Intracellular Protein Delivery System Using a Target-SpecificRepebody and Translocation Domain of Bacterial ExotoxinHee-Yeon Kim,†,∥ Jung Ae Kang,‡,∥ Jeong-Hyun Ryou,§ Gyeong Hee Lee,‡ Dae Seong Choi,‡

Dong Eun Lee,*,‡ and Hak-Sung Kim*,§

†Graduate School of Nanoscience and Technology (WCU), Korea Advanced Institute of Science and Technology (KAIST), 291Daehak-ro, Yuseong-gu, Daejeon, 305-701, Korea‡Advanced Radiation Technology Institute, Korea Atomic Energy Research Institute, 29 Geumgu-gil, Jeongeup-si, Jeollabuk-do580-185, Korea§Department of Biological Sciences, Korea Advanced Institute of Science and Technology (KAIST), 291 Daehak-ro, Yuseong-gu,Daejeon, 305-701, Korea

*S Supporting Information

ABSTRACT: With the high efficacy of protein-based therapeutics andplenty of intracellular drug targets, cytosolic protein delivery in a cell-specific manner has attracted considerable attention in the field ofprecision medicine. Herein, we present an intracellular protein deliverysystem based on a target-specific repebody and the translocationdomain of Pseudomonas aeruginosa exotoxin A. The delivery platformwas constructed by genetically fusing an EGFR-specific repebody as atargeting moiety to the translocation domain, while a protein cargo wasfused to the C-terminal end of the delivery platform. The delivery platform was revealed to efficiently translocate a protein cargoto the cytosol in a target-specific manner. We demonstrate the utility and potential of the delivery platform by showing aremarkable tumor regression with negligible toxicity in a xenograft mice model when gelonin was used as the cytotoxic proteincargo. The present platform can find wide applications to the cell-selective cytosolic delivery of diverse proteins in many areas.

Considerable advances in biological and medical scienceshave offered the understanding of molecular mechanisms

underlying the progression of pathology in various diseases.Consequently, an increasing number of promising druggabletargets have been identified, promoting the development oftherapeutic agents that can precisely control the unregulatedcell signaling processes on the molecular basis of a disease.Compared to small-molecule drugs, protein-based biologicshave attracted much attention as the most promising class oftargeted therapeutics owing to high efficacy and specificity.1,2

With a growing number of intracellular drug targets, the questfor cytosolic protein delivery in a cell-specific manner hassignificantly increased in the field of precision medicine.3,4

Diverse delivery systems for proteins have been developed,including liposomes, polymeric or metallic nanoparticles,dendrimers, cell-penetrating peptides, and micelles.5−8 Despitemany advances in drug delivery systems, however, the effectiveintracellular delivery of active proteins remains a challenge fortheir full therapeutic potential.Over the past decade, much attention has been paid to the

use of bacterial toxins for cytosolic delivery of small-moleculedrugs, oligonucleotides, and proteins.9−13 We previouslydeveloped a bacterial toxin-based protein delivery system bycombining the receptor binding domain of Escherichia coliShiga-like toxin and translocation domain of Pseudomonasaeruginosa exotoxin A.14 The delivery system was shown toefficiently translocate a protein cargo into the cytosol of

mammalian cells in a receptor-specific manner. However, theuse of the receptor-binding domain of Shiga-like toxin as atargeting moiety, which is specific for globotriaosylceramide(Gb3), has limitations to versatile applications of the deliverysystem: its engineering toward other cell surface targets isdifficult, and the relevance of Gb3 to diseases is low.Here, we present a cytosolic protein delivery system based on

an EGFR-specific repebody and the translocation domain ofPseudomonas aeruginosa exotoxin A. The repebody scaffold,which is composed of LRR (leucine-rich repeat) modules, wasshown to be easy and simple for generating repebodies specificfor diverse targets through a phage display selection.15−19 Thedelivery platform was constructed by genetic fusion of anEGFR-specific repebody as a targeting moiety to the trans-location domain of Pseudomonas aeruginosa exotoxin A.Gelonin, which is a ribosome-inactivating plant toxin (MW:29 kDa), was employed as a cytotoxic protein cargo and fusedto the C-terminus of the delivery platform.20 Free gelonincannot cross the cell membrane owing to the lack ofcarbohydrate-binding domain, thus showing a strong cytotox-icity when internalized into the cells.21−24 The developeddelivery system based on an EGFR-specific repebody exhibited

Received: July 6, 2017Accepted: October 11, 2017Published: October 11, 2017

Articles

pubs.acs.org/acschemicalbiology

© 2017 American Chemical Society 2891 DOI: 10.1021/acschembio.7b00562ACS Chem. Biol. 2017, 12, 2891−2897

Cite This: ACS Chem. Biol. 2017, 12, 2891-2897

a high translocation efficiency of a protein to the cytosol in atarget-specific manner. We demonstrate the utility and potentialof the delivery platform by investigating cell-selectivecytotoxicity and tumor regression in xenograft mice models.Details are reported herein.

■ RESULTS AND DISCUSSIONConstruction of a Cytosolic Protein Delivery Platform.

We previously developed a protein delivery system bycombining the receptor-binding domain of the Escherichia coliShiga-like toxin and translocation domain of Pseudomonasexotoxin A.14 The delivery system was shown to efficientlytranslocate diverse protein cargos into the cytosol of GB3-expressing mammalian cells. However, the receptor bindingdomain of Shiga-like toxin is specific for Globotriaosylceramide(Gb3) and was hard to diversify toward other cellular targets,limiting versatility of the previous delivery system. We thuswanted to develop a generally applicable protein deliveryplatform by replacing the Gb3-binding domain with a target-specific repebody. The repebody scaffold, which is composed ofLRR (leucine-rich repeat) modules, was previously developedas a nonantibody protein scaffold. The repebody scaffold wasshown to be stable and able to easily generate repebodiesspecific for diverse targets through a phage displayselection.15−19

A repebody (rEgH9) specific for EGFR ectodomain with abinding affinity of 301 pM was employed as a targeting moietyand genetically fused to the N-terminus of the translocationdomain of Pseudomonas aeruginosa exotoxin A.18 EGFR isknown to be overexpressed in a variety of cancer cells,25,26 and

has been used as a cell surface target for the delivery of variouscargos to the tumor sites.22,27−29 The translocation domain(residues 252 through 364) of Pseudomonas aeruginosa exotoxinA (TDP) was used owing to its favorable property of solubleexpression in E. coli. To check the performance of the deliveryplatform, we chose enhanced green fluorescent protein (EGFP)as a protein cargo and investigated the delivery efficiency(Figure 1A). EGFP was genetically fused to the C-terminus ofthe translocation domain, and the resulting construct (Rb-TDP-EGFP) was expressed as a soluble form in E. coli. The ERretention sequence (KDEL) was fused to the C-terminus of theEGFP to facilitate the retrograde transport.30 A schematicrepresentation and complete amino acid sequence of theresulting construct, Rb-TDP-EGFP, are shown in SupportingInformation Figure S1. The Rb-TDP-EGFP was eluted as amonomeric peak on a size-exclusion chromatogram and a singleband on SDS-PAGE gel (Supporting Information Figure S2).The Rb-TDP-EGFP was added to the EGFR-positive A431cells to confirm the cytosolic delivery of a protein cargo.Western blot analysis of the cell extracts showed a band of 38kDa which corresponds to the size of TDP-EGFP (Figure 1B),implying that the delivery platform follows the same pathway asbacterial exotoxin A in the cells. With the increased amount ofthe delivery platform, the EGFP level in the cells increased in adose-dependent manner. The various cell lines expressingEGFR were treated with the delivery platform, and afluorescence signal was investigated by confocal microscopy(Figure 1C). The Rb-TDP-EGFP was shown to be localized inthe cytosol of A431 and MDA-MB-468 cells expressing a highEGFR level. In contrast, MCF-7 cells, a negative control for

Figure 1. Construction of a protein delivery platform by combining a target-specific repebody and translocation domain of Pseudomonas aeruginosaexotoxin A. (A) Schematic representation of the delivery platform composed of an EGFR-specific repebody and translocation domain (TDP). EGFPwas used as a protein cargo and fused to the C-terminal end of the translocation domain. (B) Western blot analysis of intracellular delivery of EGFPin a dose dependent manner. (C) Target-specific EGFP delivery of different cancer cell lines through receptor-mediated endocytosis. (D)Intracellular delivery of Rb-TDP-EGFP in A431 cells. A431 cells were incubated with Cell tracker or ER tracker for 30 min and washed with DPBS.(E) The Z-stack images of A431 cells treated with 2 μM of Rb-TDP-EGFP for 6 h. Scale bar = 20 μm.

ACS Chemical Biology Articles

DOI: 10.1021/acschembio.7b00562ACS Chem. Biol. 2017, 12, 2891−2897

2892

EGFR expression, exhibited a negligible increase in the greenfluorescence. Based on the result, it is likely that EGFP wastranslocated to the cytosol by the delivery platform in areceptor-specific manner.To check the entrapment of EGFP in the ER compartment,

we traced the localization of EGFP using the cell and ERtrackers in A431 cells (Figure 1D). The fluorescence signalswere shown to mainly come from the cytosol, supportingefficient translocation of a cargo to the cytosol by the deliveryplatform. It is worth noting that the internalized cargo did notcolocalize with the ER trackers and distributed evenly over theentire cytosol of the cells. To further confirm the cytosolicdelivery of EGFP, the Z-stack images of A431 cells wereobtained after treatment with 2 μM of Rb-TDP-EGFP for 6 h.As a result, strong fluorescence signals were observed from thecytosol (Figure 1E). To test the effect of the concentration ofthe delivery platform, we performed an additional experimentwith 50 nM of Rb-TDP-EGFP (Supporting Information FigureS3). Distinct fluorescent signals were also detected from EGFPeven though the signal intensity was weak compared to 2 μM ofRb-TDP-EGFP. To estimate the delivery efficiency, subcellularfractionation was carried out after treatment with 2 μM of Rb-TDP-EGFP, and the amount of EGFP in the cytosol wasdetermined by Western blot using an anti-EGFP antibody(Supporting Information Figure S4). About 80% of totalinternalized EGFP was detected in the cytosolic fraction.It is well-known that the translocation domain of bacterial

exotoxin A is released from the receptor binding domain to thecytosol through the cleavage of a furin-sensitive loop(RHRQPRG) and reduction of a disulfide bond in endosomes(Figure 2A). To get insight into the translocation mechanism ofa cargo by the delivery platform, we conducted the time-courseanalysis of the localization of EGFP (Figure 2B). Following the

binding to EGFR at the cell membrane, the delivery platformwas shown to be internalized through endocytosis, exhibitingstrong fluorescence signals in endosomes after 1 h ofincubation. EGFP appeared to be colocalized with ER trackersat 3 h around the nucleus, and fluorescence signals weredistributed throughout the entire region of the cells at 5 h. Thisresult indicates the retrograde transport of EGFP by thedelivery platform as bacterial exotoxin. To examine the role ofeach domain, we constructed the delivery platform with an off-target repebody (Rboff) or without the translocation domainand tested the resulting constructs, Rboff-TDP-EGFP and Rb-EGFP, in terms of translocation of EGFP (SupportingInformation Figure S5). The Rboff-TDP-EGFP was shown toresult in a negligible delivery of EGFP to the cytosol, and mostof the Rb-EGFP was located on the plasma membrane orremained trapped in the endosomes. Based on the results, it islikely that the EGFR-specific repebody led to the receptor-mediated endocytosis of the cargo, and TDP facilitated theretrograde translocation of the cargo to the cytosol from theendosomes. It seems that protein cargo can escape from theendosomes to the golgi and translocate to the cytosol throughthe endoplasmic reticulum (ER). The translocation domain ofPseudomonas aeruginosa exotoxin A is likely to undergoproteolysis and reduction of the disulfide bond, releasing aprotein cargo from the repebody.

In Vitro Performance of the Delivery Platform. Wetested in vitro performance of the delivery platform usinggelonin as a cytotoxic protein cargo. Gelonin, which is a 29 kDasingle-chain ribosome-inactivation protein (RIP), is known toact as an N-glycosidase in cells and inactivate the 60S ribosomalsubunit.20 Gelonin (Glo) was genetically fused to the C-terminal end of the delivery platform (Figure 3A andSupporting Information Figure S6). The resulting constructed(Rb-TDP-Glo) was observed to be expressed as a soluble formin E. coli. Various cancer cell lines expressing different EGFRlevels were treated with Rb-TDP-Glo, and their viabilities wereassayed. A431 and MDA-MB-468 were known to express a highlevel of EGFR, while H1650 cells were revealed to have amoderate EGFR level. MCF7 cells were used as an EGFR-negative control. As a result, the Rb-TDP-Glo was shown toinduce dose-dependent mortality in A431 and MDA-MB-468cells, whereas no significant effect on the viability of MCF7 cellswas observed (Figure 3B).Both Rboff-TDP-Glo and free repebody were observed to

have a negligible cytotoxicity in all the tested cell lines. Thisresult indicates that the delivery platform efficiently trans-located gelonin to the cytosol of the EGFR-expressing cells in atarget-specific manner and consequently induced the cell death.We also conducted a competitive cytotoxicity assay to examinea receptor-mediated cytotoxicity (Supporting InformationFigure S7). For this, 1 μM of free repebody (rEgH9) waspreincubated as a competitor in EGFR-positive cells. Incontrast to the above result, the cell viability was significantlysuppressed in the presence of free repebody (rEgH9),indicating the EGFR-mediated endocytosis and subsequentcell death. Diverse target-specific repebodies can be easilyobtained through a phage display selection, and the presentdelivery platform seems to have the potential to be broadlyapplicable.We investigated the mode of action by Rb-TDP-Glo in

MDA-MB-468 cells using the TUNEL assay (terminaldeoxinucleotidyl transferase-mediated dUTP-fluorescein nickend labeling, Figure 3C). MDA-MB-468 and MCF-7 cells were

Figure 2. Intracellular delivery of EGFP via the retrograde transportpathway. (A) Schematic representation of cytosolic delivery of EGFP.The delivery system utilized receptor-mediated endocytosis and aretrograde transport pathway. (B) Time-course analysis of thelocalization of delivered EGFP. Fluorescence images were obtainedat different time intervals after treatment of A431 cells with Rb-TDP-EGFP (1 μM) and ER tracker (500 nM). Green, EGFP; red, ERtracker; blue, DAPI; scale bar = 20 μm.

ACS Chemical Biology Articles

DOI: 10.1021/acschembio.7b00562ACS Chem. Biol. 2017, 12, 2891−2897

2893

treated with 1 μM of Rb-TDP-Glo, followed by fluorescentlabeling of apoptotic cells by the TUNEL assay. MDA-MB-468cells exhibited a strong fluorescence signal, whereas negligiblegreen fluorescence was detected in MCF-7 cells, whichindicates that apoptosis occurred through internalization ofRb-TDP-Glo. MDA-MB-468 cells treated with only geloninshowed no green fluorescence due to the lack of EGFR-specificrepebody. Based on the results, it is likely that the cytotoxiceffect of Rb-TDP-Glo in the cells was linked to an apoptoticmechanism.In Vivo Performance of the Delivery Platform. We

investigated the in vivo antitumor activity of Rb-TDP-Glo inxenograft mice with MDA-MB-468 cells. When the tumorvolume reached about 100 mm3, the mice were subjected tointravenous injection of the Rb-TDP-Glo (10 mg/kg) fivetimes within 15 days (Figure 4A). Free EGFR-specific repebody(4.2 mg/kg) was also injected into mice intravenously every 3days as a control. Treatment of xenograft mice with Rb-TDP-Glo resulted in a remarkable suppression of tumor growth,whereas the tumor continued to grow when treated with PBSand free repebody. The tumor volume of mice treated with Rb-TDP-Glo reduced to 137 mm3 on day 30, whereas the tumorvolume increased to 465 mm3 when treated with PBS (Figure4B and C). The overall health status and body weight of micewere monitored twice a day during treatment with Rb-TDP-Glo, and no significant change in body weight was observed on

day 30 (Figure 4D). We also observed a negligible effect of Rb-TDP-Glo on the levels of aspartate and alanine amino-transferase (AST and ALT) as well as creatine and bloodurea nitrogen (BUN) even after 30 days (Figure 4E and F),implying a negligible toxic effect of Rb-TDP-Glo on the liverand kidneys. We also examined the toxicity of TDP-Glowithout a targeting moiety (Supporting Information Figure S8).No significant effects on body weight, hepatotoxicity, andnephrotoxicity were observed, implying a negligible toxicity offree gelonin. Taken together, our results demonstrate that thedelivery platform efficiently translocated the protein cargo tothe tumor site in a target-specific manner and effectivelysuppressed the tumor growth with negligible cytotoxicity inxenograft mice models.

Conclusion. We have demonstrated that the deliveryplatform enabled efficient translocation of a protein cargointo the cytosol in a target-specific manner through theendocytic pathway. Our delivery platform enabled an efficienttranslocation of a protein cargo to the cytosol in a target-specific manner. Furthermore, our platform was simple andcost-effective to produce in homogeneity by a bacterialexpression system through a genetic fusion. A protein cargo-fused delivery platform was well expressed as a soluble form inE. coli, and this seems to be due to the translocation domain ofPseudomonas aeruginosa exotoxin A. Treatment of xenograftmice with Rb-TDP-Glo led to a remarkable suppression of

Figure 3. Schematic representation for a target-specific translocation of gelonin using the delivery platform. (A) Gelonin (Glo) was genetically fusedto the C-terminus of the translocation domain of the delivery platform as a cytotoxic protein cargo. (B) Viability test for various cell lines expressingdifferent EGFR levels. (C) TUNEL assay for in situ detection of fragmented DNA on day 3 after treatment with 1 μM of Rb-TDP-Glo. MCF-7 wasused as the negative control. Scale bar = 20 μm.

ACS Chemical Biology Articles

DOI: 10.1021/acschembio.7b00562ACS Chem. Biol. 2017, 12, 2891−2897

2894

tumor growth in a target-specific manner with negligibletoxicity. An EGFR-specific repebody was employed as a modeltargeting moiety and genetically fused to the N-terminus of thetranslocation domain to construct the delivery platform. Therepebody scaffold, which is composed of LRR (leucine-richrepeat) modules, was shown to be stable and to easily generaterepebodies specific for diverse targets through a phage display.The binding affinity of a repebody could be matured to asubnanomolar range by modular engineering. Taken together,target-specificity of the delivery platform can be easilymodulated by replacing a repebody with another one, whichallows broad utility of the delivery platform.

■ MATERIALS AND METHODSProtein Expression and Purification. All genes coding for

various fusion proteins were cloned into the NdeI and XhoI restrictionenzyme sites of a pET21a vector (Novagen). DNA encoding theGelonin was PCR-amplified from a pMal-E4rGel (Addgene) anddigested with XhoI and HindIII. The fragment was then ligated intothe corresponding sites in DNA encoding the C-terminal region of arepebody (rEgH9)-TDP, which generated the expression constructrEgH9-TDP-Gelonin. The construct rEgH9-TDP-EGFP was gener-ated by replacing DNA encoding gelonin with EGFP. Each gene hasthe C-terminal hexa-histidine tag for affinity purification. The vectorswere transformed into E. coli BL21 (DE3), and IPTG (0.5 mM) wasadded for induction when the optical density reached OD600 between

0.5 and 0.7. Induced cells were further grown at 18 °C for 20 h andharvested by centrifugation at 7000 rpm. The cells were resuspendedin a lysis buffer (50 mM NaH2PO4, 300 mM NaCl, and 10 mMimidazole, at pH 8.0) and disrupted through sonication. Followingcentrifugation at 13 000 rpm for 1 h, supernatant was loaded onto aNi-NTA Superflow (Qiagen). The solution was washed with threecolumn volumes of a buffer (50 mM NaH2PO4, pH 8.0) containing300 mM NaCl and 50 mM imidazole. Hexahistidine tagged fusionproteins were eluted with elution buffer (50 mM NaH2PO4, 300 mMNaCl, and 250 mM imidazole, pH 8.0). The proteins were furtherpurified by gel permeation chromatography (Superdex200, GEHealthcare) with a phosphate-buffered saline (PBS; pH 7.4).

Mammalian Cell Culture. A431 (human epidermoid carcinoma,ATCC No. CRL-1555) cells were cultivated in a DMEM medium(Gibco) supplemented with 10% (v/v) fetal bovine serum (FBS;Hyclone). MDA-MB-468 (human adenocarcinoma, ATCC No. HTB-132) cells were cultivated in Leibovitz’s L-15 Medium (Gibco)supplemented with 10% (v/v) FBS. H1650 (human adenocarcinoma,ATCC No. CRL-5883) and MCF7 (human adenocarcinoma, ATCCNo. HTB-22) cells were cultivated in RPMI1640 Medium (Gibco)supplemented with 10% (v/v) FBS. All cell cultures were incubated ina 5% CO2 chamber at 37 °C.

Cytosolic Delivery of EGFP by the Delivery Platform. For thedelivery of proteins, the respective cells were attached to a cell cultureslide. Cells were seeded in an eight-well chamber slide (SPL) at adensity of 1 × 104 cells/well. After overnight incubation, serum-containing media with a predetermined concentration of repebody-TDP-EGFP were treated for 6 h. The Cell tracker (Invitrogen

Figure 4. In vivo antitumor activity of Rb-TDP-Glo. (A) Schedule of tumor inoculation and injection of Rb-TDP-Glo. (B) Tumor regression inxenograft mice by injection of Rb-TDP-Glo (10 mg/kg). Data represents the average and standard deviation in experiments with six mice. ***p <0.001 represents a significant differences compared with PBS control. (C) Representative image of tumors from sacrificed mice treated with Rb-TDP-Glo, free repebody, and PBS, respectively, on day 30. Values are presented as means ± SD of 6 mice per group. (D) Change in body weight ofxenograft mouse after injection of Rb-TDP-Glo (10 mg/kg). Repebody and PBS were used as a control. (E) Hepatotoxicity and (F) nephrotoxicityof Rb-TDP-Glo. Changes in the levels of aspartate and alanine aminotransferase (ALT and AST), creatine and blood urea nitrogen (BUN) wereassayed on day 30 after injection of Rb-TDP-Glo. Differences in the mean body weight, hepatotoxicity and nephrotoxicity of each group were notsignificant. Values are presented as means ± SD of 6 mice per group.

ACS Chemical Biology Articles

DOI: 10.1021/acschembio.7b00562ACS Chem. Biol. 2017, 12, 2891−2897

2895

C34551) and ER tracker (Invitrogen E34250) were treated for 30 and20 min, respectively. The cells were washed twice with DPBS and fixedwith 4% (w/v) paraformaldehyde for 20 min. Cell nuclei werecounterstained with DAPI in the mounting medium (Vector). Cellimaging was performed using a Zeiss LSM 780 confocal microscopewith a ×63 Apochromat 1.0 numerical aperture (NA) objective (CarlZeiss). The section interval of the z-stack image was 0.7 μm.In Vitro Cytotoxicity. Respective cells were seeded in a 96-well

plate (SPL) at a density of 5 × 103 cells/well, and the medium waschanged to a serum-free medium containing each protein. Differentconcentrations of the repebody-TDP-Gelonin were added to eachwell, followed by incubation for 3 days, and the serum-free mediumsupplemented with 10 μL of CCK-8 reagent (Dojindo) was added tothe wells and incubated for 2 h at 37 °C. The cell viability wasdetermined by measurement of OD450nm using an Infinite M200 platereader (Tecan).Western Blot Assay. The cells were washed with DPBS and

homogenized in a lysis buffer (Tris-HCl 50 mM, pH 8, NaCl 150 mMand 1% (v/v) Triton X-100). The supernatant fractions were separatedusing 12% SDS polyacrylamide gel electrophoresis, transferred to anitrocellulose membrane (Bio-Rad) at 100 V for 2 h in ice. Themembrane was incubated with a blocking buffer and phosphate-buffered saline with Tween 20 (PBST)-bovine serum albumin (BSA;PBS containing 0.1% (v/v) Tween-20 and 1% (v/v) BSA) andsubjected to a Western blot analysis using the goat anti-GFPpolyclonal antibody (Abcam) and rabbit antigoat IgG-HRP conjugatedantibody (Bio-Rad). The blots were visualized by an enhancedchemiluminescence solution (Millipore) using a LAS-3000 imagingsystem (Fuji-Film).TUNEL Assay. A TUNEL assay was used to evaluate the apoptotic

cells in mammalian cells using the In Situ Apoptosis Detection Kit(Takara) according to the manufacturer’s instructions. Briefly, cellswere seeded in an eight-well chamber slide (SPL) at a density of 1 ×104 cells/well. The cells were fixed in a 4% (w/v) formaldehydesolution for 15 min and washed with DPBS, followed bypermeabilization with permeabilization buffer for 20 min. The cellswere quenched with 3% (v/v) hydrogen peroxide, washed with DPBS,and labeled with TdT (Terminal deoxynucleotidyl Transferase)enzyme. The slide with cells was incubated at 37 °C humidifiedchamber for 1 h and then treated with DPBS as a stop buffer.Animals and Tumor Cells for Xenograft. The colorectal

carcinoma cell line MDA-MB-468 was obtained from ATCC. MDA-MB-468 cells were cultured in RPMI (Life Technologies, Inc.),supplemented with 10% (v/v) FBS and 1% (v/v) penicillin/streptomycin. Cell cultures were maintained at 37 °C in 5% CO2incubator. Six-week-old male BALB/c nu/nu mice were purchasedfrom Orient Bio Inc., Seoul, Korea. All animals were acclimatized tocontrolled conditions of temperature (23 ± 2 °C), humidity (55% ±5%), and light (12 h light/dark cycle) at the Central Animal Researchlaboratory, Advanced Radiation Technology Institute (ARTI), KoreaAtomic Energy Institute (KAERI). The animals were provided a sterilepellet diet and water ad libitum. The experiment was approved by theInstitutional Animal Ethical Committee and was performed in strictcompliance with the guidelines prescribed by the committee.Determination of Toxicity in Mice. The animals were divided

into three groups of six male BALB/c nu/nu mice. Each group ofanimals was given injections in the tail vein. Animal mortality wasobserved over 4 weeks.Antitumor Activity. MDA-MB-468 (5 × 106 cells with 100 μL of

Matrigel) was inoculated subcutaneously (s.c.) into the right lateralflank of the mice. After about 7 days, the mice were used in theexperiments when the tumor size was established at approximately 100mm3. The study was divided into three groups with six mice in eachgroup. To evaluate the antitumor activity of Rb-TDP-Gel fusionproteins in the MDA-MB-468 xenograft mice model, the mice wereintravenously (i.v.) injected with PBS, Rb (4.2 mg/kg), and Rb-TDP-Gel (10 mg/kg) on days 0, 3, 6, 9, and 12. The mice were injected withequimolar amounts of each protein (27 μM/dose). The body weightand tumor size was measured individually twice daily for 30 days. Thetumor volume was measured macroscopically with a caliper, and tumor

volume was calculated according to the formula (mm3) length ×width2 × 0.5. Mice were euthanized when tumors reached a volume of650 mm3 or when tumors showed skin ulcerations.

Determination of Hepatotoxicity and Nephrotoxicity. Theblood was allowed to clot, and plasma was separated by centrifugation(3000g for 20 min) and stored at 4 °C. Hepatotoxicity was evaluatedbased on aspartate, and alanine transaminases (AST and ALT) inserum were assayed using enzymatic kits (Asan Pharmaceuticals, Co.,Ltd.). Nephrotoxicity was assessed by measuring the levels of bloodurea nitrogen (BUN) and creatine using a Fuji dri-Chem NX 7000i(Fuji Photo Film Co., Ltd.).

Statistical Analysis. The values are expressed as means ± SD ofsix mice per group. Statistical analyses were performed using a one-wayanalysis of variance (ANOVA), and intergroup comparisons weremade using Tukey’s multiple comparison test. A value of P < 0.05 wasconsidered statistically significant.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acschem-bio.7b00562.

Figures S1−S8 and supplementary methods (PDF)

■ AUTHOR INFORMATIONCorresponding Authors*E-mail: delee@kaeri.re.kr.*E-mail: hskim76@kaist.ac.kr.ORCIDHee-Yeon Kim: 0000-0003-3640-5491Hak-Sung Kim: 0000-0002-4034-1892Author Contributions∥These authors contributed equally to the work.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis research was supported by the Mid-Career ResearcherProgram (NRF- 2014R1A2A1A01004198), Global ResearchLaboratory (NRF-2015K1A1A2033346), and Bio & MedicalT e c h n o l o g y D e v e l o p m e n t P r o g r a m ( N R F -2017M3A9F5031419) of the National Research Foundation(NRF) funded by the Ministry of Science, ICT & FuturePlanning.

■ REFERENCES(1) Schrama, D., Reisfeld, R. A., and Becker, J. C. (2006) Antibodytargeted drugs as cancer therapeutics. Nat. Rev. Drug Discovery 5, 147−159.(2) Reinsberg, J. (2007) Detection of human antibodies generatedagainst therapeutic antibodies used in tumor therapy. Methods Mol.Biol. 378, 195−204.(3) Dharap, S. S., Wang, Y., Chandna, P., Khandare, J. J., Qiu, B.,Gunaseelan, S., Sinko, P. J., Stein, S., Farmanfarmaian, A., and Minko,T. (2005) Tumor-specific targeting of an anticancer drug deliverysystem by LHRH peptide. Proc. Natl. Acad. Sci. U. S. A. 102, 12962−12967.(4) Liu, D., Liu, F., Liu, Z., Wang, L., and Zhang, N. (2011) Tumorspecific delivery and therapy by double-targeted nanostructured lipidcarriers with anti-VEGFR-2 antibody. Mol. Pharmaceutics 8, 2291−2301.(5) Maurer, N., Fenske, D. B., and Cullis, P. R. (2001) Developmentsin liposomal drug delivery systems. Expert Opin. Biol. Ther. 1, 923−947.

ACS Chemical Biology Articles

DOI: 10.1021/acschembio.7b00562ACS Chem. Biol. 2017, 12, 2891−2897

2896

(6) Liechty, W. B., Kryscio, D. R., Slaughter, B. V., and Peppas, N. A.(2010) Polymers for drug delivery systems. Annu. Rev. Chem. Biomol.Eng. 1, 149−173.(7) Parekh, H. S. (2007) The advance of dendrimers–a versatiletargeting platform for gene/drug delivery. Curr. Pharm. Des. 13, 2837−2850.(8) Croy, S. R., and Kwon, G. S. (2006) Polymeric micelles for drugdelivery. Curr. Pharm. Des. 12, 4669−4684.(9) Patyar, S., Joshi, R., Byrav, D. S., Prakash, A., Medhi, B., and Das,B. K. (2010) Bacteria in cancer therapy: a novel experimental strategy.J. Biomed. Sci. 17, 21.(10) Fabbri, A., Travaglione, S., Falzano, L., and Fiorentini, C. (2008)Bacterial protein toxins: current and potential clinical use. Curr. Med.Chem. 15, 1116−1125.(11) Johannes, L., and Romer, W. (2010) Shiga toxins–from cellbiology to biomedical applications. Nat. Rev. Microbiol. 8, 105−116.(12) Weldon, J. E., and Pastan, I. (2011) A guide to taming a toxin–recombinant immunotoxins constructed from Pseudomonas exotoxinA for the treatment of cancer. FEBS J. 278, 4683−4700.(13) Kreitman, R. J. (2003) Recombinant toxins for the treatment ofcancer. Curr. Opin. Mol. Ther. 5, 44−51.(14) Ryou, J. H., Sohn, Y. K., Hwang, D. E., Park, W. Y., Kim, N.,Heo, W. D., Kim, M. Y., and Kim, H. S. (2016) Engineering ofbacterial exotoxins for highly efficient and receptor-specific intra-cellular delivery of diverse cargos. Biotechnol. Bioeng. 113, 1639−1646.(15) Lee, S. C., Park, K., Han, J., Lee, J. J., Kim, H. J., Hong, S., Heu,W., Kim, Y. J., Ha, J. S., Lee, S. G., Cheong, H. K., Jeon, Y. H., Kim, D.,and Kim, H. S. (2012) Design of a binding scaffold based on variablelymphocyte receptors of jawless vertebrates by module engineering.Proc. Natl. Acad. Sci. U. S. A. 109, 3299−3304.(16) Seo, H. D., Lee, J. J., Kim, Y. J., Hantschel, O., Lee, S. G., andKim, H. S. (2017) Alkaline phosphatase-fused repebody as a newformat of immuno-reagent for an immunoassay. Anal. Chim. Acta 950,184−191.(17) Hwang, D. E., Ryou, J. H., Oh, J. R., Han, J. W., Park, T. K., andKim, H. S. (2016) Anti-Human VEGF Repebody EffectivelySuppresses Choroidal Neovascularization and Vascular Leakage.PLoS One 11, e0152522.(18) Lee, J. J., Choi, H. J., Yun, M., Kang, Y., Jung, J. E., Ryu, Y., Kim,T. Y., Cha, Y. J., Cho, H. S., Min, J. J., Chung, C. W., and Kim, H. S.(2015) Enzymatic prenylation and oxime ligation for the synthesis ofstable and homogeneous protein-drug conjugates for targeted therapy.Angew. Chem., Int. Ed. 54, 12020−12024.(19) Kim, H. Y., Lee, J. J., Kim, N., Heo, W. D., and Kim, H. S.(2016) Tracking protein-protein interaction and localization in livingcells using a high-affinity molecular binder. Biochem. Biophys. Res.Commun. 470, 857−863.(20) Stirpe, F., Olsnes, S., and Pihl, A. (1980) Gelonin, a newinhibitor of protein synthesis, nontoxic to intact cells. Isolation,characterization, and preparation of cytotoxic complexes withconcanavalin A. J. Biol. Chem. 255, 6947−6953.(21) Zhou, X., Qiu, J., Wang, Z., Huang, N., Li, X., Li, Q., Zhang, Y.,Zhao, C., Luo, C., Zhang, N., Teng, X., Chen, Z., Liu, X., Yu, X., Wu,W., Wei, Y. Q., and Li, J. (2012) In vitro and in vivo anti-tumoractivities of anti-EGFR single-chain variable fragment fused withrecombinant gelonin toxin. J. Cancer Res. Clin. Oncol. 138, 1081−1090.(22) Rosenblum, M. G., Cheung, L. H., Liu, Y., and Marks, J. W., III.(2003) Design, expression, purification, and characterization, in vitroand in vivo, of an antimelanoma single-chain Fv antibody fused to thetoxin gelonin. Cancer Res. 63, 3995−4002.(23) Veenendaal, L. M., Jin, H., Ran, S., Cheung, L., Navone, N.,Marks, J. W., Waltenberger, J., Thorpe, P., and Rosenblum, M. G.(2002) In vitro and in vivo studies of a VEGF121/rGelonin chimericfusion toxin targeting the neovasculature of solid tumors. Proc. Natl.Acad. Sci. U. S. A. 99, 7866−7871.(24) Cao, Y., Marks, J. D., Marks, J. W., Cheung, L. H., Kim, S., andRosenblum, M. G. (2009) Construction and characterization of novel,recombinant immunotoxins targeting the Her2/neu oncogeneproduct: in vitro and in vivo studies. Cancer Res. 69, 8987−8995.

(25) Seshacharyulu, P., Ponnusamy, M. P., Haridas, D., Jain, M.,Ganti, A. K., and Batra, S. K. (2012) Targeting the EGFR signalingpathway in cancer therapy. Expert Opin. Ther. Targets 16, 15−31.(26) Gan, H. K., Burgess, A. W., Clayton, A. H., and Scott, A. M.(2012) Targeting of a conformationally exposed, tumor-specificepitope of EGFR as a strategy for cancer therapy. Cancer Res. 72,2924−2930.(27) Master, A. M., Qi, Y., Oleinick, N. L., and Gupta, A. S. (2012)EGFR-mediated intracellular delivery of Pc 4 nanoformulation fortargeted photodynamic therapy of cancer: in vitro studies. Nano-medicine 8, 655−664.(28) Berstad, M. B., Cheung, L. H., Berg, K., Peng, Q., Fremstedal, A.S., Patzke, S., Rosenblum, M. G., and Weyergang, A. (2015) Design ofan EGFR-targeting toxin for photochemical delivery: in vitro and invivo selectivity and efficacy. Oncogene 34, 5582−5592.(29) Wu, G., Barth, R. F., Yang, W., Kawabata, S., Zhang, L., andGreen-Church, K. (2006) Targeted delivery of methotrexate toepidermal growth factor receptor-positive brain tumors by means ofcetuximab (IMC-C225) dendrimer bioconjugates. Mol. Cancer Ther. 5,52−59.(30) Cabrera, M., Muniz, M., Hidalgo, J., Vega, L., Martin, M. E., andVelasco, A. (2003) The retrieval function of the KDEL receptorrequires PKA phosphorylation of its C-terminus. Mol. Biol. Cell 14,4114−4125.

ACS Chemical Biology Articles

DOI: 10.1021/acschembio.7b00562ACS Chem. Biol. 2017, 12, 2891−2897

2897