Ac

XLa

Ub

4

a

ARRAA

KCRhT

1

dcmpebiwTat[

LvT

0h

Immunology Letters 154 (2013) 18– 24

Contents lists available at ScienceDirect

Immunology Letters

jou rn al hom ep age: www.elsev ier .com/ locate / immlet

new recombinant immunotoxin hscFv-ETA’ demonstrates specificytotoxicity against chronic myeloid leukemia cells in vitro

iaoying Zhua,1, Kun Taob,1, Yajuan Lia, Shenfeng Lia, Lijun Zhanga, Dong Wanga,iang Zhonga, Wenli Fenga,∗

Department of Clinical Hematology, Key Laboratory of Laboratory Medical Diagnostics Designated by the Ministry of Education, Chongqing Medicalniversity, 1 Yixueyuan Road, Yuzhong District, Chongqing 400016, PR ChinaDepartment of Immunology, Molecular Medicine and Cancer Research, Chongqing Medical University, 1 Yixueyuan Road, Yuzhong District, Chongqing00016, PR China

r t i c l e i n f o

rticle history:eceived 11 April 2013eceived in revised form 11 July 2013ccepted 5 August 2013vailable online 15 August 2013

eywords:hronic myeloid leukemiaecombinant immunotoxin

a b s t r a c t

Antibodies against cell surface antigens of tumor have attracted increasing attention in immunother-apy for tumor diagnosis and treatment. Recently, we constructed a new recombinant immunotoxin forpossible clinical application in patients with chronic myeloid leukemia (CML). A functional humanizedsingle chain variable fragment (hscFv) against CML patient cells was previously obtained from an anti-CML cell hybridoma derived monoclonal antibody. By insertion into the bacterial vector pWW20, thehscFv was fused with a deletion mutant of Pseudomonas exotoxin A (ETA’). Then the fusion fragment wasinserted into the bacterial vector pET32a(+). After isopropyl �-d-thiogalactoside (IPTG) induction, the6× His tagged hscFv-ETA’ protein was periplasmically expressed and purified by Ni2+-NTA column. The

scFvargeting therapy

characteristics of the recombinant protein were assessed by cell membrane-ELISA, flow cytometry, andtoxicity assays in CML cell lines and CML patient cells. The recombinant immunotoxin showed significanttoxicity toward the CML cell lines K562 and KU812 as tested by MTT and apoptosis assay. Approximately37% of leukemia cells of CML patients were driven into apoptosis by hscFv-ETA’ as measured by flowcytometric analysis. In conclusion, the hscFv-ETA’ is efficacious against CML in vitro, providing the basisfor a novel therapeutic strategy for the treatment of CML patients.

. Introduction

Chronic myeloid leukemia (CML) is a clonal myeloproliferativeisorder characterized by the aberrant growth of granular leuko-ytes in bone marrow and peripheral blood [1,2]. Currently, bonearrow transplantation is internationally recognized as the only

ossible cure for the treatment of CML, but this method is veryxpensive and risky. In recent years, tyrosine kinase inhibitors haveeen developed for targeted therapy of CML. The tyrosine kinase

nhibitors are efficient, but they cannot completely cure patientsith CML, and the resistance or relapse would not be ignored [3].

herefore, new strategies are needed to develop for targeted ther-

py of CML. Some selective approaches, such as antibody-basedherapy, might offer a promising tool for specific treatment of CML4].

∗ Corresponding author at: Department of Clinical Hematology, Key Laboratory ofaboratory Medical Diagnostics of Ministry of Education, Chongqing Medical Uni-ersity, 1 Yixueyuan Road, Yuzhong District, Chongqing 400016, PR China.el.: +86 023 68485938; fax: +86 023 68485005.

E-mail address: fengwlcqmu@sina.com (W. Feng).1 These authors contributed equally to the work.

165-2478/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.imlet.2013.08.002

© 2013 Elsevier B.V. All rights reserved.

A potential strategy for enhancing the anti-tumor activity ofnative antibodies and helping to eliminate dormant or drug-resistant cells is chemical conjugation with drugs, isotopes or toxins[4–8]. The potent and selective anti-tumor activity of a number ofchemical and recombinant immunotoxins (rIT) has been corrob-orated in previous reports [9]. The first products were bacterialtoxins chemically coupled to antibodies, such as gemtuzumabozogamicin (Mylotarg), which target the internalizing myelo-cyte surface marker CD33 [10]. Several of these first-generationimmunotoxins have been tested in animal models and humanpatients [11], but the major obstacles to effective therapy haveuntil recently included the heterogeneity of the conjugation prod-ucts and dose-limiting side effects such as vascular leak syndrome[12]. Compared with chemically-linked conjugates, recombinantimmunotoxins have several advantages: (a) immunotoxins aredefined and compact molecules, so they are easy to be modified interms of their cytotoxic potency, stability and affinity [13,14]; (b)they are much more economical for produce and purify [15,16]; (c)

the tumor targeting and tissue distribution properties of immuno-toxins can be significantly improved through the use of scFvs whichconsist of the monoclonal antibody (mAb) variable domains joinedby a flexibly polypeptide linker [17,18]. Thus, peptide toxins such as

ogy Letters 154 (2013) 18– 24 19

Pito

afcmgssdaa

strsFmectwa

2

2

ctvt((dop(te(PnfmgwLgw

2

hdXopt

Table 1Sequences of primers used for PCR amplification.

Primers Sequences

P1 5′ -CCCAAGCTTCAGGTGCAGCTGGTGCAGTCTG-3′

P2 5′ -GCTCTAGATGATCTCCACCTTGGTCCCTCCGC-3′

P3 5′ -CGGAATTCCAGGTGCAGCTGGTGCAGTCTG-3′

P4 5′ -TGTCCGCCGGGCCCGCGAAGCTTCCGGCGGCGACCGGGCAG-3′

′ ′

X. Zhu et al. / Immunol

seudomonas exotoxin, Diphtheria toxin and Ricin have been genet-cally coupled to monoclonal antibodies or their derivatives in ordero create specific anti-cancer drugs which can kill dormant cells andvercome drug resistance.

Exotoxin A (ETA) is a 66 kDa protein secreted by Pseudomonaseroginosa and is a commonly used toxin which consists of threeunctional domains including the cell-binding domain, the translo-ation domain and the toxic domain. Its cytotoxic effect onammalian cells depends on its ADP-ribosylation of human elon-

ation factor 2 (EF-2). EF-2 is an essential component of proteinynthesis machinery and can irreversibly inhibit protein synthe-is [19]. To improve specificity of rITs, the original cell-bindingomain of ETA is usually replaced with specific ligands or scFvsllowing targeting of certain tumor cells and efficient cell killingfter internalization [20].

Compared with whole antibodies, scFvs have many advantages.cFv consists of variable regions of heavy and light chains withouthe Fc portion, leading to low immunogenicity [21]. scFv fragmentsetain the binding specificity of the parent antibody and demon-trate better tissue penetration and faster blood clearance [22].urthermore, scFvs can be cloned and expressed in bacterial andammalian cells, making it possible to produce large quantities

asily and cost-effectively. In the present study, we report theonstruction, expression and initial characterization of immuno-oxin hscFv-ETA’, targeting CML cells specifically. The rIT, whichas expressed in the periplasm, corroborated specific cytotoxicity

gainst CML cells in vitro.

. Materials and methods

.1. Bacterial strains, oligonucleotides, and plasmids

E. coli DH5� and BL21 (DE3) were used as host strains forloning and expression experiments in E. coli expression sys-em. The pUC57-hscFv vector which originated from pUC57ector (Novagen) contained the hscFv gene fragment. The vec-or pET32a(+) carries a 6× His tag for nickel-nitrilotriacetic acidNi2+-NTA) agarose resin column purification. The plasmid pWW20amino acids 252–613) kindly provided by Dr. Winfried Wels iserived from the pUC57 vector and is used for N-terminal fusionf hscFv to a modified deletion mutant of ETA. Plasmids wereurified using plasmid kits from TIANGEN Biotechnology Co. Ltd.Shanghai, China). Restriction fragments or polymerase chain reac-ion products were separated by agarose gel electrophoresis andxtracted using gel extract kits from Watson Biotechnology Co. Ltd.Shanghai, China). Restriction enzymes, T4 DNA ligase, dNTP andrimerStar DNA polymerase were purchased from Takara Biotech-ology Co. Ltd. (Dalian, China). Ni2+-NTA agarose was purchased

rom Qiagen China Co. Ltd. (Shanghai, China). Mouse anti-6× Hisonoclonal antibody, horseradish peroxidase (HRP)-conjugated

oat anti-mouse IgG and FITC-conjugated goat anti-mouse IgGere purchased from ZhongShan GoldenBridge Biotechnology Co.

td. (Beijing, China). Primers were synthesized by Shanghai Invitro-en Biotechnology (Shanghai, China). All other chemical reagentsere of analytical grade.

.2. Construction of rIT-expressing prokaryotic vector

The cDNA encoding the hscFv was amplified from the pUC57-scFv vector by polymerase chain reaction (PCR) using P1 and P2,igested with HindIII and XbaI, and inserted into the HindIII and

baI sites of pWW20 vector [23,24]. It was fused to N-terminalf the ETA’ fragment. The resultant plasmid was designated asWW20-hscFv. Furthermore, the pWW20-hscFv plasmid was iden-ified by restriction enzyme digestion, colony PCR. Then the

P5 5 -ATAAGAATGCGGCCGCCTATTAATATTCGATTGGGCTGGCATC-3

The restriction sites, introduced for cloning, are underlined. P stands for primers.

identified plasmids were amplified by PCR using P3 and P4 tomutate the stop codon between domains II and III, digested withEcoRI and HindIII, and inserted into the EcoRI and HindIII sites ofpWW20-hscFv vector. hscFv-ETA’ fragment was amplified from themutated pWW20-hscFv plasmid by PCR using P3 and P5, digestedwith EcoRI and NotI, and inserted into the EcoRI and NotI sitesof pET32a(+) vector. The clones with an insert orientation preser-ving the direction of transcription from the T7lac promoter wereselected. The resultant plasmids were transfected into E. coli BL21(DE3). The paired primers used for PCR amplification were listed inTable 1.

2.3. Periplasmic expression and purification of the rITs

As described, rITs were expressed under the control of IPTG-inducible tac promotor in the E. coli strain BL21 (DE3). Briefly,bacteria were grown overnight at 37 ◦C in Luria-Bertani (LB) broth(1% polypeptone, 1% NaCl, and 0.5% yeast extract, pH 7.0) sup-plemented with ampicillin (100 �g/ml). The culture was diluted50-fold in 2 L of the same medium, at an OD 600 of 0.4–0.6. Thenit was incubated at 37 ◦C for additional 4 h. Thereafter, immuno-toxin production was induced by the addition of 1.0 mM IPTG at23 ◦C. 6 h later, cells were collected by centrifuging at 16,000 × gfor 5 min and identified by 8% SDS-PAGE and immunoblot.

The protein was purified with Ni2+-NTA column and purity wasexamined by 8% SDS-PAGE with Coomassie blue staining. Thenthe equivalent amounts of each protein were resolved by 8% SDS-PAGE, and then transferred to PVDF membrane and immunoblottedwith the anti-6× His primary mouse antibody at a 1:1000 dilutionand HRP-conjugated goat anti-mouse IgG secondary antibody at a1:5000 dilution. The blot was developed by ECL and imaged by theBio-Rad Gel Imaging System.

2.4. Binding analyses

The binding activity of hscFv-ETA’ was determined by cellmembrane-ELISA (CM-ELISA) using biological active membranes oftumor cells as described by Tur et al. [25]. Human erythroleukemiaK562 cells and the basophilic leukemia KU812 cells, both of whichwere leukemia cells derived from the patients with CML at blastcrisis, were obtained from the American Type Culture Collection(Manassas, VA) and maintained in RPMI 1640 medium supple-mented with 10% fetal calf serum (FCS), 2 mM l-glutamine andantibiotics (100 U/ml penicillin and 100 �g/ml streptomycin) at37 ◦C in a humidified atmosphere of 5% CO2. Blood samples werecollected from CML patients and healthy donors and peripheralblood mononuclear cells were separated by Ficoll gradient centrifu-gation (Tianjin Haoyang Biological manufacture Co. Ltd., China). Thestudy was approved by the Human Ethics Committee of ChongqingMedical University, and informed consent was obtained from allpatients and controls.

96-well plate was coated with 100 �l freshly prepared mem-brane fractions of CML cell lines K562 or KU812 and BaF3 or normalWBC as control (1 × 107 cells) in 0.02 M bicarbonate buffer (pH 9.6)overnight at 4 ◦C. Plates were washed five times with PBS (pH 7.4)

2 ogy Letters 154 (2013) 18– 24

c(fiBabcAwdr4I

2

(svwPaaas

2

mat1wtwaiU

2

hi1wdP

3

3

sftwtPciTn

Fig. 1. The amplification of hscFv fragment by PCR from pUC57-hscFv and insertioninto plasmid pWW20. (a) PCR of hscFv gene was performed, and the amplifica-tion products were analyzed with electrophoresis. M: DL2000 marker; lane 1–3:PCR products of hscFv. (b) Verification of positive clones by double endonucleasedigestion. M: DL2000 marker; lanes 1–3: positive clones digested with HindIII andXbaI.

Fig. 2. Identification of mutating the stop codon of ETA’ by PCR and restrictionenzyme digestion. (a) The stop codon between domains II and III of ETA’ was mutatedby PCR. M: DL2000 marker; lanes 1–6: PCR products of mutated ETA’. (b) Verifica-tion of positive clones by double endonuclease digestion. M: DL2000 marker; lanes1–4: positive clones digested with EcoRI and HindIII.

Fig. 3. Cloning of hscFv-ETA’ fragment into expression plasmid pET32a(+). (a) The

0 X. Zhu et al. / Immunol

ontaining 0.2% (v/v) Tween 20 and blocked with 200 �l 2% BSAw/v) in PBS. After overnight incubation at 4 ◦C, plates were washedve times as before and 0–10 �g/ml hscFv-ETA’ diluted with 0.5%SA (w/v), and 0.05% Tween 20 (v/v) in PBS was added to the platesnd incubated at 37 ◦C for 1 h. Thereafter, plates were washed, andinding of the rIT was detected with the anti-6 × His mouse mono-lonal antibody in 1:1000 dilution and incubated for 1 h at 37 ◦C.fter washing the plate three times, bound protein was detectedith HRP-conjugated goat anti-mouse IgG antibody. The assay waseveloped using an 3,3′,5,5′-Tetramethyl Benzidine dihydrochlo-ide (TMB) solution (KeHua, Shanghai, China) and the absorbance at50 nm was measured by a microplate reader (Molecular Devices,smaning, Germany).

.5. Flow cytometric binding analyses

Cell binding activity of hscFv-ETA’ expressed in E. coli BL21DE3) was also evaluated by flow cytometric analysis. Cells weretained with the affinity purified immunotoxin as described pre-iously [14]. A total of 5 × 105 cells was incubated for 1 h on iceith 50 �l of hscFv-ETA’ fusion protein. The cells were washed with

BS buffer containing 0.2% (w/v) BSA and 0.05% (w/v) sodium azidend then incubated for 30 min with anti-6× His mouse monoclonalntibody. Cells were washed and incubated with FITC-labeled goat-nti mouse IgG for 1 h at 4 ◦C. After a final wash, the cells wereubsequently analyzed by fluorescence-activated cell sorting.

.6. Cell proliferation assay

The cytotoxic effect of hscFv-ETA’ on target cells was assessed byeasuring the conversion of MTT to a water-soluble orange form-

zan dye as described previously [26]. Briefly, various dilutions ofhe rIT were distributed in 50 �l aliquots in 96-well plates. A total of

× 104 target cells in 100 �l complete medium were added to eachell, and the plates were incubated for 48 h at 37 ◦C. The cell cul-

ures were then replaced with 100 �l fresh culture medium whichas supplemented with MTT/phenazine methosulphate, and the

bsorbance at 490 nm were determined with an enzyme-linkedmmunosorbent assay reader (Molecular Devices, Sunnyvale, CA,.S.A.).

.7. Apoptosis analysis

Primary leukemic cells were directly treated with His-taggedscFv-ETA’ fusion protein. Approximately 5 × 105 cells were seeded

n flat-bottomed 12-well plates in RPMI 1640 supplemented with0% FCS. A total of 100 ng/ml immunotoxin was added into eachell, and cells were cultured for 24 h at 37 ◦C. Apoptotic cells wereetected using an annexin V-FITC apoptosis detection kit I (BDharMingen, Heidelberg, Germany) by flow cytometry.

. Results

.1. Construction of the recombinant hscFv-ETA’ immunotoxin

A scheme of the recombinant fusion proteins hscFv-ETA’ con-truction was demonstrated in Fig. S1. Briefly, PCR amplified hscFvragment (Fig. 1), which was 732 bp, was directionally cloned intohe ampicillin-resistant pWW20 with a modified ETA’ and fusedith the ETA’ to form hscFv-ETA’ fusion gene fragment. Then,

he stop codon between domains II and III was mutated usingCR (1115 bp, Fig. 2) and the mutated hscFv-ETA’ (1947 bp) was

loned into the pET32a(+) expression vector containing an IPTG-nducible lac operator and a 6× His tag for easy purification (Fig. 3).he deleted domain Ia of Pseudomonas Exotoxin responsible foronspecific cell binding was thus replaced by CML cell specific

amplification of hscFv-ETA’ by PCR at different annealing temperatures (50–60 ◦C).M: DL10000 marker; lanes 1–10: PCR products of hscFv-ETA’. (b) Verification ofpositive clones by double endonuclease digestion. M: DL5000 marker; lanes 1–6:positive clones digested with EcoRI and NotI.

X. Zhu et al. / Immunology Letters 154 (2013) 18– 24 21

Fig. 4. The expression, purification, and immunoblotting of fusion protein hscFv-ETA’. (a) The expression of hscFv-ETA’ induced with 1.0 mM IPTG at 23 ◦C for 6 h. SDS-PAGEshowed that the target protein with about 88 kDa could be detected in the supernatant and sediment. M: protein marker; lane 1: non-expression of pET32a- hscFv-ETA’; lane2 ing hsd ressioe tant e

hs

3

wfmpt4pisn

3

bfcabbtuhC

3

CdhKei1tc

: supernatant of expressing hscFv-ETA’ fusion protein; lane 3: sediment of expressetected with Coomassie blue staining. M: Protein marker; lane 1: supernatant expxpression by immunoblot. Lane 1: uninduced pET32a-hscFv-ETA’; lane 2: superna

scFv. Successful cloning was verified by DNA sequencing (data nothown).

.2. Expression and purification of fusion protein hscFv-ETA’

After transformation, recombinant E. coli BL21 (DE3) clonesere cultivated under the induction of IPTG. The recombinant

usion protein was secreted partly into the periplasmic compart-ent and the purity of functional hscFv-ETA’ (∼88 kDa with a

redicted isoelectric point of 6.53, predicted by Compute pI/Mwool) directly purified by Ni2+-NTA column was >90% (Fig. 4a andb). At least 0.5 mg of purified hscFv-ETA’ protein was routinelyrepared from 1 L of bacterial shaking cultures. As shown in Fig. 4c,

mmunoblots probed with mouse anti-6× His antibody showedpecific signals of the induced recombinant bacterial. In contrast,o band was detected in control E. coli that had not been induced.

.3. Binding properties of hscFv-ETA’

To investigate whether the fusion protein would specificallyind to CML cells, CM-ELISA and flow cytometry analysis were per-ormed. Coupling of the hscFv coding regions to the truncated ETA’oding sequences did not affect the binding activity of the VH/VLntibody format. As shown in Fig. 5a and b, purified fusion proteinound to CML cell membrane fractions including K562 and KU812ut not to BaF3 or normal WBC membranes. We also observedhat the binding efficiency of hscFv-ETA’ to primary CML cells wasp to 48.97%. These results demonstrated that the fusion proteinscFv-ETA’ exhibited specificity for CML cells, including primaryML cells.

.4. In vitro cytotoxic activity

To characterize the cytotoxic activity of the recombinant anti-ML immunotoxin in vitro, we evaluated the proliferation ofifferent target cells after incubation with different amounts ofscFv-ETA’. Growth inhibition of CML-derived cell lines K562 andU812 were documented by a MTT-based colorimetric assay. Asxpected, hscFv-ETA’ inhibited the growth of K562 and KU812 cells

n a dose-dependent manner (Fig. 6a). The addition of 1, 10, and00 nM hscFv-ETA’ was associated with a dramatic inhibition ofhe growth of CML cells, while little effect was observed in BaF3ells.

cFv-ETA’ fusion protein. (b) Purified fusion protein was subjected to SDS-PAGE andn of fusion protein; lane 2: purified fusion protein. (c) Identification of hscFv-ETA’

xpression of fusion protein; lane 3: purified fusion protein.

3.5. Analysis of apoptosis on CML cells

The effects of hscFv-ETA’ on the induction of apoptosis in afreshly prepared population of CML cells from patients were exam-ined by flow cytometry. Two color flow cytometric analysis usingannexin V-FITC and PI (Fig. 6b) discriminated four populations,viable (bottom left quadrant), early apoptotic (bottom right quad-rant), late apoptotic/necrotic (top right quadrant), and necroticcells (top left quadrant). The normal WBC and BaF3 cells as neg-ative cells treated with hscFv-ETA’ for 24 h remained mostly viable(∼90%). Cell apoptosis rates in CML cell lines K562 and KU812, andprimary patient-derived CML cells treated with the recombinantimmunotoxin were significantly increased. These data indicatedthat hscFv-ETA’ induced apoptosis of CML cells specifically.

4. Discussion

In this study, we reported the construction and functional prop-erties of the first rIT against the CML cells for possible clinicalapplication in patients with CML. To realize the construction ofthe rIT, we fused the humanized anti-CML hscFv to a truncatedPseudomonas exotoxin A. The major findings in our study were: (a)functional hscFv-ETA’ was directly isolated from the periplasmiccompartment of E. coli cultured under the induction of IPTG andadditionally purified by Ni2+-NTA column; (b) in vitro character-istics of the recombinant protein were proved by CM-ELISA, flowcytometry, and colorimetric cell proliferation assay using CML cellmembranes or intact cells; and (c) the recombinant immunotoxinexhibited specific cytotoxic activity toward CML-derived cell linesK562/KU812 and primary CML patient cells.

Drug-resistant and relapse are still tough problems in thetreatment of CML, which have provoked us to pay more atten-tion in creating novel approaches for targeted cancer therapy ofCML. Antibody-based cancer therapeutics, such as immunotox-ins, recently have attracted more attention in specific targetingand killing of residual malignant cells [27]. A significant num-ber of CML cells or other cancer cells survive when the patientsare treated with conventional anticancer agents. Most of the sur-viving cells would probably cause the recurrence of cancer andrest in the G0 phase. To overcome these problems, immunotox-ins can be used as a potential alternative tool [28]. Compared with

conventional chemotherapeutic agents, immunotoxins have sev-eral advantages: (a) immunotoxins have different mechanisms ofaction. Therefore, toxin-based therapies may overcome the natu-rally resistant or acquired resistance to chemotherapeutic agents.

22 X. Zhu et al. / Immunology Letters 154 (2013) 18– 24

norma

(mTm

tattstmti

Fig. 5. Binding properties of hscFv-ETA’ to CML cells or

b) Toxins are potentially cytotoxic for non-dividing cells in a ‘dor-ant’ state which cannot be killed by conventional chemotherapy.

hese unique properties make them attractive for use in the treat-ent of cancer including CML.Immunotoxins contain an antibody-based binding domain for

umor-specific targeting fused to either a plant or bacterial toxins an effector domain. In addition to promising data concerninghe treatment of hematological malignancies [16,29,30], immuno-oxins have been also evaluated for the treatment of advancedolid tumors [31,32]. Additional problems identified in clinical

rials with chemically coupled immunotoxins are the develop-

ent of neutralizing antibodies against both the murine IgG andhe toxic moiety resulting in a limited number of applicationsn 40–60% of the patients [33], and the unspecific cytotoxicity

l cells by CM-ELISA (a) and flow cytometry (b) analysis.

related to unspecific binding of Ricin-A-based toxins to endothelialcells [34]. These problems might, at least in part, be circumventedby using recombinant DNA technology to construct smaller andless immunogenic immunotoxins with reduced unspecific toxi-cities. Recently, it had been reported in the first clinical trials thatrecombinant scFv- or IL-immunotoxin carrying truncated ETA vari-ants show reduced antibody responses in patients [35,36]. Thecommonly used truncated forms of both toxins keep the ADP-ribosylating and translocation domains in the cytosol [37,38]. Tocreate an immunotoxin against CML cells, we fused the hscFv,

which was a humanized scFv against CML cells acquired by CDR-grafting, with a mutated ETA.

Since recombinant DNA technology allows the constructionand possible modification of immunotoxins with relative ease, we

X. Zhu et al. / Immunology Letters 154 (2013) 18– 24 23

(b) w

detpifbdtbmt

Fig. 6. The effect of hscFv-ETA’ on cell proliferation (a) and apoptosis

eveloped a new anti-CML rIT based on the truncated Pseudomonasxotoxin derivative. For this purpose, the hscFv, which consists ofhe variable regions of the native antibody linked by a (G4S)3 linkereptide, was genetically fused to ETA’. After assembly and cloning

nto the expression vector pET32a(+), we isolated the hscFv-ETA’rom periplasmic compartment with a substantial proportion ofiologically active protein. The in vitro potency of hscFv-ETA’escribed in the present study is dramatically reduced compared

o the chemically-linked immunotoxin. This loss of toxicity mighte due to reduced affinity of the scFv compared to the maternalonoclonal antibody and might be improved by affinity matura-

ion. We analyzed the periplasmically expressed, non-glycosylated

as measured by MTT assay and flow cytometry analysis, respectively.

recombinant immunotoxin, which exhibited high binding activ-ity and specific cytotoxic activity against the CML cell linesK562/KU812 in vitro. However, using the standard prokaryoticexpression method, we were not able to reproducibly generate suf-ficient amounts of functional recombinant immunotoxins neededfor in vivo experiments.

In summary, we have shown that CML-derived tumor celllines can be specifically eliminated by a novel recombinant anti-

CML immunotoxin in vitro. Having demonstrated the functionalactivity of hscFv-ETA’, this selective immunotherapeutic com-pound might be used as a potential tool for the treatmentof CML.

2 ogy Le

C

A

GTC

A

f2

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

4 X. Zhu et al. / Immunol

onflict of interest

The authors have declared that no competing interests exist.

cknowledgments

We thank Dr. Winfried Wels and the members of his group ateorg-Speyer-Haus for the pWW20 vector containing the ETA gene.his work was supported by the National Science Foundation ofhina Grant No. 30871102 to Wenli Feng.

ppendix A. Supplementary data

Supplementary data associated with this article can beound, in the online version, at http://dx.doi.org/10.1016/j.imlet.013.08.002.

eferences

[1] Deininger MW, Goldman JM, Melo JV. The molecular biology of chronic myeloidleukemia. Blood 2000;96:3343–56.

[2] Rowley JD. Letter: a new consistent chromosomal abnormality in chronic myel-ogenous leukaemia identified by quinacrine fluorescence and Giemsa staining.Nature 1973;243:290–3.

[3] Hartmann JT, Haap M, Kopp HG, Lipp HP. Tyrosine kinase inhibitors - areview on pharmacology, metabolism and side effects. Curr Drug Metab2009;10:470–81.

[4] Carter P. Improving the efficacy of antibody-based cancer therapies. Nat RevCancer 2001;1:118–29.

[5] Safavy A, Bonner JA, Waksal HW, Buchsbaum DJ, Gillespie GY, Khazaeli MB, et al.Synthesis and biological evaluation of paclitaxel-C225 conjugate as a model fortargeted drug delivery. Bioconjug Chem 2003;14:302–10.

[6] Carter PJ, Senter PD. Antibody-drug conjugates for cancer therapy. Cancer J2008;14:154–69.

[7] Stan AC, Radu DL, Casares S, Bona CA, Brumeanu TD. Antineoplastic effi-cacy of doxorubicin enzymatically assembled on galactose residues of amonoclonal antibody specific for the carcinoembryonic antigen. Cancer Res1999;59:115–21.

[8] Wels W, Harwerth IM, Mueller M, Groner B, Hynes NE. Selective inhibition oftumor cell growth by a recombinant single-chain antibody-toxin specific forthe erbB-2 receptor. Cancer Res 1992;52:6310–7.

[9] Gustin A, Pederson L, Miller R, Chan C, Vickers SM. Application of molecularbiology studies to gene therapy treatment strategies. World J Surg 2002;26:854–60.

10] Bernstein ID. Monoclonal antibodies to the myeloid stem cells: therapeuticimplications of CMA-676, a humanized anti-CD33 antibody calicheamicin con-jugate. Leukemia 2000;14:474–5.

11] Ghetie V, Vitetta ES. Chemical construction of immunotoxins. Mol Biotechnol2001;18:251–68.

12] Kreitman RJ. Immunotoxins. Expert Opin Pharmacother 2000;1:1117–29.

13] Kreitman RJ, Pastan I. Importance of the glutamate residue of KDEL in increasingthe cytotoxicity of Pseudomonas exotoxin derivatives and for increased bindingto the KDEL receptor. Biochem J 1995;307(Pt 1):29–37.

14] Kreitman RJ, Puri RK, Pastan I. Increased antitumor activity of a circularly per-muted interleukin 4-toxin in mice with interleukin 4 receptor-bearing humancarcinoma. Cancer Res 1995;55:3357–63.

15] Engebraaten O, Sivam G, Juell S, Fodstad O. Systemic immunotoxin treatmentinhibits formation of human breast cancer metastasis and tumor growth innude rats. Int J Cancer 2000;88:970–6.

16] Kreitman RJ, Pastan I. Recombinant single-chain immunotoxins against T andB cell leukemias. Leuk Lymphoma 1994;13:1–10.

[

[

tters 154 (2013) 18– 24

17] Brinkmann U, Pai LH, FitzGerald DJ, Willingham M, Pastan I. B3(Fv)-PE38KDEL,a single-chain immunotoxin that causes complete regression of a human car-cinoma in mice. Proc Natl Acad Sci U S A 1991;88:8616–20.

18] Schmidt M, Wels W. Targeted inhibition of tumour cell growth by a bispecificsingle-chain toxin containing an antibody domain and TGF alpha. Br J Cancer1996;74:853–62.

19] Lory S, Strom MS, Johnson K. Expression and secretion of the cloned Pseu-domonas aeruginosa exotoxin A by Escherichia coli. J Bacteriol 1988;170:714–9.

20] Kreitman RJ. Chimeric fusion proteins—Pseudomonas exotoxin-based. CurrOpin Investig Drugs 2001;2:1282–93.

21] Liu D, Wang C, Li C, Zhang X, Zhang B, Mi Z, et al. Production and characterizationof a humanized single-chain antibody against human integrin alphav beta3protein. J Biol Chem 2011;286:24500–7.

22] Chowdhury PS, Viner JL, Beers R, Pastan I. Isolation of a high-affinity stablesingle-chain Fv specific for mesothelin from DNA-immunized mice by phagedisplay and construction of a recombinant immunotoxin with anti-tumor activ-ity. Proc Natl Acad Sci U S A 1998;95:669–74.

23] Wang D, Zhang L, Li Y, Wang H, Xiao Q, Cao W, et al. Construction and expressionof humanized chimeric T cell receptor specific for chronic myeloid leukemiacells. Biotechnol Lett 2012;34:1193–201.

24] Zhu X, Wang D, Li S, Xiao Q, Tao K, Hu J, et al. Expression of a humanized single-chain variable fragment antibody targeting chronic myeloid leukemia cells inEscherichia coli and its characterization. Int J Mol Med 2012;29:939–45.

25] Tur MK, Rothe A, Huhn M, Goerres U, Klimka A, Stocker M, et al. Anovel approach for immunization, screening and characterization of selectedscFv libraries using membrane fractions of tumor cells. Int J Mol Med2003;11:523–7.

26] Wang T, Zhao J, Ren JL, Zhang L, Wen WH, Zhang R, et al. Recombinant immuno-proapoptotic proteins with furin site can translocate and kill HER2-positivecancer cells. Cancer Res 2007;67:11830–9.

27] Huhn M, Sasse S, Tur MK, Matthey B, Schinkothe T, Rybak SM, et al. Humanangiogenin fused to human CD30 ligand (Ang-CD30L) exhibits specific cyto-toxicity against CD30-positive lymphoma. Cancer Res 2001;61:8737–42.

28] Trail PA, King HD, Dubowchik GM. Monoclonal antibody drug immuno-conjugates for targeted treatment of cancer. Cancer Immunol Immunother2003;52:328–37.

29] Barth S, Huhn M, Wels W, Diehl V, Engert A. Construction and in vitro evalu-ation of RFT5(scFv)-ETA’, a new recombinant single-chain immunotoxin withspecific cytotoxicity toward CD25+ Hodgkin-derived cell lines. Int J Mol Med1998;1:249–56.

30] Kreitman RJ, Wilson WH, Bergeron K, Raggio M, Stetler-Stevenson M, FitzGer-ald DJ, et al. Efficacy of the anti-CD22 recombinant immunotoxin BL22 inchemotherapy-resistant hairy-cell leukemia. N Engl J Med 2001;345:241–7.

31] Bruell D, Stocker M, Huhn M, Redding N, Kupper M, Schumacher P, et al.The recombinant anti-EGF receptor immunotoxin 425(scFv)-ETA’ suppressesgrowth of a highly metastatic pancreatic carcinoma cell line. Int J Oncol2003;23:1179–86.

32] Pai LH, Wittes R, Setser A, Willingham MC, Pastan I. Treatment of advancedsolid tumors with immunotoxin LMB-1: an antibody linked to Pseudomonasexotoxin. Nat Med 1996;2:350–3.

33] Grossbard ML, Gribben JG, Freedman AS, Lambert JM, Kinsella J, Rabinowe SN,et al. Adjuvant immunotoxin therapy with anti-B4-blocked ricin after autol-ogous bone marrow transplantation for patients with B-cell non-Hodgkin’slymphoma. Blood 1993;81:2263–71.

34] Baluna R, Rizo J, Gordon BE, Ghetie V, Vitetta ES. Evidence for a structural motifin toxins and interleukin-2 that may be responsible for binding to endothe-lial cells and initiating vascular leak syndrome. Proc Natl Acad Sci U S A1999;96:3957–62.

35] Kreitman RJ, Wilson WH, Robbins D, Margulies I, Stetler-Stevenson M, Wald-mann TA, et al. Responses in refractory hairy cell leukemia to a recombinantimmunotoxin. Blood 1999;94:3340–8.

36] Kawakami K, Kawakami M, Puri RK. Overexpressed cell surface interleukin-4receptor molecules can be successfully targeted for antitumor cytotoxin ther-

apy. Crit Rev Immunol 2001;21:299–310.

37] Mathew M, Verma RS. Humanized immunotoxins: a new generation ofimmunotoxins for targeted cancer therapy. Cancer Sci 2009;100:1359–65.

38] Kreitman RJ. Recombinant immunotoxins containing truncated bacterial toxinsfor the treatment of hematologic malignancies. BioDrugs 2009;23:1–13.