lable at ScienceDirect

Biomaterials 155 (2018) 145e151

Contents lists avai

Biomaterials

journal homepage: www.elsevier .com/locate/biomater ia ls

Development of a theranostic prodrug for colon cancer therapy bycombining ligand-targeted delivery and enzyme-stimulated activation

Amit Sharma a, 1, Eun-Joong Kim b, 1, Hu Shi c, 1, Jin Yong Lee c, ***, Bong Geun Chung b, **,Jong Seung Kim a, *

a Department of Chemistry, Korea University, Seoul 02841, South Koreab Department of Mechanical Engineering, Sogang University, Seoul 04107, South Koreac Department of Chemistry, Sungkyunkwan University, Suwon 16419, South Korea

a r t i c l e i n f o

Article history:Received 24 August 2017Received in revised form13 November 2017Accepted 17 November 2017Available online 20 November 2017

Keywords:Targeted drug deliveryColon cancerTheranosticb-GalactosidaseDoxorubicin

* Corresponding author.** Corresponding author.*** Corresponding author.

E-mail addresses: jinylee@skku.edu (J.Y.(B.G. Chung), jongskim@korea.ac.kr (J.S. Kim).

1 These authors contributed equally to this work.

https://doi.org/10.1016/j.biomaterials.2017.11.0190142-9612/© 2017 Elsevier Ltd. All rights reserved.

a b s t r a c t

The high incidence of colorectal cancer worldwide is currently a major health concern. Although con-ventional chemotherapy and surgery are effective to some extent, there is always a risk of relapse due toassociated side effects, including post-surgical complications and non-discrimination between cancerand normal cells. In this study, we developed a small molecule-based theranostic system, Gal-Dox, whichis preferentially taken up by colon cancer cells through receptor-mediated endocytosis. After cancer-specific activation, the active drug Dox (doxorubicin) is released with a fluorescence turn-on response,allowing both drug localization and site of action to be monitored. The therapeutic potency of Gal-Doxwas also evaluated, both in vivo and ex vivo, thus illustrating the potential of Gal-Dox as a colorectalcancer theranostic with great specificity.

© 2017 Elsevier Ltd. All rights reserved.

1. Introduction

Colorectal cancer (CRC) is the third leading cause of cancer-related deaths worldwide, with over one million new cases inEurope and the US every year [1,2]. It is the second most commoncancer affecting women, after breast cancer, and the third mostcommon in men, after prostate and lung cancers, with the overallrisk for developing CRC being approximately 1 in 20 [3]. Typically,cancers confined to the colon are curable; however, if left un-treated, they spread to the regional lymph nodes andmetastasize todistant organs. Usually, the initial stages of the disease are curablewith surgical excision and combined chemotherapy. However, anappreciable proportion of CRC patients in early stage treatmentremain clinically remissive for a prolonged period of time, followedby approximately 50% chances of tumor recurrence with latermetastasis [4e6]. Surgery is accompanied by various

Lee), bchung@sogang.ac.kr

complications, including the formation of blood clots in the legs,bleeding at the surgical sites, and damage to nearby organs [7,8].Hence, there is an urgent need for the development of new ther-apeutic strategies for CRC, with improved clinical outcomes.

Conventional therapeutic strategies, involving the systemic de-livery of antitumor drugs, cannot distinguish between normal cellsand proliferating cancerous cells, causing collateral damage tohealthy tissue. To overcome such a formidable challenge, severaltargeted delivery systems have been developed, including smallmolecule-based drug delivery systems (DDS), liposomes, polymericsystems, aptamers, and inorganic nanoparticles [9e16]. Ideally,certain criteria must be fulfilled for the successful development of adrug delivery formulation. These include preferential targeting oftumors, maximal accumulation in tumors in vivo, and finally, anefficient drug release profile within the tumor, with minimalleakage to contiguous normal cells. Nanoformulations usually takeadvantage of leaky vasculature within the tumor mass for prefer-ential accumulation, termed as enhanced permeability and reten-tion effects (EPR). However, this significantly depends upon thestate of angiogenesis and vascularization of solid tumors [17e20].In past years, the design of nanomedicines in cancer therapeuticshas been upgraded, by the incorporation of ligands that activelytarget overexpressed receptors on tumors. However, there are still

A. Sharma et al. / Biomaterials 155 (2018) 145e151146

certain hurdles to be overcome to realize the final goal of precisionmedicine, including cost-effectiveness, pharmacokinetics, anddisease-driven formulations [20]. This is where a small molecule-based combination diagnosis and therapy can play a significantrole. In the past 20 years, several small molecule-based DDS havebeen designed and developed, for use in early diagnosis, bio-imaging, and therapeutics [10,11,15]. DDS are usually decoratedwith specific cancer targeting units, along with a drug activationtrigger moiety, linked to the chemotherapy drug. For drug activa-tion, several stimuli-responsive modes have been reported,including pH, elevated enzyme activity, redox status, light, andtemperature [14,15,21,22]. From a pharmaco-economic perspective,suitable targeting units and drug activation modes can be chosenaccording to the cancer subtype, and synthetically designed. Thisrequires the simultaneous development of new DDS with easieraccessibility and promising pharmacokinetics that can also betranslated to establish superior therapies for patients.

The lysosomal enzyme, b-galactosidase (b-gal) is known to beupregulated in various cancer subtypes, including liver, lung, andovarian cancers [23e25]. Several efforts have been made to visu-alize the upregulated activity of b-gal real time in various preclin-ical cancer models [24e26]. b-gal activity has also been utilized toactivate various cancer drug delivery formulations [27] and thera-nostic agents [28]. Considering cancer as a robust system, thepreferential targeting and cellular uptake behavior of eachdesigned DDS can be influenced by several factors, includingcomplex structural construct, inherent molecular features, andvariation in expression levels of both targeting and trigger modules[29e31]. Moreover, the overall safety and efficacy of complex DDScan be influenced by several parameters including biodistribution,potential immune toxicities and intended targeting that need to beaddressed carefully in preclinical and clinical stages.

In exploring a new direction for the advancement of antitumorchemotherapeutics, it is critical to understand the in vivo uptakeand drug activation behavior of a novel DDS. This is because theoptimal drug availability within a tumor is governed by DDS-associated features, which can alter the overall therapeutic effi-cacy and toxicity profile of the incorporated drug [30,31]. Asdetailed below, we focus on the b-gal enzyme as a target to bothdeliver and activate the DDS (Gal-Dox) for targeted drug deliveryin vitro and in vivo, in CRC tumor models. The galactose moiety ofGal-Dox serves as an excellent targeting ligand for asialoglyco-protein (ASGP) receptors, which play a significant role in tran-scriptional regulation at both cellular and molecular levels [32,33].Doxorubicin (Dox) was chosen as a model anticancer drug. Gal-Doxshowed a significant ability to target ASGP receptors on coloncancer cells, and enhanced activation behavior via lysosomal b-galenzyme. We also validated the potential of Gal-Dox in a mousetumor model and demonstrated its potential for effective cancertherapy.

2. Material and methods

2.1. UV/vis, fluorescence spectroscopy and high performance liquidchromatography methods

A stock solution of Gal-Dox was prepared in DMSO and forpreliminary solution studies, 10 mM working solution in phosphatesaline buffer (PBS) buffer solution (pH ¼ 7.4, 37 �C) containing 1%DMSO was used. Excitation was done at 480 nmwith all excitationand emission slit widths at 5 nm. The total volume for b-Galacto-sidase response and pH-dependent response tests was fixed at3.0 mL. For HPLC analyses, a reverse-phase column (C18, 5 mm,Waters) equipped YL9101S (YL-Clarity) instrument was used. AUVevis detector (480 nm) was used. The eluent used for each

analysis was a water-acetonitrile gradient (0e30 min; acetonitrilefrom 5% to 85%) with a 1.0 mL/min flow rate.

2.2. Cellular uptake studies by fluorescence imaging and flowcytometry

To evaluate the intracellular delivery of Gal-Dox, confocal laserscanning microscopy (CLSM) and flow cytometry were bothemployed on HT-29, HepG2, and HeLa cells. Cells were seeded andincubated in m-slide 8 well chamber (ibidi, Munich, Germany) at adensity of 1 � 104 cells per well. After overnight incubation, 10 mMGal-Doxwas treated into the culture media of well for various timepoints. All cells were fixed with 4% paraformaldehyde in PBS for20 min at RT, following washing with phosphate-buffered saline(PBS) for three times. Nuclei and F-actin were counterstained withDAPI (Invitrogen, Molecular Probe, Eugene, OR, USA) and AlexaFluor 488 phalloidin (Invitrogen, Molecular Probe) diluted in 3%BSA for 20 min, respectively. Furthermore, Lysotracker (Invitrogen,Molecular Probe) was added to the HT-29 cells at a concentration of50 nM for 30 min for the examination of co-localization with Gal-Dox and lysosome. The fluorescence images were acquired using aCLSM (LSM 710, Carl Zeiss, and Germany). To further verify the Gal-Dox being targeted to ASGP receptor positive cell lines, the cellularuptake was analyzed by flow cytometry (FACS Calibur, BD Bio-sciences, USA). HT-29 and HeLa cells were incubated in 12-wellplate at a density of 4 � 105/well and pretreated with 1 mM of D-galactose (Sigma-Aldrich, St. Louis, MO, USA). After incubation for4 h, the culture media were discarded, and the harvested cells wereexamined by a flow cytometry, following the treatment of 10 mMGal-Dox for 2 h. The histogram plots were performed using FlowJosoftware (TreStar, Olten, Switzerland) and mean fluorescence in-tensity values were shown in the bar graph.

2.3. In vitro anti-cancer assay

HT-29 and HeLa cells were plated to 96well plates at a density of1 � 104 cells per well for 12 h, and then treated with Dox and Gal-Dox at various concentrations for 24 h. Then, the culture mediawere replaced with fresh medium and the anti-cancer effect wasevaluated by a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazo-lium bromide (MTT, Roche Diagnostics GmbH, Mannheim, Ger-many) colorimetric assay. MTT was added and incubated for anadditional 2 h at 37 �C. The MTT absorbance was measured at570 nm using a microplate reader (EL800, Bio-Tek Instruments,Winooski, VT, USA) after the treatment of solubilization buffer(Roche Diagnostics) to dissolve formazan crystals. All experimentswere carried out with three replicates.

2.4. In vivo tumor targeting and anti-cancer effects

For in vivo tumor-targeted imaging and anti-tumor activity,colorectal cancer xenografts were established by subcutaneouslyinoculating six-week-old male BALB/c nude mice (Orient Bio,Sungnam, Korea) with HT-29 cells. All in vivo experimental pro-cedures involving animals were approved by the Korea UniversityInstitutional Animal Care and Use Committee (IACUC). A total of1 � 106 viable HT-29 cells were injected into the flanks of nudemice after anesthetization. When the tumors reached about5e10 mm in diameter, Dox or Gal-Dox was injected intravenouslyinto the tail vein in a single dose of 5 mg/kg. Whole-body fluores-cence images were continuously monitored and the dissectedmajor organs (i.e, liver, kidney, heart, lung, tumor) were visualizedor analyzed after sacrificing at 6 h, 24 h, 48 h, using an in vivoimaging system (Maestro, CRi Inc., Woburn, MA, USA). Moreover,tumors were fixed in 4% paraformaldehyde solution, embedded

A. Sharma et al. / Biomaterials 155 (2018) 145e151 147

routinely in paraffin, sectioned at 5 mm thickness, mounted usingwith DAPI containing Vecta shield (Vector Laboratories, CA, USA).To validate the therapeutic efficacy of Gal-Dox, HT-29-inoculatedxenograft mice were subcutaneously injected with Saline, and3mg/kg dose of free Dox and Gal-Dox (xenografts, n¼ 5) four timesevery other day. Tumor volumes and body weights were measuredevery 3 day using a caliper and terminated at 36 days. For in vivotoxicity study, blood samples were collected from all the mice forthe measurement of serum biochemical parameters includingblood urea nitrogen (BUN), lactate dehydrogenase (LDH), aspartateaminotransferase (AST), and alanine aminotransferase (ALT) at day30 after tumor inoculation. Also, one mouse from each group wassacrificed, and main organs such as heart, lung, liver, spleen, andkidney, were submitted for routine haematoxylin and eosin (H&E)staining.

3. Results and discussion

3.1. Synthesis, fluorescence response, and mechanism of Gal-Doxactivation

Gal-Dox was synthesized by following a series of steps in goodto excellent yield (Fig. 1A and S1, detailed in the supporting infor-mation). Following a reported procedure, galactosidase linker in-termediate 2 was synthesized. Further alcohol activation by p-nitrobenzyl chloroformate and base-mediated substitution withfree Dox resulted in the formation of intermediate 4. The acetylgroups were de-protected with sodium methoxide in anhydrousmethanol and treated with cation exchange resin, resulting in theformation of the final product, Gal-Dox. All synthesized

Fig. 1. Chemical structure, fluorescence response and activation mode for Gal-Dox. A, ChemFluorescence enhancement of Gal-Dox (10 mM) upon exposure to the b-gal enzyme (1 U/mL)upon b-gal action (10 mM, 1 U/mL enzyme, 37 �C). Fluorescence enhancement is directly rela(1 U/mL) at 37 �C for 2 h. F, Mode of activation of Gal-Dox upon b-gal action.

intermediates and products were well characterized by 1H/13Cnuclear magnetic resonance (NMR) and electrospray ionizationmass spectrometry (Figs. S14e25).

To obtain a deeper insight into binding sites at the atomic level,docking and molecular dynamics (MD) simulations were studied,including b-gal and Gal-Dox interactions. The molecular bindingsite between Gal-Dox and b-gal, and the decomposed binding freeenergies, are shown in Fig. 1B and S2, respectively. The calculatedbinding free energy between Gal-Dox and b-gal was found tobe �20.22 kcal/mol. According to the binding interaction calcula-tion, four significant interacting residues in b-galactosidase wereobserved, specified as N90 (red), P501 (orange), H528 (blue) andW987 (pink). Considering the calculated binding free energiesamong whole residues, W987 showed the strongest interactionwith the Gal-Dox ligand. Details of the simulation are provided inthe supporting information (Figs. S2e3).

To obtain insight into the preliminary stability and activationmode of Gal-Dox, the compound was incubated in phosphatebuffered saline (PBS, pH 7.4, 37 �C) for 24 h. No significantdecomposition was observed during this period (determined byfluorescence microscopy and high performance liquid chromatog-raphy, HPLC), suggesting sufficient chemical stability (Fig.1 and S4).Gal-Dox exhibited a weak fluorescence emission at 590 nm (exci-tation wavelength lexc. ¼ 480 nm). Interestingly, exposure of Gal-Dox (10 mM) to b-gal (1 U/mL) resulted in cleavage of the galacto-sidase moiety, followed by the 1,6-elimination route to release freeDox with simultaneous fluorescence enhancement at 590 nm(Fig. 1C and D). Consequently, we used the fluorescence enhance-ment at 590 nm as an activation response, for the observation ofdirect Dox release in further studies. The Gal-Dox activation and

ical structure of Gal-Dox. B, Molecular binding of Gal-Dox with the b-gal enzyme. C,(excitation wavelength lexc. ¼ 480 nm). D, Time-dependent fluorescence enhancementted to cumulative Dox release. E, RP-HPLC curves of Gal-Dox (10 mM) treated with b-gal

A. Sharma et al. / Biomaterials 155 (2018) 145e151148

drug release mechanisms were also determined using mass spec-troscopy (MS) and HPLC (Fig. 1E and S7). The HPLC profile showed apeak with an elution time of 20.4 min, corresponding to Gal-Dox.Upon treatment with b-gal, a new peak appeared at 15.3 min,corresponding to free Dox. Altogether, treatment of Gal-Dox withb-gal resulted in the complete release of active Dox, over a timeperiod of 2 h. Given the disparities in pH and the presence of otherbio-analytes between cancerous and normal healthy cells, we nextevaluated the fluorescence behavior of Gal-Dox, in the presence ofvarious bio-analytes, and at a range of pHs. No appreciable changein the fluorescence signal of Gal-Doxwas observed at different pHs,or in the presence of various bio-analytes, including reductase,GSH, H2O2, and amino acids (Figs. S5e6). These results suggestedthat Gal-Dox should exhibit high stability under biological condi-tions. Gal-Dox exhibited distinct fluorescence enhancement(590 nm) only upon treatment with the b-gal enzyme.

3.2. ASGP receptor-targeted delivery of Gal-Dox into colorectalcancer cells

The specific interaction between the galactose residue and ASGPreceptors has been used for hepatocyte-targeted DDS via receptor-mediated endocytosis [34]. In this study, Gal-Doxwas employed asa theranostic prodrug against the intestinal ASGP receptors in coloncancer [35]. To confirm the selective delivery of Gal-Dox to coloncancer cells overexpressing ASGP receptors, we evaluated its tar-geting ability both in HT-29 cells, a colon adenocarcinoma cell line,and in HepG2 cells, a well-known cancer cell line bearing the ASGPreceptor. As shown in Fig. 2, HT-29 and HepG2 cells showed pro-nounced intracellular fluorescence, while relatively low fluores-cence signals were detected in HeLa cells, an ASGP receptor-negative cell line (control) [36,37], even after 2 h incubation. Co-localization experiments confirmed that Gal-Dox was specificallylocalized with lysosomes of HT-29 cells, in which b-gal activityoccurs (Figs. S8e9), Especially, intracellular fluorescence wasclearly observed 30 min after treatment because of its targetingability for ASGP receptors, while free Dox was not detectable at the

Fig. 2. Targeted cellular uptake of Gal-Dox through ASGP receptor-mediated endocytosis. Hparaformaldehyde, and counterstained with DAPI and Alexa Fluor 488-Phalloidin to visualize(red) were obtained with a laser using an excitation wavelength of 480 nm and an emissioncolour in this figure legend, the reader is referred to the web version of this article.)

same time. Fluorescence in Gal-Dox-treated cells was reached aplateau after 6 h (Fig. S8). To further verify the ASGP receptor-targeted cellular uptake of Gal-Dox, a competition assay was con-ducted by flow cytometry analysis. Initially, HT-29 cells were pre-treated 100 times with free galactose as a competitor, followed bythe addition of Gal-Dox at a concentration of 10 mMwith incubationfor 2 h. There was no significant difference in fluorescence intensitybetween HeLa cells treated with excess galactose and control un-treated cells, whereas the fluorescence intensity in HT-29 cellspretreated with galactose remarkably decreased (Fig. S10). Alto-gether, these results support that the cellular uptake of Gal-Dox canbe specifically achieved by ASGP receptor-mediated endocytosis. Tothe best of our knowledge, this is the first report citing the specifictargeting properties of a small molecule-based prodrug system toASGP receptor-overexpressing colon cancer cells, through cellularuptake via receptor-mediated endocytosis.

3.3. b-Galactosidase mediated activity of Gal-Dox

As Gal-Dox exhibited b-gal-triggered Dox release, we nextconfirmed enzyme-specific Gal-Dox activation via the reduction ofb-gal activity within the cells. To knock down the expression of b-gal, an siRNA strategy was attempted to investigate the activity ofGal-Dox in the presence of the low level of endogenous b-galwithin HT-29 cells. For this, HT-29 cells were treated with variousconcentrations of b-gal siRNA for 48 h, and the optimal siRNAconcentration (50 nM) was determined by western blot analysis forfurther studies (Fig. 3A). After b-gal expression interference bypretreated siRNA (50 nM) for 48 h, HT-29 cells were incubated with10 mM of Gal-Dox for 0.5 h and 12 h. While fluorescence decreasedin HT-29 cells with low b-gal expression, we observed relativelystrong fluorescence in control siRNA-untreated cells (Fig. 3B andFig. S11). Enzyme-specific activation of Gal-Dox was furtherconfirmed by flow cytometric analysis, which provided a compar-ison between the mean fluorescence intensity of Gal-Dox-treatedHT-29 cells, with and without siRNA knockdown (Fig. 3C). Collec-tively, these siRNA silencing results suggest that the activation of

T-29, HepG2 and HeLa cells were incubated with 10 mM Gal-Dox for 2 h, fixed with 4%nuclei (blue) and F-actin (green), respectively. Confocal microscopy images of Gal-Doxwavelength of 560e590 nm. Scale bar: 20 mm. (For interpretation of the references to

Fig. 3. Gal-Dox activation through b-galactosidase-mediated cleavage. (A) Western blot analysis of b-gal knock-down by siRNA. HT-29 cells were transfected with b-gal siRNA(25 nM and 50 nM) for 48 h. The silencing effect of siRNAwas analyzed by western blot, using an anti-b-gal antibody and b-actin antibody as a loading control. (B) After transfectionof siRNA (50 nM) directed against b-gal for 48 h, HT-29 cells were treated with 10 mM Gal-Dox for 0.5 h, and examined with confocal microscopy. Filter sets: Doxorubicin (lex: 480/lem: 560e590), F-actin (lex: 495/lem: 518 and DAPI (lex: 358/lem: 461) Scale bar: 20 mm. (C) Flow cytometry analysis was conducted with HT-29 cells treated with 10 mM Gal-Doxafter transfection of 50 nM of b-gal siRNA.

A. Sharma et al. / Biomaterials 155 (2018) 145e151 149

Gal-Dox could be achieved by upregulated b-gal-responsive trig-gering. This sophisticated system permits simultaneous Dox releaseand fluorescence activation-based imaging, for greater precision inmonitoring the drug activation event and its location.

3.4. Anticancer effects of Gal-Dox with cell selectivity

The cytotoxicity of Gal-Dox was evaluated in HT-29 and HeLacells using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazo-lium bromide (MTT) assay, as its anti-proliferative activity can bedetermined by selective cellular uptake via ASGP receptor-mediated endocytosis. Both cells were treated with free Dox and

Fig. 4. Anticancer effects of Gal-Dox in HT-29 cancer cells (A) and HeLa (B) cells. Both cells24 h. Cell viability was then determined using the MTT assay.

Gal-Dox at concentrations ranging from 1 to 200 mM, for 24 h.Although Gal-Dox did not significantly reduce viability even at themaximum dose in both cells, similar behavior was observed forboth free Dox- and Gal-Dox-treated HT-29 cells (Fig. 4A). However,a remarkable difference in cytotoxicity was detected between freeDox- and Gal-Dox-treated HeLa cells (Fig. 4B). The comparison ofIC50 values indicated that Gal-Dox-treated HT-29 cells possessedapproximately 3-fold more potent therapeutic effects than HeLacell (Table S1). These results demonstrate that targeted Gal-Doxdelivery via the overexpressed ASGP receptors in colon cancer cellsnot only improved cellular uptake, but also enhanced its anticancereffect.

were treated with Gal-Dox and Dox as a positive control, at various concentrations for

A. Sharma et al. / Biomaterials 155 (2018) 145e151150

3.5. In vivo/ex vivo diagnosis and chemotherapeutic in xenograftmouse models

To investigate the antitumor activity of Gal-Dox, in vivo tumor-targeting ability was initially confirmed in HT-29 tumor-bearingmice. For bio-distribution studies using an in vivo fluorescenceimaging system, saline, free Dox and Gal-Dox were intravenouslyinjected into mice at a single dose of 3 mg/kg, and real-time whole-body fluorescence imaging was performed 2d after administration.While the tumor-targeted accumulation of Gal-Dox was continu-ously visualized by enhanced fluorescence signal up to 48 h in theHT-29 tumor, free Dox was observed to slightly clear from the tu-mor site in 24 h post injection (Fig. 5 and S12A). After sacrificing themice 48 h later, the tumor tissues were harvested and subsequentlysubjected to ex vivo fluorescence imaging (Fig. S12B). Collectively,whole-body fluorescence imaging results were comparable tothose for the dissected tumor tissues (Fig. 5B) and biodistributionstudies (Fig. S12C and D) from saline-, free Dox-, and Gal-Dox-treated mice. The in vivo therapeutic efficacy of Gal-Dox was alsoevaluated in nude mice bearing HT-29 cell-inoculated tumors,following intravenous administration of Gal-Dox (5 mg/kg, fourtimes every second day), and controls including saline and free Dox.As shown in Fig. 5C, tumor growth was significantly retarded inDox- and Gal-Dox-injected xenograft mice, compared to saline-injected mice. However, administration of Gal-Dox showedremarkable tumor growth inhibition (53.1%) compared to free Doxtreatment (34.9%, see Fig. 5D). Although Dox has been known tohave several adverse effects such as cardiotoxicity, no significant

Fig. 5. Tumor-targeted accumulation and in vivo antitumor activity of Gal-Dox in nude micHT-29-xenograft nude mice. Mice bearing HT29 subcutaneous xenografts were intravenousafter treatment using the Maestro in vivo imaging system. (B) Fluorescence images in paraffigroups. HT-29-bearing mice were intravenously injected with saline, Dox, or Gal-Dox at a dweek. Data are shown as mean ± SE. *: P < 0.05 versus Dox; ***: P < 0.001 versus saline, deand Gal-Dox at 36 days.

in vivo toxicities were observed during this study (Fig. S13). Webelieve that this remarkable therapeutic response of Gal-Dox re-flects its efficient tumor uptake behavior by the galactosidase unit,which not only serves to improve cancer targeting, but also as anefficient prodrug activation mode to deliver the active Dox.

4. Conclusions

In conclusion, a small molecule-based theranostic prodrug, Gal-Dox, was successfully developed for colon cancer chemotherapy,using receptor-mediated targeting and enzyme-responsive activa-tion strategies. The imaging properties and therapeutic efficacy ofGal-Dox in colorectal cancer models have been successfullydemonstrated both in vitro and in vivo. Gal-Dox is preferentiallytaken up by HT-29 cancer cells through ASGP receptor-mediatedendocytosis and was activated by elevated lysosomal b-galenzyme. In particular, the theranostic activation event is accom-panied by a simultaneous fluorescence turn-on response. This alsosupports the additional possibility of monitoring both the site ofdrug activation, and the therapeutic response towards the tumor,during early stages of treatment. Compared to the pre-reportedmulti-component DDS systems incorporating Galactosidase astrigger, drugs and cancer-specific targeting units altogether in acomplex design, Gal-Doxwith simple design possesses remarkabletumor specific targeting and therapeutic potential in ASGP receptorpositive cell lines. Additionally, complexity of pre-designed DDS asmulti-component requires detailed design and orthogonal analysiswith reproducible scale-up process for achieving a consistent

e bearing HT-29 cancer xenografts. (A) In vivo fluorescence imaging of Gal-Dox-treatedly injected with 5 mg/kg of Gal-Dox and in vivo fluorescent images were obtained 48 hn-embedded tumor tissue. Scale bar: 20 mm (C) Tumor growth inhibition in the variousose of 3 mg/kg, four times every second day. Tumor volume was measured twice pertermined using Student's t-test. (D) Tumor growth inhibition after treatment with Dox

A. Sharma et al. / Biomaterials 155 (2018) 145e151 151

commercial product for intended physiological properties, phar-macological profiles, and biological response. Gal-Dox with a sim-ple design, can be accessed with ease synthetically from pharmaco-economic point of view also. Hence Gal-Dox shows the potential asan ideal theranostic system in future personalized cancerchemotherapeutics.

Acknowledgements

This research was supported by the CRI project (No. 2009-0081566, J.S.K.) and the Bio & Medical Technology DevelopmentProgram (No. 2015M3A9D7030461, B.G.C.) of the National ResearchFoundation funded by the Ministry of Science, ICT and FuturePlanning of Korea.

Appendix A. Supplementary data

Supplementary data related to this article can be found athttps://doi.org/10.1016/j.biomaterials.2017.11.019.

References

[1] R.L. Segel, et al., Colorectal cancer statistics, 2017, CA Cancer J. Clin. 67 (2017)177e193.

[2] F.A. Haggar, R.P. Boushey, Colorectal cancer epidemiology: incidence, mor-tality, survival, and risk factors, Clin. Colon Rectal Surg. 22 (2009) 191e197.

[3] C.E. Kistler, Colorectal-cancer incidence and mortality after screening, N. Engl.J. Med. 369 (2009) 2354e2355.

[4] S.D. Markowitz, D.M. Dawson, J. Willis, J.K. Willson, Focus on colon cancer,Cancer Cell 1 (2002) 233e236.

[5] H.G. Welch, D.J. Robertson, Colorectal cancer on the decline e why screeningcan't explain it all, N. Engl. J. Med. 374 (2016) 1605e1607.

[6] M. Riihim€aki, A. Hemminki, J. Sundquist, K. Hemminki, Patterns of metastasisin colon and rectal cancer, Sci. Rep. 6 (2016) 29765.

[7] G. Grosso, A. Biondi, S. Marventano, A. Mistretta, G. Calabrese, F. Basile, Majorpostoperative complications and survival for colon cancer elderly patients,BMC Surg. 12 (Suppl 1) (2012) S20.

[8] J. Kountouras, C. Zavos, D. Chatzopoulos, Therapy for colorectal cancer, N. Engl.J. Med. 352 (2005) 1820e1821.

[9] P.T. Wong, S.K. Choi, Mechanisms of drug release in nanotherapeutic deliverysystems, Chem. Rev. 115 (2015) 3388e3432.

[10] M.H. Lee, A. Sharma, M.J. Chang, J. Lee, S. Son, J.L. Sessler, C. Kang, J.S. Kim,Fluorogenic reaction-based prodrug conjugates as targeted theranostics,Chem. Soc. Rev. (2017), https://doi.org/10.1039/C7CS00557A.

[11] M. Srinivasarao, C.V. Galliford, P.S. Low, Principles in the design of ligand-targeted cancer therapeutics and imaging agents, Nat. Rev. Drug. Discov. 14(2015) 203e219.

[12] S. Bhattacharyya, R.A. Kudgus, R. Bhattacharya, P. Mukherjee, Inorganicnanoparticles in cancer therapy, Pharm. Res. 28 (2011) 237e259.

[13] X. Wu, J. Chen, M. Wu, J.X. Zhao, Aptamers: active targeting ligands for cancerdiagnosis and therapy, Theranostics 5 (2015) 322e344.

[14] V.P. Torchilin, Multifunctional, stimuli-sensitive nanoparticulate systems fordrug delivery, Nat. Rev. Drug. Discov. 13 (2014) 813e827.

[15] J.H. Jang, H. Lee, A. Sharma, S.M. Lee, T.H. Lee, C. Kang, J.S. Kim, Indomethacin-guided cancer selective prodrug conjugate activated by histone deacetylaseand tumour-associated protease, Chem. Commun. 52 (2016) 9965e9968.

[16] M.H. Lee, E.J. Kim, H. Lee, H.M. Kim, M.J. Chang, S.Y. Park, K.S. Hong, J.S. Kim,J.L. Sessler, Liposomal texaphyrin theranostics for metastatic liver cancer,J. Am. Chem. Soc. 138 (2016) 16380e16387.

[17] D. Doleschel, et al., Erythropoietin improves the accumulation and therapeutic

effects of carboplatin by enhancing tumor vascularization and perfusion,Theranostics 5 (2015) 905e918.

[18] W. Jiang, Y. Huang, Y. An, B.Y. Kim, Remodeling tumor vasculature to enhancedelivery of intermediate-sized nanoparticles, ACS Nano 9 (2015) 8689e8696.

[19] R.K. Jain, T. Stylianopoulos, Delivering nanomedicine to solid tumors, Nat. Rev.Clin. Oncol. 7 (2010) 653e664.

[20] J.I. Hare, T. Lammers, M.B. Ashford, S. Puri, G. Storm, S.T. Barry, Challenges andstrategies in anti-cancer nanomedicine development: an industry perspective,Adv. Drug. Deliv. Rev. 108 (2017) 25e38.

[21] S. Nowag, R. Haag, pH-responsive micro- and nanocarrier systems, Angew.Chem. Int. Ed. Engl. 53 (2014) 49e51.

[22] S. Jin, J. Wan, L. Meng, X. Huang, J. Guo, L. Liu, C. Wang, Biodegradation andtoxicity of protease/redox/pH stimuli-eesponsive PEGlated PMAA nano-hydrogels for targeting drug delivery, ACS Appl. Mater. Interfaces 7 (2015)19843e19852.

[23] S.K. Chatterjee, M. Bhattacharya, J.J. Barlow, Glycosyltransferase and glycosi-dase activities in ovarian cancer patients, Cancer Res. 39 (1979) 1943e1951.

[24] D. Asanuma, M. Sakabe, M. Kamiya, K. Yamamoto, J. Hiratake, M. Ogawa,N. Kosaka, P.L. Choyke, T. Nagano, H. Kobayashi, Y. Urano, Sensitive b-galac-tosidase-targeting fluorescence probe for visualizing small peritoneal meta-static tumours in vivo, Nat. Commun. 6 (2015) 6463.

[25] E.J. Kim, R. Kumar, A. Sharma, B. Yoon, H.M. Kim, H. Lee, K.S. Hong, J.S. Kim,In vivo imaging of b-galactosidase stimulated activity in hepatocellular car-cinoma using ligand-targeted fluorescent probe, Biomaterials 122 (2017)83e90.

[26] L. Liu, R.P. Mason, Imaging beta-galactosidase activity in human tumor xe-nografts and transgenic mice using a chemiluminescent substrate, PLoS One 5(2010), e12024.

[27] T. Legigan, J. Clarhaut, I. Tranoy-Opalinski, A. Monvoisin, B. Renoux,M. Thomas, A. Le Pape, S. Lerondel, S. Papot, The first generation of b-galac-tosidase-responsive prodrugs designed for the selective treatment of solidtumors in prodrug monotherapy, Angew. Chem. Int. Ed. Engl. 51 (2012)11606e11610.

[28] J. Alsarraf, E. P�eraudeau, P. Poinot, I. Tranoy-Opalinski, J. Clarhaut, B. Renoux,S. Papot, A dendritic b-galactosidase-responsive folate-monomethylauristatinE conjugate, Chem. Commun. (Camb) 51 (2015) 15792e15795.

[29] E. Vlashi, L.E. Kelderhouse, J.E. Sturgis, P.S. Low, Effect of folate-targetednanoparticle size on their rates of penetration into solid tumors, ACS Nano7 (2013) 8573e8582.

[30] J.H. Jang, W.R. Kim, A. Sharma, S.H. Cho, T.D. James, C. Kang, J.S. Kim, Targetedtumor detection: guidelines for developing biotinylated diagnostics, Chem.Commun. 53 (2017) 2154e2157.

[31] A. Ivetac, J.A. McCammon, Mapping the druggable allosteric space of G-pro-tein coupled receptors: a fragment-based molecular dynamics approach,Chem. Biol. Drug. Des. 76 (2010) 201e217.

[32] P.P. Breitfeld, C.F. Simmons Jr., G.J. Strous, H.J. Geuze, A.L. Schwartz, Cellbiology of the asialoglycoprotein receptor system: a model of receptor-mediated endocytosis, Int. Rev. Cytol. 97 (1985) 47e95.

[33] M.D. Bider, M. Spiess, Ligand-induced endocytosis of the asialoglycoproteinreceptor: evidence for heterogeneity in subunit oligomerization, FEBS Lett.434 (1998) 37e41.

[34] Y. Zou, Y. Song, W. Yang, F. Meng, H. Liu, Z. Zhong, Galactose-installed photo-crosslinked pH-sensitive degradable micelles for active targeting chemo-therapy of hepatocellular carcinoma in mice, J. Control Release 193 (2014)154e161.

[35] J.Z. Mu, M. Gordon, J.S. Shao, D.H. Alpers, Apical expression of functionalasialoglycoprotein receptor in the human intestinal cell line HT-29, Gastro-enterology 113 (1997) 1501e1509.

[36] Y.M. Li, S.C. Xu, J. Li, K.Q. Han, H.F. Pi, L. Zheng, G.H. Zuo, X.B. Huang, H.Y. Li,H.Z. Zhao, Z.P. Yu, Z. Zhou, P. Liang, Epithelial-mesenchymal transitionmarkers expressed in circulating tumor cells in hepatocellular carcinomapatients with different stages of disease, Cell Death Dis. 4 (2013), e831.

[37] Y. Tao, J. He, M. Zhang, Y. Hao, J. Liub, P. Ni, Galactosylated biodegradablepoly(ε-caprolactone-co-phosphoester) random copolymer nanoparticles forpotent hepatoma-targeting delivery of doxorubicin, Polym. Chem. 5 (2014)3443e3452.