Recent advances in non-internalized antibody-drug conjugates (ADCs).

It is well known that traditional antibody-drug conjugates (ADCs) often need to be internalized into cancer cells to release their cytotoxic payload
.
At present, a total of 14 ADC drugs have been approved for marketing
worldwide.
At present, ADC drugs have developed into a new generation of targeted anti-cancer drugs after small molecule drugs
and monoclonal antibodies.

However, such ADC drugs often require tumor targets with high expression of internalized antigens and efficient intracellular lysis and activation efficiency, which brings many limitations
to the development of tumor-targeted therapy.
Therefore, the researchers developed a novel R&D strategy, the non-internalized ADC design strategy
.
With this approach, the payload can be released from the cleavable connectome upon binding to poorly internalized antigens or other tumor components
.
This not only eliminates the dependence on high antigen expression and avoids potentially inefficient internalization, but also greatly expands the range of tumor targets, including: extracellular tumor stroma-related targets
.

This paper introduces the latest research progress
of non-internalized ADCs in recent years.
In addition, tumor-related cell surface and extracellular proteins, tumor mesenchymal targeting, and linkage chemistry
that enables extracellular payload release are described.

1 Background

Antibody-drug conjugates (ADCs) are a class of oncology therapeutics that provide targeted drug delivery by binding to highly specific monoclonal antibodies (Figure 1).

This drug combines the targeting ability of antibodies and the cytotoxicity
of small molecule drugs.
Unlike the typical mechanism of action of ADCs, this article focuses on ADCs
that do not require intrinsic efficacy.

2 Mechanism of action of non-internalized ADCs

ADCs are typically designed to function through internalized mechanisms: after the antibody binds to the target antigen on cancer cells, the ADC is internalized by endocytosis and then degraded in lysosomes, releasing a cytotoxic payload (Figure 2A).

If the payload has proper membrane permeability, it can spread out the cells that release it, into surrounding "bystander" cancer cells that may or may not express the antigen of interest (known as the bystander effect
).
While this approach has been successful (in fact, all ADCs currently approved by the FDA use internalized antibodies), it does have certain limitations, such as reliance on overexpression of internalized antigens that are not characteristic of
many tumors.
In addition, antibodies are large proteins, so their penetration into tumors is limited
due to slow spread.

Previous theories suggested that internalization is essential for the treatment of ADCs, but a growing body of research suggests that alternative mechanisms of action may be more effective
.
Lysosomal degradation is not required to release intracellular payloads, and antibody-drug ligation can be designed to be unstable in an extracellular tumor environment (Figure 2B).

This takes advantage of cancer's unique chemical or enzymatic environment
compared to healthy tissue.
The extracellular payload released can spread into surrounding cancer cells, exerting its cytotoxic role
.
The non-internalized mechanism of action may broaden the range of tumor targets because it excludes the stringent requirements of internalized ADCs, such as highly internalized antigen expression
.
In fact, some non-internalized ADCs can target physiological tumor features common to most aggressive cancers and therefore may be a promising broad-spectrum anti-cancer approach
.
The extracellular release of the payload also allows the anticancer agent to penetrate deeper into the tumor, as the anticancer agent spreads more easily than the antibody and maximizes
the potential bystander effect.

Multiple characteristics of the target antigen are important
in determining the success of ADCs.
For example, the target antigen is highly expressed on the target cancer cells, but not on healthy cells, to ensure the specificity of the treatment and avoid off-target toxicity
.
The nature of the antibody-antigen binding interaction determines whether (and how quickly)
the construct internalizes.
To successfully internalize ADCs, antibodies must rapidly induce receptor internalization, endosomal transport, and lysosomal processing (Figure 2A)

Conversely, for non-internalized ADCs, these steps that need to be completed within the cell are negligible
since internalization does not need to occur at a considerable rate after antibody binding.
With the exploration of non-internalized ADCs, more and more cases of poor internalization of antibody-antigen interactions have been continuously discovered
.

3 Mechanism of non-internalized ADCs against B-cell antigens

As part of their study of the mechanism of action of multiple ADCs, Person et al.
examined the feasibility
of non-internalized ADCs against B-cell CD20, CD21, and CD72 antigens.

In this work, they generated a variety of ADCs
from non-cleavable or clepable connectomes containing disulfides or Val-Cit dipeptides.
They found that cleavage disulfide ADCs were effective
even against very poorly internalized antigen targets CD20, CD21, and CD72 in vivo.
It is hypothesized that disulfide bonds can be cleavaged outside the cell, thus acting as a non-internalizing mechanism
.
Also in this work, the MMAE payload of the Anticd21 ADC containing the Val-Cit linker was replaced by MMAF, an analogue containing the ends of charged carboxylic acids (Figure 3C), which hindered its ability
to spread across the cell membrane.
This change eliminates the activity of the ADC, indicating that diffusion of the released payload across the cell membrane is critical for activity, supporting the proposed mechanism
of extracellular action.
Obviously, this data contradicts the classical idea that disulfides and Val-Cit clepable connectomes need to be internalized to release the payload, and demonstrates that cleavage of connectomes outside the cell is possible
.

1 Targeting membrane proteins

       PD-L1:

Programmed cell death ligand 1 (PD-L1) is a transmembrane protein that binds to programmed cell death protein-1 on the surface of T cells, inhibiting T cell function and allowing cancer cells to evade the immune system
.
PD-L1 expression is upregulated in various types of solid carcinoma and is often associated with
poor prognosis and reduced survival.
Three antibodies against PD-L1 have been approved by the FDA: Avelumab (Merck), Atezolizumab (Roche) and Durvalumab (AstraZeneca
).
Due to the transient and heterogeneous expression of PD-L1 and the poor penetration of a large number of antibodies in the dense tumor interstitium, these antibodies have been greatly limited
in the treatment of tumors.
Therefore, the non-internalized ADC approach is attractive because extracellular payload release can overcome these problems
.
While cleavage of lysable Val-Cit linkers has been observed in previous studies to be cleavage, internalization is required in this study, possibly because of the uncertain level of cleavage of extracellular linkers
.
Levels of extracellular proteases or other linkage triggers differ
between cell lines, tumor types, and even between preclinical mouse models and humans.
Therefore, in vitro and in vivo application of extracellular lypable connexants may require case-by-case validation of the levels of
available endogenous triggers.

       CEACAM5：

Carcinoembryonic antigen cell adhesion molecule 5 (CEACAM5) is a biomarker for many types of cancer and is usually not significantly internalized
.
Targeting this antigen was successfully validated in preclinical tumor models using pH-sensitive carbonates or ester linkers to release SN-38
from ADCs in an extracellular acidic tumor microenvironment.

Goldenberg et al.
investigated the activity
of anti-CEACAM5 labituzumab ADCs in mouse models of human colon cancer, human pancreatic tumor, and systemic lymphoma xenograft tumors.
All ADCs contain acid-intolerant concaters of carbonate or ester moieties, and some substrates have additional protease-separable phenylalanine-lysine (Phe-Lys) residues (Figure 5).

Fortunately, despite the poor internalization of CEACAM5, 24 ADCs induced 100% survival in the CEACAM5-positive lymphoma tumor target mode, with ADCs with dual cleavable linker CL2 (Figure 5) performing best
in vivo.
The authors propose a non-internalized mechanism of action in which SN-38 is slowly released from the ADC extracellular, allowing high local accumulation
of free drugs.
The relatively high carbonate/ester bond activity is considered suitable for use with SN-38 because SN-38 has moderate cytotoxicity compared to aureglucin or metansin-like payloads, allowing easy drug release without significant off-target toxicity
.
In fact, FDA-approved ADC Trodelvys contain SN-38 bound via CL2A linker (Figure 5C).

Although its goal is to highly internalize the antigen troP-2, replacing the CL2A linker with a highly stable linker significantly reduces E-cacy, suggesting that the gradual release of SN-38 from the CL2A linker contributes to E-cacy
.

       NKA ：

NKA is a plasma membrane ion pump that is essential for the function of all cells: therefore, inhibition of NKA leads to cell death
.
Unlike the classical ADC approach, the authors did not develop antibodies against NKA (which are not specific for healthy cells), but instead took advantage of the proximity of other cancer-associated proteins to NKA
.
Dydapherin is a cell membrane glycoprotein that is overexpressed in many metastatic cancers and interacts
with NKA.
Binding the cardiac glycoside inhibitor of NKA to the anti-Dydapherin antibody binds the antibody antigen to Dydapherin, bringing the inhibitor very close
to NKA.
Inhibitors that bind to NKA lead to cell death by inducing cell swelling and subsequent membrane rupture (Figure 6).

Interestingly, this strategy excludes linker cleavage, and the inhibitor acts extracellular
while attaching to the antibody.

2 Target the tumor microenvironment

Alternative strategies for targeting tumors without utilizing membrane antigens are now emerging
.
Reliance on targeted membrane antigens has several drawbacks: because few cancers overexpress antigens uniformly, drug resistance can be acquired
through changes in antigen expression.
Instead, other approaches indirectly seek to eradicate malignant cells
by targeting components of the extracellular tumor microenvironment, such as secreted or extracellular proteins, or targeting connective tissues and blood vessels rich in the tumor matrix.

3 Target extracellular matrix proteins

Galectin-3-binding protein (Gal-3-BP), widely present in most tumors, is an attractive target for tumor
microenvironment-targeting ADCs.
Yakobelli et al.
developed a humanized antibody against the Gal-3-bp lectin binding domain, which binds
drugs in a "connectionless" manner by forming mixed disulfides directly with C-terminal cysteine and sulfur-containing metatanine (DM1, DM3 or DM4).
This link-free technique was first described by the Neri lab in which he demonstrated that high concentrations of reducing agents in the extracellular space help selectively lyse disulfides, resulting in the release
of extracellular drugs.

It is thought that after payload release and cell death, more and more reducing agents are released in the tumor microenvironment, leading to payload release and cell death in a "chain reaction" (Figures 7A and B).

Thus, ADCs containing anti-Gal-3-BP disulfide achieved significant tumor growth inhibition in vivo, and DM3 conjugates achieved long-term and complete remission
in the optimal dose of mouse xenograft models.

4 Targeting mechanism or tumor blood vessels

Many solid tumors exhibit dense intercellular stroma: connective tissue and vascular networks that prevent macromolecules from penetrating deep into the tumor (Figure 8).

This is often a barrier
to successful treatment of solid tumors with ADCs.
A new strategy alternative to targeting tumor cells located behind the stromal barrier is targeting the stroma itself
.
The accumulation of ADCs in the stroma allows the release of extracellular payloads in the tumor microenvironment that then spread into neighboring tumor cells
.
Since the proliferation of small molecules is greater than that of monoclonal antibodies, the permeability of active drugs to solid tumors is expected to be improved
.

This is called "cancer matrix targeted therapy" (CAST).

In addition, the treatment of several solid tumors (colorectal, lung, pancreas) by classical ADCs is limited
due to its heterogeneous target antigen expression.
Therefore, an alternative targeting approach is to modify the components of the tumor extracellular matrix including collagen, fibronectin, and fibrin, which can be very attractive targets due to their richness, stability, and selectivity in the tumor matrix or blood vessels
.

While most ADCs rely on endogenous triggers to release payloads, such as tumor-associated enzymes or different tumor pH, a new alternative method has been developed to introduce small molecule triggers from separate exogenous triggers in
ADCs.
The time to allow the ADC to bind to the target antigen after administration of the lysable, non-internalized ADC, and the time
to clear any unbound, untargeted ADC from the blood to counteract off-target effects.
Subsequently, an exogenous release trigger is given to activate only the ADC at the site of action (Figure 9).

Unlike traditional ADCs, this approach does not rely on cleavage of natural biological activation mechanisms, such as the presence or specific concentrations of certain enzymes released by the payload, which can vary greatly between cancer and patients
.
Thus, bioorthogonal activation can allow for more predictable drug release rates, spatial and temporal control
.

Tumor-associated glycoprotein 72 (TAG72) is overexpressed
on the surface of cancer cells in a variety of solid tumors.
Rossin et al.
have demonstrated its applicability
as a non-internalized ADC target.
WHO generated a non-internalized anti-TAG-72 ADC
using the Bioorthogonal Inverse Electron Requirements Diels Alder (IEDDA) reaction.
The addition of trans-cyclooctene linkage to the ADC enables exogenously administered small molecule tetrazine probes to easily "click-release" chemistry
.
A pyridazine elimination reaction is subsequently triggered, allowing allyl carbamate to cleave from trans-cyclooctene, releasing free amine-containing cytotoxins (Figure 10).

Thus, a non-internalized ADC is generated that releases the payload outside the cell after cutting the linker via an exogenous trigger
.

Unbound ADCs need to be cleared from the blood prior to the introduction of exogenous probes to avoid systemic toxicity
due to off-target activation and payload release.
Therefore, the authors used ADCs in the Diamond format due to faster clearance and deeper tumor penetration
.
With high tumor localization preserved, Diabody-conjugates with MMAE are administered tetrazine click activator after 2 days, giving extracellular activation and MMAE release
.
In healthy non-target tissues, low retention of Diabody-conjugates provides high tumor selectivity
.
In vivo, the TAG-72 non-internalized click-release Diabody ADC was compared
to an analogue of the TAG-72 Diabody-ADC with Val-Cit cleavable connectome.
Click-release ADCs were observed in vivo to improve tumor growth inhibition at EC50s of 185 and 35 pm (in colorectal and ovarian cancer xenografts).

In contrast, val-cit ADCs are ineffective in tumor growth inhibition
due to limited extracellular release of MMAEs.
The click-release strategy was well tolerated and was not toxic with ADC, tetrazine activator or a combination
thereof.
Although this approach requires the clinical development of two different components (ADC and activator), it is expected that the same activator can be applied to broad-spectrum cancer targets such as the matrix components present in many solid tumors
.

In summary, the authors demonstrate the applicability of IEDDA pyridazine elimination to allow time-controlled and trace-free ADC linker cleavage, activating payload release secretion only at extracellular sites, independent of endogenous tumor biological effects
.

Over the past decade, more and more attention has turned to the development of non-internalized ADCs with extracellular payload release mechanisms, with the aim of improving some of the shortcomings
associated with classical internalized ADCs.
For example, their dependence on high, homogeneous antigen expression excludes many other cancer types from target selection
.
Therefore, non-internalized ADCs represent a new hope that non-internalized ADCs can be used to effectively treat a wider range of cancers
lacking suitable antigen expression.

References

[1] Combination radioimmunotherapy and chemoimmunotherapy involving different or the same targets improves therapy of human pancreatic carcinoma xenograft models, Mol Cancer Ther.
2011 Jun; 10(6):1072-81.

[2] In vivo chemistry for pretargeted tumor imaging in live mice, Angew Chem Int Ed Engl, 2010 Apr 26; 49(19):3375-8.

[3] A thioether-directed palladium-cleavable linker for targeted bioorthogonal drug decaging , Chem Sci.
2018 Apr 6; 9(17):4185-4189.

[4] Eradication of Triple-Negative Breast Cancer Cells by Targeting Glycosylated PD-L1, Cancer Cell.
2018 Feb 12; 33(2):187-201.
e10.

[5] Location matters: site of conjugation modulates stability and pharmacokinetics of antibody drug conjugates, Chem Biol.
2013 Feb 21; 20(2):161-7.

[6] Non-internalising antibody-drug conjugates.
[J] .
Chem Soc Rev, 2022