Site-Specific Conjugation of Monomethyl Auristatin E to Anti-CD30

Jan 27, 2015 - E-mail: [email protected]., *(R.S.) Phone: +41 56 ... In vitro cell toxicity experiments using two different CD30-posi...
0 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/molecularpharmaceutics

Site-Specific Conjugation of Monomethyl Auristatin E to Anti-CD30 Antibodies Improves Their Pharmacokinetics and Therapeutic Index in Rodent Models F. Lhospice,*,† D. Brégeon,† C. Belmant,† P. Dennler,‡ A. Chiotellis,§ E. Fischer,‡ L. Gauthier,† A. Boed̈ ec,† H. Rispaud,† S. Savard-Chambard,† A. Represa,† N. Schneider,† C. Paturel,† M. Sapet,† C. Delcambre,† S. Ingoure,† N. Viaud,† C. Bonnafous,† R. Schibli,*,‡,§ and F. Romagné∥ †

Innate Pharma SA, F13276 Marseille, France Center for Radiopharmaceutical Sciences, ETH-PSI-USZ, Paul Scherrer Institute, 5232 Villigen, Switzerland § Institute of Pharmaceutical Sciences, Department of Chemistry and Applied Biosciences, ETH Zurich, 8093 Zurich, Switzerland ∥ MI-mAbs (C/0 CIML), Parc Scientifique et Technologique de Luminy, Avenue de Luminy case 906, F13288 Marseille Cedex 9, France ‡

S Supporting Information *

ABSTRACT: Antibody−drug conjugates (ADCs) have demonstrated clinical benefits that have led to the recent FDA approval of KADCYLA and ADCETRIS. Most ADCs that are currently in clinical use or development, including ADCETRIS, are produced by chemical conjugation of a toxin via either lysine or cysteine residues, inevitably leading to heterogeneous products with variable drug-to-antibody ratios (DARs). Here, we describe the in vitro and in vivo characterization of four novel ADCs that are based on the antiCD30 antibody cAC10, which has the same polypeptide backbone as ADCETRIS, and compare the results with the latter. Bacterial transglutaminase (BTG) was exploited to site-specifically conjugate derivatives of monomethyl auristatin E (all comprising a cleavable linker) to the glutamine at positions 295 and 297 of cAC10, thereby yielding homogeneous ADCs with a DAR of 4. In vitro cell toxicity experiments using two different CD30-positive cell lines (Karpas 299 and Raji-CD30+) revealed comparable EC50 values for ADCETRIS (1.8 ± 0.4 and 3.6 ± 0.6 ng/mL, respectively) and the four cAC10-based ADCs (2.0 ± 0.4 to 4.9 ± 1.0 ng/mL). Quantitative time-dependent in vivo biodistribution studies (3−96 h p.i.) in normal and xenografted (Karpas 299 cells) SCID mice were performed with a selected 125I-radioiodinated cAC10 ADC and compared with that of 125I-ADCETRIS. The chemoenzymatically conjugated, radioiodinated ADC showed higher tumor uptake (17.84 ± 2.2% ID/g 24 h p.i.) than 125I-ADCETRIS (10.5 ± 1.8% ID/g 24 h p.i.). Moreover, 125I-ADCETRIS exhibited higher nontargeted liver and spleen uptake. In line with these results, the maximum tolerated dose of the BTG-coupled ADC (>60 mg/kg) was significantly higher than that of ADCETRIS (18 mg/kg) in rats. These results suggest that homogeneous ADCs display improved pharmacokinetics and better therapeutic indexes compared to those of chemically modified ADCs with variable DARs. KEYWORDS: antibody−drug conjugate, site-specific modification, transglutaminase, monomethyl auristatin E, CD30, ADCETRIS

■

INTRODUCTION Antibody−drug conjugates (ADCs) belong to a rapidly growing class of pharmaceutics for targeted delivery of cytotoxic drugs. More than 25 ADCs that are equipped with a variety of toxins with different modes of action are currently in clinical development against different targets.1−4 These efforts recently led to the FDA approval of KADCYLA (anti-HER2/neu antibody equipped with mertansine) and ADCETRIS (antiCD30 antibody equipped with monomethyl auristatin E) for the treatment of breast cancer and Hodgkin’s lymphoma.5−7 However, in recent years, it has also become evident that the development of ADCs remains a challenge from both a manufacturing perspective and in terms of their preclinical characterization and assessment. Critical issues to address are © 2015 American Chemical Society

(i) the stability of ADCs in blood serum with respect to premature cleavage of the toxic drug. Significant differences in plasma half-life and stability have been reported for ADCs in murine, rat, and human plasma depending on the spacer entities and the localization of the drug on the antibody.8 This can lead to low therapeutic efficacy and high off-target toxicity, thereby limiting the maximal tolerated dose (MTD) that can be Special Issue: Antibody-Drug Conjugates Received: Revised: Accepted: Published: 1863

October 10, 2014 January 8, 2015 January 27, 2015 January 27, 2015 DOI: 10.1021/mp500666j Mol. Pharmaceutics 2015, 12, 1863−1871

Article

Molecular Pharmaceutics administered.9 (ii) Current methods to conjugate the toxin to the antibody rely on chemical coupling via lysine or cysteine, leading to a mixture of species with variable drug-to-antibody ratios (DARs). In addition, it has been reported that the maleimide linkage shows a certain level of instability in serum due to the exchange of toxins with unpaired cysteine residues on circulating albumin.10 (iii) Toxins often belong to a class of very hydrophobic compounds that tend to induce antibody aggregation, especially at high DARs.9,11 Such caveats are likely to get worse with the advent of ever more complex linkers and toxins. It is clear that the field would benefit significantly from new coupling technologies that enable the generation of homogeneous ADCs. This could facilitate the manufacturing and follow-up of ADCs in vivo, thereby allowing an unbiased comparison of different linkers and toxins. All of these factors could contribute to an accelerated translation of a new generation of ADCs into the clinic. Along this line, the so-called THIOMAB technology was the first technique that allowed the production of site-specific ADCs with a defined DAR. In vivo experiments in a rodent model revealed that conjugates bearing two drug molecules display slower clearance and equivalent efficacy compared to that of a classical ADC bearing an average of four drug molecules.12,13 Although promising, such a procedure introduces a free cysteine residue in the antibody structure that may be immunogenic, requires reduction and reoxidation to attach the drug, and still suffers from the potential instability of the maleimide−thiol linkage and oxidation of unpaired cysteine residues.14 Obtaining a DAR of more than 2 using this procedure might be feasible, but production problems cannot be excluded. Another alternative is the introduction of non-natural amino acids into the structure of an antibody, which allows bio-orthogonal, site-specific coupling of a defined number of toxins.15,16 However, the tolerance and immunogenicity of such non-natural amino acid residues remain to be evaluated. Furthermore, this technology requires a new production system consisting of transfected CHO cells expressing both an engineered tRNA and the corresponding tRNA synthetase enzyme complex.17 An alternative to obtaining homogeneous ADCs that preserves the same manufacturing process for antibodies is the use of enzymatic conjugation technologies. Various enzymatic mechanisms are currently being tested, including formylglycine enzyme, sortase, galactosyltransferase for glycoengineering, and bacterial transglutaminase (BTG).18−21 The latter approach was demonstrated to yield homogeneous DAR 2 ADC’s, which were stable in vivo, but this came at the expense of introducing significant modifications in the antibody backbone with the addition of a transglutaminase tag consisting of a minimum of 4 to 6 amino acid residues.22,23 The Schibli group has reported the functionalization of antibodies via BTG without the introduction of a specific tag. They observed that deglycosylation of various antibodies at position N297 renders native Q295 accessible to BTG and produces immunoconjugates with a DAR of 2 (one per heavy chain). Here, we expand on our previous findings that demonstrated the site-specific and stoichiometric functionalization of antibodies via BTG.24−26 Using the anti-CD30 antibody cAC10 as a scaffold, we have produced an aglycosylated variant with a N297Q mutation (cAC10Q) with four potential sites of modification for BTG. The antibody was functionalized with derivatives of valine-citrulline (val-cit) MMAE using BTG. The in vitro and in vivo stability, as well as quantitative in vivo pharmacokinetics and therapeutic efficacy, of these new ADCs

were assessed in mouse and rat models. Equivalent data was obtained for ADCETRIS, thereby allowing a direct comparison between the cAC10Q ADC and ADCETRIS. These studies demonstrated that BTG-mediated, homogeneous ADCs targeting CD30 are stable in vitro and in vivo and compare favorably to ADCETRIS in terms of pharmacokinetics, safety, and efficacy in rodent models.

■

MATERIALS AND METHODS General. All of the toxins were equipped with a valinecitrulline (val-cit) 4-aminobenzyl alcohol cleavable linker unit. Amine-PEG-val-cit-MMAE and DBCO-PEG-val-cit-MMAE (Figure S1, Supporting Information) were purchased from ADC Biotechnology (St. Asaph, UK); aminocaproyl-val-citMMAE (Figure S1, Supporting Information) was purchased from Concortis (San Diego, CA, USA). ADCETRIS (brentuximab vedotin) was purchased from Theradis Pharma (Cagnessur-Mer, France). Azido-PEG-amine was purchased from Click Chemistry Tools (Scottsdale, AZ, USA). Bacterial transglutaminase was purchased from Zedira (Darmstadt, Germany). Production of cAC10-N297Q Monoclonal Antibody (cAC10Q). cAC10 (brentuximab) is a full-length antibody specific for human CD30 and has previously been described [US Patent No. 7,090,843].27 The cAC10Q variant of antibody cAC10 contains a N297Q mutation. This antibody therefore has two acceptor glutamines per heavy chain at amino acid residues 295 and 297 (EU numbering), i.e., a total of four acceptor glutamines, and is aglycosylated. The antibody was produced in CHO cells transfected with a vector encoding cAC10-N297Q. The cAC10 antibody was purified from the harvested supernatant by Protein A affinity chromatography (Ä kta FPLC, GE Healthcare). Culture supernatant was charged on a Hi-Trap column (rProt-A FF, GE Healthcare, 2 mL/min) and washed with PBS (Gibco, 10 column volumes). The protein was eluted with 0.1 M citrate buffer, pH 3, and immediately neutralized with Tris 1 M, pH 8. The eluted protein was dialyzed into PBS and stored at 4 °C. Preparation of cAC10Q ADCs. One-Step Approach. The procedure essentially follows a published method.25 cAC10Q (1 mg/mL) was incubated with 80 mol equiv of amino-valcit-MMAE and 6 U/mL bacterial transglutaminase (Zedira, Darmstadt, Germany) for 16 h at 37 °C in PBS. Excess toxin and BTG were removed by size exclusion chromatography (SEC, Superdex 200 10/300 GL, GE Healthcare, 0.5 mL/min flow). Two-Step Approach. The procedure essentially follows a published method.25 Briefly, 5 mg/mL cAC10Q was incubated with 10 mol equiv of amino-spacer-azide per site of coupling and 2 U/mL BTG overnight at 37 °C in PBS. Excess spacer and BTG were removed by size exclusion chromatography (SEC, Superdex 200 10/300 GL, GE Healthcare, 0.5 mL/min flow). Functionalized antibody (2 mg/mL in PBS) was incubated with 1.25 mol equiv of derivatized-MMAE per site of coupling (i.e., 5 equiv for cAC10Q). The mixture was incubated for 2−6 h at RT with gentle agitation. Excess derivatized-MMAE was removed by size exclusion chromatography (SEC, Superdex 200 10/300 GL, GE Healthcare, 0.5 mL/min flow). The final ADC was formulated in citrate buffer (20 mM sodium citrate, 1 mM citric acid, pH 6.6) containing 0.2 mg/mL polysorbate 80 and 70 mg/mL trehalose, i.e., ADCETRIS buffer. LC/MS Analysis of ADCs. Entire Antibody Analysis for Determination of DAR. ADC products were eluted on a PLRP-S polymeric reverse-phase column (2.1 × 50 mm, 5 μm, 4000 Å, 1864

DOI: 10.1021/mp500666j Mol. Pharmaceutics 2015, 12, 1863−1871

Article

Molecular Pharmaceutics Agilent) heated at 80 °C, at a flow rate of 0.35 mL/min, using the following gradient: 0−1 min, 5% B; 1−4 min, 5−50% B; 4−5 min, 50% B; 5−6 min, 50−5% B; 6−7 min, 5% B (A, water + 0.1% formic acid; B, acetonitrile + 0.1% formic acid). Analytes were ionized by electrospray and detected by a microTOF QII mass spectrometer (Bruker) operating in positive TOF-MS mode. Raw data were analyzed with Data Analysis software (Bruker), and deconvolution was performed using MaxEnt1. Reduced Antibody Analysis. ADCETRIS was previously deglycosylated using DeglycIT (Genovis). ADCs were reduced with DTT. ADC products were eluted on a BEH300-C4 (2.1 × 50 mm, 1.7 μm, Waters) heated at 80 °C, at a flow rate of 0.2 mL/min, using the following gradient: 0−2 min, 10% B; 2− 20 min, 10−40% B; 20−21 min, 40−90% B; 21−24 min, 90%; 24−25 min, 90−10% B; 25−30 min, 10% B (A, water + 0.1% formic acid; B, acetonitrile + 0.1% formic acid). Analytes were ionized by electrospray and detected by a Xevo G2S_QToF mass spectrometer (Waters) operating in positive TOF-MS mode. Raw data were analyzed with MassLynx software (Waters), and deconvolution was performed using MaxEnt1. ProteoStat Thermal Shift Stability Assay. The temperature of aggregation (TAgg) was measured using the ProteoStat thermal shift stability assay (Enzo Life Sciences). The readout was detected with a UV/Fluo spectrophotometer along with a Peltier temperature controller. Products with 5 mg/mL initial concentration were diluted five times in ADCETRIS buffer to obtain sample volumes of 180 μL at a concentration of 1.0 mg/mL. 50 μL of each sample was processed in triplicate. 1000× ProteoStat thermal shift detection reagent was diluted with 1× ProteoStat thermal shift diluent (final dye dilution of 2×). Using a LS55 spectrofluorometer (PerkinElmer), fluorescence was read and recorded continuously while the temperature was ramped from 40 to 100 °C in 0.5 °C steps (5 s equilibrium step duration). TAgg was determined as the maximum point in the first derivative (slope) of the fluorescence curve (dF/dT°). In Vitro Cell Toxicity Studies of cAC10Q−(1−4). The human lymphoma cell line Karpas 299 was obtained from DSMZ (Germany). The Raji cell line was obtained from ATCC. Raji-CD30 cell line was obtained by transduction with lentiviral particles encoding CD30. All cell lines were cultured in complete RPMI (RPMI-1640-GIBCO) containing 10% FBS (GIBCO), 2 mM L-glutamine (Invitrogen), 1 mM sodium pyruvate (Invitrogen), and nonessential amino-acids at 1× (Invitrogen). The cells were grown in suspension culture at 37 °C with 5% CO2. Cytotoxicity was measured using Cell Titer Glo (Promega). Cells were plated in flat bottom 96-well plates (5000 cells per well) and treated with serially diluted ADCs or ADCETRIS ranging from 10 μg/mL to 10 pg/mL and incubated at 37 °C for 96 h under a 5% CO2 atmosphere. After 96 h, the cells were labeled with Cell Titer Glo (100 μL/ well), shaken for 2 min, and incubated for 10 min in the dark. Luminescence was measured on a plate reader (VICTOR, PerkinElmer). The raw data was processed using GraphPad Prism 5.0. EC50 values are defined as the concentration that results in a 50% reduction in cell growth relative to untreated controls. 125 I Labeling of Antibodies and ADCs. cAC10Q, cAC10Q−(1), and ADCETRIS were labeled with 125I (PerkinElmer) using the iodogen method. Briefly, 100 μg of antibody or ADC was incubated with 11 MBq 125I in an iodogencoated tube on ice for 15 min. Radiolabeled protein was separated from free 125I on a PD10 column. The immunoreactivity of 125Ilabeled proteins was assessed with a cell-binding assay using Karpas 299 cells.

Biodistribution Studies with 125I-Labeled Antibodies and ADCs in SCID Mice. These experiments were approved by the local ethics committee for animal experimentation. Biodistribution studies were performed in female CB17-SCID mice (Charles Rivers). 3.5 μg of 125I-labeled antibody (0.25− 0.3 MBq) was injected into the tail vein of mice with or without subcutaneous Karpas 299 tumors, and groups of four mice were sacrificed at 3, 24, 48, 72, and 96 h later. Tumors, organs, and a blood sample were weighed and counted in a gamma counter. Pharmacokinetics Studies in Rats. Pharmacokinetics (PK) studies were performed in male Wistar rats to evaluate the total antibody concentration and the stability of the conjugated antibody. BTGs ADC and ADCETRIS were i.v. injected at 10 mg/kg, 4 animals per group. After i.v. injection, blood samples were collected on D1 (day 1) at 5 min, 1 h, 2 h, 6 h, and 24 h and then on D3, D6, D8, D10, D15, D22, D29, D36, D43, D50, D57, D64, and D71. Blood samples were processed to obtain serum and stored frozen at −20 °C until analysis. The total serum antibody concentration was assessed by ELISA at each time point. PK parameters (Cmax, AUC, clearance, half-life) were calculated with WINnonlin software using a noncompartmental analysis. DAR was measured by LC/MS analysis after affinity capture (for detailed methods, see Supporting Information). Determination of Maximum Tolerated Dose (MTD). The experiments to determine the MTD were subcontracted to C-RIS Pharma (Saint Malo, France). Groups of two rats (Wistar) (Janvier Laboratories) were injected with 27, 40, 80, and 60 mg/kg of cAC10Q−(4) and 27, 13.5, 18, and 22.5 mg/kg of ADCTERIS via single i.v. bolus administration to determine the maximum tolerated dose. Rats were monitored daily for 15 days, and both weight and clinical observation were recorded. Rats that developed significant signs of toxicity and morbidity were euthanized. The MTD was defined as the highest dose that did not induce >20% weight loss and/or signs of distress. In Vivo Efficacy of ADCs in SCID Mice. In vivo efficacy of ADCs was performed on SCID mice engrafted with Karpas 299 cells. ALCL Karpas 299 tumor cells (5 × 106) were engrafted subcutaneously in the flank of female CB17-SCID mice (Janvier Laboratories). Tumor volume was determined using the formula (L × W2)/2. At a tumor volume of approximately 70 mm3, mice were randomized into five different groups. Therapy was initiated when the tumor size in each group of 10 animals averaged approximatively 200 mm3. Treatment consisted of a single i.v. injection of ADC at two doses, 0.3 or 1 mg/kg. Tumor volume was measured every 3 days by a caliper device and calculated using the formula described above. Mice were sacrificed when the tumor volume exceeded 2000 mm3 or if the tumor was too necrotic. Surviving mice were euthanized on day 63 postimplant.

■

RESULTS Preparation of ADCs Using BTG. Dennler et al. previously described a procedure that enables the generation of homogeneous ADCs with a DAR of 2 by exploiting Q295 on deglycosylated antibodies as the sole site of modification by BTG.24,25 They tested the efficacy of one-step enzymatic and two-step chemo-enzymatic approaches to produce site-specific and stoichiometrically functionalized ADCs of Trastuzumab (Herceptin). For the present study, we avoided the deglycosylation step by introducing a single point mutation at position 297 (N297Q), the natural N-glycosylation site of the heavy chain. This created two additional sites on the cAC10 antibody 1865

DOI: 10.1021/mp500666j Mol. Pharmaceutics 2015, 12, 1863−1871

Article

Molecular Pharmaceutics

moieties in our cAC10Q mutant on its in vitro stability. We investigated the temperature of aggregation, the formation of high molecular weight protein (HMWP), and the DAR over a 6 month incubation at 40 °C in ADCETRIS formulation buffer. The temperature of aggregation of BTG-ADCs varied from 64.4 to 66.0 °C (Table S2, Supporting Information). The results demonstrated that cAC10Q−(2/4) display comparable temperatures of aggregation as that of ADCETRIS, whereas cAC10Q−(3) exhibits a significantly lower temperature of aggregation (two-way ANOVA, P = 0.0005). The higher aggregation propensity for cAC10Q−(3) was confirmed by monitoring HMWP formation by SEC-UPLC over 6 months at +40 °C (Figure S3, Supporting Information). cAC10Q−(3) provided the highest level of soluble aggregates. Interestingly, when compared to ADCETRIS, cAC10Q−(4) displayed a significantly slower rate of HMWP formation. Finally, cAC10Q−(2) had a similar HMWP formation kinetic as that of ADCETRIS. There was no significant change in DAR over 6 months for cAC10Q−(1−4) at +40 °C irrespective of the type of conjugation and type of linker (Table S1, Supporting Information). Plasma stability was assessed in cynomolgus monkey, rat, mouse, and human plasma (Figure S4, Supporting Information). Products cAC10Q−(1−4) were stable (no loss of toxin) in human and cynomolgus monkey plasma over 1 week. All products were also stable for at least 1 week at 37 °C in Wistar rat plasma (Figure S5A, Supporting Information). However, cAC10Q−(3/4) displayed instability in nude mouse plasma. The estimated half-life of cAC10Q−(3/4) was 4.5 and 3.2 h, respectively. Mass spectrometric analysis of cAC10Q− (3/4) after incubation in nude mouse plasma revealed the presence of several degraded intermediate species ranging from DAR 0 to DAR 4 resulting from the consecutive loss of a specific fragment that corresponds to a mass shift in the range 850−870 Da (Figure S5B, Supporting Information). In contrast, cAC10Q−(1/2), with a shorter linker, remained stable over 1 week in nude mouse plasma. Assessment of the in Vitro Cell Toxicity. The cytotoxic activity of cAC10 ADCs and ADCETRIS were evaluated in two CD30-positive cell lines (Karpas 299 and Raji transfected with CD30) as well as in nontransfected Raji cells as a negative control. cAC10Q−(1−4) showed similar cell killing activity in Karpas 299 and Raji-CD30 (Figure S6A,B, Supporting Information). The EC50 values of cAC10Q−(1−4) and ADCETRIS in CD30-expressing cell lines revealed that they have comparable cytotoxicity (Figure S6D, Supporting Information). As expected, no cell killing activity was observed with CD30-negative Raji cells (Figure S6C, Supporting Information). Because production of cAC10Q−(4) seems to be more scalable, provides ADCs with an exact DAR of 4, generates a low degree of aggregation, and is completely stable in rat, cynomolgus monkey, and human plasma, it was used for the PK and MTD determination in rats. On the other hand, since the serum stability in mice is low for this product, we used the stable one-step cAC10Q−(1) product for biodistribution and in vivo efficacy studies. Quantitative Biodistribution of Radioiodinated ADCs in SCID Mice. Because of its higher stability in plasma, the biodistribution and therapy studies were performed in mice with cAC10Q−(1) and ADCETRIS. cAC10Q−(1) and ADCETRIS were randomly labeled with 125I via tyrosine side chains using the classical iodogen method. In addition, we also tested the unmodified cAC10Q antibody in order to determine the influence of the toxic cargo on the pharmacokinetics.

Figure 1. Schematic representation of different cAC10Q-based ADCs. Each antibody is armed with four toxins conjugated to Q295 and Q297 on each heavy chain. cAC10Q−(1) and cAC10Q−(2) were generated by a direct one-step approach, whereas the toxin was conjugated to cAC10Q−(3) and cAC10Q−(4) via a chemo-enzymatic two-step approach. For the structure of vcMMAE, see Supporting Information Figure S1.

amenable to enzymatic coupling of spacers or toxins using BTG. The structures of the various cAC10Q-based ADCs used for the present study are shown in Figure 1. According to LC/MS analyses, the chemo-enzymatic two-step procedure generally yielded products with higher homogeneity (cAC10Q−(3/4); DAR = 4.0) than that with the enzymatic one-step procedure (cAC10Q−(1/2); DAR = 3.7; Figure S2, Supporting Information). This is in accordance with previous findings.26 Exclusive modification of the heavy chain with two toxins (total of four toxins per antibody) was observed by LC/MS analysis (Figure 2A), whereas LC/MS analysis of deglycosylated ADCETRIS, where the toxins are conventionally conjugated via reduced disulfide bridges, revealed a heterogeneous functionalization of the light and heavy chains (Figure 2B). The average DAR was calculated to be approximately 4, which allowed a fair comparison with our products, cAC10Q−(1−4). Assessment of the in Vitro Stability of cAC10Q−(1−4) and ADCETRIS. It is well-known that the chemical nature of the spacers, the position of the drug entities on the antibody backbone, and aglycosylation can potentially affect the aggregation of ADCs.28 Therefore, we were particularly interested in exploring the influence of the absence of the carbohydrate 1866

DOI: 10.1021/mp500666j Mol. Pharmaceutics 2015, 12, 1863−1871

Article

Molecular Pharmaceutics

Figure 2. Deconvoluted mass spectra of (A) reduced cAC10Q−(3) and (B) ADCETRIS. Heavy (H) and light (L) chains armed with different numbers of MMAE are indicated (L0, L1, H0−L3). The cAC10Q−(3) MS is representative of all cAC10 ADCs and shows the homogeneous reaction yield. Asterisk (*) indicates in-source MS fragmentation of H2.

Figure 3. Blood clearance (A) and time-dependent organ-to-blood ratios (B-D) of 125I-ADCETRIS (green, circles) and 125I-cAC10Q−(1) (orange, squares) in tumor-free SCID mice (solid lines) and in SCID mice bearing subcutaneous Karpas 299 tumors (dashed lines).

mice, presumably due to a high uptake of 125I-ADCs in the subcutaneous xenografts (Figure 3A). The tumor uptake was significantly higher for 125I-cAC10Q−(1) than that determined for 125I-ADCETRIS over the entire time course, with a maximum uptake of 17.8 ± 2.2% ID/g compared to 10.5 ± 1.8% ID/g 24 h postinjection, respectively (Figure 4A). Consequently, the tumor-to-liver and tumor-to-spleen ratios were significantly better for 125I-cAC10Q−(1) than that for 125IADCETRIS (Figure 4C,D). The complete set of data from the time-dependent biodistribution studies is reported in Tables S3−S8, Supporting Information. Pharmacokinetics in Wistar Rats. We compared the PK attributes of cAC10Q−(4) with those of ADCETRIS and a nonconjugated antibody as a control following single i.v.

Radiolabeling yields were >80%, radiochemical purity was >98% after purification, and immunoreactivity was intact with >70% for all three proteins. 125I-ADCETRIS showed a slightly faster blood clearance compared to that of 125I-cAC10Q−(1) in SCID mice with and without Karpas 299 tumors (Figure 3A). This faster blood clearance of 125I-ADCETRIS was accompanied by a higher liver and spleen uptake (Figure 4A), leading to higher liver-to-blood and spleen-to-liver ratios (Figure 3B/C), whereas the kidney-to-blood ratios where not significantly different between 125I-cAC10Q−(1) and 125I-ADCETRIS (Figure 3D). At the same time, it is important to note that tissue-to-blood ratios were very similar over time for 125I-cAC10Q−(1) and 125 I-cAC10Q lacking the toxins. Furthermore, it was observed that blood clearance of both ADCs was much faster in tumor-bearing 1867

DOI: 10.1021/mp500666j Mol. Pharmaceutics 2015, 12, 1863−1871

Article

Molecular Pharmaceutics

Figure 4. Biodistribution at 24 h postinjection of radioactivity (A) and time-dependent tumor-to-tissue ratios (B−D) of 125I-ADCETRIS (green, circles) and 125I-cAC10Q−(1) (orange, squares) in SCID mice bearing subcutaneous Karpas 299 tumors. Asterisk (*) indicates statistical significance determined using the Holm−Sidak method, with alpha = 5.000%.

ADCETRIS. At 1.0 mg/kg, cAC10Q−(1) and ADCETRIS showed equivalent in vivo efficacy with a complete tumor regression for 10 of 10 animals treated, in accordance with published data (Figure 6).9 At 0.3 mg/kg, cAC10Q−(1) induced a slightly higher tumor-growth delay compared to that with ADCETRIS. The control, IgG1-AC-vcMMAE, did not show any detectable effect on tumor growth.

administration in male Wistar rats at a dose of 10 mg/kg. Consistent with the ex vivo stability study in rat plasma, the DAR of cAC10Q−(4) remained stable over the tested period, with no observable loss of toxin (Figure 5A). The in vivo terminal half-life and clearance attributes of the total antibody for cAC10Q−(4), ADCETRIS, and the nonconjugated antibody control were determined using a noncompartmental model analysis. The comparison of PK parameters reveals that the terminal half-life does not differ significantly among the ADCs, whereas the clearance of cAC10Q−(4) is significantly more favorable than the clearance of ADCETRIS (one-way ANOVA, P < 0.008). Furthermore, the clearance of cAC10Q−(4) is comparable to the clearance of the nonconjugated antibody (Figure 5B). Determination of the Maximum Tolerated Dose (MTD). To assess the MTD of the ADCs, Wistar rats were treated with different doses of cAC10Q−(4) and ADCETRIS. The MTD for ADCETRIS was evaluated through a stepwise treatment starting at 27 mg/kg. Signs of toxicity and suffering (bristly fur, whitish mucous membranes and ears, closed eyes, difficulty moving) were observed, leading to a decrease in the dose to 13.5 mg/kg at first and then increasing it to 18 mg/kg, at which no toxicity was recorded. At 22.5 mg/kg, animals showed signs of suffering (bristly fur, yelping, closed eyes, difficulty moving) and body weight loss was observed. The MTD for ADCETRIS was therefore estimated to be 18 mg/kg, equivalent to the dose already assessed in Sprague−Dawley rats and published, 4. Such species are prone to faster catabolism and reduced binding to cell surface FcγRs and to cell surface mannose receptors.30 On the other hand, aglycosylated mAbs (as in the case of cAC10Q derivatives) demonstrate loose binding to both activating and inhibitory Fc receptors, but they retain binding to neonatal Fc receptor (FcRn). The absence of binding to FcR may reduce off-target toxicity on FcR-positive cells that are known to have internalizing capabilities. This feature could also explain the decreased off-target accumulation of cAC10Q−(1) in the spleen and liver as well as the significantly higher tolerability of cAC10Q−(4) in rats compared to that for ADCETRIS. The loss of FcR-mediated effector function of the cAC10Q− (1−4) ADCs, in particular, promotion of ADCC, may be considered a drawback even if the contribution of such effector functions to the efficacy of ADCs has not been clearly demonstrated. ADCs usually target receptors that exhibit a high degree of internalization, and this might limit the recruitment of effector cells. Although more studies are necessary to clarify this issue, it should be noted that cAC10Q−(1) is at least as efficient as that of glycosylated ADCETRIS, which is known to recruit effector cells in mice xenograft models, at concentrations of 0.3 and 1 mg/kg. These results indicate that the contribution

and rat plasma for at least 1 week. In contrast, ADCs cAC10Q− (3/4), comprising longer alkyl and PEG spacers, were less stable in plasma from nude mice. This phenomenon could be explained by proteases having better access to the valinecitrulline linkage in this particular species, leading to more cleavage.22 This is in accordance with a report by Tian et al. in which it was demonstrated that stability is influenced not only by the location of the toxin on the antibody but also by the spacer between the toxin and the antibody.15 They demonstrated that a short PEG2 linker was more favorable than a longer PEG4 linker. We noticed similar characteristics with cAC10Q−(1/2) that bear a shorter spacer. On the basis of the superior plasma stability data, we decided to perform in vivo biodistribution and efficacy studies in mice with cAC10Q−(1). On the other hand, the superior features of the chemo-enzymatic two-step preparation of ADCs via BTG (less excess of toxin is necessary and thus it is more likely to be scalable) prompted us to assess PK studies with cAC10Q−(4) in rats. In vivo PK data of cAC10Q−(4) in rats were consistent with the in vitro data. The conjugate was still stable after 2 weeks in circulation, and no change in the DAR was observed. In addition, we found that cAC10Q−(4) exhibits a significantly lower clearance than that of ADCETRIS. Notably, the biodistribution study in xenografted and nonxenografted SCID mice confirmed this trend. A slower clearance for 125IcAC10Q−(1) compared to that for 125I-ADCETRIS was observed over 4 days, along with lower liver-to-blood and 1869

DOI: 10.1021/mp500666j Mol. Pharmaceutics 2015, 12, 1863−1871

Article

Molecular Pharmaceutics

(4) Chari, R. V.; Miller, M. L.; Widdison, W. C. Antibody−drug conjugates: an emerging concept in cancer therapy. Angew. Chem., Int. Ed. 2014, 53, 3796−827. (5) Younes, A.; Yasothan, U.; Kirkpatrick, P. Brentuximab vedotin. Nat. Rev. Drug Discovery 2012, 11, 19−20. (6) Francisco, J. A.; Cerveny, C. G.; Meyer, D. L.; Mixan, B. J.; Klussman, K.; Chace, D. F.; Rejniak, S. X.; Gordon, K. A.; DeBlanc, R.; Toki, B. E.; Law, C. L.; Doronina, S. O.; Siegall, C. B.; Senter, P. D.; Wahl, A. F. cAC10-vcMMAE, an anti-CD30-monomethyl auristatin E conjugate with potent and selective antitumor activity. Blood 2003, 102, 1458−65. (7) Doronina, S. O.; Mendelsohn, B. A.; Bovee, T. D.; Cerveny, C. G.; Alley, S. C.; Meyer, D. L.; Oflazoglu, E.; Toki, B. E.; Sanderson, R. J.; Zabinski, R. F.; Wahl, A. F.; Senter, P. D. Enhanced activity of monomethylauristatin F through monoclonal antibody delivery: effects of linker technology on efficacy and toxicity. Bioconjugate Chem. 2006, 17, 114−24. (8) Adem, Y. T.; Schwarz, K. A.; Duenas, E.; Patapoff, T. W.; Galush, W. J.; Esue, O. Auristatin antibody drug conjugate physical instability and the role of drug payload. Bioconjugate Chem. 2014, 25, 656−64. (9) Hamblett, K. J.; Senter, P. D.; Chace, D. F.; Sun, M. M.; Lenox, J.; Cerveny, C. G.; Kissler, K. M.; Bernhardt, S. X.; Kopcha, A. K.; Zabinski, R. F.; Meyer, D. L.; Francisco, J. A. Effects of drug loading on the antitumour activity of a monoclonal antibody drug conjugate. Clin. Cancer Res. 2004, 10, 7063−70. (10) Shen, B. Q.; Xu, K.; Liu, L.; Raab, H.; Bhakta, S.; Kenrick, M.; Parsons-Reponte, K. L.; Tien, J.; Yu, S. F.; Mai, E.; Li, D.; Tibbitts, J.; Baudys, J.; Saad, O. M.; Scales, S. J.; McDonald, P. J.; Hass, P. E.; Eigenbrot, C.; Nguyen, T.; Solis, W. A.; Fuji, R. N.; Flagella, K. M.; Patel, D.; Spencer, S. D.; Khawli, L. A.; Ebens, A.; Wong, W. L.; Vandlen, R.; Kaur, S.; Sliwkowski, M. X.; Scheller, R. H.; Polakis, P.; Junutula, J. R. Conjugation site modulates the in vivo stability and therapeutic activity of antibody−drug conjugates. Nat. Biotechnol. 2012, 30, 184−9. (11) Jeffrey, S. C.; Burke, P. J.; Lyon, R. P.; Meyer, D. W.; Sussman, D.; Anderson, M.; Hunter, J. H.; Leiske, C. I.; Miyamoto, J. B.; Nicholas, N. D.; Okeley, N. M.; Sanderson, R. J.; Stone, I. J.; Zeng, W.; Gregson, S. J.; Masterson, L.; Tiberghien, A. C.; Howard, P. W.; Thurston, D. E.; Law, C. L.; Senter, P. D. A potent anti-CD70 antibody−drug conjugate combining a dimeric pyrrolobenzodiazepine drug with site-specific conjugation technology. Bioconjugate Chem. 2013, 24, 1256−63. (12) Junutula, J. R.; Raab, H.; Clark, S.; Bhakta, S.; Leipold, D. D.; Weir, S.; Chen, Y.; Simpson, M.; Tsai, S. P.; Dennis, M. S.; Lu, Y.; Meng, Y. G.; Ng, C.; Yang, J.; Lee, C. C.; Duenas, E.; Gorrell, J.; Katta, V.; Kim, A.; McDorman, K.; Flagella, K.; Venook, R.; Ross, S.; Spencer, S. D.; Lee Wong, W.; Lowman, H. B.; Vandlen, R.; Sliwkowski, M. X.; Scheller, R. H.; Polakis, P.; Mallet, W. Site-specific conjugation of a cytotoxic drug to an antibody improves the therapeutic index. Nat. Biotechnol. 2008, 26, 925−32. (13) Boswell, C. A.; Mundo, E. E.; Zhang, C.; Bumbaca, D.; Valle, N. R.; Kozak, K. R.; Fourie, A.; Chuh, J.; Koppada, N.; Saad, O.; Gill, H.; Shen, B. Q.; Rubinfeld, B.; Tibbitts, J.; Kaur, S.; Theil, F. P.; Fielder, P. J.; Khawli, L. A.; Lin, K. Impact of drug conjugation on pharmacokinetics and tissue distribution of anti-STEAP1 antibody− drug conjugates in rats. Bioconjugate Chem. 2011, 22, 1994−2004. (14) Chumsae, C.; Gaza-Bulseco, G.; Liu, H. Identification and localization of unpaired cysteine residues in monoclonal antibodies by fluorescence labeling and mass spectrometry. Anal. Chem. 2009, 81, 6449−57. (15) Tian, F.; Lu, Y.; Manibusan, A.; Sellers, A.; Tran, H.; Sun, Y.; Phuong, T.; Barnett, R.; Hehli, B.; Song, F.; DeGuzman, M. J.; Ensari, S.; Pinkstaff, J. K.; Sullivan, L. M.; Biroc, S. L.; Cho, H.; Schultz, P. G.; DiJoseph, J.; Dougher, M.; Ma, D.; Dushin, R.; Leal, M.; Tchistiakova, L.; Feyfant, E.; Gerber, H. P.; Sapra, P. A general approach to sitespecific antibody drug conjugates. Proc. Natl. Acad. Sci. U.S.A. 2014, 111, 1766−71. (16) Axup, J. Y.; Bajjuri, K. M.; Ritland, M.; Hutchins, B. M.; Kim, C. H.; Kazane, S. A.; Halder, R.; Forsyth, J. S.; Santidrian, A. F.; Stafin, K.;

of ADCC to the efficacy of the ADCs, at least in this model, is negligible. In conclusion, we propose here a scalable technology to obtain homogeneous ADCs with exactly 2 or 4 toxins per antibody. The resulting ADCs compare very favorably to ADCETRIS in terms of in vitro efficacy, stability in plasma, and in vivo efficacy. Moreover, they exhibit an increased therapeutic window. Such conjugation technologies should ease the pharmaceutical development of ADCs with improved pharmacological characteristics and should greatly facilitate their clinical development and therapeutic utility.

■

ASSOCIATED CONTENT

S Supporting Information *

Chemical synthesis and characterization data; chemical structures of val-cit-MMAE, val-cit-MMAE derivatives, and linkers; mass spectrometric analysis of one- and two-step conjugation strategies; stability experiments; aggregation assays; in vitro cell toxicity; pharmacokinetics; and the complete data set of time-dependent biodistribution with radioiodinated ADCs. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Authors

*(F.L.) Phone: +33 4 30 30 30 30. Fax: +33 4 30 30 30 00. E-mail: fl[email protected]. *(R.S.) Phone: +41 56 3102837. Fax: +41 56 310284. E-mail: [email protected]. Author Contributions

F.L., D.B., C.B., R.S., E.F., L.G., and F.R. designed the experiments; D.B., P.D., A.C., A.B., E.F., H.R., S.S., A.R., N.S., M.S., C.D., S.I., N.V., and C.B. performed the experiments; and F.L., D.B., P.D., E.F., R.S., and F.R. wrote the paper. Notes

The authors declare the following competing financial interest(s): All Innate Pharma authors are shareholders in Innate Pharma. F.R. is a shareholder in Innate Pharma.

■

ACKNOWLEDGMENTS We thank Susan Cohrs for excellent technical assistance (PSI) and Laurent Pouyet and Sabrina Carpentier for their support in experiments and statistical expertise (MI-mabs). We thank Yannis Morel, Lukas Vollmy, Esper Boel (Innate-Pharma,) and Laura Bailey (PSI) for critical review of the manuscript. We thank Hervé Brailly for his support in the project (Innate Pharma). This work was partially funded by the Swiss National Science Foundation (grant no. 132611 to P.D.).

■

ABBREVIATIONS ADC, antibody−drug conjugate; AUC, area under curve; BTG, bacterial transglutaminase; CHO, Chinese hamster ovary; DAR, drug antibody ratio; GMP, good manufacturing practice; MTD, maximum tolerated dose

■

REFERENCES

(1) Sassoon, I.; Blanc, V. Antibody−drug conjugate (ADC) clinical pipeline: a review. Methods Mol. Biol. 2013, 1045, 1−27. (2) Lambert, J. M. Drug-conjugated antibodies for the treatment of cancer. Br. J. Clin. Pharmacol. 2013, 76, 248−62. (3) Beck, A.; Reichert, J. M. Antibody−drug conjugates: present and future. mAbs 2014, 6, 15−7. 1870

DOI: 10.1021/mp500666j Mol. Pharmaceutics 2015, 12, 1863−1871

Article

Molecular Pharmaceutics Lu, Y.; Tran, H.; Seller, A. J.; Biroc, S. L.; Szydlik, A.; Pinkstaff, J. K.; Tian, F.; Sinha, S. C.; Felding-Habermann, B.; Smider, V. V.; Schultz, P. G. Synthesis of site-specific antibody−drug conjugates using unnatural amino acids. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 16101−6. (17) Liu, C. C.; Schultz, P. G. Adding new chemistries to the genetic code. Annu. Rev. Biochem. 2010, 79, 413−44. (18) Zhou, Q.; Stefano, J. E.; Manning, C.; Kyazike, J.; Chen, B.; Gianolio, D. A.; Park, A.; Busch, M.; Bird, J.; Zheng, X.; SimondsMannes, H.; Kim, J.; Gregory, R. C.; Miller, R. J.; Brondyk, W. H.; Dhal, P. K.; Pan, C. Q. Site-Specific Antibody-Drug Conjugation through Glycoengineering. Bioconjugate Chem. 2014, 25, 510−20. (19) Warden-Rothman, R.; Caturegli, I.; Popik, V.; Tsourkas, A. Sortase-tag expressed protein ligation: combining protein purification and site-specific bioconjugation into a single step. Anal. Chem. 2013, 85, 11090−7. (20) Li, X.; Fang, T.; Boons, G. J. Preparation of well-defined antibody-drug conjugates through glycan remodeling and strainpromoted azide-alkyne cycloadditions. Angew. Chem., Int. Ed. 2014, 53, 7179−82. (21) Hudak, J. E.; Barfield, R. M.; de Hart, G. W.; Grob, P.; Nogales, E.; Bertozzi, C. R.; Rabuka, D. Synthesis of heterobifunctional protein fusions using copper-free click chemistry and the aldehyde tag. Angew. Chem., Int. Ed. 2012, 51, 4161−5. (22) Strop, P.; Liu, S. H.; Dorywalska, M.; Delaria, K.; Dushin, R. G.; Tran, T. T.; Ho, W. H.; Farias, S.; Casas, M. G.; Abdiche, Y.; Zhou, D.; Chandrasekaran, R.; Samain, C.; Loo, C.; Rossi, A.; Rickert, M.; Krimm, S.; Wong, T.; Chin, S. M.; Yu, J.; Dilley, J.; Chaparro-Riggers, J.; Filzen, G. F.; O’Donnell, C. J.; Wang, F.; Myers, J. S.; Pons, J.; Shelton, D. L.; Rajpal, A. Location matters: site of conjugation modulates stability and pharmacokinetics of antibody drug conjugates. Chem. Biol. 2013, 20, 161−7. (23) Jeger, S.; Zimmermann, K.; Blanc, A.; Grünberg, J.; Honer, M.; Hunziker, P.; Struthers, H.; Schibli, R. Site-specific and stoichiometric modification of antibodies by bacterial transglutaminase. Angew. Chem., Int. Ed. 2010, 49, 9995−7. (24) Dennler, P.; Schibli, R.; Fischer, E. Enzymatic antibody modification by bacterial transglutaminase. Methods Mol. Biol. 2013, 1045, 205−15. (25) Dennler, P.; Chiotellis, A.; Fischer, E.; Bregeon, D.; Belmant, C.; Gauthier, L.; Lhospice, F.; Romagne, F.; Schibli, R. Transglutaminasebased chemo-enzymatic conjugation approach yields homogeneous antibody−drug conjugates. Bioconjugate Chem. 2014, 25, 569−78. (26) Grünberg, J.; Jeger, S.; Sarko, D.; Dennler, P.; Zimmermann, K.; Mier, W.; Schibli, R. DOTA-functionalized polylysine: a high number of DOTA chelates positively influences the biodistribution of enzymatic conjugated anti-tumor antibody chCE7agl. PLoS One 2013, 8, e60350. (27) Wahl, A. F.; Klussman, K.; Thompson, J. D.; Chen, J. H.; Francisco, L. V.; Risdon, G.; Chace, D. F.; Siegall, C. B.; Francisco, J. A. The anti-CD30 monoclonal antibody SGN-30 promotes growth arrest and DNA fragmentation in vitro and affects antitumor activity in models of Hodgkin’s disease. Cancer Res. 2002, 62, 3736−42. (28) Ju, M. S.; Jung, S. T. Aglycosylated full-length IgG antibodies: steps toward next-generation immunotherapeutics. Curr. Opin. Biotechnol. 2014, 30C, 128−39. (29) McDonagh, C. F.; Turcott, E.; Westendorf, L.; Webster, J. B.; Alley, S. C.; Kim, K.; Andreyka, J.; Stone, I.; Hamblett, K. J.; Francisco, J. A.; Carter, P. Engineered antibody−drug conjugates with defined sites and stoichiometries of drug attachment. Protein Eng., Des. Sel. 2006, 19, 299−307. (30) Gorovits, B.; Krinos-Fiorotti, C. Proposed mechanism of offtarget toxicity for antibody−drug conjugates driven by mannose receptor uptake. Cancer Immunol. Immunother. 2013, 62, 217−23.

1871

DOI: 10.1021/mp500666j Mol. Pharmaceutics 2015, 12, 1863−1871