Russell J. Sanderson

Intracellular Activation of SGN-35, a Potent Anti-CD30 Antibody-Drug Conjugate

Purpose: SGN-35 is an antibody-drug conjugate (ADC) containing the potent antimitotic drug, monomethylauristatin E (MMAE), linked to the anti-CD30 monoclonal antibody, cAC10. As previously shown, SGN-35 treatment regresses and cures established Hodgkin lymphoma and anaplastic large cell lymphoma xenografts. Recently, the ADC has been shown to possess pronounced activity in clinical trials. Here, we investigate the molecular basis for the activities of SGN-35 by determining the extent of targeted intracellular drug release and retention, and bystander activities. Experimental Design: SGN-35 was prepared with 14C-labeled MMAE. Intracellular ADC activation on CD30+ and negative cell lines was determined using a combination of radiometric and liquid chromatograhpy/mass spectrometry-based assays. The bystander activity of SGN-35 was determined using mixed tumor cell cultures consisting of CD30+ and CD30− lines. Results: SGN-35 treatment of CD30+ cells leads to efficient intracellular release of chemically unmodified MMAE, with intracellular concentrations of MMAE in the range of 500 nmol/L. This was due to specific ADC binding, uptake, MMAE retention, and receptor recycling or resynthesis. MMAE accounts for the total detectable released drug from CD30+ cells, and has a half-life of retention of 15 to 20 h. Cytotoxicity studies with mixtures of CD30+ and CD30− cell lines indicated that diffusible released MMAE from CD30+ cells was able to kill cocultivated CD30− cells. Conclusions: MMAE is efficiently released from SGN-35 within CD30+ cancer cells and, due to its membrane permeability, is able to exert cytotoxic activity on bystander cells. This provides mechanistic insight into the pronounced preclinical and clinical antitumor activities observed with SGN-35. Clin Cancer Res; 16(3); 888–97

Crystal structure of human uracil-DNA glycosylase in complex with a protein inhibitor: Protein mimicry of DNA

Uracil-DNA glycosylase inhibitor (Ugi) is a B. subtilis bacteriophage protein that protects the uracil-containing phage DNA by irreversibly inhibiting the key DNA repair enzyme uracil-DNA glycosylase (UDG). The 1.9 A crystal structure of Ugi complexed to human UDG reveals that the Ugi structure, consisting of a twisted five-stranded antiparallel beta sheet and two alpha helices, binds by inserting a beta strand into the conserved DNA-binding groove of the enzyme without contacting the uracil specificity pocket. The resulting interface, which buries over 1200 A2 on Ugi and involves the entire beta sheet and an alpha helix, is polar and contains 22 water molecules. Ugi binds the sequence-conserved DNA-binding groove of UDG via shape and electrostatic complementarity, specific charged hydrogen bonds, and hydrophobic packing enveloping Leu-272 from a protruding UDG loop. The apparent mimicry by Ugi of DNA interactions with UDG provides both a structural mechanism for UDG binding to DNA, including the enzyme-assisted expulsion of uracil from the DNA helix, and a crystallographic basis for the design of inhibitors with scientific and therapeutic applications.

Lysosomal Trafficking and Cysteine Protease Metabolism Confer Target-specific Cytotoxicity by Peptide-linked Anti-CD30-Auristatin Conjugates *

The chimeric anti-CD30 monoclonal antibody cAC10, linked to the antimitotic agents monomethyl auristatin E (MMAE) or F (MMAF), produces potent and highly CD30-selective anti-tumor activity in vitro and in vivo. These drugs are appended via a valine-citrulline (vc) dipeptide linkage designed for high stability in serum and conditional cleavage and putative release of fully active drugs by lysosomal cathepsins. To characterize the biochemical processes leading to effective drug delivery, we examined the intracellular trafficking, internalization, and metabolism of the parent antibody and two antibody-drug conjugates, cAC10vc-MMAE and cAC10vc-MMAF, following CD30 surface antigen interaction with target cells. Both cAC10 and its conjugates bound to target cells and internalized in a similar manner. Subcellular fractionation and immunofluorescence studies demonstrated that the antibody and antibody-drug conjugates entering target cells migrated to the lysosomes. Trafficking of both species was blocked by inhibitors of clathrin-mediated endocytosis, suggesting that drug conjugation does not alter the fate of antibody-antigen complexes. Incubation of cAC10vc-MMAE or cAC10vc-MMAF with purified cathepsin B or with enriched lysosomal fractions prepared by subcellular fractionation resulted in the release of active, free drug. Cysteine protease inhibitors, but not aspartic or serine protease inhibitors, blocked antibody-drug conjugate metabolism and the ensuing cytotoxicity of target cells and yielded enhanced intracellular levels of the intact conjugates. These findings suggest that in addition to trafficking to the lysosomes, cathepsin B and perhaps other lysosomal cysteine proteases are requisite for drug release and provide a mechanistic basis for developing antibody-drug conjugates cleavable by intracellular proteases for the targeted delivery of anti-cancer therapeutics.

A Potent Anti-CD70 Antibody-Drug Conjugate Combining a Dimeric Pyrrolobenzodiazepine Drug with Site-Specific Conjugation Technology

A highly cytotoxic DNA cross-linking pyrrolobenzodiazepine (PBD) dimer with a valine-alanine dipeptide linker was conjugated to the anti-CD70 h1F6 mAb either through endogenous interchain cysteines or, site-specifically, through engineered cysteines at position 239 of the heavy chains. The h1F6239C-PBD conjugation strategy proved to be superior to interchain cysteine conjugation, affording an antibody-drug conjugate (ADC) with high uniformity in drug-loading and low levels of aggregation. In vitro cytotoxicity experiments demonstrated that the h1F6239C-PBD was potent and immunologically specific on CD70-positive renal cell carcinoma (RCC) and non-Hodgkin lymphoma (NHL) cell lines. The conjugate was resistant to drug loss in plasma and in circulation, and had a pharmacokinetic profile closely matching that of the parental h1F6239C antibody capped with N-ethylmaleimide (NEM). Evaluation in CD70-positive RCC and NHL mouse xenograft models showed pronounced antitumor activities at single or weekly doses as low as 0.1 mg/kg of ADC. The ADC was tolerated at 2.5 mg/kg. These results demonstrate that PBDs can be effectively used for antibody-targeted therapy.

Measurement of in Vivo Drug Load Distribution of Cysteine-Linked Antibody–Drug Conjugates Using Microscale Liquid Chromatography Mass Spectrometry

Analysis of samples containing intact antibody-drug conjugates (ADC) using mass spectrometry provides a direct measurement of the drug-load distribution. Once dosed, the drug load distribution changes due to a combination of biological and chemical factors. Liquid chromatography-mass spectrometry (LC-MS) methods to measure the in vivo drug load distribution have been established for ADCs containing native disulfide bonds (lysine-linked or cysteine-linked). However, because of an IgG reduction step in conjugation processes, using LC-MS to analyze intact cysteine-linked ADCs requires native conditions, thus limiting sensitivity. While this limitation has been overcome at the analytical scale, to date, these methods have not been translated to a smaller scale that is required for animal or clinical doses/sampling. In this manuscript, we describe the development of ADC specific affinity capture reagents for processing in vivo samples and optimization of native LC-MS methods at a microscale. These methods are then used to detect the changing drug load distribution over time from a set of in vivo samples, representing to our knowledge the first native mass spectra of cysteine-linked ADCs from an in vivo source.

Fidelity and Mutational Specificity of Uracil-initiated Base Excision DNA Repair Synthesis in Human Glioblastoma Cell Extracts

The fidelity of DNA synthesis associated with uracil-initiated base excision repair was measured in human whole cell extracts. An M13mp2 lacZα DNA-based reversion assay was developed to assess the error frequency of DNA repair synthesis at a site-specific uracil residue. All three possible base substitution errors were detected at the uracil target causing reversion of opal codon 14 in the Escherichia coli lacZα gene. Using human glioblastoma U251 whole cell extracts, approximately 50% of the heteroduplex uracil-containing DNA substrate was completely repaired, as determined by the insensitivity of form I DNA reaction products to cleavage by a combined treatment of E. coli uracil-DNA glycosylase and endonuclease IV. The majority of repair occurred by the uracil-initiated base excision repair pathway, since the addition of the bacteriophage PBS2 uracil-DNA glycosylase inhibitor protein to extracts significantly blocked this process. In addition, the formation of repaired form I DNA molecules occurred concurrently with limited DNA synthesis, which was largely restricted to the HinfI DNA fragment initially containing the uracil residue and specific to the uracil-containing DNA strand. Based on the reversion frequency of repaired M13mp2 DNA, the fidelity of DNA repair synthesis at the target was determined to be about one misincorporated nucleotide per 1900 repaired uracil residues. The major class of base substitutions propagated transversion mutations, which were distributed almost equally between T to G and T to A changes in the template. A similar mutation frequency was also observed using whole cell extracts from human colon adenocarcinoma LoVo cells, suggesting that mismatch repair did not interfere with the fidelity measurements.

Increased spontaneous mutation frequency in human cells expressing the phage PBS2-encoded inhibitor of uracil-DNA glycosylase

The Ugi protein inhibitor of uracil-DNA glycosylase encoded by bacteriophage PBS2 inactivates human uracil-DNA glycosylases (UDG) by forming a tight enzyme:inhibitor complex. To create human cells that are impaired for UDG activity, the human glioma U251 cell line was engineered to produce active Ugi protein. In vitro assays of crude cell extracts from several Ugi-expressing clonal lines showed UDG inactivation under standard assay conditions as compared to control cells, and four of these UDG defective cell lines were characterized for their ability to conduct in vivo uracil-DNA repair. Whereas transfected plasmid DNA containing either a U:G mispair or U:A base pairs was efficiently repaired in the control lines, uracil-DNA repair was not evident in the lines producing Ugi. Experiments using a shuttle vector to detect mutations in a target gene showed that Ugi-expressing cells exhibited a 3-fold higher overall spontaneous mutation frequency compared to control cells, due to increased C:G to T:A base pair substitutions. The growth rate and cell cycle distribution of Ugi-expressing cells did not differ appreciably from their parental cell counterpart. Further in vitro examination revealed that a thymine DNA glycosylase (TDG) previously shown to mediate Ugi-insensitive excision of uracil bases from DNA was not detected in the parental U251 cells. However, a Ugi-insensitive UDG activity of unknown origin that recognizes U:G mispairs and to a lesser extent U:A base pairs in duplex DNA, but which was inactive toward uracil residues in single-stranded DNA, was detected under assay conditions previously shown to be efficient for detecting TDG.

IDENTIFICATION OF SPECIFIC CARBOXYL GROUPS ON URACIL-DNA GLYCOSYLASE INHIBITOR PROTEIN THAT ARE REQUIRED FOR ACTIVITY

The bacteriophage PBS2 uracil-DNA glycosylase inhibitor (Ugi) protein inactivates uracil-DNA glycosylase (Ung) by forming an exceptionally stable protein-protein complex in which Ugi mimics electronegative and structural features of duplex DNA (Beger, R. D., Balasubramanian, S., Bennett, S. E., Mosbaugh, D. W., and Bolton, P. H. (1995) J. Biol. Chem. 270, 16840-16847; Mol, C. D., Arvai, A. S., Sanderson, R. J., Slupphaug, G., Kavli, B., Krokan, H. E., Mosbaugh, D. W., and Tainer, J. A. (1995) Cell 82, 701-708). The role of specific carboxylic amino acid residues in forming the Ung·Ugi complex was investigated using selective chemical modification techniques. Ugi treated with carbodiimide and glycine ethyl ester produced five discrete protein species (forms I-V) that were purified and characterized. Analysis by mass spectrometry revealed that Ugi form I escaped protein modification, and forms II-V showed increasing incremental amounts of acyl-glycine ethyl ester adduction. Ugi forms II-V retained their ability to form a Ung·Ugi complex but exhibited a reduced ability to inactivate Escherichia coli Ung, directly reflecting the extent of modification. Competition experiments using modified forms II-V with unmodified Ugi as a competitor protein revealed that unmodified Ugi preferentially formed complex. Furthermore, unmodified Ugi and poly(U) were capable of displacing forms II-V from a preformed Ung·Ugi complex but were unable to displace Ugi form I. The primary sites of acyl-glycine ethyl ester adduction were located in the α2-helix of Ugi at Glu-28 and Glu-31. We infer that these two negatively charged amino acids play an important role in mediating a conformational change in Ugi that precipitates the essentially irreversible Ung/Ugi interaction.

Uracil-initiated base excision DNA repair synthesis fidelity in human colon adenocarcinoma LoVo and Escherichia coli cell extracts.

The error frequency of uracil-initiated base excision repair (BER) DNA synthesis in human and Escherichia coli cell-free extracts was determined by an M13mp2 lacZ alpha DNA-based reversion assay. Heteroduplex M13mp2 DNA was constructed that contained a site-specific uracil target located opposite the first nucleotide position of opal codon 14 in the lacZ alpha gene. Human glioblastoma U251 and colon adenocarcinoma LoVo whole-cell extracts repaired the uracil residue to produce form I DNA that was resistant to subsequent in vitro cleavage by E. coli uracil-DNA glycosylase (Ung) and endonuclease IV, indicating that complete uracil-initiated BER repair had occurred. Characterization of the BER reactions revealed that (1) the majority of uracil-DNA repair was initiated by a uracil-DNA glycosylase-sensitive to Ugi (uracil-DNA glycosylase inhibitor protein), (2) the addition of aphidicolin did not significantly inhibit BER DNA synthesis, and (3) the BER patch size ranged from 1 to 8 nucleotides. The misincorporation frequency of BER DNA synthesis at the target site was 5.2 x 10(-4) in U251 extracts and 5.4 x 10(-4) in LoVo extracts. The most frequent base substitution errors in the U251 and LoVo mutational spectrum were T to G > T to A >> T to C. Uracil-initiated BER DNA synthesis in extracts of E. coli BH156 (ung) BH157 (dug), and BH158 (ung, dug) was also examined. Efficient BER occurred in extracts of the BH157 strain with a misincorporation frequency of 5.6 x 10(-4). A reduced, but detectable level of BER was observed in extracts of E. coli BH156 cells; however, the mutation frequency of BER DNA synthesis was elevated 6.4-fold.

Engineered cysteine antibodies: an improved antibody-drug conjugate platform with a novel mechanism of drug-linker stability

Antibody-drug conjugates (ADCs) are fulfilling the promise of targeted therapy with meaningful clinical success. An intense research effort is directed towards improving pharmacokinetic profiles, toxicity and chemical stability of ADCs. The majority of ADCs use amide and thioether chemistry to link potent cytotoxic agents to antibodies via endogenous lysine and cysteine residues. While maleimide-cysteine conjugation is used for many clinical stage ADC programs, maleimides have been shown to exhibit some degree of post-conjugation instability. Previous research with site-directed mutagenic incorporation of cysteine residues for conjugation revealed that the stability of the drug-antibody linkage depends on the site of conjugation. Here we report on a collection of engineered cysteine antibodies (S239C, E269C, K326C and A327C) that can be site-specifically conjugated to potent cytotoxic agents to produce homogenous 2-loaded ADCs. These ADCs confirm that site of conjugation impacts maleimide stability and present a novel mechanism of thioether stabilization, effectively unlinking stability from either local chemical environment or calculated solvent accessibility and expanding the current paradigm for ADC drug-linker stability. These ADCs show potent in vitro and in vivo activity while delivering half of the molar equivalent dose of drug per antibody when compared to an average 4-loaded ADC. In addition, our lead engineered site shields highly hydrophobic drugs, enabling conjugation, formulation and clinical use of otherwise intractable chemotypes.