Advances in the Development of Dual-Drug Antibody Drug Conjugates

Abstract
Antibody drug conjugates (ADCs) are touted for their ability to site selectively deliver a small molecule chemotherapeutic directly to a tumor cell, bypassing off target side effects from systemic circulation. To date, twelve ADCs have been FDA approved in the United States and >200 more in the clinical pipeline (clinicaltrials.gov).

While ADCs have proven successful in both solid and hematological cancers, resistance and tumor heterogeneity are major causes of failure clinically.[1] Tumor heterogeneity is known to lead to recurrence, metastasis, and acquired resistance to ADCs and other therapeutic strategies. Heterogenous tumors with differential drug sensitivities result in aggressive tumor growth, high relapse rates, and poor survival.[1]

To combat these challenges, the majority of chemotherapeutic regimens consist of a combination of drugs. Co-delivery of small molecules can overcome resistance, generate additive or synergistic effects, and enhance therapeutic efficacy.[2][3]

Emergence of tumors refractory to current therapies has given impetus to the evaluation of new ADC formats. This challenge has led to the exploration of dual-drug ADCs capable of delivering two mechanistically distinct payloads simultaneously. Strategies for the construction of dual-drug ADCs involve attachment of both drugs to one linker or through the use of two different conjugation sites on the antibody. Herein, we will review the synthesis and evaluation of the dual-drug ADCs reported to date.

We focus on ADCs constructed by conjugation of linkers directly to antibody scaffolds and not to other formats or targeting molecules.

Dual-drug antibody drug conjugates built on heterofunctional linkers
The most common strategy for dual-drug ADC synthesis has involved the installation of both drugs onto the same linker. In 2017, Levengood et al. disclosed the first dual-drug ADC bearing auristatin derivatives monomethyl auristatin E (MMAE) and monomethyl auristatin F (MMAF) on a short peptide linker bearing orthogonally protected cysteine amino acids (Figure 1A).[3]

To synthesize the ADC, the interchain disulfides of a native antibody are first reduced with TCEP followed by conjugation of dual-Cys linker 1 through Michael addition of the free thiols into the pendant maleimide of the linker. This results in the installation of eight linkers onto the antibody to ultimately append 16 total drugs per antibody. Subsequent sequential unmasking of the cysteine residues allows for conjugation of two different drugs. To install the first drug, the linker-mAb construct is treated with TCEP to remove the thio-isopropyl disulfide protecting group on the first linker cysteine.

Once exposed, the free thiol conjugates via a maleimide to the first drug, MMAF. Next, the mono-drug ADC is treated with aqueous mercury acetate to deprotect the acetamidomethyl group on the second linker cysteine followed by Quadrasil MP resin purification to remove any remaining mercury. Finally, reaction with the maleimide-MMAE drug construct leads to the desired dual-drug ADC with a total drug to antibody ratio (DAR) of 16 and a 1:1 ratio of drugs.

Two tubulin-targeting auristatin derivatives, MMAE and MMAF, were chosen by Levengood et al. for their differing physiochemical properties to allow for complementary activities in tumor cells.[3] MMAE is a potent chemotherapeutic already validated through its use in four approved ADCs. It is notable due to its cell permeability, leading to bystander activity in neighboring cell types that may not express high levels of target antigen. However, it is a substrate for drug exporters commonly upregulated in resistant tumor cells. In contrast, MMAF (employed in one current ADC) has minimal cell permeability but has known activity on multi-drug resistant cell lines. Together, MMAE should be able to combat differential antigen expression in heterogenous tumors while MMAF remains potent in resistant cells.

In resistant tumor models in vitro, the MMAE/F dual-drug ADC showed comparable potency relative to the MMAF single-drug ADC control (IC50 0.7 and 0.8 ng/mL respectively) where the MMAE single-drug ADC was ineffective (IC50 > 2000 ng/mL). In vivo, 3:5 cures were achieved with the dual-drug conjugate compared to only 1:5 cures with the MMAF control where a cure was determined by durable regression at the study endpoint (45-60 days post tumor implantation). In a bystander tumor model, MMAE dominated the effect both in vitro and in vivo with no significant difference between the dual-drug ADC and MMAE single-drug control.

Figure 1. Strategies for the construction of dual-drug ADCs on one multi-functional linker. A) Conjugation of a differentially protected cysteine linker to native antibody residues allows for sequential drug conjugation. B) A heterotrifunctional linker enables sequential conjugation via two different drug-moieties. C) Orthogonal azide and methyltetrazine functional groups allow for a one pot conjugation of two different drugs after linker conjugation to the antibody. Structures are shown with single modifications for figure clarity; both heavy chains are expected to have similar modification and conjugation.

In 2018, a second dual-drug ADC was reported by Kumar et al. containing two drugs with different cellular targets, MMAE and a pyrrolobenzodiazepine (PBD) dimer. [2] In this study, a heterotrifunctional linker was synthesized to again allow conjugation of both drugs to the same site on the antibody. Specifically, the linker contained a self-stabilizing N-aryl maleimide for site-specific conjugation to engineered thiols on the antibody.

For drug conjugation, alkyne and ketone moieties were installed for copper-mediated azide-alkyne cycloaddition (CuAAC) and aminooxy reaction via an oxime linkage respectively (Figure 1B). First, engineered cysteine 293 was revealed through TCEP reduction and dehydroascorbic acid (DHAA) re-oxidation before thiol-maleimide conjugation with heterotrifunctional linker 2 and purification by ceramic hydroxyapatite (CHT) chromatography. The first drug (MMAE) was conjugated by oxime ligation onto the linker ketone by reaction overnight and purified by CHT chromatography. Finally, PBD-SG3557 was installed through CuAAC and purification by CHT chromatography. In total, this strategy requires three synthetic steps to be performed on the antibody, with purification after each step. The MMAE/PBD dual-drug conjugate was evaluated in HER2-expressing MDA-MB-453 cells in comparison to the single drug controls. The dual-drug ADC was equipotent to the PBD single-drug ADC control, suggesting that the higher potency of the PBD drug dominated the effect of the ADC.

A third dual-drug ADC was constructed by Yamazaki et al. in 2021 using an alternative heterotrifunctional linker.[1] Specifically, a branched ADC linker built from a central lysine residue was designed to contain one or two azides for strain-promoted azide-dibenzocyclooctyne (DBCO) cycloaddition and an orthogonal clickable methyltetrazine for trans-cyclooctene (TCO) cycloaddition (Figure 1C). As in the Levengood study, MMAE and MMAF were chosen to evaluate the effects due to complementary mechanisms in bystander activity and resistance. To synthesize the dual drug ADC, the linker was first installed on the antibody through the pendant primary amine from the lysine residue via microbial transglutaminase (MTGase) mediated conjugation to the glutamine side chain at position Q295 within an N297A anti-HER2 monoclonal antibody (Figure 1C). Next, consecutive reactions with TCO-MMAF and DBCO-MMAE in one pot yielded dual-drug ADC. In total, dual-drug ADCs with DAR MMAE/F 2+2, 4+2, and 2+4 as well as mono-drug ADCs with a DAR of 2, 4, or 6 were synthesized.

The potency of the MMAE/F dual-drug ADCs were assessed in vitro in models for heterogenous HER2 expression and resistance. As anticipated, in J1MT-1 cells that were chosen for their known resistance to hydrophobic drugs, DAR 2 MMAE had lower efficacy than DAR 2 MMAF single-drug conjugates (IC50 1.023 and 0.213 nM, respectively). Promisingly, DAR 6 dual-drug ADCs (IC50 0.26 or 0.24 nM) had the highest efficacy in this cell line in comparison to DAR 6 MMAE single-drug ADC (IC50 0.060 nM). The MMAE/F 2+2 dual drug ADC performed comparably to DAR 4 MMAF single-drug ADC (0.045 and 0.36 nM, respectively). In JIMT-1(MDR1+) cells with artificially induced drug resistance, the MMAE/F 2+4 dual drug conjugate performed similarly to DAR 4 MMAF (IC50 0.017 and 0.012 nM, respectively). In contrast, the MMAE/F 4+2 dual-drug ADC had lower potency in JIMT-1(MDR1+) (IC50 0.027), indicating that the ratio of MMAF plays an important role in overcoming resistance.

The dual-drug ADCs were further evaluated in vivo in a resistant and heterogenous tumor model combining resistant HER2-positive JIMT-1 cells with HER2-negative MDA-MB-231 cells.[1] Tumors established from this combination of cells represent aggressive growth, heterogenous HER2 populations, and moderate resistance, requiring treatments that are capable of both evading drug efflux and exhibiting bystander effects in neighboring cell populations. Notably, MMAE/F 4+2 dual-drug ADC provided the most promising results with complete remission and no tumor regrowth in all mice at the end of the study when dosed at 3 mg/kg ADC compared to remission but only 3/5 survivals with MMAE DAR 4 ADC or partial tumor regrowth and 2/5 survivals with a 1:1 mixture of DAR 4 single-drug ADCs. At a dose of 1 mg/kg, the MMAE/F 4+2 dual-drug ADC demonstrated significantly higher antitumor activity when compared to MMAE DAR 4 ADC (P=0.0268 at Day 38) or a 1:1 mixture of DAR 4 single-drug ADCs (P=0.0006 at Day 38). Results from Yamazaki et al. showed that dual-drug ADC performed better than DAR matched single drug ADC or combination treatment with two single-drug ADCs, highlighting the importance of simultaneous dosing of two differing chemotherapeutic drugs.[1]

Dual-drug antibody drug conjugates constructed through sequential conjugation sites
In contrast to conjugation of two different drugs on the same linker, dual-drug ADCs can also be constructed through reaction of each drug at a different site on the antibody. In 2019, Nilchan et al. demonstrated this strategy in 2019 through sequential conjugation of MMAF and PNU-159682 at engineered selenocysteine and cysteine residues (Figure 2).[4] PNU-159682* is a DNA-damaging agent chosen for its high potency and activity against drug resistant and non-dividing cancer cells. The selenocysteine residue at S396U (Kabat numbering) was first reduced under mild DTT conditions before reaction with an iodoacetamide-bearing PNU-159682. Next, reduction of the interchain and engineered cysteine residues at A114C was performed with TCEP, followed by re-oxidation of the interchain disulfides with DHAA. Finally, reaction of the free engineered thiols with a pendant MSODA functional group on MMAF led to MMAF/PNU dual-drug ADC with a DARPNU of 1.9 and a DARMMAF of 1.5.

Figure 2. Sequential conjugation at engineered selenocysteine and cysteine residues yields dual-drug ADCs.

As with the combination of MMAE and PBD from Kumar et al., the MMAF/PNU dual-drug ADC in this study exhibited cellular toxicity dominated by PNU and did not enhance efficacy relative to the PNU single-drug ADC.

To further elucidate differences between dual-drug and single-drug ADCs, the researchers probed the mechanism of cell death by analyzing cell cycle changes by flow cytometry. MMAF interferes with cell cycle progression, leading to arrest at the G2/M phase. Analysis of the MMAF single-drug ADC revealed increases in the G2/M and G1 populations. In contrast, PNU single-drug ADC triggered a drastic increase in the S phase. The dual-drug ADC showed modest increases in the G2/M and G1 populations and a substantial increase in the S phase. Taken together, these data indicate that although PNU drives the toxicity of the dual-drug ADC, mechanistic effects from the addition of MMAF can be detected. These results further highlight the importance of matching drug pairs for successful complementary benefits.

Conclusions and Future Directions
The synthesis and evaluation of MMAE/MMAF dual-drug ADCs by both Levengood et al.[3] and Yamazaki et al.[1] has highlighted the therapeutic potential of ADC platforms capable of the simultaneous delivery of drugs with different physiochemical properties in tumor populations containing heterogenous antigen expression and resistant cells. Importantly the dual-drug ADC format had increased efficacy compared to co-administration of single-drug ADCs. Application of this technology to alternative drug combinations, MMAE + PBD or MMAF + PNU, has shown that the potency of dual-drug ADCs may be driven by the more potent drug in the combination. As such, the screening and evaluation of drug pairs will be essential in developing further dual-drug ADC platforms where we anticipate promising pairs to consist of two payloads with comparable toxicity and differing mechanisms of action.

Complementary pairs may also include drugs that exhibit actions other than direct tumor toxicity, such as immune checkpoint inhibitors or potentiation of ADC trafficking. The development of more modular dual-payload ADC synthesis platforms will facilitate generation of larger ADC libraries and could allow for rapid screening to uncover combinations that exert complementary or synergistic activities in vitro.

Promisingly, in vivo data from the MMAE/MMAF combinations in both studies revealed a benefit from the dual-drug ADC that was even more pronounced than observed in vitro, demonstrating the importance of the evaluation of promising dual-drug constructs in complex tumor settings.

Note
* PNU-159682 (3′-deamino-3”,4′-anhydro-[2”(S)-methoxy-3”(R)-oxy-4”-morpholinyl]doxorubicin) is a highly potent metabolite of the anthracycline nemorubicin, a DNA topoisomerase II inhibitor. PNU-159682 demonstrated inhibition in a panel of human tumor cell lines with IC70 values in the range of 0.07-0.58 nM. The drug is 2,360- to 790-fold and 6,420- to 2,100-fold more potent than MMDX and doxorubicin, respectively.[5]

References
[1] Yamazaki, C. M.; Yamaguchi, A.; Anami, Y.; Xiong, W.; Otani, Y.; Lee, J.; Ueno, N. T.; Zhang, N.; An, Z.; Tsuchikama, K. Antibody-drug conjugates with dual payloads for combating breast tumor heterogeneity and drug resistance. Nat Commun 2021, 12 (1), 3528.
[2] Kumar, A.; Kinneer, K.; Masterson, L.; Ezeadi, E.; Howard, P.; Wu, H.; Gao, C.; Dimasi, N. Synthesis of a heterotrifunctional linker for the site-specific preparation of antibody-drug conjugates with two distinct warheads. Bioorg Med Chem Lett 2018, 28 (23-24), 3617-3621.
[3]  Levengood, M. R.; Zhang, X.; Hunter, J. H.; Emmerton, K. K.; Miyamoto, J. B.; Lewis, T. S.; Senter, P. D. Orthogonal Cysteine Protection Enables Homogeneous Multi-Drug Antibody-Drug Conjugates. Angew Chem Int Ed Engl 2017, 56 (3), 733-737.
[4] Nilchan, N.; Li, X.; Pedzisa, L.; Nanna, A. R.; Roush, W. R.; Rader, C. Dual-mechanistic antibody-drug conjugate via site-specific selenocysteine/cysteine conjugation. Antib Ther 2019, 2 (4), 71-78.
[5] Quintieri L, Geroni C, Fantin M, et al. Formation and antitumor activity of PNU-159682, a major metabolite of nemorubicin in human liver microsomes. Clin Cancer Res. 2005;11(4):1608-1617. doi:10.1158/1078-0432.CCR-04-1845


Authors:

Corresponding Authors: Christine S. Nervig, Ph.D. E-mail: Christine S. Nervig, Ph.D and Shawn C. Owen, Ph.D. E-mail: Shawn C. Owen, Ph.D; Phone (801) 581-8069

Key terms: ADC, bioconjugation, differential antigen expression, simultaneous delivery of drugs
Published In: ADC Review| Journal of Antibody-drug Conjugates

DOI: https://doi.org/10.14229/jadc.2023.01.05.001.


How to cite:
Nervig C.S.1 and Owen S.C. 1, 2, 3. Advances in the Development of Dual-Drug Antibody Drug Conjugates – J. ADC. January 5, 2023. DOI: 10.14229/jadc.2023.01.05.001.

1Department of Medicinal Chemistry, University of Utah, Salt Lake City, Utah 84112;
2Department of Molecular Pharmaceutics, University of Utah, Salt Lake City, Utah 84112
3Department of Biomedical Engineering, University of Utah, Salt Lake City, Utah 84112


Last Editorial Review: December 26, 2022

Article History:

  • Original Manuscript Received October 24, 2022
  • Review results received December 13, 2022
  • Manuscript accepted for publication December 27, 2022