Tumor neoantigens
We have mentioned cancer immunotherapy in the section on small molecular drug screening. This part will overview the detailed mechanism of tumor neoantigens and cancer immunity, then discuss the application of organoids in identifying tumor antigenic peptides.
Tumor neoantigens (also known as tumor-specific antigens) are specifically present on tumor cells but not on normal cells and can be recognized by T lymphocytes 99. Neoantigens are peptides with eight to ten amino acids produced by the degradation proteasome of intracellular proteins which are the translational products of mutant oncogenesis100. These antigenic peptides are then associated with major histocompatibility complex (MHC) class Ⅰ molecules (human leukocytes antigen (HLA)) in the endoplasmic reticulum and migrate to the cell membrane. Activated CD8+ T lymphocytes recognize and bind to cancer cells and kill their target cells. This process is a cycle in which the tumor cell death results in the release of more antigens and improves the immune response to tumor cells. Notably, tumor neoantigens are captured by dendritic cells (DCs), then present to T lymphocytes for activating them. Therefore, the core of cancer immunotherapy is to activate the killing function of activated T lymphocytes or promote the presentation of neoantigens101.
Currently, a wide range of studies focuses on identifying tumor neoantigens, which is critical for the development of new treatment modalities for cancers, such as peptide-based tumor vaccines and personalized drugs to kill tumor cells102. Mass spectrometry (MS)-based immunopeptidomics and computational predictions are the commonly used approaches to identify possible antigenic peptides99. The latter is based on computer-generated algorithms to identify peptides that could be produced by mutated genes in tumors and are likely to associate with MHC molecules103. The selected antigenic peptides are synthesized and used to activate T lymphocytes. The activated cytotoxic T lymphocyte (CTL, CD8+ T lymphocytes) are then co-cultured with tumor cells to evaluate the CTL response for neoantigen screens104, 105. Owing to the limited source of patient-derived tumor tissues, studies have focused on using patient-derived organoids to find out the mutated oncogenes and investigate whether the activated CTLs can kill cancer cells. Newey et al. expanded patient-derived colorectal cancer (CRC) organoids and demonstrated the feasibility of MS-based immunopeptidomics of CRC organoids in investigating neoantigen presentation in vitro106. Wang et al. generated patient-derived hepatobiliary tumor organoids and found that they preserve most of the characteristics of their parental tissues, such as genetic features and neoantigen landscape. They used organoids as preclinical models to identify the predicted-peptide activated CTLs that exhibited anti-tumor activity107. This study provides evidence for the application of tumor organoids as preclinical models for rapid antigenic peptide validation through a prediction-based approach. Few cancer patients share the same neoantigens, and more than 99.95% of neoantigens are present in only one patient resulting from tumor heterogeneity108, 109. Patient-derived tumor organoids could avoid these issues and achieve personalized immunotherapy in the future. Since tumor heterogeneity is also characterized by different cell subgroups within the same tumor tissue, Demmers et al. cultured single cell-derived CRC organoid clones from the same patient and demonstrated that the HLA class Ⅰ peptide presentation landscape was heterogeneous even within one individual110. They also indicated that highly conserved antigenic peptides in HLA presentation could be identified using the single-cell derived clonal organoids, which may be a suitable choice for designing anti-tumor vaccines.
The current studies that used organoids to perform tumor antigenic peptide screens are 1ow-throughput. Patient-derived organoids enable high-throughput screening of a large number of computationally-predicted peptides. Furthermore, organoids can undergo extensive expansion, allowing large quantities of material for MS-based immunopeptidomics analysis (Figure 2). Therefore, we propose that the utility of patient-derived organoids as preclinical models to high-throughput identify tumor neoantigens could be a powerful approach to the development of precision therapies for cancer patients.
Cell-penetrating peptides
Cell-penetrating peptides (CPPs) can cross tissues and cell surfaces without causing lethal injury to the membranes. The mechanism of the penetration process remains controversial. Most CPPs are endogenously produced proteins and peptides, including heparin-binding proteins, DNA-binding proteins, antimicrobial peptides and viral peptides111. Given the characteristics of CPPs to cross cell membranes, more and more researchers focus on whether they can pass the blood-brain barrier (BBB)112. The BBB is a complex microvasculature system mainly consisting of brain endothelial cells (ECs) that are tightly lined in the cerebral vascular lumens. The main function of BBB is to protect the brain. In addition to ECs, astrocytic glia and neurons together are organized into well-structured neurovascular units113. The brain ECs express high levels of tight junction proteins, efflux pumps, and specific transporters. The tight junctions between ECs prevent molecules in the blood from entering the central nervous system. The efflux pumps, including P-glycoprotein (PgP), exclude foreign substances from the brain. Specific transporters deliver essential nutrients to the brain, such as glucose and amino acids. Therefore, BBB is the main obstacle to delivering drugs to neural cells and developing effective treatments for central nervous system diseases. Although CPPs are promising vectors to deliver drugs across the BBB, no CPP-based treatment is currently used clinical practice, partly due to the lack of suitable preclinical models that can accurately mimic the features of BBB to discover potential CPPs in a high throughput way114. The widely used in vitro models for BBB are co-culturing of brain ECs (top), astrocytes and pericytes (bottom) in a transwell system115-117. Microfluidics could introduce blood flow to stimulate BBB more dynamically118. However, these model systems require advanced equipment to establish the platform, increasing the experimental complexity of performing high-throughput drug screens.
Researchers recently reported three-dimensional multicellular structures through self-organization arrangement of brain ECs, pericytes and astrocytes119. The in vitro spheroids can recapitulate the complex interactions between each cell type, which is critical cell behavior to maintain the essential function of BBB. Lawler’s group modified the method and established BBB organoids that could mimic the essential function of BBB120. They, for the first time, investigate whether the BBB organoids are suitable for screening BBB-penetrating drugs. BBB organoids were used to identify several CPPs that could cross the BBB, demonstrating their feasibility and utility as models for cost-effective and high-throughput drug screens. Many peptides are susceptible to being degraded by proteolytic, and show a relatively low ability to cross the BBB. The same group found that peptide macrocyclization could increase the cell uptake of CPPs and found that one macrocyclic analog of transpotan-10 displays improved capacity to deliver across the BBB organoids121.
Together these studies demonstrated that three-dimensional multicellular BBB organoids can capitulate the complex interactions and arrangements of each cell type, and can reduce experimental complexity (Figure 2). These advantages make BBB organoids ideal preclinical models for high-throughput screening of CPPs that can cross BBB114. Furthermore, CPPs are also promising carriers to transfer therapeutic drugs into tumor cells. Therefore, it is necessary to establish tumor organoids that are amenable to the discovery of efficient CPPs with high stability, internalization ability and specificity.
Host defense peptides
The primary biological functions of the naturally-produced peptides, host defense peptides (HDPs, also known as antimicrobial peptides), are immunomodulatory, anti-inflammatory, and anti-bacterial122. HDP are peptides with 12–50 amino acids composed of cationic and hydrophobic amino acids that adopt an amphipathic conformation upon folding, usually after contact with membranes123, 124. Increasing antimicrobial resistance (AMR) organisms has become a severe issue for the treatment of inflammatory diseases due to the excessive use of antibiotics125, 126. The naturally-produced HDPs are promising candidates for developing treatments against the global threat caused by AMR organisms126. Similarly, there is also an urgent need to establish novel model systems to screen HDPs as potential drug candidates. Currently, an organoid system called air-liquid interface (ALI) construct are widely used in investigating HDPs126. ALI systems are comprised of a porous filter separating the apical and basolateral compartments. Cells cultured on the top chambers grow to multilayers and across the basal-apical threshold, where the medium remains on the bottom of the cell layers and the apical interface is surrounded by air. This system is ideal for studying the biology of tissues interacting with liquid and air in vivo127. Patients and healthy lung ALI models were used to examine the HDP expressions. The results demonstrated that they are suitable models to study the essential function of HDPs in respiratory diseases and enable the identification of drug candidates128-130. One study using primary ALI models reported that the frog skin-derived HDPs Esc (1-21) and its synthetic derivation both could protect the epithelial integrity when infected by P. aeruginosa131. Ritter et al. used ALI cultures with an aerosol delivery system to predict acute local lung toxicity through the assessment of various combinations of HDPs and nanocarriers132. This study also indicated the sensitivity of ALI models compared to submerged cultures.
Current studies introducing HDPs to organoid systems are all using ALI models, which involve complex manipulations and equipment, making them unsuitable for high-throughput screens. We believe that real three-dimensional organoid models (infected with microbes) are a potential approach to discovering effective HDPs and their derivations in a high-throughput way.