Understanding How the Stability of the Thiol ... - ACS Publications

Subscriber access provided by ORTA DOGU TEKNIK UNIVERSITESI KUTUPHANESI

Article

Understanding How the Stability of the Thiol-Maleimide Linkage Impacts the Pharmacokinetics of Lysine-linked Antibody-Maytansinoid Conjugates Jose F. Ponte, Xiuxia Sun, Nicholas C Yoder, Nathan Fishkin, Rassol Laleau, Jennifer Coccia, Leanne Lanieri, Megan Bogalhas, Lintao Wang, Sharon D. Wilhelm, Wayne C. Widdison, Thomas A Keating, Ravi V.J. Chari, Hans Erickson, and John M Lambert Bioconjugate Chem., Just Accepted Manuscript • DOI: 10.1021/acs.bioconjchem.6b00117 • Publication Date (Web): 13 May 2016 Downloaded from http://pubs.acs.org on May 15, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Bioconjugate Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Bioconjugate Chemistry

Understanding How the Stability of the Thiol-Maleimide Linkage Impacts the Pharmacokinetics of Lysine-linked Antibody-Maytansinoid Conjugates

Jose F. Ponte(1), *Xiuxia Sun(1), Nicholas C Yoder(2), Nathan Fishkin(1), Rassol Laleau(1), Jennifer Coccia(3), Leanne Lanieri(1), Megan Bogalhas(1), Lintao Wang(1), Sharon Wilhelm(1), Wayne Widdison(1), Thomas A. Keating(4), Ravi Chari(1), Hans K. Erickson(5), John M. Lambert(1)

(1) Immunogen Inc – Biochemistry, 830 winter st Waltham Massachusetts 02451-1477, United States (2) Whitehead Institute, 9 Cambridge Center, Cambridge, Massachusetts 02142 (3) Regeneron Pharmaceuticals Inc, 777 Old Saw Mill River Rd., Tarrytown, New York 10591 (4) AstraZeneka Pharmaceuticals, 35 Gatehouse Drive, Waltham, Massachuesetts, 02451 (5) Genentech – Protein Chemistry, 1 DNA Way, South San Francisco, California, 94080

Abstract: Antibody-drug conjugates (ADCs) have become a widely investigated modality for cancer therapy, in part due to the clinical findings with ado-trastuzumab emtansine (Kadcyla®). Ado-trastuzumab emtansine utilizes the Ab-SMCC-DM1 format, in which the

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

thiol-functionalized maytansinoid cytotoxic agent, DM1, is linked to the antibody (Ab) via the maleimide moiety of the heterobifunctional SMCC linker. The pharmacokinetic (PK) data for ado-trastuzumab emtansine point to a faster clearance for the ADC than for total antibody. Cytotoxic agent release in plasma has been reported with non-maytansinoid, cysteine-linked ADCs via thiol-maleimide exchange, for example, brentuximab vedotin. For Ab-SMCC-DM1 ADCs, however, the main catabolite reported is lysine-SMCC-DM1, the expected product of intracellular antibody proteolysis. To understand these observations better, we conducted a series of studies to examine the stability of the thiol-maleimide linkage, utilizing the EGFR-targeting conjugate, J2898A-SMCC-DM1, and comparing it with a control ADC made with a non-cleavable linker that lacked a thiol-maleimide adduct (J2898A-(CH2)3-DM). We employed radiolabeled ADCs to directly measure both the antibody and the ADC components in plasma. The PK properties of the conjugated antibody moiety of the two conjugates, J2898A-SMCC-DM1 and J2898A-(CH2)3-DM (each with an average of 3.0 to 3.4 maytansinoid molecules per antibody), appear to be similar to that of the unconjugated antibody. Clearance values of the intact conjugates were slightly faster than those of the Ab components. Furthermore, J2898A-SMCC-DM1 clears slightly faster than J2898A-(CH2)3-DM, suggesting that there is a fraction of maytansinoid loss from the SMCC-DM1 ADC, possibly through a thiol-maleimide dependent mechanism. Experiments on ex vivo stability confirm that some loss of maytansinoid from Ab-SMCCDM1 conjugates can occur via thiol elimination, but at a slower rate than the corresponding rate of loss reported for thiol-maleimide links formed at thiols derived by reduction of endogenous cysteine residues in antibodies, consistent with expected differences in thiol-


Page 2 of 37

Page 3 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


maleimide stability related to thiol pKa. These findings inform the design strategy for future ADCs.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Introduction Interest in the development of antibody-drug conjugates (ADCs) for cancer therapy has increased dramatically in the past 5 years, stimulated by the encouraging clinical trial data and regulatory approval of brentuximab vedotin (Adcetris®)1, 2 and ado-trastuzumab emtansine (TDM1, Kadcyla®).3-5 ADCs currently in the clinic consist of monoclonal antibodies to a multitude of targets, chemically conjugated to a variety of cytotoxic small molecules, and utilizing several different linker types.6

Thiol-maleimide linkage has been widely used in the synthesis of ADCs because of its high selectivity, rapid reaction kinetics, and compatibility with aqueous reaction conditions.7 For example, brentuximab vedotin is prepared by reducing the endogenous antibody (Ab) cystines, followed by conjugation of cysteine thiols with a maleimide-functionalized monomethyl auristatin E derivative (mc-VC-PAB-MMAE).1,2 Ado-trastuzumab emtansine (T-DM1) is prepared by reacting trastuzumab, through its lysine residues, with the N-hydroxysuccinimidyl (NHS) ester of the SMCC linker, followed by reaction of the maleimide moiety of the linker with the thiol group of DM1, a derivative of the natural cytotoxic polyketide maytansine.8

Although thiol-maleimide bonds (thiosuccinimides) have long been believed to be stable in plasma, some studies have suggested that these thioether bonds can undergo decomposition by oxidation9 or β-elimination.7 Notably, the thiol-maleimide link of mc-VC-PAB-MMAE ADCs has been shown to undergo dissociation in vivo, followed by scavenging of the released maleimide by free thiols on cysteines of serum albumin.10 Further, multiple reports on the in


Page 4 of 37

Page 5 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


vivo disposition of Ab-SMCC-DM1 conjugates have been published. In particular, clinical plasma pharmacokinetics (PK) data for T-DM1 indicate a two-fold faster clearance and two-fold shorter terminal half-life for the conjugate as compared to total Ab11, as determined by ELISA. Similar faster clearance of ADC versus Ab was reported in mice 12 and in cynomolgus monkeys.13 Based on the observation of loss of the payload moiety from cysteine-linked ADCs, this loss of DM1 from T-DM1 is sometimes ascribed to elimination of DM1 from the SMCC maleimide.10

Baldwin and Kiick have shown that the stability of the thiosuccinimide linkages is strongly dependent on the pKa of the thiol, with thiosuccinimides prepared from thiols with higher pKa s being more stable.7 Therefore, it is reasonable to propose that lysine-linked (DM1 thiol pKa, ~9.8) ADCs will be more stable than cysteine-linked (thiol pKa , ~8.6) ADCs. However, the stability of lysine-linked Ab-SMCC-DM1 can be affected by other factors besides that of the thiol-maleimide linkage stability. For example, ADCs produced by reaction with NHS esters for conjugation at lysine amino groups are expected to release low levels of carboxylate species due to a population of unstable non-amide linkages formed by reaction with other amino acids, particularly tyrosines.14 The other possible mechanisms for maytansinoid loss is the release of maysine from conjugates via β-elimination.15

In this study we examined the PK of the Ab-SMCC-DM1 conjugates to better understand the in vivo disposition of the Ab and linker components and examine the fate of the thiol-maleimide moiety used in lysine-linked DM1-conjugates. We designed an ADC with a non-cleavable linker, wherein the maytansinoid was conjugated directly to lysine residues of the Ab via a linker



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

with an all-carbon backbone lacking the thiol-maleimide bond of the SMCC-DM1 conjugate. The Ab-(CH2)3-DM ADC serves as a reference for assessing the contribution of the stability of the thiol-maleimide linkage in the overall stability of Ab-SMCC-DM1 ADC, since both conjugates utilize NHS ester chemistry for antibody conjugation and would be expected to be similarly sensitive to potential instability due to any non-amide linkages that may be formed, and to maysine release via β-elimination, as mentioned above. We compared the PK in vivo, and stability ex vivo of this new ADC construct with that of Ab-SMCC-DM1. We discuss the evidence for elimination of DM1 from the SMCC linker, and the contribution of mechanisms both dependent on, and independent of, thiol-maleimide chemistry as delineated by the control ADC lacking this linkage. We show that, in a conjugate setting, a fraction of DM1 elimination from SMCC can be attributed to the reversion of the thiol-maleimide reaction, although the thiosuccinimide linkage described here between SMCC and DM1 is more stable than the thiosuccinimides formed at reduced endogenous cystine residues in Abs as described in the literature.10


Page 6 of 37

Page 7 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Results and Discussion In order to better understand the role of the thiol-maleimide linkage in the pharmacokinetic behavior of SMCC-DM1 ADCs, we prepared a new maytansinoid-ADC, Ab-(CH2)3-DM (Figure 1). In this ADC, the maytansinoid was linked to the Ab via a non-cleavable linker with an all carbon backbone that lacked the thiol-maleimide (thiosuccinimide moiety). Chemical synthesis of the DM-(CH2)3-NHS ester from previously described maytansinoid derivatives is described in the supplementary information. The NHS ester was used to generate Ab-(CH2)3-DM conjugates

Figure 1. Structures of Ab-SMCC-DM1 and Ab-(CH2)3-DM conjugates. Both ADCs are synthesized by reacting Ab lysines with the NHS ester of either the SMCC linker or the linkerDM. Ab-(CH2)3-DM is similar in structure to Ab-SMCC-DM1 but lacks the thiosuccinimide.

linked via amino groups of lysine residues of the Abs. To evaluate if the thiosuccinimide moiety could affect the in vivo disposition of the Ab component, unconjugated anti-huEGFR Ab (Ab1) and the related ADCs, Ab1-SMCC-DM1 and Ab1-(CH2)3-DM, were trace-labeled on the Ab moiety with 3H-propionic acid NHS ester16 to provide a sensitive means of Ab detection (Figure 2a). The murine-derived anti-huEGFR Ab does not bind to the murine EGFR so that there was no involvement of antigen-mediated clearance in the experimental system. Mouse PK of the



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

radiolabeled Ab1-SMCC-DM1, Ab1-(CH2)3-DM, and Ab1 showed no measurable difference in PK curves (Figure 2a), nor in the derived t1/2 values and total exposure values (Table 1). These observations are consistent with previous studies carried out using 125I-labeled Ab and other linker-maytansinoid ADC formats.12, 17

(a)


Page 8 of 37

Page 9 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1. Pharmacokinetic parameters (t1/2, AUC and CL) for Ab and DM components of radiolabeled ADCs derived from plasma concentrations following i.v. injection (10 mg/kg, mouse). Average of three animals per time point ± S.D. T1/2 , terminal half-life; AUC, area under the curve; CL, clearance.

(b)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(c)

(d)


Page 10 of 37

Page 11 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(e)

Figure 2. Pharmacokinetic studies evaluating Ab and ADC concentrations using radiolabeled conjugates. (a) Conjugation via lysines at DAR ≈ 3.5 has no impact on Ab PK. PK of [3H]Ab1, [3H]Ab1-SMCC-DM1 and [3H]Ab1-(CH2)3-DM conjugates (mouse, 10 mg/kg injected i.v. dose) show that the Ab component of conjugate clears similarly to the clearance of naked Ab. The plasma concentration of Ab for all test articles was determined by radioactivity associated with tritium label on Ab. (b) PK of Ab1-SMCC-[3H]DM1 and Ab1-(CH2)3-[3H]DM conjugates (mouse, 10 mg/kg) showing that Ab1-SMCC-[3H]DM1 clears slightly faster than Ab1-(CH2)3[3H]DM. Concentration of maytansinoid was determined by radioactivity associated with tritium label on the maytansinoid. Tabular results with pharmacokinetic parameters for Figure 2a and 2b can be found in Table 1. (c) DAR over the 28 day experiment was estimated by dividing



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

maytansinoid concentration from Figure 2b by antibody concentration from Figure 2a. (d) ADC PK was repeated with two additional Ab-SMCC-[3H]DM1 conjugates (Ab2-SMCC-[3H]DM1, 12 mg/kg and Ab3-SMCC-[3H]DM1, 10 mg/kg) and very similar results were found. (e) Monomeric status of Ab1-SMCC-[3H]DM1 and Ab1-(CH2)3-[3H]DM conjugates from plasma (b) at indicated time analyzed by SEC-HPLC. The arrow points to a small early eluting shoulder peak which was observed for Ab1-SMCC-[3H]DM1.

In order to examine the effect of thiol-maleimide linkage stability on Ab-SMCC-DM1 disposition in vivo, the PK of ADCs incorporating radiolabeled SMCC-DM1 or (CH2)3-DM were compared. Mice were administered a single 10 mg/kg dose of Ab1-SMCC- [3H]DM1 or Ab1(CH2)3-[3H]DM and the plasma concentrations of the ADCs were determined by liquid scintillation counting (LSC) (Figure 2b). Both conjugates had similar clearance with t1/2 values in the 9-11 day range. However, both conjugates (radiolabeled maytansinoid) cleared slightly faster than the Ab component of the conjugates (radiolabeled protein), as shown by comparison of the t1/2 and AUC values in Table 1. This is consistent either with eliminative loss of [3H]maytansinoid from the Ab, or with the selective faster clearance of the species having the greater number of cytotoxic molecules per antibody, known as DAR, from the pool of ADC species present in plasma (effectively lowering the overall average DAR over time). However, the latter hypothesis is not consistent with the observation that the Ab1 components of these two ADCs were cleared from circulation with similar PK as the unconjugated Ab1 (Figure 2a and Table 1).

Further examination of the results shown in Figure 2b reveals that the measured AUC over the 28 day experiment was approximately 25% lower for the SMCC-DM1 conjugate than for the


Page 12 of 37

Page 13 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(CH2)3-DM conjugate (Table 1, 639 day·µg·mL-1 vs. 872 day·µg·mL-1). Thus, we observe somewhat greater stability in vivo with the Ab1-(CH2)3-[3H]DM conjugate which lacks the thiosuccinimide moiety contained within the linker of the Ab1-SMCC-DM1 conjugate (Figure 1). To estimate the extent of maytansinoid loss from the conjugates with respect to time in circulation, values for DAR over the 28 day experiment were estimated by dividing the linked maytansinoid concentration (from this ADC-based experiment) by the Ab concentration (in the corresponding antibody PK experiment) at each time point (Figure 2c), thus normalizing for overall antibody clearance mechanisms. In the case of Ab1- (CH2)3-DM , the starting DAR of 3.5 decreased to about 2.8 after one week, and was maintained at 2.8 after 28 days (20% reduction in DM load), the loss of maytansinoid from the conjugate appearing to plateau after about day 4 (Figure 2c). In the case of Ab1-SMCC-DM1, the DAR decreased during circulation in vivo from a starting value of 3.1 to about 2.2 after one week, and to about 1.6 after 28 days (29% and 48% reduction in DM1 load, respectively). The loss of maytansinoid from the SMCCDM1conjugate did not appear to reach a plateau over 28 days in circulation. Comparison of the two conjugates suggests that about 9% and 22% of the loss of maytansinoid could be attributed to the presence of the thiosuccinimide moiety in the linker of Ab-SMCC-DM1 after 7 and 14 days in circulation, respectively. By comparison, an ADC produced by alkylation of endogenous cysteines with a cytotoxin-maleimide showed a 50% and 70% DAR decrease after 7 and 14 days in vivo, with this loss attributed primarily to reversion of the thiol-maleimide adduct.10 Finally, to ensure the PK observed with Ab1-SMCC-[3H]DM1 were representative of this class of ADC, we repeated the experiments with two additional Ab-SMCC-[3H]DM1 conjugates and found very similar results (Figure 2d).



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Plasma samples from the above PK study comparing Ab1-SMCC-[3H]DM1 and Ab1-(CH2)3[3H]DM1 were also analyzed by size exclusion chromatography (SEC) using fractionation and subsequent LSC (Figure 2e) as described by Shen et al.8 The results showed that the major circulating species consisted of monomeric ADCs. No evidence was observed for accumulation of high molecular weight aggregated ADCs, or of albumin-linked [3H]-DM1 species (as might be expected from elimination of DM1 free thiol from the SMCC-DM1 ADC and subsequent reaction of the DM1with albumin through thiol-disulfide exchange reactions). However, a small early-eluting shoulder peak, as indicated by an arrow, was observed starting at 7 d for the Ab1SMCC-DM1-containing samples. No such peak is apparent in the samples containing Ab1(CH2)3-DM, suggesting that the minor peak observed in the plasma samples containing Ab1SMCC-DM1was formed by a thiol-maleimide dependent mechanism. We subsequently observed the formation of similar species during ex vivo plasma incubation, which were confirmed to consist of conjugate–albumin adducts (vide infra).

To compare in vivo stability of Ab-SMCC-DM1 and Ab-(CH2)3-DM using a direct, complementary method, we examined the DAR distribution of Ab4 (an Ab that recognizes human CD56) ADCs recovered from mouse plasma using affinity capture with CD56-ECD-Fc followed by LC-MS (Figure 3). Both Ab4-SMCC-DM1 and Ab4-(CH2)3-DM showed loss of high DAR species to a similar extent, suggesting that the major mechanisms for maytansinoid loss during 0 to 7 days in vivo are independent of thiol-maleimide chemistry, and consistent with the change in DAR observed from 0 to 7 days in the PK studies for conjugates with both linker structures (Figure 2c). Both conjugates can lose maytansinoid by hydrolysis of linker-payload from non-lysine linkages that can be formed in the reaction of NHS esters with proteins14 (giving


Page 14 of 37

Page 15 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


weakly cytotoxic carboxylate species; Supplementary Figure 1), which likely accounts for most of the change in DAR within the first ~ 4 days in circulation in vivo (see Figure 2c). While

Figure 3. Ab-SMCC-DM1 and Ab-(CH2)3-DM show similar loss of high DAR species. A single 10 mg/kg dose of Ab4-SMCC-DM1 and Ab4-(CH2)3-DM conjugates were administered to female CD-1 mice. ADCs were purified from plasma 7 days post-dose using affinity capture with recombinant CD56-ECD and analyzed by LC/MS.

maysine release from conjugates via β-elimination (which gives a much less potent by-product; Supplementary Figure 1) is also possible15 , this mechanism likely makes a minimal contribution to net changes in DAR given the plateau in loss of maytansinoid from Ab1-(CH2)3-DM from day 7 to day 28 (Figure 2c). The slight PK differences between SMCC-DM1 and (CH2)3-DM cannot be quantified here due to the limited sensitivity of LC-MS methods for detection of lowabundance DAR species.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

In order to help understand the potential decomposition pathways that apply to SMCC-DM1 ADCs in vivo, we monitored the fate of Ab1-SMCC-DM1 upon incubation in mouse plasma buffered with HEPES to maintain pH (7.4) ex vivo (Figure 4). Despite the limitations of ex vivo

Figure 4. Ex vivo plasma stability assessment of Ab1-SMCC-DM1 (DAR 3.5). Ab1-SMCCDM1 was incubated at 4oC or 37oC in buffered mouse plasma for the indicated times. Ab and ADC ELISAs were performed to determine the ADC concentration. Cytotoxicity assays were done to assess the potency of the samples. At 4oC, the concentration of Ab1-SMCC-DM1 was essentially stable through 35 days with no impact on potency. At 37oC, Ab1-SMCC-DM1 is stable up to 4 days, but not at 7 days. The loss of conjugate by ELISA is consistent with the observed loss of potency.

experiments (maintaining the integrity of plasma), such studies may allow identification of degradation products which may be difficult to detect in vivo due to clearance processes. The


Page 16 of 37

Page 17 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ELISA for both Ab and ADC showed no loss of material in samples incubated at 4oC, and cytotoxicity assays demonstrated that the samples retained full potency for the entire 35 day incubation period. However, after incubation in plasma at 37oC, loss of both ELISA-reactive material and cytotoxic potency was observed after day 4, with degradation becoming apparent by day 7. Prolonged incubation at 37oC should be viewed as a stress condition since it is likely that the integrity of plasma deteriorates over the 35 day incubation.

To examine the biochemical fate of these ADCs, radiolabeled Ab1-SMCC-[3H]DM1 and Ab1(CH2)3-[3H]DM incubated in mouse plasma ex vivo at 37oC were analyzed by SEC (Figure 5). By day 7, a large proportion of the SMCC-DM1 ADC was in the form of high molecular weight aggregates while the (CH2)3-DM ADC maintained a better monomeric profile. While this phenomenon was not observed in vivo (Figure 2e), we characterized the ex vivo process further. Both the SEC profile and SDS-PAGE of Ab4-SMCC-DM1 incubated in mouse plasma for 15 days, and affinity purified with recombinant CD56-ECD, clearly showed an increase in high molecular weight species (Supplementary Figure 2a). This species is proposed to be a nonreducible conjugate-albumin adduct, as determined through immunoblotting analysis against albumin under both reducing and non-reducing conditions. Similar species were formed in plasma samples of Ab4-SPP-DM1 (bearing a more labile disulfide linker17), but these adducts were reducible; purification followed by immunoblotting analysis against albumin confirms that the aggregates were indeed disulfide-linked Ab-albumin adducts (Supplementary Figure 2b). However, high molecular weight species were not detected from plasma samples of Ab4-(CH2)3DM. These observations suggest that, upon ex vivo incubation, DM1 loss from thiosuccinimide



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 5. Characterization of monomeric status of conjugate incubated with plasma ex vivo (SEC). Ab1-SMCC-[3H]DM1 and Ab1-(CH2)3-[3H]DM conjugates were incubated at 37oC in buffered mouse plasma for the indicated times. A portion of plasma was then analyzed by SECHPLC for monomeric status. The effluent was collected in 0.5 mL fractions and the radioactivity associated with each fraction was determined by LSC. Ab1-SMCC-[3H]DM1 maintains its monomeric profile up to 4 days, but loses its monomer contour by 7 days. Ab1-(CH2)3-[3H]DM maintains a better monomeric profile for at least 7 days.

or disulfide linkers can be followed by coupling to albumin, leading to formation of Ab-SMCCalbumin or observations suggest that, upon ex vivo incubation, DM1 loss from thiosuccinimide or disulfide linkers can be followed by coupling to albumin, leading to formation of Ab-SMCCalbumin or Ab-SPP-albumin conjugates. The Ab-SMCC-albumin adduct may be formed in vivo since a similar, but minor, peak corresponding to high molecular weight species was observed, either following affinity capture (data not shown), or via radioanalysis (Figure 2e, early eluting shoulder peak) of in vivo plasma samples from mice administered with Ab-SMCC-DM1.


Page 18 of 37

Page 19 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The nature of the released maytansinoid from Ab-SMCC-[3H]-DM1 could not be conclusively identified by HPLC/MS of plasma extracts from in vivo samples (data not shown), possibly due to formation of DM1 adducts with albumin, other protein thiols, or thiol compounds such as cysteine. Small molecular weight maytansinoids likely also clear quickly from circulation in vivo, further hampering their detection. We therefore measured the quantity of small molecular weight 3H-maytansinoid species released from Ab-SMCC-[3H]-DM1 and Ab-(CH2)3-[3H]DM ADCs incubated in plasma ex vivo, with and without a disulfide-reduction step (using TCEP) prior to extraction into organic solvent (Supplementary Figure 3). Small amounts of 3Hmaytansinoids were detected from the Ab-(CH2)3-[3H]DM samples, and the levels show little change with or without TCEP treatment. By contrast, for every time point, an increasingly greater amount of 3H-maytansinoid was recovered from the Ab-SMCC-[3H]-DM1 sample after TCEP reduction relative to no TCEP reduction, suggesting that DM1 could dissociate from AbSMCC-DM1, and react to form disulfides with proteins ex vivo. The overall amount of 3Hmaytansinoid released from Ab-SMCC-DM1 after TCEP reduction was 7.9% after 7 d. When normalized for the release of 3H-maytansinoid from the Ab-(CH2)3-[3H]DM sample, about 4.8% of the total release of 7.9% could be attributed to thiosuccinimide cleavage. These ex vivo results show that thiol-maleimide exchange can indeed take place, releasing the maytansinoid thiol DM1, and provide the most likely explanation for the steady loss of maytansinoid from the Ab1SMCC-DM1 conjugate in vivo that continues after day 7 through day 28 as indicated by the continuing decrease in DAR (Figure 2c).

Although the thiosuccinimide moiety formed by conjugation of a cytotoxin-maleimide to



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

endogenous cysteine residues of antibodies10 was reported to be more labile in vivo than the thiosuccinimide of Ab-SMCC-DM1 conjugates reported herein, it has also been reported that the release of the cytotoxin-maleimide from cysteine residues in antibodies could plateau because of stabilization of the conjugate by succinimide ring opening.18-20 However, prior works analyzing metabolites of Ab-SMCC-DM1 conjugates have shown no evidence for a ring-opened thiosuccinimide in released lysine-SMCC-DM112, 21, 22, suggesting that the potential for thiosuccinimide ring-opening may not be of major significance in Ab-SMCC-DM1 conjugate stability. Indeed, the PK analysis with Ab1-SMCC-DM1 suggested that the loss of maytansinoid from the SMCC-DM1conjugate did not appear to reach a plateau over 28 days in circulation (Figure 2c). Similarly, in the case of auristatin conjugates linked to endogenous cysteine residues derived from interchain cystine residues in Abs, the DAR of such conjugates continues to drop over the two week experimental period10, again suggesting that the rate of succinimide ring-opening is too slow to arrest decline in DAR. Consistent with these observations, it was reported that the rates of ring-opening in conjugates prepared with some commonly used maleimides were too slow to serve as prevention against thiol exchange.23 However, the thiosuccinimide linkages from conjugates made with certain engineered cysteine sites are clearly prone to hydrolytic ring-opening of the thiosuccinimide, with a clear plateau in the loss of payload.20 Building on this finding, rapid thiosuccinimide hydrolysis could be purposely achieved by incorporating a basic amino group adjacent to the maleimide, positioned to provide intramolecular catalysis of thiosuccinimide ring hydrolysis.18

While thiosuccinimide adducts can be reliably made from thiols of varying structure, it is known that their stability is strongly dependent on the pKa of the thiol, with thiosuccinmides


Page 20 of 37

Page 21 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


prepared from thiols with higher pKa s being more stable.7 For example, the thiosuccinimide derivative of N-acetyl cysteine (thiol pKa 9.5) was found to be less stable than the corresponding 3-mercaptopropionic acid derivative (thiol pKa 10.3).7 The pKa of the DM1 thiol was calculated to be ~9.8 (Advanced Chemistry Development (ACD/Labs) Software V11.02), and the pKa of the free cysteine thiol is ~8.6.24 We experimentally compared the buffer stability of different thiolmaleimide adducts as Ab conjugates. The same Ab was conjugated with a commercially available BODIPY-maleimide, either through endogenous cysteine residues (generated via reduction of cystines), or by maleimide labeling of 3-thiobutanoyl groups attached to the Ab via NHS-ester mediated acylation of lysine residues (Figure 6). These two conjugates were then incubated in buffer at 37oC in the presence of 5 mM cysteine as a maleimide scavenger, and the amount of BODIPY-cysteine adduct was measured by an HPLC-based assay. The thiobutanoyllinked conjugate showed minimal loss (0.5%) of BODIPY over the 7-day incubation. Over the same time period, ~8% of the linked BODIPY was lost from the BODIPY conjugate made by alkylation of Ab cysteines. The difference in the inherent chemical stability of the two types of thiol-maleimide adducts is consistent with the expected differences in the pKa of the different thiols.

It has been reported that the pKa of a cysteine thiol in a protein may be influenced by the three-dimensional protein structure.24 25

Indeed, variations in stability of mc-VC-PAB-MMAE

conjugates made with engineered antibodies with cysteine residues introduced at different positions have been observed.20 The BODIPY assay described herein could be a useful tool to evaluate and compare the susceptibility of thiosuccinimides introduced at various site-specific cysteines in Abs to cleavage.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 6. Variable stability of thiol-maleimide linkages. Antibodies were conjugated with BODIPY-maleimide, either via 3-thiobutanoyl linker (lysine conjugate, 4.3 BODIPY/Ab) or directly to the reduced interchain disulfides (cysteine conjugate, 3.9 BODIPY/Ab) as shown. BODIPY conjugates were incubated with 5 mM cysteine in PBS at 37oC for 7 days and the loss of BODIPY was monitored by RP-HPLC. Cysteine-maleimide adducts are less stable compared to the alkyl thiol-maleimide adduct.

Two mechanisms have been proposed to account for the reversibility of the thiol-maleimide linkage, as shown in Figure 7. First, a thiosuccinimide-linked ADC was reported to release maytansinoid (DM1-SO3-) following chemical oxidation and sulfoxide elimination under oxidative conditions.9 The second mechanism proposed is the cleavage of the thiosuccinimide linkage by retro-Michael reaction and exchange with albumin and other thiols present in plasma, a mechanism that has been reported to occur for ADCs generated by direct maleimide alkylation of endogenous Ab cysteine residues10 (Figure 7). The maleimide from Ab-SMCC-DM1 generated via either of these two degradation mechanisms could react with the free thiol of


Page 22 of 37

Page 23 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


albumin to form the observed Ab-SMCC-albumin (Supplementary Figure 2 and Figure 2e). These two mechanisms are not mutually exclusive, although the second mechanism may be more likely since DM1-SO3- is not detected from Ab-SMCC-DM1 samples incubated in plasma ex vivo without oxidative stress (data not shown) or from in vivo samples.8

Figure 7. Proposed mechanisms for Ab-SMCC-DM1 loss in plasma. Literature suggests two mechanisms for thiosuccinimide cleavage: oxidative cleavage,9 or retro-Michael β-elimination as has been proposed for cysteine-conjugated ADCs.10

We show here that Ab-SMCC-DM1 ADCs exhibit favorable pharmacokinetic properties, exhibiting only slightly faster clearance than a comparator ADC lacking a thiosuccinimide linkage (Figure 2 b). We also provide evidence that the DM1-thiosuccinimide in Ab-SMCCDM1 ADCs is less labile in vivo as compared with published information on the cleavage of



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

thiosuccinimides formed by reaction of maleimide-cytotoxins with thiols derived by reduction of endogenous cystines of Abs, as expected based on the established relationship between stability and thiol pKa. The experimental results described here will help inform the design of ADC linker chemistries which utilize thiol-maleimide reactions.


Page 24 of 37

Page 25 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


EXPERIMENTAL PROCEDURES

Antibodies, ADCs, and reagents. Humanized monoclonal IgG1 Abs recognizing human EGFR, CanAg, FOLR-1, and CD56 (referred to as Ab1, Ab2, Ab3, and Ab4, respectively) were generated at ImmunoGen, Inc. All lack cross-reactivity with their respective murine antigens. DM1- ADCs utilizing two linkers, SMCC and SPP, were prepared and characterized according to previously described protocols.26, 27 The Ab-(CH2)3-DM conjugate was prepared by reacting the humanized monoclonal Ab with DM-(CH2)3-NHS ester in solution buffered to pH 8 in the presence of 10% dimethylacetamide (DMA) overnight, and then purified over a G-25 Sephadex column (DNA Grade, GE Healthcare). A polyclonal chicken Ab against human albumin was obtained from Abcam (Cambridge, MA). Human antigen-Fc fusion proteins were generated at ImmunoGen, Inc., by expression of DNA constructs derived from DNA encoding extracellular antigen domains fused with a murine IgG2A Fc domain (huEGFR-ECD-Fc, huCD56-ECD-Fc).

Chemical synthesis. The DM-(CH2)3-NHS ester was prepared from maytansinol-N-methyl alanine ester by acylation with glutaric anhydride followed by activation of the resulting carboxylic acid and purification by RP-HPLC. Details are provided in the supporting information for methods. Maysine28, S-methyl-DM429, S-methyl-DM129, DM1-MCCA29 and maytansinoids bearing 3H at the C20 methoxy group21 were prepared as previously described.

Radiolabeling of Antibody and ADCs. Abs and ADCs were trace-labeled on the protein moiety with N-succinimidyl-[2,3,3H]propionate as described by Lai et al.16 All Abs and conjugates had less than 1% unconjugated tritium label, with 0.01 (Ab1), 0.009 (Ab1-SMCC-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

DM1) and 0.004 (Ab1-(CH2)3-DM) moles of the label incorporated per mole of Ab, and had specific radioactivities of 0.82 (Ab1), 0.71 (Ab1-SMCC-DM1) and 0.34 (Ab1-(CH2)3-DM) Ci/mmol Ab respectively. Preparation of ADCs with radiolabeled maytansinoid was carried out at ImmunoGen, Inc., following published procedures.17

Animals. Female CD-1 mice (Charles River Laboratory) were used for all in vivo studies. Mice were housed in a clean barrier facility in standard rodent micro isolator cages. Animal work was carried out in accordance with ImmunoGen’s Animal Care and Use Committee and the Guide for the Care and Use of Laboratory Animals of the U.S. National Institutes of Health.

Pharmacokinetic studies. Animals (n = 9 per group) were administered with maytansinoidADCs at a single dose of 10 or 12 mg/kg (Ab concentration) via intravenous bolus injection (see Figure legends for specific conjugate and dose). At selected intervals up to 35 days following dosing, blood (~70 µL) was collected on a rotational basis, with n = 3 animals per time point. Blood samples were processed for plasma by centrifugation. Plasma samples were stored at 60oC to -80oC in polypropylene collection tubes until assayed for Ab and ADC concentrations by counting of radioactivity. Pharmacokinetic analyses of the Ab and ADC plasma concentrations at various times post-injection were performed using the standard algorithms of the noncompartmental pharmacokinetic analysis program, WinNonlin, Professional version 6.1 (Pharsight, Mountain View, CA). Nominal sample collection times and dose concentrations were used in the data analysis. The values of the first order rate constant for determining conjugate t1/2 were evaluated by using the concentration data from 1 to 28 days post-administration.


Page 26 of 37

Page 27 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Liquid scintillation counting (LSC). Radioactivity in plasma was measured by first digesting samples in Solvable (1 mL, PerkinElmer, Torrance, CA). Samples were bleached by an addition of EDTA (100 mM, 0.1 mL) and 30% H2O2 (0.3 mL), and incubated at room temperature overnight. Ultima Gold scintillant (15 mL, PerkinElmer) was then added to each vial along with HCl (1 N, 0.25 mL). Vials were shaken and stored overnight in the dark. Each vial was counted for 5 min. in a Tri-Carb 2900T liquid scintillation counter (Packard BioScience, Downer Grove, IL). The Ab or ADC concentration in plasma was then calculated using the specific radioactivity of the particular Ab or ADC.

Size exclusion chromatography (SEC)-HPLC. Plasma samples were mixed with an equal volume of SEC mobile phase (170 mM potassium phosphate, 212.5 mM KCl, and 15% isopropanol, pH 7.0), and this mixture (100 µL) was injected onto an Agilent 1100 HPLC system equipped with a TSK G3000SWXL 7.8 mm x 300 mm column (Tosoh Biosciences) at a flow rate of 0.5 mL/min with a run time of 30 min. The effluent was collected in 0.5 mL fractions and the radioactivity associated with each fraction was determined by mixing with Ultima Gold liquid scintillation cocktail (4 mL) before counting for 5 min. in a Tri-Carb 2900T liquid scintillation counter.

Affinity capture LC-MS. CD56-ECD-Fc was biotinylated using EZ Link NHS-LC-Biotin (Pierce) dissolved in dry DMA at 25 mM. The reaction was performed at room temperature for 3 h at 2 mg/mL CD56-ECD-Fc in 50 mM potassium phosphate, 2 mM EDTA, pH 7.5, containing DMA (5%), using an excess of 12 moles of biotinylating reagent to CD56-ECD-Fc dimer. The biotinylated protein was purified into PBS, pH 7.4, using a G25 Sephadex column. The final



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

material was assayed for protein concentration by UV/Vis spectrometry (using an extinction coefficient of 273485 M-1 cm-1 (at 280 nm) and a molecular weight of 205464 g/mol for CD56ECD -Fc dimer), and for biotin load by the HABA assay (Pierce) according to the manufacturer’s protocol.

Commercially available xMag-Streptavidin Microparticles (Biochain, CA) were washed with washing buffer (50 mM Tris·HCl, 0.15 M NaCl, pH 8.0) twice and resuspended in the same buffer to their original volume. Biotinylated CD56-ECD-Fc (116 µg) was then added to the streptavidin particles (200 µL) and rotated at room temperature for 2 h. The beads were washed 3 times with washing buffer and re-suspended to their original volume in washing buffer containing 0.4% Tween 20.

Plasma samples containing around 70 µg of ADC were added to streptavidin-biotin- CD56ECD-Fc particles (200 µL per sample). After gentle shaking at room temperature for 2 h, the resin was washed 3 times with washing buffer (1 mL) and eluted using 1:1 (v/v) mixture of 0.1 M sodium citrate (pH 3.0) and ethylene glycol (50 µL). The eluent was immediately neutralized with Tris·HCl, pH 8.5 (1 M, 9 µL), and then analyzed by SEC or SEC-LC/MS as described previously.30

Ex vivo stability studies. The ex vivo stability of Ab-SMCC-DM1 was evaluated in pooled mouse plasma buffered by the addition of 0.025 volumes of 1 M HEPES buffer, pH 7.4. AbSMCC-DM1 conjugates were diluted to a starting concentration of 100 µg/mL (concentration of Ab), with each Ab having an average of 3.4 DM1 molecules conjugated. Aliquots were


Page 28 of 37

Page 29 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


incubated at 4 and 37oC for 0, 1, 2, 4, 7, 10, 14, 21, 28 and 35 days. Separate ELISA methods to determine total Ab and conjugate concentration were performed on samples from each temperature at all the listed time points. Additionally, some samples were assessed for potency using a WST cytotoxicity assay (Dojindo Molecular Technologies, Inc.) in accordance with the manufacturer’s protocol.

ELISA for determination of the concentration of Ab1-SMCC-DM1 (ADC ELISA). The concentration of Ab1-SMCC-DM1 was determined using a sandwich ELISA by coating Costar high binding plates (Corning Incorporated, Corning, NY) with a murine monoclonal antimaytansine Ab (ImmunoGen, Inc., 1 µg/mL) overnight at 4oC. The plates were then washed using Tris-buffered saline (TBS) containing 0.05% Tween-20 and blocked using TBS containing 0.5% bovine serum albumin at room temperature for at least one hour. The standards, internal controls and diluted samples were incubated with coated plates for one hour at 37oC. The plates were washed and then incubated with horseradish peroxidase (HRP)-conjugated donkey antihuman IgG (Jackson Immunoresearch Laboratories, West Grove, PA) at room temperature for one hour. Bound donkey anti-human IgG-HRP was detected by incubating with tetramethylbenzidine (Bio FX Laboratories, Randallstown, MD) for 5 min at room temperature before adding stop solution (BioFX), then read on a EL808 Ultra Microplate reader (Bio-Tek, Winooski, Vermont) at 450 nm. Data was analyzed using the Kineticalc software (Bio-Tek), using a four parameter logistic (4pl) fit for the absorbance data.

ELISA for determination of the concentration of the Ab component of Ab1-SMCC-DM1 (Ab ELISA). The Ab1 component of Ab1-SMCC-DM1 in plasma samples was detected using ELISA by coating Costar high binding plate with huEGFR-ECD-Fc (1.5 µg/mL) in carbonate



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

buffer (Sigma) at 4oC overnight. The plates were then washed using TBS buffer containing 0.1% Tween-20 and blocked using TBS containing 0.5% casein (Sigma) at room temperature for at least one hour. The standards, internal controls and diluted samples were incubated with coated plate for one hour at room temperature. The plate was then incubated with horseradish peroxidase (HRP)-conjugated goat anti-human IgG (Jackson Immunoresearch Laboratories) (1:20,000 in TBS, 0.5% casein). Following a one-hour incubation at room temperature, unbound secondary Ab reagent was washed from the wells, and the plate was developed by incubating with tetramethylbenzidine for 10 min at room temperature before adding stop solution, then read at 450 nm on a EL808 Ultra Microplate reader (Bio-Tek) and analyzed as described above.

Characterization of ADC-albumin adduct by Western blot. ADCs purified by affinity capture from plasma samples, or fractionated samples from SEC-HPLC of thus purified ADCs, were treated with or without 25 mM dithiothreitol (DTT) for 1 h at room temperature for reduced and non-reduced analysis, respectively. After adding SDS-gel loading buffer, samples were heated at 90oC for ~5 min, separated by a NuPAGE 4-12% Bis-Tris polyacrylamide gel, transferred to a nitrocellulose membrane, and analyzed by anti-albumin immunoblot according to standard procedures.

SEC-HPLC and free maytansinoid analysis of ex vivo plasma sample incubated with radiolabeled conjugates. The ex vivo stability of Ab-SMCC-[3H]DM1 and Ab-(CH2)3-[3H]DM were evaluated in pooled buffered mouse plasma (25 mM HEPES, pH 7.4) at 100 µg/mL (concentration of Ab protein) at 37oC. Aliquots were taken at 0, 2, 4 and 7 days and kept at 80oC. A portion of each aliquot was analyzed by SEC-HPLC for monomeric status as


Page 30 of 37

Page 31 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


aforementioned. Other portions, with or without reduction (5 mM TCEP, 25oC, 3 h) and alkylation (12 mM N-ethylmaleimide, room temperature, 1 h), were mixed with three volume equivalents of ice-cold acetone and kept at -80oC for at least 1 h or until further processing. Precipitated protein was removed by centrifugation at 3000 g for 15 min and the supernates were counted for radioactivity by LSC. The net radioactivity following subtraction of the radioactivity for free maytansinoid associated with conjugates prior to incubation (time 0, no TCEP treatment) was divided by the total radioactivity from the same portion of the plasma to give the percentage of free maytansinoid.

Preparation of BODIPY-conjugated antibodies. Ab-Cys-mal-BODIPY (“cysteine linked”): Ab (6 mg/mL) in PBS buffer, pH 7.4, was reduced with DTT (8 mM) for 45 min at 37oC to generate free thiols from the interchain disulfides. The excess reducing agent was removed using Sephadex G-25 column equilibrated in PBS, pH 7.4. The reduced Ab was immediately reacted with 15 molar equivalents of BODIPY® FL N-(2-aminoethyl)maleimide (B-10250, Invitrogen) for 30 min followed by a second purification using Sephadex G-25 column. Ab-butoxy-S-mal-BODIPY (“lysine linked”): Ab (5 mg/mL) in 50 mM EPPS buffer, pH 8.0, was reacted with 6 molar equivalents of SPDB linker for 90 min at room temperature. This was followed by treatment with 10 molar equivalents of DTT for 10 min at room temperature to remove the thiopyridine group of SPDB without reducing the interchain disulfides of the Ab. The resulting modified Ab was purified using Sephadex G-25 column equilibrated in 10 mM sodium citrate, pH 5.5. The free thiols on the Ab were then reacted immediately with 12 molar equivalents of BODIPY® FL N-(2-aminoethyl)maleimide for 30 min at room temperature and the conjugate was purified using Sephadex G-25 column equilibrated in PBS, pH 7.4.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Thiosuccinimide stability measurements of Ab-BODIPY conjugates. Ab-Cys-malBODIPY or Ab-butoxy-S-mal-BODIPY (both at 10 µM in PBS pH 7.4, average of 3.7-4 BODIPY per Ab) was incubated for 7 d at 37oC with or without 5 mM L-cysteine. Increase in free BODIPY species was quantified using a mixed-mode chromatography method.31 Briefly, a HISEP shielded hydrophobic phase column (5 µm particle size, 4.6 × 250 mm length, Supelco, Bellefonte, PA, USA) was used. Mobile phase A consisted of 100 mM ammonium acetate, pH 7.0, and mobile phase B was 100% acetonitrile. The column was equilibrated at 25% B followed by a linear gradient over 25 min to 40% B after sample injection at a flow rate of 0.7 mL/min. Intact conjugate eluted between 2-5 min while released BODIPY was detected between 10-25 min. The unconjugated BODIPY was determined as a percentage of total integrated area at 506 nm (λmax for BODIPY).


Page 32 of 37

Page 33 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Abbreviations: ADC, antibody-drug conjugate; Ab, antibody; NHS, N-hydroxysuccinimidyl; SMCC, N-succinimidyl 4-(N-maleimidomethyl)cyclohexane-1-carboxylate; PK, pharmacokinetics; SPP, N-succinimidyl 4-(2-pyridyldithio)pentanoate; SPDB, N-succinimidyl4-(2-pyridyldithio)butyrate; DMA, dimethylacetamide; LSC, liquid scintillation counting; TCEP, tris(2-carboxyethyl)phosphine; SEC, size exclusion chromatography; PBS, phosphatebuffered saline; ECD, extracellular domain; HRP, horseradish peroxidase; DTT, dithiothreitol; TBS, Tris-buffered saline; DAR, cytotoxic agent/antibody molar ratio; DM1, N2'-deacetyl-N2'(3-mercapto-1-oxopropyl) maytansine

Supporting Information The Supporting Information is available free of charge, including chemical synthesis of DM(CH2)3-NHS ester and supplemental figures.

Corresponding Author *Correspondence to Xiuxia Sun, ImmunoGen Inc 830 Winter street, Waltham, MA 02451. Tel: 781895-0758. Fax 781-895-0611. E-mail: [email protected] References (1)

Senter, P. D., and Sievers, E. L. (2012) The discovery and development of brentuximab vedotin for use in relapsed Hodgkin lymphoma and systemic anaplastic large cell lymphoma. Nature biotechnology 30, 631-7. (2) Newland, A. M., Li, J. X., Wasco, L. E., Aziz, M. T., and Lowe, D. K. (2013) Brentuximab vedotin: a CD30-directed antibody-cytotoxic drug conjugate. Pharmacotherapy 33, 93-104. (3) Ballantyne, A., and Dhillon, S. (2013) Trastuzumab emtansine: first global approval. Drugs 73, 755-65. (4) Lambert, J. M., and Chari, R. V. (2014) Ado-trastuzumab Emtansine (T-DM1): an antibody-drug conjugate (ADC) for HER2-positive



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

(14)

(15)

(16)

Page 34 of 37

breast cancer. Journal of medicinal chemistry 57, 6949-64. Amiri-Kordestani, L., Blumenthal, G. M., Xu, Q. C., Zhang, L., Tang, S. W., Ha, L., Weinberg, W. C., Chi, B., Candau-Chacon, R., Hughes, P., et al. (2014) FDA approval: ado-trastuzumab emtansine for the treatment of patients with HER2-positive metastatic breast cancer. Clinical cancer research : an official journal of the American Association for Cancer Research 20, 4436-41. Gordon, M., Canakci, M., Li, L., Zhuang, J., Osborne, B. A., and Thayumanavan, S. (2015) A Field Guide to Challenges and Opportunities in Antibody-Drug Conjugates for Chemists. Bioconjugate chemistry. Baldwin, A. D., and Kiick, K. L. (2011) Tunable degradation of maleimide-thiol adducts in reducing environments. Bioconjugate chemistry 22, 1946-53. Shen, B. Q., Bumbaca, D., Saad, O., Yue, Q., Pastuskovas, C. V., Khojasteh, S. C., Tibbitts, J., Kaur, S., Wang, B., Chu, Y. W., et al. (2012) Catabolic fate and pharmacokinetic characterization of trastuzumab emtansine (T-DM1): an emphasis on preclinical and clinical catabolism. Current drug metabolism 13, 901-10. Fishkin, N., Maloney, E. K., Chari, R. V., and Singh, R. (2011) A novel pathway for maytansinoid release from thioether linked antibody-drug conjugates (ADCs) under oxidative conditions. Chemical communications 47, 10752-4. Alley, S. C., Benjamin, D. R., Jeffrey, S. C., Okeley, N. M., Meyer, D. L., Sanderson, R. J., and Senter, P. D. (2008) Contribution of linker stability to the activities of anticancer immunoconjugates. Bioconjugate chemistry 19, 759-65. Girish, S., Gupta, M., Wang, B., Lu, D., Krop, I. E., Vogel, C. L., Burris Iii, H. A., LoRusso, P. M., Yi, J. H., Saad, O., et al. (2012) Clinical pharmacology of trastuzumab emtansine (T-DM1): an antibody-drug conjugate in development for the treatment of HER2positive cancer. Cancer chemotherapy and pharmacology 69, 1229-40. Erickson, H. K., Lewis Phillips, G. D., Leipold, D. D., Provenzano, C. A., Mai, E., Johnson, H. A., Gunter, B., Audette, C. A., Gupta, M., Pinkas, J., et al. (2012) The effect of different linkers on target cell catabolism and pharmacokinetics/pharmacodynamics of trastuzumab maytansinoid conjugates. Molecular cancer therapeutics 11, 1133-42. Bender, B., Leipold, D. D., Xu, K., Shen, B. Q., Tibbitts, J., and Friberg, L. E. (2014) A mechanistic pharmacokinetic model elucidating the disposition of trastuzumab emtansine (T-DM1), an antibody-drug conjugate (ADC) for treatment of metastatic breast cancer. The AAPS journal 16, 994-1008. Chih, H. W., Gikanga, B., Yang, Y., and Zhang, B. (2011) Identification of amino acid residues responsible for the release of free drug from an antibody-drug conjugate utilizing lysinesuccinimidyl ester chemistry. Journal of pharmaceutical sciences 100, 2518-25. Wakankar, A., Chen, Y., Gokarn, Y., and Jacobson, F. S. (2011) Analytical methods for physicochemical characterization of antibody drug conjugates. mAbs 3, 161-72. Lai, K. C., Deckert, J., Setiady, Y. Y., Shah, P., Wang, L.,


Page 35 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Chari, R., and Lambert, J. M. (2015) Evaluation of Targets for Maytansinoid ADC Therapy Using a Novel Radiochemical Assay. Pharmaceutical research. (17) Xie, H., Audette, C., Hoffee, M., Lambert, J. M., and Blattler, W. A. (2004) Pharmacokinetics and biodistribution of the antitumor immunoconjugate, cantuzumab mertansine (huC242-DM1), and its two components in mice. The Journal of pharmacology and experimental therapeutics 308, 1073-82. (18) Lyon, R. P., Setter, J. R., Bovee, T. D., Doronina, S. O., Hunter, J. H., Anderson, M. E., Balasubramanian, C. L., Duniho, S. M., Leiske, C. I., Li, F., et al. (2014) Self-hydrolyzing maleimides improve the stability and pharmacological properties of antibodydrug conjugates. Nature biotechnology 32, 1059-62. (19) Tumey, L. N., Charati, M., He, T., Sousa, E., Ma, D., Han, X., Clark, T., Casavant, J., Loganzo, F., Barletta, F., et al. (2014) Mild method for succinimide hydrolysis on ADCs: impact on ADC potency, stability, exposure, and efficacy. Bioconjugate chemistry 25, 1871-80. (20) Shen, B. Q., Xu, K., Liu, L., Raab, H., Bhakta, S., Kenrick, M., Parsons-Reponte, K. L., Tien, J., Yu, S. F., Mai, E., et al. (2012) Conjugation site modulates the in vivo stability and therapeutic activity of antibody-drug conjugates. Nature biotechnology 30, 184-9. (21) Erickson, H. K., Widdison, W. C., Mayo, M. F., Whiteman, K., Audette, C., Wilhelm, S. D., and Singh, R. (2010) Tumor delivery and in vivo processing of disulfide-linked and thioether-linked antibody-maytansinoid conjugates. Bioconjugate chemistry 21, 8492. (22) Sun, X., Widdison, W., Mayo, M., Wilhelm, S., Leece, B., Chari, R., Singh, R., and Erickson, H. (2011) Design of antibodymaytansinoid conjugates allows for efficient detoxification via liver metabolism. Bioconjugate chemistry 22, 728-35. (23) Fontaine, S. D., Reid, R., Robinson, L., Ashley, G. W., and Santi, D. V. (2015) Long-term stabilization of maleimide-thiol conjugates. Bioconjugate chemistry 26, 145-52. (24) Roos, G., Foloppe, N., and Messens, J. (2013) Understanding the pK(a) of redox cysteines: the key role of hydrogen bonding. Antioxidants & redox signaling 18, 94-127. (25) Cremers, C. M., and Jakob, U. (2013) Oxidant sensing by reversible disulfide bond formation. The Journal of biological chemistry 288, 26489-96. (26) Widdison, W. C., Wilhelm, S. D., Cavanagh, E. E., Whiteman, K. R., Leece, B. A., Kovtun, Y., Goldmacher, V. S., Xie, H., Steeves, R. M., Lutz, R. J., et al. (2006) Semisynthetic maytansine analogues for the targeted treatment of cancer. Journal of medicinal chemistry 49, 4392-408. (27) Singh, R., and Erickson, H. K. (2009) Antibody-cytotoxic agent conjugates: preparation and characterization. Methods in molecular biology 525, 445-67, xiv. (28) Kupchan, S. M., Komoda, Y., Branfman, A. R., Sneden, A. T., Court, W. A., Thomas, G. J., Hintz, H. P., Smith, R. M., Karim, A., Howie, G. A., et al. (1977) The maytansinoids. Isolation,



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

structural elucidation, and chemical interrelation of novel ansa macrolides. The Journal of organic chemistry 42, 2349-57. (29) Widdison, W., Wilhelm, S., Veale, K., Costoplus, J., Jones, G., Audette, C., Leece, B., Bartle, L., Kovtun, Y., and Chari, R. (2015) Metabolites of Antibody-Maytansinoid Conjugates: Characteristics and in Vitro Potencies. Molecular pharmaceutics. (30) Lazar, A. C., Wang, L., Blattler, W. A., Amphlett, G., Lambert, J. M., and Zhang, W. (2005) Analysis of the composition of immunoconjugates using size-exclusion chromatography coupled to mass spectrometry. Rapid communications in mass spectrometry : RCM 19, 1806-14. (31) Fleming, M. S., Zhang, W., Lambert, J. M., and Amphlett, G. (2005) A reversed-phase high-performance liquid chromatography method for analysis of monoclonal antibody-maytansinoid immunoconjugates. Analytical biochemistry 340, 272-8.


Page 36 of 37

Page 37 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For Table of Contents Only


Understanding How the Stability of the Thiol ... - ACS Publications

Recommend Documents