Antibody–drug conjugates (ADC) delivering pyrrolobenzodiazepine (PBD) DNA cross-linkers are currently being evaluated in clinical trials, with encouraging results in Hodgkin and non–Hodgkin lymphomas. The first example of an ADC delivering a PBD DNA cross-linker (loncastuximab tesirine) has been recently approved by the FDA for the treatment of relapsed and refractory diffuse large B-cell lymphoma. There has also been considerable interest in mono-alkylating PBD analogs. We conducted a head-to-head comparison of a conventional PBD bis-imine and a novel PBD mono-imine. Key Mitsunobu chemistry allowed clean and convenient access to the mono-imine class. Extensive DNA-binding studies revealed that the mono-imine mediated a type of DNA interaction that is described as “pseudo cross-linking,” as well as alkylation. The PBD mono-imine ADC demonstrated robust antitumor activity in mice bearing human tumor xenografts at doses 3-fold higher than those that were efficacious for the PBD bis-imine ADC. A single-dose toxicology study in rats demonstrated that the MTD of the PBD mono-alkylator ADC was approximately 3-fold higher than that of the ADC bearing a bis-imine payload, suggesting a comparable therapeutic index for this molecule. However, although both ADCs caused myelosuppression, renal toxicity was observed only for the bis-imine, indicating possible differences in toxicologic profiles that could influence tolerability and therapeutic index. These data show that mono-amine PBDs have physicochemical and pharmacotoxicologic properties distinct from their cross-linking analogs and support their potential utility as a novel class of ADC payload.

Antibody–drug conjugates (ADC) are becoming an increasingly important class of cancer therapeutics. There are currently 12 marketed ADCs, including the recent approvals of the pyrrolobenzodiazepine (PBD) ADCs loncastuximab tesirine, the biosimilar trastuzumab emtansine, and disitamab vedotin (approved in China). Several more approvals are expected in the next 4 years. At present, 110 different ADCs are being evaluated in clinical trials, (source: Beacon Targeted Therapies), and the diversity of targets, antibodies, linker technologies, and payloads is constantly increasing.

We have been studying the potential therapeutic use of the PBD pharmacophore. PBD monomers are naturally occurring antitumor agents that alkylate guanines in the minor groove of DNA. SJG-136 (SG2000) is a synthetic dimer that is able to form DNA interstrand cross-links (ICL), with a preference for PuGATCPy sequences (1). The cytotoxic potency of PBDs is due to both efficient formation of cross-links and their persistence compared with conventional cross-linking agents. On the basis of preclinical efficacy data, SJG-136 has been clinically evaluated as a stand-alone agent in leukemia and ovarian cancer (2–4). The development of even more potent PBD dimer agents has prompted their evaluation as ADC payloads, resulting in the development of talirine (5, 6) and tesirine (7–9). Although potent antitumor activity translated into significant activity in early-phase clinical trials, further development has been hampered in some cases. Obstacles have included lack of selectivity or insufficient expression of the target antigen, as well as antibody limitations such as short half-life in humans. In addition, lack of selectivity has resulted in toxicities linked to the DNA cross-linking PBD dimer. However, successes have been achieved when these factors are optimized. Tesirine is the payload component of loncastuximab tesirine and camidanlumab tesirine, which is currently being evaluated for the treatment of classical Hodgkin lymphoma in a phase II clinical trial (NCT04052997).

Monofunctional DNA alkylating agents, including the duocarmycin (10–13) and indolinobenzodiazepine mono-alkylating agents, are also being evaluated as ADC payloads (14–19). Although less cytotoxic than cross-linking agents, the DNA damage produced by these agents may elicit different DNA damage responses, including DNA repair and altered tolerability profiles.

In this study, we sought to directly compare two PBD dimer drug-linkers: SG3771, a bis-imine PBD drug-linker and direct analog of tesirine (20), and related molecule SG3784, which features only a single DNA alkylating moiety. The aims were to develop robust synthetic methodology to provide access to this class of payload; to investigate the DNA-binding modes of the free drugs in silico, on naked DNA, and in single-cell assays; to correlate free-drug cytotoxicity profiles with physicochemical properties and probable mode of action; and to investigate model ADCs in vivo in rodents to compare efficacies and therapeutic indexes.

Cell lines, antibodies, and reagents

The human cancer cell lines NCI-N87 (gastric carcinoma), A549 (lung carcinoma), HPAF-II (pancreatic adenocarcinoma), PC3 (prostate adenocarcinoma), Ramos (Burkitt lymphoma), and Daudi (Burkitt lymphoma) were obtained from the ATCC. Cell line identification was validated using the CellCheck assay (IDEXX Bioanalytics) and all cell lines were validated free of Mycoplasma contamination using the MycoSEQ assay (Thermo Fisher Scientific) or STAT-Myco assay (IDEXX Bioanalytics). All cell lines were cultured in media according to ATCC recommendations, supplemented with 10% HyClone FBS (SH30070HI; GE Healthcare), and maintained at 37°C in a humidified 5% CO2 atmosphere incubator.

Preparation of ADCs

ADCs trastuzumab-C239i-SG3771, with a drug-to-antibody ratio of 2 (DAR2; Tras-SG3771); R347-C239i-SG3771, DAR2 (R347-SG3771); 1C1-C239i-SG3771, DAR2 (1C1-SG3771); trastuzumab-C239i-SG3784, DAR2 (Tras-SG3784); 1C1-C239i-SG3784, DAR2 (1C1-SG3784); and R347-C239i-SG3784, DAR2 (R347-SG3784) were prepared according to previously published methods (21, 22). In brief, antibodies with an engineered cysteine inserted into the heavy chain at position 239 were reduced and then reoxidized to reform the native disulfide bridges and leave only the engineered cysteines available for conjugation by Michael addition with the maleimide-containing payloads. ADCs were purified typically by size exclusion chromatography–fast protein liquid chromatography and obtained with a DAR of approximately 1.8 and a monomeric purity of >95%. 1C1 Anti-EphA2 antibody was prepared as described previously (23, 24).

Molecular modeling

DNA ligands SG3552 and SG3553 were modeled using Molecular Operating Environment 2018.01 (Chemical Computing Group). Preferred sequence AGAATCT and ligand SG3553 were superimposed on PDB structure 2KTT before minimization. For ligand SG3552, an additional covalent bond was created with guanine N2.

Agarose gel–based determination of DNA interstrand cross-linking in naked DNA

DNA interstrand cross-linking was determined as described previously (25). Briefly, DNA plasmid pBR322 was linearized by HindIII digestion and then dephosphorylated with bacterial alkaline phosphatase. The 5′-ends of the plasmid were radiolabeled with T4 polynucleotide kinase in the presence of [32P]γ-ATP, and the plasmid was then precipitated and resuspended in nanopure water at 125 ng/μL. Solvent (DMSO) or compounds of interest were then added to 10 ng of total labeled plasmid DNA in 25 mmol/L triethanolamine, 1 mmol/L ethylenediaminetetraacetic acid (EDTA; pH 7.2) and incubated at 37°C for 2 hours before the reactions were terminated with a stop solution (0.6 mol/L sodium acetate, 20 mmol/L EDTA, 100 μg/mL transfer RNA). The DNA was then precipitated with three volumes of 95% ethanol at 4°C and lyophilized. An alkaline DNA strand separation buffer (1% sodium hydroxide, 6% sucrose, 0.04% bromophenol blue) was added to each DNA pellet, which was then vortexed for 3 minutes before being loaded onto a 0.8% agarose gel [gel and running buffer: 40 mmol/L tris(hydroxymethyl)aminomethane, 20 mmol/L acetic acid, and 2 mmol/L EDTA (pH 8.0)]. A nondenatured sample was dissolved in 10 μL of 6% sucrose, 0.04% bromophenol blue and loaded as a double-stranded control. Gels were dried at 80°C onto double-layer Whatman 3MM filter paper (Sigma-Aldrich) on a Hoefer GD 2000 gel dryer (Thermo Fisher Scientific) connected to a vacuum. Autoradiography was performed with double-sided x-ray film, which was scanned.

Warhead cytotoxicity

Solid warhead material was dissolved in DMSO to a 2 mmol/L stock solution, from which eight serial dilutions were made at a 1:10 ratio in DMSO and stored at –20°C until use. Cell density and viability were determined in duplicate by trypan blue exclusion assay, using a Luna-II automated cell counter (Logos Biosystems) before cell suspensions were diluted to 1 × 105 cells/mL in cell-specific growth medium. A total of 2 mL of cell suspensions per well was treated with 10 μL of warhead dilutions per well in sterile polypropylene plates. For controls, cell suspensions were treated with 10 μL of DMSO. A total of 100 μL of each sample was then dispensed into two replicate wells of a sterile, flat-bottomed, 96-well microplate and incubated in a 37°C CO2-gassed (5%) incubator. At the end of the incubation period (7 days for NCI-N87, A549, and HPAF-II; 6 days for PC3; 5 days for Ramos and Daudi cell lines), cell viability was measured by CellTiter 96 Aqueous One (MTS) assay (Promega), which was dispensed at 20 μL per well and incubated for 4 hours at 37°C in 5% CO2. Plates were then read on an EnVision Multilabel Plate Reader (PerkinElmer), using absorbance at 490 nm. Cell survival percentage was calculated from the mean absorbance of the two replicate wells for each sample and compared with the mean absorbance in the two control wells treated with DMSO only (100%). The IC50 was determined by fitting each dataset to sigmoidal dose–response curves with a variable slope, using the nonlinear curve fit algorithm on Prism software (GraphPad). All the experiments for this assay were carried out and tested in a minimum of three independent experiments. Data reported are the mean of the three independent replicates.

Bystander cell killing assay

The concentration and viability of NCI-N87/SK-BR-3 cells, from a subconfluent (80%–90% confluency) T75 flask, were measured by Trypan blue staining and counted with the Luna-II Automated Cell Counter. Cells were diluted to 2 × 105/mL, dispensed (50 μL/well) into 96-well, flat-bottomed plates, and left overnight at 37°C in a CO2-gassed incubator. A stock solution (1 mL) of ADC (20 μg/mL) was made by dilution of filter-sterilized ADC into cell culture medium. A set of 8 × 10-fold dilutions of stock ADC were made in a 24-well plate by serial transfer of 100 μL onto 900 μL of cell culture medium. ADC dilution was dispensed (50 μL/well) into four replicate wells of the 96-well plate, which contained a 50-μL cell suspension that had been seeded the previous day. Control wells received 50 μL of cell culture medium.

The 96-well plate containing cells and ADCs was incubated at 37°C in a CO2-gassed incubator for 7 days with NCI-N87 or 4 days with SK-BR-3.

MDA-MB-468 cells were plated at 1 × 105/mL as just described on the day before the positive cell line incubation was terminated. On the following day, 50 μL of medium from the cultured NCI-N87/SK-BR-3 was transferred onto the MDA-MB-468 cells. Plates were incubated for 4 days.

At the end of the incubation period, cell viability was measured by MTS assay. MTS (Promega) was dispensed (20 μL/well) into each well and incubated for 4 hours at 37°C in a CO2-gassed incubator. Well absorbance was measured at 490 nm. The percentage of cell survival was calculated from the mean absorbance in the four ADC-treated wells and compared with the mean absorbance in the four control untreated wells (100%). IC50 was determined from the dose–response data, using Prism software with a nonlinear curve fit algorithm [x = log(concentration)].

Animal studies

All animal experiments were conducted in a facility accredited by the Association for Assessment and Accreditation of Laboratory Animal Care under the guidelines of AstraZeneca's Institutional Animal Care and Use Committee and appropriate animal research approvals.

In vivo efficacy

ADCs Tras-SG3771 and Tras-SG3784 were evaluated in vivo for efficacy in the NCI-N87 human gastric carcinoma SCID mouse xenograft model by Charles River Discovery Services. Female severe combined immunodeficient mice (Fox Chase SCID, CB17/Icr-Prkdcscid/IcoIcrCrl; Charles River Laboratories) were 9 weeks old with body weights of 16.0–22.0 g on day 1 of the study. Human NCI-N87 gastric carcinoma cells were cultured in RPMI1640 medium supplemented with 10% FBS, 2 mmol/L glutamine, 100 U/mL penicillin, 100 μg/mL streptomycin sulfate, and 25 μg/mL gentamicin. The cells were grown in tissue culture flasks in a humidified incubator at 37°C in an atmosphere of 5% CO2 and 95% air. The NCI-N87 cells used for implantation were harvested during log phase growth and resuspended in PBS containing 50% Matrigel (BD Biosciences). On the day of tumor implant, each test mouse was injected subcutaneously in the right flank with 1 × 107 cells (0.1 mL of cell suspension), and tumor growth was monitored as the average tumor volume approached the target range of 100–150 mm3. Seventeen days later, designated as day 1 of the study, mice were sorted according to calculated tumor sizes into 13 groups, each consisting of 10 animals with individual tumor volumes of 108–172 mm3 and group mean tumor volumes of 120–122 mm3. Tumors were measured in two dimensions with calipers, and volume was calculated using the formula: tumor volume (mm3) = (w2 × l)/2, where w = width and l = length, in millimeters, of the tumor. Tumor weight was estimated with the assumption that 1 mg was equivalent to 1 mm3 of tumor volume. All agents were administered intravenously via tail vein injection once on day 1. The dosing volume was 0.2 mL per 20 g of body weight (10 mL/kg) and was scaled to the body weight of each individual animal. Tumors were measured with calipers twice per week, and each animal was euthanized when its tumor reached the endpoint volume of 800 mm3 or at the end of the study (day 82), whichever came first.

Pharmacokinetic assessment

Pharmacokinetic assessment was performed in plasma samples taken from non–tumor-bearing female CB17-SCID mice after a single intravenous injection to understand differences between bis-imine and mono-imine ADCs. Doses of 1 mg/kg for trastuzumab-C239i-SG3771 and 3 mg/kg for trastuzumab-C239i-SG3784 were selected on the basis of their ability to regress tumors in an NCI-N87 murine xenograft study. Pharmacokinetic parameters (area under the concentration–time curve from time 0 to last measurable concentration, maximum concentration, clearance, volume of distribution at steady state, and elimination half-life) were determined from plasma concentrations measured by ELISA for clearance, stability, and half-life perspective.

Rat toxicity studies

The tolerability and toxicity of bis-imine versus mono-imine PBD ADCs was investigated in male Sprague-Dawley rats. Control article/vehicle (25 mmol/L histidine, 200 mmol/L sucrose, 0.02% polysorbate 80, pH 6.0) or PBD ADC was administered to groups of 4 rats as a single slow intravenous bolus injection via tail vein at doses of 1.2 and 2 mg/kg R347-SG3771 or 4.5 and 6 mg/kg R347-SG3784, with a 27-day stagger between dose escalations. After dosing, animals were monitored daily for clinical signs of toxicity, and body weights were recorded twice weekly until necropsy on day 29. Blood samples were taken to evaluate hematology, coagulation, and serum chemistry at scheduled times throughout the study. Terminal blood samples were also collected from animals that were euthanized for humane reasons. Macroscopic examination was performed at scheduled necropsy, and gross observations and organ weights were recorded. Specific tissues and organs were collected and processed for microscopic examination based on gross observations and clinical pathology data.

Data availability

The data generated in this study are available within the article and its Supplementary Data.

Design and synthesis of the warheads and drug-linkers

For this work, we sought to take advantage of the available intermediates generated during the synthesis of tesirine (26) (SG3249) to study a monofunctional PBD dimer (SG3553) and a bifunctional dimer (SG3552). As it was previously reported that the loss of one of the imines lowered the efficacy of the drugs (ref. 14; thus requiring higher doses), we sought to compensate this effect by switching the central five-carbon tether of the molecules with a benzene ring. This central aromatic ring was shown to be superior to a pentyl linker at creating van der Waals interactions in the DNA minor groove but, crucially, conserved the five-carbon distance between the PBDs (Supplementary Fig. S1A).

SG3552, the bis-imine PBD dimer, was synthesized from monomeric intermediate SG3267 by previously reported methods (27). Similarly, SG3553, the DNA mono-imine PBD dimer, was obtained by statistical reduction of the PBD imines and careful purification in a process similar to that employed by Miller and colleagues (Supplementary Fig. S2A; ref. 14).

The drug-linkers SG3771 and SG3784 were designed to release the free drugs SG3552 and SG3553, respectively, after ADC antigen binding, internalization, enzymatic cleavage of the valine-alanine trigger, and self-immolation of the para-aminobenzyl carbamate in a sequence identical to that for a previously reported (20) tesirine (Supplementary Fig. S2B).

As the statistical reduction of one imine in the dimer is low yielding and requires stringent purification standards to avoid contamination by the remaining imine, we developed a new synthetic method to access the PBD alkylator core by Mitsunobu chemistry. Key building block SG3601 thus obtained was used to construct the mono-alkylating drug-linker SG3784 (Supplementary Fig. S2C). The method is highly versatile and may give rise to a wide range of N10-protected PBD mono-alkylators or related molecules. Synthetic schemes and associated experimental procedures can be found in the Supplementary Data (Supplementary Schemes S1–S8)

DNA interstrand cross-linking in naked DNA

The ability of SG3552 and SG3553 to produce DNA ICLs was evaluated in a gel-based plasmid assay. As expected, SG3552 cross-linked DNA with high efficiency and an XL50 of approximately 2 nmol/L (Fig. 1B). In contrast, SG3553 did not show evidence of double-stranded (cross-linked) DNA except at the highest doses used (≥100 nmol/L), at which some double-stranded DNA was observed. Evidence of DNA covalent interaction was, however, seen at lower doses by the dose-dependent increased migration of single-stranded DNA, which was probably due to alkylation-mediated distortion of the single-stranded DNA structure by SG3553.

Figure 1.

A, Structures of PBD dimer SG3552 and the alkylator SG3553. B,In vitro32P-labeled plasmid cross-linking assay showed that the PBD dimer SG3552 was markedly more efficient at forming ICLs than the alkylator SG3553. C, Photograph of AA8 cells treated for 2 hours with 50% ICL forming doses of cross-linker SG3552 and irradiated with 15 Gy of x-ray. The comet tail was greatly diminished by cross-linking. D, AA8 cells were incubated with an increment in SG3552 or SG3553 doses for 2 hours before being harvested for modified alkaline comet assays. SG3552 was significantly more efficient than SG3553 at forming ICLs in AA8 cells, as measured by the percentage of decrease in OTM. E, AA8 cells were pulse treated with 50% ICL, forming doses of SG3552 or SG3553 for 2 hours and allowed to grow in drug-free medium while samples were harvested for modified alkaline comet assay at 0, 6, 12, 24, and 48 hours. Lesions induced by both compounds persisted over 48 hours.

Figure 1.

A, Structures of PBD dimer SG3552 and the alkylator SG3553. B,In vitro32P-labeled plasmid cross-linking assay showed that the PBD dimer SG3552 was markedly more efficient at forming ICLs than the alkylator SG3553. C, Photograph of AA8 cells treated for 2 hours with 50% ICL forming doses of cross-linker SG3552 and irradiated with 15 Gy of x-ray. The comet tail was greatly diminished by cross-linking. D, AA8 cells were incubated with an increment in SG3552 or SG3553 doses for 2 hours before being harvested for modified alkaline comet assays. SG3552 was significantly more efficient than SG3553 at forming ICLs in AA8 cells, as measured by the percentage of decrease in OTM. E, AA8 cells were pulse treated with 50% ICL, forming doses of SG3552 or SG3553 for 2 hours and allowed to grow in drug-free medium while samples were harvested for modified alkaline comet assay at 0, 6, 12, 24, and 48 hours. Lesions induced by both compounds persisted over 48 hours.

Close modal

DNA ICLs in cells were then measured by the modified alkaline comet assay (Fig. 1C and D). Dose-dependent cross-linking was observed in AA8 cells after a 2-hour exposure with SG3552. Interestingly, evidence of “pseudo cross-links,” or aggregation of the DNA, was also observed with SG3553 but required doses that were approximately 50-fold higher. At doses that gave equivalent levels of cross-links or pseudo cross-links, these DNA lesions were found to persist in cells over a 48-hour period, suggesting that lesions caused by neither molecule are repaired or unhooked by the DNA repair machinery (Fig. 1E). Full experimental details can be found in the Supplementary Data.

In vitro cytotoxicity of SG3552 and SG3553

Warheads SG3552 and SG3553 were evaluated in a panel of hematologic and solid-tumor cell lines (Supplementary Table S1). Both compounds were found to be highly potent, with IC50 values of 0.008–1 nmol/L. Mono-imine SG3553 and bis-imine SG3552 showed similar activities in A549 (lung) and HPAF-II (pancreas) cell lines, whereas SG3552 was five times more active in Ramos and Daudi lymphoma cell lines. On average across the panel, mono-imine SG3553 was only three times less active than SG3552. The high level of activity displayed by SG3553 did not correspond to its lower (pseudo) cross-linking data.

We then looked at the lipophilicity of the free drugs at pH 7.4, either by chromatographic methods or by calculation (MarvinSketch, Chemaxon; Chemdraw, PerkinElmer; Supplementary Tables S2 and S3; Supplementary Fig. S3). A small but clear trend was observed, as the mono-alkylator was shown to be more lipophilic than the cross-linker. One would expect a small enhancement of in vitro IC50 values, as well as a bystander effect, as a result of improved permeability. Full experimental details can be found in the Supplementary Data.

Drug-linker conjugation

Drug-linkers SG3771 and SG3784 were conjugated to anti-Her2 trastuzumab (Tras), anti-EphA2 (1C1), and isotype control (R347) antibodies. These antibodies contain a single engineered cysteine in their CH3 domain (21) to allow site-specific conjugation of two drug molecules per antibody (DAR2). This well-described technology (22) produces stable ADCs with an average DAR of approximately 1.8 (Table 1) and yields of 70%–90%.

Table 1.

Characterization of ADCs evaluated in this study.

EntryADCDARaMonomer (%)bYield (%)c
Trastuzumab-C239i-SG3771 1.87 98 75d 
R347-C239i-SG3771 1.82 98 79 
1C1-C239i-SG3771 1.74 99 91 
Trastuzumab-C239i-SG3784 1.79 99 85 
R347-C239i-SG3784 1.87 97 75 
1C1-C239i-SG3784 1.72 96 68 
EntryADCDARaMonomer (%)bYield (%)c
Trastuzumab-C239i-SG3771 1.87 98 75d 
R347-C239i-SG3771 1.82 98 79 
1C1-C239i-SG3771 1.74 99 91 
Trastuzumab-C239i-SG3784 1.79 99 85 
R347-C239i-SG3784 1.87 97 75 
1C1-C239i-SG3784 1.72 96 68 

aDetermined by reduced reversed-phase high-performance liquid chromatography.

bDetermined by size exclusion chromatography–high-performance liquid chromatography after final purification.

cCalculated for ADC after purification. ADCs were purified by preparative size exclusion chromatography–fast protein liquid chromatography unless stated otherwise.

dPurification by spin filtration.

The mono-imine drug-linker (SG3784) was slightly more hydrophobic than its bis-imine counterpart SG3771, based on theoretical logD calculations (MarvinSketch). Analysis of the resultant ADCs by hydrophobic interaction chromatography revealed longer retention times for mono-imine ADCs compared with their bis-imine counterparts, reflecting a more hydrophobic character of SG3784 (Supplementary Fig. S4). However, the increased hydrophobic character was not found to induce aggregation during the conjugation process. Full experimental details can be found in the Supplementary Data.

In vitro cytotoxicity

Both anti-HER2 ADCs (Tras-SG3784 and Tras-SG3771) were found to be highly potent against NCI-N87, with an almost indistinguishable level of activity in single-digit ng/mL (Fig. 2A). High efficacy was also observed in PC3 with the anti-EphA2 1C1 ADCs. In this instance, it was observed that the bis-imine ADC was three times more active than the mono-imine (7 vs. 20 ng/mL, respectively; Fig. 2B). Both molecules displayed good bystander activity in a medium transfer assay, indicating efficient warhead diffusion. Full experimental details can be found in the Supplementary Data.

Figure 2.

A,In vitro activity of isotype controls R347-SG3771 (cross-linker) and R347-SG3784 (mono-alkylator) ADCs and anti-Her2 Tras-SG3771 (cross-linker) and Tras-SG3784 (mono-alkylator) in NCI-N87. B,In vitro activity of isotype controls R347-SG3771 (cross-linker) and R347-SG3784 (mono-alkylator) and anti-EphA2 ADCs 1C1-SG3771 (cross-linker) and 1C1-SG3784 (mono-alkylator) in PC3 cell line.

Figure 2.

A,In vitro activity of isotype controls R347-SG3771 (cross-linker) and R347-SG3784 (mono-alkylator) ADCs and anti-Her2 Tras-SG3771 (cross-linker) and Tras-SG3784 (mono-alkylator) in NCI-N87. B,In vitro activity of isotype controls R347-SG3771 (cross-linker) and R347-SG3784 (mono-alkylator) and anti-EphA2 ADCs 1C1-SG3771 (cross-linker) and 1C1-SG3784 (mono-alkylator) in PC3 cell line.

Close modal

In vivo efficacy, pharmacokinetics, and tolerability

In vivo efficacy was evaluated after a single injection of trastuzumab ADCs in the HER2 high-expressing (1.2 × 106 receptors per cell; ref. 28) NCI-N87 human gastric cancer xenograft model. Tumor growth inhibition (stasis) was observed at a dose of 1 mg/kg for Tras-SG3784 and 0.3 mg/kg for Tras-SG3771. Tumor regression to complete remission was observed at 3 mg/kg for Tras-SG3784 and 1 mg/kg for Tras-SG3771 (Fig. 3A). At these doses, treatment with Tras-SG3784 resulted in tumor-free survival in 10 of 10 mice, whereas Tras-SG3771 resulted in tumor-free survival in 6 of 10 mice (three partial responders and seven complete responders). At these curative doses, both ADCs were shown to be well tolerated, with no weight loss for the full length of the study (up to 80 days). Furthermore, at equally efficacious doses, both molecules led to a significant reduction in the Olive tail moment (OTM) of tumor cells (i.e., increased cross-linking), as measured by the modified alkaline comet assay, suggesting that in vivo efficacy corresponds to the extent of DNA cross-linking or pseudo cross-linking. However, an equal dose of the mono-imine–bearing ADC (i.e., Tras-SG3784 at 1 mg/kg) did not lead to a significant change in the OTM of the tumor cells. Importantly, there was also no significant change in the OTM of peripheral blood mononuclear cells of the same mice, suggesting minimal or no systemic off-target effect from the ADCs tested. SG3784 was further evaluated in a low-expression model: human prostate PC3 (16,000 EphA2 receptors per cell; ref. 23) with ADC 1C1-SG3784. A single administration of 3 mg/kg achieved complete remission in 10 of 10 mice.

Figure 3.

A,In vivo activity of Tras-SG3771 and Tras-SG3784 in an NCI-N87 (+++) murine xenograft model. Cross-linker SG3771 was three times more efficacious than alkylator SG3784, producing complete regression at 1.0 and 3.0 mg/kg, respectively. B, Average body weight change after treatment. Both cross-linker SG3771 and alkylator SG3784 were well tolerated at complete regression doses (1.0 and 3.0 mg/kg, respectively). C,In vivo activity of 1C1-SG3784 in a PC3 (++) murine xenograft model. Full tumor regressions were also obtained in this lower target expression model, with good tolerability. D, Cells derived from NCI-N87 xenografts 24 hours after treatment with a curative dose of Tras-SG3771 or Tras-SG3784 ADC showed a significant decrease in OTM. Extent of ICL formation corresponded to efficacy in vivo. E, No significant changes to the OTM of peripheral blood mononuclear cells were observed, regardless of the ADC treatment. I, Irradiated with 17.5 Gy of x-ray; SEM, standard error of the mean; UI, unirradiated. ANOVA P < 0.0001; F (7, 32) = 30.82. Asterisks represent Sidak multiple comparison test P values. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

Figure 3.

A,In vivo activity of Tras-SG3771 and Tras-SG3784 in an NCI-N87 (+++) murine xenograft model. Cross-linker SG3771 was three times more efficacious than alkylator SG3784, producing complete regression at 1.0 and 3.0 mg/kg, respectively. B, Average body weight change after treatment. Both cross-linker SG3771 and alkylator SG3784 were well tolerated at complete regression doses (1.0 and 3.0 mg/kg, respectively). C,In vivo activity of 1C1-SG3784 in a PC3 (++) murine xenograft model. Full tumor regressions were also obtained in this lower target expression model, with good tolerability. D, Cells derived from NCI-N87 xenografts 24 hours after treatment with a curative dose of Tras-SG3771 or Tras-SG3784 ADC showed a significant decrease in OTM. Extent of ICL formation corresponded to efficacy in vivo. E, No significant changes to the OTM of peripheral blood mononuclear cells were observed, regardless of the ADC treatment. I, Irradiated with 17.5 Gy of x-ray; SEM, standard error of the mean; UI, unirradiated. ANOVA P < 0.0001; F (7, 32) = 30.82. Asterisks represent Sidak multiple comparison test P values. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

Close modal

Pharmacokinetic studies in mice

ELISA of total antibody revealed that terminal elimination half-life for both bis-imine and mono-imine ADCs was greater than that for unconjugated antibody (Supplementary Tables S4–S6). However, ELISA of conjugated drug showed excellent in vivo stability for the mono-imine ADC compared with the bis-imine counterpart, which is evident from the lower clearance (9.5 and 14.2 mL/day/kg, respectively) and longer half-life (8.6 and 6.3 days, respectively; Fig. 4).

Figure 4.

Total antibody and total conjugated drug mouse plasma concentrations measured by ELISA for unconjugated trastuzumab-C239i (3 mg/kg), trastuzumab-C239i-SG3771 (1 mg/kg), and trastuzumab-C239i-SG3784 (3 mg/kg) ADCs. Values are reported as the geometric mean (SD) for each time point. tAb, total antibody; tADC, total conjugated drug.

Figure 4.

Total antibody and total conjugated drug mouse plasma concentrations measured by ELISA for unconjugated trastuzumab-C239i (3 mg/kg), trastuzumab-C239i-SG3771 (1 mg/kg), and trastuzumab-C239i-SG3784 (3 mg/kg) ADCs. Values are reported as the geometric mean (SD) for each time point. tAb, total antibody; tADC, total conjugated drug.

Close modal

Toxicity of PBD ADCs in rats

A single intravenous dose of R347-SG3771 at 2.0 mg/kg resulted in the early termination of 3 of 4 rats that were considered moribund. Administration of R347-SG3771 at 1.2 mg/kg and R347-SG3784 at doses of 4.5 and 6 mg/kg did not result in mortality, and all animals survived to scheduled necropsy on day 29; however, adverse effects were observed in all treated animals.

Clinical observations including lethargy, hunched posture, and piloerection were noted in animals administered 2.0 mg/kg R347-SG3771 and 6.0 mg/kg R347-SG3784. Decreases in body weight occurred in all PBD ADC-treated animals, with marked losses (∼20%) at the higher dose levels (Fig. 5A). Effects on body weight at doses of 1.2 and 4.5 mg/kg R347-SG3771 and R347-SG3784, respectively, were more modest, with group mean losses not exceeding 5% of pretreatment values.

Figure 5.

Toxicity of a single intravenous dose of R347-SG3771 and R347-SG3784 in rats. Animals were dosed on day 1 and sacrificed on day 29. Data are N = 4 for all groups. A, Effect on body weight relative to predose on day 1. Data are presented as individual animals (symbols) and group means (connecting lines). In the group receiving 2.0 mg/kg R347-SG3771, 3 of 4 animals were prematurely euthanized on different days (arrows). B–D, Hematology on day 8. Absolute white blood cell, platelet, and reticulocyte counts indicative of ADC-mediated myelotoxicity are presented as both individual (symbols) and group mean (—) values. E and F, Serum chemistry perturbations determined in samples taken on day 29 or at the time of premature euthanasia of rats dosed with 2.0 mg/kg R347-SG3771. Concentrations of creatinine and urea nitrogen are presented as both individual (symbols) and group mean (—) values and are indicative of renal toxicity for the rats dosed with 2.0 mg/kg R347-SG3771.

Figure 5.

Toxicity of a single intravenous dose of R347-SG3771 and R347-SG3784 in rats. Animals were dosed on day 1 and sacrificed on day 29. Data are N = 4 for all groups. A, Effect on body weight relative to predose on day 1. Data are presented as individual animals (symbols) and group means (connecting lines). In the group receiving 2.0 mg/kg R347-SG3771, 3 of 4 animals were prematurely euthanized on different days (arrows). B–D, Hematology on day 8. Absolute white blood cell, platelet, and reticulocyte counts indicative of ADC-mediated myelotoxicity are presented as both individual (symbols) and group mean (—) values. E and F, Serum chemistry perturbations determined in samples taken on day 29 or at the time of premature euthanasia of rats dosed with 2.0 mg/kg R347-SG3771. Concentrations of creatinine and urea nitrogen are presented as both individual (symbols) and group mean (—) values and are indicative of renal toxicity for the rats dosed with 2.0 mg/kg R347-SG3771.

Close modal

For both R347-PBD ADCs, hematology data were comparable and indicated myelotoxicity (decreases in all circulating cell lineages) that was apparent by day 8 after dosing (Fig. 5BD). In contrast to R347-SG3784, marked increases in serum levels of urea nitrogen and creatinine were noted in animals dosed with 2.0 mg/kg R347-SG3771 (Fig. 5E and F). These findings, indicative of renal toxicity, correlated with increased kidney weight and histologic findings of tubular degeneration and/or regeneration associated with tubular dilatation and karyocytomegaly (Supplementary Fig. S5).

Single doses of 1.2 mg/kg R347-SG3771 or 4.5 mg/kg R347-SG3784 were considered to be the respective MTD levels. Higher dose levels resulted in adverse effects, primarily affecting the bone marrow (both ADCs) and the kidney (R347-SG3771 only), that were not tolerated.

Data availability

Data generated in this study are available within this article and its Supplementary Data. Supplementary Data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org).

As multiple research groups are currently designing and testing DNA-binding warheads for ADC delivery, we conducted a comprehensive study of a molecule that is capable of covalently cross-linking DNA and compared it with its mono-alkylating equivalent. The aim of the study was not just to determine whether one warhead was superior to the other, but whether differences in mechanisms of action and activity and/or toxicity profiles could be observed.

As expected, initial plasmid DNA studies showed strong cross-linking activity for SG3552. More surprising results were obtained for alkylator SG3553: the experiment was able to semiquantify alkylation levels as the single-stranded DNA ran noticeably lower down the gel in a dose-dependent manner. In addition, a pseudo cross-linking effect was observed, with an XL50 value of approximately 100 nmol/L, and was surmised to occur as a result of hydrogen bonding between the nonalkylating end of SG3553 and DNA base acceptors in the minor groove. It appears that under the conditions of the assay, it takes 100/2.1, or approximately 50 hydrogen bonds with SG3553, to equate a single covalent bond from SG3552. This ratio is in line with published energy data for C-N single covalent bond (290–315 kJ/mol) and hydrogen bonds (12–25 kJ/mol, although this is widely influenced by context; ref. 29). Data extracted from the molecular model (Supplementary Fig. S1B) estimate the energy of the hydrogen bond at 10.6 kJ/mol, and so it would theoretically take 28 hydrogen bonds to equate a single covalent bond. Further cross-linking experiments run in a single-cell comet assay confirmed this behavior and, importantly, showed prolonged (at least 48 hours) DNA binding for both SG3552 and SG3553.

In vitro cytotoxicity evaluation of SG3552 and SG3553 on a panel of cancer cell lines confirmed previously observed results that the alkylating molecule is on average only three times less potent than the cross-linker. Although lymphoma cell lines appeared more sensitive to the cross-linker, other cell lines were sometimes equally sensitive to both. We sought a possible explanation for the relatively high activity of alkylator SG3553 to the higher lipophilicity of the molecule (as observed by chromatographic retention time), although this effect alone cannot explain why a molecule 50 times weaker at creating pseudo cross-links is sometimes equally cytotoxic to fully fledged cross-linker SG3552. A different mode of action must be at play here, and the results of a more detailed investigation will be the object of future work.

In vitro evaluation of both ADCs again showed potent activity, almost equally (in the case of NCI-N87 cells) and in the single-digit ng/mL range, independently of the mode of action. This was surprising, given the putative modes of action, and may be explained by a range of factors such as enhanced membrane permeability. Singh and colleagues (30) previously commented on the fact that three times more DNA adducts were observed with the alkylators than the cross-linkers, an observation possibly explained by greater nuclear localization.

In PC3 cells, the cross-linker ADC was three times more active than the alkylator, but it was not possible to conclude whether this ratio was observed as a result of the lower antigen surface expression in PC3 compared with NCI-N87, because it reflected well the results obtained with the warhead alone (Fig. 2B).

A favorable pharmacokinetic profile was achieved for the mono-imine class of compounds with lower clearance rates. Irrespective of the higher hydrophobicity profile, excellent in vivo stability has been achieved with a dose of 3 mg/kg, and similar observations have been reported previously (31). We postulate that the absence of an imine moiety in the SG3784 payload potentially avoids the interaction with serum albumin proteins, resulting in minimal deconjugation, a known phenomenon in ADCs generated through cysteine maleimide conjugation.

A question of interest to the field is whether DNA cross-linking molecules are intrinsically more toxic (and thus have a lower therapeutic index) than DNA alkylators, and whether this hypothesis could be broadly applied to the design of novel warheads. However, in a mouse tumor xenograft study, we were unable to reach a clear conclusion on this question, as both ADCs were found to be well tolerated at efficacious dose levels (complete responses at 1 and 3 mg/kg for Tras-SG3771 and Tras-SG3784, respectively). It was noted, however, that a consistent ratio of 3 seemed to be conserved between the molecules (the stasis doses, for which the tumors were prevented from growing for at least 21 days, were 0.3 and 1 mg/kg for Tras-SG3771 and Tras-SG3784, respectively). The complete responses observed in human prostate model PC3 with alkylator ADC 1C1-SG3784 at 3 mg/kg were impressive, considering that previous studies with benchmark cross-linker SG3249 (tesirine) in the same PC3 model did not show complete responses at the typical 1 mg/kg doses and required 1.5 mg/kg to achieve similar results (32, 33). This result demonstrated the versatility of the mono-alkylator PBD dimers to target both high and low target-expressing tumor models.

The 3-fold difference in efficacy observed in the murine tumor xenograft model was reflected in the rat model by the tolerability of SG3771 and SG3784 when conjugated to the untargeted R347 mAb (MTD of 1.2 and 4.5 mg/kg, respectively). Despite the fact that R347-SG3784 was better tolerated than R347-SG3771, the therapeutic index based on tolerability after a single dose was not improved, as the ratio of rat (MTD) to mouse minimum efficacious dose was approximately 4 for both payloads.

Both R347-SG3771 and R347-SG3784 elicited myelotoxicity consistent with that previously reported for PBD ADCs (33). Although the degree of myelotoxicity was generally comparable for both ADCs, only the administration of 2.0 mg/kg R347-SG3771 resulted in renal toxicity, as evidenced by increased serum levels of urea nitrogen and creatinine; these were accompanied by anatomic pathology changes that were indicative of renal tubular injury. Renal toxicity has been previously reported in rats, monkeys, and humans for tesirine and talirine payloads, indicating that the kidney could be a target organ for cross-linking, but not alkylating, PBD ADCs (33–35). The mechanism(s) underlying the observed differential renal toxicity is not currently understood. This effect could be related to pharmacokinetic differences that influence the biodistribution of cross-linking and alkylating PBD ADCs to the kidney. Clearly, additional studies are needed to better understand this emerging qualitative difference in toxicologic profile between cross-linking and alkylating PBD payloads and whether this could influence tolerability, and thus the therapeutic index, after repeated dosing.

The different mechanisms of action of the molecules clearly influenced the efficacy, stability, clearance, and toxicology profiles after a single dose without dramatically affecting the therapeutic index. The newly created hydrogen bond between the alkylator and the DNA minor groove was strong enough to correctly align the molecule in a nondistortive fashion; this provided durability of action, but not enough to hold two opposite strands of DNA in place under denaturing conditions. This mode of action, dubbed “stealthy alkylation,” together with the lipophilicity profile of the warhead that gave rise to efficient permeability, is worthy of further investigations. In particular, the pharmacologic effects of alkylator SG3553 at a cellular level, and why only a 3-fold difference in activity seemed to be echoed throughout the studies (in vitro and in vivo), remains an important avenue of research.

In conclusion, we developed two novel PBD drug-linkers with distinct modes of action. During development, novel chemistry was devised to create the PBD alkylator ring system in a reliable and flexible fashion. Both molecules were found to be highly active, both as stand-alone warheads and as targeted ADCs, with the cross-linker typically three times more potent than the alkylator. Both molecules were well tolerated in rodents after a single dose that induced complete therapeutic responses that were maintained for up to 60 days. The therapeutic indexes of the two molecules (4 and 4.5) were similar.

Some pharmacologic nuances exist between the molecules; the cross-linker was slightly more active in lymphoma cell lines, and the alkylator was more efficacious in some solid-tumor models. In addition, mechanistic studies have revealed similarities and some key differences in DNA repair, stress response signaling, and cell fate determination (manuscript in preparation). These nuances, as well as the differences in modes of action, will be studied in detail and will be the subject of further research.

A.C. Tiberghien reports a patent for WO2011/130598 issued; and was an AstraZeneca employee at the time of writing. E. Rosfjord reports personal fees from AstraZeneca, Black Diamond Therapeutics, and Pfizer outside the submitted work. J.A. Hartley reports grants, personal fees, and other support from ADC Therapeutics outside the submitted work. P.W. Howard reports other support from AstraZeneca outside the submitted work; in addition, P.W. Howard has a patent for WO2011/130598 issued, a patent for WO2014/096368 issued, and a patent for unpublished pending. No disclosures were reported by the other authors.

A.C. Tiberghien: Conceptualization, resources, data curation, investigation, writing–original draft. B. Vijayakrishnan: Conceptualization, resources, data curation, supervision, methodology, writing–original draft, project administration, writing–review and editing. A. Esfandiari: Data curation, formal analysis, investigation, writing–review and editing. M. Ahmed: Data curation, formal analysis. R. Pardo: Data curation, formal analysis. J. Bingham: Data curation, formal analysis. L. Adams: Data curation, formal analysis. K. Santos: Data curation, formal analysis, methodology. G.-D. Kang: Data curation, formal analysis, methodology. K.M. Pugh: Resources, data curation. S. Afif-Rider: Resources, supervision, methodology. K. Vashisht: Data curation, formal analysis, visualization. K. Haque: Conceptualization, data curation, formal analysis, methodology, writing–review and editing. R. Tammali: Data curation, formal analysis, methodology. E. Rosfjord: Conceptualization, resources, supervision. A. Savoca: Data curation, formal analysis. J.A. Hartley: Conceptualization, supervision, project administration, writing–review and editing. P.W. Howard: Conceptualization, supervision, project administration, writing–review and editing.

We thank Deborah J. Schuman, AZ, for the editorial support.

This study was supported by AstraZeneca.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/).

1.
Gregson
SJ
,
Howard
PW
,
Gullick
DR
,
Hamaguchi
A
,
Corcoran
KE
,
Brooks
NA
, et al
.
Linker length modulates DNA cross-linking reactivity and cytotoxic potency of C8/C8' ether-linked C2-exo-unsaturated pyrrolo[2,1-c][1,4]benzodiazepine (PBD) dimers
.
J Med Chem
2004
;
47
:
1161
74
.
2.
Hochhauser
D
,
Meyer
T
,
Spanswick
VJ
,
Wu
J
,
Clingen
PH
,
Loadman
P
, et al
.
Phase I study of sequence-selective minor groove DNA binding agent SJG-136 in patients with advanced solid tumors
.
Clin Cancer Res
2009
;
15
:
2140
7
.
3.
Janjigian
YY
,
Lee
W
,
Kris
MG
,
Miller
VA
,
Krug
LM
,
Azzoli
CG
, et al
.
A phase I trial of SJG-136 (NSC#694501) in advanced solid tumors
.
Cancer Chemother Pharmacol
2010
;
65
:
833
8
.
4.
Puzanov
I
,
Lee
W
,
Chen
AP
,
Calcutt
MW
,
Hachey
DL
,
Vermeulen
WL
, et al
.
Phase I pharmacokinetic and pharmacodynamic study of SJG-136, a novel DNA sequence selective minor groove cross-linking agent, in advanced solid tumors
.
Clin Cancer Res
2011
;
17
:
3794
802
.
5.
Kung Sutherland
MS
,
Walter
RB
,
Jeffrey
SC
,
Burke
PJ
,
Yu
C
,
Kostner
H
, et al
.
SGN-CD33A: a novel CD33-targeting antibody-drug conjugate using a pyrrolobenzodiazepine dimer is active in models of drug-resistant AML
.
Blood
2013
;
122
:
1455
63
.
6.
Fathi
AT
,
Erba
HP
,
Lancet
JE
,
Stein
EM
,
Ravandi
F
,
Faderl
S
, et al
.
A phase 1 trial of vadastuximab talirine combined with hypomethylating agents in patients with CD33-positive AML
.
Blood
2018
;
132
:
1125
33
.
7.
Saunders
LR
,
Bankovich
AJ
,
Anderson
WC
,
Aujay
MA
,
Bheddah
S
,
Black
K
, et al
.
A DLL3-targeted antibody-drug conjugate eradicates high-grade pulmonary neuroendocrine tumor-initiating cells in vivo
.
Sci Transl Med
2015
;
7
:
302ra136
.
8.
Gymnopoulos
M
,
Betancourt
O
,
Blot
V
,
Fujita
R
,
Galvan
D
,
Lieuw
V
, et al
.
TR1801-ADC: a highly potent cMet antibody-drug conjugate with high activity in patient-derived xenograft models of solid tumors
.
Mol Oncol
2020
;
14
:
54
-
68
.
9.
Flynn
MJ
,
Zammarchi
F
,
Tyrer
PC
,
Akarca
AU
,
Janghra
N
,
Britten
CE
, et al
.
ADCT-301, a pyrrolobenzodiazepine (PBD) dimer-containing antibody-drug conjugate (ADC) targeting CD25-expressing hematological malignancies
.
Mol Cancer Ther
2016
;
15
:
2709
21
.
10.
van der Lee
MMC
,
Groothuis
PG
,
Ubink
R
,
van der Vleuten
MAJ
,
van Achterberg
TA
,
Loosveld
EM
, et al
.
The preclinical profile of the duocarmycin-based HER2-targeting ADC SYD985 predicts for clinical benefit in low HER2-expressing breast cancers
.
Mol Cancer Ther
2015
;
14
:
692
703
.
11.
Owonikoko
TK
,
Hussain
A
,
Stadler
WM
,
Smith
DC
,
Kluger
H
,
Molina
AM
, et al
.
First-in-human multicenter phase I study of BMS-936561 (MDX-1203), an antibody-drug conjugate targeting CD70
.
Cancer Chemother Pharmacol
2016
;
77
:
155
62
.
12.
Yao
HP
,
Zhao
H
,
Hudson
R
,
Tong
XM
,
Wang
MH
.
Duocarmycin-based antibody-drug conjugates as an emerging biotherapeutic entity for targeted cancer therapy: pharmaceutical strategy and clinical progress
.
Drug Discov Today
2021
;
26
:
1857
74
.
13.
Boger
DL
,
Searcey
M
,
Tse
WC
,
Jin
Q
.
Bifunctional alkylating agents derived from duocarmycin SA: potent antitumor activity with altered sequence selectivity
.
Bioorg Med Chem Lett
2000
;
10
:
495
8
.
14.
Miller
ML
,
Fishkin
NE
,
Li
W
,
Whiteman
KR
,
Kovtun
Y
,
Reid
EE
, et al
.
A new class of antibody-drug conjugates with potent DNA alkylating activity
.
Mol Cancer Ther
2016
;
15
:
1870
8
.
15.
Miller
ML
,
Shizuka
M
,
Wilhelm
A
,
Salomon
P
,
Reid
EE
,
Lanieri
L
, et al
.
A DNA-interacting payload designed to eliminate cross-linking improves the therapeutic index of antibody-drug conjugates (ADCs)
.
Mol Cancer Ther
2018
;
17
:
650
60
.
16.
Rahman
KM
,
Jackson
PJ
,
James
CH
,
Basu
BP
,
Hartley
JA
,
de la Fuente
M
, et al
.
GC-targeted C8-linked pyrrolobenzodiazepine-biaryl conjugates with femtomolar in vitro cytotoxicity and in vivo antitumor activity in mouse models
.
J Med Chem
2013
;
56
:
2911
35
.
17.
Wells
G
,
Martin
CR
,
Howard
PW
,
Sands
ZA
,
Laughton
CA
,
Tiberghien
A
, et al
.
Design, synthesis, and biophysical and biological evaluation of a series of pyrrolobenzodiazepine-poly(N-methylpyrrole) conjugates
.
J Med Chem
2006
;
49
:
5442
61
.
18.
Thomas
J
,
Yurkovetskiy
A
,
Yin
M
,
Bodyak
N
,
Tang
J
,
Protopopova
M
, et al
.
Indole-biaryl pyrrolobenzodiazepines (I-BiPs): a potent and well-tolerated class of DNA mono-alkylating payload for antibody-drug conjugates (ADCs)
.
Eur J Cancer
2018
;
103
:
E63
E4
.
19.
Mantaj
J
,
Jackson
PJ
,
Rahman
KM
,
Thurston
DE
.
From anthramycin to pyrrolobenzodiazepine (PBD)-containing antibody-drug conjugates (ADCs)
.
Angew Chem Int Ed Engl
2017
;
56
:
462
88
.
20.
Tiberghien
AC
,
Levy
JN
,
Masterson
LA
,
Patel
NV
,
Adams
LR
,
Corbett
S
, et al
.
Design and synthesis of tesirine, a clinical antibody-drug conjugate pyrrolobenzodiazepine dimer payload
.
ACS Med Chem Lett
2016
;
7
:
983
7
.
21.
Dimasi
N
,
Fleming
R
,
Zhong
H
,
Bezabeh
B
,
Kinneer
K
,
Christie
RJ
, et al
.
Efficient preparation of site-specific antibody–drug conjugates using cysteine insertion
.
Mol Pharm
2017
;
14
:
1501
16
.
22.
Gregson
SJ
,
Barrett
AM
,
Patel
NV
,
Kang
GD
,
Schiavone
D
,
Sult
E
, et al
.
Synthesis and evaluation of pyrrolobenzodiazepine dimer antibody-drug conjugates with dual beta-glucuronide and dipeptide triggers
.
Eur J Med Chem
2019
;
179
:
591
607
.
23.
Jackson
D
,
Gooya
J
,
Mao
S
,
Kinneer
K
,
Xu
L
,
Camara
M
, et al
.
A human antibody-drug conjugate targeting EphA2 inhibits tumor growth in vivo
.
Cancer Res
2008
;
68
:
9367
74
.
24.
Thompson
P
,
Fleming
R
,
Bezabeh
B
,
Huang
F
,
Mao
S
,
Chen
C
, et al
.
Rational design, biophysical and biological characterization of site-specific antibody-tubulysin conjugates with improved stability, efficacy and pharmacokinetics
.
J Control Release
2016
;
236
:
100
16
.
25.
Kiakos
K
,
Hartley
JM
,
Hartley
JA
.
Measurement of DNA interstrand crosslinking in naked DNA using gel-based methods
.
Methods Mol Biol
2010
;
613
:
283
302
.
26.
Tiberghien
AC
,
von Bulow
C
,
Barry
C
,
Ge
H
,
Noti
C
,
Collet Leiris
F
, et al
.
Scale-up synthesis of tesirine
.
Org Process Res Dev
2018
;
22
:
1241
56
.
27.
Hartley
JA
,
Flynn
MJ
,
Bingham
JP
,
Corbett
S
,
Reinert
H
,
Tiberghien
A
, et al
.
Pre-clinical pharmacology and mechanism of action of SG3199, the pyrrolobenzodiazepine (PBD) dimer warhead component of antibody-drug conjugate (ADC) payload tesirine
.
Sci Rep
2018
;
8
:
10479
.
28.
Li
JY
,
Perry
SR
,
Muniz-Medina
V
,
Wang
X
,
Wetzel
LK
,
Rebelatto
MC
, et al
.
A biparatopic HER2-targeting antibody-drug conjugate induces tumor regression in primary models refractory to or ineligible for HER2-targeted therapy
.
Cancer Cell
2016
;
29
:
117
29
.
29.
Smith
MB
.
March's advanced organic chemistry: reactions, mechanisms, and structure
.
Hoboken
NJ
:
John Wiley & Sons
;
2020
. p.
105
32
.
30.
Singh
R
,
Reid
EE
,
Harris
L
,
Salomon
PL
,
Miller
ML
,
Chari
RVJ
, et al
.
Antibody-drug conjugates (ADCs) with indolinobenzodiazepine dimer (IGN) payloads: DNA-binding mechanism of IGN catabolites in target cancer cells
.
Mol Pharm
2019
;
17
:
50
8
.
31.
Bai
C
,
Reid
EE
,
Wilhelm
A
,
Shizuka
M
,
Maloney
EK
,
Laleau
R
, et al
.
Site-specific conjugation of the indolinobenzodiazepine DGN549 to antibodies affords antibody−drug conjugates with an improved therapeutic index as compared with lysine conjugation
.
Bioconjug Chem
2020
;
31
:
93
103
.
32.
St. Amant
AH
,
Huang
F
,
Lin
J
,
Lemen
D
,
Chakiath
C
,
Mao
S
, et al
.
A reactive antibody platform for one-step production of antibody-drug conjugates through a Diels-Alder reaction with maleimide
.
Bioconjug Chem
2019
;
30
:
2340
8
.
33.
Hinrichs
MJM
,
Ryan
PM
,
Zheng
B
,
Afif-Rider
S
,
Yu
XQ
,
Gunsior
M
, et al
.
Fractionated dosing improves preclinical therapeutic index of pyrrolobenzodiazepine-containing antibody drug conjugates
.
Clin Cancer Res
2017
;
23
:
5858
68
.
34.
Saber
H
,
Simpson
N
,
Ricks
TK
,
Leighton
JK
.
An FDA oncology analysis of toxicities associated with PBD-containing antibody-drug conjugates
.
Regul Toxicol Pharmacol
2019
;
107
:
104429
.
35.
Rudin
CM
,
Pietanza
MC
,
Bauer
TM
,
Ready
N
,
Morgensztern
D
,
Glisson
BS
, et al
.
Rovalpituzumab tesirine, a DLL3-targeted antibody-drug conjugate, in recurrent small-cell lung cancer: a first-in-human, first-in-class, open-label, phase 1 study
.
Lancet Oncol
2017
;
18
:
42
51
.