Fmoc-Lys(Fmoc)-
OH/HATU/DIEA (4:4:6 relative to each reactive amine),
coupling until a negative Kaiser test was obtained, requiring
Under physiological conditions, the synthesized drug amphiphiles
described self-assemble into nanofibrous structures with a core micelle
structure as shown in the representative transmission electron microscopy
(TEM) image of mCPT-buSS-Tau (scale bar is 100 nm).
Acta Pharmacologica Sinica

www.chinaphar.com
Cheetham AG et al
3
Scheme 1. Solid-phase synthesis of the cysteine-functionalized Tau precursor peptides. Fmoc-GVQIVYKK-Rink was created using an automated
peptide synthesizer and further modified by manual synthesis techniques. Reaction conditions: (a) (i) 20% 4-methylpiperidine in DMF, (ii). Fmoc-
Lys(Fmoc)-OH, HATU, DIEA (4:3.98:6 per amine); (b) (i). 20% 4-methylpiperidine in DMF, (ii) Fmoc-Cys(Trt)-OH, HATU, DIEA (4:3.98:6 per amine), (iii) 20%
4-methylpiperidine in DMF, (iv) 20% acetic anhydride in DMF, DIEA; (c) TFA, TIS, H2O (95:2.5:2.5); (d) TFA, TIS, H2O, EDT (90:5:2.5:2.5).
extended reaction times and on occasion repeated reaction purity confirmed by MALDI-Tof (Figures S2-S4 in Supplemen-
with fresh reagents. Cysteine incorporation for all three pep- tary Information, SI). The four cysteine-containing peptides
tides was achieved using Fmoc-Cys(Trt)-OH/HATU/DIEA were purified by initially dissolving in 1 mL AcOH and then
(4:4:6 relative to each reactive amine). Acetylation was carried diluting with 0.1% v/v aqueous TFA to 20 mL.
out manually using 20% v/v acetic anhydride in DMF after
N-terminal Fmoc deprotection.
4-(Pyridin-2-yldisulfanyl)butanoic acid (HO2C-BuSS-Pyr)
Peptides were cleaved from the resin using the standard This was synthesized using a modification of a previously
cleavage solution of TFA/TIS/H2O (95:2.5:2.5) for 2 h, with published method[8]. Briefly, 4-bromobutyric acid (2.0 g, 12.0
the exception of dCys-Tau, and qCys-Tau, which used TFA/ mmol) and thiourea (1.06 g, 14.0 mmol) were refluxed in EtOH
TIS/H2O/EDT (90:5:2.5:2.5). Cleaved peptides were isolated (50 mL) for 4 h. NaOH (4.85 g, 121 mmol) in EtOH (50 mL)
by trituration into cold diethyl ether, followed by filtration was added and reflux was continued for 16 h. After cooling
and drying under suction. All peptides were purified to >95% to room temperature and concentration in vacuo, the residue
homogeneity by preparative RP-HPLC, with their identity and was diluted to 50 mL with water that was extracted twice with
Acta Pharmacologica Sinica

www.nature.com/aps
4
Cheetham AG et al
Et2O. The aqueous portion was then acidified to pH 5 with 4 be 233 μmol/L (8.9 mg, 43%). The solution was then aliquot-
mol/L HCl, giving a cloudy solution that was extracted with ted into cryo-vials, lyophilized and stored at -30°C.
Et2O. The organic extracts were dried over anhydrous Na2SO4
and concentrated to give 4-sulfanylbutyric acid as a clear oil dCPT-buSS-Tau
(802 mg, 56%) that was used without further purification. dCys-Tau (10.8 mg, 5.0 mmol) was dissolved in an N2-purged
4-Sulfanylbutyric acid (802 mg, 6.7 mmol) was dissolved in DMSO solution of CPT-buSS-Pyr (9 mg in 500 μL, 16.1 μmol)
MeOH (5 mL) and added dropwise to a solution of 2-aldrithiol and allowed to react for 3 d. The solution was diluted to 10
(3.03 g, 13.8 mmol) in MeOH (5 mL), which developed a yel- mL with 0.1% v/v aqueous TFA and purified by RP-HPLC.
low color. After 3 h, the mixture was purified by RP-HPLC, Product fractions were combined and immediately lyophi-
collecting the major peak and removing the solvents in vacuo. lized. The pale yellow solid obtained was dissolved in 15 mL
The resultant oil was dissolved in CHCl3, dried over Na2SO4 nanopure water containing 0.1% TFA and 8% acetonitrile and
and solvent removed to give HO2C-BuSS-Pyr as a pale yellow the product concentration was determined by DTT calibration
viscous oil (1.02 g, 67%). 1H NMR (CDCl3, 400 MHz, 298 K): to be 72.9 μmol/L (2.5 mg, 22%). The solution was then ali-
δH (ppm) 8.59 (d, 3JHH=4.6, 1H), 7.91–7.81 (m, 2H), 7.30–7.25 (m, quotted into cryo-vials, lyophilized and stored at -30°C.
3
3
1H), 2.88 (t, JHH=7.1, 2H), 2.50 (t, JHH=7.2, 2H), 2.09-2.00 (m,
2H).
qCPT-buSS-Tau
qCys-Tau (3.5 mg, 1.91 μmol) was dissolved in an N2-purged
DMSO solution of CPT-buSS-Pyr (10 mg in 500 μL, 17.8 μmol)
Camptothecin-4-(pyridin-2-yldisulfanyl)butanoate (CPT-buSS-Pyr)
Camptothecin (200 mg, 574 μmol) was suspended in DCM (32 and allowed to react for 8 d. The solution was diluted to 10
mL) and dimethylaminopyridine (44 mg, 360 μmol), HO2C- mL with 0.1% v/v aqueous TFA and purified by RP-HPLC.
BuSS-Pyr (280 mg, 1.22 mmol) and diisopropylcarbodiimide Product fractions were combined and immediately lyophi-
(436 μL, 2.80 mmol) were added. The mixture was stirred for lized. The pale yellow solid obtained was dissolved in 19.5
36 h, with TLC (3% MeOH in CHCl3) showing complete con- mL nanopure water containing 0.05% TFA and 25% acetoni-
sumption. The solution was then filtered, diluted with CHCl3 trile and the product concentration was determined by DTT
(60 mL), extracted with sat. NaHCO3 (50 mL), brine (50 mL), calibration to be 14.2 μmol/L (1.0 mg, 15%). The solution
dried over Na2SO4 and concentrated in vacuo. The residue was was then aliquotted into cryo-vials, lyophilized and stored at
purified by flash chromatography using EtOAc (500 mL) then -30°C.
0.5% MeOH in EtOAc (250 mL). Product fractions were iden-
tified by TLC, combined and solvent removed in vacuo to give Concentration calibration
1
CPT-buSS-Pyr as a pale yellow solid (195 mg, 61%); H NMR To determine the CPT concentration of the purified drug
3
(CDCl3, 400 MHz, 298 K): δH (ppm) 8.43 (d, JHH=4.2, 1H), 8.40 amphiphiles, a dithiothreitol (DTT) reduction assay was devel-
3
3
(s, 1H), 8.23 (d, JHH=8.6, 1H), 7.94 (d, JHH=8.2, 1H), 7.84 (m, oped. Briefly, a solution of the conjugate with an approximate
1H), 7.70–7.65 (m, 2H), 7.60 (m, 1H), 7.20 (s, 1H), 7.04 (m, 1H), concentration in the range 5 to 500 μmol/L was adjusted to pH
5.67 (d, 2JHH=17.3, 1H), 5.40 (d, 2JHH=17.2, 1H), 5.29 (s, 2H), 2.86 7 with a known amount of 0.1 mol/L NaOH. 10 μL of freshly
3
(t, JHH=7.1, 2H), 2.75–2.57 (m, 2H), 2.31–2.03 (m, 4H), 0.97 (t, prepared 3 mol/L aqueous DTT was added to 90 μL of the
3JHH=7.5); 13C NMR (CDCl3, 100 MHz, 298 K): δC (ppm) 172.1, conjugate solution and allowed to stand for 3 h with periodic
167.7, 160.3, 157.6, 152.6, 149.9, 149.1, 146.6, 146.1, 137.3, 131.5, vortexing. 30 μL of this sample was then mixed with 20 μL
131.4, 130.9, 129.9, 128.63, 128.62, 128.4, 128.3, 120.4, 96.1, 76.2, of DMSO and analyzed by RP-HPLC (injecting 50 μL so as to
67.4, 50.2, 37.7, 32.4, 32.0, 31.2, 24.0, 7.9; MS (MALDI-TOF): completely fill the 20 μL sample loop), measuring the area of
560.065 [M+H]+.
the peak due to CPT-buSH. The CPT concentration of the ana-
lyzed solution in μmol/L is given by (peak area)/0.2964 (as
determined by a calibration curve study of DTT-reduced CPT-
Synthesis of the drug-amphiphiles
All drug amphiphiles were synthesized by reaction of the pep- buSS-Pyr, section S1.4 in SI). The conjugate concentration of
tide cysteine thiol(s) with the activated disulfide of CPT-buSS- the original solution was calculated based on the applied dilu-
Pyr in N2-purged DMSO (Scheme 2), followed by RP-HPLC tions and number of CPT molecules the conjugate possesses.
purification as described below in more detail.
Transmission electron microscopy (TEM) imaging
mCPT-buSS-Tau
A 100 μmol/L aqueous solution of drug amphiphile was pre-
mCys-Tau (14.6 mg, 13.5 μmol) was dissolved in an N2-purged pared from a stock solution of 1 mmol/L conjugate in water.
DMSO solution CPT-buSS-Pyr (10 mg in 1.50 mL, 17.8 μmol) Samples were aged overnight prior to sample preparation. A
and shaken overnight. The reaction was diluted to 30 mL sample for imaging was prepared by depositing 7 μL of the
with 0.1% v/v aqueous TFA, giving a slightly viscous solution solution onto a carbon-coated copper grid (Electron Micros-
that was then purified by RP-HPLC. Product fractions were copy Services, Hatfield, PA, USA), wicking away the excess
combined and immediately lyophilized. The pale yellow solid solution with a small piece of filter paper. Next, 7 μL of a 2%
obtained was dissolved in 25 mL nanopure water and the (wt) aqueous uranyl acetate solution was deposited and the
product concentration was determined by DTT calibration to excess solution was carefully removed as above to leave a very
Acta Pharmacologica Sinica

www.chinaphar.com
Cheetham AG et al
5
Scheme 2. (A) Directed disulfide formation method used to synthesize the described drug amphiphiles and (B) molecular structures of the three drug
amphiphiles mCPT-buSS-Tau, dCPT-buSS-Tau and qCPT-buSS-Tau and their respective drug loadings.
thin layer. The sample grid was then allowed to dry at room sample solution was produced using a Vitrobot with a con-
temperature prior to imaging. Bright-field TEM imaging was trolled humidity chamber (FEI). After loading of the sample
performed on a FEI Tecnai 12 TWIN Transmission Electron solution, the lacey carbon grid was blotted using preset
Microscope operated at an acceleration voltage of 100 kV. All parameters and plunged instantly into a liquid ethane reser-
TEM images were recorded by a SIS Megaview III wide-angle voir precooled by liquid nitrogen. Vitrified samples were then
CCD camera.
transferred to a cryo-holder and cryo-transfer stage that was
cooled by liquid nitrogen. To prevent sublimation of vitreous
water, the cryo-holder temperature was maintained below
Cryogenic-TEM (Cryo-TEM) imaging
3–5 µL of the sample solution was placed on a holey carbon -170°C during the imaging process. Imaging was performed
film supported on a TEM copper grid (Electron Microscopy on the FEI Tecnai 12 TWIN Transmission Electron Microscope,
Services, Hatfield, PA, USA) that had been pre-treated with operating at 80 kV. All images were recorded by a 16 bit
plasma air to render the film hydrophilic. A thin film of the 2K×2K FEI Eagle bottom mount camera.
Acta Pharmacologica Sinica

www.nature.com/aps
6
Cheetham AG et al
Circular dichroism (CD)
The CD spectra were recorded on a Jasco J-710 spectropolar-
acidic water to the crude solids.
imeter (JASCO, Easton, MD, USA) using a 1 mm path length Prodrug linker synthesis
quartz UV-Vis absorption cell (Thermo Fisher Scientific, Pitts- The activated disulfide prodrug, CPT-buSS-Pyr, was synthe-
burgh, PA, USA). A background spectrum of the solvent was sized by adapting a procedure used to create a similar pacli-
acquired and subtracted from the sample spectrum. Collected taxel prodrug[8]. CPT was esterified by reaction with 4-(pyr-
data were normalized with respect to sample concentration idin-2-yldisulfanyl)butanoic acid in the presence of DIC and
and β-sheet forming residues.
DMAP, giving CPT-buSS-Pyr in 61% yield after purification.
Synthesis of mCPT-buSS-Tau
Results
Peptide synthesis
Reaction of mCys-Tau with an excess of the activated disulfide
The peptides used in this study were synthesized using stan- CPT-buSS-Pyr in N2-purged DMSO was found to give rapid
dard Fmoc solid-phase peptide synthesis techniques (Scheme conversion to the desired conjugate, mCPT-buSS-Tau, by
1), employing an automated synthesizer to make the backbone both ESI-MS and HPLC analysis (Figures 2A and 3A). After
of the sequence –GVQIVYKK-Rink– and manual synthesis overnight agitation to allow the reaction to go to completion,
techniques to add the requisite cysteine residues. For the purification by reversed phase HPLC and subsequent lyophili-
mono-Cys peptide, no issues were observed for the coupling zation of product fractions gave mCPT-buSS-Tau in 43% yield,
of cysteine or during cleavage from the resin when using with >99% homogeneity (by HPLC).
standard protocols. For both the di- and quad-Cys peptides,
however, extended reaction time or double couplings were Synthesis of dCPT-buSS-Tau
required for complete reaction to occur. Cleavage also pre- The synthesis of the di-conjugate was performed in a similar
sented a problem and a solution containing a greater amount manner to that of mCPT-buSS-Tau, though a longer reaction
of the scavenger triisopropylsilane (5% compared to 2.5%), in time was required to achieve near complete reaction. The
addition to ethane-1,2-diol (2.5%), was required to obtain the formation of the semi-reacted conjugate containing one CPT
product, otherwise significant degradation to unidentifiable moiety could be clearly seen (Figures 2B and 3B), with HPLC
species was found to occur. Purification of these peptides analysis indicating the formation of the two isomeric forms.
required initial dissolution in acetic acid to avoid the forma- Purification as above gave dCPT-buSS-Tau in 22% yield, with
tion of a gelatinous solid that would result upon addition of >99% homogeneity (by HPLC).
Figure 2. ESI-MS analysis of the reaction solutions during drug amphiphile synthesis. mCPT-buSS-Tau after 1 h (A), dCPT-buSS-Tau after 3 h (B), and
qCPT-buSS-Tau after 18 h (C). 1×, 2×, and 3× CPT indicate singly, doubly and triply reacted products, respectively.
Acta Pharmacologica Sinica

www.chinaphar.com
Cheetham AG et al
7
Figure 3. Preparative HPLC chromatograms of mCPT-buSS-Tau (A), dCPT-buSS-Tau (B), and qCPT-buSS-Tau (C) showing the reaction products formed
during the synthesis of the drug amphiphiles. 1×, 2× and 3× CPT indicate incomplete reaction products.
Synthesis of qCPT-buSS-Tau
development of degradable linker groups has been ongoing
The formation of the quad-conjugate was observed to take for several decades and there are many examples in the litera-
much longer to occur, even with a larger excess of CPT-buSS- ture that can be utilized, including acid-sensitive[69], enzyme-
Pyr, ESI-MS analysis (Figure 2C) clearly showed the formation cleavable[70, 71], and reducible[72, 73]
.
of mono-, di- tri- and quad-reacted species after overnight
Third, since many chemotherapeutics can be sensitive to
agitation. Despite the decreased rate of reaction, the cysteine strongly acidic conditions, the use of conjugation methods
thiols remained reactive and after 8 d significant conversion for post-cleavage attachment to the peptide is often required
had been achieved (Figure 3C). Purification was therefore per- since the coupling cannot be performed on-resin. These con-
formed as before, giving qCPT-buSS-Tau in 15% yield, with jugation methods necessitate reaction mechanisms that must
>99% homogeneity (by HPLC).
be selective and efficient so as to minimize side-reactions. A
multitude of reactions for the conjugation of synthetic entities
to proteins have been developed that satisfy these conditions
and can be readily applied for the purpose of peptide–drug
Discussion
Drug amphiphile design and synthesis
Our approach for the synthesis of peptide-based amphiphilic conjugation. Examples of these include directed disulfide for-
drug conjugates is the attachment of the hydrophobic che- mation[74], thiolene reaction[75] or copper-assisted azide-alkyne
motherapeutic agent to a peptide with overall hydrophilicity cycloaddition (CuAAC)[76]. Amide and ester bond formation
via a biodegradable linker (Figure 1). When designing such a are also possibilities, but given the drug amphiphile requires
conjugate, there are a number of important points that must be a charged head-group (eg, lysine, glutamic or aspartic acids)
considered. First, the peptide can be chosen so as to influence there is a greater potential for unwanted side-reactions.
both the assembly properties of the conjugate and the surface
Last, the conditions required for purification of the final con-
chemistry of the assembled nanostructure[59]. The preferred jugate must not only be compatible with both the drug and the
secondary structure that the peptide adopts can strongly influ- chosen linker group, but must also disfavor self-assembly of
ence the final assembly morphology, with β-sheet-forming the conjugate since large aggregates will have poorer retention
peptides preferentially giving one-dimensional (1D) fibrous characteristics on reverse-phased silica when compared with
structures[60, 61], whereas other secondary structures could the monomer.
give rise to micelles or vesicles[62, 63]. Peptide-based epitopes
To create a self-assembling drug amphiphile, we conjugated
or other moieties could also be incorporated so as to bestow the DNA topoisomerase I inhibitor, camptothecin (Pom-
further functionality upon the drug nanostructures, such as mier, 2006 #4) to a β-sheet forming peptide derived from the
targeting[18, 64–66] and stealth capabilities[67, 68]
.
Tau protein[77] –“C”GVQIVYKK (“C” indicates a cysteine-
Second, the linker choice is important as release of the free containing segment) (Scheme 1). The reducible disulfylbutyr-
drug is essential if a cytotoxic effect is to be induced. The ate[8] moiety was chosen to link the two segments of the drug
Acta Pharmacologica Sinica

www.nature.com/aps
8
Cheetham AG et al
amphiphile, employing the activated disulfide CPT-buSS-Pyr longer reaction time and a greater reagent excess were found
to direct disulfide bond formation with a cysteine-thiol of the to promote more complete conjugation.
peptide to form the conjugate (Scheme 2A). This linker can
undergo reduction in the intracellular environment, breaking Effect of peptide segment on stability
down to release the free drug. The two lysine residues at the One property of the Tau conjugates described above that
C-terminal of the peptide serve as a hydrophilic head group to became apparent was how the solubility of the resultant conju-
promote solubility and enable purification under acidic condi- gate is reduced with the increasing number of CPT molecules
tions, where assembly would be disrupted due to electrostatic attached. mCPT-buSS-Tau was found to be very soluble,
repulsions[78]. The loading of the drug amphiphile can be pre- with solutions above 2 mol/L remaining clear, albeit with an
cisely controlled by simply modifying the peptide design to increase in viscosity that led to gel formation. dCPT-buSS-Tau
incorporate multiple attachment sites. In our design, we used remained soluble up to 1 mol/L, becoming cloudy thereaf-
the two amine functionalities of lysine to introduce branching ter. qCPT-buSS-Tau on the other hand was sparingly soluble,
points at the N-terminal that were then functionalized with giving a turbid solution even at 100 µmol/L. This observa-
cysteine, allowing access to drug amphiphiles with 1, 2 or 4 tion may explain why only relatively short nanotubes were
drug molecules per conjugate (23%, 31% and 38% loadings, observed by TEM, as the availability of soluble monomers
respectively, Scheme 2B).
From a practical standpoint, the branched peptides can be leading to formation of kinetically trapped short nanotubes.
synthesized through conventional Fmoc solid-phase peptide One method to improve solubility is to incorporate a more
may reduce the likelihood of long nanotubes being formed,
synthesis procedures (Scheme 1). The only caveat to this, hydrophilic peptide sequence. The Tau peptide is dominated
however, is that as the extent of branching increases (from one by hydrophobic residues (Val and Ile) and as such may con-
to four Cys attachment points), so too does the relative dif- tribute to the poor solubility characteristics. We have synthe-
ficulty of achieving complete coupling. This can be attributed sized analogues of dCPT-buSS-Tau and qCPT-buSS-Tau using
to the bulky Fmoc-Cys(tBu)-OH units being coupled to the a more hydrophilic β-sheet forming peptide derived from
branching lysines, with each subsequent addition increasing the Sup35 yeast prion, NNQQNY (Figure 4A)[79, 80], allowing
the steric bulk around the resin-bound peptide and hindering a direct comparison of their properties. In the case of dCPT-
the approach of the next activated Cys residue to be added. buSS-Sup35[79], we saw little difference in the morphology
Complete coupling was achieved through longer reaction formed with that of dCPT-buSS-Tau, giving nanofilamentous
times and repeated coupling with fresh reagents. Similar structures with widths of 8.9±1.3 nm—slightly wider than
issues also arose when conjugating the CPT prodrug, as each those of the Tau analogue. However, a clear difference was
addition will add a bulky CPT unit that can hinder further seen in the contour length, with the Sup35 conjugate forming
reaction for the di- and quad-substituted conjugates. Again, fibrous nanostructures that were typically longer than 1 µm
Figure 4. Comparison of CPT conjugates with Tau and Sup35 peptide segments. (A) Chemical structures of dCPT-buSS-Sup35 and qCPT-buSS-Sup35.
Representative TEM images of dCPT-buSS-Sup35 (B), dCPT-buSS-Tau (C), qCPT-buSS-Sup35 (D) and qCPT-buSS-Tau (E). All samples are 100 µmol/L in
H2O, except for dCPT-buSS-Tau which is 200 µmol/L. Scale bars are 100 nm.
Acta Pharmacologica Sinica

www.chinaphar.com
Cheetham AG et al
9
(Figure 4B), while the Tau conjugate formed predominantly the potential biocompatibility of the chosen peptide segments.
shorter filaments (Figure 4C). Cryo-TEM imaging confirmed However, for using these materials in an in vivo setting,
that these observations were not an artifact of the drying pro- more comprehensive studies are necessary to evaluate their
cess involved in sample preparation (Supplementary Figure unwanted toxicities in interfacing with other biomolecules.
S9). A further effect of the more hydrophilic peptide was Further design of self-assembling drug amphiphiles should
found in the apparent critical micellization concentrations avoid using amyloid disease-related sequences as the auxiliary
(CMC), with that of dCPT-buSS-Sup35 being around 20–30 segment. Given the supramolecular nature of these filamen-
µmol/L in comparison to submicromolar for dCPT-buSS-Tau. tous assemblies, the concern related to their biodegradability
This is borne out by the circular dichroism (CD) spectra of the is less an issue here, as we have demonstrated that these nano-
two drug amphiphiles (Supplementary Figure S10), in which structures can effectively release the incorporated therapeutic
it can be clearly seen that the β-sheet signal of the Sup35 ana- agents.
logue is no longer apparent at 5 µmol/L, whereas that of the
Tau conjugate is still present below 1 µmol/L. It is evident
Conclusion
that the greater hydrophilicity of the Sup35 sequence lends In summary, we have detailed the guiding principles for the
a greater solubility to the resulting conjugate at the cost of a synthesis of anticancer drug amphiphiles, utilizing an estab-
reduction in its propensity for self-assembly. Importantly, lished bioconjugation technique to create peptide-based pro-
this effect on monomer solubility may provide a pathway for drugs that can self-assemble into nanofibrous structures. The
exerting some degree of control over the contour length of the sequence of the conjugated peptide segment can not only be
one-dimensional nanostructures. This ability would be useful used to direct the self-assembly, but can also be utilized to
in tailoring the drug amphiphiles toward a particular applica- control the solubility characteristics and influence the physico-
tion. Longer fibers are able to more effectively entangle than chemical properties of the resulting nanostructures. Given the
shorter fibers and would be ideal candidates for the formation range of hydrophobic anticancer drugs in current clinical use
of hydrogels, which can be utilized for local and sustained and the versatility present in peptide design, the approach we
delivery of therapeutics[81–83]. On the other hand, shorter nano- are developing has enormous potential as the basis for a future
structures would be more suitable for systemic delivery as drug delivery platform.
they may fare better during circulation, being less likely to get
trapped by the body’s filtration mechanisms or in smaller cap-
Acknowledgements
illary blood vessels.
The work is supported by the National Science Foundation
Incorporation of the Sup35 peptide into the quad-CPT con- (DMR 1255281), USA. We thank National Institutes of Health
jugate results in a marked improvement in the solubility[80]
,
(NIH) for funding Andrew G CHEETHAM (T-32CA130840).
as clear solutions of the qCPT-buSS-Sup35 conjugate can be We acknowledge the use of the Johns Hopkins Department
prepared that are above 1 mol/L in concentration. These of Chemistry NMR and Mass Spectrometry Shared Facilities
solutions also exhibit greater viscosity, indicating the pres- and the Integrated Imaging Center for access to the relevant
ence of longer nanostructures that can entangle. TEM imag- instrumentation. Financial support for MALDI-Tof analysis
ing confirms the formation of nanotubes of widths 9.9±1.1 nm was provided by NSF Grant CHE-0840463. We thank Dr Phil
(Figure 4D), similar to those of qCPT-buSS-Tau (9.5±1.0 nm, MORTIMER for assistance with LC-MS and Prof Kalina HRIS-
Figure 4E), though a larger proportion were greater than 1 µm TOVA for use of the CD spectropolarimeter.
in length. The increased solubility the Sup35 peptide allows
the nanotubes to further elongate during self-assembly. The
Author contribution
hydrophobicity of the four CPT molecules clearly still plays an Andrew G CHEETHAM and Honggang CUI were respon-
important role in the self-assembly, but the incorporation of sible for the research design and data analysis; Andrew
the Sup35 peptide bestows a greater hydrophobic-hydrophilic G CHEETHAM performed all syntheses, purifications,
balance to the conjugate that increases its solubility without and molecular and spectroscopic characterizations; Yi-an
compromising its stability.
LIN carried out TEM and Cryo-TEM imaging; Andrew G
Two concerns associated with β-sheet-based peptide assem- CHEETHAM, Ran LIN, and Honggang CUI wrote the paper.
blies are their potential toxicity and degradability for body
clearance. In the case reported here, the two β-sheet segments
Supplementary Information
we chose, Tau and Sup35, are derived from natural proteins 1H and 13C NMR, HPLC chromatograms and mass spectromet-
known to form amyloid plaques. After conjugation with ther- ric characterization of the purified materials can be found in
apeutic agents and modification with charged amino acids, the Supplementary Information on the website of Acta Phar-
it is likely that their potential to induce formation of amyloid macologica Sinica.
plaques is altered. In our in vitro studies with cells, although
all the synthesized drug amphiphiles demonstrated effective
References
toxicities against a number of cancer cell lines, no adverse
effects were observed for filaments formed by self-assembly of
the peptides conjugated with a linear hydrocarbon, suggesting
1
Lien S, Lowman HB. Therapeutic peptides. Trends Biotechnol 2003;
21: 556–62.
2
Albericio F, Kruger HG. Therapeutic peptides. Future Med Chem
Acta Pharmacologica Sinica

www.nature.com/aps
10
Cheetham AG et al
2012; 4: 1527–31.
20: 37–41.
3
4
5
6
Ahrens VM, Bellmann-Sickert K, Beck-Sickinger AG. Peptides and 24 Ceylan H, Tekinay AB, Guler MO. Selective adhesion and growth of
peptide conjugates: therapeutics on the upward path. Future Med
Chem 2012; 4: 1567–86.
vascular endothelial cells on bioactive peptide nanofiber function-
alized stainless steel surface. Biomaterials 2011; 32: 8797–805.
Arap W, Pasqualini R, Ruoslahti E. Cancer treatment by targeted drug 25 Cinar G, Ceylan H, Urel M, Erkal TS, Tekin ED, Tekinay AB, et al.
delivery to tumor vasculature in a mouse model. Science 1998; 279:
Amyloid inspired self-assembled peptide nanofibers. Biomacro-
377–80.
molecules 2012; 13: 3377–87.
Burkhart DJ, Kalet BT, Coleman MP, Post GC, Koch TH. Doxorubicin- 26 Schneider JP, Pochan DJ, Ozbas B, Rajagopal K, Pakstis L, Kretsinger
formaldehyde conjugates targeting αvβ3 integrin. Mol Cancer Ther
2004; 3: 1593–604.
J. Responsive hydrogels from the intramolecular folding and self-
assembly of a designed peptide. J Am Chem Soc 2002; 124: 15030–
Wang S, Placzek WJ, Stebbins JL, Mitra S, Noberini R, Koolpe M, et al.
7.
Novel targeted system to deliver chemotherapeutic drugs to EphA2- 27 Li JY, Kuang Y, Gao Y, Du XW, Shi JF, Xu B. D-amino acids boost the
expressing cancer cells. J Med Chem 2012; 55: 2427–36.
Brooks H, Lebleu B, Vives E. Tat peptide-mediated cellular delivery:
back to basics. Adv Drug Deliv Rev 2005; 57: 559–77.
Dubikovskaya EA, Thorne SH, Pillow TH, Contag CH, Wender PA.
Overcoming multidrug resistance of small-molecule therapeutics
selectivity and confer supramolecular hydrogels of a nonsteroidal anti-
inflammatory drug (NSAID). J Am Chem Soc 2013; 135: 542–5.
28 Cheetham AG, Keith D, Zhang PC, Lin R, Su H, Cui HG. Targeting
tumors with small molecule peptides. Curr Cancer Drug Targets
2016; 16: 489–508.
7
8
through conjugation with releasable octaarginine transporters. Proc 29 Su H, Zhang PC, Cheetham AG, Koo JM, Lin R, Masood A, et al.
Natl Acad Sci U S A 2008; 105: 12128–33.
Supramolecular crafting of self-assembling camptothecin prodrugs
with enhanced efficacy against primary cancer cells. Theranostics
2016; 6: 1065–74.
9
Soukasene S, Toft DJ, Moyer TJ, Lu HM, Lee HK, Standley SM, et
al. Antitumor activity of peptide amphiphile nanofiber-encapsulated
camptothecin. ACS Nano 2011; 5: 9113–21.
30 Lock LL, Tang Z, Keith D, Reyes C, Cui HG. Enzyme-specific doxoru-
bicin drug beacon as drug-resistant theranostic molecular probes.
ACS Macro Lett 2015; 4: 552–5.
31 Sievers EL, Linenberger M. Mylotarg: Antibody-targeted chemotherapy
comes of age. Curr Opin Oncol 2001; 13: 522–7.
10 Zhang PC, Cheetham AG, Lock LL, Cui HG. Cellular uptake and
cytotoxicity of drug-peptide conjugates regulated by conjugation site.
Bioconjugate Chem 2013; 24: 604–13.
11 Wang Y, Cheetham AG, Angacian G, Su H, Xie L, Cui H. Peptide-drug
conjugates as effective prodrug strategies for targeted delivery. Adv 32 Younes A, Yasothan U, Kirkpatrick P. Brentuximab vedotin. Nat Rev
Drug Deliv Rev 2016; pii: S0169-409X(16)30208-3. Drug Discov 2012; 11: 19–20.
12 Lin R, Zhang PC, Cheetham AG, Walston J, Abadir P, Cui HG. Dual 33 Traynor K. Ado-trastuzumab emtansine approved for advanced breast
peptide conjugation strategy for improved cellular uptake and
mitochondria targeting. Bioconjugate Chem 2015; 26: 71–7.
13 Zompra AA, Galanis AS, Werbitzky O, Albericio F. Manufacturing
peptides as active pharmaceutical ingredients. Future Med Chem
2009; 1: 361–77.
cancer. Am J Health Sys Pharm 2013; 70: 562.
34 Vlieghe P, Lisowski V, Martinez J, Khrestchatisky M. Synthetic thera-
peutic peptides: science and market. Drug Discov Today 2010; 15:
40–56.
35 Tang L, Persky AM, Hochhaus G, Meibohm B. Pharmacokinetic
aspects of biotechnology products. J Pharm Sci 2004; 93: 2184–
204.
14 Su H, Koo JM, Cui HG. One-component nanomedicine. J Control
Release 2015; 219: 383–95.
15 Hartgerink JD, Beniash E, Stupp SI. Self-assembly and mineralization 36 Evans WE, Relling MV. Clinical pharmacokinetics-pharmacodynamics
of peptide-amphiphile nanofibers. Science 2001; 294: 1684–8. of anticancer drugs. Clin Pharmacokinet 1989; 16: 327–36.
16 Rajangam K, Behanna HA, Hui MJ, Han XQ, Hulvat JF, Lomasney JW, 37 Undevia SD, Gomez-Abuin G, Ratain MJ. Pharmacokinetic variability
et al. Heparin binding nanostructures to promote growth of blood
vessels. Nano Lett 2006; 6: 2086–90.
17 Standley SM, Toft DJ, Cheng H, Soukasene S, Chen J, Raja SM, et
of anticancer agents. Nat Rev Cancer 2005; 5: 447–58.
38 Duncan R. Polymer conjugates as anticancer nanomedicines. Nat
Rev Cancer 2006; 6: 688–701.
al. Induction of cancer cell death by self-assembling nanostructures 39 Tong R, Cheng J. Anticancer polymeric nanomedicines. Polym Rev
incorporating a cytotoxic peptide. Cancer Res 2010; 70: 3020–6. 2007; 47: 345–81.
18 Pakalns T, Haverstick KL, Fields GB, McCarthy JB, Mooradian DL, 40 Cheng JJ, Khin KT, Jensen GS, Liu AJ, Davis ME. Synthesis of linear,
Tirrell M. Cellular recognition of synthetic peptide amphiphiles in self-
assembled monolayer films. Biomaterials 1999; 20: 2265–79.
beta-cyclodextrin-based polymers and their camptothecin conjugates.
Bioconjugate Chem 2003; 14: 1007–17.
19 Trent A, Marullo R, Lin B, Black M, Tirrell M. Structural properties of 41 Zou J, Jafr G, Themistou E, Yap Y, Wintrob ZAP, Alexandridis P, et
soluble peptide amphiphile micelles. Soft Matter 2011; 7: 9572–82.
20 Jun HW, Yuwono V, Paramonov SE, Hartgerink JD. Enzyme-mediated
al. pH-Sensitive brush polymer-drug conjugates by ring-opening
metathesis copolymerization. Chem Commun 2011; 47: 4493–5.
degradation of peptide-amphiphile nanofiber networks. Adv Mater 42 Yu Y, Chen CK, Law WC, Mok J, Zou J, Prasad PN, et al. Well-defined
2005; 17: 2612–7.
degradable brush polymer-drug conjugates for sustained delivery of
21 Gauba V, Hartgerink JD. Self-assembled heterotrimeric collagen triple
paclitaxel. Mol Pharm 2013; 10: 867–74.
helices directed through electrostatic interactions. J Am Chem Soc 43 Gao Y, Kuang Y, Guo ZF, Guo ZH, Krauss IJ, Xu B. Enzyme-instructed
2007; 129: 2683–90.
molecular self-assembly confers nanofibers and a supramolecular
22 Toledano S, Williams RJ, Jayawarna V, Ulijn RV. Enzyme-triggered self-
hydrogel of taxol derivative. J Am Chem Soc 2009; 131: 13576–7.
assembly of peptide hydrogels via reversed hydrolysis. J Am Chem 44 Shen Y, Jin E, Zhang B, Murphy CJ, Sui M, Zhao J, et al. Prodrugs
Soc 2006; 128: 1070–1.
forming high drug loading multifunctional nanocapsules for intra-
23 Smith AM, Williams RJ, Tang C, Coppo P, Collins RF, Turner ML, et
cellular cancer drug delivery. J Am Chem Soc 2010; 132: 4259–65.
al. Fmoc-Diphenylalanine self assembles to a hydrogel via a novel 45 Cheetham AG, Zhang P, Lin YA, Lock LL, Cui H. Supramolecular nano-
architecture based on pi-pi interlocked beta-sheets. Adv Mater 2008; structures formed by anticancer drug assembly. J Am Chem Soc
Acta Pharmacologica Sinica

www.chinaphar.com
Cheetham AG et al
11
2013; 135: 2907–10.
for nanoparticle systems in cancer therapeutics. Adv Drug Deliv Rev
46 Lin R, Cheetham AG, Zhang P, Lin YA, Cui H. Supramolecular filaments
2008; 60: 1615–26.
containing a fixed 41% paclitaxel loading. Chem Commun (Camb) 66 Peng MY, Qin SY, Jia HZ, Zheng DW, Rong L, Zhang XZ. Self-delivery of
2013; 49: 4968–70.
a peptide-based prodrug for tumor-targeting therapy. Nano Res 2016;
47 Li XY, Yang CB, Zhang ZL, Wu ZD, Deng Y, Liang GL, et al. Folic acid
9: 663–73.
as a versatile motif to construct molecular hydrogelators through 67 Knop K, Hoogenboom R, Fischer D, Schubert US. Poly(ethylene
conjugations with hydrophobic therapeutic agents. J Mater Chem
glycol) in drug delivery: pros and cons as well as potential alternatives.
2012; 22: 21838–40.
Angew Chem Int Ed 2010; 49: 6288–308.
48 Wang HM, Wei J, Yang CB, Zhao HY, Li DX, Yin ZN, et al. The inhibition 68 Sosnik A. Drug self-assembly: A phenomenon at the nanometer scale
of tumor growth and metastasis by self-assembled nanofibers of taxol.
Biomaterials 2012; 33: 5848–53.
with major impact in the structure-biological properties relationship
and the treatment of disease. Prog Mater Sci 2016; 82: 39–82.
49 McDaniel JR, Bhattacharyya J, Vargo KB, Hassouneh W, Hammer DA, 69 MacKay JA, Chen M, McDaniel JR, Liu W, Simnick AJ, Chilkoti A. Self-
Chilkoti A. Self-assembly of thermally responsive nanoparticles of
a genetically encoded peptide polymer by drug conjugation. Angew
Chem Int Ed 2013; 52: 1683–7.
assembling chimeric polypeptide-doxorubicin conjugate nanoparticles
that abolish tumours after a single injection. Nat Mater 2009; 8:
993–9.
50 Lock LL, Reyes CD, Zhang PC, Cui HG. Tuning cellular uptake of 70 Dubowchik GM, Firestone RA. Cathepsin B-sensitive dipeptide
molecular probes by rational design of their assembly into supramo-
lecular nanoprobes. J Am Chem Soc 2016; 138: 3533–40.
prodrugs. 1. A model study of structural requirements for efficient
release of doxorubicin. Bioorg Med Chem Lett 1998; 8: 3341–6.
51 Lin R, Cui HG. Supramolecular nanostructures as drug carriers. Curr 71 Dubowchik GM, Mosure K, Knipe JO, Firestone RA. Cathepsin
Opin Chem Eng 2015; 7: 75–83.
52 Hu XY, Liu X, Zhang WY, Qin S, Yao CH, Li Y, et al. Controllable con-
struction of biocompatible supramolecular micelles and vesicles by
B-sensitive dipeptide prodrugs. 2. Models of anticancer drugs
paclitaxel (Ttaxol (R)), mitomycin C and doxorubicin. Bioorg Med
Chem Lett 1998; 8: 3347–52.
water-soluble phosphate pillar 5,6 arenes for selective anti-cancer 72 Cheng R, Feng F, Meng F, Deng C, Feijen J, Zhong Z. Glutathione-
drug delivery. Chem Mater 2016; 28: 3778–88.
responsive nano-vehicles as a promising platform for targeted intra-
53 Tan XY, Lu XG, Jia F, Liu XF, Sun YH, Logan JK, et al. Blurring the role
cellular drug and gene delivery. J Control Release 2011; 152: 2–12.
of oligonucleotides: spherical nucleic acids as a drug delivery vehicle. 73 Zhong T, Huang R, Tan LJ. Amphiphilic drug-drug assembly via dual-
J Am Chem Soc 2016; 138: 10834–7.
54 Sun Y, Kaplan JA, Shieh A, Sun HL, Croce CM, Grinstaff MW, et al.
responsive linkages for small-molecule anticancer drug delivery. RSC
Advances 2016; 6: 66420–30.
Self-assembly of a 5-fluorouracil-dipeptide hydrogel. Chem Commun 74 Andreu D, Albericio F, Sole NA, Munson MC, Ferrer M, Barany G.
2016; 52: 5254–7.
55 Sun Y, Shieh A, Kim SH, King S, Kim A, Sun HL, et al. The self-
Formation of disulfide bonds in synthetic peptides and proteins. Meth
Mol Biol 1994; 35: 91–169.
assembly of a camptothecin-lysine nanotube. Bioorg Med Chem Lett 75 Hoyle CE, Bowman CN. Thiol-ene click chemistry. Angew Chem Int Ed
2016; 26: 2834–8. 2010; 49: 1540–73.
56 Zhou Z, Munyaradzi O, Xia X, Green D, Bong D. High-capacity drug 76 Lutz JF, Zarafshani Z. Efficient construction of therapeutics, bio-
carriers from common polymer amphiphiles. Biomacromolecules
conjugates, biomaterials and bioactive surfaces using azide-alkyne
2016; 17: 3060–6.
“click” chemistry. Adv Drug Deliv Rev 2008; 60: 958–70.
57 Pommier Y. Topoisomerase I inhibitors: camptothecins and beyond. 77 Goux WJ, Kopplin L, Nguyen AD, Leak K, Rutkofsky M, Shanmuganan-
Nat Rev Cancer 2006; 6: 789–802.
dam VD, et al. The formation of straight and twisted filaments from
58 Cheetham AG, Ou YC, Zhang PC, Cui HG. Linker-determined drug
short tau peptides. J Biol Chem 2004; 279: 26868–75.
release mechanism of free camptothecin from self-assembling drug 78 Hu Y, Lin R, Patel K, Cheetham AG, Kan CY, Cui HG. Spatiotemporal
amphiphiles. Chem Commun 2014; 50: 6039–42.
59 Kang M, Zhang PC, Cui HG, Loverde SM. pi-pi Stacking mediated
control of the creation and immolation of peptide assemblies. Coord
Chem Rev 2016; 320: 2–17.
chirality in functional supramolecular filaments. Macromolecules 79 Cheetham AG, Zhang P, Lin YA, Lin R, Cui H. Synthesis and self-
2016; 49: 994–1001.
assembly of a mikto-arm star dual drug amphiphile containing both
60 Cui HG, Webber MJ, Stupp SI. Self-assembly of peptide amphiphiles:
paclitaxel and camptothecin. J Mater Chem B 2014; 2: 7316–26.
from molecules to nanostructures to biomaterials. Biopolymers 2010; 80 Lin YA, Cheetham AG, Zhang PC, Ou YC, Li YG, Liu GS, et al. Multi-
94: 1–18.
walled nanotubes formed by catanionic mixtures of drug amphiphiles.
ACS Nano 2014; 8: 12690–700.
81 Du XW, Zhou J, Shi JF, Xu B. Supramolecular hydrogelators and
hydrogels: from soft matter to molecular biomaterials. Chem Rev
2015; 115: 13165–307.
61 Hauser CAE, Zhang S. Designer self-assembling peptide nanofiber
biological materials. Chem Soc Rev 2010; 39: 2780–90.
62 Kimura S, Kim DH, Sugiyama J, Imanishi Y. Vesicular self-assembly of
a helical peptide in water. Langmuir 1999; 15: 4461–3.
63 Santoso S, Hwang W, Hartman H, Zhang SG. Self-assembly of 82 Luo Y, Li M, Zhao YW, Liu H, Gong TX, Hong Y, et al. Molecular
surfactant-like peptides with variable glycine tails to form nanotubes
nanofibers of paclitaxel form supramolecular hydrogel for preventing
and nanovesicles. Nano Lett 2002; 2: 687–91.
tumor growth in vivo. RSC Advances 2016; 6: 80847–50.
64 Jaracz S, Chen J, Kuznetsova LV, Ojima L. Recent advances in tumor- 83 Peltier R, Chen GC, Lei HP, Zhang M, Gao LQ, Lee SS, et al. The
targeting anticancer drug conjugates. Bioorg Med Chem 2005; 13:
5043–54.
rational design of a peptide-based hydrogel responsive to H2S. Chem
Commun 2015; 51: 17273–6.
65 Byrne JD, Betancourt T, Brannon-Peppas L. Active targeting schemes
Acta Pharmacologica Sinica