Linker Design Research Articles

Deep Lead Optimization: Leveraging Generative AI for Structural Modification.

The integration of deep learning-based molecular generation models into drug discovery has garnered significant attention for its potential to expedite the development process. Central to this is lead optimization, a critical phase where existing molecules are refined into viable drug candidates. As various methods for deep lead optimization continue to emerge, it is essential to classify these approaches more clearly. We categorize lead optimization methods into two main types: goal-directed and structure-directed. Our focus is on structure-directed optimization, which, while highly relevant to practical applications, is less explored compared to goal-directed methods. Through a systematic review of conventional computational approaches, we identify four tasks specific to structure-directed optimization: fragment replacement, linker design, scaffold hopping, and side-chain decoration. We discuss the motivations, training data construction, and current developments for each of these tasks. Additionally, we use classical optimization taxonomy to classify both goal-directed and structure-directed methods, highlighting their challenges and future development prospects. Finally, we propose a reference protocol for experimental chemists to effectively utilize Generative AI (GenAI)-based tools in structural modification tasks, bridging the gap between methodological advancements and practical applications.

Diffusion-based generative drug-like molecular editing with chemical natural language

Recently, diffusion models have emerged as a promising paradigm for molecular design and optimization. However, most diffusion-based molecular generative models focus on modeling 2D graphs or 3D geometries, with limited research on molecular sequence diffusion models. The International Union of Pure and Applied Chemistry (IUPAC) names are more akin to chemical natural language than the Simplified Molecular Input Line Entry System (SMILES) for organic compounds. In this work, we apply an IUPAC-guided conditional diffusion model to facilitate molecular editing from chemical natural language to chemical language (SMILES) and explore whether the pre-trained generative performance of diffusion models can be transferred to chemical natural language. We propose DiffIUPAC, a controllable molecular editing diffusion model that converts IUPAC names to SMILES strings. Evaluation results demonstrate that our model outperforms existing methods and successfully captures the semantic rules of both chemical languages. Chemical space and scaffold analysis show that the model can generate similar compounds with diverse scaffolds within the specified constraints. Additionally, to illustrate the model's applicability in drug design, we conducted case studies in functional group editing, analogue design and linker design.

Non-Markovian Dynamic Models Identify Non-Canonical KRAS-VHL Encounter Complex Conformations for Novel PROTAC Design.

Targeted protein degradation (TPD) is emerging as a promising therapeutic approach for cancer and other diseases, with an increasing number of programs demonstrating its efficacy in human clinical trials. One notable method for TPD is Proteolysis Targeting Chimeras (PROTACs) that selectively degrade a protein of interest (POI) through E3-ligase induced ubiquitination followed by proteasomal degradation. PROTACs utilize a warhead-linker-ligand architecture to bring the POI (bound to the warhead) and the E3 ligase (bound to the ligand) into proximity. The resulting non-native protein-protein interactions (PPIs) formed between the POI and E3 ligase lead to the formation of a stable ternary complex, enhancing cooperativity for TPD. A significant challenge in PROTAC design is the screening of the linkers to induce favorable non-native PPIs between POI and E3 ligase. Here, we present a physics-based computational protocol to predict noncanonical and metastable PPI interfaces between an E3 ligase and a given POI, aiding in the design of linkers to stabilize the ternary complex and enhance degradation. Specifically, we build the non-Markovian dynamic model using the Integrative Generalized Master equation (IGME) method from ∼1.5 ms all-atom molecular dynamics simulations of linker-less encounter complex, to systematically explore the inherent PPIs between the oncogene homologue protein and the von Hippel-Lindau E3 ligase. Our protocol revealed six metastable states each containing a different PPI interface. We selected three of these metastable states containing promising PPIs for linker design. Our selection criterion included thermodynamic and kinetic stabilities of PPIs and the accessibility between the solvent-exposed sites on the warheads and E3 ligand. One selected PPIs closely matches a recent cocrystal PPI interface structure induced by an experimentally designed PROTAC with potent degradation efficacy. We anticipate that our protocol has significant potential for widespread application in predicting metastable POI-ligase interfaces that can enable rational design of PROTACs.

Thiatruxene-Based Conductive MOF: Harnessing Sulfur Chemistry for Enhanced Proton Transport.

Functionalizing the organic building blocks of electrically conductive MOFs (EC-MOFs) can be a powerful method for adjusting the electronic structure and introducing a specific chemistry. However, designing EC-MOF linkers with reactive functional groups for postsynthetic modification is challenging due to the requirements of d-p conjugation. This work addresses such design limitations by synthesizing an EC-MOF, Cu-thiatruxene (Cu-thiaTRX). This conductive framework incorporated a truxene-based linker with heterocyclic sulfur, allowing for efficient conjugation and an electrical conductivity of 2.2 × 10-2 S cm-1. Harnessing sulfur chemistry in Cu-thiaTRX involves a two-step postsynthetic modification: oxidation and SNAr. The sulfinic groups introduced in the framework enabled tunable proton conductivity, leading to a 200-fold improvement. These results highlight the importance of a rational linker design for functionalization.

Rational Linker Design for Enhanced Performance of MOF-Derived Catalysts in Electrochemical Urea Synthesis.

2D metal-organic frameworks (2D MOFs) offer promising electrocatalytic potential for urea synthesis, yet the underlying reaction mechanisms and structure-activity relationships remain unclear. Using Cu-BDC as a model, density functional theory (DFT) calculations to elucidate these aspects are conducted. The results reveal a novel coupling mechanism involving *NO─CO and *NO─*ONCO, emphasizing the impact of linker modifications on Cu spin states and charge distribution. Notably, Cu-BDC-NH2 and Cu─BDC─OH emerge as promising catalysts. Additionally, structure-activity relationships through descriptors like d-band center, IE ratio, and L(Cu─O), providing insights for rational catalyst design is established. These findings pave the way for optimized catalysts and sustainable urea production, opening avenues for future research and technological advancements.

Unlocking comprehensive molecular design across all scenarios with large language model and unordered chemical language.

Molecular generation stands at the forefront of AI-driven technologies, playing a crucial role in accelerating the development of small molecule drugs. The intricate nature of practical drug discovery necessitates the development of a versatile molecular generation framework that can tackle diverse drug design challenges. However, existing methodologies often struggle to encompass all aspects of small molecule drug design, particularly those rooted in language models, especially in tasks like linker design, due to the autoregressive nature of large language model-based approaches. To empower a language model for a wider range of molecular design tasks, we introduce an unordered simplified molecular-input line-entry system based on fragments (FU-SMILES). Building upon this foundation, we propose FragGPT, a universal fragment-based molecular generation model. Initially pretrained on extensive molecular datasets, FragGPT utilizes FU-SMILES to facilitate efficient generation across various practical applications, such as de novo molecule design, linker design, R-group exploration, scaffold hopping, and side chain optimization. Furthermore, we integrate conditional generation and reinforcement learning (RL) methodologies to ensure that the generated molecules possess multiple desired biological and physicochemical properties. Experimental results across diverse scenarios validate FragGPT's superiority in generating molecules with enhanced properties and novel structures, outperforming existing state-of-the-art models. Moreover, its robust drug design capability is further corroborated through real-world drug design cases.

Structure-Based Drug Design with a Deep Hierarchical Generative Model.

Recently, the remarkable growth of available crystal structure data and libraries of commercially available or readily synthesizable molecules have unlocked previously inaccessible regions of chemical space for drug development. Paired with improvements in virtual ligand screening methods, these expanded libraries are having a notable impact on early drug design efforts. Yet screening-based methods still face scalability limits, due to computational constraints and the sheer scale of drug-like space. Machine learning approaches are overcoming these limitations by learning the fundamental intra- and intermolecular relationships in drug-target systems from existing data. Here, we introduce DrugHIVE, a deep hierarchical variational autoencoder that outperforms state-of-the-art autoregressive and diffusion-based methods in both speed and performance on common generative benchmarks. DrugHIVE's hierarchical design enables improved control over molecular generation. Its capabilities include dramatically increasing virtual screening efficiency and accelerating a wide range of common drug design tasks, including de novo generation, molecular optimization, scaffold hopping, linker design, and high-throughput pattern replacement. Our highly scalable method can even be applied to receptors with high-confidence AlphaFold-predicted structures, extending the ability to generate high-quality drug-like molecules to a majority of the unsolved human proteome.

Optimizing structural design in SN38 delivery: More assembly stability and activation efficiency

Dimeric prodrug self-assembly nanoparticles (DPS NPs) present promising avenues for chemotherapeutic delivery, yet optimizing the linker design for effective SN38 delivery remains challenging. We developed various SN38 dimeric prodrugs with differing linker lengths to explore how linker length impacts DPS NP performance. Our study reveals that linker length critically affects the nanoparticles assembly stability and activation efficiency. Specifically, too short linkers compromise assembly stability and lead to premature drug activation in the bloodstream, raising safety concerns. Conversely, too long linkers hinder both assembly stability and activation efficiency within tumor cells, diminishing anti-tumor effectiveness. The optimal linker (C12) achieved the best balance, ensuring robust assembly stability and high activation efficiency, thereby enhancing anti-tumor efficacy while maintaining a favorable safety profile. This work underscores the significance of linker length in designing effective DPS NPs for cancer treatment.

Photo-Claisen Rearrangement in a Coumarin-Caged Peptide Leads to a Surprising Enzyme Hyperactivation.

Protein phosphatase-1 (PP1) is a ubiquitous enzyme that counteracts hundreds of kinases in cells. PP1 interacts with regulatory proteins via an RVxF peptide motif that binds to a hydrophobic groove on the enzyme. PP1-disrupting peptides (PDPs) compete with these regulatory proteins, leading to the release of the active PP1 subunit and promoting substrate dephosphorylation. Building on previous strategies employing the ortho-nitrobenzyl (o-Nb) group as a photocage to control PDP activity, we introduced coumarin derivatives into a PDP via an ether bond to explore their effects on PP1 activity. Surprisingly, our study revealed that the coumarin-caged peptides (PDP-DEACM and PDP-CM) underwent a photo-Claisen rearrangement, resulting in an unexpected hyperactivation of PP1. Our findings underscore the importance of linker design in controlling uncaging efficiency of photocages and highlight the need for comprehensive in vitro analysis before cellular experiments.

Enhancing structural control in covalent organic frameworks through steric interaction-driven linker design.

Covalent Organic Frameworks (COFs) exhibiting kagome (kgm) structures are promising crystalline porous materials with two distinct pores. However, there are no reliable synthetic methods to exclusively target the kgm over the polymorphic square-lattice (sql) structure. To address this, we introduce a linker design strategy featuring bulky functional groups, which through steric interactions can hinder the sql net formation, thereby leading to a kgm structure. By rigid attachment of the methyl benzoate groups to a tetradentate COF linker, steric interactions with neighbouring linkers depending on the pore size become possible. The steric interaction was tuned by varying the complementary bidentate linear linker lengths, where the shorter phenylenediamine linker leads to steric hindrance and the formation of the kgm lattice, while with the longer benzidine linker, steric interaction is reduced leading to the sql lattice. Thus, control over the net can be exerted through steric interaction strengths. Additionally, structural analysis revealed the formation of the kgm COF with an unusual ABC stacking, leading to pearl string type pores instead of two distinct pore sizes. This COF system shows that steric interaction-driven design enhances control over COF structures, expanding the design toolbox, but also provides valuable insights into network formation and polymorphism.

Antibody-Calixarene Drug Conjugate: A General Drug Delivery Platform for Tumor-Targeted Therapy.

Antibody-drug conjugates (ADCs), which combine the precise targeting capabilities of antibodies with the powerful cytotoxicity of small-molecule drugs, have evolved into a promising approach for tumor treatment. However, the traditional covalent coupling method requires the design of a specific linker tailored to the properties of the small-molecule drugs, which greatly limits the development of ADCs and the range of drugs that can be used. Herein, a novel type of antibody-calixarene drug conjugates (ACDCs) that function similarly to ADCs by delivering drugs to their targets using antibodies but without the requirement of covalent conjugation of the drugs with antibodies is presented. By replacement of conventional linkers with supramolecular linkers, the ACDCs can load various chemotherapeutic drugs through host-guest interactions. Furthermore, ACDCs are readily reduced upon reaching the hypoxic microenvironment, resulting in rapid release of the drugs. With this precise drug encapsulation and controlled release mechanism, ACDCs deliver drugs to tumor tissues effectively and achieve a significantly enhanced antitumor effect. Considering that the ACDCs can be easily prepared by combining antibody-calixarene conjugates derived from tumor-targeting antibodies with various small-molecule drugs, ACDCs may provide a promising platform technology to accelerate ADC development and thus improve the therapeutic efficacy of chemotherapy.

Interplay of PROTAC Complex Dynamics for Undruggable Targets: Insights into Ternary Complex Behavior and Linker Design.

Protein degraders, such as bifunctional proteolysis-targeting chimeras (PROTACs), selectively eliminate target proteins by leveraging the natural protein degradation machinery. PROTACs bridge the target protein with an E3 ligase, which induces ubiquitination and degradation. Investigating ternary complex structures elucidates the molecular mechanisms of their formation and degradation. This study examines the binding dynamics of E3 ligases (VHL, CRBN, and cIAP) with proteins of interest, focusing on dynamics, cooperativity, selectivity, linker length, and PROTAC conformations. The influence of interface residues and linker lengths on specific conformations for target proteins and E3 ligases is highlighted. Utilizing molecular dynamics and steered molecular dynamics simulations, the study provides comprehensive parameters on the behavior and stability of diverse ternary complexes. These insights are crucial for designing PROTACs targeting disease-causing proteins and advancing the development of degradable ternary complexes for therapeutic applications.

Bismuth‐based metal‐organic frameworks and derivatives for photocatalytic applications in energy and environment: Advances and challenges

AbstractPhotocatalysis is an environmentally friendly technology for the utilizations of solar energy and has garnered significant attention in both scientific and industrial sectors. Developing cost‐effective semiconductive materials is the core issue in photocatalysis. Bismuth‐based metal‐organic frameworks (Bi‐MOFs) have emerged as attractive candidates in various photocatalytic applications, and Bi‐MOFs derivatives further expand and consolidate their promising potential in the realm of photocatalysis. Various modification strategies including in‐situ tailoring or external doping, as well as meticulous design and selection of metal nodes and organic linkers allow for fine control over the surface multifunctionality in Bi‐MOF‐based and derived photocatalytic composites with adjustable energy band structures and enhanced photocatalytic performance. In this review, the recent progress in the synthesis of diverse Bi‐MOFs‐based materials, Bi‐MOFs derivatives, and their Bi‐containing semiconductive composites were systemically analyzed and reviewed. The state‐of‐the‐art research progresses in the applications of Bi‐MOFs and derivatives, as well as composites in photocatalytic water splitting for hydrogen production, photodegradation of organic pollutants, and photocatalytic carbon dioxide reduction are comprehensively summarized. The relationships between structures, properties, and photocatalytic performance of Bi‐based semiconductive composites are discussed in detail. In addition, the perspectives and future challenges on Bi‐MOFs‐based and derived materials for photocatalytic applications are also offered.

An Introductory Guide to Protease Sensitive Linker Design Using Matrix Metalloproteinase 13 as an Example.

Proteases play a crucial role, not only in physiological, but also in pathological processes, such as cancer, inflammation, arthritis, Alzheimer's, and infections, to name but a few. Their ability to cleave peptides can be harnessed for a broad range of biotechnological purposes. To do this efficiently, it is essential to find an amino acid sequence that meets the necessary requirements, including interdependent factors like specificity, selectivity, cleavage kinetics, or synthetic accessibility. Cleavage sequences from natural substrates of the protease may not be optimal in terms of specificity and selectivity, which is why these frequently require arduous and sometimes unsuccessful optimization such as by iterative exchange of single amino acids. Hence, here we describe the systematic design of protease sensitive linkers (PSLs)─peptide sequences specifically cleaved by a target protease─guided by the mass spectrometry based determination of target protease specific cleavage sites from a proteome-based peptide library. It includes a procedure for identifying bespoke PSL sequences, their optimization, synthesis, and validation and introduces a program that can indicate potential cleavage sites by hundreds of enzymes in any arbitrary amino acid sequence. Thereby, we provide an introduction to PSL design, illustrated by the example of matrix metalloproteinase 13 (MMP13). This introduction can serve as a guide and help to greatly accelerate the development and use of protease-sensitive linkers in diverse applications.

A Copper-Based Metal-Organic Framework for Selective Separation of C2 Hydrocarbons from Methane at Ambient Conditions: Experiment and Simulation.

C2 hydrocarbon separation from methane represents a technological challenge for natural gas upgrading. Herein, we report a new metal-organic framework, [Cu2L(DEF)2]·2DEF (UNT-14; H4L = 4,4',4″,4‴-((1E,1'E,1″E,1‴E)-benzene-1,2,4,5-tetrayltetrakis(ethene-2,1-diyl))tetrabenzoic acid; DEF = N,N-diethylformamide; UNT = University of North Texas). The linker design will potentially increase the surface area and adsorption energy owing to π(hydrocarbon)-π(linker)/M interactions, hence increasing C2 hydrocarbon/CH4 separation. Crystallographic data unravel an sql topology for UNT-14, whereby [Cu2(COO)4]···[L]4- paddle-wheel units afford two-dimensional porous sheets. Activated UNT-14a exhibits moderate porosity with an experimental Brunauer-Emmett-Teller (BET) surface area of 480 m2 g-1 (vs 1868 m2 g-1 from the crystallographic data). UNT-14a exhibits considerable C2 uptake capacity under ambient conditions vs CH4. GCMC simulations reveal higher isosteric heats of adsorption (Qst) and Henry's coefficients (KH) for UNT-14a vs related literature MOFs. Ideal adsorbed solution theory yields favorable adsorption selectivity of UNT-14a for equimolar C2Hn/CH4 gas mixtures, attaining 31.1, 11.9, and 14.8 for equimolar mixtures of C2H6/CH4, C2H4/CH4, and C2H2/CH4, respectively, manifesting efficient C2 hydrocarbon/CH4 separation. The highest C2 uptake and Qst being for ethane are also desirable technologically; it is attributed to the greatest number of "agostic" or other dispersion C-H bond interactions (6) vs 4/2/4 for ethylene/acetylene/methane.

Synthesis and evaluation of antibody-drug conjugates with high drug-to-antibody ratio using dimaleimide-DM1 as a linker- payload

The notable characteristics of recently emerged Antibody-Drug Conjugates (ADCs) encompass the targeting of Human Epidermal growth factor Receptor 2 (HER2) through monoclonal antibodies (mAbs) and a high ratio of drug to antibody (DAR). The achievements of Kadcyla® (T-DM1) and Enhertu® (T-Dxd) have demonstrated that HER2-targeting antibodies, such as trastuzumab, have shown to be competitive in terms of efficacy and price for development. Furthermore, with the arrival of T-Dxd and Trodelvy®, high-DAR (7–8) ADCs, which differ from the moderate DAR (3–4) ADCs that were formerly regarded as conventional, are being acknowledged for their worth. Following this trend of drug development, we endeavored to develop a high-DAR ADC using a straightforward approach involving the utilization of DM1, a highly potent substance, in combination with the widely recognized trastuzumab. To achieve a high DAR, DM1 was conjugated to reduced cysteine through the simple design and synthesis of various dimaleimide linkers with differing lengths. Using LC and MS analysis, we have demonstrated that our synthesis methodology is uncomplicated and efficacious, yielding trastuzumab-based ADCs that exhibit a remarkable degree of uniformity. These ADCs have been experimentally substantiated to exert an inhibitory effect on cancer cells in vitro, thus affirming their value as noteworthy additions to the realm of ADCs.

Equivariant 3D-conditional diffusion model for molecular linker design

AbstractFragment-based drug discovery has been an effective paradigm in early-stage drug development. An open challenge in this area is designing linkers between disconnected molecular fragments of interest to obtain chemically relevant candidate drug molecules. In this work, we propose DiffLinker, an E(3)-equivariant three-dimensional conditional diffusion model for molecular linker design. Given a set of disconnected fragments, our model places missing atoms in between and designs a molecule incorporating all the initial fragments. Unlike previous approaches that are only able to connect pairs of molecular fragments, our method can link an arbitrary number of fragments. Additionally, the model automatically determines the number of atoms in the linker and its attachment points to the input fragments. We demonstrate that DiffLinker outperforms other methods on the standard datasets, generating more diverse and synthetically accessible molecules. We experimentally test our method in real-world applications, showing that it can successfully generate valid linkers conditioned on target protein pockets.

Diselenide-Bridged Doxorubicin Dimeric Prodrug: Synthesis and Redox-Triggered Drug Release.

The diselenide bond has attracted intense interest in redox-responsive drug delivery systems (DDSs) in tumor chemotherapy, due to its higher sensitivity than the most investigated bond, namely the disulfide bond. Here, a diselenide-bridged doxorubicin dimeric prodrug (D-DOXSeSe) was designed by coupling two doxorubicin molecules with a diselenodiacetic acid (DSeDAA) molecule via α-amidation, as a redox-triggered drug self-delivery system (DSDS) for tumor-specific chemotherapy. The drug release profiles indicated that the D-DOXSeSe could be cleaved to release the derivatives selenol (DOX-SeH) and seleninic acid (DOX-SeOOH) with the triggering of high GSH and H2O2, respectively, indicating the double-edged sword effect of the lower electronegativity of the selenide atom. The resultant solubility-controlled slow drug release performance makes it a promising candidate as a long-acting DSDS in future tumor chemotherapy. Moreover, the interaction between the conjugations in the design of self-immolation traceless linkers was also proposed for the first time as another key factor for a desired precise tumor-specific chemotherapy, besides the conjugations themselves.

Abstract LB057: Pre-clinical characterization of IBI343, a site-specifically conjugated anti-Claundin18.2 ADC, for treating solid tumors

Abstract Claudin18.2 is important for the formation and function of tight junctions between cells in epithelial tissues. Normally expressed only in gastric mucosa, Claudin18.2 is overexpressed in multiple types of solid tumors, including gastric, pancreatic, breast and colon cancers. Despite being a well-established tumor-associated antigen, the clinical value and feasibility of targeting Claudin18.2 by various modalities are yet-to-be fully demonstrated. IBI343 is a fully humanized anti-Claudin18.2 monoclonal antibody conjugated to Exatecan via site-specific glycol-conjugation with an average drug-to-antibody ratio (DAR) of 3.6. IBI343 showed robust antigen-specific in vitro cytotoxicity in cancer cell lines with varying levels of Claudin18.2 expression. In addition, due to the cleavable linker design and high hydrophilicity of the payload, IBI343 demonstrated potent bystander killing effects in vitro. The glycan-based conjugation technology leads to enhanced stability of the conjugated linker-payload and the entire ADC molecule, which results in superior pharmacokinetics. IBI343 showed potent in vivo anti-tumor efficacy with complete tumor regressions in multiple xenograft models. Furthermore, it demonstrated strong in vitro and in vivo synergistic effects with various anti-cancer agents, including immune checkpoint blockers and DNA damage repair pathway inhibitors. IBI343 displayed good safety profile in monkey GLP toxicology study (HNSTD = 30 mg/kg). These pre-clinical findings show that IBI343 is a promising anti-cancer agent with large therapeutic window, and support clinical exploration of IBI343 in patients with Claudin18.2-expressing tumors. A companion late-breaking abstract of clinical phase 1 study results (NCT05458219) will also be presented in AACR 2024. Citation Format: Shuaixiang Zhou, Xiong Yao, Jian Guan, Keke Fei, Jia Lu, Weiwei Wu, Yang Liu, Tianyu Zhu, Zhuangguang Liao, Shi Chen, Bingliang Chen, Kaijie He. Pre-clinical characterization of IBI343, a site-specifically conjugated anti-Claundin18.2 ADC, for treating solid tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB057.

Fine-tuning the activation behaviors of ternary modular cabazitaxel prodrugs for efficient and on-target oral anti-cancer therapy

The disulfide bond plays a crucial role in the design of anti-tumor prodrugs due to its exceptional tumor-specific redox responsiveness. However, premature breaking of disulfide bonds is triggered by small amounts of reducing substances (e.g., ascorbic acid, glutathione, uric acid and tea polyphenols) in the systemic circulation. This may lead to toxicity, particularly in oral prodrugs that require more frequent and high-dose treatments. Fine-tuning the activation kinetics of these prodrugs is a promising prospect for more efficient on-target cancer therapies. In this study, disulfide, steric disulfide, and ester bonds were used to bridge cabazitaxel (CTX) to an intestinal lymph vessel-directed triglyceride (TG) module. Then, synthetic prodrugs were efficiently incorporated into self-nanoemulsifying drug delivery system (corn oil and Maisine CC were used as the oil phase and Cremophor EL as the surfactant). All three prodrugs had excellent gastric stability and intestinal permeability. The oral bioavailability of the disulfide bond-based prodrugs (CTX-(C)S-(C)S-TG and CTX-S-S-TG) was 11.5- and 19.1-fold higher than that of the CTX solution, respectively, demonstrating good oral delivery efficiency. However, the excessive reduction sensitivity of the disulfide bond resulted in lower plasma stability and safety of CTX-S-S-TG than that of CTX-(C)S-(C)S-TG. Moreover, introducing steric hindrance into disulfide bonds could also modulate drug release and cytotoxicity, significantly improving the anti-tumor activity even compared to that of intravenous CTX solution at half dosage while minimizing off-target adverse effects. Our findings provide insights into the design and fine-tuning of different disulfide bond-based linkers, which may help identify oral prodrugs with more potent therapeutic efficacy and safety for cancer therapy.