Proteasomal and lysosomal degradation for specific and durable suppression of immunotherapeutic targets

Cancer immunotherapy harness the body’s immune system to eliminate cancer, by using a broad panel of soluble and membrane proteins as therapeutic targets. Immunosuppression signaling mediated by ligand-receptor interaction may be blocked by monoclonal antibodies, but because of repopulation of the membrane via intracellular organelles, targets must be eliminated in whole cells. Targeted protein degradation, as exemplified in proteolysis targeting chimera (PROTAC) studies, is a promising strategy for selective inhibition of target proteins. The recently reported use of lysosomal targeting molecules to eliminate immune checkpoint proteins has paved the way for targeted degradation of membrane proteins as crucial anti-cancer targets. Further studies on these molecules’ modes of action, target-binding “warheads”, lysosomal sorting signals, and linker design should facilitate their rational design. Modifications and derivatives may improve their cell-penetrating ability and the in vivo stability of these pro-drugs. These studies suggest the promise of alternative strategies for cancer immunotherapy, with the aim of achieving more potent and durable suppression of tumor growth. Here, the successes and limitations of antibody inhibitors in cancer immunotherapy, as well as research progress on PROTAC- and lysosomal-dependent degradation of target proteins, are reviewed.


Introduction
Immune checkpoint blockade therapy has shown promise in restoring the antitumor immune response through activating the immune system 1,2 . Multiple immunocheckpoint pathways have been described, such as those involving cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), V domaincontaining Ig suppressor of T-cell activation (VISTA), CD47, OX40, and T cell immunoglobulin and mucin domain 3 (TIM-3) [3][4][5] . The programed cell death 1 (PD-1)/programed cell death ligand 1 (PD-L1) axis is the most important pathway in terms of clinical therapeutic effects. PD-L1, a type I transmembrane protein containing a short cytoplasmic domain and 2 extracellular domains, is an important immune checkpoint molecule that is commonly expressed on the surfaces of cancer cells 6 and is encoded by the Pdcd1 gene. The promoter region of Pdcd1 has 2 transcription-factor binding sites (termed conserved regions B and C), which are critical for regulating PD-1 expression. Pdcd1 regulation occurs partly via the recruitment of nuclear factor of activated T cells 1 (NFATc1) to a novel regulatory element at the Pdcd1 locus, and it is part of the molecular mechanism underlying the induction of PD-1 in response to T cell stimulation. The interaction between PD-1 on T cells and PD-L1 on cancer cells results in the suppression of tumor-killing activity of T cells, and is a crucial mechanism of tumor immune escape 7 . In addition, tumor-intrinsic oncogenic PD-1 promotes tumor cell proliferation independently of adaptive immunity 8 .
Checkpoint inhibitors promote the anti-tumor immune response by antagonizing suppressive immune checkpoint regulatory pathways. PD-1/PD-L1 blockade therapy restores the effective activation of T-cell-mediated antitumor immunity and has substantial clinical benefits 9 . The production of antibodies targeting PD-1 and PD-L1 has led to the development of drugs targeting these pathways 10 . Multiple anti-PD-1/PD-L1 monoclonal antibodies (mAbs) have been approved as a therapy to treat many types of cancer 11 . However, PD-L1 expression varies significantly among tumor types and stages, and its level changes during therapy 12 . Moreover, the relatively low response rate, acquired resistance, and occasional fatal adverse effects also pose substantial challenges to use of this therapy 13 . The molecular mechanism regulating PD-1/PD-L1 remains largely unknown 14 . Small molecule-mediated inhibition of protein function is a canonical paradigm that enhances the efficacy of most clinical agents 15,16 . However, most small molecule inhibitors of the PD-1/PD-L1 pathway are not yet ready for widespread clinical use, and further preclinical work is required to optimize their formulation and application. Moreover, these small molecule inhibitors can achieve substantial inhibitory effects only when more than 90% of targets are engaged 17 . The high dosing level required can lead to on-target off-tumor effects. Gene knockdown methods based on RNA interference and CRISPR/Cas9 or related strategies have been applied to decrease cellular protein levels and have shown clear therapeutic potential 18,19 . Nevertheless, attaining sufficient concentrations of related agents at the target site is difficult. Safety concerns also arise from off-target effects. Moreover, these agents' poor metabolic stability causes many adverse effects 20 .
Therefore, PD-1/PD-L1 blockade therapy still faces great challenges that must be overcome. In-depth understanding of the control of PD-L1 expression in tumor cells is required for further improvement of checkpoint blockade therapy. PROTAC, a bifunctional small molecule compound, has been widely studied in the field of anti-cancer drugs and is used as a new therapeutic method 21,22 . The recently reported lysosomal targeting molecules for eliminating immune checkpoint proteins have provided a new direction for targeted degradation of target proteins as crucial anti-cancer targets 23,24 . Here, we primarily review the progress that has been made in degradation strategies for immunotherapeutic targets. These strategies have shown promise in providing an alternative strategy for cancer immunotherapy to achieve more potent and durable suppression of tumor growth.

Monoclonal antibodies
Monoclonal antibodies are a promising strategy for the blockade of PD-1/PD-L1 function. Humanized mAb targeting PD-1/PD-L1 relieves T cell immunosuppression and induces T cell activation, thus restoring the body's ability to monitor and attack tumor cells. Atezolizumab is the first licensed anti-PD-L1 mAb. Atezolizumab is designed to target PD-L1 through binding to the front beta-sheet of PD-L1 25,26 . Atezolizumab restores the anti-tumor activity of T cells by inhibiting the interaction of PD-L1 with PD-1 on the surfaces of T cells 27 . Pembrolizumab, another humanized anti-PD-L1 mAb, has low affinity for Fc receptors and C1q, and a low likelihood of host immunity stimulation 28 . Pembrolizumab has shown strong anti-tumor activity in phase I clinical trials and is widely used in patients with advanced malignant tumors 29-31 . Nivolumab, a humanized anti-PD-1 IgG4 mAb, binds an N-terminal loop outside the IgV domain of PD-1 32 . Nivolumab has been approved for application in combination with platinum-based chemotherapy 33,34 .
With the extensive development of clinical treatments, new problems have arisen in the practical use of mAb preparations. Monoclonal antibodies induce the production of anti-mAbs, thus leading to immune-related adverse events (irAEs), such as interstitial pneumonitis, colitis with gastrointestinal perforation, and severe skin reactions 35,36 . In fact, PD-L1 is transported from the plasma membrane into the cytosol and actively redistributed to the plasma membrane, thereby decreasing mAb efficacy, although mAbs can effectively block PD-L1 on the surfaces of tumor cells 8,37 (Figure 1). Emerging evidence indicates that exosomal PD-L1 mediates resistance to immunotherapy by facilitating PD-L1 evasion of the anti-PD1/PD-L1 mAbs 38 . Tumor-derived exosomes carry bioactive PD-L1 molecules on their surfaces that suppress the anti-tumor immune response. This exosomal PD-L1 can enter the blood circulation and inhibit T cells outside the primary tumor tissue, thus causing T cells to lose their anti-tumor ability before reaching the tumor 39,40 . Studies have also revealed the functions of PD-1 and PD-L1 independently of immunosuppression. The intracellular domain of PD-L1 regulates the malignant behaviors of cancer cells and mediates chemoresistance 7 . PD-1 is expressed in a broad range of tumor cells. The cancer-intrinsic PD-1 promotes malignant proliferation by upregulating mammalian target of rapamycin (mTOR) signaling 8 . These functions may explain the unexpected effects of mAbs.

Small molecule peptide inhibitors
The disadvantages of mAbs limit their application. Although much progress has been made in the development of antibodies against the PD-1/PD-L1 pathway, there is an increasing desire to use small molecules to block the PD-1/PD-L1 axis through dissociating the PD-1/PD-L1 complex. The benefits of using small molecules instead of antibodies include better oral bioavailability, fewer irAEs, better tumor permeability, and lower production costs. The long half-life is a major limitation of mAbs. Small molecule drugs are based primarily on using different therapeutic methods to target the PD-1/PD-L1 pathway. Small molecule inhibitors are more suitable for oral administration and can decrease the target occupancy time by regulating the half-life of the drug, thereby avoiding serious irAEs 41,42 . Tripeptidyl peptidase 1 (TPP1), an active small-molecule peptide, has high affinity for human PD-L1 43 . In a mouse model, TPP-1 has been found to reactivate T cells through blocking the PD-1/PD-L1 interaction and to inhibit tumor growth 44 . Nonylphenol ethoxylate (NP-12), a polypeptide antagonist of the PD-1 signaling pathway, is used as an immunomodulator for cancer treatment 44 . In mouse models of colon cancer and melanoma, NP-12 inhibits PD-1/PD-L1 interaction and suppresses tumor growth and metastasis 45 . CA-170, a small molecule that has been tested in clinical trials, inhibits both the PD-L1 pathway and the VISTA pathway 46 . Sulfamethazine and sulfamethoxazole are small molecules originally produced to inhibit the PD-1/PD-L pathway, and they have been found to rescue PD-1-mediated suppression of IFN-γ secretion 47 . Several other small molecule compounds that inhibit the PD-L1 pathway have been patented 48 . Notably, peptide inhibitors remain in an early stage of development, although they are promising in suppressing immune checkpoints. Furthermore, high drug doses are generally required, thus often leading to undesired adverse effects because of the off-target binding associated with higher drug concentrations 44,45,49 .

Targeted protein degradation as a promising strategy for drug development Small molecule protein proteolysis-targeting chimeras as antitumor agents
Small molecule inhibitors have been used to control cellular protein levels through occupancy-driven pharmacology as the mode of action. However, this strategy only temporarily inhibits the functions of regulatory proteins 50,51 . Finding new models is essential to control cellular protein levels. PROTACs, an attractive new approach for removing proteins by using cellular protein degradation systems to hijack the ubiquitin proteasome system, are playing an increasingly important role in drug discovery 52,53 . PROTACs are activators of ubiquitin ligase whose catalytic properties can be programmed (Figure 2) 54 . The heterobifunctional molecules of PROTACs recruit specific target proteins to E3 ubiquitin ligase and then reprogram the enzyme to ubiquitinate the selected target proteins, thus leading to target ubiquitination and degradation. PROTACs also activate ubiquitin ligase through mediating the formation of target protein-PROTAC-E3 ligase catalytic ternary complexes, thus providing a framework for more robust PROTAC designs 55,56 .
In 2001, the first PROTAC was reported to recruit SCFβ-TRCP E3 and subsequently induce degradation of methionine aminopeptidase 2 (MetAp-2) 52 . In 2008, small molecule-based E3 recruitment ligands were invented, and great progress was made in PROTAC technology 57 . Studies indicated the feasibility of developing PROTACs that can enter cells relatively easily 57 .
In 2013, mouse experiments provided the first demonstration that phospho-dependent PROTACs (PhosphoPROTACs) inhibit tumor growth in vivo 58 . Nonetheless, the peptidic E3 ligase ligands used in PROTACs have hindered their development into more mature chemical probes or therapeutic regimens, because peptidic E3 ligase ligands lead to high molecular weight of the entire PROTAC molecule, thus resulting in poor cell permeability 59 . In 2015, the novel PROTAC HaloPROTAC with incorporated small molecule VHL ligands was reported to successfully degrade HaloTag7 fusion proteins 60 . HaloTag7 is a modified bacterial dehalogenase that covalently reacts with hexyl chloride tags; HaloTag fusion proteins have been widely used to bioorthogonally label proteins in vivo. HaloPROTACs resulted in a 90% maximum degradation of GFP-HT7 with a low nanomolar half-maximum degradation concentration 60 .
HaloPROTACs inspired the development of future PROTACs with more drug-like properties and have become useful chemical genetic tools 61 as small molecule proteasome modulators. PROTAC is chimeric with these small molecules, and it forms a bifunctional small molecule compound that can link target proteins and E3 ubiquitin ligase in a ternary complex, thus resulting in target protein degradation through the ubiquitin-protease system 21,22 . Furthermore the development of small molecule-based PROTAC compounds with more druglike properties has allowed for potent permeable PROTACs to be generated 62,63 . Studies have shown that a PROTAC against P300/CBP-associated factor (PCAF) and general control nonderepressible 5 (GCN5) effectively regulates the expression of multiple inflammatory mediators in macrophages and dendritic cells 64 . To date, a variety of PROTAC variants have been developed, thus laying a foundation for drug advancement. Homo-PROTACs have been developed for auto-targeting of both von Hippel-Lindau (VHL) and cereblon (CRBN) 65,66 . Additional insights have been gained in the structural basis and target selectivity of PROTACs. PROTACs can be designed to target various proteins of interest, because they are programmable 67,68 . Moreover, the PROTAC technology can escape the resistance mechanisms of inhibitors, including overexpression of target proteins and resistance mutations, by enabling modulation of both the enzymatic and non-enzymatic roles of proteins 69 . PROTAC technology has many advantages, such as low manufacturing cost, low drug dosage, excellent cell permeability, broad tissue distribution, and a strong ability to regulate intracellular targets 70 . PROTACs have already been applied for the degradation of various notable targets, including abelson murine leukemia viral oncogene homolog 1 (ABL)-breakpoint cluster region (BCR) in chronic myeloid leukemia, bromodomain proteins in multiple cancers, and androgen receptor in prostate cancer (details in Table 1) 53,71,72 . PD1/PD-L1 have not been selected as target proteins for study, although multiple protein targets have successfully been modulated with PROTAC technology. A study has reported that J22352, a highly selective HDAC6 inhibitor with PROTAClike properties, decreases the immunosuppressive activity of PD-L1, thus restoring anti-tumor activity in glioblastoma 117 .

Targeted lysosomal degradation of PD-L1 by PD-LYSO
Although antibodies, small molecule inhibitors, and PROTACs have effects on immunotherapeutic target inhibition, the relatively low response rate and checkpoint blockade resistance have necessitated exploration of the molecular regulatory mechanisms of PD-L1. Studies have revealed the mechanisms that control PD-L1 transcriptional activation and post-translational modifications [118][119][120] . β transducin repeat containing protein (β-TrCP), cullin 3, and COP9 signaling body 5 (CSN5) control the degradation of PD-L1 through regulating PD-L1 ubiquitination 121,122 . Moreover, targeted blockade of PD-L1 transport from the endoplasmic reticulum to the Golgi apparatus triggers endoplasmic reticulum-related degradation of PD-L1 123 . Studies have found that chemokine-like factor (CKLF)-like MARVEL transmembrane domain-containing proteins 6 and 4 (CMTM6 and CMTM4) increase the stability of PD-L1 through downregulating ubiquitination-dependent degradation and lysosome-dependent proteolysis, thus enhancing the ability of tumor cells to suppress immune responses, and providing a new target for combinatorial immunotherapy 37,124,125 . In fact, the transport between recycling endosomes and lysosomes controls the fate of the PD-L1 protein 37,126 , although the exact mechanism of lysosomal-dependent degradation of PD-L1 is incompletely understood.
Antitumor immunity is enhanced by inhibiting PD-L1, on the basis of the molecular regulation of PD-L1 in tumor cells.
Our previous studies have shown that depletion of huntingtin-interacting protein 1-related protein (HIP1R) in tumor cells leads to significant upregulation of PD-L1, thus resulting in the suppression of T cell cytotoxicity 24 . Further research has shown that HIP1R is a regulator of PD-L1 lysosomal degradation that controls PD-L1 homeostasis. HIP1R physically interacts with PD-L1 and transports PD-L1 to lysosomes through a lysosomal targeting signal. HIP1R is an endocytic adaptor protein that contains homology domains responsible for the binding of clathrin, inositol lipids and F-actin 127 . HIP1R binds PD-L1 through its conserved C-terminal domain and uses an intrinsic sorting signal to deliver PD-L1 to lysosomes for degradation 128 . HIP1R targets PD-L1 for lysosomal degradation, thereby enhancing T cell-mediated cytotoxicity, and it is a natural regulator of lysosomal degradation. On the basis of the 'binding-sorting' model derived from the molecular roles of HIP1R, we have rationally designed the peptide PD-LYSO, incorporating the lysosome-sorting signal and the PD-L1-binding sequence of HIP1R, and used it to successfully deplete PD-L1 expression in tumor cells 129 . Other researchers have identified SA-49 as a novel regulator of PD-L1 expression from a series of novel aloperine derivatives. They have found that SA-49-induced microphthalmia transcription factor (MITF) translocation functions through activation of PKCα and subsequent suppression of GSK3β activity, thus increasing lysosome biogenesis and promoting translocation of PD-L1 to lysosomes for proteolysis 126 . In breast cancer, another study has identified a disintegrin and metalloproteinase 10 (ADAM10) and ADAM17 as enzymes mediating PD-L1 cleavage. The cleavage generates a free N-terminal fragment and a C-terminal fragment that remains associated with cells but is efficiently eliminated by lysosomal degradation 130 . Researchers are increasingly focusing on exploring more techniques for targeting proteins for lysosomal degradation, including endosome targeting chimeras (ENDTACs), lysosome targeting chimeras (LYTACs) 24 ( Table 2).
Thus, the discovery of HIP1R-mediated lysosomal degradation of PD-L1 has provided a potential new route for inhibiting PD-L1. PD-LYSO should be beneficial in the development and optimization of lysosomal targeting strategies as a crucial target for combinatorial immunotherapy.

Palmitoylation blockade triggers degradation of PD-L1 and PD-1
The intracellular storage and redistribution of PD-L1 to cell membranes minimize the therapeutic benefits 137 . The cytoplasmic domain of PD-L1 is palmitoylated, and this lipid modification stabilizes PD-L1 by preventing its ubiquitination,  139,140 . Palmitate is usually associated with cysteine residues through thioester bonds, in a process that may be catalyzed by aspartic acidhistidine-histidine-cysteine (DHHC) palmitosyltransferase 141 . Palmitoylation is a reversible lipid modification of proteins that controls a variety of protein functions, such as transport, activity, stability, and membrane association 142 . Palmitoylation has been shown to regulate the transportation and function of multiple cancer-related proteins 143,144 . Research has indicated that palmitoylation plays an important role in the regulation of PD-L1 protein stability and trafficking and has identified the palmitoyltransferase ZDHHC3 (DHHC3) as the main acetyltransferase required for the palmitoylation of PD-L1 145 . Palmitoylation decreases the lysosomal degradation of PD-L1. The compound 2-bromopalmitate, a small-molecule inhibitor of palmitoylation, blocks the palmitoylation of PD-L1 and effectively induces lysosomal degradation of PD-L1 in tumor cells, thus enhancing the cytotoxicity of tumor-specific T cells 24, 138 . The lack of specificity is a major challenge in targeting palmitoylation with existing palmitoylation inhibitors. Apart from the PD-L1-related adverse effects, 2-bromopalmitate might cause adverse effects related to its inhibitory effects on other palmitoylated proteins. PD-PALM, a PD-L1 palmitoylation inhibitor, has been designed to competitively inhibit PD-L1 palmitoylation. The application of PD-PALM decreases PD-L1 expression in tumor cells and enhances T cell activity 24 . Hence, inhibiting palmitoylation of PD-L1 may decrease PD-L1 expression on the cell membrane and deplete its storage capacity in recycling endosomes.
Palmitoylation-based targeting methods may provide more powerful and long-lasting inhibitory effects because they inhibit PD-L1 protein levels throughout the cell and therefore may represent a promising therapeutic avenue toward enhancing tumor-specific immunity.

Challenges and possibilities in future studies
Because most targeted degradation mechanisms rely on intracellular binding and sorting, the cell-penetration ability of therapeutic molecules poses a major challenge. The peptidic nature of PD-LYSO and PD-PALM also makes in vivo stability an outstanding challenge, because peptides are generally prone to rapid degradation in the serum, through the action of various enzymes. To achieve a longer half-life, commonly used approaches, such as PEGylation, bovine serum albumin fusion, and Fc fusion, may significantly increase the molecular size and prohibit intracellular delivery. Moreover, potential immunogenic effects should be considered, because peptides may stimulate the generation of neutralizing antibodies in vivo. Because of these challenges, developing small molecules to mimic the conformation and functions of peptides is highly preferable and would represent a major step toward successful drug development.