The TNF-related apoptosis-inducing ligand (TRAIL) has been considered as a promising molecule for cancer treatment. However, clinical studies with soluble TRAIL failed to show therapeutic activity, which resulted in subsequent development of more potent TRAIL-based therapeutics. In this study, we applied defined oligomerization and tumor targeting as strategies to further improve the activity of a single-chain version of TRAIL (scTRAIL). We compared three different formats of EGF receptor (EGFR)-targeting dimeric scTRAIL fusion proteins [Diabody (Db)-scTRAIL, scFv-IgE heavy chain domain 2 (EHD2)-scTRAIL, scFv-Fc-scTRAIL] as well as two nontargeted dimeric scTRAIL molecules (EHD2-scTRAIL, Fc-scTRAIL) to reveal the influence of targeting and protein format on antitumor activity. All EGFR-targeted dimeric scTRAIL molecules showed similar binding properties and comparable cell death induction in vitro, exceeding the activity of the respective nontargeted dimeric format and monomeric scTRAIL. Superior properties were observed for the Fc fusion proteins with respect to production and in vivo half-life. In vivo studies using a Colo205 xenograft model revealed potent antitumor activity of all EGFR-targeting formats and Fc-scTRAIL and furthermore highlighted the higher efficacy of fusion proteins comprising an Fc part. Despite enhanced in vitro cell death induction of targeted scTRAIL molecules, however, comparable antitumor activities were found for the EGFR-targeting scFv-Fc-scTRAIL and the nontargeting Fc-scTRAIL in vivo. Mol Cancer Ther; 16(12); 2792–802. ©2017 AACR.

This article is featured in Highlights of This Issue, p. 2639

TNF-related apoptosis-inducing ligand (TRAIL) belongs to the TNF superfamily and, in contrast to other members (e.g., TNF, FasL), selectively induces cell death in tumor cells without harming healthy tissue. Because of this unique property, TRAIL has been extensively investigated as an anticancer therapeutic, showing promising antitumor effects in preclinical models (1, 2). Besides soluble TRAIL (sTRAIL), agonistic antibodies against the two death receptors TRAIL-R1 (DR4) and TRAIL-R2 (DR5) have been studied in clinical trials. Despite a good safety profile, however, clinical results were rather disappointing with antitumor responses seen only for a small number of patients (3). This limited success has been attributed to the observation that many human tumors are resistant toward TRAIL-induced apoptosis (4–6) as well as to an insufficient activity of the analyzed therapeutics. This limited efficacy has meanwhile been linked to the deficiency of sTRAIL and agonistic antibodies in inducing higher order clustering of TRAIL receptors (7–9). While agonistic antibodies have been shown to depend on secondary crosslinking via their Fc part for efficient activity (10–13), soluble TRAIL is only able to activate apoptosis via TRAIL-R1 and requires multivalent display to induce apoptosis via TRAIL-R2 (14, 15). As these studies revealed the necessity of receptor oligomerization for efficient apoptosis induction, several strategies have been developed to improve the activity of TRAIL-based therapeutics (8).

One approach is the fusion of TRAIL to tumor-targeting antibody derivatives. ScFv-TRAIL fusion proteins directed against various tumor-associated antigens (TAA) have been generated and shown to exert considerably higher activity compared with unmodified TRAIL. Because of binding of those molecules to TAAs expressed on the cell surface, they are able to mimic the membrane-bound form of TRAIL and thus to activate TRAIL-R2 (7, 8, 16). TRAIL, however, is a homotrimeric protein, which on the one hand might allow dissociation of the fusion proteins into their monomeric subunits and on the other hand limits the types of possible combinations with fusion partners. Generation of single-chain variants of TRAIL (scTRAIL) by fusion of the extracellular part via short peptide linkers allowed the construction of a completely new set of fusion proteins with improved stability (17–20).

Another strategy to improve TRAIL-based therapeutics was built on the idea of inducing higher order receptor clustering by increasing the valency of the molecules. Various studies showed that increasing the number of TRAIL-R2–targeting antibody fragments or scaffold proteins to an at least tetravalent format strongly improved their activity (21–24). Consistent with these results, fusion of TRAIL to an isoleucine zipper hexamerization motif (25) or integration of two scTRAIL units in one molecule by fusion to dimerization modules considerably increased apoptosis induction (18, 19, 26).

In previous studies, we already showed that combination of both strategies, that is, tumor targeting and an increase of valency, leads to highly active proteins. In a first study, these properties were combined by fusing scTRAIL to the C-terminus of a diabody (Db), which mediates targeting as well as dimeric assembly (19). In another study, a targeted dimeric scTRAIL molecule was generated by fusion of a single-chain variable fragment (scFv) and scTRAIL to the N- and C-terminus of the domain 2 of the heavy chain of IgE (EHD2) as separate dimerization module (18). Both molecules, Db-scTRAIL and scFv-EHD2-scTRAIL, possess two antigen and six TRAIL receptor binding sites and comprised an antibody moiety directed against the EGFR, which is overexpressed on various tumors. The dimeric Db-scTRAIL fusion protein showed considerably better in vitro and in vivo effects not only compared with nontargeted scTRAIL, but also compared with targeted scFv-scTRAIL, possessing only one scTRAIL unit per molecule (19). Similarly, the targeted dimeric scFv-EHD2-scTRAIL fusion protein exerted potent effects, exceeding the activity of the respective nontargeting dimeric EHD2-scTRAIL, which showed better effects than scTRAIL (18).

In this study, we extended the approach of generating targeted dimeric scTRAIL fusion proteins by additionally utilizing a γ1 Fc region as dimerization module, thereby creating molecules with prolonged in vivo half-lives by exploiting FcRn-mediated recycling processes (27). We generated a nontargeted Fc-scTRAIL molecule as well as an EGFR-targeting scFv-Fc-scTRAIL fusion protein and compared their biochemical and functional properties to Db-scTRAIL, scFv-EHD2-scTRAIL and the respective nontargeted molecules (scTRAIL, EHD2-scTRAIL) in vitro and in Colo205 xenograft tumor models.

Materials

Horseradish peroxidase (HRP)- and phycoerythrin (PE)-conjugated anti-FLAG antibodies were purchased from Sigma-Aldrich (A8592) and Miltenyi Biotec (130-101-576), respectively. Mouse TRAIL-R2-Fc was obtained from Sino Biological Inc. (50412-M03H). Bortezomib was purchased from UBPBio (F1200). Clinical-grade bortezomib and cetuximab were kindly provided by Dr. Thomas Mürdter and Dr. Jens Schmid (Dr. Margarete Fischer-Bosch-Institute of Clinical Pharmacology, Stuttgart, Germany). The human colon carcinoma cell lines Colo205 and HCT116 were obtained from ATCC in 2012 without further authentication and tested negative for Mycoplasma (MycoAlert, Lonza) before making master stocks. Cells were cultured in RPMI1640 (Thermo Fisher Scientific, 11875), supplemented with 10% (v/v) FBS (PAN Biotech, 3302-P121707) for a maximum of 2 months. For animal experiments, cells were cultured for 2–3 weeks from freshly thawed master stocks. Anti-FLAG M2 affinity gel was purchased from Sigma-Aldrich (A2220). FLAG peptide was obtained from peptides & elephants (EP01741). BD OptEIA human TRAIL ELISA set was purchased from BD Biosciences (550948), Caspase-Glo 3/7 Assay and Caspase-Glo 8 Assay from Promega (G8091, G8201), and Alanine Transaminase Assay Kit and Amylase Assay Kit from Abcam (ab105134, ab102523). Human plasma was obtained from the blood bank of the Klinikum Stuttgart (Stuttgart, Germany). CD-1 and NMRI nude mice were purchased from Charles River Laboratories.

Production of recombinant proteins and biochemical characterization

All fusion proteins comprised a single-chain version of TRAIL composed of amino acid residues 118–281 of human TRAIL with the TRAIL subunits connected by a single glycine residue as linker (20). scTRAIL without fusion partner was cloned into pIRESpuro, all other fusion proteins were cloned into a modified version of pSecTagA (Invitrogen, Thermo Fisher Scientific, V90020). To remove potential protease cleavage sites of human IgG1 Fc, K447 (EU numbering scheme) was mutated to Q. All proteins were produced in stably transfected HEK293T cells cultivated in Opti-MEM (Thermo Fisher Scientific, 31985–070) containing 50 μmol/L ZnCl2 and purified from the supernatant via FLAG affinity chromatography according to the manufacturer's protocol. Optionally, size exclusion fast protein liquid chromatography was performed as additional purification step applying concentrated protein on a Superdex 200 10/300 GL column (PBS as mobile phase, flow rate of 0.5 mL/minute). Purified proteins were analyzed by SDS-PAGE and by HPLC size exclusion chromatography (SEC) using a Phenomenex Yarra 3 μm SEC-2000 or -3000 column (Phenomenex, 00H-4512-K0 or 00H-4513-K0), a Waters 2695 HPLC, and a mobile phase consisting of 0.1 mol/L Na2HPO4/NaH2PO4, 0.1 mol/L Na2SO4, pH 6.7 at a flow rate of 0.5 mL/minute. The thermal stability of the molecules was analyzed by dynamic light scattering using a ZetaSizer Nano ZS (Malvern) as described previously (18).

ELISA and flow cytometry

Binding of the molecules to receptor-Fc fusion proteins and colon carcinoma cell lines was analyzed by ELISA and flow cytometry as described previously (18). To block EGFR binding, cells were preincubated with a 200-fold molar excess of cetuximab (or PBA as control) for 30 minutes, prior to addition of the proteins to a final concentration of 20 nmol/L (scTRAIL units). Determination of median fluorescence intensities was performed using FlowJo. Relative median fluorescence intensities were calculated using the equation relative MFI = (MFIsample−(MFIdetection system−MFIcells))/MFIcells.

In vitro cytotoxicity

Colo205 (5 × 104/well) or HCT116 (1.5× 104/well) cells were seeded in 100-μL medium and incubated at 37°C, 5% CO2 overnight. After preincubation of the cells with 250 ng/mL (650 nmol/L) bortezomib or medium for 30 minutes, serial dilutions of the purified scTRAIL molecules were added. Cells were incubated for 16 hours at 37°C, 5% CO2, prior to staining of viable cells with crystal violet or by MTT assay. To analyze the effect of EGFR targeting on cell death induction, cells were preincubated with a 200-fold molar excess of cetuximab either in medium alone or medium additionally containing 250 ng/mL bortezomib for 30 minutes as described above. Experiments comparing the bioactivity of Fc-scTRAIL and Fc-scTRAIL8M were performed on Colo205 (5 × 104/well) and HCT116 cells (1 × 104/well) in the absence of bortezomib.

Caspase assay

Colo205 cells (1.5× 104/well) were cultivated for 24 hours at 37°C, 5% CO2. After preincubation with bortezomib (250 ng/mL final concentration) or medium for 30 minutes, cells were treated with protein (100 pmol/L or 1 nmol/L scTRAIL units) for different time periods. Levels of active caspase-8 and -3/7 were detected using Caspase-Glo 8 Assay and Caspase-Glo 3/7 Assay according to the manufacturer's instructions.

Plasma stability

Proteins were diluted in 50% human plasma to a concentration of 100 nmol/L (scTRAIL units). Samples were either directly stored at −20°C (0 days), or incubated at 37°C for 1, 3, and 7 days prior to storage at −20°C. Levels of intact protein were determined in ELISA via binding to human TRAIL-R2-Fc (200 ng/well) and detection with an HRP-coupled anti-FLAG antibody as described above.

Pharmacokinetics

Animal care and all experiments performed were in accordance with Federal and European guidelines and have been approved by university and state authorities. Twenty-five micrograms of protein were injected into the tail vein of female CD-1 mice (8–16 weeks, 25–35 g, 3 mice per molecule) in a total volume of 150 μL. Blood samples were collected 3 minutes, 30 minutes, 1 hour, 2 hours, 6 hours, 1 day, 3 days, and 7 days after injection. Serum concentrations of scTRAIL molecules were determined using BD OptEIA human TRAIL ELISA set according to the manufacturer's protocol. Fc-scTRAIL8M was detected using the capture antibody of BD OptEIA human TRAIL ELISA set and anti-human IgG (Fc specific)-peroxidase (Sigma-Aldrich, A0170). Serum levels of scFv-Fc and cetuximab were determined in ELISA via binding to EGFR-Fc (300 ng/well). Initial and terminal half-lives (t1/2α, t1/2β) and AUC were calculated with Microsoft Excel.

Pharmacodynamics

Animal care and all experiments performed were in accordance with Federal and European guidelines and have been approved by institutional Animal Care and Use Committee and state authorities. A total of 3 × 106 Colo205 cells (in 100 μL PBS) were injected subcutaneously into the left and right flank of female NMRI nude mice (9–11 weeks old, 6 mice per group). Tumor size was monitored by measuring the length (a) and width (b) of the tumors with a caliper, and the tumor volume was calculated by the equation a × b2/2. Treatment was started when tumors reached a volume of approximately 100 mm3. Injections of bortezomib (in 150-μL PBS) were applied intraperitoneally. Proteins (in 150-μL PBS) were injected intravenously. Control mice received respective injections of 150-μL PBS.

ALT and amylase assay

Blood samples taken 4 and 24 hours after the first and last treatment of the first pharmacodynamic experiment were analyzed for potential induction of liver and pancreatic/renal toxicity using alanine transaminase assay kit and amylase assay kit according to the manufacturer's instructions.

Statistical analysis

Except for tumor volumes that are expressed as mean ± 95% CI, all data are represented as mean ± SD of at least three independent experiments. Pairwise and multiple comparisons were performed by unpaired t test (two-tailed) and one-way ANOVA, followed by Tukey post hoc test, respectively (GraphPad Prism, GraphPad Software). P < 0.05 was considered statistically significant (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, P > 0.05). Block shift was only performed to compensate differences in OD values due to varying ELISA developing times and is clearly indicated.

EGFR-targeting and nontargeting dimeric scTRAIL fusion proteins

A single-chain TRAIL moiety consisting of three TRAIL protomers (aa 118–281) linked by single glycine residues was used as building block to create dimeric scTRAIL fusion proteins. The nontargeted dimeric scTRAIL molecules were generated by fusing scTRAIL to the C-terminus of either the heavy chain domain 2 of IgE (EHD2-scTRAIL) or a human IgG1 Fc part (Fc-scTRAIL), resulting in dimeric proteins covalently linked via disulfide bonds in the EHD2 and Fc region. EGFR-targeting molecules were generated via fusion of a humanized scFv version of C225 used in cetuximab (28) to the N-terminus of EHD2-scTRAIL or Fc-scTRAIL, yielding scFv-EHD2-scTRAIL and scFv-Fc-scTRAIL, respectively (Fig. 1). In addition, scTRAIL was fused to the C-terminus of an anti-EGFR diabody that combines targeting and dimerization and results in a noncovalently associated homodimer (Db-scTRAIL). Thus, all dimeric fusion proteins comprise two scTRAIL moieties with a total of six TRAIL receptor–binding sites. Furthermore, all targeted fusion proteins possess two antigen-binding sites for EGFR. The fusion proteins further exhibited an N-terminal FLAG-tag for purification. Proteins were produced in stably transfected HEK293T cells and purified from the supernatant by FLAG affinity chromatography. Yields were in the range of 0.7 to 10.4 mg protein per liter cell culture supernatant with highest protein amounts obtained for the Fc fusion proteins (Supplementary Table S1). In SDS-PAGE analysis, all proteins migrated according to their calculated molecular masses (Supplementary Fig. S1A and S1B; Supplementary Table S1), and all molecules containing the EHD2 or an Fc part showed the expected disulfide-linked dimers under nonreducing conditions. In SEC, all proteins eluted as one major peak (Supplementary Fig. S1C) with Stokes radii of 3.4 nm for scTRAIL and between 5.5 nm and 6.4 nm for the various dimeric fusion proteins (Supplementary Table S1). The elution profile of Db-scTRAIL, however, revealed a minor fraction of smaller size, whereas a peak of high molecular weight species was observed for scTRAIL and EHD2-scTRAIL. Thermal stability was analyzed by dynamic light scattering revealing a melting point of 52°C for scTRAIL, while all other molecules showed increased values (Supplementary Table S1), indicating further stabilization by fusion to an antibody and dimerization moiety.

Figure 1.

Overview of different formats of scTRAIL fusion proteins. Composition (A) and schematic assembly (B) of scTRAIL, EHD2-scTRAIL, Fc-scTRAIL, Db-scTRAIL, scFv-EHD2-scTRAIL, and scFv-Fc-scTRAIL. L, VH (scTRAIL), or Igκ chain leader sequence (EHD2-, Fc-scTRAIL, and all targeted formats). L1, EFGG linker; L2, AAAGGSGG linker; L3, GGSGGASSGG linker; L4, GGGGSGT linker; L5, GGSGGGSSGG linker; L6, GGGGS linker; L7, (GGGGS)3 linker. TRAIL subunits comprise aa 118–281 and are connected by a glycine residue as linker. EHD2 and Fc fusion proteins form covalently linked dimers due to disulfide bridges between the EHD2 domain and the hinge region of the Fc part, respectively.

Figure 1.

Overview of different formats of scTRAIL fusion proteins. Composition (A) and schematic assembly (B) of scTRAIL, EHD2-scTRAIL, Fc-scTRAIL, Db-scTRAIL, scFv-EHD2-scTRAIL, and scFv-Fc-scTRAIL. L, VH (scTRAIL), or Igκ chain leader sequence (EHD2-, Fc-scTRAIL, and all targeted formats). L1, EFGG linker; L2, AAAGGSGG linker; L3, GGSGGASSGG linker; L4, GGGGSGT linker; L5, GGSGGGSSGG linker; L6, GGGGS linker; L7, (GGGGS)3 linker. TRAIL subunits comprise aa 118–281 and are connected by a glycine residue as linker. EHD2 and Fc fusion proteins form covalently linked dimers due to disulfide bridges between the EHD2 domain and the hinge region of the Fc part, respectively.

Close modal

Binding to TRAIL receptors and EGFR

Binding parameters of the scTRAIL fusion proteins were analyzed in ELISA using Fc fusion proteins of the extracellular region of EGFR, human, and mouse TRAIL receptors. scTRAIL and all scTRAIL fusion proteins were able to bind to all human and mouse TRAIL receptors, except for moDcTRAIL-R1, which is consistent with previous reports (ref. 29; Supplementary Fig. S2A). Titrations of the scTRAIL molecules on human and mouse TRAIL-R2 revealed similar binding for all proteins with EC50 values in the low nanomolar range (Supplementary Fig. S2B; Supplementary Table S2). Furthermore, all antibody-targeted scTRAIL fusion proteins exhibited binding to EGFR with EC50 values in the low nanomolar range (Supplementary Fig. S2B; Supplementary Table S2). Cell binding studies were performed by flow cytometry demonstrating concentration-dependent binding to colorectal cancer cell lines Colo205 and HCT116 (Fig. 2A). EC50 values of the targeted scTRAIL fusion proteins were in the low nanomolar range, whereas the dimeric nontargeted molecules did not reach saturation levels for concentrations up to 3 μmol/L (Fig. 2A; Supplementary Table S2). To investigate the contribution of the antibody and scTRAIL moieties to cell surface binding, scTRAIL and the scTRAIL fusion proteins were analyzed at a concentration of 20 nmol/L scTRAIL units either alone or after preincubation with a 200-fold molar excess of cetuximab (Fig. 2B). While cetuximab did not affect binding of the nontargeted scTRAIL molecules, blocking of EGFR binding reduced the levels of bound antibody-scTRAIL fusion proteins to those of the nontargeted formats. This demonstrates that binding of the targeted scTRAIL fusion proteins to Colo205 and HCT116 cells is mediated via both the antibody part and scTRAIL.

Figure 2.

Cell binding of scTRAIL molecules analyzed by flow cytometry. Colo205 and HCT116 cells were incubated with either serial dilutions of the proteins (A) or at a concentration of 20 nmol/L scTRAIL (B). Analysis of 20 nmol/L scTRAIL units was performed after preincubation with medium or a 200-fold molar excess of cetuximab. Bound molecules were detected with a PE-conjugated anti-FLAG antibody. Relative MFI, relative median fluorescence intensity.

Figure 2.

Cell binding of scTRAIL molecules analyzed by flow cytometry. Colo205 and HCT116 cells were incubated with either serial dilutions of the proteins (A) or at a concentration of 20 nmol/L scTRAIL (B). Analysis of 20 nmol/L scTRAIL units was performed after preincubation with medium or a 200-fold molar excess of cetuximab. Bound molecules were detected with a PE-conjugated anti-FLAG antibody. Relative MFI, relative median fluorescence intensity.

Close modal

In vitro cytotoxicity

Cell death induction of the fusion proteins was evaluated using Colo205 and HCT116 cells. Assays were performed in the absence and presence of the proteasome inhibitor bortezomib, which is known to sensitize various tumor cell lines to TRAIL-induced apoptosis (30). Dose-dependent cell death of Colo205 and HCT116 cells was observed for all molecules (Supplementary Fig. S3A and S3B). Without bortezomib, all antibody-scTRAIL fusion proteins efficiently killed both cell lines with EC50 values (corresponding to scTRAIL units) of approximately 12 pmol/L for scFv-EHD2-scTRAIL and scFv-Fc-scTRAIL, while the Db-scTRAIL molecule exhibited higher values of 40.9 pmol/L (Colo205) and 34.9 pmol/L (HCT116; Supplementary Fig. S3A and S3B; Table 1). The nontargeted Fc-scTRAIL showed a lower activity (90.7 pmol/L on Colo205 and 165.7 pmol/L on HCT116), whereas EHD2-scTRAIL induced half-maximal killing already at 34 pmol/L and 75 pmol/L in Colo205 and HCT116 cells, respectively. Noting that the EHD2-scTRAIL preparation contains proteins of higher molecular weight (Supplementary Fig. S1C) that have been demonstrated previously to exert increased activity (31), we fractionated this peak from the dimeric fraction. In vitro experiments with Colo205 cells in the absence or presence of bortezomib showed that the high molecular weight fraction is more active than the dimeric fraction (Supplementary Fig. S4), presumably due to an efficient receptor crosslinking by the multivalent higher molecular weight proteins. The EC50 value of the dimeric fraction was approximately 8-fold higher than the EC50 value of the higher molecular fraction in the absence of bortezomib, and 4-fold higher in the presence of bortezomib, indicating that the increased activity of EHD2-scTRAIL compared with Fc-scTRAIL is linked to the presence of this high molecular weight fraction. Preincubation of Colo205 and HCT116 cells with bortezomib strongly increased the cytotoxic effects of each fusion protein, resulting in 6- to 137-fold lower EC50 values (Supplementary Fig. S3A and S3B; Table 1). Of note, scTRAIL showed the lowest bioactivity (EC50 values of 475 pmol/L on Colo205 and 1,123 pmol/L on HCT116 cells), and reached only approximately 40% to 50% cell killing in the absence of bortezomib (Supplementary Fig. S3A and S3B; Table 1). These results underline the beneficial effects of dimerizing scTRAIL. In addition, a comparison of scFv-EHD2-scTRAIL with EHD2-scTRAIL and scFv-Fc-scTRAIL with Fc-scTRAIL demonstrates the positive effects of targeted delivery to EGFR (3- to 13-fold lower EC50 values), which is in accordance with previous studies (18, 19). To further investigate the contribution of EGFR targeting to cell death induction, all targeted and nontargeted molecules were additionally analyzed after preincubation of Colo205 cells with a 200-fold molar excess of cetuximab. Blocking of EGFR binding with cetuximab increased the EC50 values of the targeted fusion proteins 2- to 6-fold in the absence and 3- to 4-fold in the presence of bortezomib highlighting the beneficial effects of targeting (Supplementary Fig. S5; Supplementary Table S3).

Table 1.

EC50 values of cell death induction

MoleculeColo205HCT116
w/o BZB+ BZBw/o BZB+ BZB
scTRAIL 474.7 ± 139.3 22.4 ± 2.3 1122.7 ± 353.1 8.2 ± 1.5 
Db-scTRAIL 40.9 ± 3.0 3.7 ± 0.4 34.9 ± 2.4 1.7 ± 0.8 
EHD2-scTRAIL 34.5 ± 6.8 2.9 ± 0.4 75.0 ± 13.5 1.3 ± 0.4 
scFv-EHD2-scTRAIL 12.6 ± 2.6 1.8 ± 0.4 11.3 ± 1.7 0.7 ± 0.3 
Fc-scTRAIL 90.7 ± 3.5 7.0 ± 0.2 165.7 ± 25.6 3.6 ± 1.5 
scFv-Fc-scTRAIL 12.8 ± 1.6 2.3 ± 0.5 12.5 ± 4.7 0.5 ± 0.1 
MoleculeColo205HCT116
w/o BZB+ BZBw/o BZB+ BZB
scTRAIL 474.7 ± 139.3 22.4 ± 2.3 1122.7 ± 353.1 8.2 ± 1.5 
Db-scTRAIL 40.9 ± 3.0 3.7 ± 0.4 34.9 ± 2.4 1.7 ± 0.8 
EHD2-scTRAIL 34.5 ± 6.8 2.9 ± 0.4 75.0 ± 13.5 1.3 ± 0.4 
scFv-EHD2-scTRAIL 12.6 ± 2.6 1.8 ± 0.4 11.3 ± 1.7 0.7 ± 0.3 
Fc-scTRAIL 90.7 ± 3.5 7.0 ± 0.2 165.7 ± 25.6 3.6 ± 1.5 
scFv-Fc-scTRAIL 12.8 ± 1.6 2.3 ± 0.5 12.5 ± 4.7 0.5 ± 0.1 

NOTE: Effects on Colo205 and HCT116 cells were analyzed in the absence and presence of bortezomib (250 ng/mL). EC50 values (pmol/L) correspond to concentration of scTRAIL.

Abbreviation: BZB, bortezomib.

Activation of caspases

The impact of targeting and dimeric assembly of scTRAIL on apoptosis induction was further quantified by measuring caspase-8 and -3/7 activity using Caspase-Glo 8 Assay and Caspase-Glo 3/7 Assay. Colo205 cells were incubated with scTRAIL, Fc-scTRAIL, and scFv-Fc-scTRAIL over a period of 20 hours. Bortezomib alone only marginally induced activation of caspase-8 and -3/7 at the analyzed concentration (250 ng/mL) and treatment intervals (Fig. 3). At a concentration of 100 pmol/L scTRAIL units, incubation of the cells with scTRAIL resulted in a gradual activation of caspase-8 and -3/7. Dimeric Fc-scTRAIL showed a faster activation kinetics with a marked caspase activity already after 8 hours. The dimeric EGFR-targeting fusion protein scFv-Fc-scTRAIL showed an even further accelerated activation reaching maximum activity already after 2 hours and inducing considerable higher levels of active caspases after 4–8 hours compared with Fc-scTRAIL and scTRAIL. At a concentration of 1 nmol/L scTRAIL units, all molecules showed a faster activation kinetics for initiator and executioner caspases. Both Fc fusion proteins displayed similar activation profiles reaching comparable levels of active caspase-8 and -3/7, however, slightly delayed for the nontargeted molecule. Despite a faster induction of caspase activity at this higher concentration, scTRAIL was not able to induce activity levels as high as those measured for the dimeric fusion proteins. In the absence of bortezomib, treatment with a concentration of 1 nmol/L scTRAIL units induced caspase-8 and -3/7 activity with a similar activation profile as compared with that measured in the presence of bortezomib. However, slightly lower activity levels were reached, especially for caspase-3/7. Thus, dimeric assembly of scTRAIL increases activation of caspases, which can be further improved by fusion to a targeting moiety.

Figure 3.

Induction of caspase-8 and -3/-7 activity in Colo205 cells. After preincubation with either bortezomib (250 ng/mL) or medium, Colo205 cells were treated with the indicated concentrations of protein for different treatment intervals. Caspase-8 and -3/7 activities were determined using Caspase-Glo 8 Assay and Caspase-Glo 3/7 Assay, respectively. Measured luminescence signal is proportional to the amount of caspase activity. Data are represented as mean ± SD of two independent experiments. RLU, relative light units.

Figure 3.

Induction of caspase-8 and -3/-7 activity in Colo205 cells. After preincubation with either bortezomib (250 ng/mL) or medium, Colo205 cells were treated with the indicated concentrations of protein for different treatment intervals. Caspase-8 and -3/7 activities were determined using Caspase-Glo 8 Assay and Caspase-Glo 3/7 Assay, respectively. Measured luminescence signal is proportional to the amount of caspase activity. Data are represented as mean ± SD of two independent experiments. RLU, relative light units.

Close modal

Plasma stability and pharmacokinetics

We first analyzed the stability of the proteins in human plasma at 37°C. Integrity of the fusion proteins was determined by ELISA via TRAIL-R2 binding (mediated by the C-terminal scTRAIL unit) and detection via the N-terminal FLAG-tag. All proteins showed a time-dependent decrease in the detected level of bioactive protein with at least 50% of the protein remaining intact after 7 days (Fig. 4A). Highest stability was observed for scTRAIL and the Fc-scTRAIL fusion protein with 10%–20% reduced binding levels after 7 days.

Figure 4.

Plasma stability and in vivo pharmacokinetics of scTRAIL molecules. A, Stability of the proteins (100 nmol/L scTRAIL) in 50% human plasma was determined at 37°C. Levels of intact protein were measured in ELISA via binding to human TRAIL-R2-Fc and detection with an anti-FLAG-HRP. Data were normalized to the concentration of the sample directly frozen. B, Pharmacokinetic profiles were determined in female CD-1 mice (3 mice per molecule). Twenty-five micrograms of protein were injected intravenously. Concentrations of serum samples collected after indicated time intervals were determined by ELISA.

Figure 4.

Plasma stability and in vivo pharmacokinetics of scTRAIL molecules. A, Stability of the proteins (100 nmol/L scTRAIL) in 50% human plasma was determined at 37°C. Levels of intact protein were measured in ELISA via binding to human TRAIL-R2-Fc and detection with an anti-FLAG-HRP. Data were normalized to the concentration of the sample directly frozen. B, Pharmacokinetic profiles were determined in female CD-1 mice (3 mice per molecule). Twenty-five micrograms of protein were injected intravenously. Concentrations of serum samples collected after indicated time intervals were determined by ELISA.

Close modal

To determine the in vivo half-lives, CD-1 mice received an intravenous injection of 25 μg protein and serum levels of the molecules were measured in ELISA. scTRAIL and Db-scTRAIL showed short terminal half-lives of 2.2 and 2.4 hours, respectively (Fig. 4B; Supplementary Table S4). ScFv-EHD2-scTRAIL and the corresponding nontargeted fusion protein exhibited significantly extended half-lives of 6.5 and 8.6 hours (P < 0.001) compared with scTRAIL and Db-scTRAIL. Best pharmacokinetic properties were observed for the Fc fusion proteins exhibiting terminal half-lives of 17.4 (scFv-Fc-scTRAIL) and 14.5 hours (Fc-scTRAIL; P < 0.01). Compared with the corresponding scFv-Fc and IgG (cetuximab), however, the Fc-comprising scTRAIL fusion proteins exhibited 6- to 15-fold shorter half-lives (P < 0.001) indicating a role of TRAIL-mediated processes in the elimination of these molecules. To investigate a potential contribution of TRAIL receptor binding in the elimination of the scTRAIL fusion proteins, Fc-scTRAIL8M was generated on the basis of a mutated version of scTRAIL carrying eight mutations (Supplementary Fig. S6A) described to interfere with TRAIL receptor binding and bioactivity (32). In fact, binding of Fc-scTRAIL8M to human and mouse TRAIL receptors was only detected for high concentrations, consequently leading to a considerably reduced capability to induce cell death (Supplementary Fig. S6). Although the terminal half-life of Fc-scTRAIL8M (Fig. 4B; Supplementary Table S4) was with 21.6 hours significantly prolonged compared with Fc-scTRAIL (P < 0.05), it remained significantly shorter than that of scFv-Fc (P < 0.01) suggesting a role of additional processes in the elimination of scTRAIL fusion proteins.

Antitumor activity

The antitumor activity of the scTRAIL fusion proteins was analyzed in Colo205 xenograft models. Nude mice with established subcutaneous tumors (approximately 100 mm3) received six intraperitoneal injections of bortezomib (5 μg) in combination with six intravenous applications of either Db-scTRAIL, scFv-EHD2-scTRAIL, scFv-Fc-scTRAIL, Fc-scTRAIL (0.5 nmol corresponding to 1 nmol scTRAIL units), or PBS as control. Treatment with scTRAIL was omitted because in previous studies we have already shown that at the doses tested, scTRAIL in combination with bortezomib is ineffective (18, 19). After three treatments every other day, application intervals were gradually increased by one day. Treatment with all proteins induced complete tumor regression, which remained stable until end of the study (day 115, Fig. 5A and B). The serum concentrations of the fusion proteins evaluated 3 minutes, 4 hours, 24 hours, and 48 hours after the first treatment were comparable with those after the last treatment and reflect the pharmacokinetic profiles determined in nontumor-bearing CD-1 mice (Fig. 5C). Despite the significantly increased drug exposure of the Fc molecules compared with the Db and EHD2 fusion proteins (P < 0.05; Supplementary Table S5), no differences in antitumoral activity were observed. Safety of treatment was analyzed by determining serum alanine transaminase (ALT) and amylase levels as a measure for liver and pancreatic/renal toxicity, respectively. Enzyme levels were measured 4 and 24 hours after the first and the last treatment and compared with serum samples of untreated tumor-bearing mice. ALT and amylase levels of all treated groups did not show a significant increase (P > 0.05) compared with the untreated controls (Supplementary Fig. S7), indicating the absence of toxic effects.

Figure 5.

In vivo activity and pharmacokinetics of different EGFR-targeting scTRAIL formats and Fc-scTRAIL in NMRI nude mice (6 mice/group) with established Colo205 tumors. A, Mice were treated with bortezomib (5 μg; i.p.) and protein (0.5 nmol; i.v.). Dotted lines indicate days of treatment (days 13, 15, 17, 20, 24, 29). B, Statistical analysis of tumor volumes at day 38 (numbering of groups I to V according to A) was performed by one-way ANOVA, followed by Tukey post hoc test. C, Concentrations of scTRAIL fusion proteins in serum samples taken 3 minutes, 4 hours, 24 hours, and 48 hours after the first and last treatment were determined by ELISA. D, Mice received six intravenous treatments with 0.2 nmol protein or PBS as control. Dotted lines indicate the days of treatment (days 14, 18, 21, 25, 28, 32). E, Statistical analysis of tumor volumes of groups I to IV (numbering of groups according to D) at day 47 was done by one-way ANOVA, followed by Tukey post hoc test. F, Mice received combinatorial treatments of 5 μg bortezomib (BZB; i.p.) and either 0.3 nmol (diamonds) or 0.1 nmol (triangles) protein (i.v.) twice a week for three weeks (days 14, 18, 21, 25, 28, 32). Days of treatment are indicated with dotted lines. G, Tumor volumes of groups II to V at day 67 are represented (numbering of groups according to F). Statistical analysis was performed by one-way ANOVA, followed by Tukey post hoc test.

Figure 5.

In vivo activity and pharmacokinetics of different EGFR-targeting scTRAIL formats and Fc-scTRAIL in NMRI nude mice (6 mice/group) with established Colo205 tumors. A, Mice were treated with bortezomib (5 μg; i.p.) and protein (0.5 nmol; i.v.). Dotted lines indicate days of treatment (days 13, 15, 17, 20, 24, 29). B, Statistical analysis of tumor volumes at day 38 (numbering of groups I to V according to A) was performed by one-way ANOVA, followed by Tukey post hoc test. C, Concentrations of scTRAIL fusion proteins in serum samples taken 3 minutes, 4 hours, 24 hours, and 48 hours after the first and last treatment were determined by ELISA. D, Mice received six intravenous treatments with 0.2 nmol protein or PBS as control. Dotted lines indicate the days of treatment (days 14, 18, 21, 25, 28, 32). E, Statistical analysis of tumor volumes of groups I to IV (numbering of groups according to D) at day 47 was done by one-way ANOVA, followed by Tukey post hoc test. F, Mice received combinatorial treatments of 5 μg bortezomib (BZB; i.p.) and either 0.3 nmol (diamonds) or 0.1 nmol (triangles) protein (i.v.) twice a week for three weeks (days 14, 18, 21, 25, 28, 32). Days of treatment are indicated with dotted lines. G, Tumor volumes of groups II to V at day 67 are represented (numbering of groups according to F). Statistical analysis was performed by one-way ANOVA, followed by Tukey post hoc test.

Close modal

In a second experiment, antitumor activity of the different formats was analyzed in the absence of bortezomib with injections of 0.2 nmol protein (corresponding to 0.4 nmol scTRAIL units) twice weekly for 3 weeks (Fig. 5D). Again, strong antitumor effects were observed for the Fc fusion proteins (Fc-scTRAIL, scFv-Fc-scTRAIL), while treatment with Db-scTRAIL and scFv-EHD2-scTRAIL resulted only in a stabilization of tumor volume during treatment and regrowth after end of treatment. Compared with the PBS-treated control, mice of all protein-treated groups showed a significantly reduced tumor volume at day 47, and a significant difference was observed between Db-scTRAIL or scFv-EHD2-scTRAIL and Fc-scTRAIL therapy (Fig. 5E).

In a further experiment, Fc-scTRAIL and scFv-Fc-scTRAIL were injected at two different doses (0.1 and 0.3 nmol corresponding to 0.2 and 0.6 nmol scTRAIL units) in the presence of bortezomib twice weekly for 3 weeks. Here, both proteins showed similar effects, with rapid and complete tumor regression at the higher dose and a reduced activity at the lower dose (Fig. 5F and G).

The bioactivity of TRAIL-based therapeutics can be improved by increasing the valency for their death receptors. Especially TRAIL receptor 2 requires higher order clustering for potent activation of apoptotic signaling (14, 15). In previous studies, we employed two strategies, a diabody-mediated dimerization approach and the IgE-derived heavy chain domain 2 (EHD2) in combination with scTRAIL to generate hexavalent TRAIL molecules (18, 19). While the diabody combines dimerization and targeting in one unit, the EHD2 represents a dimerization module of small size to which targeting moieties, such as scFv fragments, can be separately fused.

Here, we have extended the approach and compared the diabody and EHD2-derived hexavalent TRAIL fusion proteins with those based on the Fc region as dimerization module. In vitro, similar high bioactivities were observed for these different hexavalent scTRAIL fusion proteins, demonstrating that the dimerization modules and linkers incorporated in this study provide sufficient flexibility to allow efficient death receptor clustering. In accordance with results from earlier studies (19), the assembly of two scTRAIL moieties into a hexavalent molecule considerably increased activation of caspases and subsequent cell death compared with trivalent scTRAIL. Of note, recent data highlighted the importance not only of increased valency, but also of an appropriate spatial organization to allow multivalent receptor binding (33). Our results demonstrate that employing a single-chain version of the natural homotrimeric TRAIL ligand in the context of immunoglobulin-derived dimerization modules enables potent death receptor activation.

Targeting to tumor cell surface markers has been already shown to improve activity of TRAIL-based therapeutics (18, 19). We chose the EGFR as a clinically relevant target and used a humanized version of cetuximab in an scFv format. To scrutinize the relevance of targeting for TRAIL-mediated apoptosis, we have directly compared the EHD2- and the Fc-based hexavalent formats in the nontargeted and the EGFR-targeted version and confirmed in vitro data of earlier studies (18, 19), showing superior activity for the targeted fusion proteins, evident from lower EC50 values and a right shift in the dose–response curves in the presence of competing antibody cetuximab. Besides targeting scTRAIL to EGFR-expressing tumor cells, these fusion proteins have potential dual action through additional blockade of EGFR function. However, for cancers with activating mutations in EGFR downstream signaling molecules, cetuximab or antibody fragments derived thereof fail to block EGFR-induced proliferative signaling (34). This is the case for the cancer cell lines Colo205 and HCT116 used here, whereas in Caco-2 cells, which exhibit a wild-type status for MAPK and PI3K pathways, an inhibition of basal and ligand-induced proliferation has been shown for EGFR-targeting Db-scTRAIL molecules, confirming dual action of this targeted scTRAIL molecule (35). Therefore, depending on the status of the tumor, targeted scTRAIL fusion proteins comprising blocking antibody fragments potentially broaden the mode of action by exerting antitumoral activity also on a priori or secondary TRAIL-resistant tumors.

In vivo studies using the Colo205 xenograft model confirmed the potent antitumoral activity of all EGFR-targeting dimeric scTRAIL formats as well as of nontargeted dimeric Fc-scTRAIL. Previous reports of half-life extended ABD-TRAIL and HSA-TNC-TRAIL fusion proteins already demonstrated a translation of improved pharmacokinetic properties into higher efficacy (36, 37). In accordance with these data, the longer-circulating Fc fusion proteins induced stronger in vivo antitumor effects than rather short-lived Db-scTRAIL and scFv-EHD2-scTRAIL. However, contrasting the in vitro data, scFv-Fc-scTRAIL and Fc-scTRAIL exerted similar antitumor activity in vivo. These results differ from earlier studies of scFv-EHD2-scTRAIL and EHD2-scTRAIL fusion proteins, which showed significantly better effects for the EGFR-targeted molecule in vivo (18). Compared with EHD2-scTRAIL variants, the superior, yet indiscriminate action of the nontargeted and the tumor-targeted Fc-scTRAIL molecules could be related to their superior pharmacokinetic properties. In support of this, analysis of serum concentrations after the first and last injection did not reveal differences in serum levels between Fc-scTRAIL and scFv-Fc-scTRAIL. Furthermore, in the tumor model studied, the rather low expression of EGFR in Colo205 cells might be insufficient to reveal a potentially enhancing effect of tumor targeting. Therefore, our data do not rule out the relevance of targeting, for example, in tumor models with a more prominent expression of the target antigen or a wild-type status for MAPK and PI3K pathways. Irrespectively, our data highlight the importance of the pharmacokinetic properties of the TRAIL-based therapeutics and thus favor fusion proteins comprising the Fc region. Further studies in different tumor models will be required to determine the relative contribution of pharmacokinetic and targeting to the powerful in vivo activity of these hexavalent scTRAIL fusion proteins.

The rationale of creating scTRAIL fusion proteins of different formats was to generate molecules of different size, flexibility, and with respect to the Fc region, different pharmacokinetic properties. Indeed, as discussed above, significantly prolonged terminal half-lives were observed for the Fc fusion proteins. However, these half-lives (in the range of 14–17 hours) are significantly shorter than that of the respective scFv-Fc and also considerable shorter than those described for other Fc fusion proteins (38). Similarly, conjugation or fusion of TRAIL to HSA (37, 39) did not prolong the half-life to an extent as expected from data of other HSA fusion proteins reported previously (40). In line with these data, an albumin-binding domain was not capable of improving the pharmacokinetic properties of TRAIL (36) as effectively as demonstrated for example for fusion to a scDb (41). Several reasons might underlie these short in vivo half-lives, including reduced protein stability, hampered FcRn-mediated recycling, as well as target-mediated and nonspecific clearance. As investigation of protein stability in human plasma at 37°C revealed persistence of greater 50% intact protein after seven days, decreased stability of the Fc region due to fusion to scTRAIL is unlikely the limiting factor. Although fusion to scTRAIL could potentially interfere with FcRn-mediated recycling and thereby affect half-life, this appears unlikely in light of a similarly reduced terminal half-life of an scFv-EHD2-scTRAIL fusion protein compared with scFv-EHD2 (18), which are not recycled via the FcRn. This argues for other processes controlling pharmacokinetic properties. Thus, to reveal a potential role of target-mediated processes in the elimination of scTRAIL fusion proteins, we generated an Fc-scTRAIL fusion protein exhibiting a strongly reduced binding to human and mouse TRAIL receptors (Fc-scTRAIL8M). Despite an approximately 50% increased terminal half-life of Fc-scTRAIL8M compared with Fc-scTRAIL (21.6 vs. 14.5 hours), overall drug exposure of Fc-scTRAIL8M was reduced (due to a decreased initial half-life) and the terminal half-life remained significantly shorter compared with scFv-Fc. On the basis of these data, TRAIL receptors appear to contribute to some extent to clearance, but on their own cannot explain the shorter half-life of these scTRAIL fusion proteins. Recently, it was shown that nonspecific binding, for example, of positively charged patches within the protein to negatively charged structures, can affect clearance of antibodies (42, 43). Such a mechanism could play a role for the half-life of scTRAIL fusion proteins, too.

In conclusion, we confirmed superior in vitro activity of EGFR-targeted dimeric scTRAIL fusion proteins compared with scTRAIL and dimeric nontargeted formats. In vivo studies identified improved half-lives and higher efficacy of hexavalent scTRAIL fusion proteins comprising an Fc part. Furthermore, depending on the tumor studied, nontargeted Fc-scTRAIL fusion proteins may induce equally effective and complete tumor regressions as their targeted counterparts, which provides a therapeutic rationale to use these molecules when no prominent tumor marker is identified.

M. Hutt has ownership interest in a patent application. O. Seifert has ownership interest in a patent application in University of Stuttgart. M. Siegemund has ownership interest in a patent application. K. Pfizenmaier has ownership interest in a patent application. R.E. Kontermann reports receiving a commercial research grant from BMBF eBio Predict (0316186A) and BMBF eMed Melanoma Sensitivity, has ownership interest in a patent application, and is a consultant/advisory board member for Biotech companies. No potential conflicts of interest were disclosed by the other authors.

Conception and design: M. Hutt, K. Pfizenmaier, R.E. Kontermann

Development of methodology: M. Hutt

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M. Hutt, L. Marquardt, O. Seifert, M. Siegemund

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M. Hutt, O. Seifert, M. Siegemund, K. Pfizenmaier, R.E. Kontermann

Writing, review, and/or revision of the manuscript: M. Hutt, O. Seifert, M. Siegemund, I. Müller, D. Kulms, K. Pfizenmaier, R.E. Kontermann

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M. Hutt, O. Seifert, D. Kulms

Study supervision: R.E. Kontermann

We would like to thank Dr. Thomas Mürdter and Dr. Jens Schmid (Dr. Margarete Fischer-Bosch-Institute of Clinical Pharmacology, Stuttgart, Germany) for kindly providing clinical grade bortezomib and cetuximab. We thank Nadine Heidel and Doris Göttsch for technical assistance.

This work was supported by grants BMBF eBio Predict: 0316186A (to K. Pfizenmaier), BMBF eMed Melanoma sensitivity 031A423D, BMBF eBio Predict: 0316186A (to R.E. Kontermann), and BMBF eMed Melanoma sensitivity 031A423A (to D. Kulms).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Ashkenazi
A
,
Pai
RC
,
Fong
S
,
Leung
S
,
Lawrence
DA
,
Marsters
SA
, et al
Safety and antitumor activity of recombinant soluble Apo2 ligand
.
J Clin Invest
1999
;
104
:
155
62
.
2.
Walczak
H
,
Miller
RE
,
Ariail
K
,
Gliniak
B
,
Griffith
TS
,
Kubin
M
, et al
Tumoricidal activity of tumor necrosis factor-related apoptosis-inducing ligand in vivo
.
Nat Med
1999
;
5
:
157
63
.
3.
den Hollander
MW
,
Gietema
JA
,
de Jong
S
,
Walenkamp
AM
,
Reyners
AK
,
Oldenhuis
CN
, et al
Translating TRAIL-receptor targeting agents to the clinic
.
Cancer Lett
2013
;
332
:
194
201
.
4.
Dimberg
LY
,
Anderson
CK
,
Camidge
R
,
Behbakht
K
,
Thorburn
A
,
Ford
HL
. 
On the TRAIL to successful cancer therapy? Predicting and counteracting resistance against TRAIL-based therapeutics
.
Oncogene
2013
;
32
:
1341
50
.
5.
Thorburn
A
,
Behbakht
K
,
Ford
H
. 
TRAIL receptor-targeted therapeutics: resistance mechanisms and strategies to avoid them
.
Drug Resist Updat
2008
;
11
:
17
24
.
6.
Zhang
L
,
Fang
B
. 
Mechanisms of resistance to TRAIL-induced apoptosis in cancer
.
Cancer Gene Ther
2005
;
12
:
228
37
.
7.
de Bruyn
M
,
Bremer
E
,
Helfrich
W
. 
Antibody-based fusion proteins to target death receptors in cancer
.
Cancer Lett
2013
;
332
:
175
83
.
8.
De Miguel
D
,
Lemke
J
,
Anel
A
,
Walczak
H
,
Martinez-Lostao
L
. 
Onto better TRAILs for cancer treatment
.
Cell Death Differ
2016
;
4
:
1
15
.
9.
Wajant
H
,
Gerspach
J
,
Pfizenmaier
K
. 
Engineering death receptor ligands for cancer therapy
.
Cancer Lett
2013
;
332
:
163
74
.
10.
Haynes
NM
,
Hawkins
ED
,
Li
M
,
McLaughlin
NM
,
Hämmerling
GJ
,
Schwendener
R
, et al
CD11c+ dendritic cells and B cells contribute to the tumoricidal activity of anti-DR5 antibody therapy in established tumors
.
J Immunol
2010
;
185
:
532
41
.
11.
Li
F
,
Ravetch
JV
. 
Apoptotic and antitumor activity of death receptor antibodies require inhibitory Fcγ receptor engagement
.
Proc Natl Acad Sci U S A
2012
;
109
:
10966
71
.
12.
Li
F
,
Ravetch
JV
. 
Antitumor activities of agonistic anti-TNFR antibodies require differential FcγRIIB coengagement in vivo
.
Proc Natl Acad Sci
2013
;
110
:
19501
6
.
13.
Wilson
NS
,
Yang
B
,
Yang
A
,
Loeser
S
,
Marsters
S
,
Lawrence
D
, et al
An Fcγ receptor-dependent mechanism drives antibody-mediated target-receptor signaling in cancer cells
.
Cancer Cell
2011
;
19
:
101
13
.
14.
Mühlenbeck
F
,
Schneider
P
,
Bodmer
JL
,
Schwenzer
R
,
Hauser
A
,
Schubert
G
, et al
The tumor necrosis factor-related apoptosis-inducing ligand receptors TRAIL-R1 and TRAIL-R2 have distinct cross-linking requirements for initiation of apoptosis and are non-redundant in JNK activation
.
J Biol Chem
2000
;
275
:
32208
13
.
15.
Wajant
H
,
Moosmayer
D
,
Wüest
T
,
Bartke
T
,
Gerlach
E
,
Schönherr
U
, et al
Differential activation of TRAIL-R1 and -2 by soluble and membrane TRAIL allows selective surface antigen-directed activation of TRAIL-R2 by a soluble TRAIL derivative
.
Oncogene
2001
;
20
:
4101
6
.
16.
Holland
PM
. 
Death receptor agonist therapies for cancer, which is the right TRAIL?
Cytokine Growth Factor Rev
2014
;
25
:
185
93
.
17.
Schneider
B
,
Münkel
S
,
Krippner-Heidenreich
A
,
Grunwald
I
,
Wels
WS
,
Wajant
H
, et al
Potent antitumoral activity of TRAIL through generation of tumor-targeted single-chain fusion proteins
.
Cell Death Dis
2010
;
1
:
e68
.
18.
Seifert
O
,
Plappert
A
,
Fellermeier
S
,
Siegemund
M
,
Pfizenmaier
K
,
Kontermann
RE
. 
Tetravalent antibody-scTRAIL fusion proteins with improved properties
.
Mol Cancer Ther
2014
;
13
:
101
11
.
19.
Siegemund
M
,
Pollak
N
,
Seifert
O
,
Wahl
K
,
Hanak
K
,
Vogel
A
, et al
Superior antitumoral activity of dimerized targeted single-chain TRAIL fusion proteins under retention of tumor selectivity
.
Cell Death Dis
2012
;
3
:
e295
.
20.
Siegemund
M
,
Seifert
O
,
Zarani
M
,
Džinić
T
,
De Leo
V
,
Göttsch
D
, et al
An optimized antibody-single-chain TRAIL fusion protein for cancer therapy
.
MAbs
2016
;
8
:
879
91
.
21.
Huet
HA
,
Growney
JD
,
Johnson
JA
,
Li
J
,
Bilic
S
,
Ostrom
L
, et al
Multivalent nanobodies targeting death receptor 5 elicit superior tumor cell killing through efficient caspase induction
.
MAbs
2014
;
6
:
1560
70
.
22.
Liu
F
,
Si
Y
,
Liu
G
,
Li
S
,
Zhang
J
,
Ma
Y
. 
The tetravalent anti-DR5 antibody without cross-linking direct induces apoptosis of cancer cells
.
Biomed Pharmacother
2015
;
70
:
41
5
.
23.
Swers
JS
,
Grinberg
L
,
Wang
L
,
Feng
H
,
Lekstrom
K
,
Carrasco
R
, et al
Multivalent scaffold proteins as superagonists of TRAIL receptor 2-induced apoptosis
.
Mol Cancer Ther
2013
;
12
:
1235
44
.
24.
Wang
W
,
He
W
,
Wang
L
,
Zhang
G
,
Gao
B
. 
Pentamerisation of a scFv directed against TRAIL receptor 2 increases its antitumour efficacy
.
Immunol Cell Biol
2013
;
91
:
360
7
.
25.
Han
JH
,
Moon
AR
,
Chang
JH
,
Bae
J
,
Choi
JM
,
Lee
SH
, et al
Potentiation of TRAIL killing activity by multimerization through isoleucine zipper hexamerization motif
.
BMB Rep
2016
;
49
:
282
7
.
26.
Gieffers
C
,
Kluge
M
,
Merz
C
,
Sykora
J
,
Thiemann
M
,
Schaal
R
, et al
APG350 induces superior clustering of TRAIL receptors and shows therapeutic antitumor efficacy independent of cross-linking via Fcγ receptors
.
Mol Cancer Ther
2013
;
12
:
2735
47
.
27.
Kontermann
RE
. 
Strategies to extend plasma half-lives of recombinant antibodies
.
BioDrugs
2009
;
23
:
93
109
.
28.
Seifert
O
,
Plappert
A
,
Heidel
N
,
Fellermeier
S
,
Messerschmidt
SK
,
Richter
F
, et al
The IgM CH2 domain as covalently linked homodimerization module for the generation of fusion proteins with dual specificity
.
Protein Eng Des Sel
2012
;
25
:
603
12
.
29.
Bossen
C
,
Ingold
K
,
Tardivel
A
,
Bodmer
JL
,
Gaide
O
,
Hertig
S
, et al
Interactions of tumor necrosis factor (TNF) and TNF receptor family members in the mouse and human
.
J Biol Chem
2006
;
281
:
13964
71
.
30.
de Wilt
LH
,
Kroon
J
,
Jansen
G
,
de Jong
S
,
Peters
GJ
,
Kruyt
FA
. 
Bortezomib and TRAIL: a perfect match for apoptotic elimination of tumour cells?
Crit Rev Oncol Hematol
2013
;
85
:
363
72
.
31.
Koschny
R
,
Walczak
H
,
Ganten
TM
. 
The promise of TRAIL–potential and risks of a novel anticancer therapy
.
J Mol Med
2007
;
85
:
923
35
.
32.
Hymowitz
SG
,
Connell
MP
,
Ultsch
MH
,
Hurst
A
,
Totpal
K
,
Ashkenazi
A
, et al
A unique zinc-binding site revealed by a high-resolution X-ray structure of Apo2L/TRAIL
.
Biochemistry
2000
;
39
:
633
40
.
33.
Beyrath
J
,
Chekkat
N
,
Smulski
CR
,
Lombardo
CM
,
Lechner
M
,
Seguin
C
, et al
Synthetic ligands of death receptor 5 display a cell-selective agonistic effect at different oligomerization levels
.
Oncotarget
2016
;
7
:
64942
56
.
34.
Lièvre
A
,
Bachet
J-B
,
Le Corre
D
,
Boige
V
,
Landi
B
,
Emile
JF
, et al
KRAS mutation status is predictive of response to cetuximab therapy in colorectal cancer
.
Cancer Res
2006
;
66
:
3992
5
.
35.
Möller
Y
,
Siegemund
M
,
Beyes
S
,
Herr
R
,
Lecis
D
,
Delia
D
, et al
EGFR-targeted TRAIL and a Smac mimetic synergize to overcome apoptosis resistance in KRAS mutant colorectal cancer cells
.
PLoS One
2014
;
9
:
e107165
.
36.
Li
R
,
Yang
H
,
Jia
D
,
Nie
Q
,
Cai
H
,
Fan
Q
, et al
Fusion to an albumin-binding domain with a high affinity for albumin extends the circulatory half-life and enhances the in vivo antitumor effects of human TRAIL
.
J Control Release
2016
;
228
:
96
106
.
37.
Müller
N
,
Schneider
B
,
Pfizenmaier
K
,
Wajant
H
. 
Superior serum half life of albumin tagged TNF ligands
.
Biochem Biophys Res Commun
2010
;
396
:
793
9
.
38.
Unverdorben
F
,
Richter
F
,
Hutt
M
,
Seifert
O
,
Malinge
P
,
Fischer
N
, et al
Pharmacokinetic properties of IgG and various Fc fusion proteins in mice
.
MAbs
2016
;
8
:
120
8
.
39.
Byeon
HJ
,
Min
SY
,
Kim
I
,
Lee
ES
,
Oh
KT
,
Shin
BS
, et al
Human serum albumin-TRAIL conjugate for the treatment of rheumatoid arthritis
.
Bioconjug Chem
2014
;
25
:
2212
21
.
40.
Müller
D
,
Karle
A
,
Meißburger
B
,
Höfig
I
,
Stork
R
,
Kontermann
RE
. 
Improved pharmacokinetics of recombinant bispecific antibody molecules by fusion to human serum albumin
.
J Biol Chem
2007
;
282
:
12650
60
.
41.
Stork
R
,
Müller
D
,
Kontermann
RE
. 
A novel tri-functional antibody fusion protein with improved pharmacokinetic properties generated by fusing a bispecific single-chain diabody with an albumin-binding domain from streptococcal protein G
.
Protein Eng Des Sel
2007
;
20
:
569
76
.
42.
Datta-Mannan
A
,
Thangaraju
A
,
Leung
D
,
Tang
Y
,
Witcher
DR
,
Lu
J
, et al
Balancing charge in the complementarity-determining regions of humanized mAbs without affecting pI reduces non-specific binding and improves the pharmacokinetics
.
MAbs
2015
;
7
:
483
93
.
43.
Datta-Mannan
A
,
Lu
J
,
Witcher
DR
,
Leung
D
,
Tang
Y
,
Wroblewski
VJ
. 
The interplay of non-specific binding, targetmediated clearance and FcRn interactions on the pharmacokinetics of humanized antibodies
.
MAbs
2015
;
7
:
1084
93
.