Nanoscale

PAPER

Cite this: DOI: 10.1039/c7nr00172j

Received 7th January 2017,Accepted 3rd April 2017

DOI: 10.1039/c7nr00172j

rsc.li/nanoscale

Self-assembled hybrid elastin-like polypeptide/silica nanoparticles enable triggered drug release†

Wei Han, a,b Ashutosh Chilkotia,b,c and Gabriel P. López*a,b,c,d

The discovery of biomimetic polypeptides that enable the biomineralization of synthetic and biosynthetic

materials has resulted in the development of hybrid materials that incorporate inorganic components for

potential application in drug delivery, enzyme immobilization, and surface modification. Here, we describe

an approach that uses micellar assemblies of an elastin-like polypeptide (ELP) modified with silica-

promoting sequences and drug conjugates that are subsequently encapsulated within a silica matrix.

Incorporation of a lysine-rich tag derived from the silaffin R5 peptide into the N-terminus of a hydrophilic

ELP that self-assembles upon conjugation of hydrophobic molecules at the C-terminus results in the

formation of spherical micelles with a conjugated drug embedded in the core and a corona that is deco-

rated with the silaffin peptide. These micelles serve as the building blocks for the polycondensation of

silica into uniform, hybrid polypeptide–silica nanoparticles. We demonstrate proof-of-concept examples

using a model hydrophobic small molecule and doxorobucin, a small molecule chemotherapeutic, and

further show pH-dependent doxorubicin release from the hybrid nanoparticles.

Introduction

Material systems comprising silica components are widelyused in biotechnology and medicine due to their numerousattractive features – silica affords a range of comparativelysimple, yet flexible modification chemistries using commer-cially available reagents, and silica based materials exhibitrobust chemical and mechanical stability, and importantly,can be non-toxic and biocompatible.1–4 Consequently, the useof silica-based materials has grown significantly in thepast few decades for biosensing,5–7 drug delivery3,8,9 andbiomaterials.1,10–12 Notably, templated formation of meso-porous silica nanoparticles has drawn substantial attention foreffective, triggered delivery of various therapeutics13–16 withthe advantages of high and multi-drug loading capacity as wellas release rate tunability; however, these nanomaterials oftenrequire harsh reaction conditions and multi-step processing tofabricate,17,18 that may limit the types of molecules that can beencapsulated. These limitations can be alleviated by taking

cues from nature, which provides the motivation for thispaper.

Biomimetic and biologically-inspired strategies for theformation of silica stem from the observation that naturalorganisms are able to direct the formation of biosilica withinmild reaction environments rapidly and with intricate pre-cision over micro- and nano-length scales.19–21 To date, numer-ous proteins21–26 have been isolated and characterized asimportant players in nature’s exquisite silicification capabili-ties, i.e. the ability to drive the deposition of silica onto anorganic template. In addition, synthetic polymers,27–31 posses-sing similar chemical and structural features as their bio-logical counterparts, have also contributed to the silicificationtoolbox. One of the most widely studied natural siliceous com-ponents is the silaffin family of polypeptides, which was firstdiscovered by Kröger et al.21 within the 813 bp sil1 genesequence of the sea diatom species, Cylindrotheca fusiformis.Upon further analysis22,23 specific post-translational modi-fications were identified—phosphorylation of serine residuesand long-chain polyamine modification of the various lysineresidues—that play a vital role in the ability of silaffins todirect the deposition of silica. Interestingly, however, thesilaffin R5 peptide without any post-translational modifi-cations also facilitates silicification in the presence of phos-phate anions.21 It has been hypothesized that silicificationoccurs by transient self-assembly of individual silaffin peptidesthrough the charge neutralization of lysines by phosphategroups, which provides the high local cationic environmentnecessary for facile silicification.17,23,32 Silaffin R5 has since

†Electronic supplementary information (ESI) available: Experimental materials,additional methods, and supplementary figures. See DOI: 10.1039/c7nr00172j

aResearch Triangle Materials Science and Engineering Center, Durham,

North Carolina 27708, USA. E-mail: gplopez@unm.edubDepartment of Biomedical Engineering, Duke University, Durham,

North Carolina 27708, USAcDepartment of Mechanical Engineering & Materials Science, Duke University,

Durham, North Carolina 27708, USAdDepartment of Chemical & Biological Engineering, the University of New Mexico,

Albuquerque, New Mexico 87131, USA

This journal is © The Royal Society of Chemistry 2017 Nanoscale

Publ

ishe

d on

07

Apr

il 20

17. D

ownl

oade

d by

Uni

vers

ity o

f C

alif

orni

a -

Irvi

ne o

n 01

/05/

2017

23:

19:1

3.

View Article OnlineView Journal

been explored for biotechnological applications such asenzyme immobilization,33–35 cell encapsulation,36 and nano-material patterning.37,38 Although silaffin R5 has previouslybeen used in the encapsulation of drugs in particles,39 pre-vious demonstrations showed relatively imprecise control ofnanoparticle size and the formation of networks of particles(presumably due to the uncontrolled crosslinking or aggre-gation of templates), which may limit the feasibility for drugdelivery applications.

In a previous study,40 our group demonstrated that theincorporation of a silaffin R5 sequence into a diblock elastin-like-polypeptide (ELP), which self-assembles into sphericalmicelles displaying the R5 peptides in the micellar coronae,provides the necessary reaction environment to template theformation of near-monodisperse silica nanoparticles. We alsoshowed that silicification is dependent on both the presenceof silaffin R5 as well as the assembly of individual ELP mole-cules (i.e., unimers) into micelles, and envisioned potentialapplications such as drug release for the system, enabled bythe facile design and expression of desired ELP constructs.ELPs are polypeptides with the repeat pentapeptide motifVPGXG, where X may be any amino acid except proline, andexhibit a lower critical transition temperature (LCST) in waterthat is dependent on a variety of factors, including intrinsicparameters like polypeptide composition of the guest residueX, molecular weight and concentration, as well as extrinsicparameters like the type and concentration of salts.41–44 ELPsare an attractive material system because they can be encodedat the genetic level to provide complete control of the aminoacid sequence and chain length and can be expressed at ahigh level in Escherichia coli. Exploitation of the thermallyresponsive phase behavior of ELPs also allows facile and rapidpurification of both the polypeptide as well as its peptideor protein fusions following expression from E. coli.45,46

Furthermore, ELPs offer an immense range of possible designsat the gene level; they can be genetically fused to peptide andproteins,47,48 intricate block architectures can bedesigned40,46,49 and ELPs can be designed with preciselylocated conjugation sites for the post-expression conjugationof non-genetically encodable moieties that can drive their self-assembly.46,50–52 Within the research space of drug delivery,ELPs have been used as a carrier for small molecule drugs andpeptides to improve their pharmacokinetics and tissueretention.50–53

The application of silica and ELPs together to create a nano-particle drug release platform has attractive and synergisticattributes. As we previously described, the ability to geneticallydesign exact polypeptide constructs that enable drug conju-gation and incorporate silica-promoting motifs to direct andtemplate the deposition of silica under mild conditions isdesirable. On the other hand, the encapsulation of a poly-peptide drug carrier system in silica also imparts benefits, asdemonstrated by the abundance of silica and silica-encapsulat-ing delivery systems that have been widely studied.1–3,5,54 Forinstance, amorphous silica is non-toxic and elicits no immuno-genic response;3 in addition, the relatively inert nature of silica

under physiological conditions allows the encapsulated drugmaterial to remain protected for sustained periods of time.Lastly, the slow hydrolysis of silica enables release of theentrapped material and produces innocuous by-products afterdegradation.3 The use of silica to encapsulate a polypeptidedrug carrier may have a positive impact on the pharmaco-kinetic characteristics of the overall system. Specifically, theuse of silica to encapsulate a polypeptide micelle may have apositive impact on in vivo pharmacokinetics, as uncrosslinkedELP micelles have been shown to have a terminal half-life of∼12 h,40,52,55–57 presumably because upon dilution below theircritical micelle concentration in systemic circulation, themicelles dissociate into their constituent ELP unimers, leadingto clearance through the renal filtration system. On the otherhand, we hypothesize that silica stabilized micelles shouldremain in circulation longer as they are covalently stabilizednanoparticles and should hence evade clearance through thekidneys.

Herein, we report the design and characterization of silica-promoting ELPs that self-assemble into micelles throughconjugation with one of two hydrophobic molecules:N-benzylmaleimide or the chemotherapeutic, doxorubicin. Wedemonstrate that both designs are able to similarly templatethe formation of silica encapsulated nanoparticles and that itis possible to program the release of drugs from the particlesin response to a shift in pH. Furthermore, we hypothesize thatthese findings, in conjunction with the precise design poten-tial of genetically encoded ELPs, may allow for the fabricationof a hybrid ELP/silica nanoparticle platform for targeted deliv-ery of drugs.

ExperimentalMaterials

Restriction enzymes BglI, NdeI, BseRI, AcuI, BamHI and XbaI,T4 DNA ligase and calf intestinal phosphatase (CIP) were pur-chased from New England Biolabs (Ipswich, MA). BL21 E. colicompetent cells were purchased from EdgeBio (Gaithersburg,MD). Oligonucleotides were obtained from Integrated DNATechnologies, Inc. (Coralville, IA). Terrific Broth was purchasedfrom MO BIO Labs, Inc. (Carlsbad, CA) and SDS-PAGE gelsfrom BioRad, Inc. (Hercules, CA). Miniprep and Gel ExtractionKits were purchased from Qiagen (Venlo, Netherlands).Tetramethylorthosilicate (TMOS) was obtained from Alfa Aesar(Ward Hill, MA). Formvar-copper 200 mesh (FF200-Cu) gridsfor transmission electron microscopy were purchased fromElectron Microscopy Sciences (Hatfield, PA). Oregon green car-boxylic acid, succinimidyl ester, wheat germ agglutinin AlexaFluor 594 conjugate, and Image-iT Fixation/PermeabilizationKit were purchased from Thermo Fisher Scientific (Waltham,MA). HT-29 cells, McCoy’s 5a Growth Medium, Dulbecco’sphosphate buffered saline (DPBS), Hank’s balanced salt solu-tion (HBSS) and fetal bovine serum (FBS) were purchased fromATCC (Manassas, VA).

Paper Nanoscale

Nanoscale This journal is © The Royal Society of Chemistry 2017

Publ

ishe

d on

07

Apr

il 20

17. D

ownl

oade

d by

Uni

vers

ity o

f C

alif

orni

a -

Irvi

ne o

n 01

/05/

2017

23:

19:1

3.

View Article Online

Construction of sp-ELP and con-ELP genes

Genes encoding silica-promoting ELP (sp-ELP) and controlELP (con-ELP) sequences were constructed by recursivedirectional ligation by plasmid reconstruction (PRe-RDL)58

of an ELP gene that has a C-terminal cysteine-rich trailersequence followed by either a silica-promoting sequence(sp-ELP) or a control sequence (con-ELP) at the N-terminus,as illustrated in Fig. 1A. The ELP gene, which encodes a40 pentapeptide VPGAG repeat in a pET-24a(+) vector(plasmid A), and another pET-24a(+) vector that contains aDNA sequence that encodes a (CGG)8 sequence (plasmid B)were obtained from a previous study.59 The plasmid contain-ing the ELP sequence (plasmid A) was digested withAcuI and BglI restriction nucleases, the plasmid containingthe trailer sequence (plasmid B) was digested withBseRI and BglI, and both segments were gel isolated andpurified using a Qiagen Gel Extraction kit. Subsequentlyboth DNA fragments were ligated together using T4 DNAligase and transformed into BL21(DE3) cells for E. coliexpression. To insert the silica-promoting or control nucleo-tide sequences (Fig. S1†) at the 5′-end of the ELP gene, theA + B plasmid product of PRe-RDL was digested with NdeI andBseRI, dephosphorylated using calf intestinal phosphatase(CIP), and purified using a Qiagen Purification Kit.Reconstituted and annealed oligos containing either a silica-promoting tag or a control tag were incubated with thedigested vector and ligated using T4 DNA ligase, transformedinto chemically competent E. coli and expressed as previouslydescribed.47

Conjugation of N-benzylmaleimide and BMPH-activateddoxorubicin

Conjugation of N-benzylmaleimide (BM) and N-β-maleimido-propionic acid hydrazide (BMPH) activated doxorubicin(Dox) was carried out as previously described.52,60 Briefly,10 mg of lyophilized sp-ELP was resuspended in 900 µLof 10 mM phosphate buffer supplemented with 20 mMtris(2-carboxyethyl)phosphine hydrochloride (TCEP), pH 7.50 mM N-benzylmaleimide in DMSO was added to reach1 mL total reaction volume and allowed to react for 3 h. Theproduct was then purified by using a PD10 column as well asdialyzed two times for 4 h each using dialysis cassettes. Thepurified samples were then lyophilized for storage and bufferexchange.

For conjugation with BMPH-activated doxorubicin, BMPHis first reacted with doxorubicin in which the hydrazide moietyof BMPH reacts with the carbonyl moiety of the doxorubicin toform a hydrazone bond, and subsequent linkage of maleimideto doxorobucin occurs. 50 mg of BMPH was dissolved in10 mL anhydrous methanol in a glass tube and 110 mg doxo-rubicin was dissolved in anhydrous methanol in a roundbottom flask, supplemented with 100 µL trifluoroacetic acid.The BMPH solution was subsequently added to the doxorobu-cin solution while stirring and allowed to react overnight inthe dark at room temperature. The methanol was thenremoved by rotary evaporation and replaced with 10 mL freshanhydrous methanol. 250 mg lyophilized sp-ELP was resus-pended in 2 mL water and supplemented with TCEP to a finalconcentration of 30 mM, pH 7, and incubated for 15 min to

Fig. 1 Schematic representation of ELP sequence, conjugation and self-assembly of hybrid silica–ELP nanoparticles. (A) Representation of silica-promoting ELP sequence (sp-ELP) with an N-terminal silaffin R5 peptide leader (green) appended to a hydrophilic ELP block (blue) and a cysteine-rich C-terminal trailer for conjugation (black). The control ELP (con-ELP) contains the ELP and cysteine-rich trailer and lacks the silica-promoting R5leader. (B) Hydrophobic molecules (red stars) lead to the self-assembly of ELP micelles. Hydrophobes: N-benzylmaleimide (1) andN-β-maleimidopropionic acid hydrazide activated doxorubicin (2). (C) Upon hydrophobe conjugation, ELP unimers self-assemble in water intospherical micelles and serve as templates for the polycondensation of silica from hydrolyzed TMOS.

Nanoscale Paper

This journal is © The Royal Society of Chemistry 2017 Nanoscale

Publ

ishe

d on

07

Apr

il 20

17. D

ownl

oade

d by

Uni

vers

ity o

f C

alif

orni

a -

Irvi

ne o

n 01

/05/

2017

23:

19:1

3.

View Article Online

break disulfide bonds. 5 M NaCl was then added to phase sep-arate the ELP, which was further separated via ultracentrifuga-tion at 14 000g for 10 minutes, and resuspended in 5 mL0.1 M/1 mM phosphate/EDTA buffer, pH 7. The BMPH-activated doxorubicin solution was then added dropwise to theELP solution while stirring and reacted overnight in the darkat room temperature. sp-ELP-Dox was purified using AmiconUltra centrifugal filters (EMD Millipore; Billerica, MA) with a10 kDa molecular weight cutoff.

Silica encapsulation and characterization of sp-ELP-BM andsp-ELP-Dox micelles

Lyophilized sp-ELP-BM or sp-ELP-Dox micelles were first resus-pended in 900 µL pure water to a concentration of 2.5mg mL−1 in a microcentrifuge tube. To prepare hydrolyzedTMOS, 1 M TMOS was incubated in 1 mM HCl and allowed tomix via rotation for 15 minutes at room temperature.Subsequently, 100 µL hydrolyzed TMOS was combined with theELP micelle solutions under 200 rpm rotation for a set period oftime, centrifuged at 15 000g and resuspended in water.

Dynamic light scattering (DLS) was carried out on aDynaPro Plate Reader II (Wyatt Technology, Santa Barbara,CA). Measurements of hydrodynamic radii (Rh) were made intriplicate, with each measurement consisting of 10 acqui-sitions of 5 s each. The autocorrelation function was fitted by aregularization fit provided by the manufacturer to determinethe Rh. Zeta potential was measured on a Zetasizer Nano ZS(Malvern Instruments; Malvern, UK) with 90° scattering optics.900 µL of an ELP sample was placed into a mini-quartzcuvette, and measured three times for each condition. Threemeasurements were then averaged to create a single datapoint. Transmission electron microscopy (TEM) was performedon a Tecnai G2 Twin (FEI Company, Hillsboro, OR) at 200 keV.

Fluorescence microscopy of the cellular uptake of freedoxorubicin, ELP micelles and ELP/silica nanoparticles

HT-29 colorectal adenocarcinoma cells (ATCC; see the ESI† forsubculture methodology) were seeded at 1 × 105 cells per cm−2

on 35 mm culture dishes with a No. 0 coverslip window attach-ment (MatTek Corporation; Ashland, MA). Cells were incu-bated with control doxorobucin, sp-ELP-Dox micelle and sp-ELP-Dox silica nanoparticle treatment, washed with Hank’sbalanced salt solution (HBSS) twice, fixed using the fixativesolution of an Image-iT® Fixation/Permeabilization Kit,stained with a wheat germ agglutinin Alexa Fluor 594 conju-gate, and finally washed four times further. Imaging was con-ducted using a Leica SP5 inverted confocal microscope using a63×/1.20 water objective and the data were analyzed usingLeica LAS AF software.

Results and discussion

Our modular design of amphiphilic, silica-promoting ELPsbegins at the amino acid sequence level with three distinctpolypeptide subunits: an N-terminal silaffin R5 peptide motif,

a hydrophilic ELP sequence, and a C-terminal trailer contain-ing eight cysteine residues for conjugation (Fig. 1A). We uti-lized the thiol-rich trailer region of the polypeptides forcovalent conjugation of hydrophobic molecules that have beenpreviously demonstrated to result in sufficient amphiphilicityto drive the self-assembly of the conjugation product into amicellar phase.52,60 For this study, we utilized two hydrophobicmolecules: (1) N-benzylmaleimide, a model small hydrophobicmolecule used for proof-of concept purposes, and (2) doxo-rubicin, a chemotherapeutic drug (Fig. 1B). Upon micellizationof ELP unimers post-conjugation, we hypothesized that theabundance of positively-charged residues in the coronae of themicelles due to the silica-promoting R5 sequence would leadto the preferential polycondensation of silica in the presenceof silica-forming monomers (e.g., hydrolyzed tetramethylorthosilicate, TMOS) to form uniform nanoparticles encapsu-lating the micellar templates and drugs (Fig. 1C).

First, we recombinantly expressed, purified, and character-ized the silica-promoting ELP (sp-ELP) construct, as well as acontrol ELP (con-ELP) construct that is identical to sp-ELP, butwhich lacks the R5 silica-condensation-promoting motif. Weperformed SDS-PAGE on purified samples of sp-ELP and con-ELP and determined that the observed bands matched thetheoretical molecular weights of 18.7 kDa and 16.8 kDa forsp-ELP and con-ELP, respectively. Furthermore, the purifi-cation by inverse transition cycling (ITC) of both constructsresulted in highly pure samples for subsequent analysis.

To study the silicification ability of sp-ELP, we first co-valently conjugated the polypeptide to N-benzylmaleimide(sp-ELP-BM) and characterized the resulting micellar materialsusing light scattering and TEM. DLS measurements ofsp-ELP-BM at 100 µM in water show a mean Rh of 19.2 ± 0.3nm, indicative of the formation of self-assembled micelles.Next, we carried out time-course silicification experimentswith hydrolyzed TMOS at 10 mM (triangles) and 100 mM ofTMOS (circles) over the course of 4 h (Fig. 2A). For both precur-sor concentrations, the observed size of the sp-ELP-BM par-ticles increased, and in particular for the 100 mM TMOSsamples, the Rh increased from ∼19 nm to ∼24 nm over thelength of the study, while the Rh of micelles with no addedTMOS (squares) showed no significant change in size. In con-trast, the control ELP (con-ELP-BM; Fig. S2† left) exhibited nosignificant difference in measured particle size over the sameperiod of time.

Subsequently, we imaged both the R5-modified and controlsamples after reaction with 100 mM TMOS for 4 h using TEM.TEM confirmed that the sp-ELP-BM construct (Fig. 2B) formsnear-monodisperse silica nanoparticles with a mean diameterof 42.1 ± 5.5 nm as calculated by image processing of the TEMimages, which is in agreement with DLS. In contrast, no silicaparticles were visible over the TEM grid (image not shown) forthe negative control, con-ELP-BM. Interestingly, the contrastvisible in the TEM images of the spherical sp-ELP-BM silicaparticles clearly suggests a core–shell structure comprising ashell of silica with high electron density that encapsulates alower electron density core, which presumably contains the

Paper Nanoscale

Nanoscale This journal is © The Royal Society of Chemistry 2017

Publ

ishe

d on

07

Apr

il 20

17. D

ownl

oade

d by

Uni

vers

ity o

f C

alif

orni

a -

Irvi

ne o

n 01

/05/

2017

23:

19:1

3.

View Article Online

ELP and encapsulated small molecules. This observation is incontrast to our previous study40 in which we utilized a self-assembled diblock ELP system and observed uniform electrondensity contrast along the entire radial axis of the nano-particles. This suggests that the core of sp-ELP-BM micellesmay be more hydrophobic than those studied previously, suchthat it precludes the penetration of silica condensation intothe micellar core, thereby leading to a core–shell morphology.

We next measured the zeta potential of the sp-ELP-BMmicelles prior to and post-silicification (as well as for the con-ELP-BM micelles) (Fig. 2C). While sp-ELP-BM micelles (dottedbar) exhibited a zeta potential of 8.7 ± 1.2 mV – likely resultingfrom the positively-charged lysine and arginine residues at thesurface of the micelles – post-silicification particles (filled bar)exhibited a zeta potential of −31.2 ± 4.6 mV, suggesting thepresence of negatively charged surface silanol groups, whichis in agreement with previous studies with templated nano-particles by silica deposition.31,40

Satisfied with the proof-of-concept demonstration usingN-benzylmaleimide, we segued to doxorubicin, a cancer drugthat is hence of greater utility. Because the drug must bereleased from the nanoparticle to be therapeutically effective,we relied upon a strategy that Chilkoti and coworkers have pre-viously implemented to release doxorubicin from ELP nano-particles.52,57 In this approach, the drug is linked to theELP through an acid-labile hydrazone bond, which is cleavedat pH < 6 in solution and releases the drug withinthe acidic environment of the late endosomes and lysosomesof cells.61–63 To insert the internal hydrazone bondbetween the drug and ELP, a heterobifunctional linker,N-β-maleimidopropionic acid hydrazide (BMPH) was first conju-gated to the 13′-keto position of doxorubicin, and the productof this reaction was then conjugated to the ELP via reaction ofthe maleimide group in the product with the cysteine residuesin the ELP; conjugation of the drug triggers self-assembly ofthe ELP into spherical micelles that are termed sp-ELP-Dox.MacKay et al. reported that the conjugation efficiency of doxo-rubicin to an ELP of similar design is 4.8 ± 1.3 drug moleculesper peptide, and based on the structural similarity of this trail-ing sequence – eight cysteines separated by glycine–glycinespacers – we believe that the conjugation efficiency of thematerial in this manuscript is similar. We performed a time-course silicification study for sp-ELP-Dox (Fig. 3A) in which weincubated solutions of the conjugate micelles at 100 µM inwater in the presence of 10 mM (triangles) and 100 mM(circles) hydrolyzed TMOS over the span of 4 h. DLS measure-ments showed that the Rh of the 10 mM experimental groupincreased from 21.2 ± 1.1 nm to 23.2 ± 1.1 nm and that of the100 mM TMOS group increased from 21.5 ± 0.8 nm to 25.3 ±1.9 nm after 4 h while the size of the control, sp-ELP-Dox incu-bated with no silica precursor did not change significantlywith time.

We next imaged sp-ELP-Dox silica nanoparticles obtainedafter 4 h incubation with 100 mM TMOS by TEM (Fig. 3B) andagain observed spherical, core–shell nanoparticles measuring47.8 ± 7.5 nm in diameter. Both of these findings are in agree-ment with the previous sp-ELP-BM construct data and furtherdemonstrate the silicification of the sp-ELP-Dox conjugate.Lastly, con-ELP-Dox micelles did not exhibit silicification(Fig. S2† right), providing further evidence that the formationof silica nanoparticles depends on the presence of the silica-promoting R5 peptide motif.

We next studied the pH-triggered release of doxorubicinfrom sp-ELP-Dox silica nanoparticles by measuring the charac-

Fig. 2 Characterization of the silicification of sp-ELPN-benzylmaleimide (sp-ELP-BM). (A) Time course measurements of Rh

of sp-ELP-BM with different concentrations of TMOS (0, 10, 100 mM).(B) Representative TEM image of silica nanoparticles templated from sp-ELP-BM with 100 mM hydrolyzed TMOS for 4 h; scale bar: 50 nm. (C)Zeta potential measurements of sp-ELP-BM micelles (dotted bar), con-ELP-BM micelles (diagonal striped bar) and sp-ELP-BM silicified nano-particles (solid bar).

Nanoscale Paper

This journal is © The Royal Society of Chemistry 2017 Nanoscale

Publ

ishe

d on

07

Apr

il 20

17. D

ownl

oade

d by

Uni

vers

ity o

f C

alif

orni

a -

Irvi

ne o

n 01

/05/

2017

23:

19:1

3.

View Article Online

teristic spectrophotometric absorbance of doxorubicin at480 nm and characterizing the amount of drug released as afraction of initial drug loading (Fig. 3C). We incubated micelleor silica nanoparticle solutions at either pH 7.4 or 5.0 for a setamount of time, adjusted the pH to 7.4 to stop further release,separated free Dox from encapsulated Dox using a microcentri-

fuge ultrafiltration unit, and measured absorbance of doxo-rubicin (at 480 nm) in the filtrate. Percent release fromsp-ELP-Dox silica (blue), calculated as a fraction of the initialfluorescence level, for samples at pH 7.4 (square markers) in-dicated that only ≈4% of doxorubicin is released after 36 h,while samples at pH 5.0 (circle markers) showed significantlygreater, 55%, drug released after 36 h incubation. The releasedata suggest that silica encapsulation of doxorubicin loadedELP micelles does not prevent the pH-dependent release of thedrug. In comparison, we also conducted the same releaseexperiment using only sp-ELP-Dox micelles (red) without silici-fication in order to characterize the release of Dox from only amicellar nanoparticle carrier. The sp-ELP-Dox micelles demon-strated a greater initial release rate, liberating ≈45% of Doxafter 8 h and 63% after 36 h; this is in general agreement withthe previous study for the micelle release for a ELP and doxo-rubicin conjugate, which exhibited a similar release profile.52

Interestingly, the silicified particles show a delay in the pH-mediated release of Dox, potentially due to the silica shell pre-venting the cleavage of the hydrazone bond or degradation ofthe polypeptide and silica particle. However, the total accumu-lated release after 36 hours suggests that both types of par-ticles are pH responsive.

To further explore the potential of sp-ELP-Dox silica nano-particles as a tool in the delivery of therapeutics, we investi-gated their uptake in cancer cells by confocal fluorescencemicroscopy (Fig. 4). We incubated HT-29 colorectal adeno-carcinoma cells with free doxorubicin (Fig. 4A–C), sp-ELP-Doxmicelles (Fig. 4D–F), and sp-ELP-Dox/silica nanoparticles(Fig. 4G–I) for 10 min, 2 h and 24 h. The cells primarily appearin the form of cellular aggregates, each containing a numberof individual cells that may be observed using the cell stain asa guide. Free doxorubicin is observed in cell nuclei as early as10 minutes (Fig. 4A), while minute fluorescence is observedwith the sp-ELP-Dox micelle sample, suggesting an initialrelease of Dox, and no fluorescence is detected in the silicaencapsulated sp-ELP-Dox sample, which is in agreement witha previous study.52 At 2 h, the fluorescence signal fromsp-ELP-Dox micelle samples is greater than that of thesp-ELP-Dox silica nanoparticles, and lastly at 24 h incubation,the doxorubicin fluorescence within cells incubated withsp-ELP-Dox silica nanoparticles intensifies moderately, indicat-ing further release and accumulation (Fig. 4I). The cell uptakedata are in general agreement with the in vitro release charac-terization, which suggests that doxorubicin remains conju-gated to ELP and encapsulated in silica until the low pHenvironment of cellular endosomes after uptake triggers thecleavage and subsequent release of the drug.64

After release of doxorubicin from the nanoparticles, we alsocharacterized the particles using dynamic light scattering tounderstand the effects on the nanoparticles. The DLS datashowed the presence of populations of multiple sizes, makinga reasonable fit of the data and drawing insight very difficult.To the best of our understanding, the cleavage of the doxo-rubicin from ELP under suitable pH conditions results in thedegradation of the nanoparticles.

Fig. 3 Characterization of silicification and release of sp-ELP doxo-rubicin (sp-ELP-Dox). (A) Time course measurements of Rh of sp-ELP-Dox with different concentrations of TMOS (0, 10, 100 mM). (B)Representative TEM image of silica nanoparticles templated from sp-ELP-Dox with 100 mM hydrolyzed TMOS for 4 h; scale bar: 50 nm. (C)Doxorubicin release as a function of time as measured by absorbancespectrophotometry for sp-ELP-Dox micelles without silica (red) and sp-ELP-Dox silica nanoparticles (blue) at pH 7.4 (squares) and pH 5.0(circles).

Paper Nanoscale

Nanoscale This journal is © The Royal Society of Chemistry 2017

Publ

ishe

d on

07

Apr

il 20

17. D

ownl

oade

d by

Uni

vers

ity o

f C

alif

orni

a -

Irvi

ne o

n 01

/05/

2017

23:

19:1

3.

View Article Online

Conclusion

We report the design, synthesis and application of ELP tem-plated silica nanoparticles towards the programmable releaseof the chemotherapeutic, doxorubicin. By incorporating asilica-promoting motif derived from the synthetic silaffin R5peptide into a modular ELP construct, we demonstrate thathydrophobe-conjugated sp-ELP micelles can drive the polycon-densation of silica-precursors to form uniform, hybrid core–shell nanoparticles. We show that the addition of a stabilizinglayer of silica does not prevent the pH-dependent release ofthe encapsulated drug and that this material system may beutilized in future in vivo drug delivery studies. In addition, theprecise control of polypeptide sequences and the broad avail-ability of additional means to modify the peptides or silicacomponents of the nanoparticles may enable the in-corporation of moieties for targeted delivery applications.Furthermore, the discovery and integration of additional bio-mineralization tags into the design of self-assembled polypep-tide micelles may result in more complex hybrid biomaterialswith valuable biomedical applications such as imaging andtheragnostics.

Acknowledgements

This work is supported by NSF’s Research Triangle MRSEC(DMR-1121107). W. H. acknowledges support from the NIHBiotechnology Predoctoral Training Fellowship (NIH Grant No.5T32GM008555-18). The authors would like to especially thankParisa Yousefpour of the Chilkoti Lab for providing advice andtechnical assistance in the conjugation of Dox and thank thestaff of the Duke University Light Microscopy Core Facility(LMCF) for training and support with confocal fluorescencemicroscopy.

References

1 S. V. Patwardhan, Chem. Commun., 2011, 47, 7567–7582.2 D. Avnir, T. Coradin, O. Lev and J. Livage, J. Mater. Chem.,

2006, 16, 1013–1030.3 C. Barbe, J. Bartlett, L. Kong, K. Finnie, H. Q. Lin,

M. Larkin, S. Calleja, A. Bush and G. Calleja, Adv. Mater.,2004, 16, 1959–1966.

Fig. 4 Cellular uptake of doxorubicin-encapsulated ELP–silica nanoparticles by HT-29 cells. Confocal fluorescence microscopy of free doxorubicin(A–C), sp-ELP-Dox micelles without silica (D–F), and sp-ELP-Dox silica particles (G–I) at 10 min (A, D, G), 2 hours (B, E, H) and 24 hours (C, F, I).Colors: doxorubicin (red) and cell stain (blue); scale bar: 10 µm.

Nanoscale Paper

This journal is © The Royal Society of Chemistry 2017 Nanoscale

Publ

ishe

d on

07

Apr

il 20

17. D

ownl

oade

d by

Uni

vers

ity o

f C

alif

orni

a -

Irvi

ne o

n 01

/05/

2017

23:

19:1

3.

View Article Online

4 P. Yang, S. Gai and J. Lin, Chem. Soc. Rev., 2012, 41, 3679–3698.

5 I. I. Slowing, B. G. Trewyn, S. Giri and V. Y. Lin, Adv. Funct.Mater., 2007, 17, 1225–1236.

6 C. Wang, Z. Ma, T. Wang and Z. Su, Adv. Funct. Mater.,2006, 16, 1673–1678.

7 Y. Wu, C. Chen and S. Liu, Anal. Chem., 2009, 81, 1600–1607.

8 Z. Li, J. C. Barnes, A. Bosoy, J. F. Stoddart and J. I. Zink,Chem. Soc. Rev., 2012, 41, 2590–2605.

9 Q. He and J. Shi, Adv. Mater., 2014, 26, 391–411.10 S. V. Patwardhan, S. J. Clarson and C. C. Perry, Chem.

Commun., 2005, 1113–1121.11 D. K. Yi, S. T. Selvan, S. S. Lee, G. C. Papaefthymiou,

D. Kundaliya and J. Y. Ying, J. Am. Chem. Soc., 2005, 127,4990–4991.

12 L. Betancor and H. R. Luckarift, Trends Biotechnol., 2008,26, 566–572.

13 I. I. Slowing, J. L. Vivero-Escoto, C.-W. Wu and V. S. Y. Lin,Adv. Drug Delivery Rev., 2008, 60, 1278–1288.

14 H. Meng, W. X. Mai, H. Zhang, M. Xue, T. Xia, S. Lin,X. Wang, Y. Zhao, Z. Ji, J. I. Zink and A. E. Nel, ACS Nano,2013, 7, 994–1005.

15 Y. Zhao, B. G. Trewyn, I. I. Slowing and V. S.-Y. Lin, J. Am.Chem. Soc., 2009, 131, 8398–8400.

16 J. L. Vivero-Escoto, I. I. Slowing, C.-W. Wu and V. S.-Y. Lin,J. Am. Chem. Soc., 2009, 131, 3462–3463.

17 C. C. Lechner and C. F. Becker, Marine Drugs, 2015, 13,5297–5333.

18 C. C. Lechner and C. F. Becker, Bioorg. Med. Chem., 2013,21, 3533–3541.

19 D. Belton, G. Paine, S. V. Patwardhan and C. C. Perry,J. Mater. Chem., 2004, 14, 2231–2241.

20 M. B. Dickerson, K. H. Sandhage and R. R. Naik, Chem.Rev., 2008, 108, 4935–4978.

21 N. Kröger, R. Deutzmann and M. Sumper, Science, 1999,286, 1129–1132.

22 N. Kröger, R. Deutzmann, C. Bergsdorf and M. Sumper,Proceedings of the National Academy of Sciences, 2000, 97,14133–14138.

23 N. Kroger, S. Lorenz, E. Brunner and M. Sumper, Science,2002, 298, 584–586.

24 S. Wenzl, R. Hett, P. Richthammer and M. Sumper, Angew.Chem., Int. Ed., 2008, 120, 1753–1756.

25 J. N. Cha, K. Shimizu, Y. Zhou, S. C. Christiansen,B. F. Chmelka, G. D. Stucky and D. E. Morse, Proc. Natl.Acad. Sci. U. S. A., 1999, 96, 361–365.

26 K. Shimizu, J. Cha, G. D. Stucky and D. E. Morse, Proc.Natl. Acad. Sci. U. S. A., 1998, 95, 6234–6238.

27 J.-S. Jan, S. Lee, C. S. Carr and D. F. Shantz, Chem. Mater.,2005, 17, 4310–4317.

28 J.-S. Jan and D. F. Shantz, Adv. Mater., 2007, 19, 2951–2956.

29 Y. Li, J. Du and S. P. Armes, Macromol. Rapid Commun.,2009, 30, 464–468.

30 J. J. Yuan and R. H. Jin, Adv. Mater., 2005, 17, 885–888.

31 J.-J. Yuan, O. O. Mykhaylyk, A. J. Ryan and S. P. Armes,J. Am. Chem. Soc., 2007, 129, 1717–1723.

32 C. C. Lechner and C. F. W. Becker, J. Pept. Sci., 2014, 20,152–158.

33 H. R. Luckarift, J. C. Spain, R. R. Naik and M. O. Stone,Nat. Biotechnol., 2004, 22, 211–213.

34 W. D. Marner, A. S. Shaikh, S. J. Muller and J. D. Keasling,Biotechnol. Prog., 2009, 25, 417–423.

35 R. R. Naik, M. M. Tomczak, H. R. Luckarift, J. C. Spain andM. O. Stone, Chem. Commun., 2004, 1684–1685.

36 S. H. Yang, K.-B. Lee, B. Kong, J.-H. Kim, H.-S. Kim andI. S. Choi, Angew. Chem., Int. Ed., 2009, 48, 9160–9163.

37 C. Wong Po Foo, S. V. Patwardhan, D. J. Belton, B. Kitchel,D. Anastasiades, J. Huang, R. R. Naik, C. C. Perry andD. L. Kaplan, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 9428–9433.

38 L. L. Brott, R. R. Naik, D. J. Pikas, S. M. Kirkpatrick,D. W. Tomlin, P. W. Whitlock, S. J. Clarson andM. O. Stone, Nature, 2001, 413, 291–293.

39 C. C. Lechner and C. F. W. Becker, Bioorg. Med. Chem.,2013, 21, 3533–3541.

40 J. Bhattacharyya, J. J. Bellucci, I. Weitzhandler,J. R. McDaniel, I. Spasojevic, X. Li, C.-C. Lin, J.-T. A. Chiand A. Chilkoti, Nat. Commun., 2015, 6, 7939.

41 D. W. Urry, C. H. Luan, T. M. Parker, D. C. Gowda,K. U. Prasad, M. C. Reid and A. Safavy, J. Am. Chem. Soc.,1991, 113, 4346–4348.

42 D. E. Meyer and A. Chilkoti, Biomacromolecules, 2004, 5,846–851.

43 Y. Cho, Y. Zhang, T. Christensen, L. B. Sagle, A. Chilkotiand P. S. Cremer, J. Phys. Chem. B, 2008, 112, 13765–13771.

44 J. R. McDaniel, D. C. Radford and A. Chilkoti,Biomacromolecules, 2013, 14, 2866–2872.

45 W. Hassouneh, T. Christensen and A. Chilkoti, Curr. Protoc.Protein Sci., 2010, Ch. 6: Unit-6.11.

46 S. R. MacEwan and A. Chilkoti, Biopolymers, 2010, 94, 60–77.47 K. Trabbic-Carlson, L. Liu, B. Kim and A. Chilkoti, Protein

Sci., 2004, 13, 3274–3284.48 M. Amiram, K. Luginbuhl, X. Li, M. N. Feinglos and

A. Chilkoti, J. Controlled Release, 2013, 172, 144–151.49 W. Hassouneh, K. Fischer, S. R. MacEwan, R. Branscheid,

C. L. Fu, R. Liu, M. Schmidt and A. Chilkoti,Biomacromolecules, 2012, 13, 1598–1605.

50 J. R. McDaniel, D. J. Callahan and A. Chilkoti, Adv. DrugDelivery Rev., 2010, 62, 1456–1467.

51 J. R. McDaniel, S. R. MacEwan, M. Dewhirst andA. Chilkoti, J. Controlled Release, 2012, 159, 362–367.

52 J. A. MacKay, M. Chen, J. R. McDaniel, W. Liu, A. J. Simnickand A. Chilkoti, Nat. Mater., 2009, 8, 993–999.

53 S. R. MacEwan and A. Chilkoti, J. Controlled Release, 2014,190, 314–330.

54 K.-I. Sano, T. Minamisawa and K. Shiba, Langmuir, 2010,26, 2231–2234.

55 J. Bhattacharyya, I. Weitzhandler, S. B. Ho, J. R. McDaniel,X. Li, L. Tang, J. Liu, M. Dewhirst and A. Chilkoti, Adv.Funct. Mater., 2017, 1605421.

Paper Nanoscale

Nanoscale This journal is © The Royal Society of Chemistry 2017

Publ

ishe

d on

07

Apr

il 20

17. D

ownl

oade

d by

Uni

vers

ity o

f C

alif

orni

a -

Irvi

ne o

n 01

/05/

2017

23:

19:1

3.

View Article Online

56 J. R. McDaniel, S. R. MacEwan, X. Li, D. C. Radford,C. D. Landon, M. Dewhirst and A. Chilkoti, Nano Lett.,2014, 14, 2890–2895.

57 E. M. Mastria, M. Chen, J. R. McDaniel, X. Li, J. Hyun,M. W. Dewhirst and A. Chilkoti, J. Controlled Release, 2015,208, 52–58.

58 J. R. McDaniel, J. A. MacKay, F. G. a. Quiroz andA. Chilkoti, Biomacromolecules, 2010, 11, 944–952.

59 J. R. McDaniel, I. Weitzhandler, S. Prevost, K. B. Vargo,M.-S. Appavou, D. A. Hammer, M. Gradzielski andA. Chilkoti, Nano Lett., 2014, 14, 6590–6598.

60 J. R. McDaniel, J. Bhattacharyya, K. B. Vargo,W. Hassouneh, D. A. Hammer and A. Chilkoti, Angew.Chem., 2013, 52, 1683–1687.

61 R. Haag, Angew. Chem., Int. Ed., 2004, 43, 278–282.62 M. Hrubý, Č. Koňák and K. Ulbrich, J. Controlled Release,

2005, 103, 137–148.63 Y. Bae, S. Fukushima, A. Harada and K. Kataoka, Angew.

Chem., Int. Ed., 2003, 42, 4640–4643.64 Y. Bae, N. Nishiyama, S. Fukushima, H. Koyama,

M. Yasuhiro and K. Kataoka, Bioconjugate Chem., 2005, 16,122–130.

Nanoscale Paper

This journal is © The Royal Society of Chemistry 2017 Nanoscale

Publ

ishe

d on

07

Apr

il 20

17. D

ownl

oade

d by

Uni

vers

ity o

f C

alif

orni

a -

Irvi

ne o

n 01

/05/

2017

23:

19:1

3.

View Article Online