Journal of Controlled Release 173 (2014) 11–17

Contents lists available at ScienceDirect

Journal of Controlled Release

j ourna l homepage: www.e lsev ie r .com/ locate / jconre l

Photosensitive Pt(IV)–azide prodrug-loaded nanoparticles exhibitcontrolled drug release and enhanced efficacy in vivo

Haihua Xiao a, Gavin T. Noble a, Jared F. Stefanick a, Ruogu Qi b,c, Tanyel Kiziltepe a,d,Xiabin Jing b,⁎, Basar Bilgicer a,d,e,⁎⁎a Department of Chemical and Biomolecular Engineering, University of Notre Dame, Notre Dame, IN 46556, USAb State Key Laboratory of Polymer Physics and Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, People's Republic of Chinac Graduate School of Chinese Academy of Sciences, Beijing 100049, People's Republic of Chinad Advanced Diagnostics and Therapeutics, University of Notre Dame, Notre Dame, IN 46556, USAe Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, IN 46556, USA

⁎ Corresponding author.⁎⁎ Correspondence to: B. Bilgicer, Department of ChemicaUniversity of Notre Dame, Notre Dame, IN 46556, USA.

E-mail addresses: xbjing@ciac.jl.cn (X. Jing), bbilgicer@

0168-3659/$ – see front matter © 2013 Elsevier B.V. All rihttp://dx.doi.org/10.1016/j.jconrel.2013.10.020

a b s t r a c t

a r t i c l e i n f o

Article history:Received 27 July 2013Accepted 16 October 2013Available online 25 October 2013

Keywords:Photosensitive chemotherapyPlatinum drugsDrug deliveryProdrugCisplatinNanomedicine

Cisplatin has long been the first line of treatment for a variety of solid tumors. However, poor pharmacokineticsand high incidences of resistance in the clinic have motivated the production of numerous alternative Pt-basedanticancer species. Recently, photosensitive Pt(IV) complexes have garnered much interest because they offera method of selective induction of active Pt(II) at the tumor site by UVA irradiation. Here, we report the firstsynthesis, in vitro and in vivo characterization of a novel series of photosensitive Pt(IV)–azide prodrugs andmicellar nanoparticle formulations thereof. Upon mild UVA irradiation, both free Pt(IV) complexes and micellarnanoparticles rapidly released biologically active Pt(II), capable of binding to 5′-GMP, while remaining extremelystable in the dark. In vitro, uptake of photosensitive Pt(IV) prodrugs by ovarian cancer SKOV-3 cells was greatlyenhanced with the micellar nanoparticles compared to their free prodrug analogs, as well as cisplatin andoxaliplatin. Increased cytotoxicity was observed upon UVA treatment, with up to a 13-fold enhancement overoxaliplatin for the micellar nanoparticles. In vivo bioavailability of micellar nanoparticles was enhanced ~10fold over free Pt(IV) prodrugs. Importantly, micellar nanoparticles demonstrated significantly improved efficacyagainst H22murine hepatocarcinoma, showing decreased systemic toxicity and increased tumor growth inhibitionrelative to small molecule drugs. These findings establish that photosensitive Pt(IV) complexes, specifically whenformulated into micellar nanoparticles, have the potential to offer a robust platform for the controlled deliveryand selective activation of Pt-based anticancer therapeutics.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Since the discovery of its anticancer activity in the late 1960s,cisplatin has become one of themost effective chemotherapeutic agentsagainst a large variety of solid tumors, including breast, ovarian andtesticular cancers [1–3], and is currently used to treat more than 50%of all cancers [4,5]. As a Pt(II) drug, the antitumor effects of cisplatinare attributed to the disruption of DNA structure in cell nuclei throughthe formation of intra- and interstrand cross-links, ultimately leadingto tumor cell death [1,6,7]. The great success of cisplatin has led tothe biological evaluation of thousands of other Pt(II) drugs including“classical” Pt(II) drugs such as carboplatin and oxaliplatin, as well as“non-classical” Pt(II) drugs such as picoplatin and BBR3464 [8,9]. Inaddition to Pt(II) drugs, octahedral Pt(IV) drugs have also been developedas prodrugs of toxic Pt(II) species as they exhibit stability in biological

l and Biomolecular Engineering,

nd.edu (B. Bilgicer).

ghts reserved.

fluid and are passively reduced to active Pt(II) in cancerous cells due tothe high intracellular concentrations of glutathione, ascorbic acid, andmercaptans [3,4]. Photoactive Pt(IV)–azide prodrugs have also beendeveloped and are converted to the biologically active Pt(II) formin vitro throughmild UVA irradiation [10,11]. Traditional photodynamictherapy (PDT) currently has many applications in the diagnosis andtreatments of various cancers, including esophageal and non-small celllung cancers, using a wide range of light (375–732 nm) andphotosensitizers [12]. One advantage Pt(IV)–azide complexes have overtraditional PDT is that they do not need photosensitizing catalysts oroxygen-rich environments [12]. However, to date, only three Pt(II)drugs, cisplatin, carboplatin, and oxaliplatin and no Pt(IV) drugs havebeen FDA approved [1–3]. Despite the widespread use of Pt(II) drugsin oncology and their clinically beneficial outcomes, they are associatedwith high incidences of level 3–4 adverse events including nephrotoxicityand neurotoxicity [1,3,6,8]. This is primarily due to their poorpharmacokinetic properties, necessitating the use of high doses forefficacy at the tumor site. However, even at high drug doses less than5% of the total administered drugs actually reach the cancer cells andonly ~1% of the intracellular drug reacts with nuclear DNA [13,14]. As

12 H. Xiao et al. / Journal of Controlled Release 173 (2014) 11–17

a result, there is an underlying need to develop alternative strategiesfor the controlled and targeted delivery of platinum drugs to improvetherapeutic outcome in the clinic.

Nanoparticle-based drug delivery systems have drawn particularattention in recent years because of their ability to offer enhancedaccumulation in tumor tissue through the enhanced permeation andretention (EPR) effect, caused by the leaky nature of angiogenic bloodvessels of solid tumors [15,16]. As a result, many nanoparticular drugdelivery systems for Pt drugs have been developed [17–19], and someare under clinical evaluation. Examples include liposomal formulationsof cisplatin (lipoplatin) and oxaliplatin (MBP-426) as well as polymer-based systems NC-6004 (cisplatin) and ProLindac (oxaliplatin) [17–19].In advanced preclinical studies, the nanoparticle-based drug deliverysystems of Pt drugs have shown reduced systemic toxicity and at leastequivalent efficacy compared to their small molecule counterparts[17,19]. However, release of active Pt(II) species from the drug carrieris only achieved indirectly, either via acid hydrolysis (due to lowerintracellular pH) and/or chemical reduction facilitated by the presenceof a higher concentration of intracellular biological reducing agents,such as ascorbic acid and glutathione (GSH) [18], which are bothapproaches that can limit efficacy. Consequently, there is an urgentneed to develop Pt nanoparticles, which can not only exploit theEPR effect and enhance drug accumulation at the tumor, but whichcan also be employed to externally and directly trigger Pt drugrelease in a highly controlled manner at the tumor site.

Here, we report for the first time the synthesis of a novel seriesof photosensitive Pt(IV)–azide prodrug complexes C1–C4 based oncisplatin (C1 and C2) and oxaliplatin (C3 and C4) in Scheme 1a. Theywere synthesized according to procedures modified from previousreports [17], with N50% yield. ESI-MS, IR, and 1H NMR confirmed thesuccessful synthesis of the complexes (Scheme S1, Fig. S1–S3). Then weprepared their micellar nanoparticle formulations NC1–NC4 for highlycontrolled and selective release of active Pt(II) at the tumor site forenhanced efficacy and decreased systemic toxicity (Scheme 1b). Wespecifically designed these complexes such that they exist in a non-toxic Pt(IV)–azide prodrug form when kept in the dark, but are rapidlyactivated to efficiently release cytotoxic Pt(II) species upon low-energy UVA irradiation. The micellar nanoparticles of these complexessignificantly improved their circulation time in the bloodstream, showedenhanced drug efficacy and decreased systemic toxicity compared to

Scheme 1. Chemical structures of photosensitive Pt(IV) prodrugs C1–C4 (a). Preparation,self-assembly and remote triggering by UVA irradiation of micellar nanoparticles NC1–NC4 (b). Polymer P1 is a triblock copolymer comprised of mPEG (m = 114),polycaprolactone (n=20) and poly-L-lysine (p=10).

cisplatin and oxaliplatin, the current gold-standard Pt-based drugs usedin chemotherapy.

2. Materials and methods

2.1. Synthesis and characterization of Pt(IV)–azide complexes and micellarnanoparticles

The synthesis of C1–C4 and NC1–NC4 is presented in the supportinginformation. C1–C4 were characterized using IR (Fig. S1), 1H NMR(Fig. S2), and ESI-MS (Fig. S3). Physicochemical properties of NC1–NC4are detailed in Table S1.

2.2. Light source

UVA irradiation was performed using a bank of 6 UVA light tubeswith 18W light sources (1.8mW/cm2, λmax=365nm).

2.3. Photosensitivity of Pt(IV)–azide complexes and nanoparticles

Aqueous solutions of C1–C4 or NC1–NC4 were UVA irradiated forthe indicated periods of time and the UV–vis spectra of the aqueoussolutions taken. For stability in the dark, aqueous solutions of C1–C4and NC1–NC4 were kept in the dark and UV–vis spectra taken at overtime.

2.4. Pt release profiles

Lyophilized nanoparticles NC1–NC4 (5 mg) were hydrated in PBS(10 mL, 0.1 M, pH 7.4), placed into a pre-swelled dialysis bag (3500molecular weight cut-off (MWCO)), and immersed into PBS (100mL)at 37 °C. Samples were kept in the dark or under UVA lamps. At 1 hintervals, 1.5 mL was withdrawn from the dialysate and measured forPt using inductively coupled plasma optical emission spectrometry(ICP-OES). After sampling, fresh PBS (1.5mL) was added to the dialysate.The process was repeated for 8h.

2.5. Cellular uptake studies

SKOV-3 cells were obtained from ATCC and grown as previouslydescribed [18], seeded in 6-well plates and treated with Pt species(5 μM Pt eq.) at 37 °C for 2 h or 6 h. The cells were washed with PBS,lysed, and Pt content measured by inductively coupled plasma massspectrometry (ICP-MS). Total protein content in each sample wasdetermined using a bicinchoninic acid (BCA) assay [18,20,21].

2.6. Cytotoxicity

SKOV-3 cellswere treatedwith platinum-based drugs (0.1 to 432μM).For the UVA irradiated samples, cells were incubated for 4h in the dark,UVA irradiated for 1 h, and then incubated in the dark for an additional43 h. For control samples, incubation was performed in the dark for48 h. Cytotoxicity was then assessed using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay [18,21].

2.7. In vivo studies

Murine cancer models were established in Kunming mice aspreviously described [20,21], and full experimental procedures aredetailed in the supporting information.

13H. Xiao et al. / Journal of Controlled Release 173 (2014) 11–17

3. Results and discussion

3.1. Synthesis of photosensitive Pt(IV)–azide complexes

Two novel Pt(IV)–azide complexes based on cisplatin (C1 and C2,Scheme 1) and two based on oxaliplatin (C3 and C4, Scheme 1), weresynthesized according to procedures modified from previous reports[17], with N50% yield. ESI-MS, IR, and 1H NMR confirmed the successfulsynthesis of the complexes (Scheme S1, Fig. S1–S3).

3.2. Pt(IV)–azide complexes exhibited high photosensitivity to UVA lightand stability in the dark

Drugs for phototherapy should display high photo responsivenessupon UVA irradiation. To confirm this, UV–vis spectroscopy was usedto monitor the photoactivation of C1–C4. Complexes C1–C4 share thesame intense N3 → Pt ligand-to-metal charge-transfer (LMCT) UV–visabsorbance centered at 258nm,with similar UV–vis spectra to previouslyreported Pt(IV)–azide complexes [10,11]. Irradiation of Pt(IV)–azidecomplexes with UVA light (365 nm, 1.8 mW/cm2) results in thedestruction of the N3 → Pt bond, causing a decrease in the UVabsorbance at 258nm as the Pt(IV) is reduced to the biologically activePt(II) form [11], providing a useful method to monitor the photo-activation of C1–C4. Upon UVA irradiation of C1–C4 at 365 nm, thepeak at 258nm for C1–C4 dropped rapidly, indicating fast degradationof the N3 → Pt bond (Fig. S4A). By plotting normalized absorbance at258 nm (NA258) vs. UVA irradiation time (Fig. S4B), the T50 values(time for 50% degradation of Pt(IV)-N3) for C1–C4 were obtained. TheT50 values for C1 and C2 (both 72min) were almost two times that ofC3 and C4 (41 min and 47 min), suggesting that complexes bearing1,2-cyclohexanediamine (DACH) ligands were more readily activatedby UVA irradiation. These differences in sensitivity of C1–C4 to low-energy UVA light could be beneficial for controlled drug release byenabling selective release rates. We designed the photosensitive Pt(IV)drugs so that they would be reduced to active Pt(II) species by releasingsuccinic acid upon UVA irradiation through hydrolysis of the Pt\succinicacid bond. To confirm this, we examined whether complexes C1–C4,which contain either one or two succinic acid moieties, released succinicacid upon UVA activation. Upon UVA irradiation, we observed theappearance of a new peak at m/z = 116.8 in the ESI-MS spectra(negative mode) of C1–C4 confirming the release of succinic acid(representative spectra for C1 and C3 are shown in Fig. S5).

For prolonged storage, drugs for phototherapy should also displayhigh stability in the dark. To confirm this, aqueous solutions of C1–C4were prepared and monitored by UV–vis spectroscopy and the NA258

versus time plotted (Fig. S4C). The intensity for all the synthesizedcomplexes remained unchanged up to 30days in dark,with no decreasein the N3 → Pt peak at 258 nm evident even up to 90 days. Thisemphasized that complexes C1–C4 were very stable and suitable forlong-term storage in the dark, while being highly sensitive towardsUVA irradiation for phototherapy.

3.3. Pt(IV)–azide complexes released biologically active Pt(II) upon UVAactivation and chelated with 5′-GMP

To evaluate if C1–C4 Pt(IV)–azide prodrugs would form cytotoxicDNA-adducts upon UVA activation, we analyzed the formation ofPt adducts using guanosine-5′-monophosphate (5′-GMP) [20,21].A representative MALDI-TOF-MS spectrum of C3 upon UVA irradiationin the presence of 5′-GMP is shown in Fig. S6. The major peak wasassigned as (DACH)Pt(5′-GMP)2 (m/z = 1002.2), indicating chelationwith 5′-GMP after the photoactivation of complex C3. This indicatedthat C3was converted to active Pt(II) species and subsequently chelatedwith 5′-GMP in the typical manner of a DNA crosslinking agent, implyingthat the C1–C4 Pt(IV) prodrugswould exhibit biological activity similar tocisplatin and oxaliplatin once UVA activated.

3.4. Formulation of Pt(IV)–azide complexes into micellar nanoparticles

To further maximize drug efficacy and reduce systemic toxicity,selective delivery of the photosensitive Pt(IV) prodrugs to the tumorsite is essential [20,21]. Nanoparticular drug delivery systems provideenhanced accumulation in the tumor tissue due to the EPR effect,thereby enhancing anti-tumor efficacy and decreasing systemic toxicity.In order to form polymeric micellar nanoparticles, simple 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide/N-hydroxysuccinimide (EDC/NHS)chemistry was used to form amide linkages between C1–C4 and apoly-L-lysine (PLL)-block co-polymer P1 to give polymeric conjugatesof C1–C4 (Scheme 1b). P1 is a non-toxic, biodegradable methoxyl-poly(ethylene glycol)-block-poly(ε-caprolactone)-poly(L-lysine)polymer (MPEG-b-PCL-b-PLL) with an average molecular weight of~9000 Da by 1H NMR, and has been previously employed for thedelivery of siRNA [22]. P1 is composed of 114 PEG, 20 caprolactoneand 10 L-lysine units. Full synthesis and characterization of P1 havebeen described previously [18]. The polymeric Pt(IV)–azide conjugatesself-assembled in aqueous solution to form the micellar nanoparticlesNC1–NC4 (Scheme 1b). P1 also self-assembled in aqueous solution,with the polycaprolactone forming the micelle core and the polarmPEG and PLL units presented at the micelle surface (Scheme S2).Prior to each experiment, nanoparticles NC1–NC4 were prepared byhydrating polymeric conjugates of C1–C4 at ~5 mM which was wellabove the critical micelle concentration (CMC) of P1 (reported data:32 mg/L, ca. 3.6 μM) [22]. Transmission emission microscopy (TEM)and dynamic light scattering (DLS) were used to confirm the formationofmicelles and the physicochemical properties of NC1–NC4micelles areshown in Table S1. NC1–NC4 displayed a reduced diameter (96–170nmby TEM and 130–205nm by DLS) and reduced zeta potential (0mV to+8.7 mV) compared to the mean diameter (200 nm by TEM and220nmbyDLS) and zeta potential (+49.3mV) of the drug-freemicellesof P1. This is likely due to the consumption of positively charged aminogroups on the micelle scaffold after drug conjugation, diminishingunfavorable charge–charge interactions between the polymer chains,thus allowing them to pack tighter in themicellar form. It is anticipatedthat the Pt(IV) drugs are present in the micelle core, as the hydrophilicgroups of P1 (PLL unit) and Pt(IV) complexes C1–C4 (axial succinicacid) are consumed as the PLL-Pt(IV) conjugates are formed, thus theconjugated Pt(IV) species become hydrophobic.

ICP-OESmeasurements of NC1–NC4 (Table S1) showed experimentalPt loadings of 9.8–12% Pt by weight (expected 13.4–15.3% for NC1–NC4for 100% conjugation), thus giving a conjugation yield of 70–80%. Thisanalysis demonstrated the successful formulation of C1–C4 into micellarnanoparticles NC1–NC4 with a high Pt-loading efficiency and definedsize in the range at which EPR effects and cellular uptake via endocytosisoccurs.

3.5. Pt(IV)–azide-loaded micellar nanoparticles displayed high UVAphotosensitivity, and stability in the dark

Next, we evaluated the Pt release rates from the Pt(IV)–azide-loadedmicelles NC1–NC4. Aqueous solutions of NC1–NC4 were exposed toUVA light and monitored by UV–vis spectroscopy. Micellar NC1–NC4showed the same UV absorbance centered at 258 nm due to N3 → PtLMCT and a similar reduction in peak intensity upon UVA irradiationas C1–C4 (Fig. 1a). This confirmed that the conjugation of C1–C4 to P1did not affect their activation upon UVA irradiation. NA258 vs. UVAirradiation time is shown in Fig. 1b. The T50 values of NC1–NC4 werein the range of 35–90 min, in the order of NC4 (35 min) ≈ NC3(40 min) b NC1 (64 min) b NC2 (84 min). This suggested that NC3 andNC4with the DACH ligands weremore sensitive towards UVA activation,as seen for C3 and C4. The in-dark stability of NC1–NC4was also assessed;rehydration of lyophilized NC1–NC4 displayed the same N3→Pt UV peakat 258nm as the freshly synthesizedmicellar nanoparticles. Furthermore,the UV–vis spectra of aqueous solutions of NC1–NC4 stored in the dark

Fig. 1.UVA sensitivity and in-dark stability of NC1–NC4. (a) Representative UV–vis spectraof NC3 upon UVA irradiation (0 to 90 min). The peak maxima at λ= 258 nm drop overtime, demonstrating the breakdown of the N3 → UVA sensitivity and in-dark stabilityversus irradiation time of NC1–NC4. (c) In-dark stability of NC1–NC4. NA258 wasunchanged for all samples in the dark indicating that the N3→ Pt bond was stable in thedark. (d) Representative “on–off” effect of intermittent UVA irradiation on NC3.

Fig. 2. Pt drug release from NC3 and the postulated release pathways. (a) Representativerelease profiles of NC3. (b) Schematic illustration of chelation of the reduced Pt(II) specieswith 5′-GMP (upper) as evidenced by MALDI-TOF MS of NC3 in the presence of 5′-GMPupon UVA irradiation (lower). (c) Postulated drug release pathways from NC3.

14 H. Xiao et al. / Journal of Controlled Release 173 (2014) 11–17

remained unchanged up to 30 days (Fig. 1c). These results emphasizedthat the Pt(IV) of micellar nanoparticles NC1–NC4 were extremely stablein the dark and ideal for phototherapy and long-term storage.

3.6. “On–off” photo-responsiveness of Pt(IV)–azide-loaded micellarnanoparticles

In order to confirm the highly controllable nature of drug releasefrom NC1–NC4, we measured the photo responsiveness of NC3 via UV–vis upon intermittent UVA irradiation. The absorbance versus irradiationtime is shown in Fig. 1d. Upon UVA irradiation, the UV–vis absorbanceof NC3 at 258 nm dropped dramatically (0 to 15min), corresponding tothe destruction of the N3→Pt bond and the reduction of Pt(IV) to releasePt(II). However, when the UVA irradiation was interrupted (from 15 to50 min) the maximum UV absorbance at 258 nm remained constant.Subsequent UVA irradiation again resulted in a drop in UV absorbance(50 to 80 min) and cessation of the UVA irradiation produced anotherplateau (80 to 95min). The highly controllable photo responsiveness ofNC3 is an attractive property for a drug delivery system, allowing afeasibly high degree of remote control over drug release.

3.7. Pt release from Pt(IV)–azide-loaded micellar nanoparticles was furtherenhanced at pH5

We next assessed the Pt release from NC1–NC4 using a dialysistechnique (MWCO=3500Da) at pH5.0 and pH7.4. ICP-OES was usedto monitor the release of Pt post-UVA activation (Pt(II) release viaphoto activation) or in the dark (Pt(IV) release via hydrolysis). UV–visspectroscopy was also used to confirm the stability of the N3 → Ptbond and whether the Pt species released was Pt(II) or Pt(IV). AsUVA irradiation results in the destruction of the N3→ Pt bond withconcomitant release of succinic acid (Fig. S5), accelerated release ofPt from the micelles should be observed relative to non-irradiatedcontrols. The Pt release profiles of NC3 are shown in Fig. 2a. The Pt releasefrom NC3 was found to be both UVA and pH dependent. As expected, Ptrelease from NC3 was found to be more susceptible to UVA irradiationthan to pH change. Regardless of pH, Pt was released faster upon UVA

irradiation than in the dark. The presence of Pt in dialysates kept inthe dark indicated that some Pt was released from NC1–NC4. Thesedialysates also possessed a UV peak at 258 nm, suggesting that the Ptreleased was in the Pt(IV) form and is likely due to slow hydrolysis ofC1–C4 from the P1 polymer chain. Interestingly, Pt was released fasterat pH 5.0 (which mimics the endocytic vesicles) than at pH 7.4, likelydue to the acid sensitivity of the Pt\succinic acid, COO\Pt bond.More specifically, after 400min, 36% of Pt(IV) was released in the darkat pH 7.4, while 91% release occurred upon UVA irradiation at pH 5.0.Release profiles for NC1, NC2 and NC4 are shown in Fig. S7. NC1 andNC2 displayed similar release rates, while NC3 and NC4 showed fasterPt release thanNC1 andNC2. The order of drug release rates formicellesNC1–NC4 is in agreement with the order of their T50 values (Fig. 1b)and demonstrated that NC3 was indeed the most UVA-responsivenanoparticle. Moreover, the enhanced activation upon UVA light ofNC1–NC4 at reduced pH suggested that they would be highly activein intracellular conditions.

3.8. Pt(IV)–azide-loaded micellar nanoparticles released biologically activePt(II) upon UVA activation and chelated 5′-GMP

ICP-OES analysis can only identify the total platinum drug released,but not the specific Pt species formed. Because it is generally believedthat the +2 oxidation state of Pt (Pt(II)) forms cytotoxic DNA adducts[23,24], we verified whether NC1–NC4 released the biologically activePt(II). MALDI-TOF MS analysis was performed after UVA activation ofNC3 in the presence of 5′-GMP (Fig. 2b). Mass spectra showed a majorpeak, which was assigned (DACH)Pt(5′-GMP)2 (m/z = 1002.2). Thisconfirmed that the Pt(IV)–azide-loaded micellar nanoparticles releasedPt(II) species that were capable of chelatingwith 5′-GMP and suggestedthat NC1–NC4would form cytotoxicDNA adducts in a similarmanner tocisplatin and oxaliplatin [1,6,7].

3.9. Possible drug release mechanism from Pt(IV)–azide-loaded micellarnanoparticles

In accordance with the above analyses we postulated that NC1–NC4release Pt drugs based on the representativemechanism shown for NC3in Fig. 2c. We prepared solutions of model Pt(II) and Pt(IV)–azidecomplexes and compared their UV–vis spectra to dialysates of micellarnanoparticles after UVA treatment (Fig. S8A–F). Under UVA irradiation,

Table 1Representative IC50 values of drugs (all units: μM Pt).

Sample Cisplatin Oxaliplatin NC1 NC2 NC3 NC4

IC50/UVA 12±1.7 19±2.9 6.2± 1.1 7.3± 1.4 1.5± 0.23 3.0± 0.23IC50/dark 16±2.5 25±3.2 98± 8.9 75±9.6 102±11.8 95±11.8

15H. Xiao et al. / Journal of Controlled Release 173 (2014) 11–17

NC3 undergoes direct photo-reduction to release the highly toxic Pt(II)species (hydrated oxaliplatin — [(DACH)Pt(H2O)2]2+, Fig. 2c, i) andnitrogen gas, as evidenced by the diminishing N3→ Pt UV absorbanceat 258 nm in Fig. 1a and b. Dialysates of UVA irradiated micellarnanoparticles also displayed diminished N3→ Pt absorbance (Fig. S8Eand Fig. S8F), which demonstrated the release of active Pt(II) from themicellar structure. In the dark, an alternative process can occur: FirstNC3 can undergo hydrolysis to release a free form of Pt(IV) throughhydrolysis of the COO\Pt coordinate bond ([(DACH)Pt(IV)(OH2(N3)2)],Fig. 2c, ii). UV–vis spectra of dialysates of micellar nanoparticles in thedark at room temperature displayed clear N3→Pt absorbance peaks at258 nm (Fig. S8C and Fig. S8D), suggesting that slow release of Pt(IV)–azide species did indeed occur. In vivo, enzymatic hydrolysis of thepolymer main chain may also occur, giving an oligo-Pt(IV)–azidecomplex. These hydrolyzed products can subsequently undergo photo-reduction to the active Pt(II) form ([(DACH)Pt(II)(H2O)2]2+) withUVA irradiation, becoming reactive towards DNA. Combined sensitivitytowards low pH and UVA irradiation is ideal for selective intracellularrelease of the Pt prodrugs. It is plausible that this mode of action isalso applicable to NC1, NC2, and NC4; for NC1 and NC2, the product isc,c-[Pt(NH3)2(H2O)2]2+, an active Pt(II) species of cisplatin [23,24].

3.10. Pt(IV)–azide-loaded micellar nanoparticles exhibited enhanced cellularuptake in vitro

Small molecule drugs are typically internalized via passive diffusionthrough the cell membrane [25,26]. Uptake of polymeric micelles, onthe other hand, is mediated by endocytosis, which can enhance druginternalization [27,28]. Therefore, we evaluated the cellular uptake ofPt from NC1–NC4 in comparison to free drug counterparts C1–C4, aswell as cisplatin and oxaliplatin. Uptake of Pt by human ovarian cancerSKOV-3 cells wasmeasured by ICP-MS at 2h and 6h (Fig. 3a). Pt uptakewas significantly greater for NC1–NC4 than the corresponding smallmolecule complexes C1–C4, cisplatin, and oxaliplatin at both 2 h and6 h. This reached an enhancement of ~19-fold at 2 h and ~45-fold at6 h for NC3 compared to C3, demonstrating efficient and enhancedcellular uptake of Pt(IV)–azide-loaded micellar nanoparticles.

3.11. Pt(IV)–azide-loadedmicellar nanoparticles showed increased efficacyupon UVA activation in vitro

In order to evaluate whether the enhanced uptake of Pt drugs fromNC1–NC4 would translate into increased efficacy in vitro, we measuredthe cytotoxicity of Pt(IV)–azide-loadedmicelles NC1–NC4 and the smallmolecule Pt complexes (cisplatin, oxaliplatin and complexes C1–C4),both in the dark and after UVA activation. For IC50 determination inthe dark, SKOV-3 cells were exposed to increasing concentrations of

Fig. 3. In vitro study of C1–C4 and NC1–NC4. (a) Intracellular uptake of drugs by SKOV-3cells after 2 h and 6 h of incubation (ICP-MS). (b) IC50 values of C1–C4 and NC1–NC4against SKOV-3 cells as determined by MTT assay. Full and detailed experimentalprocedures for the UVA and non-UVA treated groups is detailed in the SI. Data representstriplicates and cisplatin and oxaliplatin were used as controls.

Pt-equivalent doses of each agent for 48 h in the dark. For the UVAactivated group, SKOV-3 cells were exposed to increasing concentrationsof Pt-equivalent doses of each agent for 4 h in the dark, which wasfollowed by 1h of UVA irradiation, and then incubated for an additional43 h in the dark. Polymer P1, used to form the micellar nanoparticles isnon-toxic [22], and previous studies have shown UVA light to be non-damaging at 1.8 mW/cm2 [10]. Cell viability was N98% after 1 h of UVAirradiation in the absence of any drug (Fig. S9). Cell viability wasdetermined using an MTT assay and IC50 values were calculated andshown in Fig. 3b with selected IC50 values listed in Table 1.

In the dark, the Pt(II) drugs cisplatin and oxaliplatin exhibitedhigher cytotoxicity than the Pt(IV) complexes C1–C4, and the micellarnanoparticles NC1–NC4. NC1–NC4 were more cytotoxic than theirequivalent soluble complexes C1–C4, likely because of the greaterinternalization of Pt shown by NC1–NC4 (Fig. 3a). Furthermore, thePt(IV)–azide complexes may also be reduced to the cytotoxic Pt(II)form via intracellular reductants, such as GSH, albeit at a greatly reducedrate than through UVA irradiation. Therefore the enhanced uptakeafforded by micellar nanoparticles NC1–NC4 can account for theirincreased cytotoxicity compared to the free Pt(IV) complexes C1–C4in the absence of UVA irradiation.

Upon UVA activation, all IC50 values for C1–C4 and NC1–NC4improved dramatically as anticipated, since the Pt(IV)–azide complexeswere reduced to the more toxic Pt(II) form under UVA light. Upon UVAactivation, the Pt(IV) prodrugs based on the oxaliplatin structure werefound to be roughly 2–5 fold more cytotoxic than the correspondingcisplatin-based drugs (complexes C3 and C4 versus complexes C1 andC2; micelles NC-3 and NC-4 vs. micelles NC1 and NC2). This is inagreement with the differences in photosensitivity shown in Fig. 1b andsuggested that the Pt(IV)–azide complexes bearing DACH are activatedmore rapidly with UVA irradiation. The IC50 values of micelles NC3 andNC4 improved by 62 and 32-fold relative to C3 and C4, respectively,upon UVA activation. Importantly, UVA activated NC3 was 8-fold moreeffective than cisplatin and 13-fold more effective than oxaliplatin,which are the most widely used Pt drugs in the clinic. Taken together,these results demonstrated the high efficacy of NC1–NC4 against SKOV-3 cells, and established NC3 as the most efficacious Pt(IV)–azide-loadedmicellar nanoparticle, with significantly improved cytotoxicity whencompared to the gold standard compounds cisplatin and oxaliplatin.

3.12. Pt(IV)–azide-loadedmicellar nanoparticles exhibited prolonged bloodcirculation in vivo

PEGylated nanoparticles have been shown to increase bloodcirculation of drugs in vivo and improve patient outcome in the clinic[29,30]. We next evaluated if NC1–NC4 displayed enhanced circulationtimes relative to C1–C4, cisplatin and oxaliplatin. Using a H22 murinehepatocarcinoma model [21], we injected tumor-bearing mice with a5 mg Pt/kg equivalent dose of drugs intravenously and measured thetotal Pt concentration in the blood by ICP-MS at various time points.Representative results with C1, C3, and NC1 and NC3 are shown inFig. 4a. The blood clearance half-life periods of cisplatin, oxaliplatin,C1 and C3 in mice were roughly identical, around 2.7min, while thosefor NC1, and NC3 were found to be 38 min and 30 min respectively, aN10-fold enhancement. These results demonstrated that the formulationof photosensitive Pt(IV)–azide prodrug complexes C1–C4 into micellarnanoparticles enhanced their pharmacokinetics by improving bloodcirculation time.

Fig. 4. In vivo evaluation of C1–C4 and NC1–NC4. (a) Blood clearance of C1, C3, NC1 andNC3. All formulationswere injected intravenously once at 5mgPt/kgbodyweight. Relativetumor volume (b, c) and body weight (d, e) of mice treated with oxaliplatin (5 and10 mg Pt/kg), C3 (10 mg Pt/kg), NC3 (5 mg Pt/kg) and PBS in the dark (b, d) and uponUVA irradiation (c, e). For the in-dark controls (b, d) mice were returned to the darkimmediately after each drug injection. For the UVA activated groups (c, e), mice wereUVA treated for 1 h, 1 day after each drug injection and returned to the dark. Datarepresents means of 5 mice per group.

16 H. Xiao et al. / Journal of Controlled Release 173 (2014) 11–17

3.13. Pt(IV)–azide-loaded micellar nanoparticles show enhanced tumorgrowth inhibition and decreased systemic toxicity in vivo

Finally, we tested the efficacy of the photosensitive Pt(IV)-loadedmicelles NC3 in vivo using a H22 murine hepatocarcinoma model [21].NC3 was chosen for in vivo evaluation as it demonstrated the mostsignificant enhancement in cytotoxicity in vitro. C3 (10mg Pt/kg) andNC3 (5mgPt/kg) were injected into tumor-bearingmice intratumorallyat day 0 and day 4 in the dark. Themicewere then split into two groups;one groupwas kept constantly in the dark and the other groupwasUVAirradiated for 1 h on day 1 and day 5. After UVA treatment, mice werereturned to the dark. Tumor growth inhibition in the dark and withUVA activation is shown in Fig. 4b and c respectively. Oxaliplatin(5mgPt/kg and 10mgPt/kg) was also used as a control. Oxaliplatin,at 10mg Pt/kg, though effective at inhibiting tumor growth, causedexcessive weight loss and the mice from this group were sacrificed atday 7 because of high systemic toxicity. Complex C3 was only slightlymore effective in inhibiting tumor growth upon UVA irradiation thanin the dark, likely due to the fact that its uptake is limited by passivediffusion. The anti-tumor effect of NC3 was greatly enhanced upon UVAirradiation, displaying the greatest efficacy when compared to C3 andoxaliplatin. Tumor growth was fully suppressed and after 13 daysdaysthe tumor volume was reduced to around half its original size.

The relative body weight change was also measured to determinesystemic toxicity (Fig. 4d and e). Oxaliplatin at 5 and 10mgPt/kg, bothin the dark and after UVA treatment, caused a significant decrease intotal body weight, which resulted in the sacrifice of 10 mg Pt/kg-treated mice at day 7 due to excessive weight loss. In the dark, C3(10mgPt/kg), NC3 (5mgPt/kg) and PBS had little effect on bodyweight,suggesting no systemic toxicity. However, upon UVA irradiation C3 andNC3 caused a slight body weight reduction compared to the PBS controlgroup. The weight loss was still significantly lower than the oxaliplatin-treatedmice and at day 13 the bodyweight was comparable to the initialbody weight at day 0, suggesting that even with UVA irradiation, thePt(IV)–azide complexes and micellar formulations caused significantlyless systemic toxicity than the clinically used oxaliplatin. Combined,these results establish that the Pt(IV)–azide-loadedmicellar nanoparticleNC3 demonstrates enhanced anti-tumor efficacy and decreased systemictoxicity in vivo.

4. Conclusion

In conclusion, we have reported the synthesis of a series of novelphotosensitive Pt(IV)–azide prodrug complexes derived from cisplatinand oxaliplatin. With the aim of enhancing their pharmacokineticproperties and enhancing tumor accumulation, we grafted the complexesonto polymeric drug carriers, creating self-assembling Pt(IV)–azidemicellar nanoparticles with a diameter of 100–200 nm, sizes knownto induce EPR effects in vivo [15,16,19]. The Pt(IV)–azide complexesand their micellar nanoparticles displayed high sensitivity to UVAirradiation, while retaining stability in the dark. Furthermore, wedemonstrated that our Pt(IV)–azide complexes and micelles releasedbiologically active Pt(II) species capable of forming cytotoxic DNA cross-link adducts. In vitro studies showed that the photosensitive nanoparticlesof Pt(IV)–azide prodrugs were up to 8-fold more effective than cisplatinand 13-fold more effective than oxaliplatin, the clinical standards forPt(II) anticancer drugs. In vivo studies revealed that formulation of thephotosensitive Pt(IV) complexes into micellar nanoparticles enhancedtheir blood circulation half-life 10-fold from less than 3min to N30min.Importantly, micellar formulations demonstrated enhanced tumorgrowth inhibition and much reduced systemic toxicity upon UVAactivation, outperforming an equivalent dose of oxaliplatin.

Taken together, thework presented in this study establishes that thephotosensitive Pt(IV)–azide complexes have the potential to emerge as anewclass of platinum-based cancer therapeutics. Our results demonstratethat highly controllable Pt delivery is achievable via nanoparticulardelivery of photosensitive Pt(IV)–azide complexes, through controlledconversion of stable Pt(IV) to highly active anticancer Pt(II) speciesvia UVA activation. This method of delivery addresses many issueshindering traditional drug delivery approaches including rapid clearancefrom circulation, systemic toxicity, and on-demand dosing. Delivery ofphotosensitive Pt(IV) as micellar nanoparticles enables preferential drugaccumulation at the tumor site, and the conversion of Pt(IV) to activePt(II) species can be directly controlled through external stimuli. Finally,the photoactivation of the Pt(IV)–azide complexes described hereininvolves neither photosensitizing catalysts nor toxic singlet oxygenspecies. This provides an advantage to traditional PDT where theefficacy of the treatment is hindered by the fact that in solid tumorscancer cells are embedded in a hypoxic environment [12]. In the longterm, we anticipate that the highly controllable, nanoparticle-baseddelivery of the Pt(IV)–azide prodrugs described in this study has thepotential to have a significant impact on patient outcome by improvingtherapeutic index of Pt drugs in the clinic.

Acknowledgments

We thank the following institutions for their financial support:Walther Cancer Foundation, USA, The National Natural Science

17H. Xiao et al. / Journal of Controlled Release 173 (2014) 11–17

Foundation of China (Project No. 21004062, 51103148), and TheMinistryof Science and Technology of China (“973 Project”, No. 2009CB930102).

Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.jconrel.2013.10.020.

References

[1] T. Boulikas, A. Pantos, E. Bellis, P. Christofis, Designing platinum compounds incancer: structures and mechanisms, Cancer Ther. 5 (2007) 537–583.

[2] A.S. Abu-Surrah, M. Kettunen, Platinum group antitumor chemistry: design anddevelopment of new anticancer drugs complementary to cisplatin, Curr. Med. Chem.13 (2006) 1337–1357.

[3] T. Boulikas, M. Vougiouka, Cisplatin and platinum drugs at the molecular level, Oncol.Rep. 10 (2003) 1663–1682.

[4] N.J. Wheate, S. Walker, G.E. Craig, R. Qun, The status of platinum anticancer drugs inthe clinic and in clinical trials, Dalton Trans. 39 (2010) 8113–8127.

[5] M. Galanski, M.A. Jakupec, B.K. Keppler, Update of the preclinical situation ofanticancer platinum complexes: novel design strategies and innovative analyticalapproaches, Curr. Med. Chem. 12 (2005) 2075–2094.

[6] E.R. Jamieson, S.J. Lippard, Structure, recognition, and processing of cisplatin-DNAadducts, Chem. Rev. 99 (1999) 2467–2498.

[7] R.A. Alderden, M.D. Hall, T.W. Hambley, The discovery and development of cisplatin,J. Chem. Educ. 83 (2006) 728–734.

[8] N.J. Wheate, J.G. Collins, Multi-nuclear platinum complexes as anticancer drugs,Coord. Chem. Rev. 241 (2003) 133–145.

[9] G. Quiroga, Non-classical structures among current platinum complexes with potentialas antitumor drugs, Curr. Top. Med. Chem. 11 (2011) 2612.

[10] F.S. Mackay, J.A. Woods, P. Heringová, J. Kašpárková, A.M. Pizarro, S.A. Moggach, S.Parsons, V. Brabec, P.J. Sadler, A potent cytotoxic photoactivated platinum complex,Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 20743–20748.

[11] P. Bednarski, F.S. Mackay, P.J. Sadler, Photoactivatable platinum complexes, AntiCancer Agents Med. Chem. 7 (2007) 75–93.

[12] D. Dolmans, D. Fukumura, R.K. Jain, Photodynamic therapy for cancer, Nat. Rev.Cancer 3 (2003) 380–387.

[13] M.A. Fuertes, J. Castilla, C. Alonso, J.M. Perez, Cisplatin biochemical mechanism ofaction: from cytotoxicity to induction of cell death through interconnectionsbetween apoptotic and necrotic pathways, Curr. Med. Chem. 10 (2003) 257–266.

[14] V.M. Gonzalez, M.A. Fuertes, C. Alonso, J.M. Perez, Is cisplatin-induced cell deathalways produced by apoptosis? Mol. Pharmacol. 59 (2001) 657–663.

[15] H. Maeda, Macromolecular therapeutics in cancer treatment: the EPR effect andbeyond, J. Control. Release 164 (2012) 138–144.

[16] T. Kiziltepe, J.D. Ashley, J.F. Stefanick, Y.M. Qi, N.J. Alves, M.W. Handlogten, M.A.Suckow, R.M. Navari, B. Bilgicer, VLA4 targeting multifunctional nanoparticlesovercome drug resistance and inhibit tumor growth in multiple myeloma, BloodCancer J. (2012), http://dx.doi.org/10.1038/bcj.2012.10.

[17] X.Y. Wang, Z.J. Guo, Targeting and delivery of platinum-based anticancer drugs, Chem.Soc. Rev. 42 (2013) 202–224.

[18] H. Xiao, R. Qi, S. Liu, X. Hu, T. Duan, Y. Zheng, Y. Huang, X. Jing, Biodegradablepolymer–cisplatin(IV) conjugate as a pro-drug of cisplatin(II), Biomaterials 32 (2011)7732–7739.

[19] P. Sengupta, S. Basu, S. Soni, A. Pandey, B. Roy,M.S. Oh, K.T. Chin,A.S. Paraskar, S. Sarangi,Y. Connor, V.S. Sabbisetti, J. Kopparam, A. Kulkarni, K. Muto, C. Amarasiriwardena, I.Jayawardene, N. Lupoli, D.M. Dinulescu, J.V. Bonventre, R.A. Mashelkar, S. Sengupta,Cholesterol-tethered platinum II-based supramolecular nanoparticle increasesantitumor efficacy and reduces nephrotoxicity, Proc. Natl. Acad. Sci. U. S. A.109 (2012) 11294–11299.

[20] H. Xiao, H. Song, Q. Yang, H. Cai, R. Qi, L. Yan, S. Liu, Y. Zheng, Y. Huang, T. Liu, X. Jing,A prodrug strategy to deliver cisplatin(IV) and paclitaxel in nanomicelles to improveefficacy and tolerance, Biomaterials 33 (2012) 6507–6519.

[21] H. Xiao, W. Li, R. Qi, L. Yan, R. Wang, S. Liu, Y. Zheng, Z. Xie, Y. Huang, X. Jing,Co-delivery of daunomycin and oxaliplatin by biodegradable polymers for saferand more efficacious combination therapy, J. Control. Release 163 (2012) 304–314.

[22] R. Qi, S. Liu, J. Chen, H. Xiao, L. Yan, Y. Huang, X. Jing, Biodegradable copolymers withidentical cationic segments and their performance in siRNA delivery, J. Control.Release 159 (2012) 251–260.

[23] M.D. Hall, S. Amjadi, M. Zhang, P.J. Beale, T.W. Hambley, The mechanism of action ofplatinum(IV) complexes in ovarian cancer cell lines, J. Inorg. Biochem. 98 (2004)1614–1624.

[24] M.D. Hall, T.W. Hambley, Platinum(IV) antitumour compounds: their bioinorganicchemistry, Coord. Chem. Rev. 232 (2002) 49–67.

[25] T. Skovsgaard, N.I. Nissen, Membrane transport of anthracyclines, Pharmacol. Ther.18 (1982) 293–311.

[26] V. Cepeda, M.A. Fuertes, J. Castilla, C. Alonso, C. Quevedo, J.M. Pérez, Biochemicalmechanisms of cisplatin cytotoxicity, Anti Cancer AgentsMed. Chem. 7 (2007) 3–18.

[27] T. Iversen, T. Skotland, K. Sandvig, Endocytosis and intracellular transport ofnanoparticles: present knowledge and need for future studies, Nano Today 6 (2011)176–185.

[28] R. Vácha, F.J. Martinez-Veracoechea, D. Frenkel, Receptor-mediated endocytosis ofnanoparticles of various shapes, Nano Lett. 11 (2011) 5391–5395.

[29] L. Zhang, F.X. Gu, J.M. Chan, A.Z. Wang, R.S. Langer, O.C. Farokhzad, Nanoparticles inmedicine: therapeutic applications and developments, Clin. Pharmacol. Ther. 83(2008) 761–769.

[30] G. Pasut, F.M. Veronese, PEG conjugates in clinical development or use as anticanceragents: an overview, Adv. Drug Deliv. Rev. 61 (2009) 1177–1188.