English

• 1. Introduction

  In the evolving landscape of cancer treatment, chemotherapy stands as a cornerstone, leveraging chemical agents to destroy tumor cells or hinder their growth and multiplication [1]. This method plays a crucial role in the arsenal against cancer, targeting the uncontrolled cell division characteristic of the disease [2]. However, traditional chemotherapy drugs, including paclitaxel (PTX), doxorubicin (DOX), and nitrogen mustards, have significant drawbacks [3-5]. While they effectively target rapidly dividing cancer cells, their lack of specificity means they also impact healthy cells that proliferate quickly, leading to a range of adverse effects. A notable example is DOX, which accumulates cardiac cells, inducing the production of reactive oxygen species (ROS) that mediate cellular damage, impair myocardial contractility, and can lead to congestive heart failure [6]. In response to these challenges, the past two decades have seen a concerted effort by the scientific community to develop cancer therapies with greater precision in their targeting capabilities [7]. These targeted therapies are designed to exploit specific biomarkers overexpressed in tumor cells, allowing for the concentration of therapeutic agents directly at the needed sites [8]. This approach not only maximizes the therapeutic potential of the drugs but also significantly reduces their toxic side effects [9]. The therapies can be broadly categorized into two groups: Small-molecule therapy, such as kinase inhibitors and small molecule-drug conjugates (SMDCs); and macromolecular therapy, including monoclonal antibodies, antibody-drug conjugates (ADCs), and others [9-12].

  ADCs represent a cutting-edge strategy in the realm of targeted cancer therapies. These innovative constructs are composed of a monoclonal antibody that is specifically designed to target tumor cells, chemically linked to a potent cytotoxic payload [13]. This unique configuration allows ADCs to distinguish themselves from traditional chemotherapy, which typically lacks specificity, indiscriminately binding to and destroying both cancerous and healthy cells. The selective nature of ADCs enables them to deliver their cytotoxic payloads directly to tumor cells, thus minimizing collateral damage to healthy tissues. This targeted approach not only enhances the precision of treatment but also significantly increases the therapeutic index, making ADCs a more effective and safer option for cancer patients [14]. Since the U.S. Food and Drug Administration (FDA) approved the first ADC, Mylotarg (gemtuzumab ozogamicin), for the treatment of adult acute myeloid leukemia (AML) in 2000, the field of ADCs has seen remarkable growth and development [15]. Over the years, the number of FDA-approved ADCs has steadily increased, with thirteen ADCs approved as of 2023, marking a significant milestone in the advancement of targeted cancer therapies [16]. This expansion is reflected in the commercial success of the ADC market, which surpassed ten billion dollars in revenue in 2023. Projections indicate that this market will continue to grow at a robust pace, with estimates suggesting it could reach a staggering 647 billion dollars by 2030, driven by an impressive annual growth rate of 30.0% [17,18]. However, despite these successes, ADCs are not without their challenges. The large molecular size of ADCs can complicate their pharmacokinetics, leading to issues with distribution, metabolism, and clearance in the body. Additionally, the complexity of their molecular design and synthesis contributes to high production costs, posing a significant barrier to widespread clinical application [19-21]. As clinical experience with ADCs has grown, these challenges have become increasingly apparent, prompting ongoing research and innovation to overcome these obstacles and further enhance the efficacy and accessibility of ADCs in cancer treatment.

  In response to these issues, SMDCs have emerged as a similar conceptual design [22]. By employing targeted small molecules instead of antibodies for both drug delivery and tumor targeting, SMDCs offer a precise and innovative approach [23]. These conjugates consist of a small molecule ligand, tailored to target specific biomarkers overexpressed in tumor cells and a cytotoxic payload [24]. This combination is either directly linked or connected via a linker, designed to ensure stability in the bloodstream, precise targeting, and efficient release of the cytotoxic agent near the tumor [12,25]. The shift from antibody-based targeting, which typically involves large molecules of over 1000 amino acids (~150 kDa), to small molecule ligands (<5 kDa) introduces numerous benefits (Table 1) [19,25]. SMDCs boast a smaller molecular weight, allowing for easier vascular penetration and more uniform distribution within tumor tissues without accumulating in healthy tissues, thus significantly reducing toxicity [25,26]. Additionally, their synthesis is more cost-effective, promising lower production costs for future mass production [27,28]. The absence of immunogenic components in the targeting ligands simplifies safety controls due to reduced immunogenicity [29]. Furthermore, SMDCs feature straightforward linkage points, minimizing heterogeneity and enhancing stability both in vitro and in vivo [30]. As a result, SMDCs address the limitations of ADCs and have emerged as a new generation of targeted anticancer therapies. 1 (Melphalan, Fig. 1), as the first-generation SMDC, has already received approval for the treatment of multiple myeloma, ovarian carcinoma, and uveal melanoma with unresectable liver metastases [31-33]. Additionally, other SMDCs such as 2 (TH-302, Fig. 1) for metastatic or locally advanced unresectable pancreatic adenocarcinoma, and 3 (QBS72S, Fig. 1) for the treatment of brain metastases in breast cancer, are currently undergoing Phase Ⅱ or Ⅲ clinical trials [34,35]. Given the rapid advancements and ongoing research in this field, it is crucial to critically examine the underlying mechanisms and design principles of SMDCs. This perspective explores the operational mechanisms of SMDCs, the principles behind selecting targeted biomarkers, and the design considerations for each component. Concluding with an overview of the latest progress in clinical trials of SMDCs currently under active development, emphasizing the transformative potential of SMDCs in cancer treatment, a prospective outlook on the future development of SMDCs is presented.

  Table 1

  

  Table 1.  Differences between SMDCs and ADCs.

  DownLoad: CSV

  Property	ADCs	SMDCs
  Targeting molecule type	Antibodies	Small molecules
  Molecular weight	~150 kDa	<5 kDa
  Solid tumor penetration	Low	High
  Tumor entry mechanism	Receptor-mediated endocytosis; fragmentation in tumor microenvironment	Simple diffusion; active transport; receptor-mediated endocytosis; fragmentation in tumor microenvironment
  Synthesis process and cost	Complex and expensive	Simple and cost-effective
  Heterogeneity	Present	Absent
  Immunogenicity	Present	Absent
  Pharmacokinetic profile	Complex	Simple
  Duration of action	Long	Short

  Figure 1

  

  Figure 1.  Structures of a portion of disclosed SMDCs that have been marketed or are currently active in clinical trials.

  DownLoad: Full-Size Img PowerPoint

  2. Mechanisms of SMDCs

  SMDCs represent a novel approach in targeted cancer therapy, leveraging the precision of small molecules to deliver cytotoxic payloads specifically to tumor cells [36]. By targeting biomarkers that are overexpressed on the surface of tumor cells, SMDCs not only activate the therapeutic potential of these cytotoxic agents but also significantly enhance their activity and selectivity, thereby reducing off-target toxicity [37]. Notably, employing SMDCs as carriers for drug delivery internalization significantly bolsters the activity and selectivity of these cytotoxic agents, simultaneously reducing off-target toxicity [35]. A proposed mechanism of SMDC action within the body, as illustrated in Fig. 2, underscores their sophisticated functionality. The effectiveness of SMDCs is determined by a combination of factors, including the choice of targeting ligand, the chemical nature of the linker, and crucially, the molecular weight of the conjugate [38]. The molecular weight, in particular, is instrumental in dictating how SMDCs gain entry into cells, a process pivotal for their therapeutic action [25,37,39]. In contrast to ADCs, SMDCs boast a significantly reduced molecular weight, more than 30 times lesser, facilitating their effortless penetration through the vascular barrier and ensuring a more homogeneous distribution within tumor tissues [40]. This reduced molecular size categorizes their cellular entry into two primary pathways: simple diffusion and receptor-mediated internalization, both of which are instrumental in the targeted eradication of cancer cells.

  Figure 2

  

  Figure 2.  Four distinct activation procedures of SMDCs in vivo: Simple diffusion (upper right); active transport (lower right); endocytosis (lower left); and fragmentation in the tumor microenvironment (upper left) (created with BioRender.com). The structures highlighted in green fluorescence denote the targeting ligands, those in yellow fluorescence represent the linkers, and the structures in blue fluorescence indicate the payloads.

  DownLoad: Full-Size Img PowerPoint

  In one scenario, SMDCs with a molecular weight of <0.5 kDa generally follow a mechanism similar to that of small molecule drugs when entering cells [25,41]. Initially, these SMDCs freely diffuse in and out of cells. However, they tend to accumulate in tumor microenvironments or within tumor cells that exhibit overexpression of specific biomarkers [42]. Once captured and concentrated, these SMDCs undergo cleavage under specific enzyme-catalyzed or pH conditions, thereby releasing cytotoxic payloads and triggering tumor cell apoptosis [43]. Notably, due to the low or absent expression of such receptors in normal cells or tissues, SMDCs do not accumulate there, thus mitigating toxicity to normal cells [38]. Research indicates that human aldo-keto reductase family 1 member C3 (AKR1C3) is a multifunctional enzyme primarily involved in the intracellular metabolism of steroid substances [44]. It is expressed at higher levels in various malignant solid tumors (such as breast cancer and leukemia) than in normal cells [25,45]. 4 (AST-3424, Fig. 3) is specifically designed to target AKR1C3-overexpressing solid tumors and is currently undergoing Phase Ⅰ/Ⅱ clinical trials [46]. Given its low molecular weight, 4 can efficiently diffuse into cells. Its small molecule targeting ligand exhibits specific selectivity for AKR1C3, making 4 recognizable only by AKR1C3 in tumor cells overexpressing it. Upon entry into tumor cells, the nitro group of 4 is rapidly reduced by AKR1C3/NADPH to hydroxylamine, an unstable intermediate [25]. Subsequently, cleavage by phosphatase quickly releases the cytotoxin DNA alkylating agent 5 (AST-2660). This cytotoxin interacts with the double strands of DNA in tumor cells, causing irreparable damage and leading to cell death. Importantly, in normal cells with low or absent AKR1C3 expressions, 4 remains inert and can easily exit the cells. Consequently, it does not accumulate in normal tissues and organs, thereby minimizing damage to healthy cells [42]. By exploiting the differential expression of AKR1C3 between tumor and normal cells, 4 demonstrates promising potential as a targeted therapy with reduced off-target toxicity. Furthermore, some smaller SMDCs, such as 1, can actively transport into tumor cells by targeting transport proteins like L-type amino acid transporter 1 (LAT1) [47]. This active transport mechanism leads to the generation of an activation cascade similar to that seen with 4, resulting in the release of toxic payloads that induce cytotoxic effects in tumor cells.

  Figure 3

  

  Figure 3.  The activation procedure of 4 (AST-3424) in vivo.

  DownLoad: Full-Size Img PowerPoint

  In another scenario, SMDCs with a molecular weight greater than 0.5 kDa typically utilize two distinct mechanisms for intracellular delivery, similar to those employed by ADCs [48]. Initially, the targeting ligand of these SMDCs specifically recognizes receptors that are overexpressed in the tumor microenvironment or on the surface of tumor cells. Subsequently, due to cleavage of the linker chain occurring either inside or outside the cell, two different pathways for entering target cells emerge [49,50]. The first pathway entails SMDCs entering specific intracellular compartments of target cells through receptor-mediated endocytosis (e.g., endosomes, lysosomes) [51]. Here, they undergo cleavage under specific enzyme-catalyzed or pH conditions, thereby releasing cytotoxic payloads that induce tumor cell apoptosis. An exemplary illustration of this mechanism is seen in 6 (Fig. 4A), currently undergoing Phase Ⅰ/Ⅱ clinical trials at Tarveda Therapeutics [49]. This SMDC targets late-stage solid tumors exhibiting high expression of heat shock protein 90 (HSP90) [52]. 6 consists of an HSP90 inhibitor fused to 7 (SN-38) via an ester linker. Accumulating massively around tumor cells overexpressing HSP90, 6 targets the cell surface-activated HSP90 [37]. Subsequently, it is delivered into intracellular endosomes or lysosomes where it undergoes cleavage under enzyme-catalyzed conditions, releasing the HSP90 inhibitor and 7. As the active metabolite of the topoisomerase Ⅰ (TOP1) inhibitor irinotecan, 7 causes DNA damage, inhibits DNA replication, and suppresses RNA synthesis [53,54]. The second pathway involves SMDCs being specifically cleaved by enzymes in the tumor microenvironment, thereby releasing highly cell-permeable cytotoxins into tumor cells and inducing tumor cell apoptosis [55]. 8 (Fig. 4B), developed by Vincerx Pharma and currently in Phase Ⅰ clinical trials for the treatment of late-stage solid tumors, exemplifies this approach [50]. 8 is equipped with a neutral endopeptidase (NE)-cleavable linker that connects an integrin alpha-v beta-3 (αvβ3)-targeting ligand to an optimized camptothecin (CPT) derivative (VIP-126, 9) payload. Upon diffusion into tumor tissues, 7 selectively targets tumor cells overexpressing αvβ3 integrin on their surface membranes [55]. Subsequently, it is specifically cleaved by overexpressed NE in the tumor microenvironment, liberating the CPT derivative 9. Which inhibits TOP1, causing DNA damage and ultimately leading to cell apoptosis [39].

  Figure 4

  

  Figure 4.  Mode of activation of SMDCs. (A) The activation procedure of 6 (PEN-866) in vivo. Targeting ligands facilitate homing to tumor cells with high expression of HSP90, where the linker is cleaved in the lysosomes or endosomes to release 7 (SN-38). (B) The activation procedure of 8 (VIP-236) in vivo. This compound specifically homes to the tumor microenvironment by binding to activated αvβ3 integrins and is effectively cleaved by NE to release 9 (VIP-126).

  DownLoad: Full-Size Img PowerPoint

  3. Criteria for selecting target biomarker

  In the development of SMDCs, the selection of appropriate target biomarkers is pivotal for ensuring the efficacy and safety of these therapeutics [56]. The approval of a drug hinges on two critical factors: its efficacy in comparison to existing approved medications and the ability to manage its toxicity or side effects within a safe range [57]. In the context of SMDCs, efficacy is largely determined by the activity level of the therapeutic payload and the distribution and abundance of biomarkers that are specifically overexpressed in the target tissue [23]. Safety, on the other hand, hinges on the conjugates' specificity to cancer cells [25]. SMDCs stand out for their remarkable precision in targeting tumor tissues, a departure from traditional chemotherapy that significantly reduces harm to normal cells [58]. The concentration of the payload within tumor cells is directly influenced by the distribution/quantity of biomarkers [59]. Hence, when designing SMDCs, careful attention to the distribution and positioning of biomarkers with specificity is crucial. The selection of an appropriate target biomarker markedly enhances the safety of SMDCs by directly influencing the payload concentration within tumor cells [24]. Consequently, the selection of an appropriate target biomarker is a pivotal step in maximizing the therapeutic window and minimizing off-target effects in SMDCs. The process of identifying the optimal biomarker involves careful consideration of several key factors.

  3.1 The expression level of biomarkers

  A key criterion in selecting target biomarkers for SMDCs is the differential expression of these markers in tumor cells compared to normal cells [60]. To optimize drug delivery to tumor cells, it is essential to choose biomarkers that are significantly overexpressed in the tumor microenvironment or within tumor cells while maintaining low expression levels in normal tissues [61]. Typically, biomarkers are expressed at more than three times the level in tumor microenvironments or tumor cells compared to normal cells [62]. This ensures that the concentration of drugs in cancerous tissues is sufficient to induce tumor regression while minimizing unnecessary damage to normal cells, thereby reducing off-target effects and enhancing therapeutic efficacy. Commonly enriched biomarkers in tumors, such as folate receptor alpha (FRα), AKR1C3, and prostate-specific membrane antigen (PSMA), are often expressed at levels far exceeding threefold [46,48,63]. For instance, FRα, a folate-binding protein located on cell membranes, has limited expression in healthy tissues but is significantly overexpressed in various solid tumors, including mesothelioma and epithelial ovarian cancer, with over 70% overexpression rates [64,65]. It is also prevalent in non-small cell lung cancer, triple-negative breast cancer, and endometrial cancer [66-68]. FRα influences cancer cell division and migration [69]. Leveraging these properties, researchers have developed tumor imaging agents targeting FRα, such as 10 (EC17, Fig. 5), which assist surgeons in more accurately removing lesions during ovarian, breast, and lung cancer surgeries [70]. Additionally, AKR1C3 has limited expression in healthy human tissues but is commonly overexpressed in solid tumors [71]. A study of 2490 cancer patients found the highest frequency of AKR1C3 overexpression in liver cancer, with over 50% frequency in bladder cancer, kidney cancer, and gastric cancer [46]. The Sasano team investigated AKR1C3 expression in breast cancer patients and found that the median level of AKR1C3 mRNA increased 18-fold compared to healthy individuals [72]. Therefore, using these biomarkers with specific overexpression in cancer cells can maintain effective doses in cancerous tissues while reducing drug doses, thereby enhancing drug safety, widening the therapeutic window of SMDCs, and potentially making them viable targets for various diseases.

  Figure 5

  

  Figure 5.  Structure of 10 (EC17), a fluorescent imaging agent targeting FRα. Intraoperative fluorescence imaging with 10 allowed for clear detection of ovarian cancer lesions in patients, with an average tumor-to-background ratio (TBR) of 7.0 ± 1.2. The fluorescence imaging was successfully maintained for approximately 5.5 h following the administration of 10.

  DownLoad: Full-Size Img PowerPoint

  3.2 The temporal and spatial distribution of biomarkers

  The distribution of SMDCs in vivo typically corresponds to the distribution of their target biomarkers, making the temporal and spatial distribution characteristics of these biomarkers crucial in determining the risk of off-target effects (Fig. 6) [73]. On one hand, ideal target biomarkers should peak in expression during tumor cell-specific overexpression, aligning with the optimal efficacy of cytotoxic payloads [74]. The cell division cycle of tumor cells encompasses pre-DNA synthesis (G1 phase), DNA synthesis (S phase), post-DNA synthesis (G2 phase), and mitosis (M phase) [75]. Drugs are differentiated based on tumor cell drug sensitivity into cell cycle-specific and cell cycle-nonspecific categories [76]. Cell cycle-specific drugs, such as antimetabolites active during the S phase and topoisomerase inhibitors, bleomycin affecting the G2 phase, and microtubule inhibitors targeting the M phase, demonstrate anticancer activity exclusively in these specific phases and are ineffective against G0 phase cells [77-79]. Their activity is time-dependent and does not increase beyond a certain dosage [80]. Therefore, when selecting biomarkers for targeting, it is essential to identify the phase with maximal expression and pair these biomarkers with corresponding cell cycle-specific drugs to form SMDCs, maximizing therapeutic outcomes. In contrast, cell cycle-nonspecific drugs like alkylating agents and antitumor antibiotics can eradicate tumor cells during all proliferation stages and the G0 phase [81,82]. These drugs are more potent against malignant tumor cells, and their efficacy escalates with dosage [83]. Consequently, the selection of biomarkers for these drugs does not necessitate considering the phase of highest expression.

  Figure 6

  

  Figure 6.  Temporal distribution of various types of SMDC payloads in exerting their anti-tumor effects throughout the cell division cycle (created with BioRender.com).

  DownLoad: Full-Size Img PowerPoint

  On the other hand, ideal target biomarkers should be expressed on the surface or within cancer cells and exhibit a positive correlation with tumor growth to avoid off-target binding of SMDCs, which can compromise their specificity and safety [84,85]. HSP90 is a highly conserved molecular chaperone protein widely expressed in eukaryotic cells, responsible for the maturation of over 300 client protein substrates [86]. Although HSP90 is expressed in most cells, it's often overexpressed and activated in many tumor cells, leading to its exposure on the cell surface, making it an ideal target for tumor-specific drug delivery [37]. For example, 6 targeting HSP90 shows significant accumulation in tumor tissues, with a retention time 20 times longer than in normal tissues. Subsequently, 6 releases the active metabolite 7, inducing DNA damage and apoptosis in tumor cells, thereby exerting broad-spectrum antitumor activity.

  3.3 Binding pocket properties of biomarkers

  In contrast to ADCs, which typically target receptors with broad and flat binding surfaces, SMDCs require biomarkers with well-defined small molecule binding pockets to ensure precise binding [87]. Achieving high affinity and specificity necessitates careful consideration of the non-covalent interactions that occur within the binding pocket of biomarkers, such as ionic bonds, hydrogen bonds, and hydrophobic interactions. These interactions are critical for ensuring effective targeting by SMDCs and subsequent therapeutic effects. Studies have shown that LAT1 is highly expressed at the blood-brain barrier (BBB) and placental barrier, efficiently transporting eight essential amino acids and amino acid derivatives (such as levodopa) into the brain in a sodium-independent manner [88]. Moreover, LAT1 is overexpressed in various tumor cells, including glioblastoma multiforme (GBM) and multiple myeloma, while its expression in normal tissues is very low. Therefore, LAT1 serves as a biomarker for SMDCs targeting intracranial tumors. 3, developed by Quadriga BioSciences, is an SMDC targeting LAT1 currently in Phase Ⅱ clinical trials. It efficiently traverses the BBB via LAT1, selectively entering LAT1-overexpressing tumor cells while sparing normal tissues that typically do not express LAT1. Inside tumor cells, 3 releases nitrogen mustard to interfere with rapid tumor cell division, exerting its cytotoxic activity and killing the tumor. As illustrated in Fig. 7, 3 forms a dense hydrogen bond network with surrounding amino acids (Ile63, Ser66, Gly67, Phe252) primarily through carboxyl and amino groups, stabilizing its binding within the pocket. Furthermore, while most SMDCs preferentially target the receptor's natural ligand binding site, designing ligands to target allosteric or non-functional sites may be advantageous in certain cases [89]. Allosteric/non-functional sites are less frequently occupied by endogenous ligands compared to primary ligand binding sites, reducing competition with natural ligands. For instance, the FRα, a biomarker overexpressed in tumor cells, has a natural substrate concentration ranging from 5 nmol/L to 50 nmol/L, making it challenging to design folate analogs to target this binding pocket [90].

  Figure 7

  

  Figure 7.  Crystal structure of LAT1 in complex with 3 (PDB code: 7DLS). LAT1 is depicted in a white cartoon representation, while 3 is shown as orange sticks. The amino acid residues interacting with 3 are displayed as cyan sticks. The yellow dashed lines represent hydrogen bonds between 3 and the residues Ile63, Ser66, Gly67, and Phe252. Additionally, green spheres indicate the centroid of the phenyl ring in the side chain of Phe252, which engages in a π-cation interaction with the positively charged amino group of 3, as illustrated by green dashed lines. Molecular graphics figures were prepared using PyMOL (Schrödinger, LLC).

  DownLoad: Full-Size Img PowerPoint

  Additionally, when targeting biomarkers on the surface of tumor cells, it is crucial to assess their rates of internalization and recycling [64]. Ideally, receptors should undergo rapid synthesis or recycling after degradation. Examples of receptors meeting these criteria include FRα, PSMA, and epidermal growth factor receptor (EGFR) [91-93]. Among them, the FRα receptor is the most extensively studied example—SMDCs targeting the FRα receptor observed receptor recycling approximately every 20 h in M109 tumors dosed every 40 h, while in L1210A tumors dosed every 13 h, receptor recycling occurred approximately every 6 h [94].

  4. SMDCs design strategy

  The design strategy for SMDCs focuses on delivering cytotoxic payloads into tumor cells by targeting overexpressed biomarkers specific to the tumor tissue [95]. This approach aims to achieve the desired therapeutic effect while minimizing off-target effects. Therefore, the development revolves around optimizing each structural component: small molecule ligands that specifically target overexpressed biomarkers in tumor cells, chemical linkers, and cytotoxic payloads [96]. In the following sections, we will discuss the design principles of each component, guided by SMDCs that have either been commercialized or are in clinical trials.

  4.1 Targeting ligands

  Once the biomarker for a SMDC is identified, the next critical step is to design ligands that can bind specifically and selectively to these biomarkers, while ensuring stability and effective targeted delivery within the circulatory system [97]. This process is complex and requires a balance of multiple factors, including high target binding affinity, selectivity, and the overall stability of the SMDC in biological environments [38,98,99]. Consequently, many of these ligands are derived from or inspired by natural ligands, leveraging their inherent affinity for the target biomarker. Among the various ligands studied, folate derivatives targeting the FRα have been the most extensively explored due to the receptor's overexpression in certain cancers [100]. Despite the initial promise of folate-targeted SMDCs, these candidates have faced significant challenges in clinical trials, often resulting in stagnation or termination of development programs [101]. This has highlighted the need for a paradigm shift in SMDC design, particularly in the search for non-natural ligands that might offer improved efficacy and stability [25]. The exploration of non-natural ligands—those not typically found in nature but designed to mimic or improve upon natural ligand structures—presents an exciting frontier in SMDC development [102]. These ligands could potentially overcome the limitations of natural ligand derivatives by offering greater stability, higher binding specificity, and improved pharmacokinetic profiles. To address these challenges and provide insights into effective targeting ligand design, the following sections will present detailed examples and case studies. These will illustrate the strategies employed to enhance ligand-receptor interactions, optimize ligand stability, and ultimately improve the therapeutic potential of SMDCs.

  4.1.1   High specificity and affinity of targeting ligands

  The design of targeting ligands is a pivotal distinction between SMDC and ADC therapeutics [68]. Most targeting ligands adhere to Lipinski's rule of five, particularly with molecular weights under 0.5 kDa, significantly lower than the ~150 kDa typical of ADCs [46]. This substantial difference complicates the design of ligands for specificity and affinity. Previous studies suggest that a KD of <10 nmol/L is necessary for meaningful therapeutic effects in SMDCs [38,90]. However, observations from clinical trials of SMDCs indicate that this threshold is not absolute [35]. For instance, 1 targets the LAT1 transporter for treating multiple myeloma and ovarian cancer through its L-phenylalanine moiety (molecular weight = 164 Da) [103]. Despite this, the transport rate of 1 is substantially lower compared to LAT1′s natural amino acid substrates [87]. Moreover, due to its poor selectivity, 1 also targets the L-type amino acid transporter 2 (LAT2), which is broadly expressed in normal cells such as the small intestine, ovaries, and skeletal muscle, leading to side effects like diarrhea and gastrointestinal bleeding, with severe cases potentially resulting in death. To enhance the specificity and affinity of such SMDCs, Quadriga BioSciences has advanced 3 into clinical trials [98]. 3 retains the nitrogen mustard moiety while optimizing the targeting of LAT1 to enhance affinity and reduce off-target effects on LAT2. In vitro transport assays reveal that 3 is 50 times more selective for LAT1 over LAT2 [88]. Additionally, compared to 1, 3 exhibits an eight-fold increase in affinity for LAT1, enabling it to preferentially enter tumor cells overexpressing LAT1 through the BBB [35]. GBM, for example, overexpresses LAT1 by more than 40 times compared to normal cells, and this expression increases with tumor grade and stage, making 3 a promising candidate for treating TMZ-resistant GBM [88].

  4.1.2   High stability of targeting ligands

  The natural ligands of most biomarkers are inherently prone to enzymatic degradation when present in blood or tissue, posing a significant challenge to the stability and efficacy of therapeutic agents [99]. To address this issue, particularly in the context of SMDCs, the use of synthetic ligands that mimic natural substrates has emerged as a promising strategy [102]. This approach not only enhances the stability of these conjugates during transport but also potentially improves their therapeutic index [87]. In contrast to the larger ligands used in ADCs, SMDC-targeting ligands are designed with significantly lower molecular weights, typically below 5 kDa [19,25]. This lower molecular weight confers several advantages, including reduced susceptibility to in vivo degradation and the ability to be rapidly cleared from the circulation through renal filtration, ultimately resulting in a shorter half-life (t_1/2) in the bloodstream [104]. This rapid clearance is particularly advantageous in minimizing off-target effects, thereby enhancing the safety profile of SMDCs. Among the various molecular targets being explored for SMDCs, integrins stand out as particularly promising [105]. Integrins are critical members of the cell adhesion molecule family, functioning as transmembrane heterodimeric glycoproteins composed of non-covalently associated α and β polypeptide chains [106]. To date, over 24 distinct integrins have been identified, each playing a unique role in cellular adhesion, signaling, and migration. Among these, the αvβ3 integrin has garnered significant attention due to its high expression levels in tumor cells and angiogenic endothelial cells, making it a compelling target for cancer therapy. The arginine-glycine-aspartic acid 11 (RGD) sequence, a natural ligand for the αvβ3 integrin receptor, was first identified by E. Ruoslahti in the early 1970s [107]. This discovery has since spurred extensive research aimed at enhancing the stability and efficacy of 11-based therapeutic agents [108,109]. One of the key challenges with 11 is their inherent instability in vivo, with a reported t_1/2 of just 13 min, which severely limits their therapeutic utility [110]. To overcome this limitation, researchers have developed cyclic RGD peptides, which exhibit increased rigidity due to their cyclic structure [111,112]. This structural modification significantly enhances their stability and prolongs their t_1/2, making them more effective as therapeutic agents. One notable example of this approach is the work of Kessler and colleagues, who developed a cyclic RGD pentapeptide, 12 (c(RGDfK)), incorporating a non-natural D-configuration amino acid (D-Phe) (Fig. 8A) [113]. The cyclic structure, combined with the incorporation of D-amino acids, imparts high stability against enzymatic degradation in vivo, thereby improving the therapeutic potential of this peptide [114]. For researchers focused on the development of SMDCs, peptidomimetics represent a superior alternative to natural peptides. Peptidomimetics are small, protein-like chains designed to mimic the structure and function of peptides while offering enhanced stability, increased receptor affinity, and improved pharmacokinetic properties [115]. For instance, 13 (Fig. 8B) is a SMDC that consists of a peptidomimetic αvβ3 integrin binder linked to the cytotoxic agent monofunctional monomethyl auristatin E (MMAE) via a protease-cleavable GKGEVA linker [116]. This design provides distinct advantages over the use of 11, including increased physiological t_1/2, enhanced oral bioavailability, and the elimination of peptide cleavage sites, which further contribute to the stability of the conjugate. The increased stability of 13 ensures that it can be effectively transported through the bloodstream, delivering its cytotoxic payload specifically to tumor cells, thereby maximizing its therapeutic efficacy while minimizing off-target toxicity. This exemplifies the potential of peptidomimetics in the design of next-generation SMDCs, paving the way for more effective and targeted cancer therapies.

  Figure 8

  

  Figure 8.  Molecules targeting the αvβ3 integrin receptor. (A) The transition from linear 11 (RGD) to 12 (c(RGDfK)) resulted in an approximate10-fold increase in t_1/2. (B) 13 mimicking the RGD binding mode circumvents peptide instability and utilizes the αvβ3 receptor to exert cytotoxic effects on tumor cells. It exhibits no cytotoxic effect on cells lacking the αvβ3 receptor.

  DownLoad: Full-Size Img PowerPoint

  4.1.3   Conjugatable groups on targeting ligands

  As a crucial component of SMDCs, targeting ligands must possess derivatizable functional groups that enable efficient conjugation with the payload and linker [117]. The selection of these conjugatable groups is critical, as they not only need to ensure fast and stable coupling with functional groups on the linker but also must preserve the ligand's ability to bind to its receptor. Structural information from the crystal complex of the ligand-receptor pair can play an important role in this process. By directly examining the binding mode of the ligand to the receptor, it becomes possible to identify functional groups that do not affect affinity and to design appropriate conjugation reactions to anchor the ligand onto the linker.

  In the absence of such structural data, one can refer to structure-activity relationship (SAR) studies of the ligand [47]. By analyzing the impact of different binding sites on ligand affinity, it is possible to indirectly select suitable functional groups for conjugation. Common functional groups on ligands that are used for conjugation include carboxyl, amino, hydroxyl, and thiol groups. With the advancement of click chemistry, additional options such as alkynes or azides are now available for use in conjugation reactions [118,119].

  4.2 Payloads

  As the key component directly acting on tumor tissues, the cytotoxicity of the payload and its concentration within the tumor tissue determine the therapeutic efficacy of SMDCs [120]. If the payload fails to effectively kill tumor cells, the designed SMDCs will not achieve their intended therapeutic effect. Therefore, selecting a payload with high cytotoxicity that is well-suited for SMDC design is crucial. Traditional payloads typically include alkylating agents, tubulin inhibitors, topoisomerase inhibitors, and antibiotics, while newer payloads have expanded to include RNA inhibitors, B-cell lymphoma-extra large (Bcl-X_L) inhibitors, Nicotinamide phosphoribosyltransferase (NAMPT) inhibitors, and low-toxicity small molecules such as CPT [42,55,120,121]. These novel payloads provide a broader range of options for SMDC design, laying a solid foundation for enhancing therapeutic efficacy and safety. Generally, the payload must meet several critical requirements.

  4.2.1   High cytotoxicity of payload

  The payload must possess extremely high cytotoxicity. Compared to the micromolar-level toxicity of traditional chemotherapeutic agents, the toxicity of next-generation payloads can reach sub-nanomolar or even picomolar levels [120]. This implies that only a small amount of the SMDC is required to achieve a tumor-killing effect comparable to, or even greater than, traditional chemotherapy drugs, often by orders of magnitude [49]. Furthermore, since SMDCs can precisely deliver the payload to tumor tissues, the side effects on normal cells are significantly lower than those of conventional chemotherapeutics [122]. Initially, traditional chemotherapeutic agents like methotrexate (Methotrexate), vinblastine, and DOX face the limitation of insufficient cytotoxicity against tumor cells, resulting in suboptimal therapeutic outcomes [120]. With the ongoing exploration of highly toxic payloads, a series of payloads with cytotoxicity 100–1000 times greater than traditional chemotherapeutic agents, such as tubulin inhibitors and RNA inhibitors, have emerged. For example, the newly developed payload MMAE exhibits cytotoxicity at the picomolar level, demonstrating tumor cell-killing effects far superior to traditional chemotherapeutics [123].

  MMAE has been extensively validated in various ADC therapies for its efficacy across different cancer types, making it a popular choice for developing SMDCs targeting various cancers [124-126]. Given that ligands often have a high affinity for their target receptors, the focus of SMDC design lies in selecting the appropriate payload and linker. The best payloads are those with high cytotoxicity, with MMAE standing out as a prime candidate [63]. For instance, Wang et al. leveraged this characteristic to design the 14 (PSMA-1-VcMMAE, Fig. 9A) prodrug for treating prostate cancer [127]. MMAE is a potent microtubule inhibitor that kills cells by inhibiting microtubule assembly and subsequently inducing mitotic arrest. Due to its extreme toxicity, the use of MMAE as a standalone therapeutic agent is limited. Consequently, MMAE is predominantly employed within targeted delivery systems, which enhance selectivity and minimize side effects. Olatunji and his team constructed a PSMA-targeting SMDC (15, Fig. 9B) using the Val-Cit linker and the target ligand CTT1298, which exhibited good therapeutic efficacy in vivo evaluations [63].

  Figure 9

  

  Figure 9.  Structure of SMDCs targeting PSMA. (A) The cytotoxicity of 14 (PSMA-1-VcMMAE) against PSMA-positive PC3pip cells is 48 times greater than that against PSMA-negative PC3flu cells. (B) The distribution of MMAE of 15 in the mouse brain is restricted, with tumor exposure being 8-fold higher than that in plasma.

  DownLoad: Full-Size Img PowerPoint

  4.2.2   High stability of payloads

  Given that SMDCs are typically administered via injection, it is crucial to ensure that the payload maintains its structural stability from the point of injection until it reaches the tumor tissue [128]. Due to the high cytotoxicity of these payloads, premature release during delivery could cause severe damage to normal cells, leading to significant side effects. Therefore, the payload must remain stable until it reaches the tumor tissue and avoids premature degradation within the body, ensuring sufficient drug concentration at the target site for effective tumor cell killing.

  9, a novel derivative of 7, was optimized to obtain a 20-fold cell membrane permeability and a 36-fold reduction in efflux potential [39]. Unlike 7, 9 retains its cytotoxic activity in cells transfected with P-gp or BCRP transport proteins, indicating that it is not a substrate for these transporters. This suggests that 9 may exhibit better activity profiles in tumor cells. As the payload of 8, 9 effectively kills various cancer cells in both in vitro and in vivo experiments, demonstrating potent tumor-targeting efficacy, and is now in clinical trials [129].

  4.2.3   Conjugatable groups on payloads

  The payload must possess functional groups or sites that can chemically bond with the linker or targeting ligand [43]. Through chemical reactions, the payload can be conjugated to the targeting ligand or linker, thereby imparting selectivity to the SMDCs. Therefore, the chemical structure of the payload should include groups or sites, such as amines, thiols, carboxyls, or aldehydes, that can interact with these molecules, ensuring the efficient construction and precise delivery of SMDCs [130]. Commonly used conjugation sites include azides (which enhance compound potency), carboxyl groups, Tup residues, and isobutyl groups, which not only facilitate linkage but also enhance the efficacy of the payload. Additionally, once the payload is conjugated with the remaining parts to form a complete SMDC, the payload's activity is masked, further improving the safety of the SMDC.

  For instance, the poor solubility of 7 hampers its clinical application. To address this, Soyoung used a hydrophilic oligomer ethylene glycol (OEG) linker to connect the targeting motif and 7 through a cleavable ester bond (16, Fig. 10A), significantly improving pharmacokinetic properties and enhancing permeability [63].

  Figure 10

  

  Figure 10.  (A) Conjugatable groups on payloads. The 7 (SN-38) hydroxyl group and PSMA-targeted small molecule are conjugated using OEG to form 16. (B) Intracellular action sites of the payloads. 17 (CLB) demonstrates poor cell membrane permeability, with a topological polar surface area (TPSA) of 160 Ų. The condensation of the carboxyl group of CLB with the amino group of 18 (ZW800) inactivates the carboxyl group, thereby improving cell membrane permeability.

  DownLoad: Full-Size Img PowerPoint

  MMAE's extreme cytotoxicity is effectively harnessed by conjugating it via an amine group to form 14 [63]. The uncoupled drug shows no significant cytotoxicity, indicating its strong dependence on protease activation. For anthracyclines like DOX or daunorubicin (DAU), various linker systems such as ester, hydrazone, oxime, and amide bonds have been employed to conjugate with gonadotropin releasing hormone (GnRH)-Ⅲ carriers [131]. These SMDCs exhibit significant in vitro cytotoxicity and the highest levels of chemical and enzymatic stability. The antiproliferative activity tests of these conjugated drugs showed that cleavable GnRH-Ⅲ bioconjugates inhibit the growth of A2780 ovarian cancer cells, which highly express GnRH receptors while showing reduced activity against Panc-1 pancreatic cancer cells with lower GnRH receptor levels.

  4.2.4   Intracellular action sites of the payloads

  If the payload's site of action is extracellular, such as blocking ion channels or disrupting coagulation, the payload will act prematurely before the SMDC reaches the tumor tissue, resulting in loss of tumor specificity and significant side effects [49]. Therefore, the payload's site of action should be intracellular within tumor cells. For example, 17 (chlorambucil, CLB), a well-known anticancer drug, has limited use due to its lack of specificity for targeted cancer cells [132]. 17 degrades rapidly in plasma, has low bioavailability, and has poor pharmacokinetics, which may lead to unpredictable side effects [133]. Therefore, SMDCs using 17 as a payload, such as 18 (CLB-ZW800, Fig. 10B), must enhance water solubility and bioavailability to ensure that the payload is released after being taken up by the target tumor cells, thereby providing therapeutic effects while minimizing toxic reactions [134].

  4.3 Linkers

  Unlike ADCs, which require linkers to modulate pharmacokinetic properties and control the release of the payload due to the large size of antibodies, SMDCs can often forgo the use of a linker [135]. Because of their smaller size, SMDCs can directly couple targeting ligands to the payload without the need for additional space or the prevention of steric hindrance, making linkers a non-essential component of SMDC design [42]. For example, current SMDCs with nitrogen mustard-based payloads do not use linkers. A notable example is 4, as mentioned earlier [25]. From a chemical perspective, 4 is designed by directly conjugating a DNA-alkylating agent with a specific targeting ligand for AKR1C3. The ligand masks the toxicity of the payload, transforming 4 into a stable, non-toxic prodrug. Upon entering AKR1C3-overexpressing tumor cells, the ether bond linking the payload to the ligand is cleaved, releasing the active payload. Through this clever structural design, 4 links the ligand and payload directly, without introducing a linker. This simplifies the overall molecular structure, reduces the molecular weight, and enhances both tissue penetration and intracellular delivery efficiency.

  For SMDCs with non-nitrogen mustard payloads, linkers play a critical role in connecting the payload to the targeting ligand. The design of these linkers directly affects the drug's efficacy and safety. In the following section, we will explore several key aspects of linker design.

  4.3.1   Balancing hydrophilicity and lipophilicity

  Typically, both the payload and targeting ligands in SMDCs are hydrophobic, necessitating the incorporation of hydrophilic linkers [95]. Effective SMDCs are designed to release their payload at predictable sites and rates after permeating the target cells. One primary function of the linker is to increase hydrophilicity [48]. While hydrophobicity in the targeting ligand and payload maximizes membrane permeability and receptor affinity, it can also lead to off-target interactions with non-specific proteins and membranes, thereby reducing metabolic stability [136]. For example, when the transthyretin (TTR) ligand is conjugated with the hydrophobic cytotoxin MMAE, the overall hydrophobicity of the conjugate increases significantly, diminishing the drug's developability [137]. To counterbalance this, a small hydrophilic fragment was incorporated into the linker in 19 (Fig. 11A), successfully mitigating the hydrophobic impact of MMAE. This adjustment conferred overall hydrophilicity to the SMDC, enabling passive diffusion into healthy cells while extending the drug's circulation t_1/2.

  Figure 11

  

  Figure 11.  The use of different linkers in SMDCs design can balance hydrophilicity and hydrophobicity. (A) The targeting ligands PMSA and TTR are linked to the payload MMAE via a linker to form 19. The incorporation of a PEG segment into the linker imparts overall hydrophilicity to the cytotoxic conjugate, thereby restricting its passive diffusion into healthy cells. (B) The folate derivative is conjugated with vincristine through a linker to generate 20 (Vintafolide). Notably, the introduction of an Asp-Arg segment into the linker significantly enhances the overall solubility of the compound.

  DownLoad: Full-Size Img PowerPoint

  Most small molecule compounds used in cancer therapy, such as folic acid (FA), exhibit high lipophilicity, allowing them to passively diffuse across bilayer cell membranes [138]. In the design of FA-based SMDCs, such as 20 (Vintafolide, Fig. 11B), researchers introduced aspartic acid (Asp) and arginine (Arg) residues into the linker peptide to enhance the overall hydrophilicity of the SMDC. These hydrophilic SMDCs then enter target cells through FRα-mediated endocytosis [139].

  4.3.2   Conformational optimization

  A longer linker can spatially separate the payload from the targeting ligand, enhancing the binding ability of the targeting ligand by striking an appropriate balance between flexibility and rigidity, thereby avoiding potential steric hindrance [140]. Some linkers in SMDCs function by spatially separating the ligand and payload to minimize steric interference that could impair ligand-receptor affinity, while also preventing unnecessary intramolecular interactions [141]. When the payload is too close to the targeting ligand, it can directly interfere with receptor binding. This issue can often be resolved by inserting a longer linker between the ligand and the payload. For instance, 21 (BPRDP056, Fig. 12A), composed of the Zn-DPA ligand, the payload SN38, and an inactive linker, demonstrates excellent draggability and anticancer activity [142]. Previous studies have shown that the linker's presence and payload modification do not affect Zn-DPA's in vitro target binding capability while retaining the antitumor activity of SN38. Remarkably, 21 even outperformed the free drug in inhibiting colorectal and pancreatic cancers in mouse models. Similarly, BET inhibitors, which are potential ligands for SMDCs, bind with high affinity to BET proteins, but their toxicity limits clinical application [143]. To overcome this, a research team installed linker side chains of varying lengths at the amide bond portion of I-BET762 and tested the impact of these linkers on BET protein binding. When using propyl as the linking chain (22), the activity of the compound decreased by sixfold and threefold compared to two different lengths of polyethylene glycol (PEG) chains (23, 24). This decrease in activity is primarily attributed to the entropy penalty caused by the introduction of flexible side chains (Fig. 12B). The design of such compounds establishes a foundation for a drug delivery system based on BET inhibitors. In the design of 25 (Zn8_DM1, Fig. 12C), the introduction of PEG in the linker created steric hindrance that impaired the binding ability of the targeting ligand [144]. Consequently, a small PEG unit was used to bridge the targeting ligand and the payload, optimizing the molecular design by leveraging steric hindrance to not only separate the targeting ligand and payload but also provide conformational obstacles that shielded Zn8_DM1’s in vitro cytotoxicity.

  Figure 12

  

  Figure 12.  Conformational optimization can be optimized by using different linkers in SMDCs design. (A) PS-targeting ligands Zn-DPA and SN38 (TOP1 inhibitor) are connected by a longer linker to form 21 (BPRDP056). This elongated linker does not inhibit the PS-specific tumor-targeting capability of Zn-DPA. (B) I-BET762, which targets BRD4, is linked to a fluorescent probe via different linkers. The linking chain consists of propyl (22), while the other linking chains utilize PEG of varying lengths (23, 24). It is important to note that longer linkers can lead to reduced targeting efficacy. (C) PS-targeting ligands Zn-DPA and DM1 (maytansine-analog) are conjugated through a suitably sized linker to create 25 (Zn8_DM1). The significant polymerization of the PEG moiety can substantially enhance the solubility of the conjugate; however, this may also introduce steric hindrance that can compromise the binding capability of the targeting ligands.

  DownLoad: Full-Size Img PowerPoint

  4.3.3   Extending t_1/2 and enhancing drug stability

  To achieve synergy between targeting and payload release, the conjugate must remain stable during transport from the bloodstream to the target tumor cells [43]. If the SMDC releases the cytotoxic payload outside of the tumor tissue, it can cause significant damage to normal cells. Therefore, plasma stability and in vivo, circulation stability of the conjugate are critical parameters in SMDC design. 6 uses a linker composed of a degradable peptide cleavable by NE [55]. The stability of this linker is ensured by its length, the carboxyl side chain of aspartic acid, and the steric hindrance of the valine residue, which collectively stabilizes the ester bond connecting the payload. As a result, 6 exhibits high water solubility, stability, and efficient cleavability, minimizing payload release outside tumor cells and significantly enhancing in vivo stability. CPT is a classic antitumor drug, but its non-selective high cytotoxicity poses a challenge. Researchers have attached a glutathione (GSH)-responsive conjugate TC6 to CPT to form the prodrug 26 (TC6-CPT, Fig. 13A) [145]. This conjugate demonstrates good stability under various pH conditions, exhibiting low cytotoxicity toward human gastric epithelial cells (GES-1) while maintaining high cytotoxicity against gastric tumor cells. This strategy undoubtedly enhances CPT's stability during the targeting process, significantly reducing off-target toxicity. 21 is another excellent example [142]. In pharmacokinetic studies on rodent models, the area under the plasma concentration-time curve (AUC) results and plasma metabolite concentrations indicated that the conjugate's stability was significantly improved compared to the free payload. The results suggest that 21 can slow SN38 elimination and extend its circulation t_1/2.

  Figure 13

  

  Figure 13.  (A) 26 (TC6-CPT) consists of the GSH reactive conjugate TC6 and the DNA topoisomerase CPT. 26 had a t_1/2 of 60 min at 2 mmol/L GSH and maintained high stability at a variety of pH conditions commonly found in vivo. (B) 27 (Fc1070) is constructed by linking PSMA-targeted small molecules to Monomethylauristatin F through an FC-containing linker, resulting in the formation of 28 (SMDC1070). 27 had a 3-fold higher affinity for PSMA than 28. (C) 29 ([⁶⁸Ga] Ga-PSMA-093) was developed by the incorporation of a specific linker O-(carboxymethyl)-L-tyrosine between the Lys-u-Glu and HBED-CC.

  DownLoad: Full-Size Img PowerPoint

  4.3.4   Enhancing affinity of the targeting ligand

  Due to its small molecular weight, SMDC can penetrate solid tumor tissues, but this also results in a very short t_1/2, limiting its therapeutic effect. To address this, some research teams have proposed conjugating an Fc fragment to the linker [146]. This novel SMDC design not only exhibits a longer t_1/2 but also shows superior specificity, affinity, and efficacy against tumor cells. A typical example is 27 (Fc1070, Fig. 13B), which consistently displays higher binding capacity in PSMA-positive cell lines compared to its counterpart 28 (SMDC1070, Fig. 13B), opening a new avenue for SMDC development. Linkers that enhance ligand affinity have also been applied in the design of PSMA-targeted agents. Given the presence of an auxiliary hydrophobic pocket in PSMA, hydrophobic aromatic structures can be employed to enhance PSMA binding affinity. For example, 29 ([⁶⁸Ga] Ga-PSMA-093, Fig. 13C) incorporates tyrosine between the Lys-u-Glu and HBED-CC to provide better lipophilicity, thereby achieving higher PSMA binding affinity [147].

  4.3.5   Controlled release mechanisms: Three types of cleavable linkers

  The activity of SMDCs depends on the cleavable linker's ability to release the payload at a controlled rate and specific site after permeating the target cells. Thus, linkers should enable controlled release triggered by the tumor cell's internal environment or protein-specific interactions to achieve optimal targeting. Cleavable linkers can generally be categorized into three types based on their cleavage mechanisms (Fig. 14).

  Figure 14

  

  Figure 14.  Three responsive cleavage mechanisms of SMDCs linkers: enzymatic cleavage, acidic condition cleavage, and reductive cleavage (created with BioRender.com).

  DownLoad: Full-Size Img PowerPoint

  Given that SMDCs, like ADCs, are primarily used in cancer treatment, enzyme selection for cleavage typically focuses on cathepsins overexpressed in tumors, particularly cathepsin B, which is abundant in the lysosomes of mammalian cells. A representative linker cleaved by cathepsins is the Val-Ala site. For instance, 13 with MMAE, as the payload uses GKGEVA, containing Val-Ala, as the linker, and the resulting conjugate, demonstrates high selectivity and activity, validating the value of this enzymatic cleavage site [116]. Legumain (LGMN), a member of the cysteine protease family primarily localized in lysosomes, is overexpressed in solid tumors and is involved in tumor invasion and metastasis. Research teams have explored LGMN as a cleavage enzyme for linkers and identified a series of asparagine-containing cleavable peptide linkers (AsnAsn, AsnGln, and AsnAla) [148]. The AsnAsn-PABC-MMAE conjugate based on this linker (30, Fig. 15A) can be effectively hydrolyzed by LGMN and remains stable in plasma.

  Figure 15

  

  Figure 15.  Responsive cleavage of the different linker of SMDC. (A) After incubation in the lysosome for 45 min, the cleavage rate of 30 reached 29.3%. (B) The cleavage t_1/2 of 31 varied under different conditions, showing a decrease in t_1/2 with increasing acidity. (C) At a pH of 5.0, 32 rapidly cleave and release their payload, while at pH 7.4, no cleavage occurs within 2 h. (D) 33 (TBG) undergoes continuous cleavage in the presence of GSH. This data highlights the influence of environmental factors, such as pH and redox conditions, on the performance of SMDCs, emphasizing their potential for targeted therapeutic applications.

  DownLoad: Full-Size Img PowerPoint

  Compared to normal tissues, tumor tissues convert glucose to lactic acid at a higher rate, leading to a localized pH drop of 0.5–1.0 units. The acidic nature of the tumor microenvironment can trigger payload release, providing a targeting condition for acid-cleavable linkers. For example, a linker based on phosphoramidite ester cleaves under acidic conditions to release MMAE (31, Fig. 15B). This linker shows remarkable pH selectivity, being most labile under mildly acidic conditions while remaining relatively stable under physiological conditions (pH 7.4), highlighting its potential application [149]. Among commonly used acid-labile linkers, acyl hydrazones have shown good activity in preclinical studies. Researchers also experimented with a hydrazone-linked SMDC targeting FRα (32, Fig. 15C), which cleaves under acidic conditions to release the highly active N-(5,5-acetoxy-pentyl) DOX. Compared to the free drug, this SMDC exhibited significantly reduced off-target toxicity while maintaining therapeutic activity. GSH is the primary reducing agent in cellular biochemical processes, rich in thiols. In some tumor tissues, GSH concentrations are often elevated compared to normal tissues, making reduction-sensitive linkers containing disulfide bonds common in SMDC design. For example, 33 (TBG, Fig. 15D), a drug conjugate targeting the sodium-dependent multivitamin transporter (SMVT), utilizes a disulfide bond to link the potent anticancer agent gefitinib to mertansine [150,151]. This conjugate undergoes continuous cleavage in the presence of GSH.

  5. SMDCs in clinical progress

  The primary application of SMDCs lies in targeted cancer therapy. Currently, 9 SMDC-based drugs are in various stages of clinical trials or have already reached the market (Table 2). Among these, SMDCs for solid tumors make up a significant proportion of those that have entered clinical trials.

  Table 2

  

  Table 2.  Approved or currently active SMDCs in clinical trials.

  DownLoad: CSV

  Drug name	Status	Molecular weight	Targeting ligands	Linker cleavage mode	Payload	Conditions	Clinical trial
  Melphalan	Launched	305 Da	LAT1	No linker	Chlormethine	Multiple myeloma; ovarian carcinoma; veal melanoma with unresectable liver metastases	NCT06425276
  TH-302	Phase Ⅲ	449 Da	HIF1ANa	No linker	Br-IPM	Metastatic or locally advanced unresectable pancreatic adenocarcinoma	NCT01746979
  QBS72S	Phase Ⅱ	333 Da	LAT1	No linker	Chlormethine	Brain metastases; breast cancer	NCT05305365
  AST-3424	Phase Ⅰ/Ⅱ	460 Da	AKR1C3	No linker	AST-2660	Liver cancer; relapsed and refractory acute lymphoblastic leukemia	NCT04315324
  PEN-866	Phase Ⅰ/Ⅱ	880 Da	HSP90	Enzymatic cleavage	SN38	Solid tumor	NCT04890093
  CBP-1019	Phase Ⅰ/Ⅱ	Not disclosed	FRα & TRPV6	No linker	MMAE	Solid tumor	NCT06576037
  AST-001	Phase Ⅰ/Ⅱ	Not disclosed	AKR1C3	No linker	AST-2660	Solid tumor	NCT06245330
  VIP-236	Phase Ⅰ	1516 Da	αvβ3	Enzymatic cleavage	VIP-126	Solid tumor	NCT05712889
  T-1201	Phase Ⅰ	Not disclosed	PSb	Enzymatic Cleavage	SN-38	Solid tumor	NCT04866641
  a HIF1AN, hypoxia inducible factor 1 subunit alpha. b PS, phosphatidylserine.

  One prominent example of an SMDC using a nitrogen mustard payload is 1, approved for the treatment of multiple myeloma [31]. While 1 was not specifically developed as an SMDC, its design and mechanism of action share similarities with SMDCs. 1 links its alkylating agent payload, nitrogen mustard, to L-phenylalanine, utilizing passive lipid solubility to enter tumor cells, where it is enzymatically hydrolyzed to exert cytotoxic effects. 2 is designed to target hypoxic regions that are rare in normal tissues, also has a nitrogen mustard payload [34]. 2 releases bromo‑isophosphoramide (Br-IPM) under these hypoxic conditions. Notably, the payload is directly linked to a 2-nitroimidazole hypoxia-triggering moiety, allowing for efficient masking of 2′s cytotoxicity until activated. In hypoxic tumor environments, 2 selectively converts into its active form to exert therapeutic effects. 2 has reached Phase Ⅲ clinical trials, targeting metastatic or locally advanced unresectable pancreatic cancer. Additionally, it has completed 15 clinical trials, demonstrating the drug's broad therapeutic potential. A promising SMDC, 3, is in Phase Ⅱ clinical trials for the treatment of breast cancer and brain metastases [88]. 3 is a bifunctional SMDC targeting the LAT1 transporter, which facilitates its active transport across the blood-brain barrier, allowing it to selectively enter tumor cells with high LAT1 expression and release its nitrogen mustard payload for targeted tumor therapy. Another standout example is 4, which is developed by Accendatech in Shenzhen, China. 4 is a prodrug designed to target AKR1C3, which converts it into the cytotoxic nitrogen mustard compound 5 and is currently in Phase Ⅰ/Ⅱ clinical trials for the treatment of metastatic or unresectable solid malignancies, including pancreatic cancer [122]. 4 is also being investigated for non-solid tumors, such as acute lymphoblastic leukemia, which is in Phase Ⅱ trials (NCT04315324). Beyond these leading candidates, AST-001, which shares the same payload as 4, has entered Phase Ⅱ trials [40]. In addition, 2 showed significant efficacy in the treatment of PARP inhibitor resistant solid tumors by targeting breast cancer susceptibility gene 1/2 (BRCA1/2) mutant malignancies via Br-IPM [152].

  In addition to nitrogen mustard alkylating agents, there are also many highly cytotoxic agents commonly used as conjugate drugs. The most common non-nitrogen mustard loadings were MMAE, CPT, and their derivatives. 6 is an SMDC that incorporates the active metabolite of irinotecan called 7 [49]. 7 known for its poor solubility, low bioavailability, and dose-limiting toxicities, is commonly formulated as conjugate drugs (e.g., Trodelvy) to overcome these issues. Clinical data show that 6, at therapeutic doses, did not exhibit dose-limiting toxicities; only one patient experienced mild neutropenia, a marked safety advantage compared to Trodelvy's 61% incidence rate [153]. 6 is currently undergoing Phase Ⅰ/Ⅱ clinical trials for the treatment of small-cell lung cancer, colorectal cancer, and pancreatic cancer. In 2019, it was granted orphan drug designation by the FDA for pancreatic cancer treatment. CBP-1019 is an uncommon bispecific ligand-binding SMDC that targets both FRα and Transient Receptor Potential Vanilloid 6 (TRPV6), and it also delivers a TOP1 inhibitor (NCT05830097). Compared to its predecessor, CBP-1018 (peptide-drug conjugate (PDC)), which used MMAE as a payload, CBP-1019 utilizes a trifunctional linker design that significantly enhances its stability and binding activity [154]. Preclinical studies have demonstrated CBP-1019′s excellent efficacy and safety profile, with minimal harm to test animals (NCT06576037). In 2023, CBP-1019 received orphan drug designation in the U.S. for the treatment of pancreatic cancer. There are also other SMDCs with a non-nitrogen mustard payload that have entered Phase Ⅰ trials, 8 and T-1201, both of which use modified CPT as their payloads, are both performing well in the treatment of solid tumors [41].

  While the aforementioned SMDCs are actively being developed, many other SMDCs have failed during clinical development or have been abandoned. One notable example is 20, a FRα-targeting SMDC conjugated with vinblastine [121,155,156]. 20 showed promising activity in Phase Ⅰ and Ⅱ trials, but in Phase Ⅲ ovarian cancer trial (NCT01170650), no significant improvement in progression-free survival (PFS) was observed, leading to its discontinuation. In contrast, the newer SMDCs targeting the FRα, such as CBP-1019, incorporate bispecific targeting. For instance, CBP-1019 targets FRα and TRPV6, suggesting that a dual-targeting strategy may enhance the efficacy of FRα-targeted therapies.

  6. Outlook for SMDCs development

  Due to the significant toxicity, off-target effects, and the associated harm caused to patients by traditional chemotherapy drugs, the development of targeted therapies has become a key trend in improving current cancer treatments [157]. In comparison to ADCs, SMDCs also possess unique advantages. Composed of small molecule linkers, SMDCs have a lower molecular weight, making them more capable of penetrating and accumulating in solid tumor tissues, with minimal distribution in healthy tissues. Due to their larger molecular size, ADCs can only enter tumor cells via receptor-mediated endocytosis or fragmentation in tumor microenvironment [158]. In contrast, SMDCs can penetrate tumor cells through various mechanisms, including endocytosis and passive diffusion, allowing for a more diverse design of SMDCs [35,42,49,50]. Moreover, ADCs are prone to immunogenic reactions due to their conjugation with large antibody molecules, and their pharmacokinetic profiles are highly complex issues that SMDCs entirely circumvent [159]. Lastly, from an industrial production perspective, SMDCs are less costly and complex to produce, exhibit greater drug stability, and require simpler storage and transportation conditions, advantages that also surpass those of PDCs and antibody fragment-drug conjugates (FDCs) [12,160,161]. However, compared to macromolecular conjugates such as ADCs and FDCs, which can leverage flat protein surfaces to form strong protein-protein interactions, SMDCs may struggle with targets that lack small-molecule binding pockets, such as CD30 [162]. Moreover, optimizing targeting ligands for high affinity and selectivity poses a significant challenge for SMDCs.

  The design of SMDCs can be summarized in four key steps. First, appropriate biological markers must be selected based on their expression levels and distribution characteristics, ensuring that there is a druggable binding pocket available. In contrast to ADCs, SMDCs possess a distinct advantage in molecular weight, allowing them to target intracellular biomarkers and significantly expanding the range of potential targets [163]. This could serve as a crucial breakthrough for the future of SMDC development. Second, specific ligands that can selectively bind to the targeted biomarkers need to be chosen. These ligands should exhibit high affinity and selectivity, alongside a certain degree of stability, as well as a suitable conjugation site. Compared to larger ADCs, SMDCs can be rapidly excreted from the body, thereby minimizing the adverse effects associated with the release of toxins during prolonged retention. The third step involves the selection of an effective payload that is highly cytotoxic, stable, has a conjugation site, and exerts its action intracellularly. The principles guiding the selection of payloads for SMDCs are fundamentally similar to those for ADCs. Lastly, the design of linkers must be considered. Unlike ADCs, linkers are not a mandatory component of SMDCs. However, an appropriate linker can enhance the balance between hydrophilicity and lipophilicity, optimize conformation, improve drug stability, elevate ligand affinity, and facilitate controlled release of the payload. Therefore, linker design is often given significant attention in the development of SMDCs.

  Currently, there is a wide range of molecular weights for existing SMDCs, with some being quite large and others relatively small [49,122]. High molecular weight SMDCs frequently utilize well-established targeting ligands, linkers, and cytotoxic payloads, such as those targeting folate or HSP90, representing a straightforward design strategy. On the other hand, lower molecular weight SMDCs employ ligands with high affinity for specific biomarkers even at smaller sizes, often choosing nitrogen mustard-like cytotoxic payloads, and determining the necessity of a linker based on the release process. While the former approach generally achieves higher success rates, it can result in suboptimal physicochemical properties for the drug. The latter, although more challenging in design, has the potential to produce molecules optimized for oral administration—a feat unattainable with ADCs—thereby offering a significant competitive advantage in drug development.

  Beyond cancer treatment, SMDCs also hold vast potential for broader applications. For instance, the mTOR inhibitor EC0371 for treating polycystic kidney disease and the FRα-targeted anti-inflammatory drug EC1669 have both entered preclinical studies [164,165]. SMDCs can also be used to conjugate fluorescent substances or imaging agents for tumor cell localization or receptor-targeted drug research [134]. For example, SMDCs conjugated with fluorescent dyes can be used during surgery to mark tumor areas for resection or to quantify tumor cells in patient blood samples to measure treatment efficacy. In summary, as a design strategy that combines the advantages of small molecules and conjugated drugs, SMDCs have even more application potential waiting to be explored.

  However, compared to ADCs, the current progress of SMDCs is relatively slow, with very few small molecule candidates advancing to clinical trials. Analyzing the current SMDC designs reveals significant limitations in ligand selection. The ligands in SMDCs must balance binding affinity, selectivity, and compound size, which has led to the predominant use of derivatives from natural compounds—offering limited choices and high modification difficulty [166]. This is one of the greatest challenges in current SMDC development. Currently, SMDC development is primarily focused on FRα, with a significant lack of target diversity and few active products. Despite extensive research on FRα-targeted SMDCs, no drug has yet succeeded in reaching the market. The highly anticipated drug 20, for instance, failed in its Phase Ⅲ clinical trials, which undoubtedly dampened the enthusiasm for SMDC development [121]. Despite this challenging situation, the potential for SMDCs remains immense. The substantial investment from pharmaceutical giants like Novartis in SMDC research is a testament to this potential [92]. With advances in research technology, new ligand screening techniques such as DNA-encoded library technology (DELT) could be applied to SMDC development [167,168]. Compared to traditional high-throughput screening, DELT allows for more efficient and simpler identification of molecules with extremely high affinity for target sites, thereby addressing the current limitations in SMDC ligand selection. Additionally, AI-assisted drug development could also provide significant support, such as predicting and screening ligands based on target structures and assisting in the overall design and optimization of SMDCs [169]. Given the flexibility to interchange ligands, linkers, and payloads, SMDCs could also be modularly assembled to create personalized treatment plans for patients, potentially overcoming resistance issues in targeted cancer therapy and enhancing therapeutic outcomes.

  In conclusion, as a novel approach for targeted delivery of highly cytotoxic drugs, SMDCs have demonstrated considerable drug development potential. Although current progress has been slow, the application of new drug screening and discovery technologies is expected to inject fresh momentum into SMDC design. We firmly believe that the continued efforts of researchers in this field will unlock the key to advancing SMDCs, establishing SMDCs as the optimal choice for targeted cancer therapy.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Jiawei Zhu: Writing – review & editing, Writing – original draft, Conceptualization. Yucheng Xiong: Writing – review & editing. Xiaoxue Bai: Writing – review & editing. Chenlong Xie: Writing – review & editing. Baichen Xiong: Writing – review & editing. Yao Chen: Supervision, Funding acquisition. Haopeng Sun: Writing – review & editing, Supervision, Funding acquisition.

  Acknowledgments

  We gratefully acknowledge the financial support from the National Natural Science Foundation of China (Nos. 82473781, 82173652 and 81872728) and the Natural Science Foundation of Jiangsu Province (No. BK20221522). Support from Jiangsu “333 High Level Talents Cultivation” Leading Talents (No. 2022–3–16–203) is also appreciated.

  
  
• 1. [1]

     B.A. Chabner, T.G. Roberts, Nat. Rev. Cancer 5 (2005) 65–72. doi: 10.1038/nrc1529

  2. [2]

     A.A.M. Van Der Veldt, M. Lubberink, R.H.J. Mathijssen, et al., Clin. Cancer Res. 19 (2013) 4163–4173. doi: 10.1158/1078-0432.CCR-12-3779

  3. [3]

     L. Mosca, A. Ilari, F. Fazi, et al., Drug Resist. Updat. 54 (2021) 100742. doi: 10.1016/j.drup.2020.100742

  4. [4]

     B. Kalyanaraman, Redox Biol. 29 (2020) 101394. doi: 10.1016/j.redox.2019.101394

  5. [5]

     R.K. Singh, S. Kumar, D.N. Prasad, T.R. Bhardwaj, Eur. J. Med. Chem. 151 (2018) 401–433. doi: 10.1016/j.ejmech.2018.04.001

  6. [6]

     S. Zhang, X. Liu, T. Bawa-Khalfe, et al., Nat. Med. 18 (2012) 1639–1642. doi: 10.1038/nm.2919

  7. [7]

     M.T. Manzari, Y. Shamay, H. Kiguchi, et al., Nat. Rev. Mater. 6 (2021) 351–370. doi: 10.1038/s41578-020-00269-6

  8. [8]

     E. Ortega-Forte, A. Rovira, A. Gandioso, et al., J. Med. Chem. 64 (2021) 17209–17220. doi: 10.1021/acs.jmedchem.1c01254

  9. [9]

     X. Sheng, X. Yan, L. Wang, et al., Clin. Cancer Res. 27 (2021) 43–51. doi: 10.1158/1078-0432.ccr-20-2488

  10. [10]

      J.P. Sharman, M. Egyed, W. Jurczak, et al., Lancet 395 (2020) 1278–1291. doi: 10.1016/S0140-6736(20)30262-2

  11. [11]

      R.S. Herbst, G. Giaccone, F. De Marinis, et al., N. Engl. J. Med. 383 (2020) 1328–1339. doi: 10.1056/nejmoa1917346

  12. [12]

      N. Krall, F. Pretto, W. Decurtins, et al., Angew. Chem. Int. Ed. 53 (2014) 4231–4235. doi: 10.1002/anie.201310709

  13. [13]

      J.M. Lambert, R.V.J. Chari, J. Med. Chem. 57 (2014) 6949–6964. doi: 10.1021/jm500766w

  14. [14]

      P. Tarantino, R. Carmagnani Pestana, C. Corti, et al., CA Cancer J. Clin. 72 (2022) 165–182. doi: 10.3322/caac.21705

  15. [15]

      K.J. Norsworthy, C.W. Ko, J.E. Lee, et al., Oncologist 23 (2018) 1103–1108. doi: 10.1634/theoncologist.2017-0604

  16. [16]

      K. Liu, M. Li, Y. Li, et al., Mol. Cancer 23 (2024) 62. doi: 10.1186/s12943-024-01963-7

  17. [17]

      M. Senior, Nat. Biotechnol. 42 (2024) 362–366. doi: 10.1038/s41587-024-02168-5

  18. [18]

      C.Do Pazo, K. Nawaz, R.M. Webster, Nat. Rev. Drug Discov. 20 (2021) 583–584. doi: 10.1038/d41573-021-00054-2

  19. [19]

      T. Doi, K. Shitara, Y. Naito, et al., Lancet Oncol. 18 (2017) 1512–1522. doi: 10.1016/S1470-2045(17)30604-6

  20. [20]

      Y.A. Heo, Drugs 83 (2023) 265–273. doi: 10.1007/s40265-023-01834-3

  21. [21]

      R. Bartsch, A.S. Berghoff, J. Furtner, et al., Nat. Med. 28 (2022) 1840–1847. doi: 10.1038/s41591-022-01935-8

  22. [22]

      G. Casi, D. Neri, J. Med. Chem. 58 (2015) 8751–8761. doi: 10.1021/acs.jmedchem.5b00457

  23. [23]

      A. Zana, A. Galbiati, E. Gilardoni, et al., Clin. Cancer Res. 28 (2022) 5440–5454. doi: 10.1158/1078-0432.ccr-22-1788

  24. [24]

      C.F. Lo, T.Y. Chiu, Y.T. Liu, et al., J. Med. Chem. 65 (2022) 12802–12824. doi: 10.1021/acs.jmedchem.2c00631

  25. [25]

      K. Evans, J. Duan, T. Pritchard, et al., Clin. Cancer Res. 25 (2019) 4493–4503. doi: 10.1158/1078-0432.ccr-19-0551

  26. [26]

      M.S. Dennis, H. Jin, D. Dugger, et al., Cancer Res. 67 (2007) 254–261. doi: 10.1158/0008-5472.CAN-06-2531

  27. [27]

      J. Yang, J. Han, Y. Zhang, et al., Front. Pharmacol. 13 (2022) 924126. doi: 10.3389/fphar.2022.924126

  28. [28]

      B. Oller-Salvia, G. Kym, J.W. Chin, Angew. Chem. Int. Ed. 57 (2018) 2831–2834. doi: 10.1002/anie.201712370

  29. [29]

      F. Sun, T. Wang, J. Jiang, et al., Oncotarget 8 (2017) 51238–51252. doi: 10.18632/oncotarget.17228

  30. [30]

      P. Dennler, A. Chiotellis, E. Fischer, et al., Bioconjug. Chem. 25 (2014) 569–578. doi: 10.1021/bc400574z

  31. [31]

      G. Shah, S. Giralt, P. Dahi, et al., Blood Rev. 64 (2024) 101162. doi: 10.1016/j.blre.2023.101162

  32. [32]

      U.S. Food and Drug Administration, https://www.fda.gov/drugs/resourcesinformation-approved-drugs/fda-approves-melphalan-liver-directedtreatment-uveal-melanoma.

  33. [33]

      U.S. Food and Drug Administration, https://www.fda.gov/drugs/first-generic-drug-approvals/2020-first-generic-drug-approvals.

  34. [34]

      B. Wang, F. Qi, P. Chen, et al., Int. J. Biol. Sci. 20 (2024) 1634–1651. doi: 10.7150/ijbs.92211

  35. [35]

      J. Deng, S.B. Chernikova, Y. Wang, et al., Mol. Cancer Ther. 20 (2021) 2110–2116. doi: 10.1158/1535-7163.mct-21-0140

  36. [36]

      M. Srinivasarao, P.S. Low, Chem. Rev. 117 (2017) 12133–12164. doi: 10.1021/acs.chemrev.7b00013

  37. [37]

      D.A. Proia, D.L. Smith, J. Zhang, et al., Mol. Cancer Ther. 14 (2015) 2422–2432. doi: 10.1158/1535-7163.MCT-15-0455

  38. [38]

      T.U. Amin, R. Emara, A. Pal, et al., J. Med. Chem. 65 (2022) 15473–15486. doi: 10.1021/acs.jmedchem.2c01423

  39. [39]

      H.G. Lerchen, B. Stelte-Ludwig, C. Kopitz, et al., Cancers 14 (2022) 391. doi: 10.3390/cancers14020391

  40. [40]

      A.D. Hobson, J. Xu, D.S. Welch, et al., J. Med. Chem. 66 (2023) 12544–12558. doi: 10.1021/acs.jmedchem.3c01174

  41. [41]

      C.P. Guise, M.R. Abbattista, R.S. Singleton, et al., Cancer Res. 70 (2010) 1573–1584.

  42. [42]

      F. Meng, W.F. Li, D. Jung, et al., Am. J. Cancer Res. (2021) 3645–3659.

  43. [43]

      Y. Zheng, R. Xu, H. Cheng, et al., Mol. Ther. 32 (2024) 1048–1060. doi: 10.1016/j.ymthe.2024.02.020

  44. [44]

      M. Stanbrough, G.J. Bubley, K. Ross, et al., Cancer Res. 66 (2006) 2815–2825. doi: 10.1158/0008-5472.CAN-05-4000

  45. [45]

      T. Yoda, K. Kikuchi, Y. Miki, et al., Mol. Cell. Endocrinol. 413 (2015) 236–247. doi: 10.1016/j.mce.2015.07.008

  46. [46]

      A.M. Tsimberidou, C.F. Verschraegen, R. Wesolowski, et al., Br. J. Cancer 129 (2023) 266–274. doi: 10.1038/s41416-023-02280-4

  47. [47]

      R. Yan, Y. Li, J. Müller, et al., Cell Discov. 7 (2021) 16.

  48. [48]

      C.P. Leamon, I.R. Vlahov, J.A. Reddy, et al., Bioconjug. Chem. 25 (2014) 560–568. doi: 10.1021/bc400441s

  49. [49]

      J. Bendell, G. Falchook, S. Sen, et al., Ann. Oncol. 30 (2019) 172.

  50. [50]

      B. Stelte-Ludwig, M.M. Frigault, H.G. Lerchen, et al., Cancer Res. 83 (2023) 484. doi: 10.1158/1538-7445.am2023-484

  51. [51]

      S. Chen, X. Zhao, J. Chen, et al., Bioconjug. Chem. 21 (2010) 979–987. doi: 10.1021/bc9005656

  52. [52]

      J. Ge, E. Normant, J.R. Porter, et al., J. Med. Chem. 49 (2006) 4606–4615. doi: 10.1021/jm0603116

  53. [53]

      P.F. Chiu, C.K. Chang, P.S. Huang, et al., J. Med. Chem. 66 (2023) 9684–9696. doi: 10.1021/acs.jmedchem.3c00476

  54. [54]

      P. Sapra, H. Zhao, M. Mehlig, et al., Clin. Cancer Res. 14 (2008) 1888–1896. doi: 10.1158/1078-0432.CCR-07-4456

  55. [55]

      H.G. Lerchen, B. Stelte-Ludwig, M. Heroult, et al., Cancers 15 (2023) 4381. doi: 10.3390/cancers15174381

  56. [56]

      C. Kratochwil, W. Fendler, M. Eiber, et al., Eur. J. Nucl. Med. Mol. Imaging 50 (2023) 2830–2845. doi: 10.1007/s00259-023-06255-8

  57. [57]

      A. Mullard, Nat. Rev. Drug Discov. 23 (2024) 88–95. doi: 10.1038/d41573-024-00001-x

  58. [58]

      X. Wang, J. Zhang, K. Zheng, et al., J. Pharm. Anal. 13 (2023) 776–787. doi: 10.1016/j.jpha.2023.02.010

  59. [59]

      O.C. Kulterer, S. Pfaff, W. Wadsak, et al., J. Nucl. Med. 62 (2021) 360–365. doi: 10.2967/jnumed.120.245530

  60. [60]

      M. Corpetti, C. Müller, H. Beltran, et al., Eur. Urol. 85 (2024) 193–204. doi: 10.1016/j.eururo.2023.11.018

  61. [61]

      S.J. Park, B.H. Yoon, S.K. Kim, S.Y. Kim, BMC Med. Genomics 12 (2019) 101. doi: 10.1186/s12920-019-0514-7

  62. [62]

      U.A. Matulonis, D. Lorusso, A. Oaknin, et al., J. Clin. Oncol. 41 (2023) 2436–2445. doi: 10.1200/jco.22.01900

  63. [63]

      M. Chen, L. Cai, Y. Xiang, et al., Bioorg. Chem. 141 (2023) 106889. doi: 10.1016/j.bioorg.2023.106889

  64. [64]

      G.M. Van Dam, G. Themelis, L.M.A. Crane, et al., Nat. Med. 17 (2011) 1315–1319. doi: 10.1038/nm.2472

  65. [65]

      Y. Yuan, D.A. Nymoen, H.P. Dong, et al., Hum. Pathol. 40 (2009) 1453–1460. doi: 10.1016/j.humpath.2009.02.013

  66. [66]

      R. Cé, G.K. Couto, B.Z. Pacheco, et al., Eur. J. Pharm. Sci. 165 (2021) 105943. doi: 10.1016/j.ejps.2021.105943

  67. [67]

      N. Parakasikron, C. Chaotham, P. Chanvorachote, et al., Drug Deliv. 28 (2021) 1443–1454. doi: 10.1080/10717544.2021.1943055

  68. [68]

      X. Li, S. Zhou, C.L. Abrahams, et al., Mol. Cancer Ther. 22 (2023) 155–167.

  69. [69]

      S. Karan, M.Y. Cho, H. Lee, et al., J. Med. Chem. 64 (2021) 2971–2981. doi: 10.1021/acs.jmedchem.0c02162

  70. [70]

      Q.R.J.G. Tummers, C.E.S. Hoogstins, K.N. Gaarenstroom, et al., Oncotarget 7 (2016) 32144–32155. doi: 10.18632/oncotarget.8282

  71. [71]

      Y. Liu, Y. Chen, J. Jiang, et al., Eur. J. Med. Chem. 247 (2023) 115013. doi: 10.1016/j.ejmech.2022.115013

  72. [72]

      Y. Nakamura, P.J. Hornsby, P. Casson, et al., J. Clin. Endocrinol. Metab. 94 (2009) 2192–2198. doi: 10.1210/jc.2008-2374

  73. [73]

      K. Furuuchi, K. Rybinski, J. Fulmer, et al., Cancer Sci. 112 (2021) 2467–2480. doi: 10.1111/cas.14898

  74. [74]

      X. Wu, J. Wu, J. Dai, et al., Natl. Sci. Rev. 8 (2021) nwaa306. doi: 10.1093/nsr/nwaa306

  75. [75]

      J.A. Cornwell, A. Crncec, M.M. Afifi, et al., Nature 619 (2023) 363–370. doi: 10.1038/s41586-023-06274-3

  76. [76]

      S.M. Gross, F. Mohammadi, C. Sanchez-Aguila, et al., Nat. Commun. 14 (2023) 3450. doi: 10.1038/s41467-023-39122-z

  77. [77]

      D.I. Krijbolder, M. Verstappen, B.T. Van Dijk, et al., Lancet 400 (2022) 283–294. doi: 10.1016/S0140-6736(22)01193-X

  78. [78]

      Y.H. Ling, C. Tornos, R. Perez-Soler, et al., J. Biol. Chem. 273 (1998) 18984–18991. doi: 10.1074/jbc.273.30.18984

  79. [79]

      O. Tounekti, G. Pron, J. Belehradek, et al., Cancer Res. 53 (1993) 5462–5469.

  80. [80]

      M. Shi, J. Hou, S. Shao, et al., Int. J. Mol. Sci. 24 (2023) 2498. doi: 10.3390/ijms24032498

  81. [81]

      S. Dhillon, Drugs 81 (2021) 963–969. doi: 10.1007/s40265-021-01522-0

  82. [82]

      Z. Liang, Y. Chen, Z. Wang, et al., J. Adv. Res. 40 (2022) 249–261. doi: 10.1016/j.jare.2021.12.007

  83. [83]

      L.A. Korde, M.R. Somer, E.S. Hwang, et al., J. Clin. Oncol. 39 (2021) 1485–1505. doi: 10.1200/jco.20.03399

  84. [84]

      D. Thomas, A.K. Rathinavel, P. Radhakrishnan, et al., Biochim. Biophys. Acta Rev. Cancer 1875 (2021) 188464. doi: 10.1016/j.bbcan.2020.188464

  85. [85]

      V.K. Sarhadi, G. Armengol, Biomolecules 12 (2022) 1021. doi: 10.3390/biom12081021

  86. [86]

      L. Whitesell, S.L. Lindquist, Nat. Rev. Cancer 5 (2005) 761–772. doi: 10.1038/nrc1716

  87. [87]

      Y. Kanai, Pharmacol. Ther. 230 (2022) 107964. doi: 10.1016/j.pharmthera.2021.107964

  88. [88]

      T. Ozawa, M. Rodriguez, G. Zhao, et al., Cureus 13 (2021) e17595.

  89. [89]

      S. Lu, X. He, Z. Yang, et al., Nat. Commun. 12 (2021) 4721. doi: 10.1038/s41467-021-25020-9

  90. [90]

      R.T. Morris, R.N. Joyrich, R.W. Naumann, et al., Ann. Oncol. 25 (2014) 852–858. doi: 10.1093/annonc/mdu024

  91. [91]

      K.N. Moore, A. Angelergues, G.E. Konecny, et al., Engl. J. Med. 389 (2023) 2162–2174. doi: 10.1056/nejmoa2309169

  92. [92]

      O. Sartor, J. De Bono, K.N. Chi, et al., N. Engl. J. Med. 385 (2021) 1091–1103. doi: 10.1056/nejmoa2107322

  93. [93]

      M. De Wit, Y. Gao, D. Mercieca, et al., EBioMedicine 56 (2020) 102796. doi: 10.1016/j.ebiom.2020.102796

  94. [94]

      C.M. Paulos, J.A. Reddy, C.P. Leamon, et al., Mol. Pharmacol. 66 (2004) 1406–1414. doi: 10.1124/mol.104.003723

  95. [95]

      C. Zhuang, X. Guan, H. Ma, et al., Eur. J. Med. Chem. 163 (2019) 883–895. doi: 10.1016/j.ejmech.2018.12.035

  96. [96]

      M. Srinivasarao, C.V. Galliford, P.S. Low, Nat. Rev. Drug Discov. 14 (2015) 203–219. doi: 10.1038/nrd4519

  97. [97]

      F.P. Olatunji, M. Pun, J.W. Herman, et al., Mol. Cancer Ther. 21 (2022) 1701–1709. doi: 10.1158/1535-7163.mct-22-0160

  98. [98]

      C. Murga-Zamalloa, S. Webb, J. Reneau, et al., Leuk. Res. Rep. 21 (2024) 100398.

  99. [99]

      A. De, P. Roychowdhury, N. Bhuyan, et al., Molecules 28 (2023) 1393. doi: 10.3390/molecules28031393

  100. [100]

       M. Ahmadi, C.A. Ritter, T. Von Woedtke, et al., Chem. Sci. 15 (2024) 1966–2006. doi: 10.1039/d3sc05539f

  101. [101]

       M. Scaranti, E. Cojocaru, S. Banerjee, et al., Nat. Rev. Clin. Oncol. 17 (2020) 349–359. doi: 10.1038/s41571-020-0339-5

  102. [102]

       R. Slack, S. Macdonald, J. Roper, et al., Nat. Rev. Drug Discov. 21 (2022) 60–78. doi: 10.1038/s41573-021-00284-4

  103. [103]

       R. Alexanian, A. Haut, A.U. Khan, et al., J. Am. Med. Assoc. 208 (1969) 1680–1685. doi: 10.1001/jama.1969.03160090040009

  104. [104]

       G. Bennett, A. Brown, G. Mudd, et al., Mol. Cancer Ther. 19 (2020) 1385–1394. doi: 10.1158/1535-7163.mct-19-1092

  105. [105]

       D. Cox, Expert Opin. Drug Discov. 16 (2021) 197–211. doi: 10.1080/17460441.2020.1819234

  106. [106]

       X. Pang, X. He, Z. Qiu, et al., Signal Transduct. Target. Ther. 8 (2023) 1. doi: 10.1038/s41392-022-01259-6

  107. [107]

       S. Suzuki, W. Argraves, R. Pytela, et al., Proc. Natl. Acad. Sci. U. S. A. 83 (1986) 8614–8618. doi: 10.1073/pnas.83.22.8614

  108. [108]

       K. Mishiro, S. Imai, Y. Ematsu, et al., J. Med. Chem. 65 (2022) 16741–16753. doi: 10.1021/acs.jmedchem.2c01586

  109. [109]

       M. Sabat, D. Carney, G. Hernandez-Torres, et al., J. Med. Chem. 67 (2024) 10306–10320. doi: 10.1021/acs.jmedchem.4c00743

  110. [110]

       X. Montet, M. Funovics, K. Montet-Abou, et al., J. Med. Chem. 49 (2006) 6087–6093. doi: 10.1021/jm060515m

  111. [111]

       J. Xiong, J. Chu, W. Fong, et al., J. Am. Chem. Soc. 144 (2022) 10647–10658. doi: 10.1021/jacs.2c04017

  112. [112]

       L. Zhang, P. Wang, X. Zhou, et al., J. Am. Chem. Soc. 145 (2023) 14963–14980. doi: 10.1021/jacs.3c04855

  113. [113]

       M. Kantlehner, P. Schaffner, D. Finsinger, et al., ChemBioChem 1 (2000) 107–114. doi: 10.1002/1439-7633(20000818)1:2<107::AID-CBIC107>3.0.CO;2-4

  114. [114]

       S. Konno, K. Kobayashi, M. Senda, et al., J. Med. Chem. 65 (2022) 2926–2939. doi: 10.1021/acs.jmedchem.1c00665

  115. [115]

       Y. Lamb, Drugs 82 (2022) 585–591. doi: 10.1007/s40265-022-01692-5

  116. [116]

       J. Paulus, B. Nachtigall, P. Meyer, et al., Chem. Eur. J. 29 (2023) e202203476. doi: 10.1002/chem.202203476

  117. [117]

       P. Moharil, Z. Wan, A. Pardeshi, et al., Acta Pharm. Sin. B 12 (2022) 1148–1162. doi: 10.1016/j.apsb.2021.09.024

  118. [118]

       M.P. VanBrunt, K. Shanebeck, Z. Caldwell, et al., Bioconjug. Chem. 26 (2015) 2249–2260. doi: 10.1021/acs.bioconjchem.5b00359

  119. [119]

       S. Milles, S. Tyagi, N. Banterle, et al., J. Am. Chem. Soc. 134 (2012) 5187–5195. doi: 10.1021/ja210587q

  120. [120]

       Z. Wang, H. Li, L. Gou, et al., Acta Pharm. Sin. B 13 (2023) 4025–4059. doi: 10.1016/j.apsb.2023.06.015

  121. [121]

       I. Vergote, C.P. Leamon, Ther. Adv. Med. Oncol. 7 (2015) 206–218. doi: 10.1177/1758834015584763

  122. [122]

       S. Qin, J. Li, Y. Liu, et al., J. Clin. Oncol. 42 (2024) 3121. doi: 10.1200/jco.2024.42.16_suppl.3121

  123. [123]

       B. Gray, L. Kelly, D. Ahrens, et al., Proc. Natl. Acad. Sci. U. S. A. 115 (2018) 4761–4766. doi: 10.1073/pnas.1717705115

  124. [124]

       F. Shi, Y. Liu, X. Zhou, et al., Drug Deliv. 29 (2022) 1335–1344. doi: 10.1080/10717544.2022.2069883

  125. [125]

       P. Guo, J. Huang, B. Zhu, et al., Sci. Adv. 9 (2023) eabq7866. doi: 10.1126/sciadv.abq7866

  126. [126]

       C. Zheng, D. Zhou, W. Li, et al., Cell Death Dis. 14 (2023) 295. doi: 10.1038/s41419-023-05820-1

  127. [127]

       X. Wang, A. Shirke, E. Walker, et al., Cancers 13 (2021) 417. doi: 10.3390/cancers13030417

  128. [128]

       S. Cazzamalli, A. Dal Corso, D. Neri, J. Control. Release 246 (2017) 39–45. doi: 10.1016/j.jconrel.2016.11.023

  129. [129]

       A.S. Rebstock, M. Wiedmann, B. Stelte-Ludwig, et al., Front. Pharmacol. 15 (2024) 1358393. doi: 10.3389/fphar.2024.1358393

  130. [130]

       T. Patel, N. Adhikari, S. Amin, et al., New J. Chem. 45 (2021) 5291–5321. doi: 10.1039/d0nj04134c

  131. [131]

       S. Schuster, E. Juhasz, G. Halmos, et al., Int. J. Mol. Sci. 23 (2022) 5071. doi: 10.3390/ijms23095071

  132. [132]

       P. Berk, J. Goldberg, M. Silverstein, et al., N. Engl. J. Med. 304 (1981) 441– 447. doi: 10.1056/NEJM198102193040801

  133. [133]

       N. Greig, E. Daly, D. Sweeney, et al., Cancer Chemother. Pharmacol. 25 (1990) 320–325. doi: 10.1007/BF00686230

  134. [134]

       G. Jo, E.J. Kim, H. Hyun, Int. J. Mol. Sci. 23 (2022) 14093. doi: 10.3390/ijms232214093

  135. [135]

       Y. Liubomirski, G. Tiram, A. Scomparin, et al., J. Control. Release 367 (2024) 148–157. doi: 10.1016/j.jconrel.2024.01.025

  136. [136]

       R. Mukkamala, D. Carlson, N. Miller, et al., J. Nucl. Med. 65 (2024) 1257–1263. doi: 10.2967/jnumed.124.267756

  137. [137]

       A. Pal, W. Albusairi, F. Liu, et al., Mol. Pharm. 16 (2019) 3237–3252. doi: 10.1021/acs.molpharmaceut.9b00432

  138. [138]

       P. Low, W. Henne, D. Doorneweerd, Acc. Chem. Res. 41 (2008) 120–129. doi: 10.1021/ar7000815

  139. [139]

       I.R. Vlahov, C.P. Leamon, Bioconjug. Chem. 23 (2012) 1357–1369. doi: 10.1021/bc2005522

  140. [140]

       W. Gao, J. Tian, Y. Shang, Chem. Sci. 11 (2020) 3903–3908. doi: 10.1039/d0sc01060j

  141. [141]

       M. Tomczyk, K. Matczak, M. Skonieczna, et al., Bioorganic Chem. 140 (2023) 106782. doi: 10.1016/j.bioorg.2023.106782

  142. [142]

       C.Y. Hsu, Y.Y. Chen, C.F. Lo, et al., Cancer Res. 79 (2019) 100897.

  143. [143]

       R. Traquete, E. Henderson, S. Picaud, et al., Chem. Commun. 55 (2019) 10128–10131. doi: 10.1039/c9cc05054j

  144. [144]

       C.F. Lo, T.Y. Chiu, Y.T. Liu, et al., Pharmaceutics 14 (2022) 1316. doi: 10.3390/pharmaceutics14071316

  145. [145]

       Z. Liu, S. Chen, H. Wang, et al., Bioorganic Chem. 129 (2022) 106221. doi: 10.1016/j.bioorg.2022.106221

  146. [146]

       Y. Zheng, R. Xu, S. Chen, et al., Biomaterials 290 (2022) 121820. doi: 10.1016/j.biomaterials.2022.121820

  147. [147]

       S. Debnath, N. Zhou, M. McLaughlin, et al., Int. J. Mol. Sci. 23 (2022) 1158. doi: 10.3390/ijms23031158

  148. [148]

       J.T. Miller, C.N. Vitro, S. Fang, et al., Bioconjug. Chem. 32 (2021) 842–858. doi: 10.1021/acs.bioconjchem.1c00124

  149. [149]

       F.P. Olatunji, E.A. Savoy, M. Panteah, et al., Bioconjug. Chem. 32 (2021) 2386–2396. doi: 10.1021/acs.bioconjchem.1c00435

  150. [150]

       X. Song, R. Wang, J. Gao, et al., Chin. Chem. Lett. 33 (2022) 1567–1571. doi: 10.1016/j.cclet.2021.08.111

  151. [151]

       Y. Yang, Y. Zhang, R. Wang, et al., Chin. Chem. Lett. 33 (2022) 4583–4586. doi: 10.1016/j.cclet.2022.03.040

  152. [152]

       Y. Wang, Y. Liu, C. Zhou, et al., Leuk. Lymphoma 61 (2020) 1660–1668. doi: 10.1080/10428194.2020.1728746

  153. [153]

       Y. Syed, Drugs 80 (2020) 1019–1025. doi: 10.1007/s40265-020-01337-5

  154. [154]

       K. Li, J. Wu, S. Ye, et al., Ann. Oncol. 34 (2023) S991.

  155. [155]

       J.A. Reddy, E. Westrick, H.K.R. Santhapuram, et al., Cancer Res. 67 (2007) 6376–6382. doi: 10.1158/0008-5472.CAN-06-3894

  156. [156]

       J.A. Reddy, M. Nelson, C. Dircksen, et al., Cancer Res. 77 (2017) 2133. doi: 10.1158/1538-7445.am2017-2133

  157. [157]

       E. Fokas, A. Schlenska-Lange, B. Polat, et al., JAMA Oncol. 8 (2022) e215445. doi: 10.1001/jamaoncol.2021.5445

  158. [158]

       J. Drago, S. Modi, S. Chandarlapaty, Nat. Rev. Clin. Oncol. 18 (2021) 327–344. doi: 10.1038/s41571-021-00470-8

  159. [159]

       C. Dumontet, J. Reichert, P. Senter, et al., Nat. Rev. Drug Discov. 22 (2023) 641–661. doi: 10.1038/s41573-023-00709-2

  160. [160]

       Y. Sun, H. Zhang, G. Lu, et al., Chin. Chem. Lett. 34 (2023) 107817. doi: 10.1016/j.cclet.2022.107817

  161. [161]

       E. Kim, J. Jeong, J. Lee, et al., Biomolecules 10 (2020) 955. doi: 10.3390/biom10060955

  162. [162]

       A. Younes, N. Bartlett, J. Leonard, et al., N. Engl. J. Med. 363 (2010) 1812–1821. doi: 10.1056/NEJMoa1002965

  163. [163]

       D. Wei, Y. Mao, H. Wang, et al., Chin. Chem. Lett. 34 (2023) 108091. doi: 10.1016/j.cclet.2022.108091

  164. [164]

       J.M. Shillingford, C.P. Leamon, I.R. Vlahov, et al., J. Am. Soc. Nephrol. 23 (2012) 1674–1681. doi: 10.1681/ASN.2012040367

  165. [165]

       Y. Lu, K.N. Wollak, V.A. Cross, et al., Clin. Immunol. 150 (2014) 64–77. doi: 10.1016/j.clim.2013.10.010

  166. [166]

       C. Li, Pharmacol. Res. 207 (2024) 107341. doi: 10.1016/j.phrs.2024.107341

  167. [167]

       A.A. Peterson, D.R. Liu, Nat. Rev. Drug Discov. 22 (2023) 699–722. doi: 10.1038/s41573-023-00713-6

  168. [168]

       Q. Nie, S. Zhong, Y. Li, et al., Chin. Chem. Lett. 33 (2022) 2559–2563. doi: 10.1016/j.cclet.2021.09.041

  169. [169]

       X. Zeng, H. Xiang, L. Yu, et al., Mach. Intell. 4 (2022) 1004–1016. doi: 10.1038/s42256-022-00557-6

• 

  Figure 1  Structures of a portion of disclosed SMDCs that have been marketed or are currently active in clinical trials.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Four distinct activation procedures of SMDCs in vivo: Simple diffusion (upper right); active transport (lower right); endocytosis (lower left); and fragmentation in the tumor microenvironment (upper left) (created with BioRender.com). The structures highlighted in green fluorescence denote the targeting ligands, those in yellow fluorescence represent the linkers, and the structures in blue fluorescence indicate the payloads.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  The activation procedure of 4 (AST-3424) in vivo.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Mode of activation of SMDCs. (A) The activation procedure of 6 (PEN-866) in vivo. Targeting ligands facilitate homing to tumor cells with high expression of HSP90, where the linker is cleaved in the lysosomes or endosomes to release 7 (SN-38). (B) The activation procedure of 8 (VIP-236) in vivo. This compound specifically homes to the tumor microenvironment by binding to activated αvβ3 integrins and is effectively cleaved by NE to release 9 (VIP-126).

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Structure of 10 (EC17), a fluorescent imaging agent targeting FRα. Intraoperative fluorescence imaging with 10 allowed for clear detection of ovarian cancer lesions in patients, with an average tumor-to-background ratio (TBR) of 7.0 ± 1.2. The fluorescence imaging was successfully maintained for approximately 5.5 h following the administration of 10.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  Temporal distribution of various types of SMDC payloads in exerting their anti-tumor effects throughout the cell division cycle (created with BioRender.com).

  下载: 全尺寸图片 幻灯片
  

  Figure 7  Crystal structure of LAT1 in complex with 3 (PDB code: 7DLS). LAT1 is depicted in a white cartoon representation, while 3 is shown as orange sticks. The amino acid residues interacting with 3 are displayed as cyan sticks. The yellow dashed lines represent hydrogen bonds between 3 and the residues Ile63, Ser66, Gly67, and Phe252. Additionally, green spheres indicate the centroid of the phenyl ring in the side chain of Phe252, which engages in a π-cation interaction with the positively charged amino group of 3, as illustrated by green dashed lines. Molecular graphics figures were prepared using PyMOL (Schrödinger, LLC).

  下载: 全尺寸图片 幻灯片
  

  Figure 8  Molecules targeting the αvβ3 integrin receptor. (A) The transition from linear 11 (RGD) to 12 (c(RGDfK)) resulted in an approximate10-fold increase in t_1/2. (B) 13 mimicking the RGD binding mode circumvents peptide instability and utilizes the αvβ3 receptor to exert cytotoxic effects on tumor cells. It exhibits no cytotoxic effect on cells lacking the αvβ3 receptor.

  下载: 全尺寸图片 幻灯片
  

  Figure 9  Structure of SMDCs targeting PSMA. (A) The cytotoxicity of 14 (PSMA-1-VcMMAE) against PSMA-positive PC3pip cells is 48 times greater than that against PSMA-negative PC3flu cells. (B) The distribution of MMAE of 15 in the mouse brain is restricted, with tumor exposure being 8-fold higher than that in plasma.

  下载: 全尺寸图片 幻灯片
  

  Figure 10  (A) Conjugatable groups on payloads. The 7 (SN-38) hydroxyl group and PSMA-targeted small molecule are conjugated using OEG to form 16. (B) Intracellular action sites of the payloads. 17 (CLB) demonstrates poor cell membrane permeability, with a topological polar surface area (TPSA) of 160 Ų. The condensation of the carboxyl group of CLB with the amino group of 18 (ZW800) inactivates the carboxyl group, thereby improving cell membrane permeability.

  下载: 全尺寸图片 幻灯片
  

  Figure 11  The use of different linkers in SMDCs design can balance hydrophilicity and hydrophobicity. (A) The targeting ligands PMSA and TTR are linked to the payload MMAE via a linker to form 19. The incorporation of a PEG segment into the linker imparts overall hydrophilicity to the cytotoxic conjugate, thereby restricting its passive diffusion into healthy cells. (B) The folate derivative is conjugated with vincristine through a linker to generate 20 (Vintafolide). Notably, the introduction of an Asp-Arg segment into the linker significantly enhances the overall solubility of the compound.

  下载: 全尺寸图片 幻灯片
  

  Figure 12  Conformational optimization can be optimized by using different linkers in SMDCs design. (A) PS-targeting ligands Zn-DPA and SN38 (TOP1 inhibitor) are connected by a longer linker to form 21 (BPRDP056). This elongated linker does not inhibit the PS-specific tumor-targeting capability of Zn-DPA. (B) I-BET762, which targets BRD4, is linked to a fluorescent probe via different linkers. The linking chain consists of propyl (22), while the other linking chains utilize PEG of varying lengths (23, 24). It is important to note that longer linkers can lead to reduced targeting efficacy. (C) PS-targeting ligands Zn-DPA and DM1 (maytansine-analog) are conjugated through a suitably sized linker to create 25 (Zn8_DM1). The significant polymerization of the PEG moiety can substantially enhance the solubility of the conjugate; however, this may also introduce steric hindrance that can compromise the binding capability of the targeting ligands.

  下载: 全尺寸图片 幻灯片
  

  Figure 13  (A) 26 (TC6-CPT) consists of the GSH reactive conjugate TC6 and the DNA topoisomerase CPT. 26 had a t_1/2 of 60 min at 2 mmol/L GSH and maintained high stability at a variety of pH conditions commonly found in vivo. (B) 27 (Fc1070) is constructed by linking PSMA-targeted small molecules to Monomethylauristatin F through an FC-containing linker, resulting in the formation of 28 (SMDC1070). 27 had a 3-fold higher affinity for PSMA than 28. (C) 29 ([⁶⁸Ga] Ga-PSMA-093) was developed by the incorporation of a specific linker O-(carboxymethyl)-L-tyrosine between the Lys-u-Glu and HBED-CC.

  下载: 全尺寸图片 幻灯片
  

  Figure 14  Three responsive cleavage mechanisms of SMDCs linkers: enzymatic cleavage, acidic condition cleavage, and reductive cleavage (created with BioRender.com).

  下载: 全尺寸图片 幻灯片
  

  Figure 15  Responsive cleavage of the different linker of SMDC. (A) After incubation in the lysosome for 45 min, the cleavage rate of 30 reached 29.3%. (B) The cleavage t_1/2 of 31 varied under different conditions, showing a decrease in t_1/2 with increasing acidity. (C) At a pH of 5.0, 32 rapidly cleave and release their payload, while at pH 7.4, no cleavage occurs within 2 h. (D) 33 (TBG) undergoes continuous cleavage in the presence of GSH. This data highlights the influence of environmental factors, such as pH and redox conditions, on the performance of SMDCs, emphasizing their potential for targeted therapeutic applications.

  下载: 全尺寸图片 幻灯片

  Table 1.  Differences between SMDCs and ADCs.

  Property	ADCs	SMDCs
  Targeting molecule type	Antibodies	Small molecules
  Molecular weight	~150 kDa	<5 kDa
  Solid tumor penetration	Low	High
  Tumor entry mechanism	Receptor-mediated endocytosis; fragmentation in tumor microenvironment	Simple diffusion; active transport; receptor-mediated endocytosis; fragmentation in tumor microenvironment
  Synthesis process and cost	Complex and expensive	Simple and cost-effective
  Heterogeneity	Present	Absent
  Immunogenicity	Present	Absent
  Pharmacokinetic profile	Complex	Simple
  Duration of action	Long	Short

  下载: 导出CSV

  Table 2.  Approved or currently active SMDCs in clinical trials.

  Drug name	Status	Molecular weight	Targeting ligands	Linker cleavage mode	Payload	Conditions	Clinical trial
  Melphalan	Launched	305 Da	LAT1	No linker	Chlormethine	Multiple myeloma; ovarian carcinoma; veal melanoma with unresectable liver metastases	NCT06425276
  TH-302	Phase Ⅲ	449 Da	HIF1ANa	No linker	Br-IPM	Metastatic or locally advanced unresectable pancreatic adenocarcinoma	NCT01746979
  QBS72S	Phase Ⅱ	333 Da	LAT1	No linker	Chlormethine	Brain metastases; breast cancer	NCT05305365
  AST-3424	Phase Ⅰ/Ⅱ	460 Da	AKR1C3	No linker	AST-2660	Liver cancer; relapsed and refractory acute lymphoblastic leukemia	NCT04315324
  PEN-866	Phase Ⅰ/Ⅱ	880 Da	HSP90	Enzymatic cleavage	SN38	Solid tumor	NCT04890093
  CBP-1019	Phase Ⅰ/Ⅱ	Not disclosed	FRα & TRPV6	No linker	MMAE	Solid tumor	NCT06576037
  AST-001	Phase Ⅰ/Ⅱ	Not disclosed	AKR1C3	No linker	AST-2660	Solid tumor	NCT06245330
  VIP-236	Phase Ⅰ	1516 Da	αvβ3	Enzymatic cleavage	VIP-126	Solid tumor	NCT05712889
  T-1201	Phase Ⅰ	Not disclosed	PSb	Enzymatic Cleavage	SN-38	Solid tumor	NCT04866641
  a HIF1AN, hypoxia inducible factor 1 subunit alpha. b PS, phosphatidylserine.

  下载: 导出CSV
•