Degrader-antibody conjugates (DACs) are novel entities that combine a proteolysis targeting chimera
(PROTAC) payload with a monoclonal antibody via some type of chemical linker. This review provides a
current summary of the DAC field. Many general aspects associated with the creation and biological
performance of traditional cytotoxic antibody-drug conjugates (ADCs) are initially presented. These
characteristics are subsequently compared and contrasted with related parameters that impact DAC
generation and biological activity. Several examples of DACs assembled from both the scientific and the
patent literature are utilized to highlight dif f ering strategies for DAC creation, and specific challenges
associated with DAC construction are documented. Collectively, the assembled examples demonstrate
that biologically-active DACs can be successfully prepared using a variety of PROTAC payloads which
employ diverse E3 ligases to degrade multiple protein targets.

1. Introduction

Synthetic chimeric protein degraders, often referred to as proteolysis targeting chimeras (i.e., PROTACs), are rapidly revolutionizing the fields of biology and medicinal chemistry. 1 These heterobifunctional molecules are able to specifically degrade targeted proteins inside of cells, and they thereby of f er the potential for improved and/or prolonged biological activity relative to simple small-molecule inhibitors of the same enti- ties. Such degraders are typically composed of a binding element that recognizes the targeted protein, a separate moiety that associates with an E3 ligase, and a spacer group which connects these two components (Fig. 1; note that the "spacer" nomenclature is purposefully used here in preference to the "linker" terminology often employed in the PROTAC literature to avoid confusion with the antibody-drug conjugate linkers described later in this work). Because of their chimeric nature, heterobifunctional degraders often possess physicochemical characteristics that can impart poor DMPK properties to the associated molecules such as low oral bioavailability and/or rapid in vivo clearence. 2 The degrader field is well-aware of these potential challenges and has recently focused on creating chimeric molecules with physicochemical attributes that more- closely resemble typical orally bioavailable compounds. 3,4 Such next-generation degraders often utilize cereblon (CRBN) as the E3 ligase since the corresponding binding ligands are currently considered more drug-like than those which target other ligases such as VHL (the von Hippel-Lindau tumor suppressor). 5 These approaches have af f orded a number of orally-bioavailable PRO- TAC compounds with several examples now progressing in human clinical trials. 6 As an alternate means to enable the ef f i cientin vivo delivery of chimeric degraders, several groups have explored conjugation of such entities to monoclonal antibodies. These strategies build on the growing clinical and commercial successes of antibody- drug conjugates (ADCs) bearing cytotoxic payloads. 7 The resulting entities, termed "degrader-antibody conjugates" (DACs) throughout the remainder of this review, of f er several potential advantages relative to the in vivo administration of unconjugated PROTAC molecules. These possible benefits include: (1) the in vivo delivery of chimeric degraders with poor physicochemical and/or DMPK properties (especially those which employ E3 ligases other than CRBN), (2) the avoidance ofcomplex and/or non-standard formula- tions which are often required to enable unconjugated PROTACs to achieve meaningful in vivo exposures, and (3) the ability to target a PROTAC molecule of interest to a specific tumor or tissue via the antigen that is recognized by the DAC. This reviewdescribes several general aspects of degrader-antibody conjugate composition and design and illustrates these concepts via recent examples of such entities taken from the scientific and patent literature

2. Antibody-drug conjugate composition and mechanism-of-action

Before describing DACs, it is important to discuss traditional (cytotoxic) ADCs so that similarities and dif f erences between the two modalities can be properly illustrated. Antibody-drug conjugates are typically comprised of a monoclonal antibody that is connected to a cytotoxic payload ("drug") via some type of chemical linker (Fig. 2). The linker incorporates a chemical moiety that enables its covalent attachment to the antibody. It is also covalently connected to the payload ("drug"), and it may include several optional components such as (1) a spacing element which distances the attached linker from the antibody, (2) a chemically or biologically activated trigger designed to undergo intracellular cleavage, and (3) a second (often self- immolative) spacing element which separates the trigger from the payload. 8 The color-coding ofthese optional items shown in Fig. 2 is maintained in the chemical depictions of specific degrader-antibody conjugates below to help readers identify the function of each linker component. The combined linker and payload portion of an ADC is frequently termed the "linker- drug" (Fig. 2). These linker-drug entities are usually synthesized chemically and are subsequently attached to the antibody via a variety of conjugation technologies (see below). Importantly, when joined to the antibody in this manner, the payload is usually not able to ef f i ciently interact with its biological target.

The antibody portion of an ADC specifically binds to an antigen that is expressed on the surface ofa tumor cell (or other targeted tissue) and this event results in the internalization of the conjugate (Fig. 3). 9 Antigens employed in ADC research are purposefully selected based on their high expression in the tumor/tissue of interest, their ability to rapidly internalize and properly traf f i c the bound ADC, and their relatively low expression in other non-targeted tissues. 10 The internalized ADC is sub-sequently traf f i cked to the lysosomal compartment of the cell where the payload is released in bioactive form via one of several methods. 9 In the case of a non-cleavable linker, the antibody is catabolized by lysosomal enzymes resulting in a bioactive payload that is still connected to the linker and a small portion of the antibody (typically a single amino acid; Fig. 3, path A). In contrast, if the ADC contains a cleavable linker, that moiety is activated in the lysosomal compartment in a manner that ultimately releases the free payload (Fig. 3, path B). Hybrid possibilities also exist in which a linker fragmentremains attached to the bioactive payload post-cleavage (Fig. 3, path C).

A variety of lysosomal-based linker cleavage mechanisms have been exploited in the ADC field including: acid-based hydrolysis (the lysosome is somewhat more acidic than the extracellular environment and other intracellular compartments), disulfide reduction (some debate exists regarding the exact cellular location of such cleavage events), and, most commonly, enzymatic cleavage ofpeptide, peptidomimetic, or monosaccharide- based moieties. 8 Greater detail regarding cleavage and payload release mechanisms can be found in the references associated with the specific degrader-antibody conjugates described below in Sections 3B and 3C. Importantly, all these approaches typically seek to identifycleavable linkers which are stable in circulation in vivo and which onlyrelease the attached payload after internalization into the targeted cell. The bioactive payload produced viathese events then transitions out of the lysosomal compartment, possibly via mechanisms involving active transport, and dif f uses through the cellular environment to reach its biological target (Fig. 3). 11 Critically, the payload mustbe suf f i cientlystable in the lysosomal environmentto enable such transition to occur with reasonablyhigh ef f i ciency. 12 In certain cases, it is possible for the transitioned payload to exit the cell originally targeted by the ADC and to enter a nearby cell that may or may not express the target antigen (the so-called "bystander ef f ect"; Fig. 3). 13 This behavior has the advantage of enabling a given ADC to impactbiologically-relevantcells which do notexpress the target antigen (thus possibly addressing concerns regarding heterogeneous antigen expression in the targeted tumors/tissues), but may also result in increased toxicity toward non-targeted cells. When all of the above criteria are met, an ADC is able to deliver its associated cytotoxic payload to the targeted tumors/ tissues in an antigen-dependent manner and thereby induce desired biological outcomes (typically antiproliferation activity in vitro or tumor growth inhibition in vivo). Evidence for such antigen-dependent delivery can be provided by several methods including (1) attenuated activity of the ADC toward cells which do not highly express the target antigen, (2) similarly attenu- ated activity of a control ADC bearing the same payload but which targets an antigen not highly expressed on the tumor/ tissue of interest (i.e., a "non-target" control conjugate), and (3) weakened activity ofthe targeted ADC when co-administered with an unconjugated antibody which competes for the same antigen (assuming that mAb-antigen binding does not result in meaningful biological ef f ects). Importantly, smaller absolute amounts of the attached payloads are typically delivered to cell interiors by the ADC approach relative to the passive extra- cellular dif f usion processes frequently observed for unconju- gated small molecules. For this reason, cytotoxic ADC payloads ave historically exhibited extremely potent (picomolar) in vitro antiproliferation properties, although the field is now shifting to less-active compounds and higher drug loadings as the next generation of cytotoxic ADC payloads are explored. As mentioned above, a number of diverse conjugation methods are employed to attach the linker-drug to the antibody during the creation of an ADC, and these approaches have evolved considerably over time. 14,15 For example, non-specific derivatization of mAb surface lysine residues was utilized to create the first ADCs to be approved. These chemistries produce heterogeneous conjugates (not shown) with average drug- antibody ratios (DAR; i.e., the number of linker-drugs attached to a given mAb) between two and seven. More-recently, the four disulfide bonds which connect the light and heavy chains of a typical human IgG antibody were selectively reduced to af f ord
the corresponding cysteines, and these residues were then used to attach linker-drugs (Fig. 4, top row). Maleimide-based con- jugation methods are frequently employed in such conjugation reactions, and these approaches af f orded several marketed ADCs with average DAR values of approximately four (partial reduction, somewhat heterogeneous) or eight (full reduction, homogeneous). 16 In addition, the use of bifunctional electro- philes which react with two cysteines, often termed "thiol rebridging", can increase the homogeneity of DAR4 conjugates generated via such interchain conjugation methods. 17Even more-recently, a number of"site-specific" conjugation technol- ogies were described in the ADC field which enable the con- struction of near-homogeneous conjugates with DAR values ranging from two to six. 14,15Examples of such approaches include the use of engineered cysteine residues or non- natural amino acids that are introduced at specific locations within the IgG structure and which are subsequently employed for conjugation purposes (Fig. 4, bottom row). 18The utilization ofenzyme-based conjugation methods to produce homogeneous DAR2 conjugates is also gaining prominence (not shown). 19 As was the case with the ADC linkers described above, additional details regarding selected conjugation approaches and related chemistries are provided below in the discussions associated with specific degrader-antibody conjugates.